From mhammond@skippinet.com.au  Sat Apr  1 05:17:39 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Sat, 1 Apr 2000 15:17:39 +1000
Subject: [Patches] test_winreg doesnt delete its test data.
Message-ID: <ECEPKNMJLHAPFFJHDOJBCEMMCHAA.mhammond@skippinet.com.au>

The call to remove the test data (also testing the delete functions)
was commented out!

diff -r1.1 test_winreg.py
125c125
< #    DeleteTestData(root_key)
---
>     DeleteTestData(root_key)


Release info:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that
I have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce,
distribute, perform and/or display publicly, prepare derivative
versions, and otherwise use this contribution as part of the Python
software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize
others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and
its related documentation.  I further grant CNRI permission to use
my name and other identifying information provided to CNRI by me for
use in connection with the Python software and its related
documentation.

Mark.




From fdrake@acm.org  Sat Apr  1 05:26:17 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Sat, 1 Apr 2000 00:26:17 -0500 (EST)
Subject: [Patches] test_winreg doesnt delete its test data.
In-Reply-To: <ECEPKNMJLHAPFFJHDOJBCEMMCHAA.mhammond@skippinet.com.au>
References: <ECEPKNMJLHAPFFJHDOJBCEMMCHAA.mhammond@skippinet.com.au>
Message-ID: <14565.34937.561497.697032@seahag.cnri.reston.va.us>

Mark Hammond writes:
 > The call to remove the test data (also testing the delete functions)
 > was commented out!

  Thanks Mark!


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From sjoerd@oratrix.nl  Sat Apr  1 12:47:53 2000
From: sjoerd@oratrix.nl (Sjoerd Mullender)
Date: Sat, 01 Apr 2000 14:47:53 +0200
Subject: [Patches] Lib/sre*.py modules checked in wrong
Message-ID: <20000401124754.724B1301CF9@bireme.oratrix.nl>

Please check the Lib/sre*.py modules in again, since they contain ^M's
when I check them out on Linux.  When I try to compile these modules
using py_compile, it complains about syntax errors.

Thanks.

-- Sjoerd Mullender <sjoerd.mullender@oratrix.com>


From mal@lemburg.com  Sat Apr  1 19:40:22 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Sat, 01 Apr 2000 21:40:22 +0200
Subject: [Patches] Unicode Doc Snapshot 2000-04-01
Message-ID: <38E650A6.74C5B6E1@lemburg.com>

This is a multi-part message in MIME format.
--------------AC5A3AFAAB32FC257EC8E73D
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

No, not a joke... I added a fairly complete section to 
api.tex explaining the C API side of things.

The patch also includes other patches for the different new
APIs and small changes that were done in order to make 
Unicode intergrate well with strings and other objects.

Please note that I have not checked whether the docs compile
or not... there may be piles of those nice TeX errors in there ;-)

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------AC5A3AFAAB32FC257EC8E73D
Content-Type: text/plain; charset=iso-8859-1;
 name="Unicode-Implementation-2000-04-01-Doc.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-01-Doc.patch"

diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/api/api.tex Python+Unicode/Doc/api/api.tex
--- CVS-Python/Doc/api/api.tex	Fri Mar 31 10:37:59 2000
+++ Python+Unicode/Doc/api/api.tex	Sat Apr  1 21:30:53 2000
@@ -1847,6 +1847,717 @@
 \end{cfuncdesc}
 
 
+\subsection{Unicode Objects \label{unicodeObjects}}
+
+%--- Unicode Type -------------------------------------------------------
+
+These are the basic Unicode object types used for the Unicode
+implementation in Python:
+
+\begin{ctypedesc}{Py_UNICODE}
+This type represents a 16-bit unsigned storage type which is used by
+Python internally as basis for holding Unicode ordinals. On platforms
+where \ctype{wchar_t} is available and also has 16-bits,
+\ctype{Py_UNICODE} is a typedef alias for \ctype{wchar_t} to enhance
+native platform compatibility. On all other platforms,
+\ctype{Py_UNICODE} is a typedef alias for \ctype{unsigned short}.
+\end{ctypedesc}
+
+\begin{ctypedesc}{PyUnicodeObject}
+This subtype of \ctype{PyObject} represents a Python Unicode object.
+\end{ctypedesc}
+
+\begin{cvardesc}{PyTypeObject}{PyUnicode_Type}
+This instance of \ctype{PyTypeObject} represents the Python Unicode type.
+\end{cvardesc}
+
+%--- These are really C macros... is there a macrodesc TeX macro ?
+
+The following APIs are really C macros and can be used to do fast
+checks and to access internal read-only data of Unicode objects:
+
+\begin{cfuncdesc}{int}{PyUnicode_Check}{PyObject *o}
+Returns true if the object \var{o} is a Unicode object.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{PyUnicode_GET_SIZE}{PyObject *o}
+Returns the size of the object.  o has to be a
+PyUnicodeObject (not checked).
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{PyUnicode_GET_DATA_SIZE}{PyObject *o}
+Returns the size of the object's internal buffer in bytes. o has to be
+a PyUnicodeObject (not checked).
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{PyUnicode_AS_UNICODE}{PyObject *o}
+Returns a pointer to the internal Py_UNICODE buffer of the object. o
+has to be a PyUnicodeObject (not checked).
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{PyUnicode_AS_DATA}{PyObject *o}
+Returns a (const char *) pointer to the internal buffer of the object.
+o has to be a PyUnicodeObject (not checked).
+\end{cfuncdesc}
+
+% --- Unicode character properties ---------------------------------------
+
+Unicode provides many different character properties. The most often
+needed ones are available through these macros which are mapped to C
+functions depending on the Python configuration.
+
+\begin{cfuncdesc}{int}{Py_UNICODE_ISSPACE}{Py_UNICODE ch}
+Returns 1/0 depending on whether \var{ch} is a whitespace character.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_ISLOWER}{Py_UNICODE ch}
+Returns 1/0 depending on whether \var{ch} is a lowercase character.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_ISUPPER}{Py_UNICODE ch}
+Returns 1/0 depending on whether \var{ch} is a uppercase character.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_ISTITLE}{Py_UNICODE ch}
+Returns 1/0 depending on whether \var{ch} is a titlecase character.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_ISLINEBREAK}{Py_UNICODE ch}
+Returns 1/0 depending on whether \var{ch} is a linebreak character.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_ISDECIMAL}{Py_UNICODE ch}
+Returns 1/0 depending on whether \var{ch} is a decimal character.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_ISDIGIT}{Py_UNICODE ch}
+Returns 1/0 depending on whether \var{ch} is a digit character.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_ISNUMERIC}{Py_UNICODE ch}
+Returns 1/0 depending on whether \var{ch} is a numeric character.
+\end{cfuncdesc}
+
+These APIs can be used for fast direct character conversions:
+
+\begin{cfuncdesc}{Py_UNICODE}{Py_UNICODE_TOLOWER}{Py_UNICODE ch}
+Returns the character \var{ch} converted to lower case.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{Py_UNICODE}{Py_UNICODE_TOUPPER}{Py_UNICODE ch}
+Returns the character \var{ch} converted to upper case.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{Py_UNICODE}{Py_UNICODE_TOTITLE}{Py_UNICODE ch}
+Returns the character \var{ch} converted to title case.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_TODECIMAL}{Py_UNICODE ch}
+Returns the character \var{ch} converted to a decimal positive integer.
+Returns -1 in case this is not possible. Does not raise exceptions.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{Py_UNICODE_TODIGIT}{Py_UNICODE ch}
+Returns the character \var{ch} converted to a single digit integer.
+Returns -1 in case this is not possible. Does not raise exceptions.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{double}{Py_UNICODE_TONUMERIC}{Py_UNICODE ch}
+Returns the character \var{ch} converted to a (positive) double.
+Returns -1.0 in case this is not possible. Does not raise exceptions.
+\end{cfuncdesc}
+
+% --- Plain Py_UNICODE ---------------------------------------------------
+
+To create Unicode objects and access their basic sequence properties,
+use these APIs:
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_FromUnicode}{const Py_UNICODE *u,
+                                                    int size} 
+
+Create a Unicode Object from the Py_UNICODE buffer \var{u} of the
+given size. \var{u} may be \NULL{} which causes the contents to be
+undefined. It is the user's responsibility to fill in the needed data.
+The buffer is copied into the new object.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{Py_UNICODE *}{PyUnicode_AsUnicode}{PyObject *unicode}
+Return a read-only pointer to the Unicode object's internal
+\ctype{Py_UNICODE} buffer.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{PyUnicode_GetSize}{PyObject *unicode}
+Return the length of the Unicode object.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_FromObject}{PyObject *obj}
+
+Coerce obj to an Unicode object and return a reference with
+incremented refcount.
+
+Coercion is done in the following way:
+\begin{enumerate}
+\item  Unicode objects are passed back as-is with incremented
+      refcount.
+
+\item String and other char buffer compatible objects are decoded
+      under the assumptions that they contain UTF-8 data. Decoding
+      is done in "strict" mode.
+
+\item All other objects raise an exception.
+\end{enumerate}
+The API returns NULL in case of an error. The caller is responsible
+for decref'ing the returned objects.
+\end{cfuncdesc}
+
+% --- wchar_t support for platforms which support it ---------------------
+
+If the platform supports \ctype{wchar_t} and provides a header file
+wchar.h, Python can interface directly to this type using the
+following functions. Support is optimized if Python's own
+\ctype{Py_UNICODE} type is identical to the system's \ctype{wchar_t}.
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_FromWideChar}{const wchar_t *w,
+                                                     int size}
+Create a Unicode Object from the \ctype{whcar_t} buffer \var{w} of the
+given size. Returns \NULL{} on failure.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{PyUnicode_AsWideChar}{PyUnicodeObject *unicode,
+                                             wchar_t *w,
+                                             int size}
+
+Copies the Unicode Object contents into the \ctype{whcar_t} buffer
+\var{w}.  At most \var{size} \ctype{whcar_t} characters are copied.
+Returns the number of \ctype{whcar_t} characters copied or -1 in case
+of an error.
+\end{cfuncdesc}
+
+
+\subsubsection{Builtin Codecs \label{builtinCodecs}}
+
+Python provides a set of builtin codecs which are written in C
+for speed. All of these codecs are directly usable via the
+following functions.
+
+Many of the following APIs take two arguments encoding and
+errors. These parameters encoding and errors have the same semantics
+as the ones of the builtin unicode() Unicode object constructor.
+
+Setting encoding to NULL causes the default encoding to be used which
+is UTF-8.
+
+Error handling is set by errors which may also be set to NULL meaning
+to use the default handling defined for the codec. Default error
+handling for all builtin codecs is ``strict'' (ValueErrors are raised).
+
+The codecs all use a similar interface. Only deviation from the
+following generic ones are documented for simplicity.
+
+% --- Generic Codecs -----------------------------------------------------
+
+These are the generic codec APIs:
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Decode}{const char *s,
+                                               int size,
+                                               const char *encoding,
+                                               const char *errors}
+
+Create a Unicode object by decoding \var{size} bytes of the encoded
+string \var{s}. \var{encoding} and \var{errors} have the same meaning
+as the parameters of the same name in the unicode() builtin
+function. The codec to be used is looked up using the Python codec
+registry. Returns \NULL{} in case an exception was raised by the
+codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Encode}{const Py_UNICODE *s,
+                                               int size,
+                                               const char *encoding,
+                                               const char *errors}
+
+Encodes the \ctype{Py_UNICODE} buffer of the given size and returns a
+Python string object. \var{encoding} and \var{errors} have the same
+meaning as the parameters of the same name in the Unicode .encode()
+method. The codec to be used is looked up using the Python codec
+registry. Returns \NULL{} in case an exception was raised by the
+codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsEncodedString}{PyObject *unicode,
+                                               const char *encoding,
+                                               const char *errors}
+
+Encodes a Unicode object and returns the result as Python string
+object. \var{encoding} and \var{errors} have the same meaning as the
+parameters of the same name in the Unicode .encode() method. The codec
+to be used is looked up using the Python codec registry. Returns
+\NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+% --- UTF-8 Codecs -------------------------------------------------------
+
+These are the UTF-8 codec APIs:
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_DecodeUTF8}{const char *s,
+                                               int size,
+                                               const char *errors}
+
+Creates a Unicode object by decoding \var{size} bytes of the UTF-8
+encoded string \var{s}. Returns \NULL{} in case an exception was
+raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_EncodeUTF8}{const Py_UNICODE *s,
+                                               int size,
+                                               const char *errors}
+
+Encodes the \ctype{Py_UNICODE} buffer of the given size using UTF-8
+and returns a Python string object.  Returns \NULL{} in case an
+exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsUTF8String}{PyObject *unicode}
+
+Encodes a Unicode objects using UTF-8 and returns the result as Python
+string object. Error handling is ``strict''. Returns
+\NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+% --- UTF-16 Codecs ------------------------------------------------------ */
+
+These are the UTF-16 codec APIs:
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_DecodeUTF16}{const char *s,
+                                               int size,
+                                               const char *errors,
+                                               int *byteorder}
+
+Decodes \var{length} bytes from a UTF-16 encoded buffer string and
+returns the corresponding Unicode object.
+
+\var{errors} (if non-NULL) defines the error handling. It defaults
+to ``strict''.
+
+If \var{byteorder} is non-\NULL{}, the decoder starts decoding using
+the given byte order:
+
+\begin{verbatim}
+   *byteorder == -1: little endian
+   *byteorder == 0:  native order
+   *byteorder == 1:  big endian
+\end{verbatim}
+
+and then switches according to all byte order marks (BOM) it finds in
+the input data. BOM marks are not copied into the resulting Unicode
+string.  After completion, \var{*byteorder} is set to the current byte
+order at the end of input data.
+
+If \var{byteorder} is \NULL{}, the codec starts in native order mode.
+
+Returns \NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_EncodeUTF16}{const Py_UNICODE *s,
+                                               int size,
+                                               const char *errors,
+                                               int byteorder}
+
+Returns a Python string object holding the UTF-16 encoded value of the
+Unicode data in \var{s}.
+
+If \var{byteorder} is not 0, output is written according to the
+following byte order:
+
+\begin{verbatim}
+   byteorder == -1: little endian
+   byteorder == 0:  native byte order (writes a BOM mark)
+   byteorder == 1:  big endian
+\end{verbatim}
+
+If byteorder is 0, the output string will always start with the
+Unicode BOM mark (U+FEFF). In the other two modes, no BOM mark is
+prepended.
+
+Note that \ctype{Py_UNICODE} data is being interpreted as UTF-16
+reduced to UCS-2. This trick makes it possible to add full UTF-16
+capabilities at a later point without comprimising the APIs.
+
+Returns \NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsUTF16String}{PyObject *unicode}
+
+Returns a Python string using the UTF-16 encoding in native byte
+order. The string always starts with a BOM mark. Error handling is
+``strict''. Returns \NULL{} in case an exception was raised by the
+codec.
+\end{cfuncdesc}
+
+% --- Unicode-Escape Codecs ----------------------------------------------
+
+These are the ``Unicode Esacpe'' codec APIs:
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_DecodeUnicodeEscape}{const char *s,
+                                               int size,
+                                               const char *errors}
+
+Creates a Unicode object by decoding \var{size} bytes of the Unicode-Esacpe
+encoded string \var{s}. Returns \NULL{} in case an exception was
+raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_EncodeUnicodeEscape}{const Py_UNICODE *s,
+                                               int size,
+                                               const char *errors}
+
+Encodes the \ctype{Py_UNICODE} buffer of the given size using Unicode-Escape
+and returns a Python string object.  Returns \NULL{} in case an
+exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsUnicodeEscapeString}{PyObject *unicode}
+
+Encodes a Unicode objects using Unicode-Escape and returns the result
+as Python string object. Error handling is ``strict''. Returns
+\NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+% --- Raw-Unicode-Escape Codecs ------------------------------------------
+
+These are the ``Raw Unicode Esacpe'' codec APIs:
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_DecodeRawUnicodeEscape}{const char *s,
+                                               int size,
+                                               const char *errors}
+
+Creates a Unicode object by decoding \var{size} bytes of the Raw-Unicode-Esacpe
+encoded string \var{s}. Returns \NULL{} in case an exception was
+raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_EncodeRawUnicodeEscape}{const Py_UNICODE *s,
+                                               int size,
+                                               const char *errors}
+
+Encodes the \ctype{Py_UNICODE} buffer of the given size using Raw-Unicode-Escape
+and returns a Python string object.  Returns \NULL{} in case an
+exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsRawUnicodeEscapeString}{PyObject *unicode}
+
+Encodes a Unicode objects using Raw-Unicode-Escape and returns the result
+as Python string object. Error handling is ``strict''. Returns
+\NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+% --- Latin-1 Codecs ----------------------------------------------------- 
+
+These are the Latin-1 codec APIs:
+
+Latin-1 corresponds to the first 256 Unicode ordinals and only these
+are accepted by the codecs during encoding.
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_DecodeLatin1}{const char *s,
+                                               int size,
+                                               const char *errors}
+
+Creates a Unicode object by decoding \var{size} bytes of the Latin-1
+encoded string \var{s}. Returns \NULL{} in case an exception was
+raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_EncodeLatin1}{const Py_UNICODE *s,
+                                               int size,
+                                               const char *errors}
+
+Encodes the \ctype{Py_UNICODE} buffer of the given size using Latin-1
+and returns a Python string object.  Returns \NULL{} in case an
+exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsLatin1String}{PyObject *unicode}
+
+Encodes a Unicode objects using Latin-1 and returns the result as
+Python string object. Error handling is ``strict''. Returns
+\NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+% --- ASCII Codecs ------------------------------------------------------- 
+
+These are the ASCII codec APIs:
+
+Only 7-bit ASCII data is excepted. All other codes generate errors.
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_DecodeASCII}{const char *s,
+                                               int size,
+                                               const char *errors}
+
+Creates a Unicode object by decoding \var{size} bytes of the ASCII
+encoded string \var{s}. Returns \NULL{} in case an exception was
+raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_EncodeASCII}{const Py_UNICODE *s,
+                                               int size,
+                                               const char *errors}
+
+Encodes the \ctype{Py_UNICODE} buffer of the given size using ASCII
+and returns a Python string object.  Returns \NULL{} in case an
+exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsASCIIString}{PyObject *unicode}
+
+Encodes a Unicode objects using ASCII and returns the result as Python
+string object. Error handling is ``strict''. Returns
+\NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+% --- Character Map Codecs ----------------------------------------------- 
+
+These are the mapping codec APIs:
+
+This codec is special in that it can be used to implement many
+different codecs (and this is in fact what was done to obtain most of
+the standard codecs included in the \module{encodings} package). The
+codec uses mapping to encode and decode characters.
+
+Decoding mappings must map single string characters to single Unicode
+characters, integers (which are then interpreted as Unicode ordinals)
+or None (meaning "undefined mapping" and causing an error). 
+
+Encoding mappings must map single Unicode characters to single string
+characters, integers (which are then interpreted as Latin-1 ordinals)
+or None (meaning "undefined mapping" and causing an error).
+
+The mapping objects provided must only support the __getitem__ mapping
+interface.
+
+If a character lookup fails with a LookupError, the character is
+copied as-is meaning that its ordinal value will be interpreted as
+Unicode or Latin-1 ordinal resp. Because of this, mappings only need
+to contain those mappings which map characters to different code
+points.
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_DecodeCharmap}{const char *s,
+                                               int size,
+                                               PyObject *mapping,
+                                               const char *errors}
+
+Creates a Unicode object by decoding \var{size} bytes of the encoded
+string \var{s} using the given \var{mapping} object.  Returns \NULL{}
+in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_EncodeCharmap}{const Py_UNICODE *s,
+                                               int size,
+                                               PyObject *mapping,
+                                               const char *errors}
+
+Encodes the \ctype{Py_UNICODE} buffer of the given size using the
+given \var{mapping} object and returns a Python string object.
+Returns \NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsCharmapString}{PyObject *unicode,
+                                                        PyObject *mapping}
+
+Encodes a Unicode objects using the given \var{mapping} object and
+returns the result as Python string object. Error handling is
+``strict''. Returns \NULL{} in case an exception was raised by the
+codec.
+\end{cfuncdesc}
+
+The following codec API is special in that maps Unicode to Unicode.
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_TranslateCharmap}{const Py_UNICODE *s,
+                                               int size,
+                                               PyObject *table,
+                                               const char *errors}
+
+Translates a \ctype{Py_UNICODE} buffer of the given length by applying
+a character mapping \var{table} to it and returns the resulting
+Unicode object.
+
+The \var{mapping} table must map Unicode ordinal integers to Unicode
+ordinal integers or None (causing deletion of the character).
+
+Mapping tables must only provide the __getitem__ interface,
+e.g. dictionaries or sequences. Unmapped character ordinals (ones
+which cause a LookupError) are left untouched and are copied as-is.
+
+Returns \NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+% --- MBCS codecs for Windows --------------------------------------------
+
+These are the MBCS codec APIs. They are currently only available
+Windows and use the Win32 MBCS converters to implement the
+conversions. 
+
+Note that MBCS (or DBCS) is a class of encodings, not just one.  The
+target encoding is defined by the user settings on the machine running
+the codec.
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_DecodeMBCS}{const char *s,
+                                               int size,
+                                               const char *errors}
+
+Creates a Unicode object by decoding \var{size} bytes of the MBCS
+encoded string \var{s}. Returns \NULL{} in case an exception was
+raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_EncodeMBCS}{const Py_UNICODE *s,
+                                               int size,
+                                               const char *errors}
+
+Encodes the \ctype{Py_UNICODE} buffer of the given size using MBCS
+and returns a Python string object.  Returns \NULL{} in case an
+exception was raised by the codec.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_AsMBCSString}{PyObject *unicode}
+
+Encodes a Unicode objects using MBCS and returns the result as Python
+string object. Error handling is ``strict''. Returns
+\NULL{} in case an exception was raised by the codec.
+\end{cfuncdesc}
+
+% --- Methods & Slots ----------------------------------------------------
+
+\subsubsection{Methods & Slot Functions \label{unicodeMethodsAndSlots}}
+
+The following APIs are capable of handling Unicode objects and strings
+on input (we refer to them as strings in the descriptions) and return
+Unicode objects or integers as apporpriate.
+
+They all return \NULL{} or -1 in case an exception occurrs.
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Concat}{PyObject *left,
+                                               PyObject *right}
+
+Concat two strings giving a new Unicode string.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Split}{PyObject *s,
+                                              PyObject *sep,
+                                              int maxsplit}
+
+Split a string giving a list of Unicode strings.
+
+If sep is NULL, splitting will be done at all whitespace
+substrings. Otherwise, splits occur at the given separator.
+
+At most maxsplit splits will be done. If negative, no limit is set.
+
+Separators are not included in the resulting list.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Splitlines}{PyObject *s,
+                                                   int maxsplit}
+
+Dito, but split at line breaks.
+
+CRLF is considered to be one line break. Line breaks are not
+included in the resulting list.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Translate}{PyObject *str,
+                                                  PyObject *table,
+                                                  const char *errors}
+
+Translate a string by applying a character mapping table to it and
+return the resulting Unicode object.
+
+The mapping table must map Unicode ordinal integers to Unicode ordinal
+integers or None (causing deletion of the character).
+
+Mapping tables must only provide the __getitem__ interface,
+e.g. dictionaries or sequences. Unmapped character ordinals (ones
+which cause a LookupError) are left untouched and are copied as-is.
+
+\var{errors} has the usual meaning for codecs. It may be \NULL{}
+which indicates to use the default error handling.
+
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Join}{PyObject *separator,
+                                             PyObject *seq}
+
+Join a sequence of strings using the given separator and return
+the resulting Unicode string.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Tailmatch}{PyObject *str,
+                                                  PyObject *substr,
+                                                  int start,
+                                                  int end,
+                                                  int direction}
+
+Return 1 if \var{substr} matches \var{str}[\var{start}:\var{end}] at
+the given tail end (\var{direction} == -1 means to do a prefix match,
+\var{direction} == 1 a suffix match), 0 otherwise.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Find}{PyObject *str,
+                                                  PyObject *substr,
+                                                  int start,
+                                                  int end,
+                                                  int direction}
+
+Return the first position of \var{substr} in
+\var{str}[\var{start}:\var{end}] using the given \var{direction}
+(\var{direction} == 1 means to do a forward search,
+\var{direction} == -1 a backward search), 0 otherwise.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Count}{PyObject *str,
+                                                  PyObject *substr,
+                                                  int start,
+                                                  int end}
+
+Count the number of occurrences of \var{substr} in
+\var{str}[\var{start}:\var{end}]
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Replace}{PyObject *str,
+                                                PyObject *substr,
+                                                PyObject *replstr,
+                                                int maxcount}
+
+Replace at most \var{maxcount} occurrences of \var{substr} in
+\var{str} with \var{replstr} and return the resulting Unicode object.
+\var{maxcount} == -1 means: replace all occurrences.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{PyUnicode_Compare}{PyObject *left,
+                                                PyObject *right}
+
+Compare two strings and return -1, 0, 1 for less than, equal,
+greater than resp.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{PyObject*}{PyUnicode_Format}{PyObject *format,
+                                              PyObject *args}
+Returns a new string object from \var{format} and \var{args}.  Analogous
+to \code{\var{format} \% \var{args}}.  The \var{args} argument must be
+a tuple.
+\end{cfuncdesc}
+
+\begin{cfuncdesc}{int}{PyUnicode_Contains}{PyObject *container,
+                                           PyObject *element}
+
+Checks whether \var{element} is contained in \var{container} and
+returns 1/0 accordingly.
+
+\var{element} has to coerce to an one element Unicode string. -1 is
+returned in case of an error.
+\end{cfuncdesc}
+
+
 \subsection{Buffer Objects \label{bufferObjects}}
 
 XXX need a real description of buffers and buffer ''segments.``
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/lib/lib.tex Python+Unicode/Doc/lib/lib.tex
--- CVS-Python/Doc/lib/lib.tex	Wed Mar  1 10:22:28 2000
+++ Python+Unicode/Doc/lib/lib.tex	Sat Apr  1 17:13:45 2000
@@ -128,6 +128,7 @@
 \input{libcalendar}
 \input{libcmd}
 \input{libshlex}
+\input{libcodecs}
 
 \input{liballos}		% Generic Operating System Services
 \input{libos}
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/lib/libcodecs.tex Python+Unicode/Doc/lib/libcodecs.tex
--- CVS-Python/Doc/lib/libcodecs.tex	Thu Jan  1 01:00:00 1970
+++ Python+Unicode/Doc/lib/libcodecs.tex	Sat Apr  1 17:13:45 2000
@@ -0,0 +1,123 @@
+\section{\module{codecs} ---
+         Python codec registry and base classes.}
+\declaremodule{standard}{codec}
+
+\modulesynopsis{Encode and decode data and streams.}
+
+\index{Unicode}
+\index{Codecs}
+\indexii{Codecs}{encode}
+\indexii{Codecs}{decode}
+\index{streams}
+\indexii{stackable}{streams}
+
+
+This module defines base classes for standard Python codecs (encoders
+and decoders) and provides access to the internal Python codec
+registry which manages the codec lookup process.
+
+It defines the following functions:
+
+\begin{funcdesc}{register}{search_function}
+Register a codec search function. Search functions are expected to
+take one argument, the encoding name in all lower case letters, and
+return a tuple of functions \code{(\var{encoder}, \var{decoder}, \var{stream_reader},
+\var{stream_writer})} taking the following arguments:
+
+  \var{encoder} and \var{decoder}: These must be functions or methods
+  which have the same interface as the .encode/.decode methods of
+  Codec instances (see Codec Interface). The functions/methods are
+  expected to work in a stateless mode.
+
+  \var{stream_reader} and \var{stream_writer}: These have to be
+  factory functions providing the following interface:
+
+	\code{factory(\var{stream},\var{errors}='strict')}
+
+  The factory functions must return objects providing the interfaces
+  defined by the base classes
+  \class{StreamWriter}/\class{StreamReader} resp. Stream codecs can
+  maintain state.
+
+  Possible values for errors are 'strict' (raise an exception in case
+  of an encoding error), 'replace' (replace malformed data with a
+  suitable replacement marker, e.g. '?') and 'ignore' (ignore
+  malformed data and continue without further notice).
+
+In case a search function cannot find a given encoding, it should
+return None.
+\end{funcdesc}
+
+\begin{funcdesc}{lookup}{encoding}
+Looks up a codec tuple in the Python codec registry and returns the
+function tuple as defined above.
+
+Encodings are first looked up in the registry's cache. If not found,
+the list of registered search functions is scanned. If no codecs tuple
+is found, a LookupError is raised. Otherwise, the codecs tuple is
+stored in the cache and returned to the caller.
+\end{funcdesc}
+
+To simplify working with encoded files or stream, the module
+also defines these utility functions:
+
+\begin{funcdesc}{open}{filename, mode\optional{, encoding=None, errors='strict', buffering=1}}
+Open an encoded file using the given \var{mode} and return
+a wrapped version providing transparent encoding/decoding.
+
+Note: The wrapped version will only accept the object format defined
+by the codecs, i.e. Unicode objects for most builtin codecs. Output is
+also codec dependent and will usually by Unicode as well.
+
+\var{encoding} specifies the encoding which is to be used for the
+the file.
+
+\var{errors} may be given to define the error handling. It defaults
+to 'strict' which causes a \exception{ValueError} to be raised in case
+an encoding error occurs.
+
+\var{buffering} has the same meaning as for the builtin open() API.
+It defaults to line buffered.
+\end{funcdesc}
+
+\begin{funcdesc}{EncodedFile}{file, input\optional{, output=None, errors='strict'}}
+
+Return a wrapped version of file which provides transparent
+encoding translation.
+
+Strings written to the wrapped file are interpreted according to the
+given \var{input} encoding and then written to the original file as
+string using the \var{output} encoding. The intermediate encoding will
+usually be Unicode but depends on the specified codecs.
+
+If \var{output} is not given, it defaults to input.
+
+\var{errors} may be given to define the error handling. It defaults to
+'strict' which causes a \exception{ValueError} to be raised in case
+an encoding error occurs.
+\end{funcdesc}
+
+
+
+...XXX document codec base classes...
+
+
+
+The module also provides the following constants which are useful
+for reading and writing to platform dependent files:
+
+\begin{datadesc}{BOM*}
+\dataline{BOM}
+\dataline{BOM_BE}
+\dataline{BOM_LE}
+\dataline{BOM32_BE}
+\dataline{BOM32_LE}
+\dataline{BOM64_BE}
+\dataline{BOM64_LE}
+These constants define the byte order marks (BOM) used in data
+streams to indicate the byte order used in the stream or file.
+\var{BOM} is set to either BOM_BE or BOM_LE depending on the platform's
+native byte order, while the others represent big endian (_BE) and
+little endian (_LE) byte order using 32-bit and 64-bit encodings.
+\end{datadesc}
+
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/lib/libexcs.tex Python+Unicode/Doc/lib/libexcs.tex
--- CVS-Python/Doc/lib/libexcs.tex	Fri Dec  3 13:09:30 1999
+++ Python+Unicode/Doc/lib/libexcs.tex	Sat Apr  1 17:13:45 2000
@@ -310,6 +310,11 @@
   details about the type mismatch.
 \end{excdesc}
 
+\begin{excdesc}{UnicodeError}
+  Raised when a Unicode related encoding/decoding error occurs. It is
+  a subclass of \exception{ValueError}.
+\end{excdesc}
+
 \begin{excdesc}{ValueError}
   Raised when a built-in operation or function receives an argument
   that has the right type but an inappropriate value, and the
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/lib/libfuncs.tex Python+Unicode/Doc/lib/libfuncs.tex
--- CVS-Python/Doc/lib/libfuncs.tex	Wed Mar  1 10:22:28 2000
+++ Python+Unicode/Doc/lib/libfuncs.tex	Sat Apr  1 17:13:45 2000
@@ -466,9 +466,11 @@
 \end{funcdesc}
 
 \begin{funcdesc}{ord}{c}
-  Return the \ASCII{} value of a string of one character.  E.g.,
-  \code{ord('a')} returns the integer \code{97}.  This is the inverse of
-  \function{chr()}.
+  Return the \ASCII{} value of a string of one character or a Unicode
+  character.  E.g., \code{ord('a')} returns the integer \code{97},
+  \code{ord(u'\\u2020')} returns \code{8224}.  This is the inverse of
+  \function{chr()} for strings and of \function{unichr()} for Unicode
+  characters.
 \end{funcdesc}
 
 \begin{funcdesc}{pow}{x, y\optional{, z}}
@@ -658,6 +660,21 @@
 >>> import types
 >>> if type(x) == types.StringType: print "It's a string"
 \end{verbatim}
+\end{funcdesc}
+
+\begin{funcdesc}{unichr}{i}
+  Return the Unicode string of one character whose Unicode code is the
+  integer \var{i}, e.g., \code{unichr(97)} returns the string
+  \code{u'a'}.  This is the inverse of \function{ord()} for Unicode
+  strings.  The argument must be in the range [0..65535], inclusive.
+  A \exception{ValueError} is raised otherwise.
+\end{funcdesc}
+
+\begin{funcdesc}{unicode}{string\optional{,encoding='utf-8'}\optional{,errors='strict'}}
+Decodes \var{string} using the codec found for \var{encoding}. Error
+handling is done according to \var{errors}. Default is to decode UTF-8
+in strict mode, meaning that encoding errors raise a
+\exception{ValueError}.
 \end{funcdesc}
 
 \begin{funcdesc}{vars}{\optional{object}}
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/lib/libmarshal.tex Python+Unicode/Doc/lib/libmarshal.tex
--- CVS-Python/Doc/lib/libmarshal.tex	Mon Apr 26 21:02:36 1999
+++ Python+Unicode/Doc/lib/libmarshal.tex	Sat Apr  1 17:13:45 2000
@@ -32,11 +32,11 @@
 whose value is independent from a particular invocation of Python can
 be written and read by this module.  The following types are supported:
 \code{None}, integers, long integers, floating point numbers,
-strings, tuples, lists, dictionaries, and code objects, where it
-should be understood that tuples, lists and dictionaries are only
-supported as long as the values contained therein are themselves
-supported; and recursive lists and dictionaries should not be written
-(they will cause infinite loops).
+strings, Unicode objects, tuples, lists, dictionaries, and code
+objects, where it should be understood that tuples, lists and
+dictionaries are only supported as long as the values contained
+therein are themselves supported; and recursive lists and dictionaries
+should not be written (they will cause infinite loops).
 
 \strong{Caveat:} On machines where C's \code{long int} type has more than
 32 bits (such as the DEC Alpha), it
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/lib/libpickle.tex Python+Unicode/Doc/lib/libpickle.tex
--- CVS-Python/Doc/lib/libpickle.tex	Tue Jul 27 14:12:21 1999
+++ Python+Unicode/Doc/lib/libpickle.tex	Sat Apr  1 17:13:45 2000
@@ -199,7 +199,7 @@
 
 \item integers, long integers, floating point numbers
 
-\item strings
+\item normal and Unicode strings
 
 \item tuples, lists and dictionaries containing only picklable objects
 
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/lib/libtypes.tex Python+Unicode/Doc/lib/libtypes.tex
--- CVS-Python/Doc/lib/libtypes.tex	Mon Dec 21 23:24:12 1998
+++ Python+Unicode/Doc/lib/libtypes.tex	Sat Apr  1 17:13:45 2000
@@ -56,6 +56,10 @@
 The type of character strings (e.g. \code{'Spam'}).
 \end{datadesc}
 
+\begin{datadesc}{UnicodeType}
+The type of Unicode character strings (e.g. \code{u'Spam'}).
+\end{datadesc}
+
 \begin{datadesc}{TupleType}
 The type of tuples (e.g. \code{(1, 2, 3, 'Spam')}).
 \end{datadesc}
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/ref/ref3.tex Python+Unicode/Doc/ref/ref3.tex
--- CVS-Python/Doc/ref/ref3.tex	Fri Dec  3 13:09:39 1999
+++ Python+Unicode/Doc/ref/ref3.tex	Sat Apr  1 18:45:21 2000
@@ -278,6 +278,21 @@
 \bifuncindex{chr}
 \bifuncindex{ord}
 
+\item[Unicode]
+The items of a Unicode object are Unicode characters.  A Unicode
+character is represented by a Unicode object of one item and can hold
+a 16-bit value representing a Unicode ordinal.  The built-in functions
+\function{unichr()}\bifuncindex{unichr} and
+\function{ord()}\bifuncindex{ord} convert between characters and
+nonnegative integers representing the Unicode ordinals as defined in
+the Unicode Standard 3.0. Conversion from and to other encodings are
+possible through the Unicode method \method{encode} and the built-in
+function \function{unicode()}\bifuncindex{unicode}.
+\obindex{unicode}
+\index{character}
+\index{integer}
+\index{Unicode@\UNICODE{}}
+
 \item[Tuples]
 The items of a tuple are arbitrary Python objects.
 Tuples of two or more items are formed by comma-separated lists
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/texinputs/python.sty Python+Unicode/Doc/texinputs/python.sty
--- CVS-Python/Doc/texinputs/python.sty	Fri Dec  3 13:09:39 1999
+++ Python+Unicode/Doc/texinputs/python.sty	Sat Apr  1 17:13:45 2000
@@ -660,6 +660,7 @@
 \newcommand{\UNIX}{{\sc Unix}}
 \newcommand{\POSIX}{POSIX}
 \newcommand{\ASCII}{{\sc ascii}}
+\newcommand{\UNICODE}{{\sc unicode}}
 \newcommand{\Cpp}{C\protect\raisebox{.18ex}{++}}
 \newcommand{\C}{C}
 \newcommand{\EOF}{{\sc eof}}
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS CVS-Python/Doc/tut/tut.tex Python+Unicode/Doc/tut/tut.tex
--- CVS-Python/Doc/tut/tut.tex	Fri Dec  3 13:09:39 1999
+++ Python+Unicode/Doc/tut/tut.tex	Sat Apr  1 17:13:46 2000
@@ -699,6 +699,104 @@
 34
 \end{verbatim}
 
+\subsection{Unicode Strings \label{unicodeStrings}}
+
+Starting with Python 1.6 a new data type for storing text data is
+available to the programmer: the Unicode object. It can be used to
+store and manipulate Unicode data (see <http://www.unicode.org>) and
+intergrates well with the existing string objects providing
+auto-conversions where necessary.
+
+Unicode has the advantage of providing one ordinal for every character
+in every script used in modern and ancient texts. Previously, there
+were only 256 possible ordinals for script characters and texts were
+typically bound to a code page which mapped the ordinals to script
+characters. This lead to very much confusion especially with respect
+to internalization (usually written as 'i18n' -- 'i' + 18 characters +
+'n') of software. Unicode solves these problems by defining one code
+page for all scripts.
+
+Creating Unicode strings in Python is just as simple as creating
+normal strings:
+
+\begin{verbatim}
+>>> u'Hello World !'
+u'Hello World !'
+\end{verbatim}
+
+The small 'u' in front of the quote indicates that an Unicode string
+is supposed to be created. If you want to include special characters
+in the string, you can do so by using the Python \emph{Unicode-Escape}
+encoding. The following example shows how:
+
+\begin{verbatim}
+>>> u'Hello\\u0020World !'
+u'Hello World !'
+\end{verbatim}
+
+The escape sequence \code{\\u0020} indicates to insert the Unicode
+character with the HEX ordinal 0x0020 (the space character) at the
+given position.
+
+Other characters are interpreted by using their respective ordinal
+value directly as Unicode ordinal. Due to the fact that the lower 256
+Unicode are the same as the standard Latin-1 encoding used in many
+western countries, the process of entering Unicode is greatly
+simplified.
+
+For experts, there is also a raw mode just like for normal
+strings. You have to prepend the string with a small 'r' to have
+Python use the \emph{Raw-Unicode-Escape} encoding. It will only apply
+the above \code{\\uXXXX} conversion if there is an uneven number of
+backslashes in front of the small 'u'.
+
+\begin{verbatim}
+>>> ur'Hello\u0020World !'
+u'Hello World !'
+>>> ur'Hello\\u0020World !'
+u'Hello\\\\u0020World !'
+\end{verbatim}
+
+The raw mode is most useful when you have to enter lots of backslashes
+e.g. in regular expressions.
+
+Apart from these standard encodings, Python provides a whole set of
+other ways of creating Unicod strings on the basis of a known
+encoding. 
+
+The builtin \code{unicode()} provides access to all registered Unicode
+codecs (COders and DECoders). Some of the more well known encodings
+which these codecs can convert are
+\emph{Latin-1}, \emph{ASCII}, \emph{UTF-8} and \emph{UTF-16}. The latter two
+are variable length encodings which permit to store Unicode characters
+in 8 or 16 bits. Python uses UTF-8 as default encoding. This becomes
+noticable when printing Unicode strings or writing them to files.
+
+\begin{verbatim}
+>>> u"äöü"
+u'\344\366\374'
+>>> str(u"äöü")
+'\303\244\303\266\303\274'
+\end{verbatim}
+
+If you have data in a specific encoding and want to produce a
+corresponding Unicode string from it, you can use the \code{unicode()}
+builtin with the encoding name as second argument.
+
+\begin{verbatim}
+>>> unicode('\303\244\303\266\303\274','UTF-8')
+u'\344\366\374'
+\end{verbatim}
+
+To convert the Unicode string back into a string using the original
+encoding, the objects provide an \code{.encode()} method.
+
+\begin{verbatim}
+>>> u"äöü".encode('UTF-8')
+'\303\244\303\266\303\274'
+\end{verbatim}
+
+
 \subsection{Lists \label{lists}}
 
 Python knows a number of \emph{compound} data types, used to group

--------------AC5A3AFAAB32FC257EC8E73D--




From pf@artcom-gmbh.de  Sun Apr  2 23:58:04 2000
From: pf@artcom-gmbh.de (Peter Funk)
Date: Mon, 3 Apr 2000 00:58:04 +0200 (MEST)
Subject: [Patches] New lib module: UserString.py and testcase for it
Message-ID: <m12btJl-000CnCC@artcom0.artcom-gmbh.de>

Dear Python patcher!

Please add the files appended below and commit them into the CVS tree.
(Note: the patch submission guide lines are not very clear about the 
 mail format for a suite of new module files.  I've used good old 'shar'.  
 Please tell me if I should submit a context diff against null or a 
 set of mime attachments instead.)
--=-- argument: --=--=--=--=--=--=--=--=--=--=-->8--=-
In Python 1.6 string objects have grown a bunch of methods.
Users may wish to override particular methods in self defined classes.
UserString wraps builtin string objects into a wrapper class similar 
to UserDict and UserList.  UserString object instances are still immutable.
For educational purposes a class `MutableString' derived from `UserString'
is also provided.
--=-- obligatory disclaimer: -=--=--=--=--=--=-->8--=-
I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.
--=-- dry signature: =--=--=--=--=--=--=--=--=-->8--=-
note: In order to fullfil the formal requirements I have submitted
      the wet signature form per postal mail.  This should arrive at your
      site in a week or so (I've applied air mail postage).

Regards, Peter
-- 
Peter Funk, Oldenburger Str.86, D-27777 Ganderkesee, Germany, Fax:+49 4222950260
office: +49 421 20419-0 (ArtCom GmbH, Grazer Str.8, D-28359 Bremen)
--=-- shar file: --=--=--=--=--=--=--=--=--=--=--=-->8--=-
#!/bin/sh
# This is a shell archive (produced by GNU sharutils 4.2c).
# To extract the files from this archive, save it to some FILE, remove
# everything before the `!/bin/sh' line above, then type `sh FILE'.
#
# Made on 2000-04-03 00:53 MEST by <pf@artcom0>.
# Source directory was `/home/pf/archiv/freeware/python/CVS_01_04_00/dist'.
#
# Existing files will *not* be overwritten unless `-c' is specified.
#
# This shar contains:
# length mode       name
# ------ ---------- ------------------------------------------
#   7077 -rwxrwxr-x src/Lib/UserString.py
#   6599 -rwxrwxr-x src/Lib/test/test_userstring.py
#     16 -rw-rw---- src/Lib/test/output/test_userstring
#
save_IFS="${IFS}"
IFS="${IFS}:"
gettext_dir=FAILED
locale_dir=FAILED
first_param="$1"
for dir in $PATH
do
  if test "$gettext_dir" = FAILED && test -f $dir/gettext \
     && ($dir/gettext --version >/dev/null 2>&1)
  then
    set `$dir/gettext --version 2>&1`
    if test "$3" = GNU
    then
      gettext_dir=$dir
    fi
  fi
  if test "$locale_dir" = FAILED && test -f $dir/shar \
     && ($dir/shar --print-text-domain-dir >/dev/null 2>&1)
  then
    locale_dir=`$dir/shar --print-text-domain-dir`
  fi
done
IFS="$save_IFS"
if test "$locale_dir" = FAILED || test "$gettext_dir" = FAILED
then
  echo=echo
else
  TEXTDOMAINDIR=$locale_dir
  export TEXTDOMAINDIR
  TEXTDOMAIN=sharutils
  export TEXTDOMAIN
  echo="$gettext_dir/gettext -s"
fi
if (echo "testing\c"; echo 1,2,3) | grep c >/dev/null; then
  if (echo -n testing; echo 1,2,3) | sed s/-n/xn/ | grep xn >/dev/null; then
    shar_n= shar_c='
'
  else
    shar_n=-n shar_c=
  fi
else
  shar_n= shar_c='\c'
fi
touch -am 1231235999 $$.touch >/dev/null 2>&1
if test ! -f 1231235999 && test -f $$.touch; then
  shar_touch=touch
else
  shar_touch=:
  echo
  $echo 'WARNING: not restoring timestamps.  Consider getting and'
  $echo "installing GNU \`touch', distributed in GNU File Utilities..."
  echo
fi
rm -f 1231235999 $$.touch
#
$echo $shar_n 'x -' 'lock directory' "\`_sh29194': "$shar_c
if mkdir _sh29194; then
  $echo 'created'
else
  $echo 'failed to create'
  exit 1
fi
# ============= src/Lib/UserString.py ==============
if test ! -d 'src'; then
  $echo $echo_n 'x -' 'src: '$echo_c
  if mkdir 'src'; then $echo 'created'; else $echo 'failed to create'; fi
fi
if test ! -d 'src/Lib'; then
  $echo $echo_n 'x -' 'src/Lib: '$echo_c
  if mkdir 'src/Lib'; then $echo 'created'; else $echo 'failed to create'; fi
fi
if test -f 'src/Lib/UserString.py' && test "$first_param" != -c; then
  $echo 'x -' SKIPPING 'src/Lib/UserString.py' '(file already exists)'
else
  $echo 'x -' extracting 'src/Lib/UserString.py' '(text)'
  sed 's/^X//' << 'SHAR_EOF' > 'src/Lib/UserString.py' &&
#!/usr/bin/env python
## vim:ts=4:et:nowrap
"""A user-defined wrapper around string objects
X
Note: string objects have grown methods in Python 1.6 
This module requires Python 1.6 or later.
"""
from types import StringType, UnicodeType
import sys
X
class UserString:
X    def __init__(self, seq):
X        if isinstance(seq, StringType) or isinstance(seq, UnicodeType):
X            self.data = seq
X        elif isinstance(seq, UserString):
X            self.data = seq.data[:]
X        else: 
X            self.data = str(seq)
X    def __str__(self): return str(self.data)
X    def __repr__(self): return repr(self.data)
X    def __int__(self): return int(self.data)
X    def __long__(self): return long(self.data)
X    def __float__(self): return float(self.data)
X    def __complex__(self): return complex(self.data)
X    def __hash__(self): return hash(self.data)
X
X    def __cmp__(self, string):
X        if isinstance(string, UserString):
X            return cmp(self.data, string.data)
X        else:
X            return cmp(self.data, string)
X    def __contains__(self, char):
X        return char in self.data
X
X    def __len__(self): return len(self.data)
X    def __getitem__(self, index): return self.__class__(self.data[index])
X    def __getslice__(self, start, end):
X        start = max(start, 0); end = max(end, 0)
X        return self.__class__(self.data[start:end])
X
X    def __add__(self, other):
X        if isinstance(other, UserString):
X            return self.__class__(self.data + other.data)
X        elif isinstance(other, StringType) or isinstance(other, UnicodeType):
X            return self.__class__(self.data + other)
X        else:
X            return self.__class__(self.data + str(other))
X    def __radd__(self, other):
X        if isinstance(other, StringType) or isinstance(other, UnicodeType):
X            return self.__class__(other + self.data)
X        else:
X            return self.__class__(str(other) + self.data)
X    def __mul__(self, n):
X        return self.__class__(self.data*n)
X    __rmul__ = __mul__
X
X    # the following methods are defined in alphabetical order:
X    def capitalize(self): return self.__class__(self.data.capitalize())
X    def center(self, width): return self.__class__(self.data.center(width))
X    def count(self, sub, start=0, end=sys.maxint):
X        return self.data.count(sub, start, end)
X    def encode(self, encoding=None, errors=None): # XXX improve this?
X        if encoding:
X            if errors:
X                return self.__class__(self.data.encode(encoding, errors))
X            else:
X                return self.__class__(self.data.encode(encoding))
X        else: 
X            return self.__class__(self.data.encode())
X    def endswith(self, suffix, start=0, end=sys.maxint):
X        return self.data.endswith(suffix, start, end)
X    def expandtabs(self, tabsize=8): 
X        return self.__class__(self.data.expandtabs(tabsize))
X    def find(self, sub, start=0, end=sys.maxint): 
X        return self.data.find(sub, start, end)
X    def index(self, sub, start=0, end=sys.maxint): 
X        return self.data.index(sub, start, end)
X    def isdecimal(self): return self.data.isdecimal()
X    def isdigit(self): return self.data.isdigit()
X    def islower(self): return self.data.islower()
X    def isnumeric(self): return self.data.isnumeric()
X    def isspace(self): return self.data.isspace()
X    def istitle(self): return self.data.istitle()
X    def isupper(self): return self.data.isupper()
X    def join(self, seq): return self.data.join(seq)
X    def ljust(self, width): return self.__class__(self.data.ljust(width))
X    def lower(self): return self.__class__(self.data.lower())
X    def lstrip(self): return self.__class__(self.data.lstrip())
X    def replace(self, old, new, maxsplit=-1): 
X        return self.__class__(self.data.replace(old, new, maxsplit))
X    def rfind(self, sub, start=0, end=sys.maxint): 
X        return self.data.rfind(sub, start, end)
X    def rindex(self, sub, start=0, end=sys.maxint): 
X        return self.data.rindex(sub, start, end)
X    def rjust(self, width): return self.__class__(self.data.rjust(width))
X    def rstrip(self): return self.__class__(self.data.rstrip())
X    def split(self, sep=None, maxsplit=-1): 
X        return self.data.split(sep, maxsplit)
X    def splitlines(self, maxsplit=-1): return self.data.splitlines(maxsplit)
X    def startswith(self, prefix, start=0, end=sys.maxint): 
X        return self.data.startswith(prefix, start, end)
X    def strip(self): return self.__class__(self.data.strip())
X    def swapcase(self): return self.__class__(self.data.swapcase())
X    def title(self): return self.__class__(self.data.title())
X    def translate(self, table, deletechars=""): 
X        return self.__class__(self.data.translate(table, deletechars))
X    def upper(self): return self.__class__(self.data.upper())
X
class MutableString(UserString):
X    """mutable string objects
X
X    Python strings are immutable objects.  This has the advantage, that
X    strings may be used as dictionary keys.  If this property isn't needed
X    and you insist on changing string values in place instead, you may cheat
X    and use MutableString.
X
X    But the purpose of this class is an educational one: to prevent
X    people from inventing their own mutable string class derived
X    from UserString and than forget thereby to remove (override) the
X    __hash__ method inherited from ^UserString.  This would lead to
X    errors that would be very hard to track down.
X
X    A faster and better solution is to rewrite your program using lists."""
X    def __init__(self, string=""):
X        self.data = string
X    def __hash__(self): 
X        raise TypeError, "unhashable type (it is mutable)"
X    def __setitem__(self, index, sub):
X        if index < 0 or index >= len(self.data): raise IndexError
X        self.data = self.data[:index] + sub + self.data[index+1:]
X    def __delitem__(self, index):
X        if index < 0 or index >= len(self.data): raise IndexError
X        self.data = self.data[:index] + self.data[index+1:]
X    def __setslice__(self, start, end, sub):
X        start = max(start, 0); end = max(end, 0)
X        if isinstance(sub, UserString):
X            self.data = self.data[:start]+sub.data+self.data[end:]
X        elif isinstance(sub, StringType) or isinstance(sub, UnicodeType):
X            self.data = self.data[:start]+sub+self.data[end:]
X        else:
X            self.data =  self.data[:start]+str(sub)+self.data[end:]
X    def __delslice__(self, start, end):
X        start = max(start, 0); end = max(end, 0)
X        self.data = self.data[:start] + self.data[end:]
X    def immutable(self):
X        return UserString(self.data)
X    
if __name__ == "__main__":
X    # execute the regression test to stdout, if called as a script:
X    import os
X    called_in_dir, called_as = os.path.split(sys.argv[0])
X    called_in_dir = os.path.abspath(called_in_dir)
X    called_as, py = os.path.splitext(called_as)
X    sys.path.append(os.path.join(called_in_dir, 'test'))
X    if '-q' in sys.argv:
X        import test_support
X        test_support.verbose = 0
X    __import__('test_' + called_as.lower())
SHAR_EOF
  $shar_touch -am 04030006100 'src/Lib/UserString.py' &&
  chmod 0775 'src/Lib/UserString.py' ||
  $echo 'restore of' 'src/Lib/UserString.py' 'failed'
  if ( md5sum --help </dev/null 2>&1 | grep 'sage: md5sum \[' ) >/dev/null 2>&1 \
  && ( md5sum --version </dev/null 2>&1 | grep -v 'textutils 1.12' ) >/dev/null; then
    md5sum -c << SHAR_EOF >/dev/null 2>&1 \
    || $echo 'src/Lib/UserString.py:' 'MD5 check failed'
16ba5d7ab3cefa6d75d8fe270e05957a  src/Lib/UserString.py
SHAR_EOF
  else
    shar_count="`LC_ALL=C wc -c < 'src/Lib/UserString.py'`"
    test 7077 -eq "$shar_count" ||
    $echo 'src/Lib/UserString.py:' 'original size' '7077,' 'current size' "$shar_count!"
  fi
fi
# ============= src/Lib/test/test_userstring.py ==============
if test ! -d 'src/Lib/test'; then
  $echo $echo_n 'x -' 'src/Lib/test: '$echo_c
  if mkdir 'src/Lib/test'; then $echo 'created'; else $echo 'failed to create'; fi
fi
if test -f 'src/Lib/test/test_userstring.py' && test "$first_param" != -c; then
  $echo 'x -' SKIPPING 'src/Lib/test/test_userstring.py' '(file already exists)'
else
  $echo 'x -' extracting 'src/Lib/test/test_userstring.py' '(text)'
  sed 's/^X//' << 'SHAR_EOF' > 'src/Lib/test/test_userstring.py' &&
#!/usr/bin/env python
import sys, string
from test_support import verbose
# UserString is a wrapper around the native builtin string type.
# UserString instances should behave similar to builtin string objects.
# The test cases were in part derived from 'test_string.py'.
from UserString import UserString
X
if __name__ == "__main__":
X    verbose = 0
X
tested_methods = {}
X
def test(methodname, input, *args):
X    global tested_methods
X    tested_methods[methodname] = 1
X    if verbose:
X        print '%s.%s(%s) ' % (input, methodname, args),
X    u = UserString(input)
X    objects = [input, u, UserString(u)]
X    res = [""] * 3
X    for i in range(3):
X	object = objects[i]
X	try:
X	    f = getattr(object, methodname)
X	    res[i] = apply(f, args)
X	except:
X	    res[i] = sys.exc_type
X    if res[0] != res[1]:
X        if verbose:
X            print 'no'
X        print `input`, f, `res[0]`, "<>", `res[1]`
X    else:
X        if verbose:
X            print 'yes'
X    if res[1] != res[2]:
X        if verbose:
X            print 'no'
X        print `input`, f, `res[1]`, "<>", `res[2]`
X    else:
X        if verbose:
X            print 'yes'
X
test('capitalize', ' hello ')
test('capitalize', 'hello ')
X
test('center', 'foo', 0)
test('center', 'foo', 3)
test('center', 'foo', 16)
X
test('ljust', 'foo', 0)
test('ljust', 'foo', 3)
test('ljust', 'foo', 16)
X
test('rjust', 'foo', 0)
test('rjust', 'foo', 3)
test('rjust', 'foo', 16)
X
test('count', 'abcabcabc', 'abc')
test('count', 'abcabcabc', 'abc', 1)
test('count', 'abcabcabc', 'abc', -1)
test('count', 'abcabcabc', 'abc', 7)
test('count', 'abcabcabc', 'abc', 0, 3)
test('count', 'abcabcabc', 'abc', 0, 333)
X
test('find', 'abcdefghiabc', 'abc')
test('find', 'abcdefghiabc', 'abc', 1)
test('find', 'abcdefghiabc', 'def', 4)
test('rfind', 'abcdefghiabc', 'abc')
X
test('index', 'abcabcabc', 'abc')
test('index', 'abcabcabc', 'abc', 1)
test('index', 'abcabcabc', 'abc', -1)
test('index', 'abcabcabc', 'abc', 7)
test('index', 'abcabcabc', 'abc', 0, 3)
test('index', 'abcabcabc', 'abc', 0, 333)
X
test('rindex', 'abcabcabc', 'abc')
test('rindex', 'abcabcabc', 'abc', 1)
test('rindex', 'abcabcabc', 'abc', -1)
test('rindex', 'abcabcabc', 'abc', 7)
test('rindex', 'abcabcabc', 'abc', 0, 3)
test('rindex', 'abcabcabc', 'abc', 0, 333)
X
X
test('lower', 'HeLLo')
test('lower', 'hello')
test('upper', 'HeLLo')
test('upper', 'HELLO')
X
test('title', ' hello ')
test('title', 'hello ')
test('title', "fOrMaT thIs aS titLe String")
test('title', "fOrMaT,thIs-aS*titLe;String")
test('title', "getInt")
X
test('expandtabs', 'abc\rab\tdef\ng\thi')
test('expandtabs', 'abc\rab\tdef\ng\thi', 8)
test('expandtabs', 'abc\rab\tdef\ng\thi', 4)
test('expandtabs', 'abc\r\nab\tdef\ng\thi', 4)
X
test('islower', 'a')
test('islower', 'A')
test('islower', '\n')
test('islower', 'abc')
test('islower', 'aBc')
test('islower', 'abc\n')
X
test('isupper', 'a')
test('isupper', 'A')
test('isupper', '\n')
test('isupper', 'ABC')
test('isupper', 'AbC')
test('isupper', 'ABC\n')
X
test('isdigit', '  0123456789')
test('isdigit', '56789')
test('isdigit', '567.89')
test('isdigit', '0123456789abc')
X
test('isspace', '')
test('isspace', ' ')
test('isspace', ' \t')
test('isspace', ' \t\f\n')
X
test('istitle', 'a')
test('istitle', 'A')
test('istitle', '\n')
test('istitle', 'A Titlecased Line')
test('istitle', 'A\nTitlecased Line')
test('istitle', 'A Titlecased, Line')
test('istitle', 'Not a capitalized String')
test('istitle', 'Not\ta Titlecase String')
test('istitle', 'Not--a Titlecase String')
X
test('splitlines', "abc\ndef\n\rghi")
test('splitlines', "abc\ndef\n\r\nghi")
test('splitlines', "abc\ndef\r\nghi")
test('splitlines', "abc\ndef\r\nghi\n")
test('splitlines', "abc\ndef\r\nghi\n\r")
test('splitlines', "\nabc\ndef\r\nghi\n\r")
test('splitlines', "\nabc\ndef\r\nghi\n\r")
test('splitlines', "\nabc\ndef\r\nghi\n\r")
X
test('split', 'this is the split function')
test('split', 'a|b|c|d', '|')
test('split', 'a|b|c|d', '|', 2)
test('split', 'a b c d', None, 1)
test('split', 'a b c d', None, 2)
test('split', 'a b c d', None, 3)
test('split', 'a b c d', None, 4)
test('split', 'a b c d', None, 0)
test('split', 'a  b  c  d', None, 2)
test('split', 'a b c d ')
X
# join now works with any sequence type
class Sequence:
X    def __init__(self): self.seq = 'wxyz'
X    def __len__(self): return len(self.seq)
X    def __getitem__(self, i): return self.seq[i]
X
test('join', '', ('a', 'b', 'c', 'd'))
test('join', '', Sequence())
test('join', '', 7)
X
class BadSeq(Sequence):
X    def __init__(self): self.seq = [7, 'hello', 123L]
X
test('join', '', BadSeq())
X
test('strip', '   hello   ')
test('lstrip', '   hello   ')
test('rstrip', '   hello   ')
test('strip', 'hello')
X
test('swapcase', 'HeLLo cOmpUteRs')
transtable = string.maketrans("abc", "xyz")
test('translate', 'xyzabcdef', transtable, 'def')
X
transtable = string.maketrans('a', 'A')
test('translate', 'abc', transtable)
test('translate', 'xyz', transtable)
X
test('replace', 'one!two!three!', '!', '@', 1)
test('replace', 'one!two!three!', '!', '')
test('replace', 'one!two!three!', '!', '@', 2)
test('replace', 'one!two!three!', '!', '@', 3)
test('replace', 'one!two!three!', '!', '@', 4)
test('replace', 'one!two!three!', '!', '@', 0)
test('replace', 'one!two!three!', '!', '@')
test('replace', 'one!two!three!', 'x', '@')
test('replace', 'one!two!three!', 'x', '@', 2)
X
test('startswith', 'hello', 'he')
test('startswith', 'hello', 'hello')
test('startswith', 'hello', 'hello world')
test('startswith', 'hello', '')
test('startswith', 'hello', 'ello')
test('startswith', 'hello', 'ello', 1)
test('startswith', 'hello', 'o', 4)
test('startswith', 'hello', 'o', 5)
test('startswith', 'hello', '', 5)
test('startswith', 'hello', 'lo', 6)
test('startswith', 'helloworld', 'lowo', 3)
test('startswith', 'helloworld', 'lowo', 3, 7)
test('startswith', 'helloworld', 'lowo', 3, 6)
X
test('endswith', 'hello', 'lo')
test('endswith', 'hello', 'he')
test('endswith', 'hello', '')
test('endswith', 'hello', 'hello world')
test('endswith', 'helloworld', 'worl')
test('endswith', 'helloworld', 'worl', 3, 9)
test('endswith', 'helloworld', 'world', 3, 12)
test('endswith', 'helloworld', 'lowo', 1, 7)
test('endswith', 'helloworld', 'lowo', 2, 7)
test('endswith', 'helloworld', 'lowo', 3, 7)
test('endswith', 'helloworld', 'lowo', 4, 7)
test('endswith', 'helloworld', 'lowo', 3, 8)
test('endswith', 'ab', 'ab', 0, 1)
test('endswith', 'ab', 'ab', 0, 0)
X
# TODO: test cases for: int, long, float, complex, +, * and cmp
s = ""
for builtin_method in dir(s):
X    if not tested_methods.has_key(builtin_method):
X	print "no regression test case for method '"+builtin_method+"'"
X
SHAR_EOF
  $shar_touch -am 04030025100 'src/Lib/test/test_userstring.py' &&
  chmod 0775 'src/Lib/test/test_userstring.py' ||
  $echo 'restore of' 'src/Lib/test/test_userstring.py' 'failed'
  if ( md5sum --help </dev/null 2>&1 | grep 'sage: md5sum \[' ) >/dev/null 2>&1 \
  && ( md5sum --version </dev/null 2>&1 | grep -v 'textutils 1.12' ) >/dev/null; then
    md5sum -c << SHAR_EOF >/dev/null 2>&1 \
    || $echo 'src/Lib/test/test_userstring.py:' 'MD5 check failed'
bbbdf7e85913aa80e38e2469991b3cac  src/Lib/test/test_userstring.py
SHAR_EOF
  else
    shar_count="`LC_ALL=C wc -c < 'src/Lib/test/test_userstring.py'`"
    test 6599 -eq "$shar_count" ||
    $echo 'src/Lib/test/test_userstring.py:' 'original size' '6599,' 'current size' "$shar_count!"
  fi
fi
# ============= src/Lib/test/output/test_userstring ==============
if test ! -d 'src/Lib/test/output'; then
  $echo $echo_n 'x -' 'src/Lib/test/output: '$echo_c
  if mkdir 'src/Lib/test/output'; then $echo 'created'; else $echo 'failed to create'; fi
fi
if test -f 'src/Lib/test/output/test_userstring' && test "$first_param" != -c; then
  $echo 'x -' SKIPPING 'src/Lib/test/output/test_userstring' '(file already exists)'
else
  $echo 'x -' extracting 'src/Lib/test/output/test_userstring' '(text)'
  sed 's/^X//' << 'SHAR_EOF' > 'src/Lib/test/output/test_userstring' &&
test_userstring
SHAR_EOF
  $shar_touch -am 04030030100 'src/Lib/test/output/test_userstring' &&
  chmod 0660 'src/Lib/test/output/test_userstring' ||
  $echo 'restore of' 'src/Lib/test/output/test_userstring' 'failed'
  if ( md5sum --help </dev/null 2>&1 | grep 'sage: md5sum \[' ) >/dev/null 2>&1 \
  && ( md5sum --version </dev/null 2>&1 | grep -v 'textutils 1.12' ) >/dev/null; then
    md5sum -c << SHAR_EOF >/dev/null 2>&1 \
    || $echo 'src/Lib/test/output/test_userstring:' 'MD5 check failed'
8a16c32581de42c338366cb9f7488e4c  src/Lib/test/output/test_userstring
SHAR_EOF
  else
    shar_count="`LC_ALL=C wc -c < 'src/Lib/test/output/test_userstring'`"
    test 16 -eq "$shar_count" ||
    $echo 'src/Lib/test/output/test_userstring:' 'original size' '16,' 'current size' "$shar_count!"
  fi
fi
$echo $shar_n 'x -' 'lock directory' "\`_sh29194': " $shar_c
if rm -fr _sh29194; then
  $echo 'removed'
else
  $echo 'failed to remove'
fi
exit 0


From fdrake@acm.org  Mon Apr  3 04:55:15 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Sun, 2 Apr 2000 23:55:15 -0400 (EDT)
Subject: [Patches] New lib module: UserString.py and testcase for it
In-Reply-To: <m12btJl-000CnCC@artcom0.artcom-gmbh.de>
References: <m12btJl-000CnCC@artcom0.artcom-gmbh.de>
Message-ID: <14568.5667.841020.109795@seahag.cnri.reston.va.us>

Peter Funk writes:
 > Please add the files appended below and commit them into the CVS tree.
 > (Note: the patch submission guide lines are not very clear about the 
 >  mail format for a suite of new module files.  I've used good old 'shar'.  
 >  Please tell me if I should submit a context diff against null or a 
 >  set of mime attachments instead.)

Peter,
  It's in, but I think there's a problem with the anonCVS getting
updated from my checkins.  I'll look further into that tomorrow.  ;(
  Attachments are better for me, but can deal with shar as well.  Just
hope I never have to use a non-Unix box for development!  ;)
  Thanks!


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From helsten@inconnect.com  Mon Apr  3 07:15:28 2000
From: helsten@inconnect.com (Lance Finn Helsten)
Date: Mon, 3 Apr 2000 00:15:28 -0600
Subject: [Patches] netrc module parse bug
Message-ID: <200004030615.AAA15438@helsten.dsl.inconnect.com>

<fixed><smaller><fontfamily><param>Courier</param>When a .netrc file has =
passwords that contain non-alphanumeric characters the netrc.__init__ =
throws an exception. To change this in the .netrc would pose security =
risks.


Changes:

    1) Added all 7-bit characters to the wordchars field.

    2) Turned off quotes.

    3) Turn '#' off and on as a comment character around the top level =
get_token, so it is usable in passwords.


This assumes that all comments start in column 0. End of line comments =
will probably not work.


*** netrc.py    Fri Feb  4 08:10:33 2000

--- /usr/local/lib/python1.5/netrc.py   Sun Apr  2 23:46:31 2000

***************

*** 15,24 ****

          self.hosts =3D {}

          self.macros =3D {}

          lexer =3D shlex.shlex(fp)

!         lexer.wordchars =3D lexer.wordchars + '.'

          while 1:

              # Look for a machine, default, or macdef top-level keyword

              toplevel =3D tt =3D lexer.get_token()

              if tt =3D=3D '' or tt =3D=3D None:

                  break

              elif tt =3D=3D 'machine':

--- 13,25 ----

          self.hosts =3D {}

          self.macros =3D {}

          lexer =3D shlex.shlex(fp)

!         lexer.wordchars =3D lexer.wordchars + =
'~!@#$%^&*()-=3D+`[]{}\\|;:,./<<>?\'"'

!         lexer.quotes =3D ''

          while 1:

              # Look for a machine, default, or macdef top-level keyword

+             lexer.commenters =3D '#'

              toplevel =3D tt =3D lexer.get_token()

+             lexer.commenters =3D ''

              if tt =3D=3D '' or tt =3D=3D None:

                  break

              elif tt =3D=3D 'machine':

***************

<flushleft>

I confirm that, to the best of my knowledge and belief, this

contribution is free of any claims of third parties under

copyright, patent or other rights or interests ("claims").  To

the extent that I have any such claims, I hereby grant to CNRI a

nonexclusive, irrevocable, royalty-free, worldwide license to

reproduce, distribute, perform and/or display publicly, prepare

derivative versions, and otherwise use this contribution as part

of the Python software and its related documentation, or any

derivative versions thereof, at no cost to CNRI or its licensed

users, and to authorize others to do so.


I acknowledge that CNRI may, at its sole discretion, decide

whether or not to incorporate this contribution in the Python

software and its related documentation.  I further grant CNRI

permission to use my name and other identifying information

provided to CNRI by me for use in connection with the Python

software and its related documentation.



Lance Finn Helsten

helsten@inconnect.com

helsten@ieee.org

helsten@acm.org


You cannot free a slave save he do it himself. You cannot enslave a =
freeman, the most you can do is kill him. ~Robert A. Heinlien=


From m.favas@per.dem.csiro.au  Mon Apr  3 08:14:36 2000
From: m.favas@per.dem.csiro.au (Mark Favas)
Date: Mon, 03 Apr 2000 15:14:36 +0800
Subject: [Patches] Patches for bug report 258 - mmapmodule.c compile problem
Message-ID: <38E844DC.75391280@per.dem.csiro.au>

The following pathces fix the compile error and warnings for the CVS
version of mmapmodule.c of April 3 on Tru64 Unix 4.0F with the Compaq C
compiler.

I confirm that, to the best of my knowledge and belief, this
                  contribution is free of any claims of third parties
under
                  copyright, patent or other rights or interests
("claims").  To
                  the extent that I have any such claims, I hereby grant
to CNRI a
                  nonexclusive, irrevocable, royalty-free, worldwide
license to
                  reproduce, distribute, perform and/or display
publicly, prepare
                  derivative versions, and otherwise use this
contribution as part
                  of the Python software and its related documentation,
or any
                  derivative versions thereof, at no cost to CNRI or its
licensed
                  users, and to authorize others to do so.

                  I acknowledge that CNRI may, at its sole discretion,
decide
                  whether or not to incorporate this contribution in the
Python
                  software and its related documentation.  I further
grant CNRI
                  permission to use my name and other identifying
information
                  provided to CNRI by me for use in connection with the
Python
                  software and its related documentation.

Mark


-- 
Email - m.favas@per.dem.csiro.au       Postal - Mark C Favas
Phone - +61 8 9333 6268, 041 892 6074           CSIRO Exploration &
Mining
Fax   - +61 8 9333 6121                         Private Bag No 5
                                                Wembley, Western
Australia 6913

*** mmapmodule.c.orig   Mon Apr  3 10:23:19 2000
--- mmapmodule.c        Mon Apr  3 14:57:23 2000
***************
*** 118,124 ****
        char value;
        char * where = (self->data+self->pos);
        CHECK_VALID(NULL);
!       if ((where >= 0) && (where < (self->data+self->size))) {
                value = (char) *(where);
                self->pos += 1;
                return Py_BuildValue("c", (char) *(where));
--- 118,124 ----
        char value;
        char * where = (self->data+self->pos);
        CHECK_VALID(NULL);
!       if ((where >= (char *) 0) && (where < (self->data+self->size)))
{
                value = (char) *(where);
                self->pos += 1;
                return Py_BuildValue("c", (char) *(where));
***************
*** 617,623 ****
                                "mmap slice assignment is wrong size");
                return -1;
        }
!       buf = PyString_AsString(v);
        memcpy(self->data + ilow, buf, ihigh-ilow);
        return 0;
  }
--- 617,623 ----
                                "mmap slice assignment is wrong size");
                return -1;
        }
!       buf = (unsigned char *) PyString_AsString(v);
        memcpy(self->data + ilow, buf, ihigh-ilow);
        return 0;
  }
***************
*** 640,646 ****
                        "mmap assignment must be single-character
string");
                return -1;
        }
!       buf = PyString_AsString(v);
        self->data[i] = buf[0];
        return 0;
  }
--- 640,646 ----
                        "mmap assignment must be single-character
string");
                return -1;
        }
!       buf = (unsigned char *) PyString_AsString(v);
        self->data[i] = buf[0];
        return 0;
  }


From pf@artcom-gmbh.de  Mon Apr  3 10:55:41 2000
From: pf@artcom-gmbh.de (Peter Funk)
Date: Mon, 3 Apr 2000 11:55:41 +0200 (MEST)
Subject: [Patches] [1.6] dictionary objects: new method 'supplement'
Message-ID: <m12c3a9-000CnCC@artcom0.artcom-gmbh.de>

Dear Python patcher!

Please consider to apply the patch appended below and commit into the CVS tree.
It applies to: Python 1.6a1 as released on april 1st.
--=-- argument: --=--=--=--=--=--=--=--=--=--=-->8--=-
This patch adds a new method to dictionary and UserDict objects:
'.supplement()' is a "sibling" of '.update()', but it add only
those items that are not already there instead of replacing them.

This idea has been discussed on python-dev last month.
--=-- obligatory disclaimer: -=--=--=--=--=--=-->8--=-
I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.
--=-- dry signature: =--=--=--=--=--=--=--=--=-->8--=-
Regards, Peter
-- 
Peter Funk, Oldenburger Str.86, D-27777 Ganderkesee, Germany, Fax:+49 4222950260
office: +49 421 20419-0 (ArtCom GmbH, Grazer Str.8, D-28359 Bremen)
--=-- patch: --=--=--=--=--=--=--=--=--=--=--=-->8--=-
*** ../../cvs_01_04_00_orig/dist/src/Objects/dictobject.c	Fri Mar 31 11:45:02 2000
--- src/Objects/dictobject.c	Mon Apr  3 10:30:11 2000
***************
*** 734,739 ****
--- 734,781 ----
  }
  
  static PyObject *
+ dict_supplement(mp, args)
+       register dictobject *mp;
+       PyObject *args;
+ {
+ 	register int i;
+ 	dictobject *other;
+         dictentry *entry, *oldentry;
+ 	if (!PyArg_Parse(args, "O!", &PyDict_Type, &other))
+ 		return NULL;
+ 	if (other == mp)
+ 		goto done; /* a.supplement(a); nothing to do */
+ 	/* Do one big resize at the start, rather than incrementally
+ 	   resizing as we insert new items.  Expect that there will be
+ 	   no (or few) overlapping keys. */
+ 	if ((mp->ma_fill + other->ma_used)*3 >= mp->ma_size*2) {
+ 		if (dictresize(mp, (mp->ma_used + other->ma_used)*3/2) != 0)
+ 			return NULL;
+ 	}
+ 	for (i = 0; i < other->ma_size; i++) {
+ 		entry = &other->ma_table[i];
+ 		if (entry->me_value != NULL) {
+ 			oldentry = lookdict(mp, entry->me_key, entry->me_hash);
+ 			if (oldentry->me_value == NULL) {
+ 				/* TODO: optimize:
+ 				   'insertdict' does another call to 'lookdict'.
+ 				   But for sake of readability and symmetry with
+ 				   'dict_update' I didn't tried to avoid this.
+ 				   At least not now as we go into 1.6 alpha. */
+ 				Py_INCREF(entry->me_key);
+ 				Py_INCREF(entry->me_value);
+ 				insertdict(mp, entry->me_key, entry->me_hash,
+ 					   entry->me_value);
+ 			}
+ 		}
+ 	}
+   done:
+ 	Py_INCREF(Py_None);
+ 	return Py_None;
+ }
+ 
+ 
+ static PyObject *
  dict_copy(mp, args)
        register dictobject *mp;
        PyObject *args;
***************
*** 1045,1050 ****
--- 1087,1093 ----
  	{"clear",	(PyCFunction)dict_clear},
  	{"copy",	(PyCFunction)dict_copy},
  	{"get",         (PyCFunction)dict_get,          METH_VARARGS},
+ 	{"supplement",	(PyCFunction)dict_supplement},
  	{NULL,		NULL}		/* sentinel */
  };
  
*** ../../cvs_01_04_00_orig/dist/src/Lib/test/test_types.py	Wed Feb 23 23:23:17 2000
--- src/Lib/test/test_types.py	Mon Apr  3 10:41:53 2000
***************
*** 242,247 ****
--- 242,250 ----
  d.update({2:20})
  d.update({1:1, 2:2, 3:3})
  if d != {1:1, 2:2, 3:3}: raise TestFailed, 'dict update'
+ d.supplement({1:"not", 2:"neither", 4:4})
+ if d != {1:1, 2:2, 3:3, 4:4}: raise TestFailed, 'dict supplement'
+ del d[4]
  if d.copy() != {1:1, 2:2, 3:3}: raise TestFailed, 'dict copy'
  if {}.copy() != {}: raise TestFailed, 'empty dict copy'
  # dict.get()
*** ../../cvs_01_04_00_orig/dist/src/Lib/UserDict.py	Wed Feb  2 16:10:14 2000
--- src/Lib/UserDict.py	Mon Apr  3 10:45:17 2000
***************
*** 32,36 ****
--- 32,45 ----
          else:
              for k, v in dict.items():
                  self.data[k] = v
+     def supplement(self, dict):
+         if isinstance(dict, UserDict):
+             self.data.supplement(dict.data)
+         elif isinstance(dict, type(self.data)):
+             self.data.supplement(dict)
+         else:
+             for k, v in dict.items():
+ 		if not self.data.has_key(k):
+ 		    self.data[k] = v
      def get(self, key, failobj=None):
          return self.data.get(key, failobj)
*** ../../cvs_01_04_00_orig/dist/src/Lib/test/test_userdict.py	Fri Mar 26 16:32:02 1999
--- src/Lib/test/test_userdict.py	Mon Apr  3 10:50:29 2000
***************
*** 93,101 ****
--- 93,109 ----
  t.update(u2)
  assert t == u2
  
+ # Test supplement
+ 
+ t = UserDict(d1)
+ t.supplement(u2)
+ assert t == u2
+ 
  # Test get
  
  for i in u2.keys():
      assert u2.get(i) == u2[i]
      assert u1.get(i) == d1.get(i)
      assert u0.get(i) == d0.get(i)
+ 
+ # TODO: Add a test using dir({}) to test for unimplemented methods


From gstein@lyra.org  Mon Apr  3 11:12:58 2000
From: gstein@lyra.org (Greg Stein)
Date: Mon, 3 Apr 2000 03:12:58 -0700 (PDT)
Subject: [Patches] Patches for bug report 258 - mmapmodule.c compile
 problem
In-Reply-To: <38E844DC.75391280@per.dem.csiro.au>
Message-ID: <Pine.LNX.4.10.10004030309080.19464-100000@nebula.lyra.org>

On Mon, 3 Apr 2000, Mark Favas wrote:
>...
> *** mmapmodule.c.orig   Mon Apr  3 10:23:19 2000
> --- mmapmodule.c        Mon Apr  3 14:57:23 2000
> ***************
> *** 118,124 ****
>         char value;
>         char * where = (self->data+self->pos);
>         CHECK_VALID(NULL);
> !       if ((where >= 0) && (where < (self->data+self->size))) {
>                 value = (char) *(where);
>                 self->pos += 1;
>                 return Py_BuildValue("c", (char) *(where));
> --- 118,124 ----
>         char value;
>         char * where = (self->data+self->pos);
>         CHECK_VALID(NULL);
> !       if ((where >= (char *) 0) && (where < (self->data+self->size)))
> {
>                 value = (char) *(where);
>                 self->pos += 1;
>                 return Py_BuildValue("c", (char) *(where));

Nope... that should be (where >= self->data).

The more obvious test would be something like:

  if (self->pos >= 0 && self->pos < self->size)

and throw the error before ever worrying about setting up "where".

> ***************
> *** 617,623 ****
>                                 "mmap slice assignment is wrong size");
>                 return -1;
>         }
> !       buf = PyString_AsString(v);
>         memcpy(self->data + ilow, buf, ihigh-ilow);
>         return 0;
>   }
> --- 617,623 ----
>                                 "mmap slice assignment is wrong size");
>                 return -1;
>         }
> !       buf = (unsigned char *) PyString_AsString(v);
>         memcpy(self->data + ilow, buf, ihigh-ilow);
>         return 0;
>   }

Rather than casting, the definition of "buf" should change to
"const char *buf".

> ***************
> *** 640,646 ****
>                         "mmap assignment must be single-character
> string");
>                 return -1;
>         }
> !       buf = PyString_AsString(v);
>         self->data[i] = buf[0];
>         return 0;
>   }
> --- 640,646 ----
>                         "mmap assignment must be single-character
> string");
>                 return -1;
>         }
> !       buf = (unsigned char *) PyString_AsString(v);
>         self->data[i] = buf[0];
>         return 0;
>   }

Same here.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/



From gstein@lyra.org  Mon Apr  3 11:18:30 2000
From: gstein@lyra.org (Greg Stein)
Date: Mon, 3 Apr 2000 03:18:30 -0700 (PDT)
Subject: [Patches] [1.6] dictionary objects: new method 'supplement'
In-Reply-To: <m12c3a9-000CnCC@artcom0.artcom-gmbh.de>
Message-ID: <Pine.LNX.4.10.10004030315040.19464-100000@nebula.lyra.org>

I don't recall the termination of the discussion, but I don't know that
consensus was ever reached.

Personally, I find this of little value over the similar (not exact) code:

def supplement(dict, extra):
  d = extra.copy()
  d.update(dict)
  return d

If the dictionary needs to be modified in place, then the loop from your
UserDict.supplement would be used.

Another view: why keep adding methods to service all possible needs?

Cheers,
-g

On Mon, 3 Apr 2000, Peter Funk wrote:
> Dear Python patcher!
> 
> Please consider to apply the patch appended below and commit into the CVS tree.
> It applies to: Python 1.6a1 as released on april 1st.
> --=-- argument: --=--=--=--=--=--=--=--=--=--=-->8--=-
> This patch adds a new method to dictionary and UserDict objects:
> '.supplement()' is a "sibling" of '.update()', but it add only
> those items that are not already there instead of replacing them.
> 
> This idea has been discussed on python-dev last month.
> --=-- obligatory disclaimer: -=--=--=--=--=--=-->8--=-
> I confirm that, to the best of my knowledge and belief, this
> contribution is free of any claims of third parties under
> copyright, patent or other rights or interests ("claims").  To
> the extent that I have any such claims, I hereby grant to CNRI a
> nonexclusive, irrevocable, royalty-free, worldwide license to
> reproduce, distribute, perform and/or display publicly, prepare
> derivative versions, and otherwise use this contribution as part
> of the Python software and its related documentation, or any
> derivative versions thereof, at no cost to CNRI or its licensed
> users, and to authorize others to do so.
> 
> I acknowledge that CNRI may, at its sole discretion, decide
> whether or not to incorporate this contribution in the Python
> software and its related documentation.  I further grant CNRI
> permission to use my name and other identifying information
> provided to CNRI by me for use in connection with the Python
> software and its related documentation.
> --=-- dry signature: =--=--=--=--=--=--=--=--=-->8--=-
> Regards, Peter
> -- 
> Peter Funk, Oldenburger Str.86, D-27777 Ganderkesee, Germany, Fax:+49 4222950260
> office: +49 421 20419-0 (ArtCom GmbH, Grazer Str.8, D-28359 Bremen)
> --=-- patch: --=--=--=--=--=--=--=--=--=--=--=-->8--=-
> *** ../../cvs_01_04_00_orig/dist/src/Objects/dictobject.c	Fri Mar 31 11:45:02 2000
> --- src/Objects/dictobject.c	Mon Apr  3 10:30:11 2000
> ***************
> *** 734,739 ****
> --- 734,781 ----
>   }
>   
>   static PyObject *
> + dict_supplement(mp, args)
> +       register dictobject *mp;
> +       PyObject *args;
> + {
> + 	register int i;
> + 	dictobject *other;
> +         dictentry *entry, *oldentry;
> + 	if (!PyArg_Parse(args, "O!", &PyDict_Type, &other))
> + 		return NULL;
> + 	if (other == mp)
> + 		goto done; /* a.supplement(a); nothing to do */
> + 	/* Do one big resize at the start, rather than incrementally
> + 	   resizing as we insert new items.  Expect that there will be
> + 	   no (or few) overlapping keys. */
> + 	if ((mp->ma_fill + other->ma_used)*3 >= mp->ma_size*2) {
> + 		if (dictresize(mp, (mp->ma_used + other->ma_used)*3/2) != 0)
> + 			return NULL;
> + 	}
> + 	for (i = 0; i < other->ma_size; i++) {
> + 		entry = &other->ma_table[i];
> + 		if (entry->me_value != NULL) {
> + 			oldentry = lookdict(mp, entry->me_key, entry->me_hash);
> + 			if (oldentry->me_value == NULL) {
> + 				/* TODO: optimize:
> + 				   'insertdict' does another call to 'lookdict'.
> + 				   But for sake of readability and symmetry with
> + 				   'dict_update' I didn't tried to avoid this.
> + 				   At least not now as we go into 1.6 alpha. */
> + 				Py_INCREF(entry->me_key);
> + 				Py_INCREF(entry->me_value);
> + 				insertdict(mp, entry->me_key, entry->me_hash,
> + 					   entry->me_value);
> + 			}
> + 		}
> + 	}
> +   done:
> + 	Py_INCREF(Py_None);
> + 	return Py_None;
> + }
> + 
> + 
> + static PyObject *
>   dict_copy(mp, args)
>         register dictobject *mp;
>         PyObject *args;
> ***************
> *** 1045,1050 ****
> --- 1087,1093 ----
>   	{"clear",	(PyCFunction)dict_clear},
>   	{"copy",	(PyCFunction)dict_copy},
>   	{"get",         (PyCFunction)dict_get,          METH_VARARGS},
> + 	{"supplement",	(PyCFunction)dict_supplement},
>   	{NULL,		NULL}		/* sentinel */
>   };
>   
> *** ../../cvs_01_04_00_orig/dist/src/Lib/test/test_types.py	Wed Feb 23 23:23:17 2000
> --- src/Lib/test/test_types.py	Mon Apr  3 10:41:53 2000
> ***************
> *** 242,247 ****
> --- 242,250 ----
>   d.update({2:20})
>   d.update({1:1, 2:2, 3:3})
>   if d != {1:1, 2:2, 3:3}: raise TestFailed, 'dict update'
> + d.supplement({1:"not", 2:"neither", 4:4})
> + if d != {1:1, 2:2, 3:3, 4:4}: raise TestFailed, 'dict supplement'
> + del d[4]
>   if d.copy() != {1:1, 2:2, 3:3}: raise TestFailed, 'dict copy'
>   if {}.copy() != {}: raise TestFailed, 'empty dict copy'
>   # dict.get()
> *** ../../cvs_01_04_00_orig/dist/src/Lib/UserDict.py	Wed Feb  2 16:10:14 2000
> --- src/Lib/UserDict.py	Mon Apr  3 10:45:17 2000
> ***************
> *** 32,36 ****
> --- 32,45 ----
>           else:
>               for k, v in dict.items():
>                   self.data[k] = v
> +     def supplement(self, dict):
> +         if isinstance(dict, UserDict):
> +             self.data.supplement(dict.data)
> +         elif isinstance(dict, type(self.data)):
> +             self.data.supplement(dict)
> +         else:
> +             for k, v in dict.items():
> + 		if not self.data.has_key(k):
> + 		    self.data[k] = v
>       def get(self, key, failobj=None):
>           return self.data.get(key, failobj)
> *** ../../cvs_01_04_00_orig/dist/src/Lib/test/test_userdict.py	Fri Mar 26 16:32:02 1999
> --- src/Lib/test/test_userdict.py	Mon Apr  3 10:50:29 2000
> ***************
> *** 93,101 ****
> --- 93,109 ----
>   t.update(u2)
>   assert t == u2
>   
> + # Test supplement
> + 
> + t = UserDict(d1)
> + t.supplement(u2)
> + assert t == u2
> + 
>   # Test get
>   
>   for i in u2.keys():
>       assert u2.get(i) == u2[i]
>       assert u1.get(i) == d1.get(i)
>       assert u0.get(i) == d0.get(i)
> + 
> + # TODO: Add a test using dir({}) to test for unimplemented methods
> 
> _______________________________________________
> Patches mailing list
> Patches@python.org
> http://www.python.org/mailman/listinfo/patches
> 

-- 
Greg Stein, http://www.lyra.org/



From mal@lemburg.com  Mon Apr  3 13:59:18 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Mon, 03 Apr 2000 14:59:18 +0200
Subject: [Patches] Unicode Patch Set 2000-04-03
Message-ID: <38E895A5.98FC0A4B@lemburg.com>

This is a multi-part message in MIME format.
--------------F97906822AA778CDBFA10B52
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

The attached patch set adds the following new features:

New Unicode support for int(), float(), complex() and long().

- new APIs PyInt_FromUnicode() and PyLong_FromUnicode()
- added support for Unicode to PyFloat_FromString()
- new encoding API PyUnicode_EncodeDecimal() which converts
  Unicode to a decimal char* string (used in the above new
  APIs)
- shortcuts for calls like int(<int object>) and float(<float obj>)
- tests for all of the above

Better testing support for the standard codecs.

Misc minor enhancements, such as an alias dbcs for the mbcs codec.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------F97906822AA778CDBFA10B52
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-03.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-03.patch"

diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Include/Python.h Python+Unicode/Include/Python.h
--- CVS-Python/Include/Python.h	Sat Mar 11 10:52:34 2000
+++ Python+Unicode/Include/Python.h	Mon Apr  3 13:18:05 2000
@@ -72,6 +72,7 @@
 
 #include "pydebug.h"
 
+#include "unicodeobject.h"
 #include "intobject.h"
 #include "longobject.h"
 #include "floatobject.h"
@@ -92,7 +93,6 @@
 #include "cobject.h"
 #include "traceback.h"
 #include "sliceobject.h"
-#include "unicodeobject.h"
 
 #include "codecs.h"
 #include "pyerrors.h"
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Include/intobject.h Python+Unicode/Include/intobject.h
--- CVS-Python/Include/intobject.h	Tue Mar  7 18:58:16 2000
+++ Python+Unicode/Include/intobject.h	Mon Apr  3 12:54:35 2000
@@ -60,6 +60,7 @@
 #define PyInt_Check(op) ((op)->ob_type == &PyInt_Type)
 
 extern DL_IMPORT(PyObject *) PyInt_FromString Py_PROTO((char*, char**, int));
+extern DL_IMPORT(PyObject *) PyInt_FromUnicode Py_PROTO((Py_UNICODE*, int, int));
 extern DL_IMPORT(PyObject *) PyInt_FromLong Py_PROTO((long));
 extern DL_IMPORT(long) PyInt_AsLong Py_PROTO((PyObject *));
 extern DL_IMPORT(long) PyInt_GetMax Py_PROTO((void));
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Include/longobject.h Python+Unicode/Include/longobject.h
--- CVS-Python/Include/longobject.h	Thu Jan 14 11:24:51 1999
+++ Python+Unicode/Include/longobject.h	Mon Apr  3 13:00:50 2000
@@ -82,6 +82,7 @@
 #endif /* HAVE_LONG_LONG */
 
 DL_IMPORT(PyObject *) PyLong_FromString Py_PROTO((char *, char **, int));
+DL_IMPORT(PyObject *) PyLong_FromUnicode Py_PROTO((Py_UNICODE*, int, int));
 
 #ifdef __cplusplus
 }
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Include/unicodeobject.h Python+Unicode/Include/unicodeobject.h
--- CVS-Python/Include/unicodeobject.h	Wed Mar 29 11:50:05 2000
+++ Python+Unicode/Include/unicodeobject.h	Mon Apr  3 14:17:10 2000
@@ -358,7 +358,7 @@
 
 /* --- UTF-16 Codecs ------------------------------------------------------ */
 
-/* Decodes length bytes from a UTF-16 encoded buffer string and return
+/* Decodes length bytes from a UTF-16 encoded buffer string and returns
    the corresponding Unicode object.
 
    errors (if non-NULL) defines the error handling. It defaults
@@ -397,7 +397,7 @@
     );
 
 /* Returns a Python string object holding the UTF-16 encoded value of
-   the Unicode data in s.
+   the Unicode data.
 
    If byteorder is not 0, output is written according to the following
    byte order:
@@ -586,6 +586,37 @@
     );
 
 #endif /* MS_WIN32 */
+
+/* --- Decimal Encoder ---------------------------------------------------- */
+
+/* Takes a Unicode string holding a decimal value and writes it into
+   an output buffer using standard ASCII digit codes.
+
+   The output buffer has to provide at least length+1 bytes of storage
+   area. The output string is 0-terminated.
+
+   The encoder converts whitespace to ' ', decimal characters to their
+   corresponding ASCII digit and all other Latin-1 characters as-is.
+   Characters outside the Latin-1 range (Unicode ordinals 0-256) are
+   treated as errors.
+
+   Error handling is defined by the errors argument:
+
+      NULL or "strict": raise a ValueError
+      "ignore": ignore the wrong characters (these are not copied to the
+                output buffer)
+      "replace": replaces illegal characters with '?'
+
+   Returns 0 on success, -1 on failure.
+
+*/
+
+extern DL_IMPORT(int) PyUnicode_EncodeDecimal(
+    Py_UNICODE *s,		/* Unicode buffer */
+    int length,			/* Number of Py_UNICODE chars to encode */
+    char *output,		/* Output buffer; must have size >= length */
+    const char *errors		/* error handling */
+    );
 
 /* --- Methods & Slots ----------------------------------------------------
 
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/encodings/__init__.py Python+Unicode/Lib/encodings/__init__.py
--- CVS-Python/Lib/encodings/__init__.py	Thu Mar 23 23:58:42 2000
+++ Python+Unicode/Lib/encodings/__init__.py	Mon Apr  3 13:41:40 2000
@@ -4,8 +4,8 @@
     directory.
 
     Codec modules must have names corresponding to standard lower-case
-    encoding names. Hyphens are automatically converted to
-    underscores, e.g. 'utf-8' is looked up as module utf_8.
+    encoding names with hyphens mapped to underscores, e.g. 'utf-8' is
+    implemented by the module 'utf_8.py'.
 
     Each codec module must export the following interface:
 
@@ -40,7 +40,7 @@
         return entry
 
     # Import the module
-    modname = string.replace(encoding,'-','_')
+    modname = string.replace(encoding, '-', '_')
     modname = aliases.aliases.get(modname,modname)
     try:
         mod = __import__(modname,globals(),locals(),'*')
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/encodings/aliases.py Python+Unicode/Lib/encodings/aliases.py
--- CVS-Python/Lib/encodings/aliases.py	Mon Apr  3 13:45:48 2000
+++ Python+Unicode/Lib/encodings/aliases.py	Sat Apr  1 20:14:36 2000
@@ -54,4 +54,7 @@
     'macroman': 'mac_roman',
     'macturkish': 'mac_turkish',
 
+    # MBCS
+    'dbcs': 'mbcs',
+
 }
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/test/output/test_unicode Python+Unicode/Lib/test/output/test_unicode
--- CVS-Python/Lib/test/output/test_unicode	Wed Mar 29 11:50:15 2000
+++ Python+Unicode/Lib/test/output/test_unicode	Mon Apr  3 14:46:51 2000
@@ -3,3 +3,4 @@
 Testing Unicode contains method... done.
 Testing Unicode formatting strings... done.
 Testing builtin codecs... done.
+Testing standard mapping codecs... 0-127... 128-255... done.
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/test/test_b1.py Python+Unicode/Lib/test/test_b1.py
--- CVS-Python/Lib/test/test_b1.py	Fri Mar 26 15:58:20 1999
+++ Python+Unicode/Lib/test/test_b1.py	Mon Apr  3 13:35:19 2000
@@ -95,6 +95,7 @@
 if complex(0j, 3.14) <> 3.14j: raise TestFailed, 'complex(0j, 3.14)'
 if complex(0.0, 3.14) <> 3.14j: raise TestFailed, 'complex(0.0, 3.14)'
 if complex("  3.14+J  ") <> 3.14+1j:  raise TestFailed, 'complex("  3.14+J  )"'
+if complex(u"  3.14+J  ") <> 3.14+1j:  raise TestFailed, 'complex(u"  3.14+J  )"'
 class Z:
     def __complex__(self): return 3.14j
 z = Z()
@@ -208,6 +209,9 @@
 if float(314) <> 314.0: raise TestFailed, 'float(314)'
 if float(314L) <> 314.0: raise TestFailed, 'float(314L)'
 if float("  3.14  ") <> 3.14:  raise TestFailed, 'float("  3.14  ")'
+if float(u"  3.14  ") <> 3.14:  raise TestFailed, 'float(u"  3.14  ")'
+if float(u"  \u0663.\u0661\u0664  ") <> 3.14:
+    raise TestFailed, 'float(u"  \u0663.\u0661\u0664  ")'
 
 print 'getattr'
 import sys
@@ -270,6 +274,19 @@
         ('', ValueError),
         (' ', ValueError),
         ('  \t\t  ', ValueError),
+        (u'0', 0),
+        (u'1', 1),
+        (u'9', 9),
+        (u'10', 10),
+        (u'99', 99),
+        (u'100', 100),
+        (u'314', 314),
+        (u' 314', 314),
+        (u'\u0663\u0661\u0664 ', 314),
+        (u'  \t\t  314  \t\t  ', 314),
+        (u'', ValueError),
+        (u' ', ValueError),
+        (u'  \t\t  ', ValueError),
 ]
 for s, v in L:
     for sign in "", "+", "-":
@@ -353,6 +370,8 @@
 LL = [
         ('1' + '0'*20, 10L**20),
         ('1' + '0'*100, 10L**100),
+        (u'1' + u'0'*20, 10L**20),
+        (u'1' + u'0'*100, 10L**100),
 ]
 for s, v in L + LL:
     for sign in "", "+", "-":
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/test/test_unicode.py Python+Unicode/Lib/test/test_unicode.py
--- CVS-Python/Lib/test/test_unicode.py	Wed Mar 29 11:50:14 2000
+++ Python+Unicode/Lib/test/test_unicode.py	Sat Apr  1 00:05:35 2000
@@ -270,11 +270,88 @@
     assert unicode(u.encode(encoding),encoding) == u
 
 u = u''.join(map(unichr, range(256)))
-for encoding in ('latin-1',):
+for encoding in (
+    'latin-1',
+    ):
+    try:
     assert unicode(u.encode(encoding),encoding) == u
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
 
 u = u''.join(map(unichr, range(128)))
-for encoding in ('ascii',):
+for encoding in (
+    'ascii',
+    ):
+    try:
     assert unicode(u.encode(encoding),encoding) == u
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
+
+print 'done.'
+
+print 'Testing standard mapping codecs...',
+
+print '0-127...',
+s = ''.join(map(chr, range(128)))
+for encoding in (
+    'cp037', 'cp1026',
+    'cp437', 'cp500', 'cp737', 'cp775', 'cp850',
+    'cp852', 'cp855', 'cp860', 'cp861', 'cp862',
+    'cp863', 'cp865', 'cp866', 
+    'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
+    'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
+    'mac_cyrillic', 'mac_latin2',
+
+    'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
+    'cp1256', 'cp1257', 'cp1258',
+    'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
+
+    'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
+    'cp1006', 'cp875', 'iso8859_8',
+    
+    ### These have undefined mappings:
+    #'cp424',
+    
+    ):
+    try:
+        assert unicode(s,encoding).encode(encoding) == s
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
+
+print '128-255...',
+s = ''.join(map(chr, range(128,256)))
+for encoding in (
+    'cp037', 'cp1026',
+    'cp437', 'cp500', 'cp737', 'cp775', 'cp850',
+    'cp852', 'cp855', 'cp860', 'cp861', 'cp862',
+    'cp863', 'cp865', 'cp866', 
+    'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
+    'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
+    'mac_cyrillic', 'mac_latin2',
+    
+    ### These have undefined mappings:
+    #'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
+    #'cp1256', 'cp1257', 'cp1258',
+    #'cp424', 'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
+    #'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
+    
+    ### These fail the round-trip:
+    #'cp1006', 'cp875', 'iso8859_8',
+    
+    ):
+    try:
+        assert unicode(s,encoding).encode(encoding) == s
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
 
 print 'done.'
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Misc/unicode.txt Python+Unicode/Misc/unicode.txt
--- CVS-Python/Misc/unicode.txt	Wed Mar 29 11:50:15 2000
+++ Python+Unicode/Misc/unicode.txt	Mon Apr  3 14:46:51 2000
@@ -1,5 +1,5 @@
 =============================================================================
- Python Unicode Integration                            Proposal Version: 1.2
+ Python Unicode Integration                            Proposal Version: 1.3
 -----------------------------------------------------------------------------
 
 
@@ -203,8 +203,9 @@
   codecs.register(search_function)
 
 Search functions are expected to take one argument, the encoding name
-in all lower case letters, and return a tuple of functions (encoder,
-decoder, stream_reader, stream_writer) taking the following arguments:
+in all lower case letters and with hyphens and spaces converted to
+underscores, and return a tuple of functions (encoder, decoder,
+stream_reader, stream_writer) taking the following arguments:
 
   encoder and decoder:
 	These must be functions or methods which have the same
Only in CVS-Python/Objects: .#stringobject.c.2.59
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/abstract.c Python+Unicode/Objects/abstract.c
--- CVS-Python/Objects/abstract.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/abstract.c	Mon Apr  3 14:40:06 2000
@@ -736,8 +736,16 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyInt_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (PyString_Check(o))
 		return PyInt_FromString(PyString_AS_STRING(o), NULL, 10);
+	if (PyUnicode_Check(o))
+		return PyInt_FromUnicode(PyUnicode_AS_UNICODE(o),
+					 PyUnicode_GET_SIZE(o),
+					 10);
 	m = o->ob_type->tp_as_number;
 	if (m && m->nb_int)
 		return m->nb_int(o);
@@ -814,6 +822,10 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyLong_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (PyString_Check(o))
 		/* need to do extra error checking that PyLong_FromString() 
 		 * doesn't do.  In particular long('9.5') must raise an
@@ -821,6 +833,11 @@
 		 */
 		return long_from_string(PyString_AS_STRING(o),
 					PyString_GET_SIZE(o));
+	if (PyUnicode_Check(o))
+		/* The above check is done in PyLong_FromUnicode(). */
+		return PyLong_FromUnicode(PyUnicode_AS_UNICODE(o),
+					  PyUnicode_GET_SIZE(o),
+					  10);
 	m = o->ob_type->tp_as_number;
 	if (m && m->nb_long)
 		return m->nb_long(o);
@@ -838,6 +855,10 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyFloat_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (!PyString_Check(o)) {
 		m = o->ob_type->tp_as_number;
 		if (m && m->nb_float)
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/floatobject.c Python+Unicode/Objects/floatobject.c
--- CVS-Python/Objects/floatobject.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/floatobject.c	Mon Apr  3 14:42:03 2000
@@ -163,6 +163,22 @@
 	if (PyString_Check(v)) {
 		s = PyString_AS_STRING(v);
 		len = PyString_GET_SIZE(v);
+	}
+	else if (PyUnicode_Check(v)) {
+		char s_buffer[256];
+
+		if (PyUnicode_GET_SIZE(v) >= sizeof(s_buffer)) {
+			PyErr_SetString(PyExc_ValueError,
+				 "float() literal too large to convert");
+			return NULL;
+		}
+		if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v), 
+					    PyUnicode_GET_SIZE(v),
+					    s_buffer, 
+					    NULL))
+			return NULL;
+		s = s_buffer;
+		len = strlen(s);
 	}
 	else if (PyObject_AsCharBuffer(v, &s, &len)) {
 		PyErr_SetString(PyExc_TypeError,
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/intobject.c Python+Unicode/Objects/intobject.c
--- CVS-Python/Objects/intobject.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/intobject.c	Mon Apr  3 14:42:35 2000
@@ -259,6 +259,24 @@
 	if (pend)
 		*pend = end;
 	return PyInt_FromLong(x);
+}
+
+PyObject *
+PyInt_FromUnicode(s, length, base)
+	Py_UNICODE *s;
+	int length;
+	int base;
+{
+	char buffer[256];
+	
+	if (length >= sizeof(buffer)) {
+		PyErr_SetString(PyExc_ValueError,
+				"int() literal too large to convert");
+		return NULL;
+	}
+	if (PyUnicode_EncodeDecimal(s, length, buffer, NULL))
+		return NULL;
+	return PyInt_FromString(buffer, NULL, base);
 }
 
 /* Methods */
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/longobject.c Python+Unicode/Objects/longobject.c
--- CVS-Python/Objects/longobject.c	Wed Jan 19 18:01:30 2000
+++ Python+Unicode/Objects/longobject.c	Mon Apr  3 14:43:08 2000
@@ -783,6 +783,57 @@
 	if (pend)
 		*pend = str;
 	return (PyObject *) z;
+}
+
+PyObject *
+PyLong_FromUnicode(u, length, base)
+	Py_UNICODE *u;
+	int length;
+	int base;
+{
+	char s_buffer[256];
+	char *s;
+	const char *start;
+	char *end;
+	PyObject *x;
+	char buffer[256]; /* For errors */
+
+        /* Convert to decimal string */
+	if (length >= sizeof(s_buffer)) {
+		PyErr_SetString(PyExc_ValueError,
+				"long() literal too large to convert");
+		return NULL;
+	}
+	if (PyUnicode_EncodeDecimal(u, length, s_buffer, NULL))
+		return NULL;
+	s = s_buffer;
+
+	/* Convert to long object and apply the same error handling
+	   as is done for PyInt_FromString(). */
+	start = s;
+	while (*s && isspace(Py_CHARMASK(*s)))
+		s++;
+	x = PyLong_FromString((char*)s, &end, base);
+	if (x == NULL) {
+		if (PyErr_ExceptionMatches(PyExc_ValueError))
+			goto bad;
+		return NULL;
+	}
+	while (*end && isspace(Py_CHARMASK(*end)))
+		end++;
+	if (*end != '\0') {
+  bad:
+		sprintf(buffer, "invalid literal for long(): %.200s", s);
+		PyErr_SetString(PyExc_ValueError, buffer);
+		Py_XDECREF(x);
+		return NULL;
+	}
+	else if (end != start + length) {
+		PyErr_SetString(PyExc_ValueError,
+				"null byte in argument for long()");
+		return NULL;
+	}
+	return x;
 }
 
 static PyLongObject *x_divrem
Only in CVS-Python/Objects: stringobject.c.orig
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/unicodeobject.c Python+Unicode/Objects/unicodeobject.c
--- CVS-Python/Objects/unicodeobject.c	Mon Apr  3 13:46:02 2000
+++ Python+Unicode/Objects/unicodeobject.c	Mon Apr  3 14:43:52 2000
@@ -1923,6 +1923,60 @@
     return NULL;
 }
     
+/* --- Decimal Encoder ---------------------------------------------------- */
+
+int PyUnicode_EncodeDecimal(Py_UNICODE *s,
+			    int length,
+			    char *output,
+			    const char *errors)
+{
+    Py_UNICODE *p, *end;
+
+    if (output == NULL) {
+	PyErr_BadArgument();
+	return -1;
+    }
+
+    p = s;
+    end = s + length;
+    while (p < end) {
+	register Py_UNICODE ch = *p++;
+	int decimal;
+	
+	if (Py_UNICODE_ISSPACE(ch)) {
+	    *output++ = ' ';
+	    continue;
+	}
+	decimal = Py_UNICODE_TODECIMAL(ch);
+	if (decimal >= 0) {
+	    *output++ = '0' + decimal;
+	    continue;
+	}
+	if (ch < 256) {
+	    *output++ = ch;
+	    continue;
+	}
+	/* All other characters are considered invalid */
+	if (errors == NULL || strcmp(errors, "strict") == 0) {
+	    PyErr_SetString(PyExc_ValueError,
+			    "invalid numeric string");
+	    goto onError;
+	}
+	else if (strcmp(errors, "ignore") == 0)
+	    continue;
+	else if (strcmp(errors, "replace") == 0) {
+	    *output++ = '?';
+	    continue;
+	}
+    }
+    /* 0-terminate the output string */
+    *output++ = '\0';
+    return 0;
+
+ onError:
+    return -1;
+}
+
 /* --- Helpers ------------------------------------------------------------ */
 
 static 
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Python/bltinmodule.c Python+Unicode/Python/bltinmodule.c
--- CVS-Python/Python/bltinmodule.c	Sat Mar 11 10:55:20 2000
+++ Python+Unicode/Python/bltinmodule.c	Mon Apr  3 14:41:16 2000
@@ -449,17 +449,44 @@
 	PyObject *v;
 {
 	extern double strtod Py_PROTO((const char *, char **));
-	char *s, *start, *end;
+	const char *s, *start;
+	char *end;
 	double x=0.0, y=0.0, z;
 	int got_re=0, got_im=0, done=0;
 	int digit_or_dot;
 	int sw_error=0;
 	int sign;
 	char buffer[256]; /* For errors */
+	int len;
 
-	start = s = PyString_AS_STRING(v);
+	if (PyString_Check(v)) {
+		s = PyString_AS_STRING(v);
+		len = PyString_GET_SIZE(v);
+	}
+	else if (PyUnicode_Check(v)) {
+		char s_buffer[256];
+
+		if (PyUnicode_GET_SIZE(v) >= sizeof(s_buffer)) {
+			PyErr_SetString(PyExc_ValueError,
+				 "complex() literal too large to convert");
+			return NULL;
+		}
+		if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v), 
+					    PyUnicode_GET_SIZE(v),
+					    s_buffer, 
+					    NULL))
+			return NULL;
+		s = s_buffer;
+		len = strlen(s);
+	}
+	else if (PyObject_AsCharBuffer(v, &s, &len)) {
+		PyErr_SetString(PyExc_TypeError,
+				"complex() needs a string first argument");
+		return NULL;
+	}
 
 	/* position on first nonblank */
+	start = s;
 	while (*s && isspace(Py_CHARMASK(*s)))
 		s++;
 	if (s[0] == '\0') {
@@ -475,7 +502,7 @@
 		switch (*s) {
 
 		case '\0':
-			if (s-start != PyString_GET_SIZE(v)) {
+			if (s-start != len) {
 				PyErr_SetString(
 					PyExc_ValueError,
 					"null byte in argument for complex()");
@@ -584,7 +611,7 @@
 	i = NULL;
 	if (!PyArg_ParseTuple(args, "O|O:complex", &r, &i))
 		return NULL;
-	if (PyString_Check(r))
+	if (PyString_Check(r) || PyUnicode_Check(r))
 		return complex_from_string(r);
 	if ((nbr = r->ob_type->tp_as_number) == NULL ||
 	    nbr->nb_float == NULL ||
@@ -1289,12 +1316,17 @@
 		return NULL;
 	if (base == -909)
 		return PyNumber_Int(v);
-	else if (!PyString_Check(v)) {
+	else if (PyString_Check(v))
+		return PyInt_FromString(PyString_AS_STRING(v), NULL, base);
+	else if (PyUnicode_Check(v))
+		return PyInt_FromUnicode(PyUnicode_AS_UNICODE(v),
+					 PyUnicode_GET_SIZE(v),
+					 base);
+	else {
 		PyErr_SetString(PyExc_TypeError,
 				"can't convert non-string with explicit base");
 		return NULL;
 	}
-	return PyInt_FromString(PyString_AS_STRING(v), NULL, base);
 }
 
 static char int_doc[] =
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Python/codecs.c Python+Unicode/Python/codecs.c
--- CVS-Python/Python/codecs.c	Mon Apr  3 13:46:03 2000
+++ Python+Unicode/Python/codecs.c	Sat Apr  1 19:57:52 2000
@@ -84,8 +84,11 @@
     return -1;
 }
 
+/* Convert a string to a normalized Python string: all characters are
+   converted to lower case, spaces are replaced with underscores. */
+
 static
-PyObject *lowercasestring(const char *string)
+PyObject *normalizestring(const char *string)
 {
     register int i;
     int len = strlen(string);
@@ -96,8 +99,14 @@
     if (v == NULL)
 	return NULL;
     p = PyString_AS_STRING(v);
-    for (i = 0; i < len; i++)
-	p[i] = tolower(string[i]);
+    for (i = 0; i < len; i++) {
+        register char ch = string[i];
+        if (ch == ' ')
+            ch = '-';
+        else
+            ch = tolower(ch);
+	p[i] = ch;
+    }
     return v;
 }
 
@@ -132,8 +141,10 @@
 	    goto onError;
     }
 
-    /* Convert the encoding to a lower-cased Python string */
-    v = lowercasestring(encoding);
+    /* Convert the encoding to a normalized Python string: all
+       characters are converted to lower case, spaces and hypens are
+       replaced with underscores. */
+    v = normalizestring(encoding);
     if (v == NULL)
 	goto onError;
     PyString_InternInPlace(&v);

--------------F97906822AA778CDBFA10B52--





From bwarsaw@cnri.reston.va.us  Mon Apr  3 14:01:28 2000
From: bwarsaw@cnri.reston.va.us (Barry A. Warsaw)
Date: Mon, 3 Apr 2000 09:01:28 -0400 (EDT)
Subject: [Patches] New lib module: UserString.py and testcase for it
References: <m12btJl-000CnCC@artcom0.artcom-gmbh.de>
 <14568.5667.841020.109795@seahag.cnri.reston.va.us>
Message-ID: <14568.38440.757591.172014@anthem.cnri.reston.va.us>

>>>>> "Fred" == Fred L Drake, Jr <fdrake@acm.org> writes:

    Fred> Peter, It's in, but I think there's a problem with the
    Fred> anonCVS getting updated from my checkins.  I'll look further
    Fred> into that tomorrow.  ;( Attachments are better for me, but
    Fred> can deal with shar as well.  Just hope I never have to use a
    Fred> non-Unix box for development!  ;) Thanks!

I just re-ran the sync script so those changes have been pushed out to
cvs.python.org

-Barry


From mal@lemburg.com  Mon Apr  3 15:46:28 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Mon, 03 Apr 2000 16:46:28 +0200
Subject: [Patches] Unicode Patch Set 2000-04-03
References: <38E895A5.98FC0A4B@lemburg.com>
Message-ID: <38E8AEC4.CE1E63BB@lemburg.com>

This is a multi-part message in MIME format.
--------------751C647C75D9C69DDEBDDF5E
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

While adding the string.ato?() patches I found that I had
forgotten to fix the long() builtin to also accept Unicode.

The attached patch set now includes long() Unicode support too.
I've also added some test cases which verify this (the original
test cases didn't check the base argument which caused the
tests to pass).

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------751C647C75D9C69DDEBDDF5E
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-03.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-03.patch"

diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Include/Python.h Python+Unicode/Include/Python.h
--- CVS-Python/Include/Python.h	Sat Mar 11 10:52:34 2000
+++ Python+Unicode/Include/Python.h	Mon Apr  3 13:18:05 2000
@@ -72,6 +72,7 @@
 
 #include "pydebug.h"
 
+#include "unicodeobject.h"
 #include "intobject.h"
 #include "longobject.h"
 #include "floatobject.h"
@@ -92,7 +93,6 @@
 #include "cobject.h"
 #include "traceback.h"
 #include "sliceobject.h"
-#include "unicodeobject.h"
 
 #include "codecs.h"
 #include "pyerrors.h"
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Include/intobject.h Python+Unicode/Include/intobject.h
--- CVS-Python/Include/intobject.h	Tue Mar  7 18:58:16 2000
+++ Python+Unicode/Include/intobject.h	Mon Apr  3 12:54:35 2000
@@ -60,6 +60,7 @@
 #define PyInt_Check(op) ((op)->ob_type == &PyInt_Type)
 
 extern DL_IMPORT(PyObject *) PyInt_FromString Py_PROTO((char*, char**, int));
+extern DL_IMPORT(PyObject *) PyInt_FromUnicode Py_PROTO((Py_UNICODE*, int, int));
 extern DL_IMPORT(PyObject *) PyInt_FromLong Py_PROTO((long));
 extern DL_IMPORT(long) PyInt_AsLong Py_PROTO((PyObject *));
 extern DL_IMPORT(long) PyInt_GetMax Py_PROTO((void));
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Include/longobject.h Python+Unicode/Include/longobject.h
--- CVS-Python/Include/longobject.h	Thu Jan 14 11:24:51 1999
+++ Python+Unicode/Include/longobject.h	Mon Apr  3 13:00:50 2000
@@ -82,6 +82,7 @@
 #endif /* HAVE_LONG_LONG */
 
 DL_IMPORT(PyObject *) PyLong_FromString Py_PROTO((char *, char **, int));
+DL_IMPORT(PyObject *) PyLong_FromUnicode Py_PROTO((Py_UNICODE*, int, int));
 
 #ifdef __cplusplus
 }
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Include/unicodeobject.h Python+Unicode/Include/unicodeobject.h
--- CVS-Python/Include/unicodeobject.h	Wed Mar 29 11:50:05 2000
+++ Python+Unicode/Include/unicodeobject.h	Mon Apr  3 14:47:33 2000
@@ -358,7 +358,7 @@
 
 /* --- UTF-16 Codecs ------------------------------------------------------ */
 
-/* Decodes length bytes from a UTF-16 encoded buffer string and return
+/* Decodes length bytes from a UTF-16 encoded buffer string and returns
    the corresponding Unicode object.
 
    errors (if non-NULL) defines the error handling. It defaults
@@ -397,7 +397,7 @@
     );
 
 /* Returns a Python string object holding the UTF-16 encoded value of
-   the Unicode data in s.
+   the Unicode data.
 
    If byteorder is not 0, output is written according to the following
    byte order:
@@ -586,6 +586,37 @@
     );
 
 #endif /* MS_WIN32 */
+
+/* --- Decimal Encoder ---------------------------------------------------- */
+
+/* Takes a Unicode string holding a decimal value and writes it into
+   an output buffer using standard ASCII digit codes.
+
+   The output buffer has to provide at least length+1 bytes of storage
+   area. The output string is 0-terminated.
+
+   The encoder converts whitespace to ' ', decimal characters to their
+   corresponding ASCII digit and all other Latin-1 characters as-is.
+   Characters outside the Latin-1 range (Unicode ordinals 0-256) are
+   treated as errors.
+
+   Error handling is defined by the errors argument:
+
+      NULL or "strict": raise a ValueError
+      "ignore": ignore the wrong characters (these are not copied to the
+		output buffer)
+      "replace": replaces illegal characters with '?'
+
+   Returns 0 on success, -1 on failure.
+
+*/
+
+extern DL_IMPORT(int) PyUnicode_EncodeDecimal(
+    Py_UNICODE *s,		/* Unicode buffer */
+    int length,			/* Number of Py_UNICODE chars to encode */
+    char *output,		/* Output buffer; must have size >= length */
+    const char *errors		/* error handling */
+    );
 
 /* --- Methods & Slots ----------------------------------------------------
 
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/encodings/__init__.py Python+Unicode/Lib/encodings/__init__.py
--- CVS-Python/Lib/encodings/__init__.py	Thu Mar 23 23:58:42 2000
+++ Python+Unicode/Lib/encodings/__init__.py	Mon Apr  3 13:41:40 2000
@@ -4,8 +4,8 @@
     directory.
 
     Codec modules must have names corresponding to standard lower-case
-    encoding names. Hyphens are automatically converted to
-    underscores, e.g. 'utf-8' is looked up as module utf_8.
+    encoding names with hyphens mapped to underscores, e.g. 'utf-8' is
+    implemented by the module 'utf_8.py'.
 
     Each codec module must export the following interface:
 
@@ -40,7 +40,7 @@
         return entry
 
     # Import the module
-    modname = string.replace(encoding,'-','_')
+    modname = string.replace(encoding, '-', '_')
     modname = aliases.aliases.get(modname,modname)
     try:
         mod = __import__(modname,globals(),locals(),'*')
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/encodings/aliases.py Python+Unicode/Lib/encodings/aliases.py
--- CVS-Python/Lib/encodings/aliases.py	Mon Apr  3 13:45:48 2000
+++ Python+Unicode/Lib/encodings/aliases.py	Sat Apr  1 20:14:36 2000
@@ -54,4 +54,7 @@
     'macroman': 'mac_roman',
     'macturkish': 'mac_turkish',
 
+    # MBCS
+    'dbcs': 'mbcs',
+
 }
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/string.py Python+Unicode/Lib/string.py
--- CVS-Python/Lib/string.py	Sat Mar 11 10:52:36 2000
+++ Python+Unicode/Lib/string.py	Mon Apr  3 16:09:58 2000
@@ -196,14 +196,11 @@
     Return the floating point number represented by the string s.
 
     """
-    if type(s) == _StringType:
         return _float(s)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+
 
 # Convert string to integer
-def atoi(*args):
+def atoi(s , base=10):
     """atoi(s [,base]) -> int
 
     Return the integer represented by the string s in the given
@@ -214,23 +211,11 @@
     accepted.
 
     """
-    try:
-        s = args[0]
-    except IndexError:
-        raise TypeError('function requires at least 1 argument: %d given' %
-                        len(args))
-    # Don't catch type error resulting from too many arguments to int().  The
-    # error message isn't compatible but the error type is, and this function
-    # is complicated enough already.
-    if type(s) == _StringType:
-        return _apply(_int, args)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+    return _int(s, base)
 
 
 # Convert string to long integer
-def atol(*args):
+def atol(s, base=10):
     """atol(s [,base]) -> long
 
     Return the long integer represented by the string s in the
@@ -242,19 +227,7 @@
     unless base is 0.
 
     """
-    try:
-        s = args[0]
-    except IndexError:
-        raise TypeError('function requires at least 1 argument: %d given' %
-                        len(args))
-    # Don't catch type error resulting from too many arguments to long().  The
-    # error message isn't compatible but the error type is, and this function
-    # is complicated enough already.
-    if type(s) == _StringType:
-        return _apply(_long, args)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+    return _long(s, base)
 
 
 # Left-justify a string
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/test/output/test_unicode Python+Unicode/Lib/test/output/test_unicode
--- CVS-Python/Lib/test/output/test_unicode	Wed Mar 29 11:50:15 2000
+++ Python+Unicode/Lib/test/output/test_unicode	Mon Apr  3 16:23:26 2000
@@ -3,3 +3,4 @@
 Testing Unicode contains method... done.
 Testing Unicode formatting strings... done.
 Testing builtin codecs... done.
+Testing standard mapping codecs... 0-127... 128-255... done.
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/test/test_b1.py Python+Unicode/Lib/test/test_b1.py
--- CVS-Python/Lib/test/test_b1.py	Fri Mar 26 15:58:20 1999
+++ Python+Unicode/Lib/test/test_b1.py	Mon Apr  3 16:21:34 2000
@@ -95,6 +95,7 @@
 if complex(0j, 3.14) <> 3.14j: raise TestFailed, 'complex(0j, 3.14)'
 if complex(0.0, 3.14) <> 3.14j: raise TestFailed, 'complex(0.0, 3.14)'
 if complex("  3.14+J  ") <> 3.14+1j:  raise TestFailed, 'complex("  3.14+J  )"'
+if complex(u"  3.14+J  ") <> 3.14+1j:  raise TestFailed, 'complex(u"  3.14+J  )"'
 class Z:
     def __complex__(self): return 3.14j
 z = Z()
@@ -208,6 +209,9 @@
 if float(314) <> 314.0: raise TestFailed, 'float(314)'
 if float(314L) <> 314.0: raise TestFailed, 'float(314L)'
 if float("  3.14  ") <> 3.14:  raise TestFailed, 'float("  3.14  ")'
+if float(u"  3.14  ") <> 3.14:  raise TestFailed, 'float(u"  3.14  ")'
+if float(u"  \u0663.\u0661\u0664  ") <> 3.14:
+    raise TestFailed, 'float(u"  \u0663.\u0661\u0664  ")'
 
 print 'getattr'
 import sys
@@ -254,6 +258,9 @@
 if int(-3.9) <> -3: raise TestFailed, 'int(-3.9)'
 if int(3.5) <> 3: raise TestFailed, 'int(3.5)'
 if int(-3.5) <> -3: raise TestFailed, 'int(-3.5)'
+# Different base:
+if int("10",16) <> 16L: raise TestFailed, 'int("10",16)'
+if int(u"10",16) <> 16L: raise TestFailed, 'int(u"10",16)'
 # Test conversion fron strings and various anomalies
 L = [
         ('0', 0),
@@ -270,6 +277,19 @@
         ('', ValueError),
         (' ', ValueError),
         ('  \t\t  ', ValueError),
+        (u'0', 0),
+        (u'1', 1),
+        (u'9', 9),
+        (u'10', 10),
+        (u'99', 99),
+        (u'100', 100),
+        (u'314', 314),
+        (u' 314', 314),
+        (u'\u0663\u0661\u0664 ', 314),
+        (u'  \t\t  314  \t\t  ', 314),
+        (u'', ValueError),
+        (u' ', ValueError),
+        (u'  \t\t  ', ValueError),
 ]
 for s, v in L:
     for sign in "", "+", "-":
@@ -349,10 +369,17 @@
 if long(-3.9) <> -3L: raise TestFailed, 'long(-3.9)'
 if long(3.5) <> 3L: raise TestFailed, 'long(3.5)'
 if long(-3.5) <> -3L: raise TestFailed, 'long(-3.5)'
+if long("-3") <> -3L: raise TestFailed, 'long("-3")'
+if long(u"-3") <> -3L: raise TestFailed, 'long(u"-3")'
+# Different base:
+if long("10",16) <> 16L: raise TestFailed, 'long("10",16)'
+if long(u"10",16) <> 16L: raise TestFailed, 'long(u"10",16)'
 # Check conversions from string (same test set as for int(), and then some)
 LL = [
         ('1' + '0'*20, 10L**20),
         ('1' + '0'*100, 10L**100),
+        (u'1' + u'0'*20, 10L**20),
+        (u'1' + u'0'*100, 10L**100),
 ]
 for s, v in L + LL:
     for sign in "", "+", "-":
@@ -363,11 +390,11 @@
                 vv = -v
             try:
                 if long(ss) != long(vv):
-                    raise TestFailed, "int(%s)" % `ss`
+                    raise TestFailed, "long(%s)" % `ss`
             except v:
                 pass
             except ValueError, e:
-                raise TestFailed, "int(%s) raised ValueError: %s" % (`ss`, e)
+                raise TestFailed, "long(%s) raised ValueError: %s" % (`ss`, e)
 
 print 'map'
 if map(None, 'hello world') <> ['h','e','l','l','o',' ','w','o','r','l','d']:
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Lib/test/test_unicode.py Python+Unicode/Lib/test/test_unicode.py
--- CVS-Python/Lib/test/test_unicode.py	Wed Mar 29 11:50:14 2000
+++ Python+Unicode/Lib/test/test_unicode.py	Sat Apr  1 00:05:35 2000
@@ -270,11 +270,88 @@
     assert unicode(u.encode(encoding),encoding) == u
 
 u = u''.join(map(unichr, range(256)))
-for encoding in ('latin-1',):
+for encoding in (
+    'latin-1',
+    ):
+    try:
     assert unicode(u.encode(encoding),encoding) == u
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
 
 u = u''.join(map(unichr, range(128)))
-for encoding in ('ascii',):
+for encoding in (
+    'ascii',
+    ):
+    try:
     assert unicode(u.encode(encoding),encoding) == u
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
+
+print 'done.'
+
+print 'Testing standard mapping codecs...',
+
+print '0-127...',
+s = ''.join(map(chr, range(128)))
+for encoding in (
+    'cp037', 'cp1026',
+    'cp437', 'cp500', 'cp737', 'cp775', 'cp850',
+    'cp852', 'cp855', 'cp860', 'cp861', 'cp862',
+    'cp863', 'cp865', 'cp866', 
+    'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
+    'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
+    'mac_cyrillic', 'mac_latin2',
+
+    'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
+    'cp1256', 'cp1257', 'cp1258',
+    'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
+
+    'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
+    'cp1006', 'cp875', 'iso8859_8',
+    
+    ### These have undefined mappings:
+    #'cp424',
+    
+    ):
+    try:
+        assert unicode(s,encoding).encode(encoding) == s
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
+
+print '128-255...',
+s = ''.join(map(chr, range(128,256)))
+for encoding in (
+    'cp037', 'cp1026',
+    'cp437', 'cp500', 'cp737', 'cp775', 'cp850',
+    'cp852', 'cp855', 'cp860', 'cp861', 'cp862',
+    'cp863', 'cp865', 'cp866', 
+    'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
+    'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
+    'mac_cyrillic', 'mac_latin2',
+    
+    ### These have undefined mappings:
+    #'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
+    #'cp1256', 'cp1257', 'cp1258',
+    #'cp424', 'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
+    #'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
+    
+    ### These fail the round-trip:
+    #'cp1006', 'cp875', 'iso8859_8',
+    
+    ):
+    try:
+        assert unicode(s,encoding).encode(encoding) == s
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
 
 print 'done.'
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Misc/unicode.txt Python+Unicode/Misc/unicode.txt
--- CVS-Python/Misc/unicode.txt	Wed Mar 29 11:50:15 2000
+++ Python+Unicode/Misc/unicode.txt	Mon Apr  3 16:23:25 2000
@@ -1,5 +1,5 @@
 =============================================================================
- Python Unicode Integration                            Proposal Version: 1.2
+ Python Unicode Integration                            Proposal Version: 1.3
 -----------------------------------------------------------------------------
 
 
@@ -203,8 +203,9 @@
   codecs.register(search_function)
 
 Search functions are expected to take one argument, the encoding name
-in all lower case letters, and return a tuple of functions (encoder,
-decoder, stream_reader, stream_writer) taking the following arguments:
+in all lower case letters and with hyphens and spaces converted to
+underscores, and return a tuple of functions (encoder, decoder,
+stream_reader, stream_writer) taking the following arguments:
 
   encoder and decoder:
 	These must be functions or methods which have the same
Only in CVS-Python/Objects: .#stringobject.c.2.59
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/abstract.c Python+Unicode/Objects/abstract.c
--- CVS-Python/Objects/abstract.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/abstract.c	Mon Apr  3 14:40:06 2000
@@ -736,8 +736,16 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyInt_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (PyString_Check(o))
 		return PyInt_FromString(PyString_AS_STRING(o), NULL, 10);
+	if (PyUnicode_Check(o))
+		return PyInt_FromUnicode(PyUnicode_AS_UNICODE(o),
+					 PyUnicode_GET_SIZE(o),
+					 10);
 	m = o->ob_type->tp_as_number;
 	if (m && m->nb_int)
 		return m->nb_int(o);
@@ -814,6 +822,10 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyLong_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (PyString_Check(o))
 		/* need to do extra error checking that PyLong_FromString() 
 		 * doesn't do.  In particular long('9.5') must raise an
@@ -821,6 +833,11 @@
 		 */
 		return long_from_string(PyString_AS_STRING(o),
 					PyString_GET_SIZE(o));
+	if (PyUnicode_Check(o))
+		/* The above check is done in PyLong_FromUnicode(). */
+		return PyLong_FromUnicode(PyUnicode_AS_UNICODE(o),
+					  PyUnicode_GET_SIZE(o),
+					  10);
 	m = o->ob_type->tp_as_number;
 	if (m && m->nb_long)
 		return m->nb_long(o);
@@ -838,6 +855,10 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyFloat_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (!PyString_Check(o)) {
 		m = o->ob_type->tp_as_number;
 		if (m && m->nb_float)
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/floatobject.c Python+Unicode/Objects/floatobject.c
--- CVS-Python/Objects/floatobject.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/floatobject.c	Mon Apr  3 14:42:03 2000
@@ -163,6 +163,22 @@
 	if (PyString_Check(v)) {
 		s = PyString_AS_STRING(v);
 		len = PyString_GET_SIZE(v);
+	}
+	else if (PyUnicode_Check(v)) {
+		char s_buffer[256];
+
+		if (PyUnicode_GET_SIZE(v) >= sizeof(s_buffer)) {
+			PyErr_SetString(PyExc_ValueError,
+				 "float() literal too large to convert");
+			return NULL;
+		}
+		if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v), 
+					    PyUnicode_GET_SIZE(v),
+					    s_buffer, 
+					    NULL))
+			return NULL;
+		s = s_buffer;
+		len = strlen(s);
 	}
 	else if (PyObject_AsCharBuffer(v, &s, &len)) {
 		PyErr_SetString(PyExc_TypeError,
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/intobject.c Python+Unicode/Objects/intobject.c
--- CVS-Python/Objects/intobject.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/intobject.c	Mon Apr  3 14:42:35 2000
@@ -259,6 +259,24 @@
 	if (pend)
 		*pend = end;
 	return PyInt_FromLong(x);
+}
+
+PyObject *
+PyInt_FromUnicode(s, length, base)
+	Py_UNICODE *s;
+	int length;
+	int base;
+{
+	char buffer[256];
+	
+	if (length >= sizeof(buffer)) {
+		PyErr_SetString(PyExc_ValueError,
+				"int() literal too large to convert");
+		return NULL;
+	}
+	if (PyUnicode_EncodeDecimal(s, length, buffer, NULL))
+		return NULL;
+	return PyInt_FromString(buffer, NULL, base);
 }
 
 /* Methods */
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/longobject.c Python+Unicode/Objects/longobject.c
--- CVS-Python/Objects/longobject.c	Wed Jan 19 18:01:30 2000
+++ Python+Unicode/Objects/longobject.c	Mon Apr  3 14:43:08 2000
@@ -783,6 +783,57 @@
 	if (pend)
 		*pend = str;
 	return (PyObject *) z;
+}
+
+PyObject *
+PyLong_FromUnicode(u, length, base)
+	Py_UNICODE *u;
+	int length;
+	int base;
+{
+	char s_buffer[256];
+	char *s;
+	const char *start;
+	char *end;
+	PyObject *x;
+	char buffer[256]; /* For errors */
+
+        /* Convert to decimal string */
+	if (length >= sizeof(s_buffer)) {
+		PyErr_SetString(PyExc_ValueError,
+				"long() literal too large to convert");
+		return NULL;
+	}
+	if (PyUnicode_EncodeDecimal(u, length, s_buffer, NULL))
+		return NULL;
+	s = s_buffer;
+
+	/* Convert to long object and apply the same error handling
+	   as is done for PyInt_FromString(). */
+	start = s;
+	while (*s && isspace(Py_CHARMASK(*s)))
+		s++;
+	x = PyLong_FromString((char*)s, &end, base);
+	if (x == NULL) {
+		if (PyErr_ExceptionMatches(PyExc_ValueError))
+			goto bad;
+		return NULL;
+	}
+	while (*end && isspace(Py_CHARMASK(*end)))
+		end++;
+	if (*end != '\0') {
+  bad:
+		sprintf(buffer, "invalid literal for long(): %.200s", s);
+		PyErr_SetString(PyExc_ValueError, buffer);
+		Py_XDECREF(x);
+		return NULL;
+	}
+	else if (end != start + length) {
+		PyErr_SetString(PyExc_ValueError,
+				"null byte in argument for long()");
+		return NULL;
+	}
+	return x;
 }
 
 static PyLongObject *x_divrem
Only in CVS-Python/Objects: stringobject.c.orig
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Objects/unicodeobject.c Python+Unicode/Objects/unicodeobject.c
--- CVS-Python/Objects/unicodeobject.c	Mon Apr  3 13:46:02 2000
+++ Python+Unicode/Objects/unicodeobject.c	Mon Apr  3 14:43:52 2000
@@ -1923,6 +1923,60 @@
     return NULL;
 }
     
+/* --- Decimal Encoder ---------------------------------------------------- */
+
+int PyUnicode_EncodeDecimal(Py_UNICODE *s,
+			    int length,
+			    char *output,
+			    const char *errors)
+{
+    Py_UNICODE *p, *end;
+
+    if (output == NULL) {
+	PyErr_BadArgument();
+	return -1;
+    }
+
+    p = s;
+    end = s + length;
+    while (p < end) {
+	register Py_UNICODE ch = *p++;
+	int decimal;
+	
+	if (Py_UNICODE_ISSPACE(ch)) {
+	    *output++ = ' ';
+	    continue;
+	}
+	decimal = Py_UNICODE_TODECIMAL(ch);
+	if (decimal >= 0) {
+	    *output++ = '0' + decimal;
+	    continue;
+	}
+	if (ch < 256) {
+	    *output++ = ch;
+	    continue;
+	}
+	/* All other characters are considered invalid */
+	if (errors == NULL || strcmp(errors, "strict") == 0) {
+	    PyErr_SetString(PyExc_ValueError,
+			    "invalid numeric string");
+	    goto onError;
+	}
+	else if (strcmp(errors, "ignore") == 0)
+	    continue;
+	else if (strcmp(errors, "replace") == 0) {
+	    *output++ = '?';
+	    continue;
+	}
+    }
+    /* 0-terminate the output string */
+    *output++ = '\0';
+    return 0;
+
+ onError:
+    return -1;
+}
+
 /* --- Helpers ------------------------------------------------------------ */
 
 static 
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Python/bltinmodule.c Python+Unicode/Python/bltinmodule.c
--- CVS-Python/Python/bltinmodule.c	Sat Mar 11 10:55:20 2000
+++ Python+Unicode/Python/bltinmodule.c	Mon Apr  3 16:14:02 2000
@@ -449,17 +449,44 @@
 	PyObject *v;
 {
 	extern double strtod Py_PROTO((const char *, char **));
-	char *s, *start, *end;
+	const char *s, *start;
+	char *end;
 	double x=0.0, y=0.0, z;
 	int got_re=0, got_im=0, done=0;
 	int digit_or_dot;
 	int sw_error=0;
 	int sign;
 	char buffer[256]; /* For errors */
+	int len;
 
-	start = s = PyString_AS_STRING(v);
+	if (PyString_Check(v)) {
+		s = PyString_AS_STRING(v);
+		len = PyString_GET_SIZE(v);
+	}
+	else if (PyUnicode_Check(v)) {
+		char s_buffer[256];
+
+		if (PyUnicode_GET_SIZE(v) >= sizeof(s_buffer)) {
+			PyErr_SetString(PyExc_ValueError,
+				 "complex() literal too large to convert");
+			return NULL;
+		}
+		if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v), 
+					    PyUnicode_GET_SIZE(v),
+					    s_buffer, 
+					    NULL))
+			return NULL;
+		s = s_buffer;
+		len = strlen(s);
+	}
+	else if (PyObject_AsCharBuffer(v, &s, &len)) {
+		PyErr_SetString(PyExc_TypeError,
+				"complex() needs a string first argument");
+		return NULL;
+	}
 
 	/* position on first nonblank */
+	start = s;
 	while (*s && isspace(Py_CHARMASK(*s)))
 		s++;
 	if (s[0] == '\0') {
@@ -475,7 +502,7 @@
 		switch (*s) {
 
 		case '\0':
-			if (s-start != PyString_GET_SIZE(v)) {
+			if (s-start != len) {
 				PyErr_SetString(
 					PyExc_ValueError,
 					"null byte in argument for complex()");
@@ -584,7 +611,7 @@
 	i = NULL;
 	if (!PyArg_ParseTuple(args, "O|O:complex", &r, &i))
 		return NULL;
-	if (PyString_Check(r))
+	if (PyString_Check(r) || PyUnicode_Check(r))
 		return complex_from_string(r);
 	if ((nbr = r->ob_type->tp_as_number) == NULL ||
 	    nbr->nb_float == NULL ||
@@ -1289,12 +1316,17 @@
 		return NULL;
 	if (base == -909)
 		return PyNumber_Int(v);
-	else if (!PyString_Check(v)) {
+	else if (PyString_Check(v))
+		return PyInt_FromString(PyString_AS_STRING(v), NULL, base);
+	else if (PyUnicode_Check(v))
+		return PyInt_FromUnicode(PyUnicode_AS_UNICODE(v),
+					 PyUnicode_GET_SIZE(v),
+					 base);
+	else {
 		PyErr_SetString(PyExc_TypeError,
 				"can't convert non-string with explicit base");
 		return NULL;
 	}
-	return PyInt_FromString(PyString_AS_STRING(v), NULL, base);
 }
 
 static char int_doc[] =
@@ -1319,12 +1351,17 @@
 		return NULL;
 	if (base == -909)
 		return PyNumber_Long(v);
-	else if (!PyString_Check(v)) {
+	else if (PyString_Check(v))
+		return PyLong_FromString(PyString_AS_STRING(v), NULL, base);
+	else if (PyUnicode_Check(v))
+		return PyLong_FromUnicode(PyUnicode_AS_UNICODE(v),
+					  PyUnicode_GET_SIZE(v),
+					  base);
+	else {
 		PyErr_SetString(PyExc_TypeError,
 				"can't convert non-string with explicit base");
 		return NULL;
 	}
-	return PyLong_FromString(PyString_AS_STRING(v), NULL, base);
 }
 
 static char long_doc[] =
diff -u -b -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x Demo -x CVS -x Doc CVS-Python/Python/codecs.c Python+Unicode/Python/codecs.c
--- CVS-Python/Python/codecs.c	Mon Apr  3 13:46:03 2000
+++ Python+Unicode/Python/codecs.c	Sat Apr  1 19:57:52 2000
@@ -84,8 +84,11 @@
     return -1;
 }
 
+/* Convert a string to a normalized Python string: all characters are
+   converted to lower case, spaces are replaced with underscores. */
+
 static
-PyObject *lowercasestring(const char *string)
+PyObject *normalizestring(const char *string)
 {
     register int i;
     int len = strlen(string);
@@ -96,8 +99,14 @@
     if (v == NULL)
 	return NULL;
     p = PyString_AS_STRING(v);
-    for (i = 0; i < len; i++)
-	p[i] = tolower(string[i]);
+    for (i = 0; i < len; i++) {
+        register char ch = string[i];
+        if (ch == ' ')
+            ch = '-';
+        else
+            ch = tolower(ch);
+	p[i] = ch;
+    }
     return v;
 }
 
@@ -132,8 +141,10 @@
 	    goto onError;
     }
 
-    /* Convert the encoding to a lower-cased Python string */
-    v = lowercasestring(encoding);
+    /* Convert the encoding to a normalized Python string: all
+       characters are converted to lower case, spaces and hypens are
+       replaced with underscores. */
+    v = normalizestring(encoding);
     if (v == NULL)
 	goto onError;
     PyString_InternInPlace(&v);

--------------751C647C75D9C69DDEBDDF5E--




From jeremy@cnri.reston.va.us  Mon Apr  3 17:36:24 2000
From: jeremy@cnri.reston.va.us (Jeremy Hylton)
Date: Mon, 3 Apr 2000 12:36:24 -0400 (EDT)
Subject: [Patches] [1.6] dictionary objects: new method 'supplement'
In-Reply-To: <Pine.LNX.4.10.10004030315040.19464-100000@nebula.lyra.org>
References: <m12c3a9-000CnCC@artcom0.artcom-gmbh.de>
 <Pine.LNX.4.10.10004030315040.19464-100000@nebula.lyra.org>
Message-ID: <14568.51336.811523.937351@bitdiddle.cnri.reston.va.us>

I agree with Greg.

Jeremy



From bwarsaw@cnri.reston.va.us  Mon Apr  3 18:20:19 2000
From: bwarsaw@cnri.reston.va.us (Barry A. Warsaw)
Date: Mon, 3 Apr 2000 13:20:19 -0400 (EDT)
Subject: [Python-Dev] Re: [Patches] [1.6] dictionary objects: new method 'supplement'
References: <m12c3a9-000CnCC@artcom0.artcom-gmbh.de>
 <Pine.LNX.4.10.10004030315040.19464-100000@nebula.lyra.org>
 <14568.51336.811523.937351@bitdiddle.cnri.reston.va.us>
Message-ID: <14568.53971.777162.624760@anthem.cnri.reston.va.us>

-0 on dict.supplement(), not the least because I'll always missspell
it :)

-Barry


From Mark.Favas@per.dem.csiro.au  Tue Apr  4 04:07:13 2000
From: Mark.Favas@per.dem.csiro.au (Favas, Mark (EM, Floreat))
Date: Tue, 4 Apr 2000 11:07:13 +0800
Subject: [Patches] RE: Revised Patches for bug report 258 - mmapmodule.c compile pro
 blem
Message-ID: <1508AAEAF1D9D111B65300A0C944CCF11632BF@mailhost.per.its.CSIRO.AU>

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_000_01BF9DE2.DFE78026
Content-Type: text/plain;
	charset="windows-1252"

OK - attached are revised patches based on the comments below. Compiles
without error/warning and passes test_mmap... Thanks, Greg.

Cheers,
	Mark

-----Original Message-----
From: Greg Stein [mailto:gstein@lyra.org]
Sent: Monday, 3 April 2000 6:13 PM
To: Mark Favas
Cc: patches@python.org
Subject: Re: [Patches] Patches for bug report 258 - mmapmodule.c compile
problem


On Mon, 3 Apr 2000, Mark Favas wrote:
>...
> *** mmapmodule.c.orig   Mon Apr  3 10:23:19 2000
> --- mmapmodule.c        Mon Apr  3 14:57:23 2000
> ***************
> *** 118,124 ****
>         char value;
>         char * where = (self->data+self->pos);
>         CHECK_VALID(NULL);
> !       if ((where >= 0) && (where < (self->data+self->size))) {
>                 value = (char) *(where);
>                 self->pos += 1;
>                 return Py_BuildValue("c", (char) *(where));
> --- 118,124 ----
>         char value;
>         char * where = (self->data+self->pos);
>         CHECK_VALID(NULL);
> !       if ((where >= (char *) 0) && (where < (self->data+self->size)))
> {
>                 value = (char) *(where);
>                 self->pos += 1;
>                 return Py_BuildValue("c", (char) *(where));

Nope... that should be (where >= self->data).

The more obvious test would be something like:

  if (self->pos >= 0 && self->pos < self->size)

and throw the error before ever worrying about setting up "where".

> ***************
> *** 617,623 ****
>                                 "mmap slice assignment is wrong size");
>                 return -1;
>         }
> !       buf = PyString_AsString(v);
>         memcpy(self->data + ilow, buf, ihigh-ilow);
>         return 0;
>   }
> --- 617,623 ----
>                                 "mmap slice assignment is wrong size");
>                 return -1;
>         }
> !       buf = (unsigned char *) PyString_AsString(v);
>         memcpy(self->data + ilow, buf, ihigh-ilow);
>         return 0;
>   }

Rather than casting, the definition of "buf" should change to
"const char *buf".

> ***************
> *** 640,646 ****
>                         "mmap assignment must be single-character
> string");
>                 return -1;
>         }
> !       buf = PyString_AsString(v);
>         self->data[i] = buf[0];
>         return 0;
>   }
> --- 640,646 ----
>                         "mmap assignment must be single-character
> string");
>                 return -1;
>         }
> !       buf = (unsigned char *) PyString_AsString(v);
>         self->data[i] = buf[0];
>         return 0;
>   }

Same here.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/


------_=_NextPart_000_01BF9DE2.DFE78026
Content-Type: application/octet-stream;
	name="mmap-patch"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment;
	filename="mmap-patch"

*** mmapmodule.c.orig	Mon Apr  3 10:23:19 2000=0A=
--- mmapmodule.c	Tue Apr  4 10:58:26 2000=0A=
***************=0A=
*** 116,124 ****=0A=
  			   PyObject * args)=0A=
  {=0A=
  	char value;=0A=
! 	char * where =3D (self->data+self->pos);=0A=
  	CHECK_VALID(NULL);=0A=
! 	if ((where >=3D 0) && (where < (self->data+self->size))) {=0A=
  		value =3D (char) *(where);=0A=
  		self->pos +=3D 1;=0A=
  		return Py_BuildValue("c", (char) *(where));=0A=
--- 116,125 ----=0A=
  			   PyObject * args)=0A=
  {=0A=
  	char value;=0A=
! 	char * where;=0A=
  	CHECK_VALID(NULL);=0A=
! 	if (self->pos >=3D 0 && self->pos < self->size) {=0A=
! 	        where =3D self->data + self->pos;=0A=
  		value =3D (char) *(where);=0A=
  		self->pos +=3D 1;=0A=
  		return Py_BuildValue("c", (char) *(where));=0A=
***************=0A=
*** 593,599 ****=0A=
  	int ilow, ihigh;=0A=
  	PyObject *v;=0A=
  {=0A=
! 	unsigned char *buf;=0A=
  =0A=
  	CHECK_VALID(-1);=0A=
  	if (ilow < 0)=0A=
--- 594,600 ----=0A=
  	int ilow, ihigh;=0A=
  	PyObject *v;=0A=
  {=0A=
! 	const char *buf;=0A=
  =0A=
  	CHECK_VALID(-1);=0A=
  	if (ilow < 0)=0A=
***************=0A=
*** 628,634 ****=0A=
  	int i;=0A=
  	PyObject *v;=0A=
  {=0A=
! 	unsigned char *buf;=0A=
   =0A=
  	CHECK_VALID(-1);=0A=
  	if (i < 0 || i >=3D self->size) {=0A=
--- 629,635 ----=0A=
  	int i;=0A=
  	PyObject *v;=0A=
  {=0A=
! 	const char *buf;=0A=
   =0A=
  	CHECK_VALID(-1);=0A=
  	if (i < 0 || i >=3D self->size) {=0A=

------_=_NextPart_000_01BF9DE2.DFE78026--


From gansevle@cs.utwente.nl  Tue Apr  4 15:51:07 2000
From: gansevle@cs.utwente.nl (Fred Gansevles)
Date: Tue, 04 Apr 2000 16:51:07 +0200
Subject: [Patches] (no subject)
Message-ID: <200004041451.QAA07735@localhost.localdomain>

This patch solves 2 problems of the os module.
1) Bug ID #50 (case-mismatch wiht "environ.get(..,..)" and "del environ[..]")
2) os.environ.update (dict) doesn't propagate changes to the 'real'
   environment (i.e doesn't call putenv)

This patches also has minor changes specific for 1.6a
The string module isn't used anymore, instead the strings own methods are
used.

----------------------------------------------------------------
I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part

of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
FAQ about legal
I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

----------------------------------------------------------------
Index: os.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/os.py,v
retrieving revision 1.30
diff -c -r1.30 os.py
*** os.py	2000/02/04 15:39:30	1.30
--- os.py	2000/04/04 14:34:35
***************
*** 220,227 ****
          envpath = env['PATH']
      else:
          envpath = defpath
!     import string
!     PATH = string.splitfields(envpath, pathsep)
      if not _notfound:
          import tempfile
          # Exec a file that is guaranteed not to exist
--- 220,226 ----
          envpath = env['PATH']
      else:
          envpath = defpath
!     PATH = envpath.split (pathsep)
      if not _notfound:
          import tempfile
          # Exec a file that is guaranteed not to exist
***************
*** 248,269 ****
  
      if name in ('os2', 'nt', 'dos'):  # Where Env Var Names Must Be UPPERCASE
          # But we store them as upper case
-         import string
          class _Environ(UserDict.UserDict):
              def __init__(self, environ):
                  UserDict.UserDict.__init__(self)
                  data = self.data
-                 upper = string.upper
                  for k, v in environ.items():
!                     data[upper(k)] = v
              def __setitem__(self, key, item):
                  putenv(key, item)
!                 key = string.upper(key)
!                 self.data[key] = item
              def __getitem__(self, key):
!                 return self.data[string.upper(key)]
              def has_key(self, key):
!                 return self.data.has_key(string.upper(key))
  
      else:  # Where Env Var Names Can Be Mixed Case
          class _Environ(UserDict.UserDict):
--- 247,272 ----
  
      if name in ('os2', 'nt', 'dos'):  # Where Env Var Names Must Be UPPERCASE
          # But we store them as upper case
          class _Environ(UserDict.UserDict):
              def __init__(self, environ):
                  UserDict.UserDict.__init__(self)
                  data = self.data
                  for k, v in environ.items():
!                     data[k.upper()] = v
              def __setitem__(self, key, item):
                  putenv(key, item)
!                 self.data[key.upper()] = item
              def __getitem__(self, key):
!                 return self.data[key.upper()]
!             def __delitem__(self, key):
!                 del self.data[key.upper()]
              def has_key(self, key):
!                 return self.data.has_key(key.upper())
!             def get(self, key, failobj=None):
!                 return self.data.get(key.upper(), failobj)
!             def update(self, dict):
!                 for k, v in dict.items():
!                     self[k] = v
  
      else:  # Where Env Var Names Can Be Mixed Case
          class _Environ(UserDict.UserDict):
***************
*** 273,278 ****
--- 276,284 ----
              def __setitem__(self, key, item):
                  putenv(key, item)
                  self.data[key] = item
+             def update(self, dict):
+                 for k, v in dict.items():
+                     self[k] = v
  
      environ = _Environ(environ)
  


From hniksic@iskon.hr  Tue Apr  4 19:11:32 2000
From: hniksic@iskon.hr (Hrvoje Niksic)
Date: 04 Apr 2000 20:11:32 +0200
Subject: [Patches] Fix for readline() method of mmap objects
Message-ID: <dpd7o5u46j.fsf@mraz.iskon.hr>

I'm glad that mmapmodule.c made it into Python 1.6.  Some time ago
I've found a bug in it, and it's still there.

The bug is in mmap_read_line_method(), and its loop that searches for
newlines.  After the loop reaches EOF, eol is incremented and points
after the end of the memory.  This results in readline() method
sometimes picking up and returning a byte after the end of the string.
This is usually a bogus \0, but it could cause SIGSEGV if it's after
the end of the page).

The patch fixes the problem.  Also, it uses memchr() for finding a
character, which is in fact the "strnchr" the comment is asking for.
memchr() is already used in Python sources, so there should be no
portability problems.

*** Modules/mmapmodule.c.old	Tue Apr  4 15:57:07 2000
--- Modules/mmapmodule.c	Tue Apr  4 15:57:40 2000
***************
*** 132,151 ****
  mmap_read_line_method (mmap_object * self,
  			   PyObject * args)
  {
! 	char * start;
  	char * eof = self->data+self->size;
  	char * eol;
  	PyObject * result;
  
  	CHECK_VALID(NULL);
- 	start = self->data+self->pos;
  
! 	/* strchr was a bad idea here - there's no way to range
! 	   check it.  there is no 'strnchr' */
! 	for (eol = start; (eol < eof) && (*eol != '\n') ; eol++)
! 	{ /* do nothing */ }
! 
! 	result = Py_BuildValue("s#", start, (long) (++eol - start));
  	self->pos += (eol - start);
  	return (result);
  }
--- 132,151 ----
  mmap_read_line_method (mmap_object * self,
  			   PyObject * args)
  {
! 	char * start = self->data+self->pos;
  	char * eof = self->data+self->size;
  	char * eol;
  	PyObject * result;
  
  	CHECK_VALID(NULL);
  
! 	eol = memchr(start, '\n', self->size - self->pos);
! 	if (!eol)
! 		eol = eof;
! 	else
! 		++eol;		/* we're interested in the position after the
! 				   newline. */
! 	result = PyString_FromStringAndSize(start, (long) (eol - start));
  	self->pos += (eol - start);
  	return (result);
  }


I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.


From guido@python.org  Tue Apr  4 22:43:07 2000
From: guido@python.org (Guido van Rossum)
Date: Tue, 04 Apr 2000 17:43:07 -0400
Subject: [Patches] Unicode Patch Set 2000-04-03
In-Reply-To: Your message of "Mon, 03 Apr 2000 16:46:28 +0200."
 <38E8AEC4.CE1E63BB@lemburg.com>
References: <38E895A5.98FC0A4B@lemburg.com>
 <38E8AEC4.CE1E63BB@lemburg.com>
Message-ID: <200004042143.RAA14149@eric.cnri.reston.va.us>

Your patch set introduced a syntax error in string.py, because a line
("return _float(s)" in atof()) had changed only in indentation, and
"diff -b" doesn't flag that as a difference.  A similar thing happened
in test_unicode.py, where two asserts weren't indented.

I wonder if there are other places where this might have happened
without introducing a syntax error...

I've also found another problem (which may have existed before???):
in test.test_string.py, string.atol('  1x ',) is supposed to raise
ValueError, but actually returns 1L.

I'll hold off on checkin in these changes...

--Guido van Rossum (home page: http://www.python.org/~guido/)


From mal@lemburg.com  Tue Apr  4 23:10:17 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Wed, 05 Apr 2000 00:10:17 +0200
Subject: [Patches] Unicode Patch Set 2000-04-03
References: <38E895A5.98FC0A4B@lemburg.com>
 <38E8AEC4.CE1E63BB@lemburg.com> <200004042143.RAA14149@eric.cnri.reston.va.us>
Message-ID: <38EA6849.F3EA5F24@lemburg.com>

Guido van Rossum wrote:
> 
> Your patch set introduced a syntax error in string.py, because a line
> ("return _float(s)" in atof()) had changed only in indentation, and
> "diff -b" doesn't flag that as a difference.  A similar thing happened
> in test_unicode.py, where two asserts weren't indented.

Hmm, you're right. I'll drop the "-b" and resend the patch
along with the compare updates.
 
> I wonder if there are other places where this might have happened
> without introducing a syntax error...
> 
> I've also found another problem (which may have existed before???):
> in test.test_string.py, string.atol('  1x ',) is supposed to raise
> ValueError, but actually returns 1L.
> 
> I'll hold off on checkin in these changes...

Ouch, I only checked test_builtin, not test_string when I
added the new Unicode support to int() et al. 

The different semantics in string.atol() are due to the change
I did to the string.py ato?() APIs. I asked on python-dev about
this and noone objected, so I simply tried the change. While
doing so I found a rather strange semantic difference in the
string->long conversion:

Quote from python-dev:
"""
One thing I noticed, which needs some discussion:

There are two separate APIs which convert long string literals
to long objects: PyNumber_Long() and PyLong_FromString().
The first applies the same error checking as does the
PyInt_FromString() API, while the latter does not apply
this check... 

Question is: shouldn't the check for truncated
data ("9.5" -> 9L) be moved into PyLong_FromString() ?
"""

This is the origin of long('  1x  ') causing an exception,
while long('   1x   ',10) masks the error. IMHO, these two
should be synchronised to int()'s behaviour which is to
raise errors in both cases.

Since I changed string.atol to use long() directly rather
than trying to emulate the old error string, the different
behaviour now shines through :-)

Should I leave the string.py patch out of the set until
there has been some more discussion on python-dev ?

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/



From guido@python.org  Tue Apr  4 23:07:34 2000
From: guido@python.org (Guido van Rossum)
Date: Tue, 04 Apr 2000 18:07:34 -0400
Subject: [Patches] Unicode Patch Set 2000-04-03
In-Reply-To: Your message of "Wed, 05 Apr 2000 00:10:17 +0200."
 <38EA6849.F3EA5F24@lemburg.com>
References: <38E895A5.98FC0A4B@lemburg.com> <38E8AEC4.CE1E63BB@lemburg.com> <200004042143.RAA14149@eric.cnri.reston.va.us>
 <38EA6849.F3EA5F24@lemburg.com>
Message-ID: <200004042207.SAA14508@eric.cnri.reston.va.us>

> The different semantics in string.atol() are due to the change
> I did to the string.py ato?() APIs. I asked on python-dev about
> this and noone objected, so I simply tried the change. While
> doing so I found a rather strange semantic difference in the
> string->long conversion:
> 
> Quote from python-dev:
> """
> One thing I noticed, which needs some discussion:
> 
> There are two separate APIs which convert long string literals
> to long objects: PyNumber_Long() and PyLong_FromString().
> The first applies the same error checking as does the
> PyInt_FromString() API, while the latter does not apply
> this check... 
> 
> Question is: shouldn't the check for truncated
> data ("9.5" -> 9L) be moved into PyLong_FromString() ?
> """
> 
> This is the origin of long('  1x  ') causing an exception,
> while long('   1x   ',10) masks the error. IMHO, these two
> should be synchronised to int()'s behaviour which is to
> raise errors in both cases.
> 
> Since I changed string.atol to use long() directly rather
> than trying to emulate the old error string, the different
> behaviour now shines through :-)
> 
> Should I leave the string.py patch out of the set until
> there has been some more discussion on python-dev ?

I was too busy to follow this, but I think that the error should be
raised in all cases -- no further discussion needed.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From mal@lemburg.com  Tue Apr  4 23:23:18 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Wed, 05 Apr 2000 00:23:18 +0200
Subject: [Patches] string.atol() [Unicode Patch Set 2000-04-03]
References: <38E895A5.98FC0A4B@lemburg.com>
 <38E8AEC4.CE1E63BB@lemburg.com> <200004042143.RAA14149@eric.cnri.reston.va.us> <38EA6849.F3EA5F24@lemburg.com>
Message-ID: <38EA6B56.B6C6557D@lemburg.com>

"M.-A. Lemburg" wrote:
> 
> Guido van Rossum wrote:
> >
> > I've also found another problem (which may have existed before???):
> > in test.test_string.py, string.atol('  1x ',) is supposed to raise
> > ValueError, but actually returns 1L.
> >
> > I'll hold off on checkin in these changes...
> 
> Ouch, I only checked test_builtin, not test_string when I
> added the new Unicode support to int() et al.
> 
> The different semantics in string.atol() are due to the change
> I did to the string.py ato?() APIs. I asked on python-dev about
> this and noone objected, so I simply tried the change. While
> doing so I found a rather strange semantic difference in the
> string->long conversion:
>
> Should I leave the string.py patch out of the set until
> there has been some more discussion on python-dev ?

BTW, to get the regr. test to pass and string.atol() work
exactly like before, you only have to change the default
base to the magic number -909.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/



From mal@lemburg.com  Wed Apr  5 01:53:10 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Wed, 05 Apr 2000 02:53:10 +0200
Subject: [Patches] Unicode Patch Set 2000-04-05
Message-ID: <38EA8E76.66359797@lemburg.com>

This is a multi-part message in MIME format.
--------------962125C15898CA694E9D8F57
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

After not having much luck with the last patch set, here's a new
one which hopefully fixes all the problems with the other two
versions.

Here's what's in the bag:

New Unicode support for int(), float(), complex() and long().

- new APIs PyInt_FromUnicode() and PyLong_FromUnicode()
- added support for Unicode to PyFloat_FromString()
- new encoding API PyUnicode_EncodeDecimal() which converts
  Unicode to a decimal char* string (used in the above new
  APIs)
- shortcuts for calls like int(<int object>) and float(<float obj>)
- tests for all of the above

Unicode compares and contains checks:
- comparing Unicode and non-string types now works; TypeErrors
  are masked, all other errors such as ValueError during
  Unicode coercion are passed through (note that PyUnicode_Compare
  does not implement the masking -- PyObject_Compare does this)
- contains now works for non-string types too; TypeErrors are
  masked and 0 returned; all other errors are passed through

Better testing support for the standard codecs.

Misc minor enhancements, such as an alias dbcs for the mbcs codec.

Changes:
- PyLong_FromString() now applies the same error checks as
  does PyInt_FromString(): trailing garbage is reported
  as error and not longer silently ignored. The only characters
  which may be trailing the digits are 'L' and 'l' -- these
  are still silently ignored.
- string.ato?() now directly interface to int(), long() and
  float(). The error strings are now a little different, but
  the type still remains the same. These functions are now
  ready to get declared obsolete ;-)
- PyNumber_Int() now also does a check for embedded NULL chars
  in the input string; PyNumber_Long() already did this (and
  still does)

--

Some minor things I noticed while compiling the CVS Version
on Linux2/glibc2 using -Wall which may or may not need
tweaking:

fileobject.c: In function `file_seek':
fileobject.c:285: warning: implicit declaration of function `fseeko'
ceval.c: In function `eval_code2':
ceval.c:355: warning: `opcode' might be used uninitialized in this function
ceval.c:356: warning: `oparg' might be used uninitialized in this function
./pypcre.c: In function `compile_branch':
./pypcre.c:1222: warning: `c' might be used uninitialized in this function
./pypcre.c: In function `pcre_exec':
./pypcre.c:4524: warning: variable `start_bits' might be clobbered by `longjmp' or `vfork'
./pypcre.c:4525: warning: variable `start_match' might be clobbered by `longjmp' or `vfork'
./posixmodule.c:504: warning: `posix_strint' defined but not used
./readline.c: In function `call_readline':
./readline.c:386: warning: implicit declaration of function `sigrelse'
./timemodule.c:413: warning: `/*' within comment
./timemodule.c: In function `time_strptime':
./timemodule.c:428: warning: implicit declaration of function `strptime'
./timemodule.c:428: warning: assignment makes pointer from integer without a cast
./mmapmodule.c: In function `new_mmap_object':
./mmapmodule.c:699: warning: unused variable `namelen'
./mmapmodule.c:698: warning: unused variable `filename'

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------962125C15898CA694E9D8F57
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-05.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-05.patch"

diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Include/Python.h Python+Unicode/Include/Python.h
--- CVS-Python/Include/Python.h	Sat Mar 11 10:52:34 2000
+++ Python+Unicode/Include/Python.h	Mon Apr  3 13:18:05 2000
@@ -72,6 +72,7 @@
 
 #include "pydebug.h"
 
+#include "unicodeobject.h"
 #include "intobject.h"
 #include "longobject.h"
 #include "floatobject.h"
@@ -92,7 +93,6 @@
 #include "cobject.h"
 #include "traceback.h"
 #include "sliceobject.h"
-#include "unicodeobject.h"
 
 #include "codecs.h"
 #include "pyerrors.h"
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Include/intobject.h Python+Unicode/Include/intobject.h
--- CVS-Python/Include/intobject.h	Tue Mar  7 18:58:16 2000
+++ Python+Unicode/Include/intobject.h	Mon Apr  3 12:54:35 2000
@@ -60,6 +60,7 @@
 #define PyInt_Check(op) ((op)->ob_type == &PyInt_Type)
 
 extern DL_IMPORT(PyObject *) PyInt_FromString Py_PROTO((char*, char**, int));
+extern DL_IMPORT(PyObject *) PyInt_FromUnicode Py_PROTO((Py_UNICODE*, int, int));
 extern DL_IMPORT(PyObject *) PyInt_FromLong Py_PROTO((long));
 extern DL_IMPORT(long) PyInt_AsLong Py_PROTO((PyObject *));
 extern DL_IMPORT(long) PyInt_GetMax Py_PROTO((void));
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Include/longobject.h Python+Unicode/Include/longobject.h
--- CVS-Python/Include/longobject.h	Thu Jan 14 11:24:51 1999
+++ Python+Unicode/Include/longobject.h	Mon Apr  3 13:00:50 2000
@@ -82,6 +82,7 @@
 #endif /* HAVE_LONG_LONG */
 
 DL_IMPORT(PyObject *) PyLong_FromString Py_PROTO((char *, char **, int));
+DL_IMPORT(PyObject *) PyLong_FromUnicode Py_PROTO((Py_UNICODE*, int, int));
 
 #ifdef __cplusplus
 }
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Include/unicodeobject.h Python+Unicode/Include/unicodeobject.h
--- CVS-Python/Include/unicodeobject.h	Wed Mar 29 11:50:05 2000
+++ Python+Unicode/Include/unicodeobject.h	Wed Apr  5 00:56:31 2000
@@ -358,7 +358,7 @@
 
 /* --- UTF-16 Codecs ------------------------------------------------------ */
 
-/* Decodes length bytes from a UTF-16 encoded buffer string and return
+/* Decodes length bytes from a UTF-16 encoded buffer string and returns
    the corresponding Unicode object.
 
    errors (if non-NULL) defines the error handling. It defaults
@@ -397,7 +397,7 @@
     );
 
 /* Returns a Python string object holding the UTF-16 encoded value of
-   the Unicode data in s.
+   the Unicode data.
 
    If byteorder is not 0, output is written according to the following
    byte order:
@@ -586,6 +586,37 @@
     );
 
 #endif /* MS_WIN32 */
+
+/* --- Decimal Encoder ---------------------------------------------------- */
+
+/* Takes a Unicode string holding a decimal value and writes it into
+   an output buffer using standard ASCII digit codes.
+
+   The output buffer has to provide at least length+1 bytes of storage
+   area. The output string is 0-terminated.
+
+   The encoder converts whitespace to ' ', decimal characters to their
+   corresponding ASCII digit and all other Latin-1 characters except
+   \0 as-is. Characters outside this range (Unicode ordinals 1-256)
+   are treated as errors. This includes embedded NULL bytes.
+
+   Error handling is defined by the errors argument:
+
+      NULL or "strict": raise a ValueError
+      "ignore": ignore the wrong characters (these are not copied to the
+		output buffer)
+      "replace": replaces illegal characters with '?'
+
+   Returns 0 on success, -1 on failure.
+
+*/
+
+extern DL_IMPORT(int) PyUnicode_EncodeDecimal(
+    Py_UNICODE *s,		/* Unicode buffer */
+    int length,			/* Number of Py_UNICODE chars to encode */
+    char *output,		/* Output buffer; must have size >= length */
+    const char *errors		/* error handling */
+    );
 
 /* --- Methods & Slots ----------------------------------------------------
 
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Lib/test/output/test_unicode Python+Unicode/Lib/test/output/test_unicode
--- CVS-Python/Lib/test/output/test_unicode	Wed Mar 29 11:50:15 2000
+++ Python+Unicode/Lib/test/output/test_unicode	Wed Apr  5 02:40:54 2000
@@ -3,3 +3,4 @@
 Testing Unicode contains method... done.
 Testing Unicode formatting strings... done.
 Testing builtin codecs... done.
+Testing standard mapping codecs... 0-127... 128-255... done.
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/abstract.c Python+Unicode/Objects/abstract.c
--- CVS-Python/Objects/abstract.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/abstract.c	Wed Apr  5 00:54:48 2000
@@ -726,6 +726,27 @@
 	return type_error("bad operand type for abs()");
 }
 
+/* Add a check for embedded NULL-bytes in the argument. */
+static PyObject *
+int_from_string(s, len)
+	const char *s;
+	int len;
+{
+	char *end;
+	PyObject *x;
+
+	x = PyInt_FromString((char*)s, &end, 10);
+	if (x == NULL)
+		return NULL;
+	if (end != s + len) {
+		PyErr_SetString(PyExc_ValueError,
+				"null byte in argument for int()");
+		Py_DECREF(x);
+		return NULL;
+	}
+	return x;
+}
+
 PyObject *
 PyNumber_Int(o)
 	PyObject *o;
@@ -736,69 +757,42 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyInt_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (PyString_Check(o))
-		return PyInt_FromString(PyString_AS_STRING(o), NULL, 10);
+		return int_from_string(PyString_AS_STRING(o), 
+				       PyString_GET_SIZE(o));
+	if (PyUnicode_Check(o))
+		return PyInt_FromUnicode(PyUnicode_AS_UNICODE(o),
+					 PyUnicode_GET_SIZE(o),
+					 10);
 	m = o->ob_type->tp_as_number;
 	if (m && m->nb_int)
 		return m->nb_int(o);
 	if (!PyObject_AsCharBuffer(o, &buffer, &buffer_len))
-		return PyInt_FromString((char*)buffer, NULL, 10);
+		return int_from_string((char*)buffer, buffer_len);
 
 	return type_error("object can't be converted to int");
 }
 
-/* There are two C API functions for converting a string to a long,
- * PyNumber_Long() and PyLong_FromString().  Both are used in builtin_long, 
- * reachable from Python with the built-in function long().
- *
- * The difference is this: PyNumber_Long will raise an exception when the
- * string cannot be converted to a long.  The most common situation is
- * where a float string is passed in; this raises a ValueError.
- * PyLong_FromString does not raise an exception; it silently truncates the 
- * float to an integer.
- *
- * You can see the different behavior from Python with the following:
- *
- * long('9.5')
- * => ValueError: invalid literal for long(): 9.5
- *
- * long('9.5', 10)
- * => 9L
- *
- * The first example ends up calling PyNumber_Long(), while the second one
- * calls PyLong_FromString().
- */
+/* Add a check for embedded NULL-bytes in the argument. */
 static PyObject *
 long_from_string(s, len)
 	const char *s;
 	int len;
 {
-	const char *start;
 	char *end;
 	PyObject *x;
-	char buffer[256]; /* For errors */
 
-	start = s;
-	while (*s && isspace(Py_CHARMASK(*s)))
-		s++;
 	x = PyLong_FromString((char*)s, &end, 10);
-	if (x == NULL) {
-		if (PyErr_ExceptionMatches(PyExc_ValueError))
-			goto bad;
+	if (x == NULL)
 		return NULL;
-	}
-	while (*end && isspace(Py_CHARMASK(*end)))
-		end++;
-	if (*end != '\0') {
-  bad:
-		sprintf(buffer, "invalid literal for long(): %.200s", s);
-		PyErr_SetString(PyExc_ValueError, buffer);
-		Py_XDECREF(x);
-		return NULL;
-	}
-	else if (end != start + len) {
+	if (end != s + len) {
 		PyErr_SetString(PyExc_ValueError,
 				"null byte in argument for long()");
+		Py_DECREF(x);
 		return NULL;
 	}
 	return x;
@@ -814,6 +808,10 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyLong_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (PyString_Check(o))
 		/* need to do extra error checking that PyLong_FromString() 
 		 * doesn't do.  In particular long('9.5') must raise an
@@ -821,6 +819,11 @@
 		 */
 		return long_from_string(PyString_AS_STRING(o),
 					PyString_GET_SIZE(o));
+	if (PyUnicode_Check(o))
+		/* The above check is done in PyLong_FromUnicode(). */
+		return PyLong_FromUnicode(PyUnicode_AS_UNICODE(o),
+					  PyUnicode_GET_SIZE(o),
+					  10);
 	m = o->ob_type->tp_as_number;
 	if (m && m->nb_long)
 		return m->nb_long(o);
@@ -838,6 +841,10 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyFloat_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (!PyString_Check(o)) {
 		m = o->ob_type->tp_as_number;
 		if (m && m->nb_float)
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/floatobject.c Python+Unicode/Objects/floatobject.c
--- CVS-Python/Objects/floatobject.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/floatobject.c	Mon Apr  3 14:42:03 2000
@@ -163,6 +163,22 @@
 	if (PyString_Check(v)) {
 		s = PyString_AS_STRING(v);
 		len = PyString_GET_SIZE(v);
+	}
+	else if (PyUnicode_Check(v)) {
+		char s_buffer[256];
+
+		if (PyUnicode_GET_SIZE(v) >= sizeof(s_buffer)) {
+			PyErr_SetString(PyExc_ValueError,
+				 "float() literal too large to convert");
+			return NULL;
+		}
+		if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v), 
+					    PyUnicode_GET_SIZE(v),
+					    s_buffer, 
+					    NULL))
+			return NULL;
+		s = s_buffer;
+		len = strlen(s);
 	}
 	else if (PyObject_AsCharBuffer(v, &s, &len)) {
 		PyErr_SetString(PyExc_TypeError,
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/intobject.c Python+Unicode/Objects/intobject.c
--- CVS-Python/Objects/intobject.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/intobject.c	Mon Apr  3 14:42:35 2000
@@ -259,6 +259,24 @@
 	if (pend)
 		*pend = end;
 	return PyInt_FromLong(x);
+}
+
+PyObject *
+PyInt_FromUnicode(s, length, base)
+	Py_UNICODE *s;
+	int length;
+	int base;
+{
+	char buffer[256];
+	
+	if (length >= sizeof(buffer)) {
+		PyErr_SetString(PyExc_ValueError,
+				"int() literal too large to convert");
+		return NULL;
+	}
+	if (PyUnicode_EncodeDecimal(s, length, buffer, NULL))
+		return NULL;
+	return PyInt_FromString(buffer, NULL, base);
 }
 
 /* Methods */
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/longobject.c Python+Unicode/Objects/longobject.c
--- CVS-Python/Objects/longobject.c	Wed Jan 19 18:01:30 2000
+++ Python+Unicode/Objects/longobject.c	Wed Apr  5 01:15:23 2000
@@ -724,7 +724,7 @@
 	int base;
 {
 	int sign = 1;
-	char *start;
+	char *start, *orig_str = str;
 	PyLongObject *z;
 	
 	if ((base != 0 && base < 2) || base > 36) {
@@ -772,17 +772,44 @@
 	}
 	if (z == NULL)
 		return NULL;
-	if (str == start) {
-		PyErr_SetString(PyExc_ValueError,
-				"no digits in long int constant");
-		Py_DECREF(z);
-		return NULL;
-	}
+	if (str == start)
+		goto onError;
 	if (sign < 0 && z != NULL && z->ob_size != 0)
 		z->ob_size = -(z->ob_size);
+	if (*str == 'L' || *str == 'l')
+		str++;
+	while (*str && isspace(Py_CHARMASK(*str)))
+		str++;
+	if (*str != '\0')
+		goto onError;
 	if (pend)
 		*pend = str;
 	return (PyObject *) z;
+
+ onError:
+	PyErr_Format(PyExc_ValueError, 
+		     "invalid literal for long(): %.200s", orig_str);
+	Py_XDECREF(z);
+	return NULL;
+}
+
+PyObject *
+PyLong_FromUnicode(u, length, base)
+	Py_UNICODE *u;
+	int length;
+	int base;
+{
+	char buffer[256];
+
+	if (length >= sizeof(buffer)) {
+		PyErr_SetString(PyExc_ValueError,
+				"long() literal too large to convert");
+		return NULL;
+	}
+	if (PyUnicode_EncodeDecimal(u, length, buffer, NULL))
+		return NULL;
+
+	return PyLong_FromString(buffer, NULL, base);
 }
 
 static PyLongObject *x_divrem
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/object.c Python+Unicode/Objects/object.c
--- CVS-Python/Objects/object.c	Tue Mar 28 09:19:17 2000
+++ Python+Unicode/Objects/object.c	Mon Apr  3 23:17:04 2000
@@ -347,8 +347,21 @@
 				return cmp;
 			}
 		}
-		else if (PyUnicode_Check(v) || PyUnicode_Check(w))
-			return PyUnicode_Compare(v, w);
+		else if (PyUnicode_Check(v) || PyUnicode_Check(w)) {
+			int result = PyUnicode_Compare(v, w);
+			if (result == -1 && PyErr_Occurred() && 
+			    PyErr_ExceptionMatches(PyExc_TypeError))
+				/* TypeErrors are ignored: if Unicode coercion
+				fails due to one of the arguments not
+			 	having the right type, we continue as
+				defined by the coercion protocol (see
+				above). Luckily, decoding errors are
+				reported as ValueErrors and are not masked
+				by this technique. */
+				PyErr_Clear();
+			else
+				return result;
+		}
 		else if (vtp->tp_as_number != NULL)
 			vname = "";
 		else if (wtp->tp_as_number != NULL)
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/unicodeobject.c Python+Unicode/Objects/unicodeobject.c
--- CVS-Python/Objects/unicodeobject.c	Mon Apr  3 13:46:02 2000
+++ Python+Unicode/Objects/unicodeobject.c	Wed Apr  5 01:08:15 2000
@@ -329,8 +329,14 @@
 	s = PyString_AS_STRING(obj);
 	len = PyString_GET_SIZE(obj);
     }
-    else if (PyObject_AsCharBuffer(obj, &s, &len))
+    else if (PyObject_AsCharBuffer(obj, &s, &len)) {
+	/* Overwrite the error message with something more useful in
+	   case of a TypeError. */
+	if (PyErr_ExceptionMatches(PyExc_TypeError))
+	    PyErr_SetString(PyExc_TypeError,
+			    "coercing to Unicode: need string or charbuffer");
 	return NULL;
+    }
     if (len == 0) {
 	Py_INCREF(unicode_empty);
 	return (PyObject *)unicode_empty;
@@ -1923,6 +1929,60 @@
     return NULL;
 }
     
+/* --- Decimal Encoder ---------------------------------------------------- */
+
+int PyUnicode_EncodeDecimal(Py_UNICODE *s,
+			    int length,
+			    char *output,
+			    const char *errors)
+{
+    Py_UNICODE *p, *end;
+
+    if (output == NULL) {
+	PyErr_BadArgument();
+	return -1;
+    }
+
+    p = s;
+    end = s + length;
+    while (p < end) {
+	register Py_UNICODE ch = *p++;
+	int decimal;
+	
+	if (Py_UNICODE_ISSPACE(ch)) {
+	    *output++ = ' ';
+	    continue;
+	}
+	decimal = Py_UNICODE_TODECIMAL(ch);
+	if (decimal >= 0) {
+	    *output++ = '0' + decimal;
+	    continue;
+	}
+	if (0 < ch < 256) {
+	    *output++ = ch;
+	    continue;
+	}
+	/* All other characters are considered invalid */
+	if (errors == NULL || strcmp(errors, "strict") == 0) {
+	    PyErr_SetString(PyExc_ValueError,
+			    "invalid decimal Unicode string");
+	    goto onError;
+	}
+	else if (strcmp(errors, "ignore") == 0)
+	    continue;
+	else if (strcmp(errors, "replace") == 0) {
+	    *output++ = '?';
+	    continue;
+	}
+    }
+    /* 0-terminate the output string */
+    *output++ = '\0';
+    return 0;
+
+ onError:
+    return -1;
+}
+
 /* --- Helpers ------------------------------------------------------------ */
 
 static 
@@ -2810,13 +2870,26 @@
     register const Py_UNICODE *p, *e;
     register Py_UNICODE ch;
 
-    /* Coerce the two arguments */
+    /* Coerce the two arguments; if coercion fails with a TypeError
+       this is not seen as error, but instead interpreted as not
+       having the containment property. */
     u = (PyUnicodeObject *)PyUnicode_FromObject(container);
-    if (u == NULL)
+    if (u == NULL) {
+	if (PyErr_ExceptionMatches(PyExc_TypeError)) {
+	    PyErr_Clear();
+	    return 0;
+	}
 	goto onError;
+    }
     v = (PyUnicodeObject *)PyUnicode_FromObject(element);
-    if (v == NULL)
+    if (v == NULL) {
+	if (PyErr_ExceptionMatches(PyExc_TypeError)) {
+	    PyErr_Clear();
+	    Py_DECREF(u);
+	    return 0;
+	}
 	goto onError;
+    }
 
     /* Check v in u */
     if (PyUnicode_GET_SIZE(v) != 1) {
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Python/bltinmodule.c Python+Unicode/Python/bltinmodule.c
--- CVS-Python/Python/bltinmodule.c	Sat Mar 11 10:55:20 2000
+++ Python+Unicode/Python/bltinmodule.c	Mon Apr  3 16:14:02 2000
@@ -449,17 +449,44 @@
 	PyObject *v;
 {
 	extern double strtod Py_PROTO((const char *, char **));
-	char *s, *start, *end;
+	const char *s, *start;
+	char *end;
 	double x=0.0, y=0.0, z;
 	int got_re=0, got_im=0, done=0;
 	int digit_or_dot;
 	int sw_error=0;
 	int sign;
 	char buffer[256]; /* For errors */
+	int len;
 
-	start = s = PyString_AS_STRING(v);
+	if (PyString_Check(v)) {
+		s = PyString_AS_STRING(v);
+		len = PyString_GET_SIZE(v);
+	}
+	else if (PyUnicode_Check(v)) {
+		char s_buffer[256];
+
+		if (PyUnicode_GET_SIZE(v) >= sizeof(s_buffer)) {
+			PyErr_SetString(PyExc_ValueError,
+				 "complex() literal too large to convert");
+			return NULL;
+		}
+		if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v), 
+					    PyUnicode_GET_SIZE(v),
+					    s_buffer, 
+					    NULL))
+			return NULL;
+		s = s_buffer;
+		len = strlen(s);
+	}
+	else if (PyObject_AsCharBuffer(v, &s, &len)) {
+		PyErr_SetString(PyExc_TypeError,
+				"complex() needs a string first argument");
+		return NULL;
+	}
 
 	/* position on first nonblank */
+	start = s;
 	while (*s && isspace(Py_CHARMASK(*s)))
 		s++;
 	if (s[0] == '\0') {
@@ -475,7 +502,7 @@
 		switch (*s) {
 
 		case '\0':
-			if (s-start != PyString_GET_SIZE(v)) {
+			if (s-start != len) {
 				PyErr_SetString(
 					PyExc_ValueError,
 					"null byte in argument for complex()");
@@ -584,7 +611,7 @@
 	i = NULL;
 	if (!PyArg_ParseTuple(args, "O|O:complex", &r, &i))
 		return NULL;
-	if (PyString_Check(r))
+	if (PyString_Check(r) || PyUnicode_Check(r))
 		return complex_from_string(r);
 	if ((nbr = r->ob_type->tp_as_number) == NULL ||
 	    nbr->nb_float == NULL ||
@@ -1289,12 +1316,17 @@
 		return NULL;
 	if (base == -909)
 		return PyNumber_Int(v);
-	else if (!PyString_Check(v)) {
+	else if (PyString_Check(v))
+		return PyInt_FromString(PyString_AS_STRING(v), NULL, base);
+	else if (PyUnicode_Check(v))
+		return PyInt_FromUnicode(PyUnicode_AS_UNICODE(v),
+					 PyUnicode_GET_SIZE(v),
+					 base);
+	else {
 		PyErr_SetString(PyExc_TypeError,
 				"can't convert non-string with explicit base");
 		return NULL;
 	}
-	return PyInt_FromString(PyString_AS_STRING(v), NULL, base);
 }
 
 static char int_doc[] =
@@ -1319,12 +1351,17 @@
 		return NULL;
 	if (base == -909)
 		return PyNumber_Long(v);
-	else if (!PyString_Check(v)) {
+	else if (PyString_Check(v))
+		return PyLong_FromString(PyString_AS_STRING(v), NULL, base);
+	else if (PyUnicode_Check(v))
+		return PyLong_FromUnicode(PyUnicode_AS_UNICODE(v),
+					  PyUnicode_GET_SIZE(v),
+					  base);
+	else {
 		PyErr_SetString(PyExc_TypeError,
 				"can't convert non-string with explicit base");
 		return NULL;
 	}
-	return PyLong_FromString(PyString_AS_STRING(v), NULL, base);
 }
 
 static char long_doc[] =
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Python/codecs.c Python+Unicode/Python/codecs.c
--- CVS-Python/Python/codecs.c	Mon Apr  3 13:46:03 2000
+++ Python+Unicode/Python/codecs.c	Sat Apr  1 19:57:52 2000
@@ -84,8 +84,11 @@
     return -1;
 }
 
+/* Convert a string to a normalized Python string: all characters are
+   converted to lower case, spaces are replaced with underscores. */
+
 static
-PyObject *lowercasestring(const char *string)
+PyObject *normalizestring(const char *string)
 {
     register int i;
     int len = strlen(string);
@@ -96,8 +99,14 @@
     if (v == NULL)
 	return NULL;
     p = PyString_AS_STRING(v);
-    for (i = 0; i < len; i++)
-	p[i] = tolower(string[i]);
+    for (i = 0; i < len; i++) {
+        register char ch = string[i];
+        if (ch == ' ')
+            ch = '-';
+        else
+            ch = tolower(ch);
+	p[i] = ch;
+    }
     return v;
 }
 
@@ -132,8 +141,10 @@
 	    goto onError;
     }
 
-    /* Convert the encoding to a lower-cased Python string */
-    v = lowercasestring(encoding);
+    /* Convert the encoding to a normalized Python string: all
+       characters are converted to lower case, spaces and hypens are
+       replaced with underscores. */
+    v = normalizestring(encoding);
     if (v == NULL)
 	goto onError;
     PyString_InternInPlace(&v);
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/encodings/__init__.py Python+Unicode/Lib/encodings/__init__.py
--- CVS-Python/Lib/encodings/__init__.py	Thu Mar 23 23:58:42 2000
+++ Python+Unicode/Lib/encodings/__init__.py	Mon Apr  3 13:41:40 2000
@@ -4,8 +4,8 @@
     directory.
 
     Codec modules must have names corresponding to standard lower-case
-    encoding names. Hyphens are automatically converted to
-    underscores, e.g. 'utf-8' is looked up as module utf_8.
+    encoding names with hyphens mapped to underscores, e.g. 'utf-8' is
+    implemented by the module 'utf_8.py'.
 
     Each codec module must export the following interface:
 
@@ -40,7 +40,7 @@
         return entry
 
     # Import the module
-    modname = string.replace(encoding,'-','_')
+    modname = string.replace(encoding, '-', '_')
     modname = aliases.aliases.get(modname,modname)
     try:
         mod = __import__(modname,globals(),locals(),'*')
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/encodings/aliases.py Python+Unicode/Lib/encodings/aliases.py
--- CVS-Python/Lib/encodings/aliases.py	Mon Apr  3 13:45:48 2000
+++ Python+Unicode/Lib/encodings/aliases.py	Sat Apr  1 20:14:36 2000
@@ -54,4 +54,7 @@
     'macroman': 'mac_roman',
     'macturkish': 'mac_turkish',
 
+    # MBCS
+    'dbcs': 'mbcs',
+
 }
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/string.py Python+Unicode/Lib/string.py
--- CVS-Python/Lib/string.py	Sat Mar 11 10:52:36 2000
+++ Python+Unicode/Lib/string.py	Wed Apr  5 00:24:43 2000
@@ -196,14 +196,11 @@
     Return the floating point number represented by the string s.
 
     """
-    if type(s) == _StringType:
-        return _float(s)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+    return _float(s)
+
 
 # Convert string to integer
-def atoi(*args):
+def atoi(s , base=10):
     """atoi(s [,base]) -> int
 
     Return the integer represented by the string s in the given
@@ -214,23 +211,11 @@
     accepted.
 
     """
-    try:
-        s = args[0]
-    except IndexError:
-        raise TypeError('function requires at least 1 argument: %d given' %
-                        len(args))
-    # Don't catch type error resulting from too many arguments to int().  The
-    # error message isn't compatible but the error type is, and this function
-    # is complicated enough already.
-    if type(s) == _StringType:
-        return _apply(_int, args)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+    return _int(s, base)
 
 
 # Convert string to long integer
-def atol(*args):
+def atol(s, base=10):
     """atol(s [,base]) -> long
 
     Return the long integer represented by the string s in the
@@ -242,19 +227,7 @@
     unless base is 0.
 
     """
-    try:
-        s = args[0]
-    except IndexError:
-        raise TypeError('function requires at least 1 argument: %d given' %
-                        len(args))
-    # Don't catch type error resulting from too many arguments to long().  The
-    # error message isn't compatible but the error type is, and this function
-    # is complicated enough already.
-    if type(s) == _StringType:
-        return _apply(_long, args)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+    return _long(s, base)
 
 
 # Left-justify a string
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_b1.py Python+Unicode/Lib/test/test_b1.py
--- CVS-Python/Lib/test/test_b1.py	Fri Mar 26 15:58:20 1999
+++ Python+Unicode/Lib/test/test_b1.py	Wed Apr  5 01:11:33 2000
@@ -95,6 +95,7 @@
 if complex(0j, 3.14) <> 3.14j: raise TestFailed, 'complex(0j, 3.14)'
 if complex(0.0, 3.14) <> 3.14j: raise TestFailed, 'complex(0.0, 3.14)'
 if complex("  3.14+J  ") <> 3.14+1j:  raise TestFailed, 'complex("  3.14+J  )"'
+if complex(u"  3.14+J  ") <> 3.14+1j:  raise TestFailed, 'complex(u"  3.14+J  )"'
 class Z:
     def __complex__(self): return 3.14j
 z = Z()
@@ -208,6 +209,9 @@
 if float(314) <> 314.0: raise TestFailed, 'float(314)'
 if float(314L) <> 314.0: raise TestFailed, 'float(314L)'
 if float("  3.14  ") <> 3.14:  raise TestFailed, 'float("  3.14  ")'
+if float(u"  3.14  ") <> 3.14:  raise TestFailed, 'float(u"  3.14  ")'
+if float(u"  \u0663.\u0661\u0664  ") <> 3.14:
+    raise TestFailed, 'float(u"  \u0663.\u0661\u0664  ")'
 
 print 'getattr'
 import sys
@@ -254,6 +258,9 @@
 if int(-3.9) <> -3: raise TestFailed, 'int(-3.9)'
 if int(3.5) <> 3: raise TestFailed, 'int(3.5)'
 if int(-3.5) <> -3: raise TestFailed, 'int(-3.5)'
+# Different base:
+if int("10",16) <> 16L: raise TestFailed, 'int("10",16)'
+if int(u"10",16) <> 16L: raise TestFailed, 'int(u"10",16)'
 # Test conversion fron strings and various anomalies
 L = [
         ('0', 0),
@@ -267,9 +274,28 @@
         ('314 ', 314),
         ('  \t\t  314  \t\t  ', 314),
         (`sys.maxint`, sys.maxint),
+        ('  1x', ValueError),
+        ('  1  ', 1),
+        ('  1\02  ', ValueError),
         ('', ValueError),
         (' ', ValueError),
         ('  \t\t  ', ValueError),
+        (u'0', 0),
+        (u'1', 1),
+        (u'9', 9),
+        (u'10', 10),
+        (u'99', 99),
+        (u'100', 100),
+        (u'314', 314),
+        (u' 314', 314),
+        (u'\u0663\u0661\u0664 ', 314),
+        (u'  \t\t  314  \t\t  ', 314),
+        (u'  1x', ValueError),
+        (u'  1  ', 1),
+        (u'  1\02  ', ValueError),
+        (u'', ValueError),
+        (u' ', ValueError),
+        (u'  \t\t  ', ValueError),
 ]
 for s, v in L:
     for sign in "", "+", "-":
@@ -349,10 +375,17 @@
 if long(-3.9) <> -3L: raise TestFailed, 'long(-3.9)'
 if long(3.5) <> 3L: raise TestFailed, 'long(3.5)'
 if long(-3.5) <> -3L: raise TestFailed, 'long(-3.5)'
+if long("-3") <> -3L: raise TestFailed, 'long("-3")'
+if long(u"-3") <> -3L: raise TestFailed, 'long(u"-3")'
+# Different base:
+if long("10",16) <> 16L: raise TestFailed, 'long("10",16)'
+if long(u"10",16) <> 16L: raise TestFailed, 'long(u"10",16)'
 # Check conversions from string (same test set as for int(), and then some)
 LL = [
         ('1' + '0'*20, 10L**20),
         ('1' + '0'*100, 10L**100),
+        (u'1' + u'0'*20, 10L**20),
+        (u'1' + u'0'*100, 10L**100),
 ]
 for s, v in L + LL:
     for sign in "", "+", "-":
@@ -363,11 +396,11 @@
                 vv = -v
             try:
                 if long(ss) != long(vv):
-                    raise TestFailed, "int(%s)" % `ss`
+                    raise TestFailed, "long(%s)" % `ss`
             except v:
                 pass
             except ValueError, e:
-                raise TestFailed, "int(%s) raised ValueError: %s" % (`ss`, e)
+                raise TestFailed, "long(%s) raised ValueError: %s" % (`ss`, e)
 
 print 'map'
 if map(None, 'hello world') <> ['h','e','l','l','o',' ','w','o','r','l','d']:
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_unicode.py Python+Unicode/Lib/test/test_unicode.py
--- CVS-Python/Lib/test/test_unicode.py	Wed Mar 29 11:50:14 2000
+++ Python+Unicode/Lib/test/test_unicode.py	Mon Apr  3 23:06:27 2000
@@ -221,15 +221,23 @@
 
 # Contains:
 print 'Testing Unicode contains method...',
-assert ('a' in 'abdb') == 1
-assert ('a' in 'bdab') == 1
-assert ('a' in 'bdaba') == 1
-assert ('a' in 'bdba') == 1
+assert ('a' in u'abdb') == 1
+assert ('a' in u'bdab') == 1
+assert ('a' in u'bdaba') == 1
+assert ('a' in u'bdba') == 1
 assert ('a' in u'bdba') == 1
 assert (u'a' in u'bdba') == 1
 assert (u'a' in u'bdb') == 0
 assert (u'a' in 'bdb') == 0
 assert (u'a' in 'bdba') == 1
+assert (u'a' in ('a',1,None)) == 1
+assert (u'a' in (1,None,'a')) == 1
+assert (u'a' in (1,None,u'a')) == 1
+assert ('a' in ('a',1,None)) == 1
+assert ('a' in (1,None,'a')) == 1
+assert ('a' in (1,None,u'a')) == 1
+assert ('a' in ('x',1,u'y')) == 0
+assert ('a' in ('x',1,None)) == 0
 print 'done.'
 
 # Formatting:
@@ -270,11 +278,88 @@
     assert unicode(u.encode(encoding),encoding) == u
 
 u = u''.join(map(unichr, range(256)))
-for encoding in ('latin-1',):
-    assert unicode(u.encode(encoding),encoding) == u
+for encoding in (
+    'latin-1',
+    ):
+    try:
+        assert unicode(u.encode(encoding),encoding) == u
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
 
 u = u''.join(map(unichr, range(128)))
-for encoding in ('ascii',):
-    assert unicode(u.encode(encoding),encoding) == u
+for encoding in (
+    'ascii',
+    ):
+    try:
+        assert unicode(u.encode(encoding),encoding) == u
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
+
+print 'done.'
+
+print 'Testing standard mapping codecs...',
+
+print '0-127...',
+s = ''.join(map(chr, range(128)))
+for encoding in (
+    'cp037', 'cp1026',
+    'cp437', 'cp500', 'cp737', 'cp775', 'cp850',
+    'cp852', 'cp855', 'cp860', 'cp861', 'cp862',
+    'cp863', 'cp865', 'cp866', 
+    'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
+    'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
+    'mac_cyrillic', 'mac_latin2',
+
+    'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
+    'cp1256', 'cp1257', 'cp1258',
+    'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
+
+    'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
+    'cp1006', 'cp875', 'iso8859_8',
+    
+    ### These have undefined mappings:
+    #'cp424',
+    
+    ):
+    try:
+        assert unicode(s,encoding).encode(encoding) == s
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
+
+print '128-255...',
+s = ''.join(map(chr, range(128,256)))
+for encoding in (
+    'cp037', 'cp1026',
+    'cp437', 'cp500', 'cp737', 'cp775', 'cp850',
+    'cp852', 'cp855', 'cp860', 'cp861', 'cp862',
+    'cp863', 'cp865', 'cp866', 
+    'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
+    'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
+    'mac_cyrillic', 'mac_latin2',
+    
+    ### These have undefined mappings:
+    #'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
+    #'cp1256', 'cp1257', 'cp1258',
+    #'cp424', 'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
+    #'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
+    
+    ### These fail the round-trip:
+    #'cp1006', 'cp875', 'iso8859_8',
+    
+    ):
+    try:
+        assert unicode(s,encoding).encode(encoding) == s
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
 
 print 'done.'
Only in CVS-Python/Lib/test: test_userstring.py

--------------962125C15898CA694E9D8F57--




From guido@python.org  Wed Apr  5 02:20:24 2000
From: guido@python.org (Guido van Rossum)
Date: Tue, 04 Apr 2000 21:20:24 -0400
Subject: [Patches] Unicode Patch Set 2000-04-05
In-Reply-To: Your message of "Wed, 05 Apr 2000 02:53:10 +0200."
 <38EA8E76.66359797@lemburg.com>
References: <38EA8E76.66359797@lemburg.com>
Message-ID: <200004050120.VAA15817@eric.cnri.reston.va.us>

Much better, except:

> - contains now works for non-string types too; TypeErrors are
>   masked and 0 returned; all other errors are passed through

This breaks a test in test_contains.py: (None in 'abc') should raise
an TypeError, but doesn't.  I noticed that any non-string left
argument is now okay.  This is not according to spec...

--Guido van Rossum (home page: http://www.python.org/~guido/)


From nascheme@enme.ucalgary.ca  Wed Apr  5 09:50:38 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Wed, 5 Apr 2000 02:50:38 -0600
Subject: [Patches] portable garbage collection for Python
Message-ID: <20000405025037.B16978@acs.ucalgary.ca>

I have a new version of my garbage collection patch available.
If you have time please test it out and send me comments.  The
patch against the current CVS version of Python is available
from:

    http://www.enme.ucalgary.ca/~nascheme/python/gc-cycle.diff
    
I will give some more details on my approach tomorrow after I get
some rest. :)


    Neil

-- 
Real programmers don't make mistrakes


From mal@lemburg.com  Wed Apr  5 10:50:21 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Wed, 05 Apr 2000 11:50:21 +0200
Subject: [Patches] Unicode Patch Set 2000-04-05
References: <38EA8E76.66359797@lemburg.com> <200004050120.VAA15817@eric.cnri.reston.va.us>
Message-ID: <38EB0C5D.72253EE6@lemburg.com>

This is a multi-part message in MIME format.
--------------ADCA470C18D9DED714E274DB
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Guido van Rossum wrote:
> 
> Much better, except:
> 
> > - contains now works for non-string types too; TypeErrors are
> >   masked and 0 returned; all other errors are passed through
> 
> This breaks a test in test_contains.py: (None in 'abc') should raise
> an TypeError, but doesn't.  I noticed that any non-string left
> argument is now okay.  This is not according to spec...

Looks like I've gone a step too far there... (and test_contains.py
seem to have a bug too).

I've changed back to reporting all errors in PyUnicode_Contains()
and added a few more test cases to test_contains.py (plus corrected
the join() NameError).

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------ADCA470C18D9DED714E274DB
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-05.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-05.patch"

diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Include/Python.h Python+Unicode/Include/Python.h
--- CVS-Python/Include/Python.h	Sat Mar 11 10:52:34 2000
+++ Python+Unicode/Include/Python.h	Wed Apr  5 02:44:28 2000
@@ -72,6 +72,7 @@
 
 #include "pydebug.h"
 
+#include "unicodeobject.h"
 #include "intobject.h"
 #include "longobject.h"
 #include "floatobject.h"
@@ -92,7 +93,6 @@
 #include "cobject.h"
 #include "traceback.h"
 #include "sliceobject.h"
-#include "unicodeobject.h"
 
 #include "codecs.h"
 #include "pyerrors.h"
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Include/intobject.h Python+Unicode/Include/intobject.h
--- CVS-Python/Include/intobject.h	Tue Mar  7 18:58:16 2000
+++ Python+Unicode/Include/intobject.h	Wed Apr  5 02:44:28 2000
@@ -60,6 +60,7 @@
 #define PyInt_Check(op) ((op)->ob_type == &PyInt_Type)
 
 extern DL_IMPORT(PyObject *) PyInt_FromString Py_PROTO((char*, char**, int));
+extern DL_IMPORT(PyObject *) PyInt_FromUnicode Py_PROTO((Py_UNICODE*, int, int));
 extern DL_IMPORT(PyObject *) PyInt_FromLong Py_PROTO((long));
 extern DL_IMPORT(long) PyInt_AsLong Py_PROTO((PyObject *));
 extern DL_IMPORT(long) PyInt_GetMax Py_PROTO((void));
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Include/longobject.h Python+Unicode/Include/longobject.h
--- CVS-Python/Include/longobject.h	Thu Jan 14 11:24:51 1999
+++ Python+Unicode/Include/longobject.h	Wed Apr  5 02:44:28 2000
@@ -82,6 +82,7 @@
 #endif /* HAVE_LONG_LONG */
 
 DL_IMPORT(PyObject *) PyLong_FromString Py_PROTO((char *, char **, int));
+DL_IMPORT(PyObject *) PyLong_FromUnicode Py_PROTO((Py_UNICODE*, int, int));
 
 #ifdef __cplusplus
 }
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Include/unicodeobject.h Python+Unicode/Include/unicodeobject.h
--- CVS-Python/Include/unicodeobject.h	Wed Mar 29 11:50:05 2000
+++ Python+Unicode/Include/unicodeobject.h	Wed Apr  5 02:44:28 2000
@@ -358,7 +358,7 @@
 
 /* --- UTF-16 Codecs ------------------------------------------------------ */
 
-/* Decodes length bytes from a UTF-16 encoded buffer string and return
+/* Decodes length bytes from a UTF-16 encoded buffer string and returns
    the corresponding Unicode object.
 
    errors (if non-NULL) defines the error handling. It defaults
@@ -397,7 +397,7 @@
     );
 
 /* Returns a Python string object holding the UTF-16 encoded value of
-   the Unicode data in s.
+   the Unicode data.
 
    If byteorder is not 0, output is written according to the following
    byte order:
@@ -586,6 +586,37 @@
     );
 
 #endif /* MS_WIN32 */
+
+/* --- Decimal Encoder ---------------------------------------------------- */
+
+/* Takes a Unicode string holding a decimal value and writes it into
+   an output buffer using standard ASCII digit codes.
+
+   The output buffer has to provide at least length+1 bytes of storage
+   area. The output string is 0-terminated.
+
+   The encoder converts whitespace to ' ', decimal characters to their
+   corresponding ASCII digit and all other Latin-1 characters except
+   \0 as-is. Characters outside this range (Unicode ordinals 1-256)
+   are treated as errors. This includes embedded NULL bytes.
+
+   Error handling is defined by the errors argument:
+
+      NULL or "strict": raise a ValueError
+      "ignore": ignore the wrong characters (these are not copied to the
+		output buffer)
+      "replace": replaces illegal characters with '?'
+
+   Returns 0 on success, -1 on failure.
+
+*/
+
+extern DL_IMPORT(int) PyUnicode_EncodeDecimal(
+    Py_UNICODE *s,		/* Unicode buffer */
+    int length,			/* Number of Py_UNICODE chars to encode */
+    char *output,		/* Output buffer; must have size >= length */
+    const char *errors		/* error handling */
+    );
 
 /* --- Methods & Slots ----------------------------------------------------
 
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Lib/test/output/test_unicode Python+Unicode/Lib/test/output/test_unicode
--- CVS-Python/Lib/test/output/test_unicode	Wed Mar 29 11:50:15 2000
+++ Python+Unicode/Lib/test/output/test_unicode	Wed Apr  5 11:25:24 2000
@@ -3,3 +3,4 @@
 Testing Unicode contains method... done.
 Testing Unicode formatting strings... done.
 Testing builtin codecs... done.
+Testing standard mapping codecs... 0-127... 128-255... done.
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/abstract.c Python+Unicode/Objects/abstract.c
--- CVS-Python/Objects/abstract.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/abstract.c	Wed Apr  5 02:44:28 2000
@@ -726,6 +726,27 @@
 	return type_error("bad operand type for abs()");
 }
 
+/* Add a check for embedded NULL-bytes in the argument. */
+static PyObject *
+int_from_string(s, len)
+	const char *s;
+	int len;
+{
+	char *end;
+	PyObject *x;
+
+	x = PyInt_FromString((char*)s, &end, 10);
+	if (x == NULL)
+		return NULL;
+	if (end != s + len) {
+		PyErr_SetString(PyExc_ValueError,
+				"null byte in argument for int()");
+		Py_DECREF(x);
+		return NULL;
+	}
+	return x;
+}
+
 PyObject *
 PyNumber_Int(o)
 	PyObject *o;
@@ -736,69 +757,42 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyInt_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (PyString_Check(o))
-		return PyInt_FromString(PyString_AS_STRING(o), NULL, 10);
+		return int_from_string(PyString_AS_STRING(o), 
+				       PyString_GET_SIZE(o));
+	if (PyUnicode_Check(o))
+		return PyInt_FromUnicode(PyUnicode_AS_UNICODE(o),
+					 PyUnicode_GET_SIZE(o),
+					 10);
 	m = o->ob_type->tp_as_number;
 	if (m && m->nb_int)
 		return m->nb_int(o);
 	if (!PyObject_AsCharBuffer(o, &buffer, &buffer_len))
-		return PyInt_FromString((char*)buffer, NULL, 10);
+		return int_from_string((char*)buffer, buffer_len);
 
 	return type_error("object can't be converted to int");
 }
 
-/* There are two C API functions for converting a string to a long,
- * PyNumber_Long() and PyLong_FromString().  Both are used in builtin_long, 
- * reachable from Python with the built-in function long().
- *
- * The difference is this: PyNumber_Long will raise an exception when the
- * string cannot be converted to a long.  The most common situation is
- * where a float string is passed in; this raises a ValueError.
- * PyLong_FromString does not raise an exception; it silently truncates the 
- * float to an integer.
- *
- * You can see the different behavior from Python with the following:
- *
- * long('9.5')
- * => ValueError: invalid literal for long(): 9.5
- *
- * long('9.5', 10)
- * => 9L
- *
- * The first example ends up calling PyNumber_Long(), while the second one
- * calls PyLong_FromString().
- */
+/* Add a check for embedded NULL-bytes in the argument. */
 static PyObject *
 long_from_string(s, len)
 	const char *s;
 	int len;
 {
-	const char *start;
 	char *end;
 	PyObject *x;
-	char buffer[256]; /* For errors */
 
-	start = s;
-	while (*s && isspace(Py_CHARMASK(*s)))
-		s++;
 	x = PyLong_FromString((char*)s, &end, 10);
-	if (x == NULL) {
-		if (PyErr_ExceptionMatches(PyExc_ValueError))
-			goto bad;
+	if (x == NULL)
 		return NULL;
-	}
-	while (*end && isspace(Py_CHARMASK(*end)))
-		end++;
-	if (*end != '\0') {
-  bad:
-		sprintf(buffer, "invalid literal for long(): %.200s", s);
-		PyErr_SetString(PyExc_ValueError, buffer);
-		Py_XDECREF(x);
-		return NULL;
-	}
-	else if (end != start + len) {
+	if (end != s + len) {
 		PyErr_SetString(PyExc_ValueError,
 				"null byte in argument for long()");
+		Py_DECREF(x);
 		return NULL;
 	}
 	return x;
@@ -814,6 +808,10 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyLong_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (PyString_Check(o))
 		/* need to do extra error checking that PyLong_FromString() 
 		 * doesn't do.  In particular long('9.5') must raise an
@@ -821,6 +819,11 @@
 		 */
 		return long_from_string(PyString_AS_STRING(o),
 					PyString_GET_SIZE(o));
+	if (PyUnicode_Check(o))
+		/* The above check is done in PyLong_FromUnicode(). */
+		return PyLong_FromUnicode(PyUnicode_AS_UNICODE(o),
+					  PyUnicode_GET_SIZE(o),
+					  10);
 	m = o->ob_type->tp_as_number;
 	if (m && m->nb_long)
 		return m->nb_long(o);
@@ -838,6 +841,10 @@
 
 	if (o == NULL)
 		return null_error();
+	if (PyFloat_Check(o)) {
+		Py_INCREF(o);
+		return o;
+	}
 	if (!PyString_Check(o)) {
 		m = o->ob_type->tp_as_number;
 		if (m && m->nb_float)
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/floatobject.c Python+Unicode/Objects/floatobject.c
--- CVS-Python/Objects/floatobject.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/floatobject.c	Wed Apr  5 02:44:28 2000
@@ -164,6 +164,22 @@
 		s = PyString_AS_STRING(v);
 		len = PyString_GET_SIZE(v);
 	}
+	else if (PyUnicode_Check(v)) {
+		char s_buffer[256];
+
+		if (PyUnicode_GET_SIZE(v) >= sizeof(s_buffer)) {
+			PyErr_SetString(PyExc_ValueError,
+				 "float() literal too large to convert");
+			return NULL;
+		}
+		if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v), 
+					    PyUnicode_GET_SIZE(v),
+					    s_buffer, 
+					    NULL))
+			return NULL;
+		s = s_buffer;
+		len = strlen(s);
+	}
 	else if (PyObject_AsCharBuffer(v, &s, &len)) {
 		PyErr_SetString(PyExc_TypeError,
 				"float() needs a string argument");
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/intobject.c Python+Unicode/Objects/intobject.c
--- CVS-Python/Objects/intobject.c	Sat Mar 11 10:55:08 2000
+++ Python+Unicode/Objects/intobject.c	Wed Apr  5 02:44:28 2000
@@ -261,6 +261,24 @@
 	return PyInt_FromLong(x);
 }
 
+PyObject *
+PyInt_FromUnicode(s, length, base)
+	Py_UNICODE *s;
+	int length;
+	int base;
+{
+	char buffer[256];
+	
+	if (length >= sizeof(buffer)) {
+		PyErr_SetString(PyExc_ValueError,
+				"int() literal too large to convert");
+		return NULL;
+	}
+	if (PyUnicode_EncodeDecimal(s, length, buffer, NULL))
+		return NULL;
+	return PyInt_FromString(buffer, NULL, base);
+}
+
 /* Methods */
 
 /* ARGSUSED */
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/longobject.c Python+Unicode/Objects/longobject.c
--- CVS-Python/Objects/longobject.c	Wed Jan 19 18:01:30 2000
+++ Python+Unicode/Objects/longobject.c	Wed Apr  5 02:44:28 2000
@@ -724,7 +724,7 @@
 	int base;
 {
 	int sign = 1;
-	char *start;
+	char *start, *orig_str = str;
 	PyLongObject *z;
 	
 	if ((base != 0 && base < 2) || base > 36) {
@@ -772,17 +772,44 @@
 	}
 	if (z == NULL)
 		return NULL;
-	if (str == start) {
-		PyErr_SetString(PyExc_ValueError,
-				"no digits in long int constant");
-		Py_DECREF(z);
-		return NULL;
-	}
+	if (str == start)
+		goto onError;
 	if (sign < 0 && z != NULL && z->ob_size != 0)
 		z->ob_size = -(z->ob_size);
+	if (*str == 'L' || *str == 'l')
+		str++;
+	while (*str && isspace(Py_CHARMASK(*str)))
+		str++;
+	if (*str != '\0')
+		goto onError;
 	if (pend)
 		*pend = str;
 	return (PyObject *) z;
+
+ onError:
+	PyErr_Format(PyExc_ValueError, 
+		     "invalid literal for long(): %.200s", orig_str);
+	Py_XDECREF(z);
+	return NULL;
+}
+
+PyObject *
+PyLong_FromUnicode(u, length, base)
+	Py_UNICODE *u;
+	int length;
+	int base;
+{
+	char buffer[256];
+
+	if (length >= sizeof(buffer)) {
+		PyErr_SetString(PyExc_ValueError,
+				"long() literal too large to convert");
+		return NULL;
+	}
+	if (PyUnicode_EncodeDecimal(u, length, buffer, NULL))
+		return NULL;
+
+	return PyLong_FromString(buffer, NULL, base);
 }
 
 static PyLongObject *x_divrem
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/object.c Python+Unicode/Objects/object.c
--- CVS-Python/Objects/object.c	Tue Mar 28 09:19:17 2000
+++ Python+Unicode/Objects/object.c	Wed Apr  5 02:44:28 2000
@@ -347,8 +347,21 @@
 				return cmp;
 			}
 		}
-		else if (PyUnicode_Check(v) || PyUnicode_Check(w))
-			return PyUnicode_Compare(v, w);
+		else if (PyUnicode_Check(v) || PyUnicode_Check(w)) {
+			int result = PyUnicode_Compare(v, w);
+			if (result == -1 && PyErr_Occurred() && 
+			    PyErr_ExceptionMatches(PyExc_TypeError))
+				/* TypeErrors are ignored: if Unicode coercion
+				fails due to one of the arguments not
+			 	having the right type, we continue as
+				defined by the coercion protocol (see
+				above). Luckily, decoding errors are
+				reported as ValueErrors and are not masked
+				by this technique. */
+				PyErr_Clear();
+			else
+				return result;
+		}
 		else if (vtp->tp_as_number != NULL)
 			vname = "";
 		else if (wtp->tp_as_number != NULL)
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Objects/unicodeobject.c Python+Unicode/Objects/unicodeobject.c
--- CVS-Python/Objects/unicodeobject.c	Mon Apr  3 13:46:02 2000
+++ Python+Unicode/Objects/unicodeobject.c	Wed Apr  5 11:18:57 2000
@@ -329,8 +329,14 @@
 	s = PyString_AS_STRING(obj);
 	len = PyString_GET_SIZE(obj);
     }
-    else if (PyObject_AsCharBuffer(obj, &s, &len))
+    else if (PyObject_AsCharBuffer(obj, &s, &len)) {
+	/* Overwrite the error message with something more useful in
+	   case of a TypeError. */
+	if (PyErr_ExceptionMatches(PyExc_TypeError))
+	    PyErr_SetString(PyExc_TypeError,
+			    "coercing to Unicode: need string or charbuffer");
 	return NULL;
+    }
     if (len == 0) {
 	Py_INCREF(unicode_empty);
 	return (PyObject *)unicode_empty;
@@ -1923,6 +1929,60 @@
     return NULL;
 }
     
+/* --- Decimal Encoder ---------------------------------------------------- */
+
+int PyUnicode_EncodeDecimal(Py_UNICODE *s,
+			    int length,
+			    char *output,
+			    const char *errors)
+{
+    Py_UNICODE *p, *end;
+
+    if (output == NULL) {
+	PyErr_BadArgument();
+	return -1;
+    }
+
+    p = s;
+    end = s + length;
+    while (p < end) {
+	register Py_UNICODE ch = *p++;
+	int decimal;
+	
+	if (Py_UNICODE_ISSPACE(ch)) {
+	    *output++ = ' ';
+	    continue;
+	}
+	decimal = Py_UNICODE_TODECIMAL(ch);
+	if (decimal >= 0) {
+	    *output++ = '0' + decimal;
+	    continue;
+	}
+	if (0 < ch < 256) {
+	    *output++ = ch;
+	    continue;
+	}
+	/* All other characters are considered invalid */
+	if (errors == NULL || strcmp(errors, "strict") == 0) {
+	    PyErr_SetString(PyExc_ValueError,
+			    "invalid decimal Unicode string");
+	    goto onError;
+	}
+	else if (strcmp(errors, "ignore") == 0)
+	    continue;
+	else if (strcmp(errors, "replace") == 0) {
+	    *output++ = '?';
+	    continue;
+	}
+    }
+    /* 0-terminate the output string */
+    *output++ = '\0';
+    return 0;
+
+ onError:
+    return -1;
+}
+
 /* --- Helpers ------------------------------------------------------------ */
 
 static 
@@ -2811,12 +2871,14 @@
     register Py_UNICODE ch;
 
     /* Coerce the two arguments */
-    u = (PyUnicodeObject *)PyUnicode_FromObject(container);
-    if (u == NULL)
-	goto onError;
     v = (PyUnicodeObject *)PyUnicode_FromObject(element);
     if (v == NULL)
 	goto onError;
+    u = (PyUnicodeObject *)PyUnicode_FromObject(container);
+    if (u == NULL) {
+	Py_DECREF(v);
+	goto onError;
+    }
 
     /* Check v in u */
     if (PyUnicode_GET_SIZE(v) != 1) {
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Python/bltinmodule.c Python+Unicode/Python/bltinmodule.c
--- CVS-Python/Python/bltinmodule.c	Sat Mar 11 10:55:20 2000
+++ Python+Unicode/Python/bltinmodule.c	Wed Apr  5 02:44:28 2000
@@ -449,17 +449,44 @@
 	PyObject *v;
 {
 	extern double strtod Py_PROTO((const char *, char **));
-	char *s, *start, *end;
+	const char *s, *start;
+	char *end;
 	double x=0.0, y=0.0, z;
 	int got_re=0, got_im=0, done=0;
 	int digit_or_dot;
 	int sw_error=0;
 	int sign;
 	char buffer[256]; /* For errors */
+	int len;
 
-	start = s = PyString_AS_STRING(v);
+	if (PyString_Check(v)) {
+		s = PyString_AS_STRING(v);
+		len = PyString_GET_SIZE(v);
+	}
+	else if (PyUnicode_Check(v)) {
+		char s_buffer[256];
+
+		if (PyUnicode_GET_SIZE(v) >= sizeof(s_buffer)) {
+			PyErr_SetString(PyExc_ValueError,
+				 "complex() literal too large to convert");
+			return NULL;
+		}
+		if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v), 
+					    PyUnicode_GET_SIZE(v),
+					    s_buffer, 
+					    NULL))
+			return NULL;
+		s = s_buffer;
+		len = strlen(s);
+	}
+	else if (PyObject_AsCharBuffer(v, &s, &len)) {
+		PyErr_SetString(PyExc_TypeError,
+				"complex() needs a string first argument");
+		return NULL;
+	}
 
 	/* position on first nonblank */
+	start = s;
 	while (*s && isspace(Py_CHARMASK(*s)))
 		s++;
 	if (s[0] == '\0') {
@@ -475,7 +502,7 @@
 		switch (*s) {
 
 		case '\0':
-			if (s-start != PyString_GET_SIZE(v)) {
+			if (s-start != len) {
 				PyErr_SetString(
 					PyExc_ValueError,
 					"null byte in argument for complex()");
@@ -584,7 +611,7 @@
 	i = NULL;
 	if (!PyArg_ParseTuple(args, "O|O:complex", &r, &i))
 		return NULL;
-	if (PyString_Check(r))
+	if (PyString_Check(r) || PyUnicode_Check(r))
 		return complex_from_string(r);
 	if ((nbr = r->ob_type->tp_as_number) == NULL ||
 	    nbr->nb_float == NULL ||
@@ -1289,12 +1316,17 @@
 		return NULL;
 	if (base == -909)
 		return PyNumber_Int(v);
-	else if (!PyString_Check(v)) {
+	else if (PyString_Check(v))
+		return PyInt_FromString(PyString_AS_STRING(v), NULL, base);
+	else if (PyUnicode_Check(v))
+		return PyInt_FromUnicode(PyUnicode_AS_UNICODE(v),
+					 PyUnicode_GET_SIZE(v),
+					 base);
+	else {
 		PyErr_SetString(PyExc_TypeError,
 				"can't convert non-string with explicit base");
 		return NULL;
 	}
-	return PyInt_FromString(PyString_AS_STRING(v), NULL, base);
 }
 
 static char int_doc[] =
@@ -1319,12 +1351,17 @@
 		return NULL;
 	if (base == -909)
 		return PyNumber_Long(v);
-	else if (!PyString_Check(v)) {
+	else if (PyString_Check(v))
+		return PyLong_FromString(PyString_AS_STRING(v), NULL, base);
+	else if (PyUnicode_Check(v))
+		return PyLong_FromUnicode(PyUnicode_AS_UNICODE(v),
+					  PyUnicode_GET_SIZE(v),
+					  base);
+	else {
 		PyErr_SetString(PyExc_TypeError,
 				"can't convert non-string with explicit base");
 		return NULL;
 	}
-	return PyLong_FromString(PyString_AS_STRING(v), NULL, base);
 }
 
 static char long_doc[] =
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.py CVS-Python/Python/codecs.c Python+Unicode/Python/codecs.c
--- CVS-Python/Python/codecs.c	Mon Apr  3 13:46:03 2000
+++ Python+Unicode/Python/codecs.c	Wed Apr  5 02:44:28 2000
@@ -84,8 +84,11 @@
     return -1;
 }
 
+/* Convert a string to a normalized Python string: all characters are
+   converted to lower case, spaces are replaced with underscores. */
+
 static
-PyObject *lowercasestring(const char *string)
+PyObject *normalizestring(const char *string)
 {
     register int i;
     int len = strlen(string);
@@ -96,8 +99,14 @@
     if (v == NULL)
 	return NULL;
     p = PyString_AS_STRING(v);
-    for (i = 0; i < len; i++)
-	p[i] = tolower(string[i]);
+    for (i = 0; i < len; i++) {
+        register char ch = string[i];
+        if (ch == ' ')
+            ch = '-';
+        else
+            ch = tolower(ch);
+	p[i] = ch;
+    }
     return v;
 }
 
@@ -132,8 +141,10 @@
 	    goto onError;
     }
 
-    /* Convert the encoding to a lower-cased Python string */
-    v = lowercasestring(encoding);
+    /* Convert the encoding to a normalized Python string: all
+       characters are converted to lower case, spaces and hypens are
+       replaced with underscores. */
+    v = normalizestring(encoding);
     if (v == NULL)
 	goto onError;
     PyString_InternInPlace(&v);
Binary files CVS-Python/foo and Python+Unicode/foo differ
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/encodings/__init__.py Python+Unicode/Lib/encodings/__init__.py
--- CVS-Python/Lib/encodings/__init__.py	Thu Mar 23 23:58:42 2000
+++ Python+Unicode/Lib/encodings/__init__.py	Wed Apr  5 02:44:28 2000
@@ -4,8 +4,8 @@
     directory.
 
     Codec modules must have names corresponding to standard lower-case
-    encoding names. Hyphens are automatically converted to
-    underscores, e.g. 'utf-8' is looked up as module utf_8.
+    encoding names with hyphens mapped to underscores, e.g. 'utf-8' is
+    implemented by the module 'utf_8.py'.
 
     Each codec module must export the following interface:
 
@@ -40,7 +40,7 @@
         return entry
 
     # Import the module
-    modname = string.replace(encoding,'-','_')
+    modname = string.replace(encoding, '-', '_')
     modname = aliases.aliases.get(modname,modname)
     try:
         mod = __import__(modname,globals(),locals(),'*')
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/encodings/aliases.py Python+Unicode/Lib/encodings/aliases.py
--- CVS-Python/Lib/encodings/aliases.py	Mon Apr  3 13:45:48 2000
+++ Python+Unicode/Lib/encodings/aliases.py	Wed Apr  5 02:44:28 2000
@@ -54,4 +54,7 @@
     'macroman': 'mac_roman',
     'macturkish': 'mac_turkish',
 
+    # MBCS
+    'dbcs': 'mbcs',
+
 }
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/string.py Python+Unicode/Lib/string.py
--- CVS-Python/Lib/string.py	Sat Mar 11 10:52:36 2000
+++ Python+Unicode/Lib/string.py	Wed Apr  5 02:44:28 2000
@@ -196,14 +196,11 @@
     Return the floating point number represented by the string s.
 
     """
-    if type(s) == _StringType:
-        return _float(s)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+    return _float(s)
+
 
 # Convert string to integer
-def atoi(*args):
+def atoi(s , base=10):
     """atoi(s [,base]) -> int
 
     Return the integer represented by the string s in the given
@@ -214,23 +211,11 @@
     accepted.
 
     """
-    try:
-        s = args[0]
-    except IndexError:
-        raise TypeError('function requires at least 1 argument: %d given' %
-                        len(args))
-    # Don't catch type error resulting from too many arguments to int().  The
-    # error message isn't compatible but the error type is, and this function
-    # is complicated enough already.
-    if type(s) == _StringType:
-        return _apply(_int, args)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+    return _int(s, base)
 
 
 # Convert string to long integer
-def atol(*args):
+def atol(s, base=10):
     """atol(s [,base]) -> long
 
     Return the long integer represented by the string s in the
@@ -242,19 +227,7 @@
     unless base is 0.
 
     """
-    try:
-        s = args[0]
-    except IndexError:
-        raise TypeError('function requires at least 1 argument: %d given' %
-                        len(args))
-    # Don't catch type error resulting from too many arguments to long().  The
-    # error message isn't compatible but the error type is, and this function
-    # is complicated enough already.
-    if type(s) == _StringType:
-        return _apply(_long, args)
-    else:
-        raise TypeError('argument 1: expected string, %s found' %
-                        type(s).__name__)
+    return _long(s, base)
 
 
 # Left-justify a string
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_b1.py Python+Unicode/Lib/test/test_b1.py
--- CVS-Python/Lib/test/test_b1.py	Fri Mar 26 15:58:20 1999
+++ Python+Unicode/Lib/test/test_b1.py	Wed Apr  5 02:44:28 2000
@@ -95,6 +95,7 @@
 if complex(0j, 3.14) <> 3.14j: raise TestFailed, 'complex(0j, 3.14)'
 if complex(0.0, 3.14) <> 3.14j: raise TestFailed, 'complex(0.0, 3.14)'
 if complex("  3.14+J  ") <> 3.14+1j:  raise TestFailed, 'complex("  3.14+J  )"'
+if complex(u"  3.14+J  ") <> 3.14+1j:  raise TestFailed, 'complex(u"  3.14+J  )"'
 class Z:
     def __complex__(self): return 3.14j
 z = Z()
@@ -208,6 +209,9 @@
 if float(314) <> 314.0: raise TestFailed, 'float(314)'
 if float(314L) <> 314.0: raise TestFailed, 'float(314L)'
 if float("  3.14  ") <> 3.14:  raise TestFailed, 'float("  3.14  ")'
+if float(u"  3.14  ") <> 3.14:  raise TestFailed, 'float(u"  3.14  ")'
+if float(u"  \u0663.\u0661\u0664  ") <> 3.14:
+    raise TestFailed, 'float(u"  \u0663.\u0661\u0664  ")'
 
 print 'getattr'
 import sys
@@ -254,6 +258,9 @@
 if int(-3.9) <> -3: raise TestFailed, 'int(-3.9)'
 if int(3.5) <> 3: raise TestFailed, 'int(3.5)'
 if int(-3.5) <> -3: raise TestFailed, 'int(-3.5)'
+# Different base:
+if int("10",16) <> 16L: raise TestFailed, 'int("10",16)'
+if int(u"10",16) <> 16L: raise TestFailed, 'int(u"10",16)'
 # Test conversion fron strings and various anomalies
 L = [
         ('0', 0),
@@ -267,9 +274,28 @@
         ('314 ', 314),
         ('  \t\t  314  \t\t  ', 314),
         (`sys.maxint`, sys.maxint),
+        ('  1x', ValueError),
+        ('  1  ', 1),
+        ('  1\02  ', ValueError),
         ('', ValueError),
         (' ', ValueError),
         ('  \t\t  ', ValueError),
+        (u'0', 0),
+        (u'1', 1),
+        (u'9', 9),
+        (u'10', 10),
+        (u'99', 99),
+        (u'100', 100),
+        (u'314', 314),
+        (u' 314', 314),
+        (u'\u0663\u0661\u0664 ', 314),
+        (u'  \t\t  314  \t\t  ', 314),
+        (u'  1x', ValueError),
+        (u'  1  ', 1),
+        (u'  1\02  ', ValueError),
+        (u'', ValueError),
+        (u' ', ValueError),
+        (u'  \t\t  ', ValueError),
 ]
 for s, v in L:
     for sign in "", "+", "-":
@@ -349,10 +375,17 @@
 if long(-3.9) <> -3L: raise TestFailed, 'long(-3.9)'
 if long(3.5) <> 3L: raise TestFailed, 'long(3.5)'
 if long(-3.5) <> -3L: raise TestFailed, 'long(-3.5)'
+if long("-3") <> -3L: raise TestFailed, 'long("-3")'
+if long(u"-3") <> -3L: raise TestFailed, 'long(u"-3")'
+# Different base:
+if long("10",16) <> 16L: raise TestFailed, 'long("10",16)'
+if long(u"10",16) <> 16L: raise TestFailed, 'long(u"10",16)'
 # Check conversions from string (same test set as for int(), and then some)
 LL = [
         ('1' + '0'*20, 10L**20),
         ('1' + '0'*100, 10L**100),
+        (u'1' + u'0'*20, 10L**20),
+        (u'1' + u'0'*100, 10L**100),
 ]
 for s, v in L + LL:
     for sign in "", "+", "-":
@@ -363,11 +396,11 @@
                 vv = -v
             try:
                 if long(ss) != long(vv):
-                    raise TestFailed, "int(%s)" % `ss`
+                    raise TestFailed, "long(%s)" % `ss`
             except v:
                 pass
             except ValueError, e:
-                raise TestFailed, "int(%s) raised ValueError: %s" % (`ss`, e)
+                raise TestFailed, "long(%s) raised ValueError: %s" % (`ss`, e)
 
 print 'map'
 if map(None, 'hello world') <> ['h','e','l','l','o',' ','w','o','r','l','d']:
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_contains.py Python+Unicode/Lib/test/test_contains.py
--- CVS-Python/Lib/test/test_contains.py	Tue Mar  7 16:52:01 2000
+++ Python+Unicode/Lib/test/test_contains.py	Wed Apr  5 11:23:44 2000
@@ -17,7 +17,7 @@
 
 def check(ok, *args):
     if not ok:
-        raise TestFailed, join(map(str, args), " ")
+        raise TestFailed, " ".join(map(str, args))
 
 a = base_set(1)
 b = set(1)
@@ -60,5 +60,62 @@
 try:
 	None in 'abc'
 	check(0, "None in 'abc' did not raise error")
+except TypeError:
+	pass
+
+# Test char in Unicode
+
+check('c' in u'abc', "'c' not in u'abc'")
+check('d' not in u'abc', "'d' in u'abc'")
+
+try:
+	'' in u'abc'
+	check(0, "'' in u'abc' did not raise error")
+except TypeError:
+	pass
+
+try:
+	'ab' in u'abc'
+	check(0, "'ab' in u'abc' did not raise error")
+except TypeError:
+	pass
+
+try:
+	None in u'abc'
+	check(0, "None in u'abc' did not raise error")
+except TypeError:
+	pass
+
+# Test Unicode char in Unicode
+
+check(u'c' in u'abc', "u'c' not in u'abc'")
+check(u'd' not in u'abc', "u'd' in u'abc'")
+
+try:
+	u'' in u'abc'
+	check(0, "u'' in u'abc' did not raise error")
+except TypeError:
+	pass
+
+try:
+	u'ab' in u'abc'
+	check(0, "u'ab' in u'abc' did not raise error")
+except TypeError:
+	pass
+
+# Test Unicode char in string
+
+check(u'c' in 'abc', "u'c' not in 'abc'")
+check(u'd' not in 'abc', "u'd' in 'abc'")
+
+try:
+	u'' in 'abc'
+	check(0, "u'' in 'abc' did not raise error")
+except TypeError:
+	pass
+
+try:
+	u'ab' in 'abc'
+	check(0, "u'ab' in 'abc' did not raise error")
 except TypeError:
 	pass
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_unicode.py Python+Unicode/Lib/test/test_unicode.py
--- CVS-Python/Lib/test/test_unicode.py	Wed Mar 29 11:50:14 2000
+++ Python+Unicode/Lib/test/test_unicode.py	Wed Apr  5 02:44:28 2000
@@ -221,15 +221,23 @@
 
 # Contains:
 print 'Testing Unicode contains method...',
-assert ('a' in 'abdb') == 1
-assert ('a' in 'bdab') == 1
-assert ('a' in 'bdaba') == 1
-assert ('a' in 'bdba') == 1
+assert ('a' in u'abdb') == 1
+assert ('a' in u'bdab') == 1
+assert ('a' in u'bdaba') == 1
+assert ('a' in u'bdba') == 1
 assert ('a' in u'bdba') == 1
 assert (u'a' in u'bdba') == 1
 assert (u'a' in u'bdb') == 0
 assert (u'a' in 'bdb') == 0
 assert (u'a' in 'bdba') == 1
+assert (u'a' in ('a',1,None)) == 1
+assert (u'a' in (1,None,'a')) == 1
+assert (u'a' in (1,None,u'a')) == 1
+assert ('a' in ('a',1,None)) == 1
+assert ('a' in (1,None,'a')) == 1
+assert ('a' in (1,None,u'a')) == 1
+assert ('a' in ('x',1,u'y')) == 0
+assert ('a' in ('x',1,None)) == 0
 print 'done.'
 
 # Formatting:
@@ -270,11 +278,88 @@
     assert unicode(u.encode(encoding),encoding) == u
 
 u = u''.join(map(unichr, range(256)))
-for encoding in ('latin-1',):
-    assert unicode(u.encode(encoding),encoding) == u
+for encoding in (
+    'latin-1',
+    ):
+    try:
+        assert unicode(u.encode(encoding),encoding) == u
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
 
 u = u''.join(map(unichr, range(128)))
-for encoding in ('ascii',):
-    assert unicode(u.encode(encoding),encoding) == u
+for encoding in (
+    'ascii',
+    ):
+    try:
+        assert unicode(u.encode(encoding),encoding) == u
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
+
+print 'done.'
+
+print 'Testing standard mapping codecs...',
+
+print '0-127...',
+s = ''.join(map(chr, range(128)))
+for encoding in (
+    'cp037', 'cp1026',
+    'cp437', 'cp500', 'cp737', 'cp775', 'cp850',
+    'cp852', 'cp855', 'cp860', 'cp861', 'cp862',
+    'cp863', 'cp865', 'cp866', 
+    'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
+    'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
+    'mac_cyrillic', 'mac_latin2',
+
+    'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
+    'cp1256', 'cp1257', 'cp1258',
+    'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
+
+    'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
+    'cp1006', 'cp875', 'iso8859_8',
+    
+    ### These have undefined mappings:
+    #'cp424',
+    
+    ):
+    try:
+        assert unicode(s,encoding).encode(encoding) == s
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
+
+print '128-255...',
+s = ''.join(map(chr, range(128,256)))
+for encoding in (
+    'cp037', 'cp1026',
+    'cp437', 'cp500', 'cp737', 'cp775', 'cp850',
+    'cp852', 'cp855', 'cp860', 'cp861', 'cp862',
+    'cp863', 'cp865', 'cp866', 
+    'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
+    'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
+    'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
+    'mac_cyrillic', 'mac_latin2',
+    
+    ### These have undefined mappings:
+    #'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
+    #'cp1256', 'cp1257', 'cp1258',
+    #'cp424', 'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
+    #'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
+    
+    ### These fail the round-trip:
+    #'cp1006', 'cp875', 'iso8859_8',
+    
+    ):
+    try:
+        assert unicode(s,encoding).encode(encoding) == s
+    except AssertionError:
+        print '*** codec "%s" failed round-trip' % encoding
+    except ValueError,why:
+        print '*** codec for "%s" failed: %s' % (encoding, why)
 
 print 'done.'
Binary files CVS-Python/foo and Python+Unicode/foo differ

--------------ADCA470C18D9DED714E274DB--





From mhammond@skippinet.com.au  Wed Apr  5 14:54:50 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Wed, 5 Apr 2000 23:54:50 +1000
Subject: [Patches] Patch for mmap module on Win9x
Message-ID: <ECEPKNMJLHAPFFJHDOJBGECOCIAA.mhammond@skippinet.com.au>

This patch fixes the mmap module on Windows 9x.

Also updates the mmap test to remove the test file.

Release info:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that
I have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce,
distribute, perform and/or display publicly, prepare derivative
versions, and otherwise use this contribution as part of the Python
software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize
others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and
its related documentation.  I further grant CNRI permission to use
my name and other identifying information provided to CNRI by me for
use in connection with the Python software and its related
documentation.

Mark.

diff -c -3 -r2.3 mmapmodule.c
*** mmapmodule.c	2000/03/31 15:04:26	2.3
--- mmapmodule.c	2000/04/05 13:49:44
***************
*** 752,763 ****
  		    PyErr_SetFromErrno(mmap_module_error);
  		    return NULL;
  		}
! //
! //		fh = OpenFile (filename, &file_info, OF_READWRITE);
! //		if (fh == HFILE_ERROR) {
! //			PyErr_SetFromWindowsErr(GetLastError());
! //			return NULL;
! //		}
  	}

  	m_obj = PyObject_NEW (mmap_object, &mmap_object_type);
--- 752,759 ----
  		    PyErr_SetFromErrno(mmap_module_error);
  		    return NULL;
  		}
! 		/* Win9x appears to need us seeked to zero */
! 		fseek(&_iob[fileno], 0, SEEK_SET);
  	}

  	m_obj = PyObject_NEW (mmap_object, &mmap_object_type);

diff -c -3 -r1.3 test_mmap.py
*** test_mmap.py	2000/03/31 01:20:33	1.3
--- test_mmap.py	2000/04/05 13:52:04
***************
*** 59,64 ****
--- 59,66 ----
          assert start == PAGESIZE
          assert end == PAGESIZE + 6

+     m.close()
+     os.unlink("foo")
      print ' Test passed'

  test_both()




From fdrake@acm.org  Wed Apr  5 15:17:39 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Wed, 5 Apr 2000 10:17:39 -0400 (EDT)
Subject: [Patches] Patch for mmap module on Win9x
In-Reply-To: <ECEPKNMJLHAPFFJHDOJBGECOCIAA.mhammond@skippinet.com.au>
References: <ECEPKNMJLHAPFFJHDOJBGECOCIAA.mhammond@skippinet.com.au>
Message-ID: <14571.19203.167311.530879@seahag.cnri.reston.va.us>

Mark Hammond writes:
 > This patch fixes the mmap module on Windows 9x.
 > 
 > Also updates the mmap test to remove the test file.


  Thanks Mark!


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From tismer@tismer.com  Wed Apr  5 16:22:13 2000
From: tismer@tismer.com (Christian Tismer)
Date: Wed, 05 Apr 2000 17:22:13 +0200
Subject: [Patches] portable garbage collection for Python
References: <20000405025037.B16978@acs.ucalgary.ca>
Message-ID: <38EB5A25.70A03160@tismer.com>


Neil Schemenauer wrote:
> 
> I have a new version of my garbage collection patch available.
> If you have time please test it out and send me comments.  The
> patch against the current CVS version of Python is available
> from:
> 
>     http://www.enme.ucalgary.ca/~nascheme/python/gc-cycle.diff
> 
> I will give some more details on my approach tomorrow after I get
> some rest. :)

This looks really great!

One minor bug from just reading the diff:

***************
*** 215,220 ****
--- 219,227 ----
        PyListObject *op;
  {
        int i;
+ #ifdef WITH_CYCLE_GC
+       PyGC_Remove((PyObject *)op);
+ #endif
        Py_TRASHCAN_SAFE_BEGIN(op)
        if (op->ob_item != NULL) {
                /* Do it backwards, for Christian Tismer.
***************
*** 227,233 ****
                }
                free((ANY *)op->ob_item);
        }

Code must not be inserted before Py_TRASHCAN_SAFE_BEGIN,
this construct is supposed to enclose the whole destruction
process, whatever it is.

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From guido@python.org  Wed Apr  5 20:59:01 2000
From: guido@python.org (Guido van Rossum)
Date: Wed, 05 Apr 2000 15:59:01 -0400
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: Your message of "Wed, 05 Apr 0100 19:07:55 +0200."
 <20000405170755.30583.qmail@lannert.rz.uni-duesseldorf.de>
References: <20000405170755.30583.qmail@lannert.rz.uni-duesseldorf.de>
Message-ID: <200004051959.PAA18356@eric.cnri.reston.va.us>

I kind of like the chop switch.  Anybody else got an opinion?

+1

--Guido van Rossum (home page: http://www.python.org/~guido/)


From Fredrik Lundh" <effbot@telia.com  Wed Apr  5 22:13:42 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Wed, 5 Apr 2000 23:13:42 +0200
Subject: [Patches] minor unicodeobject.c match
Message-ID: <002101bf9f43$e61c41e0$34aab5d4@hagrid>

This is a multi-part message in MIME format.

------=_NextPart_000_001B_01BF9F54.95FFC5E0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

the attached patch eliminates a MSVC compiler warning.

</F>

I confirm that, to the best of my knowledge and belief, this =
contribution
is free of any claims of third parties under copyright, patent or
other rights or interests ("claims").  To the extent that I have
any such claims, I hereby grant to CNRI a nonexclusive, irrevocable,
royalty-free, worldwide license to reproduce, distribute, perform and/or
display publicly, prepare derivative versions, and otherwise use this
contribution as part of the Python software and its related =
documentation,
or any derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
        =20
I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.


------=_NextPart_000_001B_01BF9F54.95FFC5E0
Content-Type: application/octet-stream;
	name="unicodeobject-patch-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename="unicodeobject-patch-1"

--- Objects\unicodeobject.c.bak	Wed Apr 05 23:06:31 2000
+++ Objects\unicodeobject.c	Wed Apr 05 23:08:35 2000
@@ -1959,7 +1959,7 @@
 	    continue;
 	}
 	if (0 < ch < 256) {
-	    *output++ = ch;
+	    *output++ = (char) ch;
 	    continue;
 	}
 	/* All other characters are considered invalid */

------=_NextPart_000_001B_01BF9F54.95FFC5E0--



From DavidA@ActiveState.com  Wed Apr  5 23:05:04 2000
From: DavidA@ActiveState.com (David Ascher)
Date: Wed, 5 Apr 2000 15:05:04 -0700
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <200004051959.PAA18356@eric.cnri.reston.va.us>
Message-ID: <NDBBJPNCJLKKIOBLDOMJEEIHCEAA.DavidA@ActiveState.com>

> I kind of like the chop switch.  Anybody else got an opinion?
> 
> +1

+0

How about also adding it to readlines()?

--david	


From nascheme@enme.ucalgary.ca  Thu Apr  6 02:44:35 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Wed, 5 Apr 2000 19:44:35 -0600
Subject: [Patches] New builtin function repeat?
Message-ID: <20000405194435.A31381@acs.ucalgary.ca>

--G4iJoqBmSsgzjUCe
Content-Type: text/plain; charset=us-ascii

I made this module some time about but I still think it is neat.
It creates a new function called repeat (I'm not stuck on the
name).  You can use it like this:

    for line in repeat(sys.stdin.readline):
        ...

The repeat function creates a new sequence object.  The function
supplied is called for __getitem__.  By default, a false value
returned from the function will raise an IndexError.  You can
specifiy an alternate termination value.  For example:

    for line in repeat(sys.stdin.readline, ''):
        ...

would work as well.

While I'm throwing out crazy ideas, how about adding some
sequence methods to PyFileObject?  You could then do:

    file = open(...)
    for line in file:
        ...

Too weird?  Here is a patch and the module.

    Neil

-- 
"The lyf so short, the craft so long to lerne." -- Chaucer

--G4iJoqBmSsgzjUCe
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="file-seq.diff"

diff -cr Python-cvs/Objects/fileobject.c Python-file/Objects/fileobject.c
*** Python-cvs/Objects/fileobject.c	Fri Mar 24 22:32:30 2000
--- Python-file/Objects/fileobject.c	Wed Apr  5 19:37:13 2000
***************
*** 1040,1045 ****
--- 1040,1099 ----
  	return PyMember_Set((char *)f, file_memberlist, name, v);
  }
  
+ /* as_sequence */
+ 
+ static PyObject *
+ file_getitem(self, index)
+ 	PyObject *self;
+ 	int index;
+ {
+ 	PyObject *line;
+ 	if ((line = PyFile_GetLine(self, 0)) == NULL) {
+ 		return NULL;
+ 	}
+ 	if (!PyObject_IsTrue(line)) {
+ 		Py_DECREF(line);
+ 		PyErr_SetString(PyExc_IndexError, "file index out of range");
+ 		return NULL;
+ 	}
+ 	return line;
+ }
+ 
+ static int
+ file_length(self)
+ 	PyObject *self;
+ {
+ 	return 0;
+ }
+ 
+ static PyObject *
+ file_concat(self, other)
+ 	PyObject *self;
+ 	PyObject *other;
+ {
+ 	PyErr_SetString(PyExc_AttributeError, "__concat__");
+ 	return NULL;
+ }
+ 
+ static PyObject *
+ file_repeat(self, count)
+ 	PyObject *self;
+ 	int count;
+ {
+ 	PyErr_SetString(PyExc_AttributeError, "__repeat__");
+ 	return NULL;
+ }
+ 
+ static PySequenceMethods file_as_sequence = {
+ 	(inquiry)file_length, /*sq_length*/
+ 	(binaryfunc)file_concat, /*sq_concat*/
+ 	(intargfunc)file_repeat, /*sq_repeat*/
+ 	(intargfunc)file_getitem, /*sq_item*/
+ 	(intintargfunc)0, /*sq_slice*/
+ 	(intobjargproc)0, /*sq_ass_item*/
+ 	(intintobjargproc)0, /*sq_ass_slice*/
+ };
+ 
  PyTypeObject PyFile_Type = {
  	PyObject_HEAD_INIT(&PyType_Type)
  	0,
***************
*** 1052,1057 ****
--- 1106,1113 ----
  	(setattrfunc)file_setattr, /*tp_setattr*/
  	0,		/*tp_compare*/
  	(reprfunc)file_repr, /*tp_repr*/
+ 	0,		/*tp_as_number*/
+ 	&file_as_sequence,	/*tp_as_sequence*/
  };
  
  /* Interface for the 'soft space' between print items. */

--G4iJoqBmSsgzjUCe
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="repeatmodule.c"

/*
 * A little Python extension module so you can do stuff like:
 * 	
 * 	import sys
 * 	from repeat import repeat
 * 	for line in repeat(sys.stdin.readline):
 * 		...
 *
 * repeat(<function>, [<until>]) -> RepeatObject
 *
 * RepeatObjects are sequences that only support the method __getitem__.
 * <function> is called on __getitem__.  If <until> is not specified a
 * false value from <function> raises an IndexError.  If <until> is
 * specified, if the value is equal to <until> then IndexError is
 * raised.
 * 
 *
 * Neil Schemenauer <nascheme@ucalgary.ca> April, 2000
 * 
 */

#include "Python.h"

typedef struct {
	PyObject_HEAD
	PyObject	*function;
	PyObject	*until;
} repeatObject;

staticforward PyTypeObject repeat_Type;

static repeatObject *
newrepeatObject(args)
	PyObject *args;
{
	repeatObject *self;
	PyObject *function;
	PyObject *until;

	until = NULL;
	if (!PyArg_ParseTuple(args, "O|O", &function, &until))
		return NULL;
	if (!PyCallable_Check(function)) {
		PyErr_SetString(PyExc_TypeError, "callable object required");
		return NULL;
	}
	self = PyObject_NEW(repeatObject, &repeat_Type);
	if (self == NULL)
		return NULL;
	Py_INCREF(function);
	self->function = function;
	if (until != NULL)
		Py_INCREF(until);
	self->until = until;
	return self;
}

static void
repeat_dealloc(self)
	repeatObject *self;
{
	Py_XDECREF(self->function);
	Py_XDECREF(self->until);
	PyMem_DEL(self);
}

static PyObject *
repeat_getitem(self, index)
	repeatObject *self;
	int index;
{
	PyObject *item;

	item = PyEval_CallObject(self->function, NULL);
	if ( item == NULL ) {
		return NULL;
	}
	if ( (self->until && PyObject_Compare(item, self->until) == 0) ||
			(!self->until && !PyObject_IsTrue(item)) ) {
		PyErr_SetString(PyExc_IndexError, "repeat completed");
		return NULL;
	}
	return item;
}

static int
repeat_length(self)
    PyObject *self;
{
    return 0;
}

static PyObject *
repeat_concat(self, other)
    PyObject *self;
    PyObject *other;
{
    PyErr_SetString(PyExc_AttributeError, "__concat__");
    return NULL;
}

static PyObject *
repeat_repeat(self, count)
    PyObject *self;
    int count;
{
    PyErr_SetString(PyExc_AttributeError, "__repeat__");
    return NULL;
}

static PyMethodDef repeat_methods[] = {
	{NULL,		NULL}		/* sentinel */
};

static PySequenceMethods repeat_as_sequence = {
	(inquiry)repeat_length, /*sq_length*/
	(binaryfunc)repeat_concat, /*sq_concat*/
	(intargfunc)repeat_repeat, /*sq_repeat*/
	(intargfunc)repeat_getitem, /*sq_item*/
	(intintargfunc)0, /*sq_slice*/
	(intobjargproc)0, /*sq_ass_item*/
	(intintobjargproc)0, /*sq_ass_slice*/
};

staticforward PyTypeObject repeat_Type = {
	PyObject_HEAD_INIT(NULL)
	0,			/*ob_size*/
	"repeat",		/*tp_name*/
	sizeof(repeatObject),	/*tp_basicsize*/
	0,			/*tp_itemsize*/
	/* methods */
	(destructor)repeat_dealloc, /*tp_dealloc*/
	0,			/*tp_print*/
	(getattrfunc)0, 	/*tp_getattr*/
	0,			/*tp_setattr*/
	0,			/*tp_compare*/
	0,			/*tp_repr*/
	0,			/*tp_as_number*/
	&repeat_as_sequence,	/*tp_as_sequence*/
	0,			/*tp_as_mapping*/
	0,			/*tp_hash*/
};

static PyObject *
repeat_new(self, args)
	PyObject *self; /* Not used */
	PyObject *args;
{
	repeatObject *rv;
	
	rv = newrepeatObject(args);
	if ( rv == NULL )
	    return NULL;
	return (PyObject *)rv;
}

/* List of functions defined in the module */
static PyMethodDef repeatmodule_methods[] = {
	{"repeat",		repeat_new,		1},
	{NULL,		NULL}		/* sentinel */
};


/* Initialization function for the module (*must* be called initrepeat) */
void
initrepeat()
{
	PyObject *m, *d;

	/* Create the module and add the functions */
	repeat_Type.ob_type = &PyType_Type;
	m = Py_InitModule("repeat", repeatmodule_methods);

	/* Add some symbolic constants to the module */
	d = PyModule_GetDict(m);
}

--G4iJoqBmSsgzjUCe--


From mhammond@skippinet.com.au  Thu Apr  6 02:45:59 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Thu, 6 Apr 2000 11:45:59 +1000
Subject: [Patches] zlib build problems.
Message-ID: <ECEPKNMJLHAPFFJHDOJBEEEACIAA.mhammond@skippinet.com.au>

Im trying to build the zlib module I came across some wierdness.  If
the problem is as simple as I think it is, then patches is
appropriate for the discussion :-)

Im building zlib itself from sources, rather than using the
binaries.  Im using the static zlib library.  This combination
produced linker errors.

The following patch made everything work:

diff -c -2 -r2.30 zlibmodule.c
*** zlibmodule.c	2000/02/29 13:59:24	2.30
--- zlibmodule.c	2000/04/06 01:36:18
***************
*** 5,9 ****
  #include "Python.h"
  #ifdef MS_WIN32
- #define ZLIB_DLL
  #endif
  #include "zlib.h"
--- 5,8 ----

ie, we do not want to define ZLIB_DLL, as we are using the static
ZLIB library.

[I assume we _are_ trying to use the static lib - checkin of
revision 1.6 says we are :-]

So, the question I had was "how did this work for Guido et al"?

It appears that zlib.dsp is set to point to inconsistent
directories.  The include files come from "..\..\zlib113", while the
.lib file comes from "..\..\zlib113dll\static32\zlibstat.lib" - note
that these are different directory structures.  That is the only
reason I could come up with??

I have not added a patch for the .dsp, but it does appear this needs
to be changed, so that the .h and .lib files come from the same
branch.  Further, once I applied this change, I added a
"/nod:msvcrt" option to the Debug linker switch (not needed on the
release build).  This silences a linker warning.

Release info:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that
I have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce,
distribute, perform and/or display publicly, prepare derivative
versions, and otherwise use this contribution as part of the Python
software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize
others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and
its related documentation.  I further grant CNRI permission to use
my name and other identifying information provided to CNRI by me for
use in connection with the Python software and its related
documentation.

Mark.




From guido@python.org  Thu Apr  6 14:21:23 2000
From: guido@python.org (Guido van Rossum)
Date: Thu, 06 Apr 2000 09:21:23 -0400
Subject: [Patches] zlib build problems.
In-Reply-To: Your message of "Thu, 06 Apr 2000 11:45:59 +1000."
 <ECEPKNMJLHAPFFJHDOJBEEEACIAA.mhammond@skippinet.com.au>
References: <ECEPKNMJLHAPFFJHDOJBEEEACIAA.mhammond@skippinet.com.au>
Message-ID: <200004061321.JAA23839@eric.cnri.reston.va.us>

> Im trying to build the zlib module I came across some wierdness.  If
> the problem is as simple as I think it is, then patches is
> appropriate for the discussion :-)
> 
> Im building zlib itself from sources, rather than using the
> binaries.  Im using the static zlib library.  This combination
> produced linker errors.
> 
> The following patch made everything work:
> 
> diff -c -2 -r2.30 zlibmodule.c
> *** zlibmodule.c	2000/02/29 13:59:24	2.30
> --- zlibmodule.c	2000/04/06 01:36:18
> ***************
> *** 5,9 ****
>   #include "Python.h"
>   #ifdef MS_WIN32
> - #define ZLIB_DLL
>   #endif
>   #include "zlib.h"
> --- 5,8 ----

I got rid of the #ifdef MS_WIN32 ... #endif as well :-)

> ie, we do not want to define ZLIB_DLL, as we are using the static
> ZLIB library.
> 
> [I assume we _are_ trying to use the static lib - checkin of
> revision 1.6 says we are :-]

Yes we are.  Using the DLL version is madness -- too many incomaptible
zlib.dll files around.

> So, the question I had was "how did this work for Guido et al"?

Beats me.

> It appears that zlib.dsp is set to point to inconsistent
> directories.  The include files come from "..\..\zlib113", while the
> .lib file comes from "..\..\zlib113dll\static32\zlibstat.lib" - note
> that these are different directory structures.  That is the only
> reason I could come up with??

I doubt it.  The zlib113dll directory *only* contains binaries (.dll
and .lib files).  The zlib113 directory contains the sources from
which these binaries apparently were built.  Both were downloaded from
the websites mentioned in the top of the zlibmodule.c source file, and
extracted in the directory names suggested by the archive names.
Just to be sure I downloaded them again; the build goes fine.  I
decided not to change the extraction directory.

> I have not added a patch for the .dsp, but it does appear this needs
> to be changed, so that the .h and .lib files come from the same
> branch.  Further, once I applied this change, I added a
> "/nod:msvcrt" option to the Debug linker switch (not needed on the
> release build).  This silences a linker warning.

I've added that one to make you happy.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Thu Apr  6 14:26:42 2000
From: guido@python.org (Guido van Rossum)
Date: Thu, 06 Apr 2000 09:26:42 -0400
Subject: [Patches] New builtin function repeat?
In-Reply-To: Your message of "Wed, 05 Apr 2000 19:44:35 MDT."
 <20000405194435.A31381@acs.ucalgary.ca>
References: <20000405194435.A31381@acs.ucalgary.ca>
Message-ID: <200004061326.JAA23867@eric.cnri.reston.va.us>

> I made this module some time about but I still think it is neat.
> It creates a new function called repeat (I'm not stuck on the
> name).  You can use it like this:
> 
>     for line in repeat(sys.stdin.readline):
>         ...
> 
> The repeat function creates a new sequence object.  The function
> supplied is called for __getitem__.  By default, a false value
> returned from the function will raise an IndexError.  You can
> specifiy an alternate termination value.  For example:
> 
>     for line in repeat(sys.stdin.readline, ''):
>         ...
> 
> would work as well.
> 
> While I'm throwing out crazy ideas, how about adding some
> sequence methods to PyFileObject?  You could then do:
> 
>     file = open(...)
>     for line in file:
>         ...
> 
> Too weird?  Here is a patch and the module.

All Python 1.7 or Py3K material...  To be discussed on python-dev.

I expect to redesign the way for loops interact with sequences; there
will be a separate protocol, probably a method to extract a generator
from an arbitrary object, where the generator has first/next methods.
Then "for line in list" will create a list generator which keeps track
of the loop index, while "for line in file" will create a file
generator which leaves the state up to the file object.  Some b/w
compat issues arise.

--Guido van Rossum (home page: http://www.python.org/~guido/)



From guido@python.org  Thu Apr  6 20:07:02 2000
From: guido@python.org (Guido van Rossum)
Date: Thu, 06 Apr 2000 15:07:02 -0400
Subject: [Patches] zlib build problems.
In-Reply-To: Your message of "Thu, 06 Apr 2000 09:21:23 EDT."
 <200004061321.JAA23839@eric.cnri.reston.va.us>
References: <ECEPKNMJLHAPFFJHDOJBEEEACIAA.mhammond@skippinet.com.au>
 <200004061321.JAA23839@eric.cnri.reston.va.us>
Message-ID: <200004061907.PAA25761@eric.cnri.reston.va.us>

> > The following patch made everything work:
> > 
> > diff -c -2 -r2.30 zlibmodule.c
> > *** zlibmodule.c	2000/02/29 13:59:24	2.30
> > --- zlibmodule.c	2000/04/06 01:36:18
> > ***************
> > *** 5,9 ****
> >   #include "Python.h"
> >   #ifdef MS_WIN32
> > - #define ZLIB_DLL
> >   #endif
> >   #include "zlib.h"
> > --- 5,8 ----
> 
> I got rid of the #ifdef MS_WIN32 ... #endif as well :-)

Strange...  This suddenly broke for me.  I can now only get it to
compile by define ZLIB_DLL again -- but I'm not using the DLL!

I can revert the checkin, or I can define the symbol in the setup
dialog.  What would you prefer?

--Guido van Rossum (home page: http://www.python.org/~guido/)


From mhammond@skippinet.com.au  Fri Apr  7 00:11:24 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Fri, 7 Apr 2000 09:11:24 +1000
Subject: [Patches] zlib build problems.
In-Reply-To: <200004061907.PAA25761@eric.cnri.reston.va.us>
Message-ID: <ECEPKNMJLHAPFFJHDOJBAEFKCIAA.mhammond@skippinet.com.au>

> > I got rid of the #ifdef MS_WIN32 ... #endif as well :-)
>
> Strange...  This suddenly broke for me.  I can now only get it to
> compile by define ZLIB_DLL again -- but I'm not using the DLL!
>
> I can revert the checkin, or I can define the symbol in the setup
> dialog.  What would you prefer?

Very strange - I wonder if changing the referenced directories to
the .h and the .lib come from the same tree solve it?

Otherwise - In the setup dialog - I need to change this project
locally to reflect the local path to zlib, so then I can also change
the define - ie, if the change is in the .dsp, I can make it work
without hacking the source code.

Thanks,

Mark.




From Vladimir.Marangozov@inrialpes.fr  Fri Apr  7 01:16:31 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Fri, 7 Apr 2000 02:16:31 +0200 (CEST)
Subject: [Patches] code_repr lineno fix for -O
Message-ID: <200004070016.CAA21719@python.inrialpes.fr>

This fixes the line number in the string representation of code objects
when optimization is on (python -O). It was always reported as -1
instead of the real lineno.

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--
I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------  
*** Python/compile.c-1.6a1	Fri Apr  7 01:56:18 2000
--- Python/compile.c	Fri Apr  7 02:06:44 2000
***************
*** 121,133 ****
  {
  	char buf[500];
  	int lineno = -1;
- 	unsigned char *p;
  	char *filename = "???";
  	char *name = "???";
  
! 	_PyCode_GETCODEPTR(co, &p);
! 	if (*p == SET_LINENO)
! 		lineno = (p[1] & 0xff) | ((p[2] & 0xff) << 8);
  	if (co->co_filename && PyString_Check(co->co_filename))
  		filename = PyString_AsString(co->co_filename);
  	if (co->co_name && PyString_Check(co->co_name))
--- 121,131 ----
  {
  	char buf[500];
  	int lineno = -1;
  	char *filename = "???";
  	char *name = "???";
  
! 	if (co->co_firstlineno != 0)
! 		lineno = co->co_firstlineno;
  	if (co->co_filename && PyString_Check(co->co_filename))
  		filename = PyString_AsString(co->co_filename);
  	if (co->co_name && PyString_Check(co->co_name))


From gansevle@cs.utwente.nl  Fri Apr  7 14:55:15 2000
From: gansevle@cs.utwente.nl (Fred Gansevles)
Date: Fri, 07 Apr 2000 15:55:15 +0200
Subject: [Patches] (no subject)
Message-ID: <200004071355.PAA23203@localhost.localdomain>

The copytree function doesn't pass the symlinks parameter in recursicve
calls
----------------------------------------------------------------
I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part

of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
FAQ about legal
I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

----------------------------------------------------------------
Index: shutil.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/shutil.py,v
retrieving revision 1.15
diff -c -r1.15 shutil.py
*** shutil.py	2000/02/04 15:28:41	1.15
--- shutil.py	2000/04/07 13:48:42
***************
*** 88,94 ****
                  linkto = os.readlink(srcname)
                  os.symlink(linkto, dstname)
              elif os.path.isdir(srcname):
!                 copytree(srcname, dstname)
              else:
                  copy2(srcname, dstname)
              # XXX What about devices, sockets etc.?
--- 88,94 ----
                  linkto = os.readlink(srcname)
                  os.symlink(linkto, dstname)
              elif os.path.isdir(srcname):
!                 copytree(srcname, dstname, symlinks)
              else:
                  copy2(srcname, dstname)
              # XXX What about devices, sockets etc.?


From fdrake@acm.org  Fri Apr  7 15:35:17 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Fri, 7 Apr 2000 10:35:17 -0400 (EDT)
Subject: [Patches] (no subject)
In-Reply-To: <200004071355.PAA23203@localhost.localdomain>
References: <200004071355.PAA23203@localhost.localdomain>
Message-ID: <14573.61989.402432.716689@seahag.cnri.reston.va.us>

Fred Gansevles writes:
 > The copytree function doesn't pass the symlinks parameter in recursicve
 > calls

Fred,
  I've applied your fix; thanks!


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From tismer@tismer.com  Fri Apr  7 20:19:26 2000
From: tismer@tismer.com (Christian Tismer)
Date: Fri, 07 Apr 2000 21:19:26 +0200
Subject: [Patches] Long Multiplication is not commutative.
References: <38ED0291.65CB7B5F@tismer.com>
Message-ID: <38EE34BE.6C452AD3@tismer.com>

> The following factorial loops differ by a remarkable
> factor of 1.8, and we can gain this speed by changing
> long_mult to always put the lower multiplicand into the left.
> 
> This was reported to me by Lenny Kneler, who thought he had
> found a Stackless bug, but he was actually testing long math. :-)
> 
> This buddy...
> 
> >>> def ifact3(n) :
> ...  p = 1L
> ...  for i in range(1,n+1) :
> ...    p = i*p
> ...  return p
> 
> performs better by a factor of 1.8 than this one:
> 
> >>> def ifact1(n) :
> ...  p = 1L
> ...  for i in range(1,n+1) :
> ...    p = p*i
> ...  return p

Checkin-message:
long multiplication is more efficient if the smaller
argument is at the left side. This patch swaps the
arguments accordingly. Thanks to Lenny Kneler who
let me find this out by chance.

Changed files:
longobject.c  test_long.py

Diff:
*** //d/cvsroot/python-pu/python/dist/src/lib/test/test_long.py	Thu Dec 23
15:36:42 1999
--- //d/python/python-1.5.2/lib/test/test_long.py	Fri Apr 07 20:55:38 2000
***************
*** 77,82 ****
--- 77,84 ----
  def test_division_2(x, y):
      q, r = divmod(x, y)
      q2, r2 = x/y, x%y
+     pab, pba = x*y, y*x
+     check(pab == pba, "multiplication does not commute for", x, y)
      check(q == q2, "divmod returns different quotient than / for", x, y)
      check(r == r2, "divmod returns different mod than % for", x, y)
      check(x == q*y + r, "x != q*y + r after divmod on", x, y)
*** //d/cvsroot/python-pu/python/dist/src/objects/longobject.c	Thu Dec 23
15:41:28 1999
--- //d/python/python-1.5.2/objects/longobject.c	Fri Apr 07 20:59:16 2000
***************
*** 1196,1201 ****
--- 1196,1210 ----
  	
  	size_a = ABS(a->ob_size);
  	size_b = ABS(b->ob_size);
+ 	if (size_a > size_b) {
+ 		/* we are faster with the small object on the left */
+ 		int hold_sa = size_a;
+ 		PyLongObject *hold_a = a;
+ 		size_a = size_b;
+ 		size_b = hold_sa;
+ 		a = b;
+ 		b = hold_a;
+ 	}
  	z = _PyLong_New(size_a + size_b);
  	if (z == NULL)
  		return NULL;
%%%%%%%%%%%%%%%%%%%%%%% end patch %%%%%%%%%%%%%%%%%%%%%%%%%%
Disclaimer:
I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From gward@mems-exchange.org  Fri Apr  7 23:17:33 2000
From: gward@mems-exchange.org (Greg Ward)
Date: Fri, 7 Apr 2000 18:17:33 -0400
Subject: [Patches] Fix main Makefile to create prefix, exec_prefix
Message-ID: <20000407181732.A11360@mems-exchange.org>

This patch fixes the top Python Makefile so it creates $(prefix) or
$(exec_prefix) before attempting to create directories under them.
Handy when you use prefixes like "/usr/local/python-1.5.2" and
"/usr/local/python-1.6a1", which don't exist until you install Python
into them.

diff -c -r1.85 Makefile.in
*** Makefile.in	2000/03/31 15:13:17	1.85
--- Makefile.in	2000/04/07 22:14:10
***************
*** 257,263 ****
  # Install the interpreter with $(VERSION) affixed
  # This goes into $(exec_prefix)
  altbininstall:	python$(EXE)
! 		@for i in $(BINDIR); \
  		do \
  			if test ! -d $$i; then \
  				echo "Creating directory $$i"; \
--- 257,263 ----
  # Install the interpreter with $(VERSION) affixed
  # This goes into $(exec_prefix)
  altbininstall:	python$(EXE)
! 		@for i in $(exec_prefix) $(BINDIR); \
  		do \
  			if test ! -d $$i; then \
  				echo "Creating directory $$i"; \
***************
*** 274,280 ****
  
  # Install the manual page
  maninstall:
! 		@for i in $(MANDIR) $(MANDIR)/man1; \
  		do \
  			if test ! -d $$i; then \
  				echo "Creating directory $$i"; \
--- 274,280 ----
  
  # Install the manual page
  maninstall:
! 		@for i in $(prefix) $(MANDIR) $(MANDIR)/man1; \
  		do \
  			if test ! -d $$i; then \
  				echo "Creating directory $$i"; \
***************
*** 292,298 ****
  LIBSUBDIRS=	lib-old lib-tk test test/output encodings \
  		distutils distutils/command $(MACHDEPS)
  libinstall:	python $(srcdir)/Lib/$(PLATDIR)
! 		@for i in $(SCRIPTDIR) $(LIBDEST); \
  		do \
  			if test ! -d $$i; then \
  				echo "Creating directory $$i"; \
--- 292,298 ----
  LIBSUBDIRS=	lib-old lib-tk test test/output encodings \
  		distutils distutils/command $(MACHDEPS)
  libinstall:	python $(srcdir)/Lib/$(PLATDIR)
! 		@for i in $(prefix) $(SCRIPTDIR) $(LIBDEST); \
  		do \
  			if test ! -d $$i; then \
  				echo "Creating directory $$i"; \


-- 
Greg Ward - software developer                gward@mems-exchange.org
MEMS Exchange / CNRI                           voice: +1-703-262-5376
Reston, Virginia, USA                            fax: +1-703-262-5367


From fdrake@acm.org  Sat Apr  8 06:05:20 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Sat, 8 Apr 2000 01:05:20 -0400 (EDT)
Subject: [Patches] Fix main Makefile to create prefix, exec_prefix
In-Reply-To: <20000407181732.A11360@mems-exchange.org>
References: <20000407181732.A11360@mems-exchange.org>
Message-ID: <14574.48656.133609.387194@seahag.cnri.reston.va.us>

Greg Ward writes:
 > This patch fixes the top Python Makefile so it creates $(prefix) or
 > $(exec_prefix) before attempting to create directories under them.
 > Handy when you use prefixes like "/usr/local/python-1.5.2" and
 > "/usr/local/python-1.6a1", which don't exist until you install Python

  I say check it in.  If Guido disagrees he can yell/email about it
later; this is quite reasonable given the recent discussion of "stow"
and the need of developers to keep multiple versions (including patch
levels!) installed.


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From Vladimir.Marangozov@inrialpes.fr  Sat Apr  8 12:05:39 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sat, 8 Apr 2000 13:05:39 +0200 (CEST)
Subject: [Patches] code_compare
Message-ID: <200004081105.NAA28779@python.inrialpes.fr>

Add the co_name field comparison to differentiate named code objects,
even when their code is the same. This caused side effects for -O
because of constant folding.

A side note:  anonymous code objects (co_name == "?") would still be
subject to constant folding optimization.

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--
I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------  
*** Python/compile.c-orig	Sat Apr  8 11:39:23 2000
--- Python/compile.c	Sat Apr  8 11:39:28 2000
***************
*** 153,158 ****
--- 153,160 ----
  	cmp = PyObject_Compare(co->co_names, cp->co_names);
  	if (cmp) return cmp;
  	cmp = PyObject_Compare(co->co_varnames, cp->co_varnames);
+ 	if (cmp) return cmp;
+ 	cmp = PyObject_Compare(co->co_name, cp->co_name);
  	return cmp;
  }
  


From gstein@lyra.org  Sat Apr  8 13:05:30 2000
From: gstein@lyra.org (Greg Stein)
Date: Sat, 8 Apr 2000 05:05:30 -0700 (PDT)
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <200004051959.PAA18356@eric.cnri.reston.va.us>
Message-ID: <Pine.LNX.4.10.10004080502410.4140-100000@nebula.lyra.org>

On Wed, 5 Apr 2000, Guido van Rossum wrote:
> I kind of like the chop switch.  Anybody else got an opinion?
> 
> +1

+1 here, too. I usually find it a pain to worry about that darn newline.

I might go one further, and ask for a way to *strip* the line, rather than
just take the newline off the end. I can't think of a clean way to do so.
For example:

  fileinput.input()
  fileinput.input(strip=fileinput.STRIP_NL)
  fileinput.input(strip=fileinput.STRIP_ALL)

It just seems a bit wordy, especially compared to the "chop=1" thing.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/




From guido@python.org  Sat Apr  8 16:40:06 2000
From: guido@python.org (Guido van Rossum)
Date: Sat, 08 Apr 2000 11:40:06 -0400
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: Your message of "Sat, 08 Apr 2000 05:05:30 PDT."
 <Pine.LNX.4.10.10004080502410.4140-100000@nebula.lyra.org>
References: <Pine.LNX.4.10.10004080502410.4140-100000@nebula.lyra.org>
Message-ID: <200004081540.LAA28124@eric.cnri.reston.va.us>

> On Wed, 5 Apr 2000, Guido van Rossum wrote:
> > I kind of like the chop switch.  Anybody else got an opinion?
> > 
> > +1

[Greg Stein]
> +1 here, too. I usually find it a pain to worry about that darn newline.
> 
> I might go one further, and ask for a way to *strip* the line, rather than
> just take the newline off the end. I can't think of a clean way to do so.
> For example:
> 
>   fileinput.input()
>   fileinput.input(strip=fileinput.STRIP_NL)
>   fileinput.input(strip=fileinput.STRIP_ALL)
> 
> It just seems a bit wordy, especially compared to the "chop=1" thing.

fileinput.input(strip=1)?

But I think this is going too far.  Next we get

fileinput.input(lower=1)

fileinput.input(split=1)

fileinput.input(split=1, sep=':')

fileinput.input(split=1, sep=':', maxplit=3)

fileinput.input(strip=1, capitalize=1, expandtabs=1)

fileinput.input(startswith="#")

fileinput.input(find="spam")

... :-)

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Sat Apr  8 16:41:45 2000
From: guido@python.org (Guido van Rossum)
Date: Sat, 08 Apr 2000 11:41:45 -0400
Subject: [Patches] code_compare
In-Reply-To: Your message of "Sat, 08 Apr 2000 13:05:39 +0200."
 <200004081105.NAA28779@python.inrialpes.fr>
References: <200004081105.NAA28779@python.inrialpes.fr>
Message-ID: <200004081541.LAA28136@eric.cnri.reston.va.us>

> *** Python/compile.c-orig	Sat Apr  8 11:39:23 2000
> --- Python/compile.c	Sat Apr  8 11:39:28 2000
> ***************
> *** 153,158 ****
> --- 153,160 ----
>   	cmp = PyObject_Compare(co->co_names, cp->co_names);
>   	if (cmp) return cmp;
>   	cmp = PyObject_Compare(co->co_varnames, cp->co_varnames);
> + 	if (cmp) return cmp;
> + 	cmp = PyObject_Compare(co->co_name, cp->co_name);
>   	return cmp;
>   }

If you compare the name first, sorting a list of code objects would
sort them by name first -- whgich is kind of nice.

Also, strictly speaking, the hash function should be updated.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From Vladimir.Marangozov@inrialpes.fr  Sat Apr  8 17:18:55 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sat, 8 Apr 2000 18:18:55 +0200 (CEST)
Subject: [Patches] code_compare
In-Reply-To: <200004081541.LAA28136@eric.cnri.reston.va.us> from "Guido van Rossum" at Apr 08, 2000 11:41:45 AM
Message-ID: <200004081618.SAA01225@python.inrialpes.fr>

Guido van Rossum wrote:
> 
> If you compare the name first, sorting a list of code objects would
> sort them by name first -- whgich is kind of nice.
> 
> Also, strictly speaking, the hash function should be updated.
> 

Right -- done.

--

Add the co_name field comparison to differentiate named code objects,
even when their code is the same. This caused side effects for -O
because of constant folding, like this one:

>>> class C:
...     def foo(self): pass
...     def bar(self): pass    
... 
>>> C.bar
<unbound method C.foo>
>>> 

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--
I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------  
*** Python/compile.c-orig	Sat Apr  8 11:39:23 2000
--- Python/compile.c	Sat Apr  8 18:06:10 2000
***************
*** 140,145 ****
--- 140,147 ----
  	PyCodeObject *co, *cp;
  {
  	int cmp;
+ 	cmp = PyObject_Compare(co->co_name, cp->co_name);
+ 	if (cmp) return cmp;
  	cmp = co->co_argcount - cp->co_argcount;
  	if (cmp) return cmp;
  	cmp = co->co_nlocals - cp->co_nlocals;
***************
*** 160,175 ****
  code_hash(co)
  	PyCodeObject *co;
  {
! 	long h, h1, h2, h3, h4;
! 	h1 = PyObject_Hash(co->co_code);
  	if (h1 == -1) return -1;
! 	h2 = PyObject_Hash(co->co_consts);
  	if (h2 == -1) return -1;
! 	h3 = PyObject_Hash(co->co_names);
  	if (h3 == -1) return -1;
! 	h4 = PyObject_Hash(co->co_varnames);
  	if (h4 == -1) return -1;
! 	h = h1 ^ h2 ^ h3 ^ h4 ^
  		co->co_argcount ^ co->co_nlocals ^ co->co_flags;
  	if (h == -1) h = -2;
  	return h;
--- 162,179 ----
  code_hash(co)
  	PyCodeObject *co;
  {
! 	long h, h1, h2, h3, h4, h5;
! 	h1 = PyObject_Hash(co->co_name);
  	if (h1 == -1) return -1;
! 	h2 = PyObject_Hash(co->co_code);
  	if (h2 == -1) return -1;
! 	h3 = PyObject_Hash(co->co_consts);
  	if (h3 == -1) return -1;
! 	h4 = PyObject_Hash(co->co_names);
  	if (h4 == -1) return -1;
! 	h5 = PyObject_Hash(co->co_varnames);
! 	if (h5 == -1) return -1;
! 	h = h1 ^ h2 ^ h3 ^ h4 ^ h5 ^
  		co->co_argcount ^ co->co_nlocals ^ co->co_flags;
  	if (h == -1) h = -2;
  	return h;


From Moshe Zadka <mzadka@geocities.com>  Sat Apr  8 18:16:43 2000
From: Moshe Zadka <mzadka@geocities.com> (Moshe Zadka)
Date: Sat, 8 Apr 2000 19:16:43 +0200 (IST)
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <200004081540.LAA28124@eric.cnri.reston.va.us>
Message-ID: <Pine.GSO.4.10.10004081911470.6478-100000@sundial>

On Sat, 8 Apr 2000, Guido van Rossum wrote:

> fileinput.input(strip=1)?
> 
> But I think this is going too far.  Next we get
> 
> fileinput.input(lower=1)
> 
> fileinput.input(split=1)
> 
> fileinput.input(split=1, sep=':')
> 
> fileinput.input(split=1, sep=':', maxplit=3)
> 
> fileinput.input(strip=1, capitalize=1, expandtabs=1)
> 
> fileinput.input(startswith="#")
> 
> fileinput.input(find="spam")

Hmmmm....I'm sure Python is expressive enough to solve this problem
in a nice way?

How about having the FileInput class documentedy subclassable, and 
make available some nice mixins? I'm sure such a scheme could work.
I'll have to think about it some more -- I'm hoping to come up with
something soon.

too-much-time-without-a-cute-design-problem-ly y'rs, Z.
--
Moshe Zadka <mzadka@geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com



From skip@mojam.com (Skip Montanaro)  Sun Apr  9 14:57:59 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Sun, 9 Apr 2000 08:57:59 -0500 (CDT)
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <200004081540.LAA28124@eric.cnri.reston.va.us>
References: <Pine.LNX.4.10.10004080502410.4140-100000@nebula.lyra.org>
 <200004081540.LAA28124@eric.cnri.reston.va.us>
Message-ID: <14576.35943.88104.517797@beluga.mojam.com>

    Greg> fileinput.input(strip=fileinput.STRIP_NL)
    Greg> fileinput.input(strip=fileinput.STRIP_ALL)

    Guido> fileinput.input(strip=1)?

    Guido> But I think this is going too far.

Agreed.  "chop" is a special case where you have to detect different line
termination cases on a platform-specific basis.  Plain old line stripping is 
handled just fine by existing functions.

Speaking of chop, If I am on a platform X box (X in Unix, Windows, Mac, Be)
and try to fileinput.input(chop=1) a file from a different box, will it do
the right thing (e.g. recognize ^M as a line termination character on a Unix 
or Windows machine)?

-- 
Skip Montanaro | http://www.mojam.com/
skip@mojam.com | http://www.musi-cal.com/


From guido@python.org  Sun Apr  9 23:18:10 2000
From: guido@python.org (Guido van Rossum)
Date: Sun, 09 Apr 2000 18:18:10 -0400
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: Your message of "Sun, 09 Apr 2000 08:57:59 CDT."
 <14576.35943.88104.517797@beluga.mojam.com>
References: <Pine.LNX.4.10.10004080502410.4140-100000@nebula.lyra.org> <200004081540.LAA28124@eric.cnri.reston.va.us>
 <14576.35943.88104.517797@beluga.mojam.com>
Message-ID: <200004092218.SAA28627@eric.cnri.reston.va.us>

> Agreed.  "chop" is a special case where you have to detect different line
> termination cases on a platform-specific basis.  Plain old line stripping is 
> handled just fine by existing functions.

Not really.  Chop is special because it doesn't exist as a built-in.
What we really want is Perl's chomp, which can be coded in Python as

def chomp(s):
    if s[-1:] == '\n': return s[:-1]
    else: return s

Perl's chop() is like s[:-1], and steals the last character of the
file if it is not a newline.

The CR-LF issue is taken care of at a different level:

> Speaking of chop, If I am on a platform X box (X in Unix, Windows, Mac, Be)
> and try to fileinput.input(chop=1) a file from a different box, will it do
> the right thing (e.g. recognize ^M as a line termination character on a Unix 
> or Windows machine)?

Currently, it won't.  A good change for Python would be for files
opened for reading in text mode (on all platforms!) to use Java's
algorithm (line terminators are either CR, LF or CRLF) but to
translate all three kinds of terminator into \n, so that Python code
can be written to look for \n everywhere.  (This is the intention of
the current text mode too, but it fails when a file uses another
platform's convention.)

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Sun Apr  9 23:40:21 2000
From: guido@python.org (Guido van Rossum)
Date: Sun, 09 Apr 2000 18:40:21 -0400
Subject: [Patches] Fix main Makefile to create prefix, exec_prefix
In-Reply-To: Your message of "Sat, 08 Apr 2000 01:05:20 EDT."
 <14574.48656.133609.387194@seahag.cnri.reston.va.us>
References: <20000407181732.A11360@mems-exchange.org>
 <14574.48656.133609.387194@seahag.cnri.reston.va.us>
Message-ID: <200004092240.SAA28776@eric.cnri.reston.va.us>

> Greg Ward writes:
>  > This patch fixes the top Python Makefile so it creates $(prefix) or
>  > $(exec_prefix) before attempting to create directories under them.
>  > Handy when you use prefixes like "/usr/local/python-1.5.2" and
>  > "/usr/local/python-1.6a1", which don't exist until you install Python
> 
>   I say check it in.  If Guido disagrees he can yell/email about it
> later; this is quite reasonable given the recent discussion of "stow"
> and the need of developers to keep multiple versions (including patch
> levels!) installed.

I still don't like it.  Let's discuss this face-to-face.

My argument why it doesn't do this (where I know that it easily
could!) is that if root makes a typo and says

  make install prefix=/usr/kocal

it will happily create /usr/kocal.  I think this is bad.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From mhammond@skippinet.com.au  Mon Apr 10 04:06:14 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Mon, 10 Apr 2000 13:06:14 +1000
Subject: [Patches] Change to attribute error message for instances.
Message-ID: <ECEPKNMJLHAPFFJHDOJBIEIHCIAA.mhammond@skippinet.com.au>

In line with a similar checkin to object.c a while ago, this patch
gives a more descriptive error message for an attribute error on a
class instance.  The message now looks like:

AttributeError: 'Descriptor' instance has no attribute
'GetReturnType'

[Maybe the module name would also be useful, but IMO this is fine
and a nice improvement]

Release info:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that
I have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce,
distribute, perform and/or display publicly, prepare derivative
versions, and otherwise use this contribution as part of the Python
software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize
others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and
its related documentation.  I further grant CNRI permission to use
my name and other identifying information provided to CNRI by me for
use in connection with the Python software and its related
documentation.

Mark.

diff -c -2 -r2.83 classobject.c
*** classobject.c	2000/02/28 15:03:15	2.83
--- classobject.c	2000/04/10 03:01:08
***************
*** 593,597 ****
  		v = class_lookup(inst->in_class, name, &class);
  		if (v == NULL) {
! 			PyErr_SetObject(PyExc_AttributeError, name);
  			return NULL;
  		}
--- 593,600 ----
  		v = class_lookup(inst->in_class, name, &class);
  		if (v == NULL) {
! 			PyErr_Format(PyExc_AttributeError,
! 				     "'%.50s' instance has no attribute '%.400s'",
! 				     PyString_AsString(inst->in_class->cl_name),
! 				     sname);
  			return NULL;
  		}




From mal@lemburg.com  Mon Apr 10 10:51:44 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Mon, 10 Apr 2000 11:51:44 +0200
Subject: [Patches] Unicode Patch Set 2000-04-10
Message-ID: <38F1A430.D70DF89@lemburg.com>

This is a multi-part message in MIME format.
--------------3642FC2D466120215DDB5BCE
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

The attached patch includes the following fixes and additions:

* More test cases for test_contains.py.

* New exported API PyUnicode_Resize()

* '...%s...' % u"abc" now coerces to Unicode just like 
  string methods. Care is taken not to reevaluate already formatted
  arguments -- only the first Unicode object appearing in the
  argument mapping is looked up twice. Added test cases for
  this to test_unicode.py.

* TypeErrors during comparing of mixed type arguments including
  a Unicode object are now masked (just like they are for all
  other combinations).

* The experimental Keep-Alive optimization was turned back
  on after some tweaks to the implementation. It should now
  work without causing core dumps... this has yet to tested
  though (switching it off is easy: see the unicodeobject.c
  file for details).

* Fixed a memory leak in the Unicode freelist cleanup code.

* Added tests to correctly process the return code from
  _PyUnicode_Resize().

* Fixed a bug in the 'ignore' error handling routines
  of some builtin codecs. Added test cases for these to
  test_unicode.py.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------3642FC2D466120215DDB5BCE
Content-Type: text/plain; charset=iso-8859-1;
 name="Unicode-Implementation-2000-04-10.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-10.patch"

diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Include/unicodeobject.h Python+Unicode/Include/unicodeobject.h
--- CVS-Python/Include/unicodeobject.h	Thu Apr  6 10:00:22 2000
+++ Python+Unicode/Include/unicodeobject.h	Sat Apr  8 12:56:27 2000
@@ -237,6 +237,25 @@
     PyObject *unicode	 	/* Unicode object */
     );
 
+/* Resize an already allocated Unicode object to the new size length.
+
+   *unicode is modified to point to the new (resized) object and 0
+   returned on success.
+
+   This API may only be called by the function which also called the
+   Unicode constructor. The refcount on the object must be 1. Otherwise,
+   an error is returned.
+
+   Error handling is implemented as follows: an exception is set, -1
+   is returned and *unicode left untouched.
+
+*/
+
+extern DL_IMPORT(int) PyUnicode_Resize(
+    PyObject **unicode,		/* Pointer to the Unicode object */
+    int length			/* New length */
+    );
+
 /* Coerce obj to an Unicode object and return a reference with
    *incremented* refcount.
 
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Objects/object.c Python+Unicode/Objects/object.c
--- CVS-Python/Objects/object.c	Tue Mar 28 09:19:17 2000
+++ Python+Unicode/Objects/object.c	Wed Apr  5 02:44:28 2000
@@ -347,8 +347,21 @@
 				return cmp;
 			}
 		}
-		else if (PyUnicode_Check(v) || PyUnicode_Check(w))
-			return PyUnicode_Compare(v, w);
+		else if (PyUnicode_Check(v) || PyUnicode_Check(w)) {
+			int result = PyUnicode_Compare(v, w);
+			if (result == -1 && PyErr_Occurred() && 
+			    PyErr_ExceptionMatches(PyExc_TypeError))
+				/* TypeErrors are ignored: if Unicode coercion
+				fails due to one of the arguments not
+			 	having the right type, we continue as
+				defined by the coercion protocol (see
+				above). Luckily, decoding errors are
+				reported as ValueErrors and are not masked
+				by this technique. */
+				PyErr_Clear();
+			else
+				return result;
+		}
 		else if (vtp->tp_as_number != NULL)
 			vname = "";
 		else if (wtp->tp_as_number != NULL)
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Objects/stringobject.c Python+Unicode/Objects/stringobject.c
--- CVS-Python/Objects/stringobject.c	Fri Mar 24 00:01:30 2000
+++ Python+Unicode/Objects/stringobject.c	Sat Apr  8 01:04:08 2000
@@ -389,9 +389,9 @@
 {
 	register char *s, *end;
 	register char c;
-	if (!PyString_Check(el))
+	if (PyUnicode_Check(el))
 		return PyUnicode_Contains(a, el);
-	if (PyString_Size(el) != 1) {
+	if (!PyString_Check(el) || PyString_Size(el) != 1) {
 		PyErr_SetString(PyExc_TypeError,
 				"string member test needs char left operand");
 		return -1;
@@ -2384,12 +2384,13 @@
 	char *fmt, *res;
 	int fmtcnt, rescnt, reslen, arglen, argidx;
 	int args_owned = 0;
-	PyObject *result;
+	PyObject *result, *orig_args;
 	PyObject *dict = NULL;
 	if (format == NULL || !PyString_Check(format) || args == NULL) {
 		PyErr_BadInternalCall();
 		return NULL;
 	}
+	orig_args = args;
 	fmt = PyString_AsString(format);
 	fmtcnt = PyString_Size(format);
 	reslen = rescnt = fmtcnt + 100;
@@ -2434,6 +2435,8 @@
 			int sign;
 			int len;
 			char tmpbuf[120]; /* For format{float,int,char}() */
+			char *fmt_start = fmt;
+			
 			fmt++;
 			if (*fmt == '(') {
 				char *keystart;
@@ -2584,6 +2587,10 @@
 				break;
 			case 's':
 			case 'r':
+				if (PyUnicode_Check(v)) {
+					fmt = fmt_start;
+					goto unicode;
+				}
 				if (c == 's')
 					temp = PyObject_Str(v);
 				else
@@ -2716,6 +2723,47 @@
 	}
 	_PyString_Resize(&result, reslen - rescnt);
 	return result;
+
+ unicode:
+	if (args_owned) {
+		Py_DECREF(args);
+		args_owned = 0;
+	}
+	/* Fiddle args right (remove the first argidx-1 arguments) */
+	--argidx;
+	if (PyTuple_Check(orig_args) && argidx > 0) {
+		PyObject *v;
+		int n = PyTuple_GET_SIZE(orig_args) - argidx;
+		v = PyTuple_New(n);
+		if (v == NULL)
+			goto error;
+		while (--n >= 0) {
+			PyObject *w = PyTuple_GET_ITEM(orig_args, n + argidx);
+			Py_INCREF(w);
+			PyTuple_SET_ITEM(v, n, w);
+		}
+		args = v;
+	} else {
+		Py_INCREF(orig_args);
+		args = orig_args;
+	}
+	/* Paste rest of format string to what we have of the result
+	   string; we reuse result for this */
+	rescnt = res - PyString_AS_STRING(result);
+	fmtcnt = PyString_GET_SIZE(format) - \
+		 (fmt - PyString_AS_STRING(format));
+	if (_PyString_Resize(&result, rescnt + fmtcnt)) {
+		Py_DECREF(args);
+		goto error;
+	}
+	memcpy(PyString_AS_STRING(result) + rescnt, fmt, fmtcnt);
+	format = result;
+	/* Let Unicode do its magic */
+	result = PyUnicode_Format(format, args);
+	Py_DECREF(format);
+	Py_DECREF(args);
+	return result;
+	
  error:
 	Py_DECREF(result);
 	if (args_owned) {
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Objects/unicodeobject.c Python+Unicode/Objects/unicodeobject.c
--- CVS-Python/Objects/unicodeobject.c	Mon Apr 10 11:22:21 2000
+++ Python+Unicode/Objects/unicodeobject.c	Sat Apr  8 14:23:58 2000
@@ -76,6 +76,7 @@
 #ifdef MS_WIN32
 #include <windows.h>
 #endif
+
 /* Limit for the Unicode object free list */
 
 #define MAX_UNICODE_FREELIST_SIZE       1024
@@ -87,18 +88,17 @@
    limit. This reduces malloc() overhead for small Unicode objects.  
 
    At worst this will result in MAX_UNICODE_FREELIST_SIZE *
-   (sizeof(PyUnicodeObject) + STAYALIVE_SIZE_LIMIT +
+   (sizeof(PyUnicodeObject) + KEEPALIVE_SIZE_LIMIT +
    malloc()-overhead) bytes of unused garbage.
 
    Setting the limit to 0 effectively turns the feature off.
 
-   XXX The feature is currently turned off because there are
-       apparently some lingering bugs in its implementation which I
-       haven't yet been able to sort out.
+   Note: This is an experimental feature ! If you get core dumps when
+   using Unicode objects, turn this feature off.
 
 */
 
-#define STAYALIVE_SIZE_LIMIT       0
+#define KEEPALIVE_SIZE_LIMIT       9
 
 /* Endianness switches; defaults to little endian */
 
@@ -125,9 +125,9 @@
 {
     void *oldstr;
     
-    /* Shortcut if there's nothing to do. */
+    /* Shortcut if there's nothing much to do. */
     if (unicode->length == length)
-	return 0;
+	goto reset;
 
     /* Resizing unicode_empty is not allowed. */
     if (unicode == unicode_empty) {
@@ -148,6 +148,7 @@
     unicode->str[length] = 0;
     unicode->length = length;
 
+ reset:
     /* Reset the object caches */
     if (unicode->utf8str) {
         Py_DECREF(unicode->utf8str);
@@ -158,6 +159,23 @@
     return 0;
 }
 
+int PyUnicode_Resize(PyObject **unicode,
+		     int length)
+{
+    PyUnicodeObject *v;
+
+    if (unicode == NULL) {
+	PyErr_BadInternalCall();
+	return -1;
+    }
+    v = (PyUnicodeObject *)*unicode;
+    if (v == NULL || !PyUnicode_Check(v) || v->ob_refcnt != 1) {
+	PyErr_BadInternalCall();
+	return -1;
+    }
+    return _PyUnicode_Resize(v, length);
+}
+
 /* We allocate one more byte to make sure the string is
    Ux0000 terminated -- XXX is this needed ? 
 
@@ -185,7 +203,9 @@
         unicode->ob_type = &PyUnicode_Type;
         _Py_NewReference((PyObject *)unicode);
 	if (unicode->str) {
-	    if (unicode->length < length &&
+	    /* Keep-Alive optimization: we only upsize the buffer,
+	       never downsize it. */
+	    if ((unicode->length < length) &&
 		_PyUnicode_Resize(unicode, length)) {
 		free(unicode->str);
 		PyMem_DEL(unicode);
@@ -220,19 +240,25 @@
 static
 void _PyUnicode_Free(register PyUnicodeObject *unicode)
 {
-    Py_XDECREF(unicode->utf8str);
     if (unicode_freelist_size < MAX_UNICODE_FREELIST_SIZE) {
-	if (unicode->length >= STAYALIVE_SIZE_LIMIT) {
+        /* Keep-Alive optimization */
+	if (unicode->length >= KEEPALIVE_SIZE_LIMIT) {
 	    free(unicode->str);
 	    unicode->str = NULL;
 	    unicode->length = 0;
 	}
+	if (unicode->utf8str) {
+	    Py_DECREF(unicode->utf8str);
+	    unicode->utf8str = NULL;
+	}
+	/* Add to free list */
         *(PyUnicodeObject **)unicode = unicode_freelist;
         unicode_freelist = unicode;
         unicode_freelist_size++;
     }
     else {
 	free(unicode->str);
+	Py_XDECREF(unicode->utf8str);
         PyMem_DEL(unicode);
     }
 }
@@ -665,7 +691,8 @@
         }
     }
     *p = '\0';
-    _PyString_Resize(&v, p - q);
+    if (_PyString_Resize(&v, p - q))
+	goto onError;
 
  done:
     return v;
@@ -1047,7 +1074,8 @@
             break;
         }
     }
-    _PyUnicode_Resize(v, (int)(p - buf));
+    if (_PyUnicode_Resize(v, (int)(p - buf)))
+	goto onError;
     return (PyObject *)v;
     
  onError:
@@ -1119,9 +1147,14 @@
         *p++ = q[1];
 
     *p = '\0';
-    _PyString_Resize(&repr, p - q);
+    if (_PyString_Resize(&repr, p - q))
+	goto onError;
 
     return repr;
+
+ onError:
+    Py_DECREF(repr);
+    return NULL;
 }
 
 PyObject *PyUnicode_EncodeUnicodeEscape(const Py_UNICODE *s,
@@ -1209,7 +1242,8 @@
 	s += i;
 	*p++ = x;
     }
-    _PyUnicode_Resize(v, (int)(p - buf));
+    if (_PyUnicode_Resize(v, (int)(p - buf)))
+	goto onError;
     return (PyObject *)v;
     
  onError:
@@ -1247,9 +1281,14 @@
             *p++ = (char) ch;
     }
     *p = '\0';
-    _PyString_Resize(&repr, p - q);
+    if (_PyString_Resize(&repr, p - q))
+	goto onError;
 
     return repr;
+
+ onError:
+    Py_DECREF(repr);
+    return NULL;
 }
 
 PyObject *PyUnicode_AsRawUnicodeEscapeString(PyObject *unicode)
@@ -1305,6 +1344,7 @@
     }
     else if (strcmp(errors,"replace") == 0) {
 	**dest = '?';
+	(*dest)++;
 	return 0;
     }
     else {
@@ -1321,12 +1361,13 @@
 				 const char *errors)
 {
     PyObject *repr;
-    char *s;
+    char *s, *start;
     repr = PyString_FromStringAndSize(NULL, size);
     if (repr == NULL)
         return NULL;
 
     s = PyString_AS_STRING(repr);
+    start = s;
     while (size-- > 0) {
         Py_UNICODE ch = *p++;
 	if (ch >= 256) {
@@ -1337,6 +1378,10 @@
 	else
             *s++ = (char)ch;
     }
+    /* Resize if error handling skipped some characters */
+    if (s - start < PyString_GET_SIZE(repr))
+	if (_PyString_Resize(&repr, s - start))
+	    goto onError;
     return repr;
 
  onError:
@@ -1411,8 +1456,9 @@
 				      "ordinal not in range(128)"))
 		goto onError;
     }
-    if (p - PyUnicode_AS_UNICODE(v) < size)
-	_PyUnicode_Resize(v, (int)(p - PyUnicode_AS_UNICODE(v)));	
+    if (p - PyUnicode_AS_UNICODE(v) < PyString_GET_SIZE(v))
+	if (_PyUnicode_Resize(v, (int)(p - PyUnicode_AS_UNICODE(v))))
+	    goto onError;
     return (PyObject *)v;
     
  onError:
@@ -1438,6 +1484,7 @@
     }
     else if (strcmp(errors,"replace") == 0) {
 	**dest = '?';
+	(*dest)++;
 	return 0;
     }
     else {
@@ -1454,12 +1501,13 @@
 				const char *errors)
 {
     PyObject *repr;
-    char *s;
+    char *s, *start;
     repr = PyString_FromStringAndSize(NULL, size);
     if (repr == NULL)
         return NULL;
 
     s = PyString_AS_STRING(repr);
+    start = s;
     while (size-- > 0) {
         Py_UNICODE ch = *p++;
 	if (ch >= 128) {
@@ -1470,6 +1518,10 @@
 	else
             *s++ = (char)ch;
     }
+    /* Resize if error handling skipped some characters */
+    if (s - start < PyString_GET_SIZE(repr))
+	if (_PyString_Resize(&repr, s - start))
+	    goto onError;
     return repr;
 
  onError:
@@ -1898,7 +1950,8 @@
 	Py_DECREF(x);
     }
     if (p - PyUnicode_AS_UNICODE(v) < PyUnicode_GET_SIZE(v))
-	_PyUnicode_Resize(v, (int)(p - PyUnicode_AS_UNICODE(v)));	
+	if (_PyUnicode_Resize(v, (int)(p - PyUnicode_AS_UNICODE(v))))
+	    goto onError;
 
  done:
     return (PyObject *)v;
@@ -1959,7 +2012,7 @@
 	    continue;
 	}
 	if (0 < ch && ch < 256) {
-	    *output++ = (char) ch;
+	    *output++ = ch;
 	    continue;
 	}
 	/* All other characters are considered invalid */
@@ -4539,7 +4592,8 @@
 	Py_DECREF(args);
     }
     Py_DECREF(uformat);
-    _PyUnicode_Resize(result, reslen - rescnt);
+    if (_PyUnicode_Resize(result, reslen - rescnt))
+	goto onError;
     return (PyObject *)result;
 
  onError:
@@ -4605,6 +4659,9 @@
     while (u != NULL) {
 	PyUnicodeObject *v = u;
 	u = *(PyUnicodeObject **)u;
+	if (v->str)
+	    free(v->str);
+	Py_XDECREF(v->utf8str);
 	free(v);
     }
     Py_XDECREF(unicode_empty);
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_contains.py Python+Unicode/Lib/test/test_contains.py
--- CVS-Python/Lib/test/test_contains.py	Tue Mar  7 16:52:01 2000
+++ Python+Unicode/Lib/test/test_contains.py	Wed Apr  5 11:23:44 2000
@@ -17,7 +17,7 @@
 
 def check(ok, *args):
     if not ok:
-        raise TestFailed, join(map(str, args), " ")
+        raise TestFailed, " ".join(map(str, args))
 
 a = base_set(1)
 b = set(1)
@@ -60,5 +60,62 @@
 try:
 	None in 'abc'
 	check(0, "None in 'abc' did not raise error")
+except TypeError:
+	pass
+
+# Test char in Unicode
+
+check('c' in u'abc', "'c' not in u'abc'")
+check('d' not in u'abc', "'d' in u'abc'")
+
+try:
+	'' in u'abc'
+	check(0, "'' in u'abc' did not raise error")
+except TypeError:
+	pass
+
+try:
+	'ab' in u'abc'
+	check(0, "'ab' in u'abc' did not raise error")
+except TypeError:
+	pass
+
+try:
+	None in u'abc'
+	check(0, "None in u'abc' did not raise error")
+except TypeError:
+	pass
+
+# Test Unicode char in Unicode
+
+check(u'c' in u'abc', "u'c' not in u'abc'")
+check(u'd' not in u'abc', "u'd' in u'abc'")
+
+try:
+	u'' in u'abc'
+	check(0, "u'' in u'abc' did not raise error")
+except TypeError:
+	pass
+
+try:
+	u'ab' in u'abc'
+	check(0, "u'ab' in u'abc' did not raise error")
+except TypeError:
+	pass
+
+# Test Unicode char in string
+
+check(u'c' in 'abc', "u'c' not in 'abc'")
+check(u'd' not in 'abc', "u'd' in 'abc'")
+
+try:
+	u'' in 'abc'
+	check(0, "u'' in 'abc' did not raise error")
+except TypeError:
+	pass
+
+try:
+	u'ab' in 'abc'
+	check(0, "u'ab' in 'abc' did not raise error")
 except TypeError:
 	pass
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_unicode.py Python+Unicode/Lib/test/test_unicode.py
--- CVS-Python/Lib/test/test_unicode.py	Thu Apr  6 10:00:25 2000
+++ Python+Unicode/Lib/test/test_unicode.py	Sat Apr  8 00:52:42 2000
@@ -255,6 +255,15 @@
 assert u"%r, %r" % (u"abc", "abc") == u"u'abc', 'abc'"
 assert u"%(x)s, %(y)s" % {'x':u"abc", 'y':"def"} == u'abc, def'
 assert u"%(x)s, %(ä)s" % {'x':u"abc", u'ä'.encode('utf-8'):"def"} == u'abc, def'
+# formatting jobs delegated from the string implementation:
+assert '...%(foo)s...' % {'foo':u"abc"} == u'...abc...'
+assert '...%(foo)s...' % {'foo':"abc"} == '...abc...'
+assert '...%(foo)s...' % {u'foo':"abc"} == '...abc...'
+assert '...%(foo)s...' % {u'foo':u"abc"} == u'...abc...'
+assert '...%(foo)s...' % {u'foo':u"abc",'def':123} ==  u'...abc...'
+assert '...%(foo)s...' % {u'foo':u"abc",u'def':123} == u'...abc...'
+assert '...%s...%s...%s...%s...' % (1,2,3,u"abc") == u'...1...2...3...abc...'
+assert '...%s...' % u"abc" == u'...abc...'
 print 'done.'
 
 # Test builtin codecs
@@ -264,6 +273,26 @@
 assert unicode('hello','utf-8') == u'hello'
 assert unicode('hello','utf8') == u'hello'
 assert unicode('hello','latin-1') == u'hello'
+
+try:
+    u'Andr\202 x'.encode('ascii')
+    u'Andr\202 x'.encode('ascii','strict')
+except ValueError:
+    pass
+else:
+    raise AssertionError, "u'Andr\202'.encode('ascii') failed to raise an exception"
+assert u'Andr\202 x'.encode('ascii','ignore') == "Andr x"
+assert u'Andr\202 x'.encode('ascii','replace') == "Andr? x"
+
+try:
+    unicode('Andr\202 x','ascii')
+    unicode('Andr\202 x','ascii','strict')
+except ValueError:
+    pass
+else:
+    raise AssertionError, "unicode('Andr\202') failed to raise an exception"
+assert unicode('Andr\202 x','ascii','ignore') == u"Andr x"
+assert unicode('Andr\202 x','ascii','replace') == u'Andr\uFFFD x'
 
 assert u'hello'.encode('ascii') == 'hello'
 assert u'hello'.encode('utf-8') == 'hello'

--------------3642FC2D466120215DDB5BCE--




From gstein@lyra.org  Mon Apr 10 10:53:48 2000
From: gstein@lyra.org (Greg Stein)
Date: Mon, 10 Apr 2000 02:53:48 -0700 (PDT)
Subject: [Patches] Unicode Patch Set 2000-04-10
In-Reply-To: <38F1A430.D70DF89@lemburg.com>
Message-ID: <Pine.LNX.4.10.10004100250560.11297-100000@nebula.lyra.org>

On Mon, 10 Apr 2000, M.-A. Lemburg wrote:
> The attached patch includes the following fixes and additions:
>...
> * '...%s...' % u"abc" now coerces to Unicode just like 
>   string methods. Care is taken not to reevaluate already formatted
>   arguments -- only the first Unicode object appearing in the
>   argument mapping is looked up twice. Added test cases for
>   this to test_unicode.py.
>...

I missed a chance to bring this up on the first round of discussion, but
is this really the right thing to do? We never coerce the string on the
left based on operands. For example: if the operands are class instances,
we call __str__ -- we don't call __coerce__.

It seems a bit weird to magically revise the left operand.

In many cases, a Unicode used as a string is used as a UTF-8 value. Why is
that different in this case? Seems like a wierd special case.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/



From mal@lemburg.com  Mon Apr 10 11:55:50 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Mon, 10 Apr 2000 12:55:50 +0200
Subject: [Python-Dev] Re: [Patches] Unicode Patch Set 2000-04-10
References: <Pine.LNX.4.10.10004100250560.11297-100000@nebula.lyra.org>
Message-ID: <38F1B336.12B6707@lemburg.com>

Greg Stein wrote:
> 
> On Mon, 10 Apr 2000, M.-A. Lemburg wrote:
> > The attached patch includes the following fixes and additions:
> >...
> > * '...%s...' % u"abc" now coerces to Unicode just like
> >   string methods. Care is taken not to reevaluate already formatted
> >   arguments -- only the first Unicode object appearing in the
> >   argument mapping is looked up twice. Added test cases for
> >   this to test_unicode.py.
> >...
> 
> I missed a chance to bring this up on the first round of discussion, but
> is this really the right thing to do? We never coerce the string on the
> left based on operands. For example: if the operands are class instances,
> we call __str__ -- we don't call __coerce__.
> 
> It seems a bit weird to magically revise the left operand.
> 
> In many cases, a Unicode used as a string is used as a UTF-8 value. Why is
> that different in this case? Seems like a wierd special case.

It's not a special case: % works just like a method call
and all string methods auto-coerce to Unicode in case a
Unicode object is found among the arguments.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/


From skip@mojam.com (Skip Montanaro)  Mon Apr 10 12:13:24 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Mon, 10 Apr 2000 06:13:24 -0500
Subject: [Patches] add string precisions to PyErr_Format calls
Message-ID: <200004101113.GAA09190@beluga.mojam.com>

The following attachments add string precisions to a number of calls to
PyErr_Format to prevent possible buffer overruns.

    I confirm that, to the best of my knowledge and belief, this
    contribution is free of any claims of third parties under copyright,
    patent or other rights or interests ("claims").  To the extent that I
    have any such claims, I hereby grant to CNRI a nonexclusive,
    irrevocable, royalty-free, worldwide license to reproduce, distribute,
    perform and/or display publicly, prepare derivative versions, and
    otherwise use this contribution as part of the Python software and its
    related documentation, or any derivative versions thereof, at no cost to
    CNRI or its licensed users, and to authorize others to do so.

    I acknowledge that CNRI may, at its sole discretion, decide whether or
    not to incorporate this contribution in the Python software and its
    related documentation.  I further grant CNRI permission to use my name
    and other identifying information provided to CNRI by me for use in
    connection with the Python software and its related documentation.

-- 
Skip Montanaro | http://www.mojam.com/
skip@mojam.com | http://www.musi-cal.com/

[ATTACHMENT ~/src/python/dist/src/Objects/object.c.diff, text/plain]
[ATTACHMENT ~/src/python/dist/src/Objects/unicodeobject.c.diff, text/plain]
[ATTACHMENT ~/src/python/dist/src/Modules/pyexpat.c.diff, text/plain]
[ATTACHMENT ~/src/python/dist/src/Python/dynload_next.c.diff, text/plain]
[ATTACHMENT ~/src/python/dist/src/Python/ceval.c.diff, text/plain]


From skip@mojam.com (Skip Montanaro)  Mon Apr 10 12:22:58 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Mon, 10 Apr 2000 06:22:58 -0500 (CDT)
Subject: [Patches] add string precisions to PyErr_Format calls
In-Reply-To: <200004101113.GAA09190@beluga.mojam.com>
References: <200004101113.GAA09190@beluga.mojam.com>
Message-ID: <14577.47506.14149.790979@beluga.mojam.com>

    Skip> [ATTACHMENT ~/src/python/dist/src/Objects/object.c.diff, text/plain]
    Skip> [ATTACHMENT ~/src/python/dist/src/Objects/unicodeobject.c.diff, text/plain]
    Skip> [ATTACHMENT ~/src/python/dist/src/Modules/pyexpat.c.diff, text/plain]
    Skip> [ATTACHMENT ~/src/python/dist/src/Python/dynload_next.c.diff, text/plain]
    Skip> [ATTACHMENT ~/src/python/dist/src/Python/ceval.c.diff, text/plain]

Ack!  The attachments didn't attach!  Here they are as plain text appended
one after the other...

Skip

*** /tmp/object.c.~2.65~	Mon Apr 10 06:01:08 2000
--- /tmp/object.c	Mon Apr 10 06:01:08 2000
***************
*** 236,242 ****
  			return NULL;
  		if (!PyString_Check(res)) {
  			PyErr_Format(PyExc_TypeError,
! 				     "__repr__ returned non-string (type %s)",
  				     res->ob_type->tp_name);
  			Py_DECREF(res);
  			return NULL;
--- 236,242 ----
  			return NULL;
  		if (!PyString_Check(res)) {
  			PyErr_Format(PyExc_TypeError,
! 				     "__repr__ returned non-string (type %.200s)",
  				     res->ob_type->tp_name);
  			Py_DECREF(res);
  			return NULL;
***************
*** 273,279 ****
  		return NULL;
  	if (!PyString_Check(res)) {
  		PyErr_Format(PyExc_TypeError,
! 			     "__str__ returned non-string (type %s)",
  			     res->ob_type->tp_name);
  		Py_DECREF(res);
  		return NULL;
--- 273,279 ----
  		return NULL;
  	if (!PyString_Check(res)) {
  		PyErr_Format(PyExc_TypeError,
! 			     "__str__ returned non-string (type %.200s)",
  			     res->ob_type->tp_name);
  		Py_DECREF(res);
  		return NULL;

*** /tmp/unicodeobject.c.~2.8~	Mon Apr 10 06:04:11 2000
--- /tmp/unicodeobject.c	Mon Apr 10 06:04:11 2000
***************
*** 365,371 ****
          goto onError;
      if (!PyUnicode_Check(unicode)) {
          PyErr_Format(PyExc_TypeError,
!                      "decoder did not return an unicode object (type=%s)",
                       unicode->ob_type->tp_name);
          Py_DECREF(unicode);
          goto onError;
--- 365,371 ----
          goto onError;
      if (!PyUnicode_Check(unicode)) {
          PyErr_Format(PyExc_TypeError,
!                      "decoder did not return an unicode object (type=%.400s)",
                       unicode->ob_type->tp_name);
          Py_DECREF(unicode);
          goto onError;
***************
*** 416,422 ****
      /* XXX Should we really enforce this ? */
      if (!PyString_Check(v)) {
          PyErr_Format(PyExc_TypeError,
!                      "encoder did not return a string object (type=%s)",
                       v->ob_type->tp_name);
          Py_DECREF(v);
          goto onError;
--- 416,422 ----
      /* XXX Should we really enforce this ? */
      if (!PyString_Check(v)) {
          PyErr_Format(PyExc_TypeError,
!                      "encoder did not return a string object (type=%.400s)",
                       v->ob_type->tp_name);
          Py_DECREF(v);
          goto onError;
***************
*** 484,490 ****
      if ((errors == NULL) ||
          (strcmp(errors,"strict") == 0)) {
          PyErr_Format(PyExc_UnicodeError,
!                      "UTF-8 decoding error: %s",
                       details);
          return -1;
      }
--- 484,490 ----
      if ((errors == NULL) ||
          (strcmp(errors,"strict") == 0)) {
          PyErr_Format(PyExc_UnicodeError,
!                      "UTF-8 decoding error: %.400s",
                       details);
          return -1;
      }
***************
*** 500,506 ****
      }
      else {
          PyErr_Format(PyExc_ValueError,
!                      "UTF-8 decoding error; unknown error handling code: %s",
                       errors);
          return -1;
      }
--- 500,506 ----
      }
      else {
          PyErr_Format(PyExc_ValueError,
!                      "UTF-8 decoding error; unknown error handling code: %.400s",
                       errors);
          return -1;
      }
***************
*** 607,613 ****
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "UTF-8 encoding error: %s",
  		     details);
  	return -1;
      }
--- 607,613 ----
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "UTF-8 encoding error: %.400s",
  		     details);
  	return -1;
      }
***************
*** 622,628 ****
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "UTF-8 encoding error; "
! 		     "unknown error handling code: %s",
  		     errors);
  	return -1;
      }
--- 622,628 ----
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "UTF-8 encoding error; "
! 		     "unknown error handling code: %.400s",
  		     errors);
  	return -1;
      }
***************
*** 728,734 ****
      if ((errors == NULL) ||
          (strcmp(errors,"strict") == 0)) {
          PyErr_Format(PyExc_UnicodeError,
!                      "UTF-16 decoding error: %s",
                       details);
          return -1;
      }
--- 728,734 ----
      if ((errors == NULL) ||
          (strcmp(errors,"strict") == 0)) {
          PyErr_Format(PyExc_UnicodeError,
!                      "UTF-16 decoding error: %.400s",
                       details);
          return -1;
      }
***************
*** 744,750 ****
      }
      else {
          PyErr_Format(PyExc_ValueError,
!                      "UTF-16 decoding error; unknown error handling code: %s",
                       errors);
          return -1;
      }
--- 744,750 ----
      }
      else {
          PyErr_Format(PyExc_ValueError,
!                      "UTF-16 decoding error; unknown error handling code: %.400s",
                       errors);
          return -1;
      }
***************
*** 918,924 ****
      if ((errors == NULL) ||
          (strcmp(errors,"strict") == 0)) {
          PyErr_Format(PyExc_UnicodeError,
!                      "Unicode-Escape decoding error: %s",
                       details);
          return -1;
      }
--- 918,924 ----
      if ((errors == NULL) ||
          (strcmp(errors,"strict") == 0)) {
          PyErr_Format(PyExc_UnicodeError,
!                      "Unicode-Escape decoding error: %.400s",
                       details);
          return -1;
      }
***************
*** 932,938 ****
      else {
          PyErr_Format(PyExc_ValueError,
                       "Unicode-Escape decoding error; "
!                      "unknown error handling code: %s",
                       errors);
          return -1;
      }
--- 932,938 ----
      else {
          PyErr_Format(PyExc_ValueError,
                       "Unicode-Escape decoding error; "
!                      "unknown error handling code: %.400s",
                       errors);
          return -1;
      }
***************
*** 1296,1302 ****
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "Latin-1 encoding error: %s",
  		     details);
  	return -1;
      }
--- 1296,1302 ----
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "Latin-1 encoding error: %.400s",
  		     details);
  	return -1;
      }
***************
*** 1310,1316 ****
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "Latin-1 encoding error; "
! 		     "unknown error handling code: %s",
  		     errors);
  	return -1;
      }
--- 1310,1316 ----
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "Latin-1 encoding error; "
! 		     "unknown error handling code: %.400s",
  		     errors);
  	return -1;
      }
***************
*** 1366,1372 ****
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "ASCII decoding error: %s",
  		     details);
  	return -1;
      }
--- 1366,1372 ----
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "ASCII decoding error: %.400s",
  		     details);
  	return -1;
      }
***************
*** 1381,1387 ****
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "ASCII decoding error; "
! 		     "unknown error handling code: %s",
  		     errors);
  	return -1;
      }
--- 1381,1387 ----
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "ASCII decoding error; "
! 		     "unknown error handling code: %.400s",
  		     errors);
  	return -1;
      }
***************
*** 1429,1435 ****
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "ASCII encoding error: %s",
  		     details);
  	return -1;
      }
--- 1429,1435 ----
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "ASCII encoding error: %.400s",
  		     details);
  	return -1;
      }
***************
*** 1443,1449 ****
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "ASCII encoding error; "
! 		     "unknown error handling code: %s",
  		     errors);
  	return -1;
      }
--- 1443,1449 ----
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "ASCII encoding error; "
! 		     "unknown error handling code: %.400s",
  		     errors);
  	return -1;
      }
***************
*** 1558,1564 ****
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "charmap decoding error: %s",
  		     details);
  	return -1;
      }
--- 1558,1564 ----
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "charmap decoding error: %.400s",
  		     details);
  	return -1;
      }
***************
*** 1573,1579 ****
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "charmap decoding error; "
! 		     "unknown error handling code: %s",
  		     errors);
  	return -1;
      }
--- 1573,1579 ----
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "charmap decoding error; "
! 		     "unknown error handling code: %.400s",
  		     errors);
  	return -1;
      }
***************
*** 1674,1680 ****
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "charmap encoding error: %s",
  		     details);
  	return -1;
      }
--- 1674,1680 ----
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "charmap encoding error: %.400s",
  		     details);
  	return -1;
      }
***************
*** 1689,1695 ****
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "charmap encoding error; "
! 		     "unknown error handling code: %s",
  		     errors);
  	return -1;
      }
--- 1689,1695 ----
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "charmap encoding error; "
! 		     "unknown error handling code: %.400s",
  		     errors);
  	return -1;
      }
***************
*** 1806,1812 ****
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "translate error: %s",
  		     details);
  	return -1;
      }
--- 1806,1812 ----
      if ((errors == NULL) ||
  	(strcmp(errors,"strict") == 0)) {
  	PyErr_Format(PyExc_UnicodeError,
! 		     "translate error: %.400s",
  		     details);
  	return -1;
      }
***************
*** 1821,1827 ****
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "translate error; "
! 		     "unknown error handling code: %s",
  		     errors);
  	return -1;
      }
--- 1821,1827 ----
      else {
  	PyErr_Format(PyExc_ValueError,
  		     "translate error; "
! 		     "unknown error handling code: %.400s",
  		     errors);
  	return -1;
      }

*** /tmp/pyexpat.c.~2.2~	Mon Apr 10 06:01:33 2000
--- /tmp/pyexpat.c	Mon Apr 10 06:01:33 2000
***************
*** 307,313 ****
                  return NULL;
          }
          else if (rv == 0) {
!                 PyErr_Format(ErrorObject, "%s: line %i, column %i",
                               XML_ErrorString( XML_GetErrorCode(self->itself) ),
                               XML_GetErrorLineNumber(self->itself),
                               XML_GetErrorColumnNumber(self->itself) );
--- 307,313 ----
                  return NULL;
          }
          else if (rv == 0) {
!                 PyErr_Format(ErrorObject, "%.200s: line %i, column %i",
                               XML_ErrorString( XML_GetErrorCode(self->itself) ),
                               XML_GetErrorLineNumber(self->itself),
                               XML_GetErrorColumnNumber(self->itself) );

*** /tmp/dynload_next.c.~2.2~-YCj-2	Mon Apr 10 06:00:09 2000
--- /tmp/dynload_next.c-YCVIG	Mon Apr 10 06:00:09 2000
***************
*** 185,191 ****
  		if (!NSIsSymbolNameDefined(funcname)) {
  			/* UnlinkModule() isn't implimented in current versions, but calling it does no harm */
  			NSUnLinkModule(newModule, FALSE);
! 			PyErr_Format(PyExc_ImportError, "Loaded module does not contain symbol %s", funcname);
  			return NULL;
  		}
  		theSym = NSLookupAndBindSymbol(funcname);
--- 185,193 ----
  		if (!NSIsSymbolNameDefined(funcname)) {
  			/* UnlinkModule() isn't implimented in current versions, but calling it does no harm */
  			NSUnLinkModule(newModule, FALSE);
! 			PyErr_Format(PyExc_ImportError,
! 				     "Loaded module does not contain symbol %.200s",
! 				     funcname);
  			return NULL;
  		}
  		theSym = NSLookupAndBindSymbol(funcname);

*** /tmp/ceval.c.~2.174~-YCW7k	Mon Apr 10 06:00:41 2000
--- /tmp/ceval.c-YCjFr	Mon Apr 10 06:00:41 2000
***************
*** 2513,2519 ****
  	else {
  		if (!PyFunction_Check(func)) {
  			PyErr_Format(PyExc_TypeError,
! 				     "call of non-function (type %s)",
  				     func->ob_type->tp_name);
  			return NULL;
  		}
--- 2513,2519 ----
  	else {
  		if (!PyFunction_Check(func)) {
  			PyErr_Format(PyExc_TypeError,
! 				     "call of non-function (type %.200s)",
  				     func->ob_type->tp_name);
  			return NULL;
  		}


From gstein@lyra.org  Mon Apr 10 13:30:29 2000
From: gstein@lyra.org (Greg Stein)
Date: Mon, 10 Apr 2000 05:30:29 -0700 (PDT)
Subject: [Patches] add string precisions to PyErr_Format calls
In-Reply-To: <14577.47506.14149.790979@beluga.mojam.com>
Message-ID: <Pine.LNX.4.10.10004100529580.11297-100000@nebula.lyra.org>

Wouldn't it be best to simply fix PyErr_Format so that we don't have to
continue to worry about buffer overruns?

Cheers,
-g

On Mon, 10 Apr 2000, Skip Montanaro wrote:

> 
>     Skip> [ATTACHMENT ~/src/python/dist/src/Objects/object.c.diff, text/plain]
>     Skip> [ATTACHMENT ~/src/python/dist/src/Objects/unicodeobject.c.diff, text/plain]
>     Skip> [ATTACHMENT ~/src/python/dist/src/Modules/pyexpat.c.diff, text/plain]
>     Skip> [ATTACHMENT ~/src/python/dist/src/Python/dynload_next.c.diff, text/plain]
>     Skip> [ATTACHMENT ~/src/python/dist/src/Python/ceval.c.diff, text/plain]
> 
> Ack!  The attachments didn't attach!  Here they are as plain text appended
> one after the other...
> 
> Skip
> 
> *** /tmp/object.c.~2.65~	Mon Apr 10 06:01:08 2000
> --- /tmp/object.c	Mon Apr 10 06:01:08 2000
> ***************
> *** 236,242 ****
>   			return NULL;
>   		if (!PyString_Check(res)) {
>   			PyErr_Format(PyExc_TypeError,
> ! 				     "__repr__ returned non-string (type %s)",
>   				     res->ob_type->tp_name);
>   			Py_DECREF(res);
>   			return NULL;
> --- 236,242 ----
>   			return NULL;
>   		if (!PyString_Check(res)) {
>   			PyErr_Format(PyExc_TypeError,
> ! 				     "__repr__ returned non-string (type %.200s)",
>   				     res->ob_type->tp_name);
>   			Py_DECREF(res);
>   			return NULL;
> ***************
> *** 273,279 ****
>   		return NULL;
>   	if (!PyString_Check(res)) {
>   		PyErr_Format(PyExc_TypeError,
> ! 			     "__str__ returned non-string (type %s)",
>   			     res->ob_type->tp_name);
>   		Py_DECREF(res);
>   		return NULL;
> --- 273,279 ----
>   		return NULL;
>   	if (!PyString_Check(res)) {
>   		PyErr_Format(PyExc_TypeError,
> ! 			     "__str__ returned non-string (type %.200s)",
>   			     res->ob_type->tp_name);
>   		Py_DECREF(res);
>   		return NULL;
> 
> *** /tmp/unicodeobject.c.~2.8~	Mon Apr 10 06:04:11 2000
> --- /tmp/unicodeobject.c	Mon Apr 10 06:04:11 2000
> ***************
> *** 365,371 ****
>           goto onError;
>       if (!PyUnicode_Check(unicode)) {
>           PyErr_Format(PyExc_TypeError,
> !                      "decoder did not return an unicode object (type=%s)",
>                        unicode->ob_type->tp_name);
>           Py_DECREF(unicode);
>           goto onError;
> --- 365,371 ----
>           goto onError;
>       if (!PyUnicode_Check(unicode)) {
>           PyErr_Format(PyExc_TypeError,
> !                      "decoder did not return an unicode object (type=%.400s)",
>                        unicode->ob_type->tp_name);
>           Py_DECREF(unicode);
>           goto onError;
> ***************
> *** 416,422 ****
>       /* XXX Should we really enforce this ? */
>       if (!PyString_Check(v)) {
>           PyErr_Format(PyExc_TypeError,
> !                      "encoder did not return a string object (type=%s)",
>                        v->ob_type->tp_name);
>           Py_DECREF(v);
>           goto onError;
> --- 416,422 ----
>       /* XXX Should we really enforce this ? */
>       if (!PyString_Check(v)) {
>           PyErr_Format(PyExc_TypeError,
> !                      "encoder did not return a string object (type=%.400s)",
>                        v->ob_type->tp_name);
>           Py_DECREF(v);
>           goto onError;
> ***************
> *** 484,490 ****
>       if ((errors == NULL) ||
>           (strcmp(errors,"strict") == 0)) {
>           PyErr_Format(PyExc_UnicodeError,
> !                      "UTF-8 decoding error: %s",
>                        details);
>           return -1;
>       }
> --- 484,490 ----
>       if ((errors == NULL) ||
>           (strcmp(errors,"strict") == 0)) {
>           PyErr_Format(PyExc_UnicodeError,
> !                      "UTF-8 decoding error: %.400s",
>                        details);
>           return -1;
>       }
> ***************
> *** 500,506 ****
>       }
>       else {
>           PyErr_Format(PyExc_ValueError,
> !                      "UTF-8 decoding error; unknown error handling code: %s",
>                        errors);
>           return -1;
>       }
> --- 500,506 ----
>       }
>       else {
>           PyErr_Format(PyExc_ValueError,
> !                      "UTF-8 decoding error; unknown error handling code: %.400s",
>                        errors);
>           return -1;
>       }
> ***************
> *** 607,613 ****
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "UTF-8 encoding error: %s",
>   		     details);
>   	return -1;
>       }
> --- 607,613 ----
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "UTF-8 encoding error: %.400s",
>   		     details);
>   	return -1;
>       }
> ***************
> *** 622,628 ****
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "UTF-8 encoding error; "
> ! 		     "unknown error handling code: %s",
>   		     errors);
>   	return -1;
>       }
> --- 622,628 ----
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "UTF-8 encoding error; "
> ! 		     "unknown error handling code: %.400s",
>   		     errors);
>   	return -1;
>       }
> ***************
> *** 728,734 ****
>       if ((errors == NULL) ||
>           (strcmp(errors,"strict") == 0)) {
>           PyErr_Format(PyExc_UnicodeError,
> !                      "UTF-16 decoding error: %s",
>                        details);
>           return -1;
>       }
> --- 728,734 ----
>       if ((errors == NULL) ||
>           (strcmp(errors,"strict") == 0)) {
>           PyErr_Format(PyExc_UnicodeError,
> !                      "UTF-16 decoding error: %.400s",
>                        details);
>           return -1;
>       }
> ***************
> *** 744,750 ****
>       }
>       else {
>           PyErr_Format(PyExc_ValueError,
> !                      "UTF-16 decoding error; unknown error handling code: %s",
>                        errors);
>           return -1;
>       }
> --- 744,750 ----
>       }
>       else {
>           PyErr_Format(PyExc_ValueError,
> !                      "UTF-16 decoding error; unknown error handling code: %.400s",
>                        errors);
>           return -1;
>       }
> ***************
> *** 918,924 ****
>       if ((errors == NULL) ||
>           (strcmp(errors,"strict") == 0)) {
>           PyErr_Format(PyExc_UnicodeError,
> !                      "Unicode-Escape decoding error: %s",
>                        details);
>           return -1;
>       }
> --- 918,924 ----
>       if ((errors == NULL) ||
>           (strcmp(errors,"strict") == 0)) {
>           PyErr_Format(PyExc_UnicodeError,
> !                      "Unicode-Escape decoding error: %.400s",
>                        details);
>           return -1;
>       }
> ***************
> *** 932,938 ****
>       else {
>           PyErr_Format(PyExc_ValueError,
>                        "Unicode-Escape decoding error; "
> !                      "unknown error handling code: %s",
>                        errors);
>           return -1;
>       }
> --- 932,938 ----
>       else {
>           PyErr_Format(PyExc_ValueError,
>                        "Unicode-Escape decoding error; "
> !                      "unknown error handling code: %.400s",
>                        errors);
>           return -1;
>       }
> ***************
> *** 1296,1302 ****
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "Latin-1 encoding error: %s",
>   		     details);
>   	return -1;
>       }
> --- 1296,1302 ----
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "Latin-1 encoding error: %.400s",
>   		     details);
>   	return -1;
>       }
> ***************
> *** 1310,1316 ****
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "Latin-1 encoding error; "
> ! 		     "unknown error handling code: %s",
>   		     errors);
>   	return -1;
>       }
> --- 1310,1316 ----
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "Latin-1 encoding error; "
> ! 		     "unknown error handling code: %.400s",
>   		     errors);
>   	return -1;
>       }
> ***************
> *** 1366,1372 ****
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "ASCII decoding error: %s",
>   		     details);
>   	return -1;
>       }
> --- 1366,1372 ----
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "ASCII decoding error: %.400s",
>   		     details);
>   	return -1;
>       }
> ***************
> *** 1381,1387 ****
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "ASCII decoding error; "
> ! 		     "unknown error handling code: %s",
>   		     errors);
>   	return -1;
>       }
> --- 1381,1387 ----
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "ASCII decoding error; "
> ! 		     "unknown error handling code: %.400s",
>   		     errors);
>   	return -1;
>       }
> ***************
> *** 1429,1435 ****
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "ASCII encoding error: %s",
>   		     details);
>   	return -1;
>       }
> --- 1429,1435 ----
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "ASCII encoding error: %.400s",
>   		     details);
>   	return -1;
>       }
> ***************
> *** 1443,1449 ****
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "ASCII encoding error; "
> ! 		     "unknown error handling code: %s",
>   		     errors);
>   	return -1;
>       }
> --- 1443,1449 ----
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "ASCII encoding error; "
> ! 		     "unknown error handling code: %.400s",
>   		     errors);
>   	return -1;
>       }
> ***************
> *** 1558,1564 ****
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "charmap decoding error: %s",
>   		     details);
>   	return -1;
>       }
> --- 1558,1564 ----
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "charmap decoding error: %.400s",
>   		     details);
>   	return -1;
>       }
> ***************
> *** 1573,1579 ****
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "charmap decoding error; "
> ! 		     "unknown error handling code: %s",
>   		     errors);
>   	return -1;
>       }
> --- 1573,1579 ----
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "charmap decoding error; "
> ! 		     "unknown error handling code: %.400s",
>   		     errors);
>   	return -1;
>       }
> ***************
> *** 1674,1680 ****
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "charmap encoding error: %s",
>   		     details);
>   	return -1;
>       }
> --- 1674,1680 ----
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "charmap encoding error: %.400s",
>   		     details);
>   	return -1;
>       }
> ***************
> *** 1689,1695 ****
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "charmap encoding error; "
> ! 		     "unknown error handling code: %s",
>   		     errors);
>   	return -1;
>       }
> --- 1689,1695 ----
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "charmap encoding error; "
> ! 		     "unknown error handling code: %.400s",
>   		     errors);
>   	return -1;
>       }
> ***************
> *** 1806,1812 ****
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "translate error: %s",
>   		     details);
>   	return -1;
>       }
> --- 1806,1812 ----
>       if ((errors == NULL) ||
>   	(strcmp(errors,"strict") == 0)) {
>   	PyErr_Format(PyExc_UnicodeError,
> ! 		     "translate error: %.400s",
>   		     details);
>   	return -1;
>       }
> ***************
> *** 1821,1827 ****
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "translate error; "
> ! 		     "unknown error handling code: %s",
>   		     errors);
>   	return -1;
>       }
> --- 1821,1827 ----
>       else {
>   	PyErr_Format(PyExc_ValueError,
>   		     "translate error; "
> ! 		     "unknown error handling code: %.400s",
>   		     errors);
>   	return -1;
>       }
> 
> *** /tmp/pyexpat.c.~2.2~	Mon Apr 10 06:01:33 2000
> --- /tmp/pyexpat.c	Mon Apr 10 06:01:33 2000
> ***************
> *** 307,313 ****
>                   return NULL;
>           }
>           else if (rv == 0) {
> !                 PyErr_Format(ErrorObject, "%s: line %i, column %i",
>                                XML_ErrorString( XML_GetErrorCode(self->itself) ),
>                                XML_GetErrorLineNumber(self->itself),
>                                XML_GetErrorColumnNumber(self->itself) );
> --- 307,313 ----
>                   return NULL;
>           }
>           else if (rv == 0) {
> !                 PyErr_Format(ErrorObject, "%.200s: line %i, column %i",
>                                XML_ErrorString( XML_GetErrorCode(self->itself) ),
>                                XML_GetErrorLineNumber(self->itself),
>                                XML_GetErrorColumnNumber(self->itself) );
> 
> *** /tmp/dynload_next.c.~2.2~-YCj-2	Mon Apr 10 06:00:09 2000
> --- /tmp/dynload_next.c-YCVIG	Mon Apr 10 06:00:09 2000
> ***************
> *** 185,191 ****
>   		if (!NSIsSymbolNameDefined(funcname)) {
>   			/* UnlinkModule() isn't implimented in current versions, but calling it does no harm */
>   			NSUnLinkModule(newModule, FALSE);
> ! 			PyErr_Format(PyExc_ImportError, "Loaded module does not contain symbol %s", funcname);
>   			return NULL;
>   		}
>   		theSym = NSLookupAndBindSymbol(funcname);
> --- 185,193 ----
>   		if (!NSIsSymbolNameDefined(funcname)) {
>   			/* UnlinkModule() isn't implimented in current versions, but calling it does no harm */
>   			NSUnLinkModule(newModule, FALSE);
> ! 			PyErr_Format(PyExc_ImportError,
> ! 				     "Loaded module does not contain symbol %.200s",
> ! 				     funcname);
>   			return NULL;
>   		}
>   		theSym = NSLookupAndBindSymbol(funcname);
> 
> *** /tmp/ceval.c.~2.174~-YCW7k	Mon Apr 10 06:00:41 2000
> --- /tmp/ceval.c-YCjFr	Mon Apr 10 06:00:41 2000
> ***************
> *** 2513,2519 ****
>   	else {
>   		if (!PyFunction_Check(func)) {
>   			PyErr_Format(PyExc_TypeError,
> ! 				     "call of non-function (type %s)",
>   				     func->ob_type->tp_name);
>   			return NULL;
>   		}
> --- 2513,2519 ----
>   	else {
>   		if (!PyFunction_Check(func)) {
>   			PyErr_Format(PyExc_TypeError,
> ! 				     "call of non-function (type %.200s)",
>   				     func->ob_type->tp_name);
>   			return NULL;
>   		}
> 
> _______________________________________________
> Patches mailing list
> Patches@python.org
> http://www.python.org/mailman/listinfo/patches
> 

-- 
Greg Stein, http://www.lyra.org/



From skip@mojam.com (Skip Montanaro)  Mon Apr 10 12:29:58 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Mon, 10 Apr 2000 06:29:58 -0500 (CDT)
Subject: [Patches] add string precisions to PyErr_Format calls
In-Reply-To: <Pine.LNX.4.10.10004100529580.11297-100000@nebula.lyra.org>
References: <14577.47506.14149.790979@beluga.mojam.com>
 <Pine.LNX.4.10.10004100529580.11297-100000@nebula.lyra.org>
Message-ID: <14577.47926.702463.105644@beluga.mojam.com>

    Greg> Wouldn't it be best to simply fix PyErr_Format so that we don't
    Greg> have to continue to worry about buffer overruns?

Yeah, probably, but checking for missing string precisions seemed to be the
simplest way to close these particular loopholes.  Whenever anyone decides
to fix PyErr_Format, please also fix sysmodule.c:mywrite.

Skip


From gward@mems-exchange.org  Mon Apr 10 14:20:56 2000
From: gward@mems-exchange.org (Greg Ward)
Date: Mon, 10 Apr 2000 09:20:56 -0400
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <Pine.LNX.4.10.10004080502410.4140-100000@nebula.lyra.org>; from gstein@lyra.org on Sat, Apr 08, 2000 at 05:05:30AM -0700
References: <200004051959.PAA18356@eric.cnri.reston.va.us> <Pine.LNX.4.10.10004080502410.4140-100000@nebula.lyra.org>
Message-ID: <20000410092055.A562@mems-exchange.org>

On 08 April 2000, Greg Stein said:
> I might go one further, and ask for a way to *strip* the line, rather than
> just take the newline off the end. I can't think of a clean way to do so.
> For example:
> 
>   fileinput.input()
>   fileinput.input(strip=fileinput.STRIP_NL)
>   fileinput.input(strip=fileinput.STRIP_ALL)
> 
> It just seems a bit wordy, especially compared to the "chop=1" thing.

See the distutils.text_file module for a rather overblown approach to
this problem: it takes care of stripping newlines, deleting blank lines,
stripping comments, trimming whitespace (leading, trailing, or both),
collapsing lines joined with trailing backslashes, and maybe a few other
things.  All features are optional.  The code is a bit hairy, and I make
no guarantees about performance (except that it probably sucks).  But it
works, and I never have to write all that damn logic again.

I'd vote for keeping fileinput.py relatively simple (adding chop sounds
like a good idea, but anything else is overkill), and putting the
complicated logic in a higher-level module like my text_file.py.  Or
maybe in a higher-level class also provided by fileinput.py.  Whatever.

        Greg
-- 
Greg Ward - software developer                gward@mems-exchange.org
MEMS Exchange / CNRI                           voice: +1-703-262-5376
Reston, Virginia, USA                            fax: +1-703-262-5367


From guido@python.org  Mon Apr 10 14:33:00 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 10 Apr 2000 09:33:00 -0400
Subject: [Patches] Change to attribute error message for instances.
In-Reply-To: Your message of "Mon, 10 Apr 2000 13:06:14 +1000."
 <ECEPKNMJLHAPFFJHDOJBIEIHCIAA.mhammond@skippinet.com.au>
References: <ECEPKNMJLHAPFFJHDOJBIEIHCIAA.mhammond@skippinet.com.au>
Message-ID: <200004101333.JAA29910@eric.cnri.reston.va.us>

> In line with a similar checkin to object.c a while ago, this patch
> gives a more descriptive error message for an attribute error on a
> class instance.  The message now looks like:
> 
> AttributeError: 'Descriptor' instance has no attribute
> 'GetReturnType'
> 
> [Maybe the module name would also be useful, but IMO this is fine
> and a nice improvement]
[...]
> diff -c -2 -r2.83 classobject.c
> *** classobject.c	2000/02/28 15:03:15	2.83
> --- classobject.c	2000/04/10 03:01:08
> ***************
> *** 593,597 ****
>   		v = class_lookup(inst->in_class, name, &class);
>   		if (v == NULL) {
> ! 			PyErr_SetObject(PyExc_AttributeError, name);
>   			return NULL;
>   		}
> --- 593,600 ----
>   		v = class_lookup(inst->in_class, name, &class);
>   		if (v == NULL) {
> ! 			PyErr_Format(PyExc_AttributeError,
> ! 				     "'%.50s' instance has no attribute '%.400s'",
> ! 				     PyString_AsString(inst->in_class->cl_name),
> ! 				     sname);
>   			return NULL;
>   		}

One problem with this: it breaks test_extcall.py!!!  That code was
expecting just the attribute name as the exception message.

I wonder if there's more (user) code that makes a similar exception?

I think we may have to be careful with this (kind of) change...  At
the very least I'll mention it in the list of code breakage on
python.org/1.6/.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From DavidA@ActiveState.com  Mon Apr 10 18:49:04 2000
From: DavidA@ActiveState.com (David Ascher)
Date: Mon, 10 Apr 2000 10:49:04 -0700
Subject: [Patches] Change to attribute error message for instances.
In-Reply-To: <200004101333.JAA29910@eric.cnri.reston.va.us>
Message-ID: <NDBBJPNCJLKKIOBLDOMJEEPOCEAA.DavidA@ActiveState.com>

> One problem with this: it breaks test_extcall.py!!!  That code was
> expecting just the attribute name as the exception message.
>
> I wonder if there's more (user) code that makes a similar exception?

Probably.  IMO, tough.

> I think we may have to be careful with this (kind of) change...  At
> the very least I'll mention it in the list of code breakage on
> python.org/1.6/.

Sure, but I want to point out that this kind of change is what helps newbies
more than almost anything else -- Python's error messages are liked by folks
who have to deal with bad C compilers on a daily basis, but hated by real
newbies.

As far as I know, the content of the message of AttributeError was never
documented, right?

--david



From mal@lemburg.com  Mon Apr 10 19:05:42 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Mon, 10 Apr 2000 20:05:42 +0200
Subject: [Patches] Change to attribute error message for instances.
References: <NDBBJPNCJLKKIOBLDOMJEEPOCEAA.DavidA@ActiveState.com>
Message-ID: <38F217F6.EE171DB2@lemburg.com>

David Ascher wrote:
> 
> > One problem with this: it breaks test_extcall.py!!!  That code was
> > expecting just the attribute name as the exception message.
> >
> > I wonder if there's more (user) code that makes a similar exception?
> 
> Probably.  IMO, tough.
> 
> > I think we may have to be careful with this (kind of) change...  At
> > the very least I'll mention it in the list of code breakage on
> > python.org/1.6/.
> 
> Sure, but I want to point out that this kind of change is what helps newbies
> more than almost anything else -- Python's error messages are liked by folks
> who have to deal with bad C compilers on a daily basis, but hated by real
> newbies.
> 
> As far as I know, the content of the message of AttributeError was never
> documented, right?

Not sure, but aren't the error objects supposed to carry
this kind of additional infomation in error instance attributes ?!
The text should really never be used in other ways than to
write it to a log file or print it to stderr, IMHO...

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From skip@mojam.com (Skip Montanaro)  Mon Apr 10 18:04:43 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Mon, 10 Apr 2000 12:04:43 -0500 (CDT)
Subject: [Patches] Change to attribute error message for instances.
In-Reply-To: <NDBBJPNCJLKKIOBLDOMJEEPOCEAA.DavidA@ActiveState.com>
References: <200004101333.JAA29910@eric.cnri.reston.va.us>
 <NDBBJPNCJLKKIOBLDOMJEEPOCEAA.DavidA@ActiveState.com>
Message-ID: <14578.2475.904738.937642@beluga.mojam.com>

    >> I think we may have to be careful with this (kind of) change...  At
    >> the very least I'll mention it in the list of code breakage on
    >> python.org/1.6/.

    David> Sure, but I want to point out that this kind of change is what
    David> helps newbies more than almost anything else -- Python's error
    David> messages are liked by folks who have to deal with bad C compilers
    David> on a daily basis, but hated by real newbies.

I will re-raise my suggestion that somewhere along the way Python needs
error codes as well as error messages (maybe not until Py3k).  I know it
won't be easy to do, but if people are really parsing exception messages in
their scripts, it suggests a need for error codes to make that job easier.

-- 
Skip Montanaro | http://www.mojam.com/
skip@mojam.com | http://www.musi-cal.com/


From guido@python.org  Mon Apr 10 19:44:01 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 10 Apr 2000 14:44:01 -0400
Subject: [Patches] Getting ready for 1.6 alpha 2
Message-ID: <200004101844.OAA02610@eric.cnri.reston.va.us>

I'm getting ready for the release of alpha 2.  Tomorrow afternoon
(around 5:30pm east coast time) I'm going on vacation for the rest of
the week, followed by a business trip most of the week after.

Obviously, I'm anxious to release a solid alpha tomorrow.

Please, send only simple or essential patches between now and the
release date!

--Guido van Rossum (home page: http://www.python.org/~guido/)


From gstein@lyra.org  Mon Apr 10 20:01:19 2000
From: gstein@lyra.org (Greg Stein)
Date: Mon, 10 Apr 2000 12:01:19 -0700 (PDT)
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <20000410092055.A562@mems-exchange.org>
Message-ID: <Pine.LNX.4.10.10004101200130.11297-100000@nebula.lyra.org>

On Mon, 10 Apr 2000, Greg Ward wrote:
>...
> See the distutils.text_file module for a rather overblown approach to
> this problem: it takes care of stripping newlines, deleting blank lines,
> stripping comments, trimming whitespace (leading, trailing, or both),
> collapsing lines joined with trailing backslashes, and maybe a few other
> things.  All features are optional.  The code is a bit hairy, and I make
> no guarantees about performance (except that it probably sucks).  But it
> works, and I never have to write all that damn logic again.
> 
> I'd vote for keeping fileinput.py relatively simple (adding chop sounds
> like a good idea, but anything else is overkill), and putting the
> complicated logic in a higher-level module like my text_file.py.  Or
> maybe in a higher-level class also provided by fileinput.py.  Whatever.

+1 on adding the new class to fileinput.py. I've implemented similar code
several times... and would like to avoid that :-)

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/



From gward@mems-exchange.org  Mon Apr 10 20:16:27 2000
From: gward@mems-exchange.org (Greg Ward)
Date: Mon, 10 Apr 2000 15:16:27 -0400
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <Pine.LNX.4.10.10004101200130.11297-100000@nebula.lyra.org>; from gstein@lyra.org on Mon, Apr 10, 2000 at 12:01:19PM -0700
References: <20000410092055.A562@mems-exchange.org> <Pine.LNX.4.10.10004101200130.11297-100000@nebula.lyra.org>
Message-ID: <20000410151626.A1168@mems-exchange.org>

On 10 April 2000, Greg Stein said:
> +1 on adding the new class to fileinput.py. I've implemented similar code
> several times... and would like to avoid that :-)

Well, that's one.  Is anyone else interested in seeing a super-beefed up
(possibly over-engineered) "read a `text' file with some subset of the
usual Unix conventions" class added to fileinput.py?

Here's the docstring for my nominee (text_file.py, which you presently
all have in your Python source tree squirreled away in Lib/distutils):

"""Provides a file-like object that takes care of all the things you
   commonly want to do when processing a text file that has some
   line-by-line syntax: strip comments (as long as "#" is your comment
   character), skip blank lines, join adjacent lines by escaping the
   newline (ie. backslash at end of line), strip leading and/or
   trailing whitespace, and collapse internal whitespace.  All of these
   are optional and independently controllable.

   Provides a 'warn()' method so you can generate warning messages that
   report physical line number, even if the logical line in question
   spans multiple physical lines.  Also provides 'unreadline()' for
   implementing line-at-a-time lookahead.

   Constructor is called as:

       TextFile (filename=None, file=None, **options)

   It bombs (RuntimeError) if both 'filename' and 'file' are None;
   'filename' should be a string, and 'file' a file object (or
   something that provides 'readline()' and 'close()' methods).  It is
   recommended that you supply at least 'filename', so that TextFile
   can include it in warning messages.  If 'file' is not supplied,
   TextFile creates its own using the 'open()' builtin.

   The options are all boolean, and affect the value returned by
   'readline()':
     strip_comments [default: true]
       strip from "#" to end-of-line, as well as any whitespace
       leading up to the "#" -- unless it is escaped by a backslash
     lstrip_ws [default: false]
       strip leading whitespace from each line before returning it
     rstrip_ws [default: true]
       strip trailing whitespace (including line terminator!) from
       each line before returning it
     skip_blanks [default: true}
       skip lines that are empty *after* stripping comments and
       whitespace.  (If both lstrip_ws and rstrip_ws are true,
       then some lines may consist of solely whitespace: these will
       *not* be skipped, even if 'skip_blanks' is true.)
     join_lines [default: false]
       if a backslash is the last non-newline character on a line
       after stripping comments and whitespace, join the following line
       to it to form one "logical line"; if N consecutive lines end
       with a backslash, then N+1 physical lines will be joined to
       form one logical line.
     collapse_ws [default: false]  
       after stripping comments and whitespace and joining physical
       lines into logical lines, all internal whitespace (strings of
       whitespace surrounded by non-whitespace characters, and not at
       the beginning or end of the logical line) will be collapsed
       to a single space.

   Note that since 'rstrip_ws' can strip the trailing newline, the
   semantics of 'readline()' must differ from those of the builtin file
   object's 'readline()' method!  In particular, 'readline()' returns
   None for end-of-file: an empty string might just be a blank line (or
   an all-whitespace line), if 'rstrip_ws' is true but 'skip_blanks' is
   not."""

Is this too much?  Not enough?  Silly to put it in fileinput.py when
I'll already have to include it with the Distutils for pre-1.6 Pythons?

        Greg
-- 
Greg Ward - software developer                gward@mems-exchange.org
MEMS Exchange / CNRI                           voice: +1-703-262-5376
Reston, Virginia, USA                            fax: +1-703-262-5367


From guido@python.org  Mon Apr 10 20:27:45 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 10 Apr 2000 15:27:45 -0400
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: Your message of "Mon, 10 Apr 2000 15:16:27 EDT."
 <20000410151626.A1168@mems-exchange.org>
References: <20000410092055.A562@mems-exchange.org> <Pine.LNX.4.10.10004101200130.11297-100000@nebula.lyra.org>
 <20000410151626.A1168@mems-exchange.org>
Message-ID: <200004101927.PAA02984@eric.cnri.reston.va.us>

> Well, that's one.  Is anyone else interested in seeing a super-beefed up
> (possibly over-engineered) "read a `text' file with some subset of the
> usual Unix conventions" class added to fileinput.py?

To me, text_file.py really works at a higher level than fileinput.py.
It deals with lexical analysis issues: comments, line continuations,
whitespace stripping.  But it isn't very configurable: it has fixed
notions of what's a continuation line, what's a comment, and so on,
and you can only define a simple "profile" of options.

For another approach, see Eric Raymond's shlex.py, which is a standard
module since Python 1.5.2.  It deals with splitting into tokens and
handling quotes -- again, using rather fixed rules, although you can
change what's a quote and what's in a word (a step up from
text_file.py).

Together with ConfigParser.py, I think we've got a rather eclectic and
haphazard set of modules.  Before we standardize on any particular
one, we should analyze the needs of our users more!

--Guido van Rossum (home page: http://www.python.org/~guido/)


From Moshe Zadka <mzadka@geocities.com>  Tue Apr 11 07:43:56 2000
From: Moshe Zadka <mzadka@geocities.com> (Moshe Zadka)
Date: Tue, 11 Apr 2000 08:43:56 +0200 (IST)
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <20000410151626.A1168@mems-exchange.org>
Message-ID: <Pine.GSO.4.10.10004110842170.4465-100000@sundial>

I really like the text_file.py. Can't it just be in the standard library
and have a__getitem__ method (or is there one already? I only looked at
the doc you sent). This will at least solve the FAQ about reading a file
line-by-line in a nice way.

--
Moshe Zadka <mzadka@geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com



From gansevle@cs.utwente.nl  Tue Apr 11 09:14:34 2000
From: gansevle@cs.utwente.nl (Fred Gansevles)
Date: Tue, 11 Apr 2000 10:14:34 +0200
Subject: [Patches] (no subject)
Message-ID: <200004110814.KAA28088@localhost.localdomain>

The format of Doc/api/refcount.dat has changed but the parsing-code didn't.
This patch updates the Doc/tools/refcount.py parsing script
----------------------------------------------------------------
I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part

of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
FAQ about legal
I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

----------------------------------------------------------------
Index: refcounts.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Doc/tools/refcounts.py,v
retrieving revision 1.1
diff -r1.1 refcounts.py
36c36
<         if refcount:
---
>         if not refcount in ('', 'null'):


From mal@lemburg.com  Tue Apr 11 11:03:09 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Tue, 11 Apr 2000 12:03:09 +0200
Subject: [Patches] Unicode Patch Set 2000-04-11
Message-ID: <38F2F85D.446DA68E@lemburg.com>

This is a multi-part message in MIME format.
--------------16779387E85086A13F2E7D09
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

This patch changes the semantics of the .splitlines() argument.
This method previously had an argument maxsplit which worked
just like the .split() method's argument of the same name.
For line splitting, maxsplit doesn't make much sense, so instead,
the argument is now called "keepends" and defaults to 0. When 
given and true, line end markers are included in the list entries.

I hope its not too late for this change: .splitlines() isn't
documented anywhere and probably hasn't had too much exposure 
yet -- besides, this is alpha software ;-)

BTW, The change was needed in order to make file .readline() et al. 
methods work with arbitrary string-like objects.


Patch Set Contents:
-------------------

Include/unicodeobject.h:

Changed PyUnicode_Splitlines() maxsplit argument to keepends.
The maxsplit functionality was replaced by the keepends
functionality which allows keeping the line end markers together
with the string.

Objects/stringobject.c:

The maxsplit functionality in .splitlines() was replaced by the keepends
functionality which allows keeping the line end markers together
with the string.

Added support for '%r' % obj: this inserts repr(obj) rather
than str(obj).

Objects/unicodectype.c:

Added a few missing whitespace Unicode char mappings.
Thanks to Brian Hooper.

Objects/unicodeobject.c:

The maxsplit functionality in .splitlines() was replaced by the keepends
functionality which allows keeping the line end markers together
with the string.

Python/bltinmodule.c:

Added special case to unicode(): when being passed a
Unicode object as first argument, return the object as-is.
Raises an exception when given a Unicode object *and* an
encoding name.

Lib/UserString.py:

The maxsplit functionality in .splitlines() was replaced by the keepends
functionality which allows keeping the line end markers together
with the string.

Lib/codecs.py:

Added .writelines(), .readlines() and .readline() to all
codec classes.

Added encoding name attributes to wrapper classes which
allow applications to check the used encoding names.

Lib/test/test_string.py:

Modified .splitlines() tests according to the changes
in stringobject.c.

Lib/test/test_unicode.py:

Modified .splitlines() tests according to the changes
in unicodeobject.c.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------16779387E85086A13F2E7D09
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-11.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-11.patch"

diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Include/unicodeobject.h Python+Unicode/Include/unicodeobject.h
--- CVS-Python/Include/unicodeobject.h	Mon Apr 10 15:41:41 2000
+++ Python+Unicode/Include/unicodeobject.h	Mon Apr 10 21:23:21 2000
@@ -674,7 +674,7 @@
     
 extern DL_IMPORT(PyObject*) PyUnicode_Splitlines(
     PyObject *s,		/* String to split */
-    int maxsplit		/* Maxsplit count */
+    int keepends		/* If true, line end markers are included */
     );		
 
 /* Translate a string by applying a character mapping table to it and
Only in CVS-Python/Lib/test/output: test_zipfile
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Objects/stringobject.c Python+Unicode/Objects/stringobject.c
--- CVS-Python/Objects/stringobject.c	Mon Apr 10 15:47:21 2000
+++ Python+Unicode/Objects/stringobject.c	Tue Apr 11 11:20:24 2000
@@ -2072,11 +2072,11 @@
 
 
 static char splitlines__doc__[] =
-"S.splitlines([maxsplit]]) -> list of strings\n\
+"S.splitlines([keepends]]) -> list of strings\n\
 \n\
 Return a list of the lines in S, breaking at line boundaries.\n\
-If maxsplit is given, at most maxsplit are done. Line breaks are not\n\
-included in the resulting list.";
+Line breaks are not included in the resulting list unless keepends\n\
+is given and true.";
 
 #define SPLIT_APPEND(data, left, right)					\
 	str = PyString_FromStringAndSize(data + left, right - left);	\
@@ -2092,43 +2092,43 @@
 static PyObject*
 string_splitlines(PyStringObject *self, PyObject *args)
 {
-    int maxcount = -1;
     register int i;
     register int j;
     int len;
+    int keepends = 0;
     PyObject *list;
     PyObject *str;
     char *data;
 
-    if (!PyArg_ParseTuple(args, "|i:splitlines", &maxcount))
+    if (!PyArg_ParseTuple(args, "|i:splitlines", &keepends))
         return NULL;
 
     data = PyString_AS_STRING(self);
     len = PyString_GET_SIZE(self);
 
-    if (maxcount < 0)
-        maxcount = INT_MAX;
-
     list = PyList_New(0);
     if (!list)
         goto onError;
 
     for (i = j = 0; i < len; ) {
+	int eol;
+
 	/* Find a line and append it */
 	while (i < len && data[i] != '\n' && data[i] != '\r')
 	    i++;
-	if (maxcount-- <= 0)
-	    break;
-	SPLIT_APPEND(data, j, i);
 
 	/* Skip the line break reading CRLF as one line break */
+	eol = i;
 	if (i < len) {
 	    if (data[i] == '\r' && i + 1 < len &&
 		data[i+1] == '\n')
 		i += 2;
 	    else
 		i++;
+	    if (keepends)
+		eol = i;
 	}
+	SPLIT_APPEND(data, j, eol);
 	j = i;
     }
     if (j < len) {
@@ -2591,7 +2591,10 @@
 					fmt = fmt_start;
 					goto unicode;
 				}
+				if (c == 's')
 				temp = PyObject_Str(v);
+				else
+					temp = PyObject_Repr(v);
 				if (temp == NULL)
 					goto error;
 				if (!PyString_Check(temp)) {
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Objects/unicodectype.c Python+Unicode/Objects/unicodectype.c
--- CVS-Python/Objects/unicodectype.c	Fri Mar 10 23:52:46 2000
+++ Python+Unicode/Objects/unicodectype.c	Tue Apr 11 10:40:08 2000
@@ -633,8 +633,8 @@
 
 #ifndef WANT_WCTYPE_FUNCTIONS
 
-/* Returns 1 for Unicode characters having the type 'WS', 'B' or 'S',
-   0 otherwise. */
+/* Returns 1 for Unicode characters having the bidirectional type
+   'WS', 'B' or 'S' or the category 'Zs', 0 otherwise. */
 
 int _PyUnicode_IsWhitespace(register const Py_UNICODE ch)
 {
@@ -649,6 +649,8 @@
     case 0x001E: /* RECORD SEPARATOR */
     case 0x001F: /* UNIT SEPARATOR */
     case 0x0020: /* SPACE */
+    case 0x0085: /* NEXT LINE */
+    case 0x00A0: /* NO-BREAK SPACE */
     case 0x1680: /* OGHAM SPACE MARK */
     case 0x2000: /* EN QUAD */
     case 0x2001: /* EM QUAD */
@@ -661,7 +663,9 @@
     case 0x2008: /* PUNCTUATION SPACE */
     case 0x2009: /* THIN SPACE */
     case 0x200A: /* HAIR SPACE */
+    case 0x200B: /* ZERO WIDTH SPACE */
     case 0x2028: /* LINE SEPARATOR */
+    case 0x2029: /* PARAGRAPH SEPARATOR */
     case 0x202F: /* NARROW NO-BREAK SPACE */
     case 0x3000: /* IDEOGRAPHIC SPACE */
 	return 1;
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Objects/unicodeobject.c Python+Unicode/Objects/unicodeobject.c
--- CVS-Python/Objects/unicodeobject.c	Mon Apr 10 15:51:10 2000
+++ Python+Unicode/Objects/unicodeobject.c	Tue Apr 11 11:47:03 2000
@@ -2516,7 +2516,7 @@
 }
 
 PyObject *PyUnicode_Splitlines(PyObject *string,
-			       int maxcount)
+			       int keepends)
 {
     register int i;
     register int j;
@@ -2531,29 +2531,29 @@
     data = PyUnicode_AS_UNICODE(string);
     len = PyUnicode_GET_SIZE(string);
 
-    if (maxcount < 0)
-        maxcount = INT_MAX;
-
     list = PyList_New(0);
     if (!list)
         goto onError;
 
     for (i = j = 0; i < len; ) {
+	int eol;
+	
 	/* Find a line and append it */
 	while (i < len && !Py_UNICODE_ISLINEBREAK(data[i]))
 	    i++;
-	if (maxcount-- <= 0)
-	    break;
-	SPLIT_APPEND(data, j, i);
 
 	/* Skip the line break reading CRLF as one line break */
+	eol = i;
 	if (i < len) {
 	    if (data[i] == '\r' && i + 1 < len &&
 		data[i+1] == '\n')
 		i += 2;
 	    else
 		i++;
+	    if (keepends)
+		eol = i;
 	}
+	SPLIT_APPEND(data, j, eol);
 	j = i;
     }
     if (j < len) {
@@ -3785,21 +3785,21 @@
 }
 
 static char splitlines__doc__[] =
-"S.splitlines([maxsplit]]) -> list of strings\n\
+"S.splitlines([keepends]]) -> list of strings\n\
 \n\
 Return a list of the lines in S, breaking at line boundaries.\n\
-If maxsplit is given, at most maxsplit are done. Line breaks are not\n\
-included in the resulting list.";
+Line breaks are not included in the resulting list unless keepends\n\
+is given and true.";
 
 static PyObject*
 unicode_splitlines(PyUnicodeObject *self, PyObject *args)
 {
-    int maxcount = -1;
+    int keepends = 0;
 
-    if (!PyArg_ParseTuple(args, "|i:splitlines", &maxcount))
+    if (!PyArg_ParseTuple(args, "|i:splitlines", &keepends))
         return NULL;
 
-    return PyUnicode_Splitlines((PyObject *)self, maxcount);
+    return PyUnicode_Splitlines((PyObject *)self, keepends);
 }
 
 static
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.py CVS-Python/Python/bltinmodule.c Python+Unicode/Python/bltinmodule.c
--- CVS-Python/Python/bltinmodule.c	Wed Apr  5 22:11:21 2000
+++ Python+Unicode/Python/bltinmodule.c	Tue Apr 11 11:12:50 2000
@@ -165,15 +165,28 @@
 	PyObject *self;
 	PyObject *args;
 {
-	char *s;
+        PyObject *v;
+	const void *buffer;
 	int len;
 	char *encoding = NULL;
 	char *errors = NULL;
 
-	if ( !PyArg_ParseTuple(args, "s#|ss:unicode", &s, &len, 
-			       &encoding, &errors) )
+	if ( !PyArg_ParseTuple(args, "O|ss:unicode", &v, &encoding, &errors) )
 	    return NULL;
-	return PyUnicode_Decode(s, len, encoding, errors);
+	/* Special case: Unicode will stay Unicode */
+	if (PyUnicode_Check(v)) {
+	    if (encoding) {
+		PyErr_SetString(PyExc_TypeError,
+		  "unicode() does not support decoding of Unicode objects");
+		return NULL;
+	    }
+	    Py_INCREF(v);
+	    return v;
+	}
+	/* Read raw data and decode it */
+	if (PyObject_AsReadBuffer(v, &buffer, &len))
+	    return NULL;
+	return PyUnicode_Decode((const char *)buffer, len, encoding, errors);
 }
 
 static char unicode_doc[] =
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Lib/UserString.py Python+Unicode/Lib/UserString.py
--- CVS-Python/Lib/UserString.py	Mon Apr  3 05:51:49 2000
+++ Python+Unicode/Lib/UserString.py	Tue Apr 11 11:36:56 2000
@@ -96,7 +96,7 @@
     def rstrip(self): return self.__class__(self.data.rstrip())
     def split(self, sep=None, maxsplit=-1): 
         return self.data.split(sep, maxsplit)
-    def splitlines(self, maxsplit=-1): return self.data.splitlines(maxsplit)
+    def splitlines(self, keepends=0): return self.data.splitlines(keepends)
     def startswith(self, prefix, start=0, end=sys.maxint): 
         return self.data.startswith(prefix, start, end)
     def strip(self): return self.__class__(self.data.strip())
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Lib/codecs.py Python+Unicode/Lib/codecs.py
--- CVS-Python/Lib/codecs.py	Fri Mar 31 19:22:29 2000
+++ Python+Unicode/Lib/codecs.py	Mon Apr 10 22:29:00 2000
@@ -127,14 +127,19 @@
         self.stream = stream
         self.errors = errors
 
-    def write(self,object):
+    def write(self, object):
 
         """ Writes the object's contents encoded to self.stream.
         """
         data, consumed = self.encode(object,self.errors)
         self.stream.write(data)
 
-    # XXX .writelines() ?
+    def writelines(self, list):
+
+        """ Writes the concatenated list of strings to the stream
+            using .write().
+        """
+        self.write(''.join(list))
         
     def reset(self):
 
@@ -179,7 +184,7 @@
         self.stream = stream
         self.errors = errors
 
-    def read(self,size=-1):
+    def read(self, size=-1):
 
         """ Decodes data from the stream self.stream and returns the
             resulting object.
@@ -221,8 +226,44 @@
             else:
                 return object
 
-    # XXX .readline() and .readlines() (these are hard to implement
-    #     without using buffers for keeping read-ahead data)
+    def readline(self, size=None):
+
+        """ Read one line from the input stream and return the
+            decoded data.
+
+            Note: Unlike the .readlines() method, line breaking must
+            be implemented by the underlying stream's .readline()
+            method -- there is currently no support for line breaking
+            using the codec decoder due to lack of line buffering.
+
+            size, if given, is passed as size argument to the stream's
+            .readline() method.
+            
+        """
+        if size is None:
+            line = self.stream.readline()
+        else:
+            line = self.stream.readline(size)
+        return self.decode(line)[0]
+            
+
+    def readlines(self, sizehint=0):
+
+        """ Read all lines available on the input stream
+            and return them as list of lines.
+
+            Line breaks are implemented using the codec's decoder
+            method and are included in the list entries.
+            
+            sizehint, if given, is passed as size argument to the
+            stream's .read() method.
+
+        """
+        if sizehint is None:
+            data = self.stream.read()
+        else:
+            data = self.stream.read(sizehint)
+        return self.decode(data)[0].splitlines(1)
 
     def reset(self):
 
@@ -247,6 +288,9 @@
 
 class StreamReaderWriter:
 
+    # Optional attributes set by the file wrappers below
+    encoding = 'unknown'
+
     def __init__(self,stream,Reader,Writer,errors='strict'):
 
         """ Creates a StreamReaderWriter instance.
@@ -269,10 +313,22 @@
 
         return self.reader.read(size)
 
+    def readline(size=None):
+
+        return self.reader.readline(size)
+
+    def readlines(sizehint=None):
+
+        return self.reader.readlines(sizehint)
+
     def write(self,data):
 
         return self.writer.write(data)
 
+    def writelines(self,list):
+
+        return self.writer.writelines(list)
+
     def reset(self):
 
         self.reader.reset()
@@ -290,6 +346,10 @@
 
 class StreamRecoder:
 
+    # Optional attributes set by the file wrappers below
+    data_encoding = 'unknown'
+    file_encoding = 'unknown'
+
     def __init__(self,stream,encode,decode,Reader,Writer,errors='strict'):
 
         """ Creates a StreamRecoder instance which implements a two-way
@@ -328,13 +388,34 @@
         data, bytesencoded = self.encode(data, self.errors)
         return data
 
+    def readline(self,size=None):
+
+        if size is None:
+            data = self.reader.readline()
+        else:
+            data = self.reader.readline(size)
+        data, bytesencoded = self.encode(data, self.errors)
+        return data
+
+    def readlines(self,sizehint=None):
+
+        if sizehint is None:
+            data = self.reader.read()
+        else:
+            data = self.reader.read(sizehint)
+        data, bytesencoded = self.encode(data, self.errors)
+        return data.splitlines(1)
+
     def write(self,data):
 
         data, bytesdecoded = self.decode(data, self.errors)
         return self.writer.write(data)
 
-    # .writelines(), .readline() and .readlines() ... see notes
-    # above.
+    def writelines(self,list):
+
+        data = ''.join(list)
+        data, bytesdecoded = self.decode(data, self.errors)
+        return self.writer.write(data)
 
     def reset(self):
 
@@ -380,33 +461,45 @@
     if encoding is None:
         return file
     (e,d,sr,sw) = lookup(encoding)
-    return StreamReaderWriter(file, sr, sw, errors)
+    srw = StreamReaderWriter(file, sr, sw, errors)
+    # Add attributes to simplify introspection
+    srw.encoding = encoding
+    return srw
 
-def EncodedFile(file, input, output=None, errors='strict'):
+def EncodedFile(file, data_encoding, file_encoding=None, errors='strict'):
 
     """ Return a wrapped version of file which provides transparent
         encoding translation.
 
         Strings written to the wrapped file are interpreted according
-        to the given input encoding and then written to the original
-        file as string using the output encoding. The intermediate
-        encoding will usually be Unicode but depends on the specified
-        codecs.
+        to the given data_encoding and then written to the original
+        file as string using file_encoding. The intermediate encoding
+        will usually be Unicode but depends on the specified codecs.
 
-        If output is not given, it defaults to input.
+        Strings are read from the file using file_encoding and then
+        passed back to the caller as string using data_encoding.
+
+        If file_encoding is not given, it defaults to data_encoding.
 
         errors may be given to define the error handling. It defaults
         to 'strict' which causes ValueErrors to be raised in case an
         encoding error occurs.
 
+        data_encoding and file_encoding are added to the wrapped file
+        object as attributes .data_encoding and .file_encoding resp.
+
     """
-    if output is None:
-        output = input
-    encode, decode = lookup(input)[:2]
-    Reader, Writer = lookup(output)[2:]
-    return StreamRecoder(file,
-                         encode,decode,Reader,Writer,
-                         errors)
+    if file_encoding is None:
+        file_encoding = data_encoding
+    encode, decode = lookup(data_encoding)[:2]
+    Reader, Writer = lookup(file_encoding)[2:]
+    sr = StreamRecoder(file,
+                       encode,decode,Reader,Writer,
+                       errors)
+    # Add attributes to simplify introspection
+    sr.data_encoding = data_encoding
+    sr.file_encoding = file_encoding
+    return sr
 
 ### Tests
     
@@ -414,5 +507,8 @@
 
     import sys
     
-    # Make stdout translate Latin-1 into Unicode-Escape
-    sys.stdout = EncodedFile(sys.stdout, 'latin-1', 'unicode-escape')
+    # Make stdout translate Latin-1 output into UTF-8 output
+    sys.stdout = EncodedFile(sys.stdout, 'latin-1', 'utf-8')
+    
+    # Have stdin translate Latin-1 input into UTF-8 input
+    sys.stdin = EncodedFile(sys.stdin, 'utf-8', 'latin-1')
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_string.py Python+Unicode/Lib/test/test_string.py
--- CVS-Python/Lib/test/test_string.py	Mon Mar 20 17:36:30 2000
+++ Python+Unicode/Lib/test/test_string.py	Tue Apr 11 11:20:50 2000
@@ -90,8 +90,7 @@
 test('splitlines', "abc\ndef\r\nghi\n", ['abc', 'def', 'ghi'])
 test('splitlines', "abc\ndef\r\nghi\n\r", ['abc', 'def', 'ghi', ''])
 test('splitlines', "\nabc\ndef\r\nghi\n\r", ['', 'abc', 'def', 'ghi', ''])
-test('splitlines', "\nabc\ndef\r\nghi\n\r", ['', 'abc\012def\015\012ghi\012\015'], 1)
-test('splitlines', "\nabc\ndef\r\nghi\n\r", ['', 'abc', 'def\015\012ghi\012\015'], 2)
+test('splitlines', "\nabc\ndef\r\nghi\n\r", ['\n', 'abc\n', 'def\r\n', 'ghi\n', '\r'], 1)
 
 transtable = '\000\001\002\003\004\005\006\007\010\011\012\013\014\015\016\017\020\021\022\023\024\025\026\027\030\031\032\033\034\035\036\037 !"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`xyzdefghijklmnopqrstuvwxyz{|}~\177\200\201\202\203\204\205\206\207\210\211\212\213\214\215\216\217\220\221\222\223\224\225\226\227\230\231\232\233\234\235\236\237\240\241\242\243\244\245\246\247\250\251\252\253\254\255\256\257\260\261\262\263\264\265\266\267\270\271\272\273\274\275\276\277\300\301\302\303\304\305\306\307\310\311\312\313\314\315\316\317\320\321\322\323\324\325\326\327\330\331\332\333\334\335\336\337\340\341\342\343\344\345\346\347\350\351\352\353\354\355\356\357\360\361\362\363\364\365\366\367\370\371\372\373\374\375\376\377'
 
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x *.txt -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_unicode.py Python+Unicode/Lib/test/test_unicode.py
--- CVS-Python/Lib/test/test_unicode.py	Mon Apr 10 15:52:48 2000
+++ Python+Unicode/Lib/test/test_unicode.py	Tue Apr 11 11:23:02 2000
@@ -212,8 +212,7 @@
 test('splitlines', u"abc\ndef\r\nghi\n", [u'abc', u'def', u'ghi'])
 test('splitlines', u"abc\ndef\r\nghi\n\r", [u'abc', u'def', u'ghi', u''])
 test('splitlines', u"\nabc\ndef\r\nghi\n\r", [u'', u'abc', u'def', u'ghi', u''])
-test('splitlines', u"\nabc\ndef\r\nghi\n\r", [u'', u'abc\012def\015\012ghi\012\015'], 1)
-test('splitlines', u"\nabc\ndef\r\nghi\n\r", [u'', u'abc', u'def\015\012ghi\012\015'], 2)
+test('splitlines', u"\nabc\ndef\r\nghi\n\r", [u'\n', u'abc\n', u'def\r\n', u'ghi\n', u'\r'], 1)
 
 test('translate', u"abababc", u'bbbc', {ord('a'):None})
 test('translate', u"abababc", u'iiic', {ord('a'):None, ord('b'):ord('i')})
Only in CVS-Python/Lib/test: test_zipfile.py

--------------16779387E85086A13F2E7D09--




From Markovitch@ISO.RU  Tue Apr 11 13:11:55 2000
From: Markovitch@ISO.RU (Markovitch Yakov)
Date: Tue, 11 Apr 2000 16:11:55 +0400
Subject: [Patches] Patch for thread_nt.h, radically increasing Python perfomance
Message-ID: <119FEF4BC996D311869E0050044D464D05D6FC@EXCHANGE>

Hello Guido.

Here's the new version of thread_nt.h. More particular, there is a new
version of thread lock that uses kernel object (e.g. semaphore) only in case
of contention; in other case it simply uses interlocked functions, which are
faster by the order of magnitude. It doesn't make much difference without
threads present, but as soon as thread machinery initialised and (mostly)
the interpreter global lock is on, difference becomes tremendous. I've
included a small script, which initialises threads and launches pystone.
With original thread_nt.h, Pystone results with initialised threads are
twofold worse then w/o threads. With the new version, only 10% worse.
I have used this patch for about 6 months (with threaded and non-threaded
applications). It works remarkably well (though I'd desperately prefer
Python was free-threaded; I hope, it will soon).

Here's the legal mumbo-yumbo:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

Here's the diff against the current CVS version:

diff -r2.6 thread_nt.h
5c5
<                         All Rights Reserved
---
>             All Rights Reserved
32a33
> /* Fast NonRecursiveMutex support by Yakov Markovitch, markovitch@iso.ru
*/
37a39,140
> typedef struct NRMUTEX {
>        LONG   owned ;
>        DWORD  thread_id ;
>        HANDLE hevent ;
> 
> } NRMUTEX, *PNRMUTEX ;
> 
> 
> typedef PVOID WINAPI interlocked_cmp_xchg_t(PVOID *dest, PVOID exc, PVOID
comperand) ;
> 
> /* Sorry mate, but we haven't got InterlockedCompareExchange in Win95! */
> static PVOID WINAPI interlocked_cmp_xchg(PVOID *dest, PVOID exc, PVOID
comperand)
> {
> 	static LONG spinlock = 0 ;
> 	PVOID result ;
> 
> 	/* Acqire spinlock (yielding control to other threads if cant aquire
for the moment) */
> 	while(InterlockedExchange(&spinlock, 1)) Sleep(0) ;
> 	result = *dest ;
> 	if (result == comperand)
> 		*dest = exc ;
> 	/* Release spinlock */
> 	spinlock = 0 ;
> 	return result ;
> } ;
> 
> static interlocked_cmp_xchg_t *ixchg ;
> BOOL InitializeNonRecursiveMutex(PNRMUTEX mutex)
> {
> 	if (!ixchg)
> 	{
> 		/* Sorely, Win95 has no InterlockedCompareExchange API
(Win98 has), so we have to use emulation */
> 		HANDLE kernel = GetModuleHandle("kernel32.dll") ;
> 		if (!kernel || (ixchg = (interlocked_cmp_xchg_t
*)GetProcAddress(kernel, "InterlockedCompareExchange")) == NULL)
> 			ixchg = interlocked_cmp_xchg ;
> 	}
> 
> 	mutex->owned = -1 ;  /* No threads have entered NonRecursiveMutex */
> 	mutex->thread_id = 0 ;
> 	mutex->hevent = CreateEvent(NULL, FALSE, FALSE, NULL) ;
> 	return mutex->hevent != NULL ;	/* TRUE if the mutex is created */
> }
> 
> #define InterlockedCompareExchange(dest,exchange,comperand) (ixchg((dest),
(exchange), (comperand)))
> 
> VOID DeleteNonRecursiveMutex(PNRMUTEX mutex)
> {
> 	/* No in-use check */
> 	CloseHandle(mutex->hevent) ;
> 	mutex->hevent = NULL ; /* Just in case */
> }
> 
> DWORD EnterNonRecursiveMutex(PNRMUTEX mutex, BOOL wait)
> {
> 	/* Assume that the thread waits successfully */
> 	DWORD ret ;
> 
> 	/* InterlockedIncrement(&mutex->owned) == 0 means that no thread
currently owns the mutex */
> 	if (!wait)
> 	{
> 		if (InterlockedCompareExchange((PVOID *)&mutex->owned,
(PVOID)0, (PVOID)-1) != (PVOID)-1)
> 			return WAIT_TIMEOUT ;
> 		ret = WAIT_OBJECT_0 ;
> 	}
> 	else
> 		ret = InterlockedIncrement(&mutex->owned) ?
> 			/* Some thread owns the mutex, let's wait... */
> 			WaitForSingleObject(mutex->hevent, INFINITE) :
WAIT_OBJECT_0 ;
> 
> 	mutex->thread_id = GetCurrentThreadId() ; /* We own it */
> 	return ret ;
> }
> 
> BOOL LeaveNonRecursiveMutex(PNRMUTEX mutex)
> {
> 	/* We don't own the mutex */
> 	mutex->thread_id = 0 ;
> 	return
> 		InterlockedDecrement(&mutex->owned) < 0 ||
> 		SetEvent(mutex->hevent) ; /* Other threads are waiting, wake
one on them up */
> }
> 
> PNRMUTEX AllocNonRecursiveMutex()
> {
> 	PNRMUTEX mutex = (PNRMUTEX)malloc(sizeof(NRMUTEX)) ;
> 	if (mutex && !InitializeNonRecursiveMutex(mutex))
> 	{
> 		free(mutex) ;
> 		mutex = NULL ;
> 	}
> 	return mutex ;
> }
> 
> void FreeNonRecursiveMutex(PNRMUTEX mutex)
> {
> 	if (mutex)
> 	{
> 		DeleteNonRecursiveMutex(mutex) ;
> 		free(mutex) ;
> 	}
> }
> 
65c168
<  
---
> 
68c171
< 		dprintf(("%ld: PyThread_start_new_thread succeeded: %ld\n",
PyThread_get_thread_ident(), rv));
---
> 		dprintf(("%ld: PyThread_start_new_thread succeeded: %ld\n",
PyThread_get_thread_ident(), aThreadId));
82c185
<         
---
> 
136c239
< 	HANDLE aLock;
---
> 	PNRMUTEX aLock;
142,146c245
< 		aLock = CreateSemaphore(NULL,           /* Security
attributes          */
<                          1,                     /* Initial value
*/
<                          1,                     /* Maximum value
*/
<                          NULL);       
<   /* Name of semaphore            */
---
> 	aLock = AllocNonRecursiveMutex() ;
157c256
< 	CloseHandle((HANDLE) aLock);
---
> 	FreeNonRecursiveMutex(aLock) ;
168,169c267
< 	int success = 1;
< 	DWORD waitResult;
---
> 	int success ;
173,177c271
< 	waitResult = WaitForSingleObject((HANDLE) aLock, (waitflag == 1 ?
INFINITE : 0));
< 
< 	if (waitResult != WAIT_OBJECT_0) {
< 		success = 0;    /* We failed */
< 	}
---
> 	success = aLock && EnterNonRecursiveMutex((PNRMUTEX) aLock,
(waitflag == 1 ? INFINITE : 0)) == WAIT_OBJECT_0 ;
188,192c282
< 	if (!ReleaseSemaphore(
<                         (HANDLE) aLock,                         /* Handle
of semaphore                          */
<                         1,                                      /*
increment count by one                       */
<                         NULL))                                  /* not
interested in previous count             */
< 		{
---
> 	if (!(aLock && LeaveNonRecursiveMutex((PNRMUTEX) aLock)))
194d283
< 		}




Yours sincerely 
                Yakov Markovitch, Lead Programmer 
                InterSoft Lab 
                Phone: 7 095 218-4014 
                e-mail: markovitch@iso.ru 


begin 600 thread_nt.h
M+RHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ*BHJ
M*BHJ*BHJ*BHJ*BHJ*BHJ#0I#;W!Y<FEG:'0@,3DY,2TQ.3DU(&)Y(%-T:6-H
M=&EN9R!-871H96UA=&ES8V@@0V5N=')U;2P@06US=&5R9&%M+`T*5&AE($YE
M=&AE<FQA;F1S+@T*#0H@("`@("`@("`@("!!;&P@4FEG:'1S(%)E<V5R=F5D
M#0H-"E!E<FUI<W-I;VX@=&\@=7-E+"!C;W!Y+"!M;V1I9GDL(&%N9"!D:7-T
M<FEB=71E('1H:7,@<V]F='=A<F4@86YD(&ET<PT*9&]C=6UE;G1A=&EO;B!F
M;W(@86YY('!U<G!O<V4@86YD('=I=&AO=70@9F5E(&ES(&AE<F5B>2!G<F%N
M=&5D+`T*<')O=FED960@=&AA="!T:&4@86)O=F4@8V]P>7)I9VAT(&YO=&EC
M92!A<'!E87(@:6X@86QL(&-O<&EE<R!A;F0@=&AA=`T*8F]T:"!T:&%T(&-O
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M5$A%34%425-#2`T*0T5.5%)532!/4B!#3E))($)%($Q)04),12!&3U(@04Y9
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M<F]C97-S+F@^#0H-"G1Y<&5D968@<W1R=6-T($Y2355415@@>PT*("`@("`@
M($Q/3D<@("!O=VYE9"`[#0H@("`@("`@1%=/4D0@('1H<F5A9%]I9"`[#0H@
M("`@("`@2$%.1$Q%(&AE=F5N="`[#0H-"GT@3E)-551%6"P@*E!.4DU55$58
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M1R!S<&EN;&]C:R`](#`@.PT*"5!63TE$(')E<W5L="`[#0H-"@DO*B!!8W%I
M<F4@<W!I;FQO8VL@*'EI96QD:6YG(&-O;G1R;VP@=&\@;W1H97(@=&AR96%D
M<R!I9B!C86YT(&%Q=6ER92!F;W(@=&AE(&UO;65N="D@*B\-"@EW:&EL92A)
M;G1E<FQO8VME9$5X8VAA;F=E*"9S<&EN;&]C:RP@,2DI(%-L965P*#`I(#L-
M"@ER97-U;'0@/2`J9&5S="`[#0H):68@*')E<W5L="`]/2!C;VUP97)A;F0I
M#0H)"2ID97-T(#T@97AC(#L-"@DO*B!296QE87-E('-P:6YL;V-K("HO#0H)
M<W!I;FQO8VL@/2`P(#L-"@ER971U<FX@<F5S=6QT(#L-"GT@.PT*#0IS=&%T
M:6,@:6YT97)L;V-K961?8VUP7WAC:&=?="`J:7AC:&<@.PT*0D]/3"!);FET
M:6%L:7IE3F]N4F5C=7)S:79E375T97@H4$Y2355415@@;75T97@I#0I[#0H)
M:68@*"%I>&-H9RD-"@E[#0H)"2\J(%-O<F5L>2P@5VEN.34@:&%S(&YO($EN
M=&5R;&]C:V5D0V]M<&%R945X8VAA;F=E($%022`H5VEN.3@@:&%S*2P@<V\@
M=V4@:&%V92!T;R!U<V4@96UU;&%T:6]N("HO#0H)"4A!3D1,12!K97)N96P@
M/2!'971-;V1U;&5(86YD;&4H(FME<FYE;#,R+F1L;"(I(#L-"@D):68@*"%K
M97)N96P@?'P@*&EX8VAG(#T@*&EN=&5R;&]C:V5D7V-M<%]X8VAG7W0@*BE'
M9710<F]C061D<F5S<RAK97)N96PL("));G1E<FQO8VME9$-O;7!A<F5%>&-H
M86YG92(I*2`]/2!.54Q,*0T*"0D):7AC:&<@/2!I;G1E<FQO8VME9%]C;7!?
M>&-H9R`[#0H)?0T*#0H);75T97@M/F]W;F5D(#T@+3$@.R`@+RH@3F\@=&AR
M96%D<R!H879E(&5N=&5R960@3F]N4F5C=7)S:79E375T97@@*B\-"@EM=71E
M>"T^=&AR96%D7VED(#T@,"`[#0H);75T97@M/FAE=F5N="`]($-R96%T945V
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M<V4@8VAE8VL@*B\-"@E#;&]S94AA;F1L92AM=71E>"T^:&5V96YT*2`[#0H)
M;75T97@M/FAE=F5N="`]($Y53$P@.R`O*B!*=7-T(&EN(&-A<V4@*B\-"GT-
M"@T*1%=/4D0@16YT97).;VY296-U<G-I=F5-=71E>"A03E)-551%6"!M=71E
M>"P@0D]/3"!W86ET*0T*>PT*"2\J($%S<W5M92!T:&%T('1H92!T:')E860@
M=V%I=',@<W5C8V5S<V9U;&QY("HO#0H)1%=/4D0@<F5T(#L-"@T*"2\J($EN
M=&5R;&]C:V5D26YC<F5M96YT*"9M=71E>"T^;W=N960I(#T](#`@;65A;G,@
M=&AA="!N;R!T:')E860@8W5R<F5N=&QY(&]W;G,@=&AE(&UU=&5X("HO#0H)
M:68@*"%W86ET*0T*"7L-"@D):68@*$EN=&5R;&]C:V5D0V]M<&%R945X8VAA
M;F=E*"A05D])1"`J*29M=71E>"T^;W=N960L("A05D])1"DP+"`H4%9/240I
M+3$I("$]("A05D])1"DM,2D-"@D)"7)E='5R;B!704E47U1)345/550@.PT*
M"0ER970@/2!704E47T]"2D5#5%\P(#L-"@E]#0H)96QS90T*"0ER970@/2!)
M;G1E<FQO8VME9$EN8W)E;65N="@F;75T97@M/F]W;F5D*2`_#0H)"0DO*B!3
M;VUE('1H<F5A9"!O=VYS('1H92!M=71E>"P@;&5T)W,@=V%I="XN+B`J+PT*
M"0D)5V%I=$9O<E-I;F=L94]B:F5C="AM=71E>"T^:&5V96YT+"!)3D9)3DE4
M12D@.B!704E47T]"2D5#5%\P(#L-"@T*"6UU=&5X+3YT:')E861?:60@/2!'
M971#=7)R96YT5&AR96%D260H*2`[("\J(%=E(&]W;B!I="`J+PT*"7)E='5R
M;B!R970@.PT*?0T*#0I"3T],($QE879E3F]N4F5C=7)S:79E375T97@H4$Y2
M355415@@;75T97@I#0I[#0H)+RH@5V4@9&]N)W0@;W=N('1H92!M=71E>"`J
M+PT*"6UU=&5X+3YT:')E861?:60@/2`P(#L-"@ER971U<FX-"@D)26YT97)L
M;V-K961$96-R96UE;G0H)FUU=&5X+3YO=VYE9"D@/"`P('Q\#0H)"5-E=$5V
M96YT*&UU=&5X+3YH979E;G0I(#L@+RH@3W1H97(@=&AR96%D<R!A<F4@=V%I
M=&EN9RP@=V%K92!O;F4@;VX@=&AE;2!U<"`J+PT*?0T*#0I03E)-551%6"!!
M;&QO8TYO;E)E8W5R<VEV94UU=&5X*"D-"GL-"@E03E)-551%6"!M=71E>"`]
M("A03E)-551%6"EM86QL;V,H<VEZ96]F*$Y2355415@I*2`[#0H):68@*&UU
M=&5X("8F("%);FET:6%L:7IE3F]N4F5C=7)S:79E375T97@H;75T97@I*0T*
M"7L-"@D)9G)E92AM=71E>"D@.PT*"0EM=71E>"`]($Y53$P@.PT*"7T-"@ER
M971U<FX@;75T97@@.PT*?0T*#0IV;VED($9R965.;VY296-U<G-I=F5-=71E
M>"A03E)-551%6"!M=71E>"D-"GL-"@EI9B`H;75T97@I#0H)>PT*"0E$96QE
M=&5.;VY296-U<G-I=F5-=71E>"AM=71E>"D@.PT*"0EF<F5E*&UU=&5X*2`[
M#0H)?0T*?0T*#0IL;VYG(%!Y5&AR96%D7V=E=%]T:')E861?:61E;G0H=F]I
M9"D[#0H-"B\J#0H@*B!#:&%N9V4@86QL(&AE861E<G,@=&\@<'5R92!!3E-)
M(&%S(&YO(&]N92!W:6QL('5S92!+)E(@<W1Y;&4@;VX@86X-"B`J($Y4#0H@
M*B\-"@T*+RH-"B`J($EN:71I86QI>F%T:6]N(&]F('1H92!#('!A8VMA9V4L
M('-H;W5L9"!N;W0@8F4@;F5E9&5D+@T*("HO#0IS=&%T:6,@=F]I9"!0>51H
M<F5A9%]?:6YI=%]T:')E860H=F]I9"D-"GL-"GT-"@T*+RH-"B`J(%1H<F5A
M9"!S=7!P;W)T+@T*("HO#0II;G0@4'E4:')E861?<W1A<G1?;F5W7W1H<F5A
M9"AV;VED("@J9G5N8RDH=F]I9"`J*2P@=F]I9"`J87)G*0T*>PT*"6QO;F<@
M<G8[#0H):6YT('-U8V-E<W,@/2`P.PT*#0H)9'!R:6YT9B@H(B5L9#H@4'E4
M:')E861?<W1A<G1?;F5W7W1H<F5A9"!C86QL961<;B(L(%!Y5&AR96%D7V=E
M=%]T:')E861?:61E;G0H*2DI.PT*"6EF("@A:6YI=&EA;&EZ960I#0H)"5!Y
M5&AR96%D7VEN:71?=&AR96%D*"D[#0H-"@ER=B`](%]B96=I;G1H<F5A9"AF
M=6YC+"`P+"!A<F<I.R`O*B!U<V4@9&5F875L="!S=&%C:R!S:7IE("HO#0H-
M"@EI9B`H<G8@(3T@+3$I('L-"@D)<W5C8V5S<R`](#$[#0H)"61P<FEN=&8H
M*"(E;&0Z(%!Y5&AR96%D7W-T87)T7VYE=U]T:')E860@<W5C8V5E9&5D.B`E
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M:61E;G0H=F]I9"D-"GL-"@EI9B`H(6EN:71I86QI>F5D*0T*"0E0>51H<F5A
M9%]I;FET7W1H<F5A9"@I.PT*#0H)<F5T=7)N($=E=$-U<G)E;G14:')E861)
M9"@I.PT*?0T*#0IS=&%T:6,@=F]I9"!D;U]0>51H<F5A9%]E>&ET7W1H<F5A
M9"AI;G0@;F]?8VQE86YU<"D-"GL-"@ED<')I;G1F*"@B)6QD.B!0>51H<F5A
M9%]E>&ET7W1H<F5A9"!C86QL961<;B(L(%!Y5&AR96%D7V=E=%]T:')E861?
M:61E;G0H*2DI.PT*"6EF("@A:6YI=&EA;&EZ960I#0H)"6EF("AN;U]C;&5A
M;G5P*0T*"0D)7V5X:70H,"D[#0H)"65L<V4-"@D)"65X:70H,"D[#0H)7V5N
M9'1H<F5A9"@I.PT*?0T*#0IV;VED(%!Y5&AR96%D7V5X:71?=&AR96%D*'9O
M:60I#0I[#0H)9&]?4'E4:')E861?97AI=%]T:')E860H,"D[#0I]#0H-"G9O
M:60@4'E4:')E861?7V5X:71?=&AR96%D*'9O:60I#0I[#0H)9&]?4'E4:')E
M861?97AI=%]T:')E860H,2D[#0I]#0H-"B-I9FYD968@3D]?15A)5%]04D]'
M#0IS=&%T:6,@=F]I9"!D;U]0>51H<F5A9%]E>&ET7W!R;V<H:6YT('-T871U
M<RP@:6YT(&YO7V-L96%N=7`I#0I[#0H)9'!R:6YT9B@H(E!Y5&AR96%D7V5X
M:71?<')O9R@E9"D@8V%L;&5D7&XB+"!S=&%T=7,I*3L-"@EI9B`H(6EN:71I
M86QI>F5D*0T*"0EI9B`H;F]?8VQE86YU<"D-"@D)"5]E>&ET*'-T871U<RD[
M#0H)"65L<V4-"@D)"65X:70H<W1A='5S*3L-"GT-"@T*=F]I9"!0>51H<F5A
M9%]E>&ET7W!R;V<H:6YT('-T871U<RD-"GL-"@ED;U]0>51H<F5A9%]E>&ET
M7W!R;V<H<W1A='5S+"`P*3L-"GT-"@T*=F]I9"!0>51H<F5A9%]?97AI=%]P
M<F]G(%]0,2AI;G0@<W1A='5S*0T*>PT*"61O7U!Y5&AR96%D7V5X:71?<')O
M9RAS=&%T=7,L(#$I.PT*?0T*(V5N9&EF("\J($Y/7T58251?4%)/1R`J+PT*
M#0HO*@T*("H@3&]C:R!S=7!P;W)T+B!)="!H87,@=&]O(&)E(&EM<&QE;65N
M=&5D(&%S('-E;6%P:&]R97,N#0H@*B!)(%M$86==('1R:65D('1O(&EM<&QE
M;65N="!I="!W:71H(&UU=&5X(&)U="!)(&-O=6QD(&9I;F0@82!W87D@=&\-
M"B`J('1E;&P@=VAE=&AE<B!A('1H<F5A9"!A;')E861Y(&]W;B!T:&4@;&]C
M:R!O<B!N;W0N#0H@*B\-"E!Y5&AR96%D7W1Y<&5?;&]C:R!0>51H<F5A9%]A
M;&QO8V%T95]L;V-K*'9O:60I#0I[#0H)4$Y2355415@@84QO8VL[#0H-"@ED
M<')I;G1F*"@B4'E4:')E861?86QL;V-A=&5?;&]C:R!C86QL961<;B(I*3L-
M"@EI9B`H(6EN:71I86QI>F5D*0T*"0E0>51H<F5A9%]I;FET7W1H<F5A9"@I
M.PT*#0H)84QO8VL@/2!!;&QO8TYO;E)E8W5R<VEV94UU=&5X*"D@.PT*#0H)
M9'!R:6YT9B@H(B5L9#H@4'E4:')E861?86QL;V-A=&5?;&]C:R@I("T^("5L
M>%QN(BP@4'E4:')E861?9V5T7W1H<F5A9%]I9&5N="@I+"`H;&]N9REA3&]C
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M:&4@;&]C:R!W87,@86-Q=6ER960-"B`J#0H@*B!A;F0@,"!I9B!T:&4@;&]C
M:R!W87,@;F]T(&%C<75I<F5D+B!4:&ES(&UE86YS(&$@,"!I<R!R971U<FYE
M9`T*("H@:68@=&AE(&QO8VL@:&%S(&%L<F5A9'D@8F5E;B!A8W%U:7)E9"!B
M>2!T:&ES('1H<F5A9"$-"B`J+PT*:6YT(%!Y5&AR96%D7V%C<75I<F5?;&]C
M:RA0>51H<F5A9%]T>7!E7VQO8VL@84QO8VLL(&EN="!W86ET9FQA9RD-"GL-
M"@EI;G0@<W5C8V5S<R`[#0H-"@ED<')I;G1F*"@B)6QD.B!0>51H<F5A9%]A
M8W%U:7)E7VQO8VLH)6QX+"`E9"D@8V%L;&5D7&XB+"!0>51H<F5A9%]G971?
M=&AR96%D7VED96YT*"DL*&QO;F<I84QO8VLL('=A:71F;&%G*2D[#0H-"@ES
M=6-C97-S(#T@84QO8VL@)B8@16YT97).;VY296-U<G-I=F5-=71E>"@H4$Y2
M355415@I(&%,;V-K+"`H=V%I=&9L86<@/3T@,2`_($E.1DE.251%(#H@,"DI
M(#T](%=!251?3T)*14-47S`@.PT*#0H)9'!R:6YT9B@H(B5L9#H@4'E4:')E
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M;V-K*2D[#0H-"@EI9B`H(2AA3&]C:R`F)B!,96%V94YO;E)E8W5R<VEV94UU
M=&5X*"A03E)-551%6"D@84QO8VLI*2D-"@D)9'!R:6YT9B@H(B5L9#H@0V]U
M;&0@;F]T(%!Y5&AR96%D7W)E;&5A<V5?;&]C:R@E;'@I(&5R<F]R.B`E;%QN
M(BP@4'E4:')E861?9V5T7W1H<F5A9%]I9&5N="@I+"`H;&]N9REA3&]C:RP@
M1V5T3&%S=$5R<F]R*"DI*3L-"GT-"@T*+RH-"B`J(%-E;6%P:&]R92!S=7!P
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M96UA<&AO<F4@/2!#<F5A=&5396UA<&AO<F4H($Y53$PL"2`@+RH@4V5C=7)I
M='D@871T<FEB=71E<PD@("HO#0H@("`@("`@("`@("`@("`@("`@("`@("`@
M("`@("`@("`@("`@('9A;'5E+"`@("`@("`@+RH@26YI=&EA;"!V86QU90D)
M("`J+PT*("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("!)
M3E1?34%8+"`@("`@("\J($UA>&EM=6T@=F%L=64)"2`@*B\-"B`@("`@("`@
M("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@3E5,3"D["2`@("`O*B!.
M86UE(&]F('-E;6%P:&]R90D@("HO#0H-"@ED<')I;G1F*"@B)6QD.B!0>51H
M<F5A9%]A;&QO8V%T95]S96UA*"D@+3X@)6QX7&XB+"!0>51H<F5A9%]G971?
M=&AR96%D7VED96YT*"DL("AL;VYG*6%396UA<&AO<F4I*3L-"@T*"7)E='5R
M;B`H4'E4:')E861?='EP95]S96UA*2!A4V5M87!H;W)E.PT*?0T*#0IV;VED
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M7W1H<F5A9%]I9&5N="@I+"`H;&]N9REA4V5M87!H;W)E*2D[#0H-"@EW86ET
M4F5S=6QT(#T@5V%I=$9O<E-I;F=L94]B:F5C="@@*$A!3D1,12D@85-E;6%P
M:&]R92P@24Y&24Y)5$4I.PT*#0H)9'!R:6YT9B@H(B5L9#H@4'E4:')E861?
M9&]W;E]S96UA*"5L>"D@<F5T=7)N.B`E;%QN(BP@4'E4:')E861?9V5T7W1H
M<F5A9%]I9&5N="@I+"AL;VYG*2!A4V5M87!H;W)E+"!W86ET4F5S=6QT*2D[
M#0H)<F5T=7)N(#`[#0I]#0H-"G9O:60@4'E4:')E861?=7!?<V5M82A0>51H
M<F5A9%]T>7!E7W-E;6$@85-E;6%P:&]R92D-"GL-"@E296QE87-E4V5M87!H
M;W)E*`T*"0DH2$%.1$Q%*2!A4V5M87!H;W)E+`D@("`@+RH@2&%N9&QE(&]F
M('-E;6%P:&]R90D)*B\-"@D),2P)"2\J(&EN8W)E;65N="!C;W5N="!B>2!O
M;F4)("`@("HO#0H)"4Y53$PI.PD)("`@("\J(&YO="!I;G1E<F5S=&5D(&EN
M('!R979I;W5S(&-O=6YT"2`@("`J+PT*#0H)9'!R:6YT9B@H(B5L9#H@4'E4
M:')E861?=7!?<V5M82@E;'@I7&XB+"!0>51H<F5A9%]G971?=&AR96%D7VED
?96YT*"DL("AL;VYG*6%396UA<&AO<F4I*3L-"GT-"@==
`
end

begin 600 thr_stone.py
M:6UP;W)T('1H<F5A9`T*(R!3=&%R="!E;7!T>2!T:')E860@=&\@:6YI=&EA
M;&ES92!T:')E860@;65C:&%N:6-S("AA;F0@9VQO8F%L(&QO8VLA*0T*(R!4
M:&ES('1H<F5A9"!W:6QL(&9I;FES:"!I;6UE9&EA=&5L>2!T:'5S('=O;B=T
M(&UA:V4@;75C:"!I;F9L=65N8V4@;VX-"B,@=&5S="!R97-U;'1S(&)Y(&ET
M<V5L9BP@;VYL>2!B>2!T:&%T(&9A8W0@=&AA="!I="!I;FET:6%L:7-E<R!G
M;&]B86P@;&]C:PT*=&AR96%D+G-T87)T7VYE=U]T:')E860H;&%M8F1A(#H@
M,2P@*"DI#0H-"FEM<&]R="!T97-T+G!Y<W1O;F4-"G1E<W0N<'ES=&]N92YM
'86EN*"D-"@==
`
end


From gansevle@cs.utwente.nl  Tue Apr 11 13:25:36 2000
From: gansevle@cs.utwente.nl (Fred Gansevles)
Date: Tue, 11 Apr 2000 14:25:36 +0200
Subject: [Patches] Please review before applying
Message-ID: <200004111225.OAA31234@localhost.localdomain>

For the reviewer(s):

Most of the changes are simple translations, i.e.:

    string.split (s, sep)	--> s.split (sep)

Some of the changes also take the context into account, see:
    - cgi.py
    - formatter.py
    - imaplib.py

Both 'make test' and 'Lib/compileall.py Lib' produced no error-output
so I'd expect no big problems.

The following migt go into the csv-log:
________________________________________________________________________
This (mega-)patch updates all the modules in the standard-module (all
Lib/*.py files) to use the new string-methods instead of the old string
module.

----------------------------------------------------------------
I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part

of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
FAQ about legal
I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

----------------------------------------------------------------
Index: Lib/BaseHTTPServer.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/BaseHTTPServer.py,v
retrieving revision 1.9
diff -c -r1.9 BaseHTTPServer.py
*** BaseHTTPServer.py	1999/10/26 13:01:36	1.9
--- BaseHTTPServer.py	2000/04/11 11:56:38
***************
*** 67,73 ****
  import sys
  import time
  import socket # For gethostbyaddr()
- import string
  import mimetools
  import SocketServer
  
--- 67,72 ----
***************
*** 212,218 ****
      """
  
      # The Python system version, truncated to its first component.
!     sys_version = "Python/" + string.split(sys.version)[0]
  
      # The server software version.  You may want to override this.
      # The format is multiple whitespace-separated strings,
--- 211,217 ----
      """
  
      # The Python system version, truncated to its first component.
!     sys_version = "Python/" + sys.version.split[0]
  
      # The server software version.  You may want to override this.
      # The format is multiple whitespace-separated strings,
***************
*** 237,243 ****
          elif requestline[-1:] == '\n':
              requestline = requestline[:-1]
          self.requestline = requestline
!         words = string.split(requestline)
          if len(words) == 3:
              [command, path, version] = words
              if version[:5] != 'HTTP/':
--- 236,242 ----
          elif requestline[-1:] == '\n':
              requestline = requestline[:-1]
          self.requestline = requestline
!         words = requestline.split()
          if len(words) == 3:
              [command, path, version] = words
              if version[:5] != 'HTTP/':
***************
*** 485,491 ****
      """
  
      if sys.argv[1:]:
!         port = string.atoi(sys.argv[1])
      else:
          port = 8000
      server_address = ('', port)
--- 484,490 ----
      """
  
      if sys.argv[1:]:
!         port = int (sys.argv[1])
      else:
          port = 8000
      server_address = ('', port)
Index: Lib/CGIHTTPServer.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/CGIHTTPServer.py,v
retrieving revision 1.12
diff -c -r1.12 CGIHTTPServer.py
*** CGIHTTPServer.py	1999/10/18 13:43:44	1.12
--- CGIHTTPServer.py	2000/04/11 11:56:38
***************
*** 82,93 ****
      def run_cgi(self):
          """Execute a CGI script."""
          dir, rest = self.cgi_info
!         i = string.rfind(rest, '?')
          if i >= 0:
              rest, query = rest[:i], rest[i+1:]
          else:
              query = ''
!         i = string.find(rest, '/')
          if i >= 0:
              script, rest = rest[:i], rest[i:]
          else:
--- 82,93 ----
      def run_cgi(self):
          """Execute a CGI script."""
          dir, rest = self.cgi_info
!         i = rest.rfind('?')
          if i >= 0:
              rest, query = rest[:i], rest[i+1:]
          else:
              query = ''
!         i = rest.find('/')
          if i >= 0:
              script, rest = rest[:i], rest[i:]
          else:
***************
*** 149,166 ****
              accept = []
              for line in self.headers.getallmatchingheaders('accept'):
                  if line[:1] in string.whitespace:
!                     accept.append(string.strip(line))
                  else:
!                     accept = accept + string.split(line[7:], ',')
!             env['HTTP_ACCEPT'] = string.joinfields(accept, ',')
              ua = self.headers.getheader('user-agent')
              if ua:
                  env['HTTP_USER_AGENT'] = ua
              co = filter(None, self.headers.getheaders('cookie'))
              if co:
!                 env['HTTP_COOKIE'] = string.join(co, ', ')
              # XXX Other HTTP_* headers
!             decoded_query = string.replace(query, '+', ' ')
              try:
                  os.setuid(nobody)
              except os.error:
--- 149,166 ----
              accept = []
              for line in self.headers.getallmatchingheaders('accept'):
                  if line[:1] in string.whitespace:
!                     accept.append(line.strip())
                  else:
!                     accept = accept + line[7:].split(',')
!             env['HTTP_ACCEPT'] = ','.join(accept)
              ua = self.headers.getheader('user-agent')
              if ua:
                  env['HTTP_USER_AGENT'] = ua
              co = filter(None, self.headers.getheaders('cookie'))
              if co:
!                 env['HTTP_COOKIE'] = ', '.join(co)
              # XXX Other HTTP_* headers
!             decoded_query = query.replace('+', ' ')
              try:
                  os.setuid(nobody)
              except os.error:
Index: Lib/ConfigParser.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/ConfigParser.py,v
retrieving revision 1.17
diff -c -r1.17 ConfigParser.py
*** ConfigParser.py	2000/03/03 20:43:57	1.17
--- ConfigParser.py	2000/04/11 11:56:38
***************
*** 259,265 ****
          depth = 0                       
          while depth < 10:               # Loop through this until it's done
              depth = depth + 1
!             if string.find(value, "%(") >= 0:
                  try:
                      value = value % d
                  except KeyError, key:
--- 259,265 ----
          depth = 0                       
          while depth < 10:               # Loop through this until it's done
              depth = depth + 1
!             if value.find("%(") >= 0:
                  try:
                      value = value % d
                  except KeyError, key:
***************
*** 271,290 ****
          return conv(self.get(section, option))
  
      def getint(self, section, option):
!         return self.__get(section, string.atoi, option)
  
      def getfloat(self, section, option):
!         return self.__get(section, string.atof, option)
  
      def getboolean(self, section, option):
          v = self.get(section, option)
!         val = string.atoi(v)
          if val not in (0, 1):
              raise ValueError, 'Not a boolean: %s' % v
          return val
  
      def optionxform(self, optionstr):
!         return string.lower(optionstr)
  
      #
      # Regular expressions for parsing section headers and options.  Note a
--- 271,290 ----
          return conv(self.get(section, option))
  
      def getint(self, section, option):
!         return self.__get(section, int, option)
  
      def getfloat(self, section, option):
!         return self.__get(section, float, option)
  
      def getboolean(self, section, option):
          v = self.get(section, option)
!         val = int(v)
          if val not in (0, 1):
              raise ValueError, 'Not a boolean: %s' % v
          return val
  
      def optionxform(self, optionstr):
!         return optionstr.lower()
  
      #
      # Regular expressions for parsing section headers and options.  Note a
***************
*** 324,337 ****
                  break
              lineno = lineno + 1
              # comment or blank line?
!             if string.strip(line) == '' or line[0] in '#;':
                  continue
!             if string.lower(string.split(line)[0]) == 'rem' \
                 and line[0] in "rR":      # no leading whitespace
                  continue
              # continuation line?
              if line[0] in ' \t' and cursect is not None and optname:
!                 value = string.strip(line)
                  if value:
                      cursect[optname] = cursect[optname] + '\n ' + value
              # a section header or option header?
--- 324,337 ----
                  break
              lineno = lineno + 1
              # comment or blank line?
!             if line.strip() == '' or line[0] in '#;':
                  continue
!             if line.split()[0].lower() == 'rem' \
                 and line[0] in "rR":      # no leading whitespace
                  continue
              # continuation line?
              if line[0] in ' \t' and cursect is not None and optname:
!                 value = line.strip()
                  if value:
                      cursect[optname] = cursect[optname] + '\n ' + value
              # a section header or option header?
***************
*** 357,370 ****
                      mo = self.OPTCRE.match(line)
                      if mo:
                          optname, vi, optval = mo.group('option', 'vi', 'value')
!                         optname = string.lower(optname)
                          if vi in ('=', ':') and ';' in optval:
                              # ';' is a comment delimiter only if it follows
                              # a spacing character
!                             pos = string.find(optval, ';')
                              if pos and optval[pos-1] in string.whitespace:
                                  optval = optval[:pos]
!                         optval = string.strip(optval)
                          # allow empty values
                          if optval == '""':
                              optval = ''
--- 357,370 ----
                      mo = self.OPTCRE.match(line)
                      if mo:
                          optname, vi, optval = mo.group('option', 'vi', 'value')
!                         optname = optname.lower()
                          if vi in ('=', ':') and ';' in optval:
                              # ';' is a comment delimiter only if it follows
                              # a spacing character
!                             pos = optval.find(';')
                              if pos and optval[pos-1] in string.whitespace:
                                  optval = optval[:pos]
!                         optval = optval.strip()
                          # allow empty values
                          if optval == '""':
                              optval = ''
Index: Lib/MimeWriter.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/MimeWriter.py,v
retrieving revision 1.4
diff -c -r1.4 MimeWriter.py
*** MimeWriter.py	1998/04/23 13:34:57	1.4
--- MimeWriter.py	2000/04/11 11:56:38
***************
*** 7,13 ****
  """
  
  
- import string
  import mimetools
  
  
--- 7,12 ----
***************
*** 86,97 ****
          self._headers = []
  
      def addheader(self, key, value, prefix=0):
!         lines = string.splitfields(value, "\n")
          while lines and not lines[-1]: del lines[-1]
          while lines and not lines[0]: del lines[0]
          for i in range(1, len(lines)):
!             lines[i] = "    " + string.strip(lines[i])
!         value = string.joinfields(lines, "\n") + "\n"
          line = key + ": " + value
          if prefix:
              self._headers.insert(0, line)
--- 85,96 ----
          self._headers = []
  
      def addheader(self, key, value, prefix=0):
!         lines = value.split("\n")
          while lines and not lines[-1]: del lines[-1]
          while lines and not lines[0]: del lines[0]
          for i in range(1, len(lines)):
!             lines[i] = "    " + lines[i].strip()
!         value = "\n".join(lines) + "\n"
          line = key + ": " + value
          if prefix:
              self._headers.insert(0, line)
Index: Lib/SimpleHTTPServer.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/SimpleHTTPServer.py,v
retrieving revision 1.8
diff -c -r1.8 SimpleHTTPServer.py
*** SimpleHTTPServer.py	1999/11/16 19:04:32	1.8
--- SimpleHTTPServer.py	2000/04/11 11:56:38
***************
*** 10,16 ****
  
  
  import os
- import string
  import posixpath
  import BaseHTTPServer
  import urllib
--- 10,15 ----
***************
*** 79,85 ****
  
          """
          path = posixpath.normpath(urllib.unquote(path))
!         words = string.splitfields(path, '/')
          words = filter(None, words)
          path = os.getcwd()
          for word in words:
--- 78,84 ----
  
          """
          path = posixpath.normpath(urllib.unquote(path))
!         words = path.split('/')
          words = filter(None, words)
          path = os.getcwd()
          for word in words:
***************
*** 128,134 ****
          base, ext = posixpath.splitext(path)
          if self.extensions_map.has_key(ext):
              return self.extensions_map[ext]
!         ext = string.lower(ext)
          if self.extensions_map.has_key(ext):
              return self.extensions_map[ext]
          else:
--- 127,133 ----
          base, ext = posixpath.splitext(path)
          if self.extensions_map.has_key(ext):
              return self.extensions_map[ext]
!         ext = ext.lower()
          if self.extensions_map.has_key(ext):
              return self.extensions_map[ext]
          else:
Index: Lib/StringIO.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/StringIO.py,v
retrieving revision 1.9
diff -c -r1.9 StringIO.py
*** StringIO.py	2000/02/28 15:12:25	1.9
--- StringIO.py	2000/04/11 11:56:38
***************
*** 28,35 ****
  - There's a simple test set (see end of this file).
  """
  
- import string
- 
  class StringIO:
  	def __init__(self, buf = ''):
  		self.buf = buf
--- 28,33 ----
***************
*** 50,56 ****
  		if self.closed:
  			raise ValueError, "I/O operation on closed file"
  		if self.buflist:
! 			self.buf = self.buf + string.joinfields(self.buflist, '')
  			self.buflist = []
  		if mode == 1:
  			pos = pos + self.pos
--- 48,54 ----
  		if self.closed:
  			raise ValueError, "I/O operation on closed file"
  		if self.buflist:
! 			self.buf = self.buf + ''.join(self.buflist)
  			self.buflist = []
  		if mode == 1:
  			pos = pos + self.pos
***************
*** 65,71 ****
  		if self.closed:
  			raise ValueError, "I/O operation on closed file"
  		if self.buflist:
! 			self.buf = self.buf + string.joinfields(self.buflist, '')
  			self.buflist = []
  		if n < 0:
  			newpos = self.len
--- 63,69 ----
  		if self.closed:
  			raise ValueError, "I/O operation on closed file"
  		if self.buflist:
! 			self.buf = self.buf + ''.join(self.buflist)
  			self.buflist = []
  		if n < 0:
  			newpos = self.len
***************
*** 78,86 ****
  		if self.closed:
  			raise ValueError, "I/O operation on closed file"
  		if self.buflist:
! 			self.buf = self.buf + string.joinfields(self.buflist, '')
  			self.buflist = []
! 		i = string.find(self.buf, '\n', self.pos)
  		if i < 0:
  			newpos = self.len
  		else:
--- 76,84 ----
  		if self.closed:
  			raise ValueError, "I/O operation on closed file"
  		if self.buflist:
! 			self.buf = self.buf + ''.join(self.buflist)
  			self.buflist = []
! 		i = self.buf.find('\n', self.pos)
  		if i < 0:
  			newpos = self.len
  		else:
***************
*** 108,114 ****
  		newpos = self.pos + len(s)
  		if self.pos < self.len:
  			if self.buflist:
! 				self.buf = self.buf + string.joinfields(self.buflist, '')
  				self.buflist = []
  			self.buflist = [self.buf[:self.pos], s, self.buf[newpos:]]
  			self.buf = ''
--- 106,112 ----
  		newpos = self.pos + len(s)
  		if self.pos < self.len:
  			if self.buflist:
! 				self.buf = self.buf + ''.join(self.buflist)
  				self.buflist = []
  			self.buflist = [self.buf[:self.pos], s, self.buf[newpos:]]
  			self.buf = ''
***************
*** 117,129 ****
  			self.len = newpos
  		self.pos = newpos
  	def writelines(self, list):
! 		self.write(string.joinfields(list, ''))
  	def flush(self):
  		if self.closed:
  			raise ValueError, "I/O operation on closed file"
  	def getvalue(self):
  		if self.buflist:
! 			self.buf = self.buf + string.joinfields(self.buflist, '')
  			self.buflist = []
  		return self.buf
  
--- 115,127 ----
  			self.len = newpos
  		self.pos = newpos
  	def writelines(self, list):
! 		self.write(''.joinfields(list))
  	def flush(self):
  		if self.closed:
  			raise ValueError, "I/O operation on closed file"
  	def getvalue(self):
  		if self.buflist:
! 			self.buf = self.buf + ''.join(self.buflist)
  			self.buflist = []
  		return self.buf
  
Index: Lib/asynchat.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/asynchat.py,v
retrieving revision 1.5
diff -c -r1.5 asynchat.py
*** asynchat.py	2000/02/04 15:39:29	1.5
--- asynchat.py	2000/04/11 11:56:38
***************
*** 48,54 ****
  
  import socket
  import asyncore
- import string
  
  class async_chat (asyncore.dispatcher):
  	"""This is an abstract class.  You must derive from this class, and add
--- 48,53 ----
***************
*** 120,126 ****
  				# 3) end of buffer does not match any prefix:
  				#    collect data
  				terminator_len = len(terminator)
! 				index = string.find (self.ac_in_buffer, terminator)
  				if index != -1:
  					# we found the terminator
  					self.collect_incoming_data (self.ac_in_buffer[:index])
--- 119,125 ----
  				# 3) end of buffer does not match any prefix:
  				#    collect data
  				terminator_len = len(terminator)
! 				index = self.ac_in_buffer.find (terminator)
  				if index != -1:
  					# we found the terminator
  					self.collect_incoming_data (self.ac_in_buffer[:index])
Index: Lib/asyncore.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/asyncore.py,v
retrieving revision 1.5
diff -c -r1.5 asyncore.py
*** asyncore.py	2000/02/25 11:48:42	1.5
--- asyncore.py	2000/04/11 11:56:39
***************
*** 48,54 ****
  
  import select
  import socket
- import string
  import sys
  
  import os
--- 48,53 ----
***************
*** 152,158 ****
  				status.append ('%s:%d' % self.addr)
  			return '<%s %s at %x>' % (
  				self.__class__.__name__,
! 				string.join (status, ' '),
  				id(self)
  				)
  		except:
--- 151,157 ----
  				status.append ('%s:%d' % self.addr)
  			return '<%s %s at %x>' % (
  				self.__class__.__name__,
! 				' '.join (status),
  				id(self)
  				)
  		except:
***************
*** 407,418 ****
  	del tb
  
  	file, function, line = tbinfo[-1]
! 	info = '[' + string.join (
  		map (
! 			lambda x: string.join (x, '|'),
  			tbinfo
  			),
! 		'] ['
  		) + ']'
  	return (file, function, line), t, v, info
  
--- 406,417 ----
  	del tb
  
  	file, function, line = tbinfo[-1]
! 	info = '[' + '] ['.join (
  		map (
! 			lambda x: '|'.join (x),
  			tbinfo
  			),
! 		
  		) + ']'
  	return (file, function, line), t, v, info
  
Index: Lib/bdb.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/bdb.py,v
retrieving revision 1.27
diff -c -r1.27 bdb.py
*** bdb.py	2000/02/02 15:10:14	1.27
--- bdb.py	2000/04/11 11:56:39
***************
*** 304,310 ****
      # 
      
      def format_stack_entry(self, frame_lineno, lprefix=': '):
!         import linecache, repr, string
          frame, lineno = frame_lineno
          filename = self.canonic(frame.f_code.co_filename)
          s = filename + '(' + `lineno` + ')'
--- 304,310 ----
      # 
      
      def format_stack_entry(self, frame_lineno, lprefix=': '):
!         import linecache, repr
          frame, lineno = frame_lineno
          filename = self.canonic(frame.f_code.co_filename)
          s = filename + '(' + `lineno` + ')'
***************
*** 325,331 ****
              s = s + '->'
              s = s + repr.repr(rv)
          line = linecache.getline(filename, lineno)
!         if line: s = s + lprefix + string.strip(line)
          return s
      
      # The following two methods can be called by clients to use
--- 325,331 ----
              s = s + '->'
              s = s + repr.repr(rv)
          line = linecache.getline(filename, lineno)
!         if line: s = s + lprefix + line.strip()
          return s
      
      # The following two methods can be called by clients to use
***************
*** 532,543 ****
          if not name: name = '???'
          print '+++ call', name, args
      def user_line(self, frame):
!         import linecache, string
          name = frame.f_code.co_name
          if not name: name = '???'
          fn = self.canonic(frame.f_code.co_filename)
          line = linecache.getline(fn, frame.f_lineno)
!         print '+++', fn, frame.f_lineno, name, ':', string.strip(line)
      def user_return(self, frame, retval):
          print '+++ return', retval
      def user_exception(self, frame, exc_stuff):
--- 532,543 ----
          if not name: name = '???'
          print '+++ call', name, args
      def user_line(self, frame):
!         import linecache
          name = frame.f_code.co_name
          if not name: name = '???'
          fn = self.canonic(frame.f_code.co_filename)
          line = linecache.getline(fn, frame.f_lineno)
!         print '+++', fn, frame.f_lineno, name, ':', line.strip()
      def user_return(self, frame, retval):
          print '+++ return', retval
      def user_exception(self, frame, exc_stuff):
Index: Lib/binhex.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/binhex.py,v
retrieving revision 1.12
diff -c -r1.12 binhex.py
*** binhex.py	2000/02/04 15:39:29	1.12
--- binhex.py	2000/04/11 11:56:39
***************
*** 99,105 ****
          dsize = fp.tell()
          fp.close()
          dir, file = os.path.split(name)
!         file = string.replace(file, ':', '-', 1)
          return file, finfo, dsize, 0
  
      class openrsrc:
--- 99,105 ----
          dsize = fp.tell()
          fp.close()
          dir, file = os.path.split(name)
!         file = file.replace(':', '-', 1)
          return file, finfo, dsize, 0
  
      class openrsrc:
Index: Lib/cgi.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/cgi.py,v
retrieving revision 1.47
diff -c -r1.47 cgi.py
*** cgi.py	2000/02/28 15:12:25	1.47
--- cgi.py	2000/04/11 11:56:40
***************
*** 518,524 ****
          if ctype == 'multipart/form-data':
              return parse_multipart(fp, pdict)
          elif ctype == 'application/x-www-form-urlencoded':
!             clength = string.atoi(environ['CONTENT_LENGTH'])
              if maxlen and clength > maxlen:
                  raise ValueError, 'Maximum content length exceeded'
              qs = fp.read(clength)
--- 518,524 ----
          if ctype == 'multipart/form-data':
              return parse_multipart(fp, pdict)
          elif ctype == 'application/x-www-form-urlencoded':
!             clength = int(environ['CONTENT_LENGTH'])
              if maxlen and clength > maxlen:
                  raise ValueError, 'Maximum content length exceeded'
              qs = fp.read(clength)
***************
*** 589,604 ****
  
         Returns a list, as God intended.
      """
!     name_value_pairs = string.splitfields(qs, '&')
      r=[]
      for name_value in name_value_pairs:
!         nv = string.splitfields(name_value, '=')
          if len(nv) != 2:
              if strict_parsing:
                  raise ValueError, "bad query field: %s" % `name_value`
              continue
!         name = urllib.unquote(string.replace(nv[0], '+', ' '))
!         value = urllib.unquote(string.replace(nv[1], '+', ' '))
          r.append((name, value))
  
      return r
--- 589,604 ----
  
         Returns a list, as God intended.
      """
!     name_value_pairs = qs.split('&')
      r=[]
      for name_value in name_value_pairs:
!         nv = name_value.split('=')
          if len(nv) != 2:
              if strict_parsing:
                  raise ValueError, "bad query field: %s" % `name_value`
              continue
!         name = urllib.unquote(nv[0].replace('+', ' '))
!         value = urllib.unquote(nv[1].replace('+', ' '))
          r.append((name, value))
  
      return r
***************
*** 643,650 ****
              clength = headers.getheader('content-length')
              if clength:
                  try:
!                     bytes = string.atoi(clength)
!                 except string.atoi_error:
                      pass
              if bytes > 0:
                  if maxlen and bytes > maxlen:
--- 643,650 ----
              clength = headers.getheader('content-length')
              if clength:
                  try:
!                     bytes = int(clength)
!                 except ValueError:
                      pass
              if bytes > 0:
                  if maxlen and bytes > maxlen:
***************
*** 660,666 ****
                  terminator = lastpart # End outer loop
                  break
              if line[:2] == "--":
!                 terminator = string.strip(line)
                  if terminator in (nextpart, lastpart):
                      break
              lines.append(line)
--- 660,666 ----
                  terminator = lastpart # End outer loop
                  break
              if line[:2] == "--":
!                 terminator = line.strip()
                  if terminator in (nextpart, lastpart):
                      break
              lines.append(line)
***************
*** 676,682 ****
                  elif line[-1:] == "\n":
                      line = line[:-1]
                  lines[-1] = line
!                 data = string.joinfields(lines, "")
          line = headers['content-disposition']
          if not line:
              continue
--- 676,682 ----
                  elif line[-1:] == "\n":
                      line = line[:-1]
                  lines[-1] = line
!                 data = "".join(lines)
          line = headers['content-disposition']
          if not line:
              continue
***************
*** 701,715 ****
      Return the main content-type and a dictionary of options.
  
      """
!     plist = map(string.strip, string.splitfields(line, ';'))
!     key = string.lower(plist[0])
      del plist[0]
      pdict = {}
      for p in plist:
!         i = string.find(p, '=')
          if i >= 0:
!             name = string.lower(string.strip(p[:i]))
!             value = string.strip(p[i+1:])
              if len(value) >= 2 and value[0] == value[-1] == '"':
                  value = value[1:-1]
              pdict[name] = value
--- 701,715 ----
      Return the main content-type and a dictionary of options.
  
      """
!     plist = map(lambda s: s.strip(), line.split(';'))
!     key = plist[0].lower()
      del plist[0]
      pdict = {}
      for p in plist:
!         i = p.find('=')
          if i >= 0:
!             name = p[:i].strip().lower ()
!             value = p[i+1:].strip()
              if len(value) >= 2 and value[0] == value[-1] == '"':
                  value = value[1:-1]
              pdict[name] = value
***************
*** 820,826 ****
          self.keep_blank_values = keep_blank_values
          self.strict_parsing = strict_parsing
          if environ.has_key('REQUEST_METHOD'):
!             method = string.upper(environ['REQUEST_METHOD'])
          if method == 'GET' or method == 'HEAD':
              if environ.has_key('QUERY_STRING'):
                  qs = environ['QUERY_STRING']
--- 820,826 ----
          self.keep_blank_values = keep_blank_values
          self.strict_parsing = strict_parsing
          if environ.has_key('REQUEST_METHOD'):
!             method = environ['REQUEST_METHOD'].upper()
          if method == 'GET' or method == 'HEAD':
              if environ.has_key('QUERY_STRING'):
                  qs = environ['QUERY_STRING']
***************
*** 884,890 ****
          clen = -1
          if self.headers.has_key('content-length'):
              try:
!                 clen = string.atoi(self.headers['content-length'])
              except:
                  pass
              if maxlen and clen > maxlen:
--- 884,890 ----
          clen = -1
          if self.headers.has_key('content-length'):
              try:
!                 clen = int(self.headers['content-length'])
              except:
                  pass
              if maxlen and clen > maxlen:
***************
*** 1033,1039 ****
                  break
              self.lines.append(line)
              if line[:2] == "--":
!                 strippedline = string.strip(line)
                  if strippedline == next:
                      break
                  if strippedline == last:
--- 1033,1039 ----
                  break
              self.lines.append(line)
              if line[:2] == "--":
!                 strippedline = line.strip()
                  if strippedline == next:
                      break
                  if strippedline == last:
***************
*** 1063,1069 ****
                  break
              self.lines.append(line)
              if line[:2] == "--":
!                 strippedline = string.strip(line)
                  if strippedline == next:
                      break
                  if strippedline == last:
--- 1063,1069 ----
                  break
              self.lines.append(line)
              if line[:2] == "--":
!                 strippedline = line.strip()
                  if strippedline == next:
                      break
                  if strippedline == last:
***************
*** 1171,1181 ****
      def __getitem__( self, key ):
          v = SvFormContentDict.__getitem__( self, key )
          if v[0] in string.digits+'+-.' : 
!             try:  return  string.atoi( v ) 
              except ValueError:
!                 try:    return string.atof( v )
                  except ValueError: pass
!         return string.strip(v)
      def values( self ):
          lis = [] 
          for key in self.keys():
--- 1171,1181 ----
      def __getitem__( self, key ):
          v = SvFormContentDict.__getitem__( self, key )
          if v[0] in string.digits+'+-.' : 
!             try:  return  int( v ) 
              except ValueError:
!                 try:    return float( v )
                  except ValueError: pass
!         return v.strip()
      def values( self ):
          lis = [] 
          for key in self.keys():
***************
*** 1211,1217 ****
      def length(self, key):
          return len(self.dict[key])
      def stripped(self, key):
!         if self.dict.has_key(key): return string.strip(self.dict[key][0])
          else: return None
      def pars(self):
          return self.dict
--- 1211,1217 ----
      def length(self, key):
          return len(self.dict[key])
      def stripped(self, key):
!         if self.dict.has_key(key): return self.dict[key][0].strip()
          else: return None
      def pars(self):
          return self.dict
***************
*** 1269,1275 ****
      list = traceback.format_tb(tb, limit) + \
             traceback.format_exception_only(type, value)
      print "<PRE>%s<B>%s</B></PRE>" % (
!         escape(string.join(list[:-1], "")),
          escape(list[-1]),
          )
      del tb
--- 1269,1275 ----
      list = traceback.format_tb(tb, limit) + \
             traceback.format_exception_only(type, value)
      print "<PRE>%s<B>%s</B></PRE>" % (
!         escape("".join(list[:-1])),
          escape(list[-1]),
          )
      del tb
***************
*** 1369,1379 ****
  
  def escape(s, quote=None):
      """Replace special characters '&', '<' and '>' by SGML entities."""
!     s = string.replace(s, "&", "&amp;") # Must be done first!
!     s = string.replace(s, "<", "&lt;")
!     s = string.replace(s, ">", "&gt;",)
      if quote:
!         s = string.replace(s, '"', "&quot;")
      return s
  
  
--- 1369,1378 ----
  
  def escape(s, quote=None):
      """Replace special characters '&', '<' and '>' by SGML entities."""
!     # Replacing "&" with "&amp" must be done first!
!     s = s.replace("&", "&amp;").replace("<", "&lt;").replace(">", "&gt;")
      if quote:
!         s = s.replace('"', "&quot;")
      return s
  
  
Index: Lib/cmd.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/cmd.py,v
retrieving revision 1.17
diff -c -r1.17 cmd.py
*** cmd.py	2000/02/02 15:10:14	1.17
--- cmd.py	2000/04/11 11:56:41
***************
*** 89,95 ****
          pass
  
      def onecmd(self, line):
!         line = string.strip(line)
          if line == '?':
              line = 'help'
          elif line == '!':
--- 89,95 ----
          pass
  
      def onecmd(self, line):
!         line = line.strip()
          if line == '?':
              line = 'help'
          elif line == '!':
***************
*** 102,108 ****
          self.lastcmd = line
          i, n = 0, len(line)
          while i < n and line[i] in self.identchars: i = i+1
!         cmd, arg = line[:i], string.strip(line[i:])
          if cmd == '':
              return self.default(line)
          else:
--- 102,108 ----
          self.lastcmd = line
          i, n = 0, len(line)
          while i < n and line[i] in self.identchars: i = i+1
!         cmd, arg = line[:i], line[i:].strip()
          if cmd == '':
              return self.default(line)
          else:
Index: Lib/code.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/code.py,v
retrieving revision 1.9
diff -c -r1.9 code.py
*** code.py	2000/02/28 15:12:25	1.9
--- code.py	2000/04/11 11:56:41
***************
*** 6,12 ****
  
  
  import sys
- import string
  import traceback
  from codeop import compile_command
  
--- 6,11 ----
***************
*** 242,248 ****
  
          """
          self.buffer.append(line)
!         source = string.join(self.buffer, "\n")
          more = self.runsource(source, self.filename)
          if not more:
              self.resetbuffer()
--- 241,247 ----
  
          """
          self.buffer.append(line)
!         source = "\n".join(self.buffer)
          more = self.runsource(source, self.filename)
          if not more:
              self.resetbuffer()
Index: Lib/codeop.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/codeop.py,v
retrieving revision 1.1
diff -c -r1.1 codeop.py
*** codeop.py	1998/10/22 21:56:15	1.1
--- codeop.py	2000/04/11 11:56:41
***************
*** 1,7 ****
  """Utility to compile possibly incomplete Python source code."""
  
  import sys
- import string
  import traceback
  
  def compile_command(source, filename="<input>", symbol="single"):
--- 1,6 ----
***************
*** 47,54 ****
      """
  
      # Check for source consisting of only blank lines and comments
!     for line in string.split(source, "\n"):
!         line = string.strip(line)
          if line and line[0] != '#':
              break               # Leave it alone
      else:
--- 46,53 ----
      """
  
      # Check for source consisting of only blank lines and comments
!     for line in source.split("\n"):
!         line = line.strip()
          if line and line[0] != '#':
              break               # Leave it alone
      else:
Index: Lib/dis.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/dis.py,v
retrieving revision 1.21
diff -c -r1.21 dis.py
*** dis.py	2000/03/30 15:02:11	1.21
--- dis.py	2000/04/11 11:56:41
***************
*** 1,7 ****
  """Disassembler of Python byte code into mnemonics."""
  
  import sys
- import string
  import types
  
  def dis(x=None):
--- 1,6 ----
***************
*** 64,76 ****
  		else: print '   ',
  		if i in labels: print '>>',
  		else: print '  ',
! 		print string.rjust(`i`, 4),
! 		print string.ljust(opname[op], 20),
  		i = i+1
  		if op >= HAVE_ARGUMENT:
  			oparg = ord(code[i]) + ord(code[i+1])*256
  			i = i+2
! 			print string.rjust(`oparg`, 5),
  			if op in hasconst:
  				print '(' + `co.co_consts[oparg]` + ')',
  			elif op in hasname:
--- 63,75 ----
  		else: print '   ',
  		if i in labels: print '>>',
  		else: print '  ',
! 		print `i`.rjust(4),
! 		print opname[op].ljust(20),
  		i = i+1
  		if op >= HAVE_ARGUMENT:
  			oparg = ord(code[i]) + ord(code[i+1])*256
  			i = i+2
! 			print `oparg`.rjust(5),
  			if op in hasconst:
  				print '(' + `co.co_consts[oparg]` + ')',
  			elif op in hasname:
Index: Lib/dospath.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/dospath.py,v
retrieving revision 1.12
diff -c -r1.12 dospath.py
*** dospath.py	2000/02/04 15:39:29	1.12
--- dospath.py	2000/04/11 11:56:41
***************
*** 15,21 ****
      single '_', but this has been removed.  This functionality should 
      possibly be added as a new function."""
  
!     return string.lower(string.replace(s, "/", "\\"))
  
  
  def isabs(s):
--- 15,21 ----
      single '_', but this has been removed.  This functionality should 
      possibly be added as a new function."""
  
!     return s.replace(s, "/", "\\").lower()
  
  
  def isabs(s):
***************
*** 248,256 ****
              path = path[index + 1:]
              pathlen = len(path)
              try:
!                 index = string.index(path, '\'')
                  res = res + '\'' + path[:index + 1]
!             except string.index_error:
                  res = res + path
                  index = pathlen -1
          elif c == '$':  # variable or '$$'
--- 248,256 ----
              path = path[index + 1:]
              pathlen = len(path)
              try:
!                 index = path.index('\'')
                  res = res + '\'' + path[:index + 1]
!             except ValueError:
                  res = res + path
                  index = pathlen -1
          elif c == '$':  # variable or '$$'
***************
*** 261,271 ****
                  path = path[index+2:]
                  pathlen = len(path)
                  try:
!                     index = string.index(path, '}')
                      var = path[:index]
                      if os.environ.has_key(var):
                          res = res + os.environ[var]
!                 except string.index_error:
                      res = res + path
                      index = pathlen - 1
              else:
--- 261,271 ----
                  path = path[index+2:]
                  pathlen = len(path)
                  try:
!                     index = path.index('}')
                      var = path[:index]
                      if os.environ.has_key(var):
                          res = res + os.environ[var]
!                 except ValueError:
                      res = res + path
                      index = pathlen - 1
              else:
***************
*** 290,301 ****
      """Normalize a path, e.g. A//B, A/./B and A/foo/../B all become A/B.
      Also, components of the path are silently truncated to 8+3 notation."""
  
!     path = string.replace(path, "/", "\\")
      prefix, path = splitdrive(path)
      while path[:1] == os.sep:
          prefix = prefix + os.sep
          path = path[1:]
!     comps = string.splitfields(path, os.sep)
      i = 0
      while i < len(comps):
          if comps[i] == '.':
--- 290,301 ----
      """Normalize a path, e.g. A//B, A/./B and A/foo/../B all become A/B.
      Also, components of the path are silently truncated to 8+3 notation."""
  
!     path = path.replace("/", "\\")
      prefix, path = splitdrive(path)
      while path[:1] == os.sep:
          prefix = prefix + os.sep
          path = path[1:]
!     comps = path.split(os.sep)
      i = 0
      while i < len(comps):
          if comps[i] == '.':
***************
*** 307,313 ****
          elif comps[i] == '' and i > 0 and comps[i-1] <> '':
              del comps[i]
          elif '.' in comps[i]:
!             comp = string.splitfields(comps[i], '.')
              comps[i] = comp[0][:8] + '.' + comp[1][:3]
              i = i+1
          elif len(comps[i]) > 8:
--- 307,313 ----
          elif comps[i] == '' and i > 0 and comps[i-1] <> '':
              del comps[i]
          elif '.' in comps[i]:
!             comp = comps[i].split('.')
              comps[i] = comp[0][:8] + '.' + comp[1][:3]
              i = i+1
          elif len(comps[i]) > 8:
***************
*** 318,324 ****
      # If the path is now empty, substitute '.'
      if not prefix and not comps:
          comps.append('.')
!     return prefix + string.joinfields(comps, os.sep)
  
  
  
--- 318,324 ----
      # If the path is now empty, substitute '.'
      if not prefix and not comps:
          comps.append('.')
!     return prefix + os.sep.join(comps)
  
  
  
Index: Lib/formatter.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/formatter.py,v
retrieving revision 1.15
diff -c -r1.15 formatter.py
*** formatter.py	2000/02/04 15:39:29	1.15
--- formatter.py	2000/04/11 11:56:42
***************
*** 157,175 ****
                  label = s + label
              index = index + 1
          if case == 'I':
!             return string.upper(label)
          return label
  
      def add_flowing_data(self, data,
                           # These are only here to load them into locals:
!                          whitespace = string.whitespace,
!                          join = string.join, split = string.split):
          if not data: return
          # The following looks a bit convoluted but is a great improvement over
          # data = regsub.gsub('[' + string.whitespace + ']+', ' ', data)
          prespace = data[:1] in whitespace
          postspace = data[-1:] in whitespace
!         data = join(split(data))
          if self.nospace and not data:
              return
          elif prespace or self.softspace:
--- 157,174 ----
                  label = s + label
              index = index + 1
          if case == 'I':
!             return label.upper()
          return label
  
      def add_flowing_data(self, data,
                           # These are only here to load them into locals:
!                          whitespace = string.whitespace):
          if not data: return
          # The following looks a bit convoluted but is a great improvement over
          # data = regsub.gsub('[' + string.whitespace + ']+', ' ', data)
          prespace = data[:1] in whitespace
          postspace = data[-1:] in whitespace
!         data = ' '.join (data.split())
          if self.nospace and not data:
              return
          elif prespace or self.softspace:
***************
*** 369,379 ****
  
      def send_literal_data(self, data):
          self.file.write(data)
!         i = string.rfind(data, '\n')
          if i >= 0:
              self.col = 0
              data = data[i+1:]
!         data = string.expandtabs(data)
          self.col = self.col + len(data)
          self.atbreak = 0
  
--- 368,378 ----
  
      def send_literal_data(self, data):
          self.file.write(data)
!         i = data.rfind('\n')
          if i >= 0:
              self.col = 0
              data = data[i+1:]
!         data = data.expandtabs()
          self.col = self.col + len(data)
          self.atbreak = 0
  
***************
*** 383,389 ****
          col = self.col
          maxcol = self.maxcol
          write = self.file.write
!         for word in string.split(data):
              if atbreak:
                  if col + len(word) >= maxcol:
                      write('\n')
--- 382,388 ----
          col = self.col
          maxcol = self.maxcol
          write = self.file.write
!         for word in data.split():
              if atbreak:
                  if col + len(word) >= maxcol:
                      write('\n')
Index: Lib/ftplib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/ftplib.py,v
retrieving revision 1.42
diff -c -r1.42 ftplib.py
*** ftplib.py	2000/03/28 21:45:45	1.42
--- ftplib.py	2000/04/11 11:56:42
***************
*** 37,43 ****
  
  import os
  import sys
- import string
  
  # Import SOCKS module if it exists, else standard socket module socket
  try:
--- 37,42 ----
***************
*** 235,244 ****
  
  	def sendport(self, host, port):
  		'''Send a PORT command with the current host and the given port number.'''
! 		hbytes = string.splitfields(host, '.')
  		pbytes = [`port/256`, `port%256`]
  		bytes = hbytes + pbytes
! 		cmd = 'PORT ' + string.joinfields(bytes, ',')
  		return self.voidcmd(cmd)
  
  	def makeport(self):
--- 234,243 ----
  
  	def sendport(self, host, port):
  		'''Send a PORT command with the current host and the given port number.'''
! 		hbytes = host.split('.')
  		pbytes = [`port/256`, `port%256`]
  		bytes = hbytes + pbytes
! 		cmd = 'PORT ' + ','.join(bytes)
  		return self.voidcmd(cmd)
  
  	def makeport(self):
***************
*** 448,454 ****
  		# Note that the RFC doesn't say anything about 'SIZE'
  		resp = self.sendcmd('SIZE ' + filename)
  		if resp[:3] == '213':
! 			return string.atoi(string.strip(resp[3:]))
  
  	def mkd(self, dirname):
  		'''Make a directory, return its full pathname.'''
--- 447,453 ----
  		# Note that the RFC doesn't say anything about 'SIZE'
  		resp = self.sendcmd('SIZE ' + filename)
  		if resp[:3] == '213':
! 			return int(resp[3:].strip())
  
  	def mkd(self, dirname):
  		'''Make a directory, return its full pathname.'''
***************
*** 492,498 ****
  		_150_re = re.compile("150 .* \((\d+) bytes\)", re.IGNORECASE)
  	m = _150_re.match(resp)
  	if m:
! 		return string.atoi(m.group(1))
  	return None
  
  
--- 491,497 ----
  		_150_re = re.compile("150 .* \((\d+) bytes\)", re.IGNORECASE)
  	m = _150_re.match(resp)
  	if m:
! 		return int(m.group(1))
  	return None
  
  
***************
*** 503,518 ****
  
  	if resp[:3] <> '227':
  		raise error_reply, resp
! 	left = string.find(resp, '(')
  	if left < 0: raise error_proto, resp
! 	right = string.find(resp, ')', left + 1)
  	if right < 0:
  		raise error_proto, resp	# should contain '(h1,h2,h3,h4,p1,p2)'
! 	numbers = string.split(resp[left+1:right], ',')
  	if len(numbers) <> 6:
  		raise error_proto, resp
! 	host = string.join(numbers[:4], '.')
! 	port = (string.atoi(numbers[4]) << 8) + string.atoi(numbers[5])
  	return host, port
  
  
--- 502,517 ----
  
  	if resp[:3] <> '227':
  		raise error_reply, resp
! 	left = resp.find('(')
  	if left < 0: raise error_proto, resp
! 	right = resp.find(')', left + 1)
  	if right < 0:
  		raise error_proto, resp	# should contain '(h1,h2,h3,h4,p1,p2)'
! 	numbers = resp[left+1:right].split(',')
  	if len(numbers) <> 6:
  		raise error_proto, resp
! 	host = '.'.join(numbers[:4])
! 	port = (int(numbers[4]) << 8) + int(numbers[5])
  	return host, port
  
  
***************
*** 590,602 ****
  		while 1:
  			line = fp.readline()
  			if not line: break
! 			if in_macro and string.strip(line):
  				macro_lines.append(line)
  				continue
  			elif in_macro:
  				self.__macros[macro_name] = tuple(macro_lines)
  				in_macro = 0
! 			words = string.split(line)
  			host = user = passwd = acct = None
  			default = 0
  			i = 0
--- 589,601 ----
  		while 1:
  			line = fp.readline()
  			if not line: break
! 			if in_macro and line.strip():
  				macro_lines.append(line)
  				continue
  			elif in_macro:
  				self.__macros[macro_name] = tuple(macro_lines)
  				in_macro = 0
! 			words = line.split()
  			host = user = passwd = acct = None
  			default = 0
  			i = 0
***************
*** 609,615 ****
  				if w1 == 'default':
  					default = 1
  				elif w1 == 'machine' and w2:
! 					host = string.lower(w2)
  					i = i + 1
  				elif w1 == 'login' and w2:
  					user = w2
--- 608,614 ----
  				if w1 == 'default':
  					default = 1
  				elif w1 == 'machine' and w2:
! 					host = w2.lower()
  					i = i + 1
  				elif w1 == 'login' and w2:
  					user = w2
***************
*** 651,657 ****
  		password, and the accounting field.
  
  		"""
! 		host = string.lower(host)
  		user = passwd = acct = None
  		if self.__hosts.has_key(host):
  			user, passwd, acct = self.__hosts[host]
--- 650,656 ----
  		password, and the accounting field.
  
  		"""
! 		host = host.lower()
  		user = passwd = acct = None
  		if self.__hosts.has_key(host):
  			user, passwd, acct = self.__hosts[host]
Index: Lib/gopherlib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/gopherlib.py,v
retrieving revision 1.7
diff -c -r1.7 gopherlib.py
*** gopherlib.py	2000/03/28 21:45:45	1.7
--- gopherlib.py	2000/04/11 11:56:42
***************
*** 1,7 ****
  """Gopher protocol client interface."""
  
- import string
- 
  # Default selector, host and port
  DEF_SELECTOR = '1/'
  DEF_HOST     = 'gopher.micro.umn.edu'
--- 1,5 ----
***************
*** 56,70 ****
  def send_selector(selector, host, port = 0):
      """Send a selector to a given host and port, return a file with the reply."""
      import socket
-     import string
      if not port:
!         i = string.find(host, ':')
          if i >= 0:
!             host, port = host[:i], string.atoi(host[i+1:])
      if not port:
          port = DEF_PORT
      elif type(port) == type(''):
!         port = string.atoi(port)
      s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
      s.connect((host, port))
      s.send(selector + CRLF)
--- 54,67 ----
  def send_selector(selector, host, port = 0):
      """Send a selector to a given host and port, return a file with the reply."""
      import socket
      if not port:
!         i = host.find(':')
          if i >= 0:
!             host, port = host[:i], int(host[i+1:])
      if not port:
          port = DEF_PORT
      elif type(port) == type(''):
!         port = int(port)
      s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
      s.connect((host, port))
      s.send(selector + CRLF)
***************
*** 96,102 ****
  
  def get_directory(f):
      """Get a directory in the form of a list of entries."""
-     import string
      list = []
      while 1:
          line = f.readline()
--- 93,98 ----
***************
*** 113,119 ****
              print '(Empty line from server)'
              continue
          gtype = line[0]
!         parts = string.splitfields(line[1:], TAB)
          if len(parts) < 4:
              print '(Bad line from server:', `line`, ')'
              continue
--- 109,115 ----
              print '(Empty line from server)'
              continue
          gtype = line[0]
!         parts = line[1:].split(TAB)
          if len(parts) < 4:
              print '(Bad line from server:', `line`, ')'
              continue
Index: Lib/gzip.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/gzip.py,v
retrieving revision 1.17
diff -c -r1.17 gzip.py
*** gzip.py	2000/02/04 15:39:29	1.17
--- gzip.py	2000/04/11 11:56:42
***************
*** 6,12 ****
  # based on Andrew Kuchling's minigzip.py distributed with the zlib module
  
  import time
- import string
  import zlib
  import struct
  import __builtin__
--- 6,11 ----
***************
*** 138,144 ****
              self.fileobj.write( self.compress.compress(data) )
  
      def writelines(self,lines):
!         self.write(string.join(lines))
  
      def read(self, size=-1):
          if self.extrasize <= 0 and self.fileobj is None:
--- 137,143 ----
              self.fileobj.write( self.compress.compress(data) )
  
      def writelines(self,lines):
!         self.write(' '.join (lines))
  
      def read(self, size=-1):
          if self.extrasize <= 0 and self.fileobj is None:
***************
*** 274,290 ****
          readsize = 100
          while 1:
              c = self.read(readsize)
!             i = string.find(c, '\n')
              if i >= 0 or c == '':
                  bufs.append(c[:i+1])
                  self._unread(c[i+1:])
!                 return string.join(bufs, '')
              bufs.append(c)
              readsize = readsize * 2
  
      def readlines(self, ignored=None):
          buf = self.read()
!         lines = string.split(buf, '\n')
          for i in range(len(lines)-1):
              lines[i] = lines[i] + '\n'
          if lines and not lines[-1]:
--- 273,289 ----
          readsize = 100
          while 1:
              c = self.read(readsize)
!             i = c.find('\n')
              if i >= 0 or c == '':
                  bufs.append(c[:i+1])
                  self._unread(c[i+1:])
!                 return ''.join(bufs)
              bufs.append(c)
              readsize = readsize * 2
  
      def readlines(self, ignored=None):
          buf = self.read()
!         lines = buf.split('\n')
          for i in range(len(lines)-1):
              lines[i] = lines[i] + '\n'
          if lines and not lines[-1]:
Index: Lib/htmllib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/htmllib.py,v
retrieving revision 1.14
diff -c -r1.14 htmllib.py
*** htmllib.py	1999/05/03 18:10:15	1.14
--- htmllib.py	2000/04/11 11:56:42
***************
*** 5,11 ****
  """
  
  
- import string
  from sgmllib import SGMLParser
  from formatter import AS_IS
  
--- 5,10 ----
***************
*** 49,55 ****
          data = self.savedata
          self.savedata = None
          if not self.nofill:
!             data = string.join(string.split(data))
          return data
  
      # --- Hooks for anchors; should probably be overridden
--- 48,54 ----
          data = self.savedata
          self.savedata = None
          if not self.nofill:
!             data = ''.join(data.split())
          return data
  
      # --- Hooks for anchors; should probably be overridden
***************
*** 320,332 ****
          name = ''
          type = ''
          for attrname, value in attrs:
!             value = string.strip(value)
              if attrname == 'href':
                  href = value
              if attrname == 'name':
                  name = value
              if attrname == 'type':
!                 type = string.lower(value)
          self.anchor_bgn(href, name, type)
  
      def end_a(self):
--- 319,331 ----
          name = ''
          type = ''
          for attrname, value in attrs:
!             value = value.strip()
              if attrname == 'href':
                  href = value
              if attrname == 'name':
                  name = value
              if attrname == 'type':
!                 type = value.lower()
          self.anchor_bgn(href, name, type)
  
      def end_a(self):
***************
*** 361,370 ****
              if attrname == 'src':
                  src = value
              if attrname == 'width':
!                 try: width = string.atoi(value)
                  except: pass
              if attrname == 'height':
!                 try: height = string.atoi(value)
                  except: pass
          self.handle_image(src, alt, ismap, align, width, height)
  
--- 360,369 ----
              if attrname == 'src':
                  src = value
              if attrname == 'width':
!                 try: width = int(value)
                  except: pass
              if attrname == 'height':
!                 try: height = int(value)
                  except: pass
          self.handle_image(src, alt, ismap, align, width, height)
  
Index: Lib/httplib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/httplib.py,v
retrieving revision 1.15
diff -c -r1.15 httplib.py
*** httplib.py	2000/03/28 21:45:45	1.15
--- httplib.py	2000/04/11 11:56:43
***************
*** 30,36 ****
  
  import os
  import socket
- import string
  import mimetools
  
  try:
--- 30,35 ----
***************
*** 100,110 ****
  
          """
          if not port:
!             i = string.find(host, ':')
              if i >= 0:
                  host, port = host[:i], host[i+1:]
!                 try: port = string.atoi(port)
!                 except string.atoi_error:
                      raise socket.error, "nonnumeric port"
          if not port: port = HTTP_PORT
          self.sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
--- 99,109 ----
  
          """
          if not port:
!             i = host.find(':')
              if i >= 0:
                  host, port = host[:i], host[i+1:]
!                 try: port = int(port)
!                 except ValueError:
                      raise socket.error, "nonnumeric port"
          if not port: port = HTTP_PORT
          self.sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
***************
*** 134,140 ****
          For example: h.putheader('Accept', 'text/html')
  
          """
!         str = '%s: %s\r\n' % (header, string.joinfields(args,'\r\n\t'))
          self.send(str)
  
      def endheaders(self):
--- 133,139 ----
          For example: h.putheader('Accept', 'text/html')
  
          """
!         str = '%s: %s\r\n' % (header, '\r\n\t'.join(args))
          self.send(str)
  
      def endheaders(self):
***************
*** 154,163 ****
          line = self.file.readline()
          if self.debuglevel > 0: print 'reply:', `line`
          try:
!             [ver, code, msg] = string.split(line, None, 2)
          except ValueError:
            try:
!               [ver, code] = string.split(line, None, 1)
                msg = ""
            except ValueError:
                self.headers = None
--- 153,162 ----
          line = self.file.readline()
          if self.debuglevel > 0: print 'reply:', `line`
          try:
!             [ver, code, msg] = line.split(None, 2)
          except ValueError:
            try:
!               [ver, code] = line.split(None, 1)
                msg = ""
            except ValueError:
                self.headers = None
***************
*** 165,172 ****
          if ver[:5] != 'HTTP/':
              self.headers = None
              return -1, line, self.headers
!         errcode = string.atoi(code)
!         errmsg = string.strip(msg)
          self.headers = mimetools.Message(self.file, 0)
          return errcode, errmsg, self.headers
  
--- 164,171 ----
          if ver[:5] != 'HTTP/':
              self.headers = None
              return -1, line, self.headers
!         errcode = int(code)
!         errmsg = msg.strip()
          self.headers = mimetools.Message(self.file, 0)
          return errcode, errmsg, self.headers
  
***************
*** 200,210 ****
  
              """
              if not port:
!                 i = string.find(host, ':')
                  if i >= 0:
                      host, port = host[:i], host[i+1:]
!                     try: port = string.atoi(port)
!                     except string.atoi_error:
                          raise socket.error, "nonnumeric port"
              if not port: port = HTTPS_PORT
              sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
--- 199,209 ----
  
              """
              if not port:
!                 i = host.find(':')
                  if i >= 0:
                      host, port = host[:i], host[i+1:]
!                     try: port = int(port)
!                     except ValueError:
                          raise socket.error, "nonnumeric port"
              if not port: port = HTTPS_PORT
              sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
***************
*** 243,249 ****
      print 'errmsg  =', errmsg
      print
      if headers:
!         for header in headers.headers: print string.strip(header)
      print
      print h.getfile().read()
      if hasattr(socket, "ssl"):
--- 242,248 ----
      print 'errmsg  =', errmsg
      print
      if headers:
!         for header in headers.headers: print header.strip()
      print
      print h.getfile().read()
      if hasattr(socket, "ssl"):
***************
*** 263,269 ****
          print 'errmsg  =', errmsg
          print
          if headers:
!             for header in headers.headers: print string.strip(header)
          print
          print h.getfile().read()
  
--- 262,268 ----
          print 'errmsg  =', errmsg
          print
          if headers:
!             for header in headers.headers: print header.strip()
          print
          print h.getfile().read()
  
Index: Lib/ihooks.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/ihooks.py,v
retrieving revision 1.9
diff -c -r1.9 ihooks.py
*** ihooks.py	1998/06/29 20:31:16	1.9
--- ihooks.py	2000/04/11 11:56:43
***************
*** 55,61 ****
  import imp
  import os
  import sys
- import string
  
  
  VERBOSE = 0
--- 55,60 ----
***************
*** 410,416 ****
              assert globals is parent.__dict__
              return parent
          if '.' in pname:
!             i = string.rfind(pname, '.')
              pname = pname[:i]
              parent = self.modules[pname]
              assert parent.__name__ == pname
--- 409,415 ----
              assert globals is parent.__dict__
              return parent
          if '.' in pname:
!             i = pname.rfind('.')
              pname = pname[:i]
              parent = self.modules[pname]
              assert parent.__name__ == pname
***************
*** 419,425 ****
  
      def find_head_package(self, parent, name):
          if '.' in name:
!             i = string.find(name, '.')
              head = name[:i]
              tail = name[i+1:]
          else:
--- 418,424 ----
  
      def find_head_package(self, parent, name):
          if '.' in name:
!             i = name.find('.')
              head = name[:i]
              tail = name[i+1:]
          else:
***************
*** 441,447 ****
      def load_tail(self, q, tail):
          m = q
          while tail:
!             i = string.find(tail, '.')
              if i < 0: i = len(tail)
              head, tail = tail[:i], tail[i+1:]
              mname = "%s.%s" % (m.__name__, head)
--- 440,446 ----
      def load_tail(self, q, tail):
          m = q
          while tail:
!             i = tail.find('.')
              if i < 0: i = len(tail)
              head, tail = tail[:i], tail[i+1:]
              mname = "%s.%s" % (m.__name__, head)
***************
*** 490,496 ****
          name = module.__name__
          if '.' not in name:
              return self.import_it(name, name, None)
!         i = string.rfind(name, '.')
          pname = name[:i]
          parent = self.modules[pname]
          return self.import_it(name[i+1:], name, parent)
--- 489,495 ----
          name = module.__name__
          if '.' not in name:
              return self.import_it(name, name, None)
!         i = name.rfind('.')
          pname = name[:i]
          parent = self.modules[pname]
          return self.import_it(name[i+1:], name, parent)
Index: Lib/imaplib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/imaplib.py,v
retrieving revision 1.20
diff -c -r1.20 imaplib.py
*** imaplib.py	2000/03/28 21:45:46	1.20
--- imaplib.py	2000/04/11 11:56:43
***************
*** 17,23 ****
  
  __version__ = "2.36"
  
! import binascii, re, socket, string, time, random, sys
  
  #	Globals
  
--- 17,23 ----
  
  __version__ = "2.36"
  
! import binascii, re, socket, time, random, sys
  
  #	Globals
  
***************
*** 162,168 ****
  		self._simple_command(cap)
  		if not self.untagged_responses.has_key(cap):
  			raise self.error('no CAPABILITY response from server')
! 		self.capabilities = tuple(string.split(string.upper(self.untagged_responses[cap][-1])))
  
  		if __debug__:
  			if self.debug >= 3:
--- 162,168 ----
  		self._simple_command(cap)
  		if not self.untagged_responses.has_key(cap):
  			raise self.error('no CAPABILITY response from server')
! 		self.capabilities = tuple(self.untagged_responses[cap][-1].upper().split())
  
  		if __debug__:
  			if self.debug >= 3:
***************
*** 180,186 ****
  	def __getattr__(self, attr):
  		#	Allow UPPERCASE variants of IMAP4 command methods.
  		if Commands.has_key(attr):
! 			return eval("self.%s" % string.lower(attr))
  		raise AttributeError("Unknown IMAP4 command: '%s'" % attr)
  
  
--- 180,186 ----
  	def __getattr__(self, attr):
  		#	Allow UPPERCASE variants of IMAP4 command methods.
  		if Commands.has_key(attr):
! 			return eval("self.%s" % attr.lower())
  		raise AttributeError("Unknown IMAP4 command: '%s'" % attr)
  
  
***************
*** 219,225 ****
  
  		(code, [data]) = <instance>.response(code)
  		"""
! 		return self._untagged_response(code, [None], string.upper(code))
  
  
  	def socket(self):
--- 219,225 ----
  
  		(code, [data]) = <instance>.response(code)
  		"""
! 		return self._untagged_response(code, [None], code.upper())
  
  
  	def socket(self):
***************
*** 273,279 ****
  		It should return None if the client abort response '*' should
  		be sent instead.
  		"""
! 		mech = string.upper(mechanism)
  		cap = 'AUTH=%s' % mech
  		if not cap in self.capabilities:
  			raise self.error("Server doesn't allow %s authentication." % mech)
--- 273,279 ----
  		It should return None if the client abort response '*' should
  		be sent instead.
  		"""
! 		mech = mechanism.upper()
  		cap = 'AUTH=%s' % mech
  		if not cap in self.capabilities:
  			raise self.error("Server doesn't allow %s authentication." % mech)
***************
*** 527,533 ****
  
  		Returns response appropriate to 'command'.
  		"""
! 		command = string.upper(command)
  		if not Commands.has_key(command):
  			raise self.error("Unknown IMAP4 UID command: %s" % command)
  		if self.state not in Commands[command]:
--- 527,533 ----
  
  		Returns response appropriate to 'command'.
  		"""
! 		command = command.upper()
  		if not Commands.has_key(command):
  			raise self.error("Unknown IMAP4 UID command: %s" % command)
  		if self.state not in Commands[command]:
***************
*** 719,725 ****
  
  				# Read literal direct from connection.
  
! 				size = string.atoi(self.mo.group('size'))
  				if __debug__:
  					if self.debug >= 4:
  						_mesg('read literal size %s' % size)
--- 719,725 ----
  
  				# Read literal direct from connection.
  
! 				size = int(self.mo.group('size'))
  				if __debug__:
  					if self.debug >= 4:
  						_mesg('read literal size %s' % size)
***************
*** 822,829 ****
  
  	def _quote(self, arg):
  
! 		arg = string.replace(arg, '\\', '\\\\')
! 		arg = string.replace(arg, '"', '\\"')
  
  		return '"%s"' % arg
  
--- 822,828 ----
  
  	def _quote(self, arg):
  
! 		arg = arg.replace('\\', '\\\\').replace('"', '\\"')
  
  		return '"%s"' % arg
  
***************
*** 910,916 ****
  	zonen = mo.group('zonen')
  
  	for name in ('day', 'year', 'hour', 'min', 'sec', 'zoneh', 'zonem'):
! 		exec "%s = string.atoi(mo.group('%s'))" % (name, name)
  
  	# INTERNALDATE timezone must be subtracted to get UT
  
--- 909,915 ----
  	zonen = mo.group('zonen')
  
  	for name in ('day', 'year', 'hour', 'min', 'sec', 'zoneh', 'zonem'):
! 		exec "%s = int(mo.group('%s'))" % (name, name)
  
  	# INTERNALDATE timezone must be subtracted to get UT
  
***************
*** 956,962 ****
  	if not mo:
  		return ()
  
! 	return tuple(string.split(mo.group('flags')))
  
  
  def Time2Internaldate(date_time):
--- 955,961 ----
  	if not mo:
  		return ()
  
! 	return tuple(mo.group('flags').split())
  
  
  def Time2Internaldate(date_time):
***************
*** 1000,1008 ****
  		l = dict.items()
  		if not l: return
  		t = '\n\t\t'
! 		j = string.join
! 		l = map(lambda x,j=j:'%s: "%s"' % (x[0], x[1][0] and j(x[1], '" "') or ''), l)
! 		_mesg('untagged responses dump:%s%s' % (t, j(l, t)))
  
  	_cmd_log = []		# Last `_cmd_log_len' interactions
  	_cmd_log_len = 10
--- 999,1006 ----
  		l = dict.items()
  		if not l: return
  		t = '\n\t\t'
! 		l = map(lambda x:'%s: "%s"' % (x[0], x[1][0] and '" "'.join (x[1]) or ''), l)
! 		_mesg('untagged responses dump:%s%s' % (t, t.join(l)))
  
  	_cmd_log = []		# Last `_cmd_log_len' interactions
  	_cmd_log_len = 10
***************
*** 1083,1089 ****
  		for ml in run('list', ('/tmp/', 'yy%')):
  			mo = re.match(r'.*"([^"]+)"$', ml)
  			if mo: path = mo.group(1)
! 			else: path = string.split(ml)[-1]
  			run('delete', (path,))
  
  		for cmd,args in test_seq2:
--- 1081,1087 ----
  		for ml in run('list', ('/tmp/', 'yy%')):
  			mo = re.match(r'.*"([^"]+)"$', ml)
  			if mo: path = mo.group(1)
! 			else: path = ml.split()[-1]
  			run('delete', (path,))
  
  		for cmd,args in test_seq2:
***************
*** 1092,1098 ****
  			if (cmd,args) != ('uid', ('SEARCH', 'ALL')):
  				continue
  
! 			uid = string.split(dat[-1])
  			if not uid: continue
  			run('uid', ('FETCH', '%s' % uid[-1],
  				'(FLAGS INTERNALDATE RFC822.SIZE RFC822.HEADER RFC822.TEXT)'))
--- 1090,1096 ----
  			if (cmd,args) != ('uid', ('SEARCH', 'ALL')):
  				continue
  
! 			uid = dat[-1].split()
  			if not uid: continue
  			run('uid', ('FETCH', '%s' % uid[-1],
  				'(FLAGS INTERNALDATE RFC822.SIZE RFC822.HEADER RFC822.TEXT)'))
Index: Lib/keyword.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/keyword.py,v
retrieving revision 1.8
diff -c -r1.8 keyword.py
*** keyword.py	2000/02/04 15:10:33	1.8
--- keyword.py	2000/04/11 11:56:43
***************
*** 50,56 ****
  iskeyword = kwdict.has_key
  
  def main():
!     import sys, re, string
  
      args = sys.argv[1:]
      iptfile = args and args[0] or "Python/graminit.c"
--- 50,56 ----
  iskeyword = kwdict.has_key
  
  def main():
!     import sys, re
  
      args = sys.argv[1:]
      iptfile = args and args[0] or "Python/graminit.c"
***************
*** 64,70 ****
      while 1:
          line = fp.readline()
          if not line: break
!         if string.find(line, '{1, "') > -1:
              match = strprog.search(line)
              if match:
                  lines.append("        '" + match.group(1) + "',\n")
--- 64,70 ----
      while 1:
          line = fp.readline()
          if not line: break
!         if line.find('{1, "') > -1:
              match = strprog.search(line)
              if match:
                  lines.append("        '" + match.group(1) + "',\n")
***************
*** 87,93 ****
  
      # write the output file
      fp = open(optfile, 'w')
!     fp.write(string.join(format, ''))
      fp.close()
  
  if __name__ == "__main__":
--- 87,93 ----
  
      # write the output file
      fp = open(optfile, 'w')
!     fp.write(''.join(format))
      fp.close()
  
  if __name__ == "__main__":
Index: Lib/knee.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/knee.py,v
retrieving revision 1.4
diff -c -r1.4 knee.py
*** knee.py	1998/03/26 21:12:20	1.4
--- knee.py	2000/04/11 11:56:44
***************
*** 7,13 ****
  
  """
  
! import sys, imp, __builtin__, string
  
  
  # Replacement for __import__()
--- 7,13 ----
  
  """
  
! import sys, imp, __builtin__
  
  
  # Replacement for __import__()
***************
*** 30,36 ****
          assert globals is parent.__dict__
          return parent
      if '.' in pname:
!         i = string.rfind(pname, '.')
          pname = pname[:i]
          parent = sys.modules[pname]
          assert parent.__name__ == pname
--- 30,36 ----
          assert globals is parent.__dict__
          return parent
      if '.' in pname:
!         i = pname.rfind('.')
          pname = pname[:i]
          parent = sys.modules[pname]
          assert parent.__name__ == pname
***************
*** 39,45 ****
  
  def find_head_package(parent, name):
      if '.' in name:
!         i = string.find(name, '.')
          head = name[:i]
          tail = name[i+1:]
      else:
--- 39,45 ----
  
  def find_head_package(parent, name):
      if '.' in name:
!         i = name.find('.')
          head = name[:i]
          tail = name[i+1:]
      else:
***************
*** 61,67 ****
  def load_tail(q, tail):
      m = q
      while tail:
!         i = string.find(tail, '.')
          if i < 0: i = len(tail)
          head, tail = tail[:i], tail[i+1:]
          mname = "%s.%s" % (m.__name__, head)
--- 61,67 ----
  def load_tail(q, tail):
      m = q
      while tail:
!         i = tail.find('.')
          if i < 0: i = len(tail)
          head, tail = tail[:i], tail[i+1:]
          mname = "%s.%s" % (m.__name__, head)
***************
*** 111,117 ****
      name = module.__name__
      if '.' not in name:
          return import_module(name, name, None)
!     i = string.rfind(name, '.')
      pname = name[:i]
      parent = sys.modules[pname]
      return import_module(name[i+1:], name, parent)
--- 111,117 ----
      name = module.__name__
      if '.' not in name:
          return import_module(name, name, None)
!     i = name.rfind('.')
      pname = name[:i]
      parent = sys.modules[pname]
      return import_module(name[i+1:], name, parent)
Index: Lib/locale.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/locale.py,v
retrieving revision 1.4
diff -c -r1.4 locale.py
*** locale.py	2000/02/04 15:39:29	1.4
--- locale.py	2000/04/11 11:56:44
***************
*** 3,9 ****
  # Author: Martin von Loewis
  
  from _locale import *
- import string
  
  #perform the grouping from right to left
  def _group(s):
--- 3,8 ----
***************
*** 34,40 ****
      but takes the current locale into account. 
      Grouping is applied if the third parameter is true."""
      result = f % val
!     fields = string.splitfields(result,".")
      if grouping:
          fields[0]=_group(fields[0])
      if len(fields)==2:
--- 33,39 ----
      but takes the current locale into account. 
      Grouping is applied if the third parameter is true."""
      result = f % val
!     fields = result.split(".")
      if grouping:
          fields[0]=_group(fields[0])
      if len(fields)==2:
***************
*** 48,67 ****
      """Convert float to integer, taking the locale into account."""
      return format("%.12g",val)
  
! def atof(str,func=string.atof):
      "Parses a string as a float according to the locale settings."
      #First, get rid of the grouping
!     s=string.splitfields(str,localeconv()['thousands_sep'])
!     str=string.join(s,"")
      #next, replace the decimal point with a dot
!     s=string.splitfields(str,localeconv()['decimal_point'])
!     str=string.join(s,'.')
      #finally, parse the string
      return func(str)
  
  def atoi(str):
      "Converts a string to an integer according to the locale settings."
!     return atof(str,string.atoi)
  
  def test():
      setlocale(LC_ALL,"")
--- 47,64 ----
      """Convert float to integer, taking the locale into account."""
      return format("%.12g",val)
  
! def atof(str,func=float):
      "Parses a string as a float according to the locale settings."
      #First, get rid of the grouping
!     str=str.replace(localeconv()['thousands_sep'], '')
      #next, replace the decimal point with a dot
!     str=str.replace(localeconv()['decimal_point'], '.')
      #finally, parse the string
      return func(str)
  
  def atoi(str):
      "Converts a string to an integer according to the locale settings."
!     return atof(str,int)
  
  def test():
      setlocale(LC_ALL,"")
Index: Lib/macpath.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/macpath.py,v
retrieving revision 1.21
diff -c -r1.21 macpath.py
*** macpath.py	2000/02/04 15:10:33	1.21
--- macpath.py	2000/04/11 11:56:44
***************
*** 1,15 ****
  """Pathname and path-related operations for the Macintosh."""
  
- import string
  import os
  from stat import *
  
  
  # Normalize the case of a pathname.  Dummy in Posix, but string.lower here.
  
! normcase = string.lower
  
- 
  def isabs(s):
      """Return true if a path is absolute.
      On the Mac, relative paths begin with a colon,
--- 1,14 ----
  """Pathname and path-related operations for the Macintosh."""
  
  import os
  from stat import *
  
  
  # Normalize the case of a pathname.  Dummy in Posix, but string.lower here.
  
! def normcase(path):
!     return path.lower ()
  
  def isabs(s):
      """Return true if a path is absolute.
      On the Mac, relative paths begin with a colon,
***************
*** 163,172 ****
      # XXX The Unix version doesn't raise an exception but simply
      # returns an unnormalized path.  Should do so here too.
  
-     import string
      if ':' not in s:
          return ':' + s
!     f = string.splitfields(s, ':')
      pre = []
      post = []
      if not f[0]:
--- 162,170 ----
      # XXX The Unix version doesn't raise an exception but simply
      # returns an unnormalized path.  Should do so here too.
  
      if ':' not in s:
          return ':' + s
!     f = s.split(':')
      pre = []
      post = []
      if not f[0]:
Index: Lib/macurl2path.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/macurl2path.py,v
retrieving revision 1.7
diff -c -r1.7 macurl2path.py
*** macurl2path.py	2000/02/04 15:39:29	1.7
--- macurl2path.py	2000/04/11 11:56:44
***************
*** 2,8 ****
  
  Do not import directly; use urllib instead."""
  
- import string
  import urllib
  import os
  
--- 2,7 ----
***************
*** 19,25 ****
      	pathname = pathname[2:]
      elif pathname[:2] == '//':
          raise RuntimeError, 'Cannot convert non-local URL to pathname'
!     components = string.split(pathname, '/')
      # Remove . and embedded ..
      i = 0
      while i < len(components):
--- 18,24 ----
      	pathname = pathname[2:]
      elif pathname[:2] == '//':
          raise RuntimeError, 'Cannot convert non-local URL to pathname'
!     components = pathname.split('/')
      # Remove . and embedded ..
      i = 0
      while i < len(components):
***************
*** 35,41 ****
              i = i+1
      if not components[0]:
          # Absolute unix path, don't start with colon
!         rv = string.join(components[1:], ':')
      else:
          # relative unix path, start with colon. First replace
          # leading .. by empty strings (giving ::file)
--- 34,40 ----
              i = i+1
      if not components[0]:
          # Absolute unix path, don't start with colon
!         rv = ':'.join(components[1:])
      else:
          # relative unix path, start with colon. First replace
          # leading .. by empty strings (giving ::file)
***************
*** 43,49 ****
          while i < len(components) and components[i] == '..':
              components[i] = ''
              i = i + 1
!         rv = ':' + string.join(components, ':')
      # and finally unquote slashes and other funny characters
      return urllib.unquote(rv)
  
--- 42,48 ----
          while i < len(components) and components[i] == '..':
              components[i] = ''
              i = i + 1
!         rv = ':' + ':'.join(components)
      # and finally unquote slashes and other funny characters
      return urllib.unquote(rv)
  
***************
*** 51,57 ****
      "convert mac pathname to /-delimited pathname"
      if '/' in pathname:
          raise RuntimeError, "Cannot convert pathname containing slashes"
!     components = string.split(pathname, ':')
      # Remove empty first and/or last component
      if components[0] == '':
          del components[0]
--- 50,56 ----
      "convert mac pathname to /-delimited pathname"
      if '/' in pathname:
          raise RuntimeError, "Cannot convert pathname containing slashes"
!     components = pathname.split(':')
      # Remove empty first and/or last component
      if components[0] == '':
          del components[0]
***************
*** 65,73 ****
      components = map(_pncomp2url, components)
  
      if os.path.isabs(pathname):
!         return '/' + string.join(components, '/')
      else:
!         return string.join(components, '/')
          
  def _pncomp2url(component):
  	component = urllib.quote(component[:31], safe='')  # We want to quote slashes
--- 64,72 ----
      components = map(_pncomp2url, components)
  
      if os.path.isabs(pathname):
!         return '/' + '/'.join(components)
      else:
!         return '/'.join(components)
          
  def _pncomp2url(component):
  	component = urllib.quote(component[:31], safe='')  # We want to quote slashes
Index: Lib/mailbox.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/mailbox.py,v
retrieving revision 1.19
diff -c -r1.19 mailbox.py
*** mailbox.py	2000/04/04 03:31:39	1.19
--- mailbox.py	2000/04/11 11:56:44
***************
*** 177,196 ****
          # Qmail directory mailbox
  
          def __init__(self, dirname):
-                 import string
                  self.dirname = dirname
                  self.boxes = []
  
                  # check for new mail
                  newdir = os.path.join(self.dirname, 'new')
                  for file in os.listdir(newdir):
!                         if len(string.split(file, '.')) > 2:
                                  self.boxes.append(os.path.join(newdir, file))
  
                  # Now check for current mail in this maildir
                  curdir = os.path.join(self.dirname, 'cur')
                  for file in os.listdir(curdir):
!                         if len(string.split(file, '.')) > 2:
                                  self.boxes.append(os.path.join(curdir, file))
  
          def next(self):
--- 177,195 ----
          # Qmail directory mailbox
  
          def __init__(self, dirname):
                  self.dirname = dirname
                  self.boxes = []
  
                  # check for new mail
                  newdir = os.path.join(self.dirname, 'new')
                  for file in os.listdir(newdir):
!                         if len(file.split('.')) > 2:
                                  self.boxes.append(os.path.join(newdir, file))
  
                  # Now check for current mail in this maildir
                  curdir = os.path.join(self.dirname, 'cur')
                  for file in os.listdir(curdir):
!                         if len(file.split('.')) > 2:
                                  self.boxes.append(os.path.join(curdir, file))
  
          def next(self):
***************
*** 225,231 ****
  def _test():
          import time
          import sys
-         import string
          import os
  
          args = sys.argv[1:]
--- 224,229 ----
***************
*** 261,267 ****
                  if len(args) <= 1:
                          msg.fp = None
          if len(args) > 1:
!                 num = string.atoi(args[1])
                  print 'Message %d body:'%num
                  msg = msgs[num-1]
                  msg.rewindbody()
--- 259,265 ----
                  if len(args) <= 1:
                          msg.fp = None
          if len(args) > 1:
!                 num = int(args[1])
                  print 'Message %d body:'%num
                  msg = msgs[num-1]
                  msg.rewindbody()
Index: Lib/mailcap.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/mailcap.py,v
retrieving revision 1.5
diff -c -r1.5 mailcap.py
*** mailcap.py	2000/02/04 15:10:33	1.5
--- mailcap.py	2000/04/11 11:56:44
***************
*** 1,7 ****
  """Mailcap file handling.  See RFC 1524."""
  
  import os
- import string
  
  
  # Part 1: top-level interface.
--- 1,6 ----
***************
*** 36,42 ****
      # XXX Actually, this is Unix-specific
      if os.environ.has_key('MAILCAPS'):
          str = os.environ['MAILCAPS']
!         mailcaps = string.splitfields(str, ':')
      else:
          if os.environ.has_key('HOME'):
              home = os.environ['HOME']
--- 35,41 ----
      # XXX Actually, this is Unix-specific
      if os.environ.has_key('MAILCAPS'):
          str = os.environ['MAILCAPS']
!         mailcaps = str.split(':')
      else:
          if os.environ.has_key('HOME'):
              home = os.environ['HOME']
***************
*** 64,70 ****
          line = fp.readline()
          if not line: break
          # Ignore comments and blank lines
!         if line[0] == '#' or string.strip(line) == '':
              continue
          nextline = line
          # Join continuation lines
--- 63,69 ----
          line = fp.readline()
          if not line: break
          # Ignore comments and blank lines
!         if line[0] == '#' or line.strip() == '':
              continue
          nextline = line
          # Join continuation lines
***************
*** 77,86 ****
          if not (key and fields):
              continue
          # Normalize the key
!         types = string.splitfields(key, '/')
          for j in range(len(types)):
!             types[j] = string.strip(types[j])
!         key = string.lower(string.joinfields(types, '/'))
          # Update the database
          if caps.has_key(key):
              caps[key].append(fields)
--- 76,85 ----
          if not (key and fields):
              continue
          # Normalize the key
!         types = key.split('/')
          for j in range(len(types)):
!             types[j] = types[j].strip()
!         key = '/'.join(types).lower()
          # Update the database
          if caps.has_key(key):
              caps[key].append(fields)
***************
*** 105,117 ****
      key, view, rest = fields[0], fields[1], fields[2:]
      fields = {'view': view}
      for field in rest:
!         i = string.find(field, '=')
          if i < 0:
              fkey = field
              fvalue = ""
          else:
!             fkey = string.strip(field[:i])
!             fvalue = string.strip(field[i+1:])
          if fields.has_key(fkey):
              # Ignore it
              pass
--- 104,116 ----
      key, view, rest = fields[0], fields[1], fields[2:]
      fields = {'view': view}
      for field in rest:
!         i = field.find('=')
          if i < 0:
              fkey = field
              fvalue = ""
          else:
!             fkey = field[:i].strip()
!             fvalue = field[i+1:].strip()
          if fields.has_key(fkey):
              # Ignore it
              pass
***************
*** 130,136 ****
              i = i+2
          else:
              i = i+1
!     return string.strip(line[start:i]), i
  
  
  # Part 3: using the database.
--- 129,135 ----
              i = i+2
          else:
              i = i+1
!     return line[start:i].strip(), i
  
  
  # Part 3: using the database.
***************
*** 159,165 ****
      entries = []
      if caps.has_key(MIMEtype):
          entries = entries + caps[MIMEtype]
!     MIMEtypes = string.splitfields(MIMEtype, '/')
      MIMEtype = MIMEtypes[0] + '/*'
      if caps.has_key(MIMEtype):
          entries = entries + caps[MIMEtype]
--- 158,164 ----
      entries = []
      if caps.has_key(MIMEtype):
          entries = entries + caps[MIMEtype]
!     MIMEtypes = MIMEtype.split('/')
      MIMEtype = MIMEtypes[0] + '/*'
      if caps.has_key(MIMEtype):
          entries = entries + caps[MIMEtype]
***************
*** 200,209 ****
      return res
  
  def findparam(name, plist):
!     name = string.lower(name) + '='
      n = len(name)
      for p in plist:
!         if string.lower(p[:n]) == name:
              return p[n:]
      return ''
  
--- 199,208 ----
      return res
  
  def findparam(name, plist):
!     name = name.lower() + '='
      n = len(name)
      for p in plist:
!         if p[:n].lower() == name:
              return p[n:]
      return ''
  
Index: Lib/mhlib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/mhlib.py,v
retrieving revision 1.19
diff -c -r1.19 mhlib.py
*** mhlib.py	2000/02/04 15:10:33	1.19
--- mhlib.py	2000/04/11 11:56:45
***************
*** 214,220 ****
          """Create a new folder (or raise os.error if it cannot be created)."""
          protect = pickline(self.profile, 'Folder-Protect')
          if protect and isnumeric(protect):
!             mode = string.atoi(protect, 8)
          else:
              mode = FOLDER_PROTECT
          os.mkdir(os.path.join(self.getpath(), name), mode)
--- 214,220 ----
          """Create a new folder (or raise os.error if it cannot be created)."""
          protect = pickline(self.profile, 'Folder-Protect')
          if protect and isnumeric(protect):
!             mode = int(protect, 8)
          else:
              mode = FOLDER_PROTECT
          os.mkdir(os.path.join(self.getpath(), name), mode)
***************
*** 284,290 ****
          for name in os.listdir(self.getfullname()):
              if match(name):
                  append(name)
!         messages = map(string.atoi, messages)
          messages.sort()
          if messages:
              self.last = messages[-1]
--- 284,290 ----
          for name in os.listdir(self.getfullname()):
              if match(name):
                  append(name)
!         messages = map(int, messages)
          messages.sort()
          if messages:
              self.last = messages[-1]
***************
*** 303,314 ****
          while 1:
              line = f.readline()
              if not line: break
!             fields = string.splitfields(line, ':')
              if len(fields) <> 2:
                  self.error('bad sequence in %s: %s' %
!                           (fullname, string.strip(line)))
!             key = string.strip(fields[0])
!             value = IntSet(string.strip(fields[1]), ' ').tolist()
              sequences[key] = value
          return sequences
  
--- 303,314 ----
          while 1:
              line = f.readline()
              if not line: break
!             fields = line.split(':')
              if len(fields) <> 2:
                  self.error('bad sequence in %s: %s' %
!                           (fullname, line.strip()))
!             key = fields[0].strip()
!             value = IntSet(fields[1].strip(), ' ').tolist()
              sequences[key] = value
          return sequences
  
***************
*** 358,364 ****
          if seq == 'all':
              return all
          # Test for X:Y before X-Y because 'seq:-n' matches both
!         i = string.find(seq, ':')
          if i >= 0:
              head, dir, tail = seq[:i], '', seq[i+1:]
              if tail[:1] in '-+':
--- 358,364 ----
          if seq == 'all':
              return all
          # Test for X:Y before X-Y because 'seq:-n' matches both
!         i = seq.find(':')
          if i >= 0:
              head, dir, tail = seq[:i], '', seq[i+1:]
              if tail[:1] in '-+':
***************
*** 366,372 ****
              if not isnumeric(tail):
                  raise Error, "bad message list %s" % seq
              try:
!                 count = string.atoi(tail)
              except (ValueError, OverflowError):
                  # Can't use sys.maxint because of i+count below
                  count = len(all)
--- 366,372 ----
              if not isnumeric(tail):
                  raise Error, "bad message list %s" % seq
              try:
!                 count = int(tail)
              except (ValueError, OverflowError):
                  # Can't use sys.maxint because of i+count below
                  count = len(all)
***************
*** 396,402 ****
                      i = bisect(all, anchor-1)
                      return all[i:i+count]
          # Test for X-Y next
!         i = string.find(seq, '-')
          if i >= 0:
              begin = self._parseindex(seq[:i], all)
              end = self._parseindex(seq[i+1:], all)
--- 396,402 ----
                      i = bisect(all, anchor-1)
                      return all[i:i+count]
          # Test for X-Y next
!         i = seq.find('-')
          if i >= 0:
              begin = self._parseindex(seq[:i], all)
              end = self._parseindex(seq[i+1:], all)
***************
*** 429,435 ****
          """Internal: parse a message number (or cur, first, etc.)."""
          if isnumeric(seq):
              try:
!                 return string.atoi(seq)
              except (OverflowError, ValueError):
                  return sys.maxint
          if seq in ('cur', '.'):
--- 429,435 ----
          """Internal: parse a message number (or cur, first, etc.)."""
          if isnumeric(seq):
              try:
!                 return int(seq)
              except (OverflowError, ValueError):
                  return sys.maxint
          if seq in ('cur', '.'):
***************
*** 679,694 ****
          decide which headers to return (its argument is the header
          name converted to lower case)."""
          if not pred:
!             return string.joinfields(self.headers, '')
          headers = []
          hit = 0
          for line in self.headers:
              if line[0] not in string.whitespace:
!                 i = string.find(line, ':')
                  if i > 0:
!                     hit = pred(string.lower(line[:i]))
              if hit: headers.append(line)
!         return string.joinfields(headers, '')
  
      def getbodytext(self, decode = 1):
          """Return the message's body text as string.  This undoes a
--- 679,694 ----
          decide which headers to return (its argument is the header
          name converted to lower case)."""
          if not pred:
!             return ''.join(self.headers)
          headers = []
          hit = 0
          for line in self.headers:
              if line[0] not in string.whitespace:
!                 i = line.find(':')
                  if i > 0:
!                     hit = pred(line[:i].lower())
              if hit: headers.append(line)
!         return ''.join(headers)
  
      def getbodytext(self, decode = 1):
          """Return the message's body text as string.  This undoes a
***************
*** 883,895 ****
          self.normalize()
  
      def fromstring(self, data):
-         import string
          new = []
!         for part in string.splitfields(data, self.sep):
              list = []
!             for subp in string.splitfields(part, self.rng):
!                 s = string.strip(subp)
!                 list.append(string.atoi(s))
              if len(list) == 1:
                  new.append((list[0], list[0]))
              elif len(list) == 2 and list[0] <= list[1]:
--- 883,894 ----
          self.normalize()
  
      def fromstring(self, data):
          new = []
!         for part in data.split(self.sep):
              list = []
!             for subp in part.split(self.rng):
!                 s = subp.strip()
!                 list.append(int(s))
              if len(list) == 1:
                  new.append((list[0], list[0]))
              elif len(list) == 2 and list[0] <= list[1]:
***************
*** 919,925 ****
                  if not line or line[0] not in string.whitespace:
                      break
                  text = text + line
!             return string.strip(text)
      return None
  
  def updateline(file, key, value, casefold = 1):
--- 918,924 ----
                  if not line or line[0] not in string.whitespace:
                      break
                  text = text + line
!             return text.strip()
      return None
  
  def updateline(file, key, value, casefold = 1):
***************
*** 994,1000 ****
          except Error, msg:
              print "Error:", msg
          stuff = os.popen("pick %s 2>/dev/null" % `seq`).read()
!         list = map(string.atoi, string.split(stuff))
          print list, "<-- pick"
      do('f.listmessages()')
  
--- 993,999 ----
          except Error, msg:
              print "Error:", msg
          stuff = os.popen("pick %s 2>/dev/null" % `seq`).read()
!         list = map(int, stuff.split())
          print list, "<-- pick"
      do('f.listmessages()')
  
Index: Lib/mimetools.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/mimetools.py,v
retrieving revision 1.17
diff -c -r1.17 mimetools.py
*** mimetools.py	2000/04/04 20:53:07	1.17
--- mimetools.py	2000/04/11 11:56:45
***************
*** 3,9 ****
  
  import os
  import rfc822
- import string
  import tempfile
  
  
--- 3,8 ----
***************
*** 25,41 ****
  		if str == None:
  			str = 'text/plain'
  		if ';' in str:
! 			i = string.index(str, ';')
  			self.plisttext = str[i:]
  			str = str[:i]
  		else:
  			self.plisttext = ''
! 		fields = string.splitfields(str, '/')
  		for i in range(len(fields)):
! 			fields[i] = string.lower(string.strip(fields[i]))
! 		self.type = string.joinfields(fields, '/')
  		self.maintype = fields[0]
! 		self.subtype = string.joinfields(fields[1:], '/')
  
  	def parseplist(self):
  		str = self.plisttext
--- 24,40 ----
  		if str == None:
  			str = 'text/plain'
  		if ';' in str:
! 			i = str.index(';')
  			self.plisttext = str[i:]
  			str = str[:i]
  		else:
  			self.plisttext = ''
! 		fields = str.split('/')
  		for i in range(len(fields)):
! 			fields[i] = fields[i].strip().lower()
! 		self.type = '/'.join(fields)
  		self.maintype = fields[0]
! 		self.subtype = '/'.join(fields[1:])
  
  	def parseplist(self):
  		str = self.plisttext
***************
*** 44,65 ****
  			str = str[1:]
  			if ';' in str:
  				# XXX Should parse quotes!
! 				end = string.index(str, ';')
  			else:
  				end = len(str)
  			f = str[:end]
  			if '=' in f:
! 				i = string.index(f, '=')
! 				f = string.lower(string.strip(f[:i])) + \
! 					'=' + string.strip(f[i+1:])
! 			self.plist.append(string.strip(f))
  			str = str[end:]
  
  	def getplist(self):
  		return self.plist
  
  	def getparam(self, name):
! 		name = string.lower(name) + '='
  		n = len(name)
  		for p in self.plist:
  			if p[:n] == name:
--- 43,64 ----
  			str = str[1:]
  			if ';' in str:
  				# XXX Should parse quotes!
! 				end = str.index(';')
  			else:
  				end = len(str)
  			f = str[:end]
  			if '=' in f:
! 				i = f.index('=')
! 				f = f[:i].strip().lower() + \
! 					'=' + f[i+1:].strip()
! 			self.plist.append(f.strip())
  			str = str[end:]
  
  	def getplist(self):
  		return self.plist
  
  	def getparam(self, name):
! 		name = name.lower() + '='
  		n = len(name)
  		for p in self.plist:
  			if p[:n] == name:
***************
*** 69,83 ****
  	def getparamnames(self):
  		result = []
  		for p in self.plist:
! 			i = string.find(p, '=')
  			if i >= 0:
! 				result.append(string.lower(p[:i]))
  		return result
  
  	def getencoding(self):
  		if self.encodingheader == None:
  			return '7bit'
! 		return string.lower(self.encodingheader)
  
  	def gettype(self):
  		return self.type
--- 68,82 ----
  	def getparamnames(self):
  		result = []
  		for p in self.plist:
! 			i = p.find('=')
  			if i >= 0:
! 				result.append(p[:i].lower())
  		return result
  
  	def getencoding(self):
  		if self.encodingheader == None:
  			return '7bit'
! 		return self.encodingheader.lower()
  
  	def gettype(self):
  		return self.type
Index: Lib/mimetypes.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/mimetypes.py,v
retrieving revision 1.11
diff -c -r1.11 mimetypes.py
*** mimetypes.py	2000/02/10 17:17:13	1.11
--- mimetypes.py	2000/04/11 11:56:45
***************
*** 23,29 ****
  
  """
  
- import string
  import posixpath
  import urllib
  
--- 23,28 ----
***************
*** 62,72 ****
          # data      := *urlchar
          # parameter := attribute "=" value
          # type/subtype defaults to "text/plain"
!         comma = string.find(url, ',')
          if comma < 0:
              # bad data URL
              return None, None
!         semi = string.find(url, ';', 0, comma)
          if semi >= 0:
              type = url[:semi]
          else:
--- 61,71 ----
          # data      := *urlchar
          # parameter := attribute "=" value
          # type/subtype defaults to "text/plain"
!         comma = url.find(',')
          if comma < 0:
              # bad data URL
              return None, None
!         semi = url.find(';', 0, comma)
          if semi >= 0:
              type = url[:semi]
          else:
***************
*** 84,91 ****
          encoding = None
      if types_map.has_key(ext):
          return types_map[ext], encoding
!     elif types_map.has_key(string.lower(ext)):
!         return types_map[string.lower(ext)], encoding
      else:
          return None, encoding
  
--- 83,90 ----
          encoding = None
      if types_map.has_key(ext):
          return types_map[ext], encoding
!     elif types_map.has_key(ext.lower()):
!         return types_map[ext.lower()], encoding
      else:
          return None, encoding
  
***************
*** 101,107 ****
      global inited
      if not inited:
          init()
!     type = string.lower(type)
      for ext, stype in types_map.items():
          if type == stype:
              return ext
--- 100,106 ----
      global inited
      if not inited:
          init()
!     type = type.lower()
      for ext, stype in types_map.items():
          if type == stype:
              return ext
***************
*** 125,131 ****
      while 1:
          line = f.readline()
          if not line: break
!         words = string.split(line)
          for i in range(len(words)):
              if words[i][0] == '#':
                  del words[i:]
--- 124,130 ----
      while 1:
          line = f.readline()
          if not line: break
!         words = line.split()
          for i in range(len(words)):
              if words[i][0] == '#':
                  del words[i:]
Index: Lib/mimify.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/mimify.py,v
retrieving revision 1.15
diff -c -r1.15 mimify.py
*** mimify.py	2000/02/04 15:39:29	1.15
--- mimify.py	2000/04/11 11:56:45
***************
*** 27,33 ****
  QUOTE = '> '		# string replies are quoted with
  # End configure
  
! import re, string
  
  qp = re.compile('^content-transfer-encoding:\\s*quoted-printable', re.I)
  base64_re = re.compile('^content-transfer-encoding:\\s*base64', re.I)
--- 27,33 ----
  QUOTE = '> '		# string replies are quoted with
  # End configure
  
! import re
  
  qp = re.compile('^content-transfer-encoding:\\s*quoted-printable', re.I)
  base64_re = re.compile('^content-transfer-encoding:\\s*base64', re.I)
***************
*** 94,100 ****
  		if res is None:
  			break
  		newline = newline + line[pos:res.start(0)] + \
! 			  chr(string.atoi(res.group(1), 16))
  		pos = res.end(0)
  	return newline + line[pos:]
  
--- 94,100 ----
  		if res is None:
  			break
  		newline = newline + line[pos:res.start(0)] + \
! 			  chr(int(res.group(1), 16))
  		pos = res.end(0)
  	return newline + line[pos:]
  
***************
*** 108,114 ****
  			break
  		match = res.group(1)
  		# convert underscores to spaces (before =XX conversion!)
! 		match = string.join(string.split(match, '_'), ' ')
  		newline = newline + line[pos:res.start(0)] + mime_decode(match)
  		pos = res.end(0)
  	return newline + line[pos:]
--- 108,114 ----
  			break
  		match = res.group(1)
  		# convert underscores to spaces (before =XX conversion!)
! 		match = match.replace('_', ' ')
  		newline = newline + line[pos:res.start(0)] + mime_decode(match)
  		pos = res.end(0)
  	return newline + line[pos:]
***************
*** 230,243 ****
  	pos = 0
  	if len(line) >= 5 and line[:5] == 'From ':
  		# quote 'From ' at the start of a line for stupid mailers
! 		newline = string.upper('=%02x' % ord('F'))
  		pos = 1
  	while 1:
  		res = reg.search(line, pos)
  		if res is None:
  			break
  		newline = newline + line[pos:res.start(0)] + \
! 			  string.upper('=%02x' % ord(res.group(0)))
  		pos = res.end(0)
  	line = newline + line[pos:]
  
--- 230,243 ----
  	pos = 0
  	if len(line) >= 5 and line[:5] == 'From ':
  		# quote 'From ' at the start of a line for stupid mailers
! 		newline = str('=%02x' % ord('F')).upper()
  		pos = 1
  	while 1:
  		res = reg.search(line, pos)
  		if res is None:
  			break
  		newline = newline + line[pos:res.start(0)] + \
! 			  str('=%02x' % ord(res.group(0))).upper()
  		pos = res.end(0)
  	line = newline + line[pos:]
  
***************
*** 344,350 ****
  		if chrset_res:
  			if has_iso_chars:
  				# change us-ascii into iso-8859-1
! 				if string.lower(chrset_res.group(2)) == 'us-ascii':
  					line = '%s%s%s' % (chrset_res.group(1),
  							   CHARSET,
  							   chrset_res.group(3))
--- 344,350 ----
  		if chrset_res:
  			if has_iso_chars:
  				# change us-ascii into iso-8859-1
! 				if chrset_res.group(2).lower() == 'us-ascii':
  					line = '%s%s%s' % (chrset_res.group(1),
  							   CHARSET,
  							   chrset_res.group(3))
***************
*** 445,451 ****
  			encode = unmimify
  		elif o == '-l':
  			try:
! 				MAXLEN = string.atoi(a)
  			except:
  				print usage
  				sys.exit(1)
--- 445,451 ----
  			encode = unmimify
  		elif o == '-l':
  			try:
! 				MAXLEN = int(a)
  			except:
  				print usage
  				sys.exit(1)
Index: Lib/multifile.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/multifile.py,v
retrieving revision 1.9
diff -c -r1.9 multifile.py
*** multifile.py	2000/02/04 15:10:33	1.9
--- multifile.py	2000/04/11 11:56:46
***************
*** 118,124 ****
  		return list
  
  	def read(self): # Note: no size argument -- read until EOF only!
! 		return string.joinfields(self.readlines(), '')
  
  	def next(self):
  		while self.readline(): pass
--- 118,124 ----
  		return list
  
  	def read(self): # Note: no size argument -- read until EOF only!
! 		return ''.join(self.readlines())
  
  	def next(self):
  		while self.readline(): pass
Index: Lib/nntplib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/nntplib.py,v
retrieving revision 1.19
diff -c -r1.19 nntplib.py
*** nntplib.py	2000/03/28 21:45:46	1.19
--- nntplib.py	2000/04/11 11:56:46
***************
*** 31,37 ****
  # Imports
  import re
  import socket
- import string
  
  
  
--- 31,36 ----
***************
*** 244,250 ****
  		resp, list = self.longcmd('LIST')
  		for i in range(len(list)):
  			# Parse lines into "group last first flag"
! 			list[i] = tuple(string.split(list[i]))
  		return resp, list
  
  	def group(self, name):
--- 243,249 ----
  		resp, list = self.longcmd('LIST')
  		for i in range(len(list)):
  			# Parse lines into "group last first flag"
! 			list[i] = tuple(list[i].split())
  		return resp, list
  
  	def group(self, name):
***************
*** 260,266 ****
  		resp = self.shortcmd('GROUP ' + name)
  		if resp[:3] <> '211':
  			raise NNTPReplyError(resp)
! 		words = string.split(resp)
  		count = first = last = 0
  		n = len(words)
  		if n > 1:
--- 259,265 ----
  		resp = self.shortcmd('GROUP ' + name)
  		if resp[:3] <> '211':
  			raise NNTPReplyError(resp)
! 		words = resp.split()
  		count = first = last = 0
  		n = len(words)
  		if n > 1:
***************
*** 270,276 ****
  				if n > 3:
  					last = words[3]
  					if n > 4:
! 						name = string.lower(words[4])
  		return resp, count, first, last, name
  
  	def help(self):
--- 269,275 ----
  				if n > 3:
  					last = words[3]
  					if n > 4:
! 						name = words[4].lower()
  		return resp, count, first, last, name
  
  	def help(self):
***************
*** 284,290 ****
  		"""Internal: parse the response of a STAT, NEXT or LAST command."""
  		if resp[:2] <> '22':
  			raise NNTPReplyError(resp)
! 		words = string.split(resp)
  		nr = 0
  		id = ''
  		n = len(words)
--- 283,289 ----
  		"""Internal: parse the response of a STAT, NEXT or LAST command."""
  		if resp[:2] <> '22':
  			raise NNTPReplyError(resp)
! 		words = resp.split()
  		nr = 0
  		id = ''
  		n = len(words)
***************
*** 391,404 ****
  		resp, lines = self.longcmd('XOVER ' + start + '-' + end)
  		xover_lines = []
  		for line in lines:
! 			elem = string.splitfields(line,"\t")
  			try:
  				xover_lines.append((elem[0],
  						    elem[1],
  						    elem[2],
  						    elem[3],
  						    elem[4],
! 						    string.split(elem[5]),
  						    elem[6],
  						    elem[7]))
  			except IndexError:
--- 390,403 ----
  		resp, lines = self.longcmd('XOVER ' + start + '-' + end)
  		xover_lines = []
  		for line in lines:
! 			elem = line.split("\t")
  			try:
  				xover_lines.append((elem[0],
  						    elem[1],
  						    elem[2],
  						    elem[3],
  						    elem[4],
! 						    elem[5].split(),
  						    elem[6],
  						    elem[7]))
  			except IndexError:
***************
*** 416,422 ****
  		resp, raw_lines = self.longcmd('XGTITLE ' + group)
  		lines = []
  		for raw_line in raw_lines:
! 			match = line_pat.search(string.strip(raw_line))
  			if match:
  				lines.append(match.group(1, 2))
  		return resp, lines
--- 415,421 ----
  		resp, raw_lines = self.longcmd('XGTITLE ' + group)
  		lines = []
  		for raw_line in raw_lines:
! 			match = line_pat.search(raw_line.strip())
  			if match:
  				lines.append(match.group(1, 2))
  		return resp, lines
***************
*** 432,438 ****
  		if resp[:3] <> '223':
  			raise NNTPReplyError(resp)
  		try:
! 			[resp_num, path] = string.split(resp)
  		except ValueError:
  			raise NNTPReplyError(resp)
  		else:
--- 431,437 ----
  		if resp[:3] <> '223':
  			raise NNTPReplyError(resp)
  		try:
! 			[resp_num, path] = resp.split()
  		except ValueError:
  			raise NNTPReplyError(resp)
  		else:
***************
*** 449,455 ****
  		resp = self.shortcmd("DATE")
  		if resp[:3] <> '111':
  			raise NNTPReplyError(resp)
! 		elem = string.split(resp)
  		if len(elem) != 2:
  			raise NNTPDataError(resp)
  		date = elem[1][2:8]
--- 448,454 ----
  		resp = self.shortcmd("DATE")
  		if resp[:3] <> '111':
  			raise NNTPReplyError(resp)
! 		elem = resp.split()
  		if len(elem) != 2:
  			raise NNTPDataError(resp)
  		date = elem[1][2:8]
Index: Lib/ntpath.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/ntpath.py,v
retrieving revision 1.23
diff -c -r1.23 ntpath.py
*** ntpath.py	2000/02/17 17:30:40	1.23
--- ntpath.py	2000/04/11 11:56:46
***************
*** 18,24 ****
      """Normalize case of pathname.
  
      Makes all characters lowercase and all slashes into backslashes."""
!     return string.lower(string.replace(s, "/", "\\"))
  
  
  # Return whether a path is absolute.
--- 18,24 ----
      """Normalize case of pathname.
  
      Makes all characters lowercase and all slashes into backslashes."""
!     return s.replace("/", "\\").lower()
  
  
  # Return whether a path is absolute.
***************
*** 77,87 ****
          # \\machine\mountpoint\directories...
          #           directory ^^^^^^^^^^^^^^^
          normp = normcase(p)
!         index = string.find(normp, '\\', 2)
          if index == -1:
              ##raise RuntimeError, 'illegal UNC path: "' + p + '"'
              return ("", p)
!         index = string.find(normp, '\\', index + 1)
          if index == -1:
              index = len(p)
          return p[:index], p[index:]
--- 77,87 ----
          # \\machine\mountpoint\directories...
          #           directory ^^^^^^^^^^^^^^^
          normp = normcase(p)
!         index = normp.find('\\', 2)
          if index == -1:
              ##raise RuntimeError, 'illegal UNC path: "' + p + '"'
              return ("", p)
!         index = normp.find('\\', index + 1)
          if index == -1:
              index = len(p)
          return p[:index], p[index:]
***************
*** 331,339 ****
              path = path[index + 1:]
              pathlen = len(path)
              try:
!                 index = string.index(path, '\'')
                  res = res + '\'' + path[:index + 1]
!             except string.index_error:
                  res = res + path
                  index = pathlen -1
          elif c == '$':  # variable or '$$'
--- 331,339 ----
              path = path[index + 1:]
              pathlen = len(path)
              try:
!                 index = path.index('\'')
                  res = res + '\'' + path[:index + 1]
!             except ValueError:
                  res = res + path
                  index = pathlen -1
          elif c == '$':  # variable or '$$'
***************
*** 344,354 ****
                  path = path[index+2:]
                  pathlen = len(path)
                  try:
!                     index = string.index(path, '}')
                      var = path[:index]
                      if os.environ.has_key(var):
                          res = res + os.environ[var]
!                 except string.index_error:
                      res = res + path
                      index = pathlen - 1
              else:
--- 344,354 ----
                  path = path[index+2:]
                  pathlen = len(path)
                  try:
!                     index = path.index('}')
                      var = path[:index]
                      if os.environ.has_key(var):
                          res = res + os.environ[var]
!                 except ValueError:
                      res = res + path
                      index = pathlen - 1
              else:
***************
*** 375,386 ****
  
  def normpath(path):
      """Normalize path, eliminating double slashes, etc."""
!     path = string.replace(path, "/", "\\")
      prefix, path = splitdrive(path)
      while path[:1] == os.sep:
          prefix = prefix + os.sep
          path = path[1:]
!     comps = string.splitfields(path, os.sep)
      i = 0
      while i < len(comps):
          if comps[i] == '.':
--- 375,386 ----
  
  def normpath(path):
      """Normalize path, eliminating double slashes, etc."""
!     path = path.replace("/", "\\")
      prefix, path = splitdrive(path)
      while path[:1] == os.sep:
          prefix = prefix + os.sep
          path = path[1:]
!     comps = path.split(os.sep)
      i = 0
      while i < len(comps):
          if comps[i] == '.':
***************
*** 395,401 ****
      # If the path is now empty, substitute '.'
      if not prefix and not comps:
          comps.append('.')
!     return prefix + string.joinfields(comps, os.sep)
  
  
  # Return an absolute path.
--- 395,401 ----
      # If the path is now empty, substitute '.'
      if not prefix and not comps:
          comps.append('.')
!     return prefix + os.sep.join(comps)
  
  
  # Return an absolute path.
Index: Lib/nturl2path.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/nturl2path.py,v
retrieving revision 1.5
diff -c -r1.5 nturl2path.py
*** nturl2path.py	2000/02/04 15:10:33	1.5
--- nturl2path.py	2000/04/11 11:56:46
***************
*** 16,30 ****
  	        # convert this to \\host\path\on\remote\host
  	        # (notice halving of slashes at the start of the path)
  	        url = url[2:]
! 	    components = string.split(url, '/')
  	    # make sure not to convert quoted slashes :-)
! 	    return urllib.unquote(string.join(components, '\\'))
! 	comp = string.split(url, '|')
  	if len(comp) != 2 or comp[0][-1] not in string.letters:
  		error = 'Bad URL: ' + url
  		raise IOError, error
! 	drive = string.upper(comp[0][-1])
! 	components = string.split(comp[1], '/')
  	path = drive + ':'
  	for  comp in components:
  		if comp:
--- 16,30 ----
  	        # convert this to \\host\path\on\remote\host
  	        # (notice halving of slashes at the start of the path)
  	        url = url[2:]
! 	    components = url.split('/')
  	    # make sure not to convert quoted slashes :-)
! 	    return urllib.unquote('\\'.join(components))
! 	comp = url.split('|')
  	if len(comp) != 2 or comp[0][-1] not in string.letters:
  		error = 'Bad URL: ' + url
  		raise IOError, error
! 	drive = comp[0][-1].upper()
! 	components = comp[1].split('/')
  	path = drive + ':'
  	for  comp in components:
  		if comp:
***************
*** 40,46 ****
  		///C|/foo/bar/spam.foo
  	"""
  
! 	import string, urllib
  	if not ':' in p:
  	    # No drive specifier, just convert slashes and quote the name
  	    if p[:2] == '\\\\':
--- 40,46 ----
  		///C|/foo/bar/spam.foo
  	"""
  
! 	import urllib
  	if not ':' in p:
  	    # No drive specifier, just convert slashes and quote the name
  	    if p[:2] == '\\\\':
***************
*** 48,62 ****
  	        # convert this to ////host/path/on/remote/host
  	        # (notice doubling of slashes at the start of the path)
  	        p = '\\\\' + p
! 	    components = string.split(p, '\\')
! 	    return urllib.quote(string.join(components, '/'))
! 	comp = string.split(p, ':')
  	if len(comp) != 2 or len(comp[0]) > 1:
  		error = 'Bad path: ' + p
  		raise IOError, error
  
! 	drive = urllib.quote(string.upper(comp[0]))
! 	components = string.split(comp[1], '\\')
  	path = '///' + drive + '|'
  	for comp in components:
  		if comp:
--- 48,62 ----
  	        # convert this to ////host/path/on/remote/host
  	        # (notice doubling of slashes at the start of the path)
  	        p = '\\\\' + p
! 	    components = p.split('\\')
! 	    return urllib.quote('/'.join(components))
! 	comp = p.split(':')
  	if len(comp) != 2 or len(comp[0]) > 1:
  		error = 'Bad path: ' + p
  		raise IOError, error
  
! 	drive = urllib.quote(comp[0].upper())
! 	components = comp[1].split('\\')
  	path = '///' + drive + '|'
  	for comp in components:
  		if comp:
Index: Lib/pdb.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/pdb.py,v
retrieving revision 1.43
diff -c -r1.43 pdb.py
*** pdb.py	2000/03/06 20:39:59	1.43
--- pdb.py	2000/04/11 11:56:47
***************
*** 4,10 ****
  
  # (See pdb.doc for documentation.)
  
- import string
  import sys
  import linecache
  import cmd
--- 4,9 ----
***************
*** 151,176 ****
  		"""Handle alias expansion and ';;' separator."""
  		if not line:
  			return line
! 		args = string.split(line)
  		while self.aliases.has_key(args[0]):
  			line = self.aliases[args[0]]
  			ii = 1
  			for tmpArg in args[1:]:
! 				line = string.replace(line, "%" + str(ii),
! 						      tmpArg)
  				ii = ii + 1
! 			line = string.replace(line, "%*",
! 					      string.join(args[1:], ' '))
! 			args = string.split(line)
  		# split into ';;' separated commands
  		# unless it's an alias command
  		if args[0] != 'alias':
! 			marker = string.find(line, ';;')
  			if marker >= 0:
  				# queue up everything after marker
! 				next = string.lstrip(line[marker+2:])
  				self.cmdqueue.append(next)
! 				line = string.rstrip(line[:marker])
  		return line
  
  	# Command definitions, called by cmdloop()
--- 150,173 ----
  		"""Handle alias expansion and ';;' separator."""
  		if not line:
  			return line
! 		args = line.split()
  		while self.aliases.has_key(args[0]):
  			line = self.aliases[args[0]]
  			ii = 1
  			for tmpArg in args[1:]:
! 				line = line.replace("%" + str(ii), tmpArg)
  				ii = ii + 1
! 			line = line.replace("%*", ' '.join(args[1:]))
! 			args = line.split()
  		# split into ';;' separated commands
  		# unless it's an alias command
  		if args[0] != 'alias':
! 			marker = line.find(';;')
  			if marker >= 0:
  				# queue up everything after marker
! 				next = line[marker+2:].lstrip()
  				self.cmdqueue.append(next)
! 				line = line[:marker].rstrip()
  		return line
  
  	# Command definitions, called by cmdloop()
***************
*** 196,210 ****
  		filename = None
  		lineno = None
  		cond = None
! 		comma = string.find(arg, ',')
  		if comma > 0:
  			# parse stuff after comma: "condition"
! 			cond = string.lstrip(arg[comma+1:])
! 			arg = string.rstrip(arg[:comma])
  		# parse stuff before comma: [filename:]lineno | function
! 		colon = string.rfind(arg, ':')
  		if colon >= 0:
! 			filename = string.rstrip(arg[:colon])
  			f = self.lookupmodule(filename)
  			if not f:
  				print '*** ', `filename`,
--- 193,207 ----
  		filename = None
  		lineno = None
  		cond = None
! 		comma = arg.find(',')
  		if comma > 0:
  			# parse stuff after comma: "condition"
! 			cond = arg[comma+1:].lstrip()
! 			arg = arg[:comma].rstrip()
  		# parse stuff before comma: [filename:]lineno | function
! 		colon = arg.rfind(':')
  		if colon >= 0:
! 			filename = arg[:colon].rstrip()
  			f = self.lookupmodule(filename)
  			if not f:
  				print '*** ', `filename`,
***************
*** 212,218 ****
  				return
  			else:
  				filename = f
! 			arg = string.lstrip(arg[colon+1:])
  			try:
  				lineno = int(arg)
  			except ValueError, msg:
--- 209,215 ----
  				return
  			else:
  				filename = f
! 			arg = arg[colon+1:].lstrip()
  			try:
  				lineno = int(arg)
  			except ValueError, msg:
***************
*** 276,292 ****
  	def lineinfo(self, identifier):
  		failed = (None, None, None)
  		# Input is identifier, may be in single quotes
! 		idstring = string.split(identifier, "'")
  		if len(idstring) == 1:
  			# not in single quotes
! 			id = string.strip(idstring[0])
  		elif len(idstring) == 3:
  			# quoted
! 			id = string.strip(idstring[1])
  		else:
  			return failed
  		if id == '': return failed
! 		parts = string.split(id, '.')
  		# Protection for derived debuggers
  		if parts[0] == 'self':
  			del parts[0]
--- 273,289 ----
  	def lineinfo(self, identifier):
  		failed = (None, None, None)
  		# Input is identifier, may be in single quotes
! 		idstring = identifier.split("'")
  		if len(idstring) == 1:
  			# not in single quotes
! 			id = idstring[0].strip()
  		elif len(idstring) == 3:
  			# quoted
! 			id = idstring[1].strip()
  		else:
  			return failed
  		if id == '': return failed
! 		parts = id.split('.')
  		# Protection for derived debuggers
  		if parts[0] == 'self':
  			del parts[0]
***************
*** 318,324 ****
  		if not line:
  			print 'End of file'
  			return 0
! 		line = string.strip(line)
  		# Don't allow setting breakpoint at a blank line
  		if ( not line or (line[0] == '#') or
  		     (line[:3] == '"""') or line[:3] == "'''" ):
--- 315,321 ----
  		if not line:
  			print 'End of file'
  			return 0
! 		line = line.strip()
  		# Don't allow setting breakpoint at a blank line
  		if ( not line or (line[0] == '#') or
  		     (line[:3] == '"""') or line[:3] == "'''" ):
***************
*** 354,374 ****
  				if not line:
  					print 'end of file'
  					return 0
! 				line = string.strip(line)
  				if not line: continue	# Blank line
  				if brackets <= 0 and line[0] not in ('#','"',"'"):
  					break
  		return lineno
  
  	def do_enable(self, arg):
! 		args = string.split(arg)
  		for i in args:
  			bp = bdb.Breakpoint.bpbynumber[int(i)]
  			if bp:
  				bp.enable()
  
  	def do_disable(self, arg):
! 		args = string.split(arg)
  		for i in args:
  			bp = bdb.Breakpoint.bpbynumber[int(i)]
  			if bp:
--- 351,371 ----
  				if not line:
  					print 'end of file'
  					return 0
! 				line = line.strip()
  				if not line: continue	# Blank line
  				if brackets <= 0 and line[0] not in ('#','"',"'"):
  					break
  		return lineno
  
  	def do_enable(self, arg):
! 		args = arg.split()
  		for i in args:
  			bp = bdb.Breakpoint.bpbynumber[int(i)]
  			if bp:
  				bp.enable()
  
  	def do_disable(self, arg):
! 		args = arg.split()
  		for i in args:
  			bp = bdb.Breakpoint.bpbynumber[int(i)]
  			if bp:
***************
*** 376,383 ****
  
  	def do_condition(self, arg):
  		# arg is breakpoint number and condition
! 		args = string.split(arg, ' ', 1)
! 		bpnum = int(string.strip(args[0]))
  		try:
  			cond = args[1]
  		except:
--- 373,380 ----
  
  	def do_condition(self, arg):
  		# arg is breakpoint number and condition
! 		args = arg.split(' ', 1)
! 		bpnum = int(args[0].strip())
  		try:
  			cond = args[1]
  		except:
***************
*** 391,400 ****
  
  	def do_ignore(self,arg):
  		"""arg is bp number followed by ignore count."""
! 		args = string.split(arg)
! 		bpnum = int(string.strip(args[0]))
  		try:
! 			count = int(string.strip(args[1]))
  		except:
  			count = 0
  		bp = bdb.Breakpoint.bpbynumber[bpnum]
--- 388,397 ----
  
  	def do_ignore(self,arg):
  		"""arg is bp number followed by ignore count."""
! 		args = arg.split()
! 		bpnum = int(args[0].strip())
  		try:
! 			count = int(args[1].strip())
  		except:
  			count = 0
  		bp = bdb.Breakpoint.bpbynumber[bpnum]
***************
*** 421,433 ****
  				reply = raw_input('Clear all breaks? ')
  			except EOFError:
  				reply = 'no'
! 			reply = string.lower(string.strip(reply))
  			if reply in ('y', 'yes'):
  				self.clear_all_breaks()
  			return
  		if ':' in arg:
  			# Make sure it works for "clear C:\foo\bar.py:12"
! 			i = string.rfind(arg, ':')
  			filename = arg[:i]
  			arg = arg[i+1:]
  			try:
--- 418,430 ----
  				reply = raw_input('Clear all breaks? ')
  			except EOFError:
  				reply = 'no'
! 			reply = reply.strip().lower()
  			if reply in ('y', 'yes'):
  				self.clear_all_breaks()
  			return
  		if ':' in arg:
  			# Make sure it works for "clear C:\foo\bar.py:12"
! 			i = arg.rfind(':')
  			filename = arg[:i]
  			arg = arg[i+1:]
  			try:
***************
*** 438,444 ****
  				err = self.clear_break(filename, lineno)
  			if err: print '***', err
  			return
! 		numberlist = string.split(arg)
  		for i in numberlist:
  			err = self.clear_bpbynumber(i)
  			if err:
--- 435,441 ----
  				err = self.clear_break(filename, lineno)
  			if err: print '***', err
  			return
! 		numberlist = arg.split()
  		for i in numberlist:
  			err = self.clear_bpbynumber(i)
  			if err:
***************
*** 564,570 ****
  					print '[EOF]'
  					break
  				else:
! 					s = string.rjust(`lineno`, 3)
  					if len(s) < 4: s = s + ' '
  					if lineno in breaklist: s = s + 'B'
  					else: s = s + ' '
--- 561,567 ----
  					print '[EOF]'
  					break
  				else:
! 					s = `lineno`.rjust(3)
  					if len(s) < 4: s = s + ' '
  					if lineno in breaklist: s = s + 'B'
  					else: s = s + ' '
***************
*** 604,610 ****
  		print type(value)
  
  	def do_alias(self, arg):
! 		args = string.split (arg)
  		if len(args) == 0:
  			keys = self.aliases.keys()
  			keys.sort()
--- 601,607 ----
  		print type(value)
  
  	def do_alias(self, arg):
! 		args = arg.split ()
  		if len(args) == 0:
  			keys = self.aliases.keys()
  			keys.sort()
***************
*** 614,623 ****
  		if self.aliases.has_key(args[0]) and len (args) == 1:
  			print "%s = %s" % (args[0], self.aliases[args[0]])
  		else:
! 			self.aliases[args[0]] = string.join(args[1:], ' ')
  
  	def do_unalias(self, arg):
! 		args = string.split (arg)
  		if len(args) == 0: return
  		if self.aliases.has_key(args[0]):
  			del self.aliases[args[0]]
--- 611,620 ----
  		if self.aliases.has_key(args[0]) and len (args) == 1:
  			print "%s = %s" % (args[0], self.aliases[args[0]])
  		else:
! 			self.aliases[args[0]] = ' '.join(args[1:])
  
  	def do_unalias(self, arg):
! 		args = arg.split ()
  		if len(args) == 0: return
  		if self.aliases.has_key(args[0]):
  			del self.aliases[args[0]]
Index: Lib/pickle.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/pickle.py,v
retrieving revision 1.37
diff -c -r1.37 pickle.py
*** pickle.py	2000/03/10 23:20:09	1.37
--- pickle.py	2000/04/11 11:56:48
***************
*** 27,33 ****
  
  from types import *
  from copy_reg import dispatch_table, safe_constructors
- import string
  import marshal
  import sys
  import struct
--- 27,32 ----
***************
*** 548,554 ****
      dispatch[NONE] = load_none
  
      def load_int(self):
!         self.append(string.atoi(self.readline()[:-1]))
      dispatch[INT] = load_int
  
      def load_binint(self):
--- 547,553 ----
      dispatch[NONE] = load_none
  
      def load_int(self):
!         self.append(int(self.readline()[:-1]))
      dispatch[INT] = load_int
  
      def load_binint(self):
***************
*** 564,574 ****
      dispatch[BININT2] = load_binint2
   
      def load_long(self):
!         self.append(string.atol(self.readline()[:-1], 0))
      dispatch[LONG] = load_long
  
      def load_float(self):
!         self.append(string.atof(self.readline()[:-1]))
      dispatch[FLOAT] = load_float
  
      def load_binfloat(self, unpack=struct.unpack):
--- 563,573 ----
      dispatch[BININT2] = load_binint2
   
      def load_long(self):
!         self.append(long(self.readline()[:-1], 0))
      dispatch[LONG] = load_long
  
      def load_float(self):
!         self.append(float(self.readline()[:-1]))
      dispatch[FLOAT] = load_float
  
      def load_binfloat(self, unpack=struct.unpack):
Index: Lib/popen2.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/popen2.py,v
retrieving revision 1.9
diff -c -r1.9 popen2.py
*** popen2.py	2000/02/04 15:10:34	1.9
--- popen2.py	2000/04/11 11:56:48
***************
*** 8,14 ****
  
  import os
  import sys
- import string
  
  MAXFD = 256     # Max number of file descriptors (os.getdtablesize()???)
  
--- 8,13 ----
Index: Lib/poplib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/poplib.py,v
retrieving revision 1.8
diff -c -r1.8 poplib.py
*** poplib.py	2000/03/28 21:45:46	1.8
--- poplib.py	2000/04/11 11:56:48
***************
*** 15,21 ****
  
  # Imports
  
! import regex, socket, string
  
  # Exception raised when an error or invalid response is received:
  
--- 15,21 ----
  
  # Imports
  
! import regex, socket
  
  # Exception raised when an error or invalid response is received:
  
***************
*** 189,198 ****
  		Result is tuple of 2 ints (message count, mailbox size)
  		"""
  		retval = self._shortcmd('STAT')
! 		rets = string.split(retval)
  		#if self._debugging: print '*stat*', `rets`
! 		numMessages = string.atoi(rets[1])
! 		sizeMessages = string.atoi(rets[2])
  		return (numMessages, sizeMessages)
  
  
--- 189,198 ----
  		Result is tuple of 2 ints (message count, mailbox size)
  		"""
  		retval = self._shortcmd('STAT')
! 		rets = retval.split()
  		#if self._debugging: print '*stat*', `rets`
! 		numMessages = int(rets[1])
! 		sizeMessages = int(rets[2])
  		return (numMessages, sizeMessages)
  
  
***************
*** 277,283 ****
  			raise error_proto('-ERR APOP not supported by server')
  		import md5
  		digest = md5.new(self.timestamp.group(1)+secret).digest()
! 		digest = string.join(map(lambda x:'%02x'%ord(x), digest), '')
  		return self._shortcmd('APOP %s %s' % (user, digest))
  
  
--- 277,283 ----
  			raise error_proto('-ERR APOP not supported by server')
  		import md5
  		digest = md5.new(self.timestamp.group(1)+secret).digest()
! 		digest = ''.join(map(lambda x:'%02x'%ord(x), digest))
  		return self._shortcmd('APOP %s %s' % (user, digest))
  
  
Index: Lib/posixpath.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/posixpath.py,v
retrieving revision 1.30
diff -c -r1.30 posixpath.py
*** posixpath.py	2000/02/29 13:31:16	1.30
--- posixpath.py	2000/04/11 11:56:48
***************
*** 57,64 ****
  def split(p):
      """Split a pathname.  Returns tuple "(head, tail)" where "tail" is 
  everything after the final slash.  Either part may be empty"""
!     import string
!     i = string.rfind(p, '/') + 1
      head, tail = p[:i], p[i:]
      if head and head <> '/'*len(head):
          while head[-1] == '/':
--- 57,63 ----
  def split(p):
      """Split a pathname.  Returns tuple "(head, tail)" where "tail" is 
  everything after the final slash.  Either part may be empty"""
!     i = p.rfind('/') + 1
      head, tail = p[:i], p[i:]
      if head and head <> '/'*len(head):
          while head[-1] == '/':
***************
*** 342,354 ****
  
  def normpath(path):
      """Normalize path, eliminating double slashes, etc."""
-     import string
      # Treat initial slashes specially
      slashes = ''
      while path[:1] == '/':
          slashes = slashes + '/'
          path = path[1:]
!     comps = string.splitfields(path, '/')
      i = 0
      while i < len(comps):
          if comps[i] == '.':
--- 341,352 ----
  
  def normpath(path):
      """Normalize path, eliminating double slashes, etc."""
      # Treat initial slashes specially
      slashes = ''
      while path[:1] == '/':
          slashes = slashes + '/'
          path = path[1:]
!     comps = path.split('/')
      i = 0
      while i < len(comps):
          if comps[i] == '.':
***************
*** 365,371 ****
      # If the path is now empty, substitute '.'
      if not comps and not slashes:
          comps.append('.')
!     return slashes + string.joinfields(comps, '/')
  
  
  def abspath(path):
--- 363,369 ----
      # If the path is now empty, substitute '.'
      if not comps and not slashes:
          comps.append('.')
!     return slashes + '/'.join(comps)
  
  
  def abspath(path):
Index: Lib/pstats.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/pstats.py,v
retrieving revision 1.7
diff -c -r1.7 pstats.py
*** pstats.py	2000/02/04 15:10:34	1.7
--- pstats.py	2000/04/11 11:56:49
***************
*** 34,40 ****
  
  import os
  import time
- import string
  import marshal
  import re
  
--- 34,39 ----
***************
*** 392,402 ****
  		return self
  
  	def print_call_heading(self, name_size, column_title):
! 		print string.ljust("Function ", name_size) + column_title
  
  
  	def print_call_line(self, name_size, source, call_dict):
! 		print string.ljust(func_std_string(source), name_size),
  		if not call_dict:
  			print "--"
  			return
--- 391,401 ----
  		return self
  
  	def print_call_heading(self, name_size, column_title):
! 		print "Function ".ljust(name_size) + column_title
  
  
  	def print_call_line(self, name_size, source, call_dict):
! 		print func_std_string(source).ljust(name_size),
  		if not call_dict:
  			print "--"
  			return
***************
*** 414,424 ****
  
  
  	def print_title(self):
! 		print string.rjust('ncalls', 9),
! 		print string.rjust('tottime', 8),
! 		print string.rjust('percall', 8),
! 		print string.rjust('cumtime', 8),
! 		print string.rjust('percall', 8),
  		print 'filename:lineno(function)'
  
  
--- 413,423 ----
  
  
  	def print_title(self):
! 		print 'ncalls'.rjust(9),
! 		print 'tottime'.rjust(8),
! 		print 'percall'.rjust(8),
! 		print 'cumtime'.rjust(8),
! 		print 'percall'.rjust(8),
  		print 'filename:lineno(function)'
  
  
***************
*** 427,433 ****
  		c = `nc`
  		if nc != cc:
  			c = c + '/' + `cc`
! 		print string.rjust(c, 9),
  		print f8(tt),
  		if nc == 0:
  			print ' '*8,
--- 426,432 ----
  		c = `nc`
  		if nc != cc:
  			c = c + '/' + `cc`
! 		print c.rjust(9),
  		print f8(tt),
  		if nc == 0:
  			print ' '*8,
***************
*** 522,526 ****
  #**************************************************************************
  
  def f8(x):
! 	return string.rjust(fpformat.fix(x, 3), 8)
  
--- 521,525 ----
  #**************************************************************************
  
  def f8(x):
! 	return fpformat.fix(x, 3).rjust(8)
  
Index: Lib/py_compile.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/py_compile.py,v
retrieving revision 1.15
diff -c -r1.15 py_compile.py
*** py_compile.py	2000/02/04 15:10:34	1.15
--- py_compile.py	2000/04/11 11:56:49
***************
*** 54,63 ****
      try:
          codeobject = __builtin__.compile(codestring, dfile or file, 'exec')
      except SyntaxError, detail:
!         import traceback, sys, string
          lines = traceback.format_exception_only(SyntaxError, detail)
          for line in lines:
!             sys.stderr.write(string.replace(line, 'File "<string>"',
                                              'File "%s"' % (dfile or file)))
          return
      if not cfile:
--- 54,63 ----
      try:
          codeobject = __builtin__.compile(codestring, dfile or file, 'exec')
      except SyntaxError, detail:
!         import traceback, sys
          lines = traceback.format_exception_only(SyntaxError, detail)
          for line in lines:
!             sys.stderr.write(line.replace('File "<string>"',
                                              'File "%s"' % (dfile or file)))
          return
      if not cfile:
Index: Lib/pyclbr.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/pyclbr.py,v
retrieving revision 1.15
diff -c -r1.15 pyclbr.py
*** pyclbr.py	2000/02/04 15:39:30	1.15
--- pyclbr.py	2000/04/11 11:56:49
***************
*** 57,63 ****
  import sys
  import imp
  import re
- import string
  
  TABWIDTH = 8
  
--- 57,62 ----
***************
*** 160,170 ****
  
  	dict = {}
  
! 	i = string.rfind(module, '.')
  	if i >= 0:
  		# Dotted module name
! 		package = string.strip(module[:i])
! 		submodule = string.strip(module[i+1:])
  		parent = readmodule(package, path, inpackage)
  		child = readmodule(submodule, parent['__path__'], 1)
  		return child
--- 159,169 ----
  
  	dict = {}
  
! 	i = module.rfind('.')
  	if i >= 0:
  		# Dotted module name
! 		package = module[:i].strip()
! 		submodule = module[i+1:].strip()
  		parent = readmodule(package, path, inpackage)
  		child = readmodule(submodule, parent['__path__'], 1)
  		return child
***************
*** 210,218 ****
  	# when we need a line number we simply string.count the number of
  	# newlines in the string since the last time we did this; i.e.,
  	#    lineno = lineno + \
! 	#             string.count(src, '\n', last_lineno_pos, here)
  	#    last_lineno_pos = here
- 	countnl = string.count
  	lineno, last_lineno_pos = 1, 0
  	i = 0
  	while 1:
--- 209,216 ----
  	# when we need a line number we simply string.count the number of
  	# newlines in the string since the last time we did this; i.e.,
  	#    lineno = lineno + \
! 	#             src.count('\n', last_lineno_pos, here)
  	#    last_lineno_pos = here
  	lineno, last_lineno_pos = 1, 0
  	i = 0
  	while 1:
***************
*** 226,233 ****
  			thisindent = _indent(m.group("MethodIndent"))
  			meth_name = m.group("MethodName")
  			lineno = lineno + \
! 				 countnl(src, '\n',
! 					 last_lineno_pos, start)
  			last_lineno_pos = start
  			# close all classes indented at least as much
  			while classstack and \
--- 224,230 ----
  			thisindent = _indent(m.group("MethodIndent"))
  			meth_name = m.group("MethodName")
  			lineno = lineno + \
! 				 src.count('\n', last_lineno_pos, start)
  			last_lineno_pos = start
  			# close all classes indented at least as much
  			while classstack and \
***************
*** 254,274 ****
  			      classstack[-1][1] >= thisindent:
  				del classstack[-1]
  			lineno = lineno + \
! 				 countnl(src, '\n', last_lineno_pos, start)
  			last_lineno_pos = start
  			class_name = m.group("ClassName")
  			inherit = m.group("ClassSupers")
  			if inherit:
  				# the class inherits from other classes
! 				inherit = string.strip(inherit[1:-1])
  				names = []
! 				for n in string.splitfields(inherit, ','):
! 					n = string.strip(n)
  					if dict.has_key(n):
  						# we know this super class
  						n = dict[n]
  					else:
! 						c = string.splitfields(n, '.')
  						if len(c) > 1:
  							# super class
  							# is of the
--- 251,271 ----
  			      classstack[-1][1] >= thisindent:
  				del classstack[-1]
  			lineno = lineno + \
! 				 src.count('\n', last_lineno_pos, start)
  			last_lineno_pos = start
  			class_name = m.group("ClassName")
  			inherit = m.group("ClassSupers")
  			if inherit:
  				# the class inherits from other classes
! 				inherit = inherit[1:-1].strip()
  				names = []
! 				for n in inherit.split(','):
! 					n = n.strip()
  					if dict.has_key(n):
  						# we know this super class
  						n = dict[n]
  					else:
! 						c = n.split('.')
  						if len(c) > 1:
  							# super class
  							# is of the
***************
*** 291,298 ****
  
  		elif m.start("Import") >= 0:
  			# import module
! 			for n in string.split(m.group("ImportList"), ','):
! 				n = string.strip(n)
  				try:
  					# recursively read the imported module
  					d = readmodule(n, path, inpackage)
--- 288,295 ----
  
  		elif m.start("Import") >= 0:
  			# import module
! 			for n in m.group("ImportList").split(','):
! 				n = n.strip()
  				try:
  					# recursively read the imported module
  					d = readmodule(n, path, inpackage)
***************
*** 303,309 ****
  		elif m.start("ImportFrom") >= 0:
  			# from module import stuff
  			mod = m.group("ImportFromPath")
! 			names = string.split(m.group("ImportFromList"), ',')
  			try:
  				# recursively read the imported module
  				d = readmodule(mod, path, inpackage)
--- 300,306 ----
  		elif m.start("ImportFrom") >= 0:
  			# from module import stuff
  			mod = m.group("ImportFromPath")
! 			names = m.group("ImportFromList").split(',')
  			try:
  				# recursively read the imported module
  				d = readmodule(mod, path, inpackage)
***************
*** 314,320 ****
  			# imported module to our name space if they
  			# were mentioned in the list
  			for n in names:
! 				n = string.strip(n)
  				if d.has_key(n):
  					dict[n] = d[n]
  				elif n == '*':
--- 311,317 ----
  			# imported module to our name space if they
  			# were mentioned in the list
  			for n in names:
! 				n = n.strip()
  				if d.has_key(n):
  					dict[n] = d[n]
  				elif n == '*':
***************
*** 332,336 ****
  
  	return dict
  
! def _indent(ws, _expandtabs=string.expandtabs):
! 	return len(_expandtabs(ws, TABWIDTH))
--- 329,333 ----
  
  	return dict
  
! def _indent(ws):
! 	return len(ws.expandtabs(TABWIDTH))
Index: Lib/re.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/re.py,v
retrieving revision 1.35
diff -c -r1.35 re.py
*** re.py	1999/11/15 14:19:15	1.35
--- re.py	2000/04/11 11:56:49
***************
*** 229,235 ****
          if char not in alphanum:
              if char=='\000': result[i] = '\\000'
              else: result[i] = '\\'+char
!     return string.join(result, '')
  
  def compile(pattern, flags=0):
      """compile(pattern[, flags]) -> RegexObject
--- 229,235 ----
          if char not in alphanum:
              if char=='\000': result[i] = '\\000'
              else: result[i] = '\\'+char
!     return ''.join(result)
  
  def compile(pattern, flags=0):
      """compile(pattern[, flags]) -> RegexObject
***************
*** 398,404 ****
                  append(source[lastmatch:pos])
              n = n + 1
          append(source[pos:])
!         return (string.join(results, ''), n)
                                                                              
      def split(self, source, maxsplit=0):
          """split(source[, maxsplit=0]) -> list of strings
--- 398,404 ----
                  append(source[lastmatch:pos])
              n = n + 1
          append(source[pos:])
!         return (''.join(results), n)
                                                                              
      def split(self, source, maxsplit=0):
          """split(source[, maxsplit=0]) -> list of strings
Index: Lib/regsub.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/regsub.py,v
retrieving revision 1.7
diff -c -r1.7 regsub.py
*** regsub.py	2000/02/04 15:28:40	1.7
--- regsub.py	2000/04/11 11:56:49
***************
*** 100,110 ****
  # Capitalize words split using a pattern
  
  def capwords(str, pat='[^a-zA-Z0-9_]+'):
- 	import string
  	words = splitx(str, pat)
  	for i in range(0, len(words), 2):
! 		words[i] = string.capitalize(words[i])
! 	return string.joinfields(words, "")
  
  
  # Internal subroutines:
--- 100,109 ----
  # Capitalize words split using a pattern
  
  def capwords(str, pat='[^a-zA-Z0-9_]+'):
  	words = splitx(str, pat)
  	for i in range(0, len(words), 2):
! 		words[i] = words[i].capitalize()
! 	return "".join(words)
  
  
  # Internal subroutines:
Index: Lib/repr.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/repr.py,v
retrieving revision 1.6
diff -c -r1.6 repr.py
*** repr.py	2000/02/04 15:28:40	1.6
--- repr.py	2000/04/11 11:56:50
***************
*** 1,7 ****
  """Redo the `...` (representation) but with limits on most sizes."""
  
- import string
- 
  class Repr:
  	def __init__(self):
  		self.maxlevel = 6
--- 1,5 ----
***************
*** 16,23 ****
  	def repr1(self, x, level):
  		typename = `type(x)`[7:-2] # "<type '......'>"
  		if ' ' in typename:
! 			parts = string.split(typename)
! 			typename = string.joinfields(parts, '_')
  		if hasattr(self, 'repr_' + typename):
  			return getattr(self, 'repr_' + typename)(x, level)
  		else:
--- 14,21 ----
  	def repr1(self, x, level):
  		typename = `type(x)`[7:-2] # "<type '......'>"
  		if ' ' in typename:
! 			parts = typename.split()
! 			typename = '_'.join(parts)
  		if hasattr(self, 'repr_' + typename):
  			return getattr(self, 'repr_' + typename)(x, level)
  		else:
Index: Lib/rfc822.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/rfc822.py,v
retrieving revision 1.46
diff -c -r1.46 rfc822.py
*** rfc822.py	2000/02/10 17:17:14	1.46
--- rfc822.py	2000/04/11 11:56:50
***************
*** 145,152 ****
              if headerseen and line[0] in ' \t':
                  # It's a continuation line.
                  list.append(line)
!                 x = (self.dict[headerseen] + "\n " + string.strip(line))
!                 self.dict[headerseen] = string.strip(x)
                  continue
              elif self.iscomment(line):
                  # It's a comment.  Ignore it.
--- 145,152 ----
              if headerseen and line[0] in ' \t':
                  # It's a continuation line.
                  list.append(line)
!                 x = (self.dict[headerseen] + "\n " + line.strip())
!                 self.dict[headerseen] = x.strip()
                  continue
              elif self.iscomment(line):
                  # It's a comment.  Ignore it.
***************
*** 158,164 ****
              if headerseen:
                  # It's a legal header line, save it.
                  list.append(line)
!                 self.dict[headerseen] = string.strip(line[len(headerseen)+1:])
                  continue
              else:
                  # It's not a header line; throw it back and stop here.
--- 158,164 ----
              if headerseen:
                  # It's a legal header line, save it.
                  list.append(line)
!                 self.dict[headerseen] = line[len(headerseen)+1:].strip()
                  continue
              else:
                  # It's not a header line; throw it back and stop here.
***************
*** 182,190 ****
          You may override this method in order to use Message parsing
          on tagged data in RFC822-like formats with special header formats.
          """
!         i = string.find(line, ':')
          if i > 0:
!             return string.lower(line[:i])
          else:
              return None
      
--- 182,190 ----
          You may override this method in order to use Message parsing
          on tagged data in RFC822-like formats with special header formats.
          """
!         i = line.find(':')
          if i > 0:
!             return line[:i].lower()
          else:
              return None
      
***************
*** 219,230 ****
          times, all occurrences are returned.  Case is not
          important in the header name.
          """
!         name = string.lower(name) + ':'
          n = len(name)
          list = []
          hit = 0
          for line in self.headers:
!             if string.lower(line[:n]) == name:
                  hit = 1
              elif line[:1] not in string.whitespace:
                  hit = 0
--- 219,230 ----
          times, all occurrences are returned.  Case is not
          important in the header name.
          """
!         name = name.lower() + ':'
          n = len(name)
          list = []
          hit = 0
          for line in self.headers:
!             if line[:n].lower() == name:
                  hit = 1
              elif line[:1] not in string.whitespace:
                  hit = 0
***************
*** 239,245 ****
          only the first matching header (and its continuation
          lines).
          """
!         name = string.lower(name) + ':'
          n = len(name)
          list = []
          hit = 0
--- 239,245 ----
          only the first matching header (and its continuation
          lines).
          """
!         name = name.lower() + ':'
          n = len(name)
          list = []
          hit = 0
***************
*** 247,253 ****
              if hit:
                  if line[:1] not in string.whitespace:
                      break
!             elif string.lower(line[:n]) == name:
                  hit = 1
              if hit:
                  list.append(line)
--- 247,253 ----
              if hit:
                  if line[:1] not in string.whitespace:
                      break
!             elif line[:n].lower() == name:
                  hit = 1
              if hit:
                  list.append(line)
***************
*** 267,273 ****
          if not list:
              return None
          list[0] = list[0][len(name) + 1:]
!         return string.joinfields(list, '')
      
      def getheader(self, name, default=None):
          """Get the header value for a name.
--- 267,273 ----
          if not list:
              return None
          list[0] = list[0][len(name) + 1:]
!         return ''.join(list)
      
      def getheader(self, name, default=None):
          """Get the header value for a name.
***************
*** 278,284 ****
          version which finds the *last* such header.
          """
          try:
!             return self.dict[string.lower(name)]
          except KeyError:
              return default
      get = getheader
--- 278,284 ----
          version which finds the *last* such header.
          """
          try:
!             return self.dict[name.lower()]
          except KeyError:
              return default
      get = getheader
***************
*** 297,309 ****
          for s in self.getallmatchingheaders(name):
              if s[0] in string.whitespace:
                  if current:
!                     current = "%s\n %s" % (current, string.strip(s))
                  else:
!                     current = string.strip(s)
              else:
                  if have_header:
                      result.append(current)
!                 current = string.strip(s[string.find(s, ":") + 1:])
                  have_header = 1
          if have_header:
              result.append(current)
--- 297,309 ----
          for s in self.getallmatchingheaders(name):
              if s[0] in string.whitespace:
                  if current:
!                     current = "%s\n %s" % (current, s.strip())
                  else:
!                     current = s.strip()
              else:
                  if have_header:
                      result.append(current)
!                 current = s[s.find(":") + 1:].strip()
                  have_header = 1
          if have_header:
              result.append(current)
***************
*** 337,347 ****
              else:
                  if raw:
                      raw.append(', ')
!                 i = string.find(h, ':')
                  if i > 0:
                      addr = h[i+1:]
                  raw.append(addr)
!         alladdrs = string.join(raw, '')
          a = AddrlistClass(alladdrs)
          return a.getaddrlist()
      
--- 337,347 ----
              else:
                  if raw:
                      raw.append(', ')
!                 i = h.find(':')
                  if i > 0:
                      addr = h[i+1:]
                  raw.append(addr)
!         alladdrs = ''.join(raw)
          a = AddrlistClass(alladdrs)
          return a.getaddrlist()
      
***************
*** 379,385 ****
      
      def __getitem__(self, name):
          """Get a specific header, as from a dictionary."""
!         return self.dict[string.lower(name)]
  
      def __setitem__(self, name, value):
          """Set the value of a header.
--- 379,385 ----
      
      def __getitem__(self, name):
          """Get a specific header, as from a dictionary."""
!         return self.dict[name.lower()]
  
      def __setitem__(self, name, value):
          """Set the value of a header.
***************
*** 389,403 ****
          rather than where the altered header was.
          """
          del self[name] # Won't fail if it doesn't exist
!         self.dict[string.lower(name)] = value
          text = name + ": " + value
!         lines = string.split(text, "\n")
          for line in lines:
              self.headers.append(line + "\n")
      
      def __delitem__(self, name):
          """Delete all occurrences of a specific header, if it is present."""
!         name = string.lower(name)
          if not self.dict.has_key(name):
              return
          del self.dict[name]
--- 389,403 ----
          rather than where the altered header was.
          """
          del self[name] # Won't fail if it doesn't exist
!         self.dict[name.lower()] = value
          text = name + ": " + value
!         lines = text.split("\n")
          for line in lines:
              self.headers.append(line + "\n")
      
      def __delitem__(self, name):
          """Delete all occurrences of a specific header, if it is present."""
!         name = name.lower()
          if not self.dict.has_key(name):
              return
          del self.dict[name]
***************
*** 407,413 ****
          hit = 0
          for i in range(len(self.headers)):
              line = self.headers[i]
!             if string.lower(line[:n]) == name:
                  hit = 1
              elif line[:1] not in string.whitespace:
                  hit = 0
--- 407,413 ----
          hit = 0
          for i in range(len(self.headers)):
              line = self.headers[i]
!             if line[:n].lower() == name:
                  hit = 1
              elif line[:1] not in string.whitespace:
                  hit = 0
***************
*** 419,425 ****
  
      def has_key(self, name):
          """Determine whether a message contains the named header."""
!         return self.dict.has_key(string.lower(name))
      
      def keys(self):
          """Get all of a message's header field names."""
--- 419,425 ----
  
      def has_key(self, name):
          """Determine whether a message contains the named header."""
!         return self.dict.has_key(name.lower())
      
      def keys(self):
          """Get all of a message's header field names."""
***************
*** 462,476 ****
  
  def quote(str):
      """Add quotes around a string."""
!     return '"%s"' % string.join(
!     string.split(
!     string.join(
!     string.split(str, '\\'),
!     '\\\\'),
!     '"'),
!     '\\"')
  
- 
  def parseaddr(address):
      """Parse an address into a (realname, mailaddr) tuple."""
      a = AddrlistClass(address)
--- 462,469 ----
  
  def quote(str):
      """Add quotes around a string."""
!     return '"%s"' % str.replace ('\\', '\\\\').replace ('"', '\\"')
  
  def parseaddr(address):
      """Parse an address into a (realname, mailaddr) tuple."""
      a = AddrlistClass(address)
***************
*** 539,545 ****
          if self.pos >= len(self.field):
              # Bad email address technically, no domain.
              if plist:
!                 returnlist = [(string.join(self.commentlist), plist[0])]
              
          elif self.field[self.pos] in '.@':
              # email address is just an addrspec
--- 532,538 ----
          if self.pos >= len(self.field):
              # Bad email address technically, no domain.
              if plist:
!                 returnlist = [(' '.join(self.commentlist), plist[0])]
              
          elif self.field[self.pos] in '.@':
              # email address is just an addrspec
***************
*** 547,553 ****
              self.pos = oldpos
              self.commentlist = oldcl
              addrspec = self.getaddrspec()
!             returnlist = [(string.join(self.commentlist), addrspec)]
              
          elif self.field[self.pos] == ':':
              # address is a group
--- 540,546 ----
              self.pos = oldpos
              self.commentlist = oldcl
              addrspec = self.getaddrspec()
!             returnlist = [(' '.join(self.commentlist), addrspec)]
              
          elif self.field[self.pos] == ':':
              # address is a group
***************
*** 567,579 ****
              routeaddr = self.getrouteaddr()
              
              if self.commentlist:
!                 returnlist = [(string.join(plist) + ' (' + \
!                          string.join(self.commentlist) + ')', routeaddr)]
!             else: returnlist = [(string.join(plist), routeaddr)]
              
          else:
              if plist:
!                 returnlist = [(string.join(self.commentlist), plist[0])]
              elif self.field[self.pos] in self.specials:
                  self.pos = self.pos + 1
          
--- 560,572 ----
              routeaddr = self.getrouteaddr()
              
              if self.commentlist:
!                 returnlist = [(' '.join(plist) + ' (' + \
!                          ' '.join(self.commentlist) + ')', routeaddr)]
!             else: returnlist = [(' '.join(plist), routeaddr)]
              
          else:
              if plist:
!                 returnlist = [(' '.join(self.commentlist), plist[0])]
              elif self.field[self.pos] in self.specials:
                  self.pos = self.pos + 1
          
***************
*** 632,643 ****
              self.gotonext()
          
          if self.pos >= len(self.field) or self.field[self.pos] != '@':
!             return string.join(aslist, '')
          
          aslist.append('@')
          self.pos = self.pos + 1
          self.gotonext()
!         return string.join(aslist, '') + self.getdomain()
      
      def getdomain(self):
          """Get the complete domain name from an address."""
--- 625,636 ----
              self.gotonext()
          
          if self.pos >= len(self.field) or self.field[self.pos] != '@':
!             return ''.join(aslist)
          
          aslist.append('@')
          self.pos = self.pos + 1
          self.gotonext()
!         return ''.join(aslist) + self.getdomain()
      
      def getdomain(self):
          """Get the complete domain name from an address."""
***************
*** 655,661 ****
              elif self.field[self.pos] in self.atomends:
                  break
              else: sdlist.append(self.getatom())
!         return string.join(sdlist, '')
      
      def getdelimited(self, beginchar, endchars, allowcomments = 1):
          """Parse a header fragment delimited by special characters.
--- 648,654 ----
              elif self.field[self.pos] in self.atomends:
                  break
              else: sdlist.append(self.getatom())
!         return ''.join(sdlist)
      
      def getdelimited(self, beginchar, endchars, allowcomments = 1):
          """Parse a header fragment delimited by special characters.
***************
*** 691,697 ****
                  slist.append(self.field[self.pos])
              self.pos = self.pos + 1
          
!         return string.join(slist, '')
      
      def getquote(self):
          """Get a quote-delimited fragment from self's field."""
--- 684,690 ----
                  slist.append(self.field[self.pos])
              self.pos = self.pos + 1
          
!         return ''.join(slist)
      
      def getquote(self):
          """Get a quote-delimited fragment from self's field."""
***************
*** 715,721 ****
              else: atomlist.append(self.field[self.pos])
              self.pos = self.pos + 1
          
!         return string.join(atomlist, '')
      
      def getphraselist(self):
          """Parse a sequence of RFC-822 phrases.
--- 708,714 ----
              else: atomlist.append(self.field[self.pos])
              self.pos = self.pos + 1
          
!         return ''.join(atomlist)
      
      def getphraselist(self):
          """Parse a sequence of RFC-822 phrases.
***************
*** 752,758 ****
          return len(self.addresslist)
  
      def __str__(self):
!         return string.joinfields(map(dump_address_pair, self.addresslist),", ")
  
      def __add__(self, other):
          # Set union
--- 745,751 ----
          return len(self.addresslist)
  
      def __str__(self):
!         return ", ".join(map(dump_address_pair, self.addresslist))
  
      def __add__(self, other):
          # Set union
***************
*** 810,826 ****
      
      Accounts for military timezones.
      """
!     data = string.split(data)
!     if data[0][-1] in (',', '.') or string.lower(data[0]) in _daynames:
          # There's a dayname here. Skip it
          del data[0]
      if len(data) == 3: # RFC 850 date, deprecated
!         stuff = string.split(data[0], '-')
          if len(stuff) == 3:
              data = stuff + data[1:]
      if len(data) == 4:
          s = data[3]
!         i = string.find(s, '+')
          if i > 0:
              data[3:] = [s[:i], s[i+1:]]
          else:
--- 803,819 ----
      
      Accounts for military timezones.
      """
!     data = data.split()
!     if data[0][-1] in (',', '.') or data[0].lower() in _daynames:
          # There's a dayname here. Skip it
          del data[0]
      if len(data) == 3: # RFC 850 date, deprecated
!         stuff = data[0].split('-')
          if len(stuff) == 3:
              data = stuff + data[1:]
      if len(data) == 4:
          s = data[3]
!         i = s.find('+')
          if i > 0:
              data[3:] = [s[:i], s[i+1:]]
          else:
***************
*** 829,844 ****
          return None
      data = data[:5]
      [dd, mm, yy, tm, tz] = data
!     mm = string.lower(mm)
      if not mm in _monthnames:
!         dd, mm = mm, string.lower(dd)
          if not mm in _monthnames:
              return None
      mm = _monthnames.index(mm)+1
      if mm > 12: mm = mm - 12
      if dd[-1] == ',':
          dd = dd[:-1]
!     i = string.find(yy, ':')
      if i > 0:
          yy, tm = tm, yy
      if yy[-1] == ',':
--- 822,837 ----
          return None
      data = data[:5]
      [dd, mm, yy, tm, tz] = data
!     mm = mm.lower()
      if not mm in _monthnames:
!         dd, mm = mm, dd.lower()
          if not mm in _monthnames:
              return None
      mm = _monthnames.index(mm)+1
      if mm > 12: mm = mm - 12
      if dd[-1] == ',':
          dd = dd[:-1]
!     i = yy.find(':')
      if i > 0:
          yy, tm = tm, yy
      if yy[-1] == ',':
***************
*** 847,853 ****
          yy, tz = tz, yy
      if tm[-1] == ',':
          tm = tm[:-1]
!     tm = string.splitfields(tm, ':')
      if len(tm) == 2:
          [thh, tmm] = tm
          tss = '0'
--- 840,846 ----
          yy, tz = tz, yy
      if tm[-1] == ',':
          tm = tm[:-1]
!     tm = tm.split(':')
      if len(tm) == 2:
          [thh, tmm] = tm
          tss = '0'
***************
*** 856,876 ****
      else:
          return None
      try:
!         yy = string.atoi(yy)
!         dd = string.atoi(dd)
!         thh = string.atoi(thh)
!         tmm = string.atoi(tmm)
!         tss = string.atoi(tss)
!     except string.atoi_error:
          return None
      tzoffset=None
!     tz=string.upper(tz)
      if _timezones.has_key(tz):
          tzoffset=_timezones[tz]
      else:
          try: 
!             tzoffset=string.atoi(tz)
!         except string.atoi_error: 
              pass
      # Convert a timezone offset into seconds ; -0500 -> -18000
      if tzoffset:
--- 849,869 ----
      else:
          return None
      try:
!         yy = int(yy)
!         dd = int(dd)
!         thh = int(thh)
!         tmm = int(tmm)
!         tss = int(tss)
!     except ValueError:
          return None
      tzoffset=None
!     tz=tz.upper()
      if _timezones.has_key(tz):
          tzoffset=_timezones[tz]
      else:
          try: 
!             tzoffset=int(tz)
!         except ValueError: 
              pass
      # Convert a timezone offset into seconds ; -0500 -> -18000
      if tzoffset:
Index: Lib/robotparser.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/robotparser.py,v
retrieving revision 1.3
diff -c -r1.3 robotparser.py
*** robotparser.py	2000/03/27 19:29:31	1.3
--- robotparser.py	2000/04/11 11:56:50
***************
*** 30,36 ****
  
      def parse(self, lines):
          """parse the input lines from a robot.txt file"""
!         import string, re
          active = []
          for line in lines:
              if self.debug: print '>', line,
--- 30,36 ----
  
      def parse(self, lines):
          """parse the input lines from a robot.txt file"""
!         import re
          active = []
          for line in lines:
              if self.debug: print '>', line,
***************
*** 39,50 ****
                  active = []
                  continue
              # remove optional comment and strip line
!             line = string.strip(line[:string.find(line, '#')])
              if not line:
                  continue
              line = re.split(' *: *', line)
              if len(line) == 2:
!                 line[0] = string.lower(line[0])
                  if line[0] == 'user-agent':
                      # this record applies to this user agent
                      if self.debug: print '>> user-agent:', line[1]
--- 39,50 ----
                  active = []
                  continue
              # remove optional comment and strip line
!             line = line[:line.find('#')].strip()
              if not line:
                  continue
              line = re.split(' *: *', line)
              if len(line) == 2:
!                 line[0] = line[0].lower()
                  if line[0] == 'user-agent':
                      # this record applies to this user agent
                      if self.debug: print '>> user-agent:', line[1]
Index: Lib/sgmllib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/sgmllib.py,v
retrieving revision 1.19
diff -c -r1.19 sgmllib.py
*** sgmllib.py	2000/02/04 15:28:40	1.19
--- sgmllib.py	2000/04/11 11:56:51
***************
*** 220,226 ****
              if not match:
                  return -1
              tag, data = match.group(1, 2)
!             tag = string.lower(tag)
              self.finish_shorttag(tag, data)
              k = match.end(0)
              return k
--- 220,226 ----
              if not match:
                  return -1
              tag, data = match.group(1, 2)
!             tag = tag.lower()
              self.finish_shorttag(tag, data)
              k = match.end(0)
              return k
***************
*** 240,246 ****
              if not match:
                  raise RuntimeError, 'unexpected call to parse_starttag'
              k = match.end(0)
!             tag = string.lower(rawdata[i+1:k])
              self.lasttag = tag
          while k < j:
              match = attrfind.match(rawdata, k)
--- 240,246 ----
              if not match:
                  raise RuntimeError, 'unexpected call to parse_starttag'
              k = match.end(0)
!             tag = rawdata[i+1:k].lower()
              self.lasttag = tag
          while k < j:
              match = attrfind.match(rawdata, k)
***************
*** 251,257 ****
              elif attrvalue[:1] == '\'' == attrvalue[-1:] or \
                   attrvalue[:1] == '"' == attrvalue[-1:]:
                  attrvalue = attrvalue[1:-1]
!             attrs.append((string.lower(attrname), attrvalue))
              k = match.end(0)
          if rawdata[j] == '>':
              j = j+1
--- 251,257 ----
              elif attrvalue[:1] == '\'' == attrvalue[-1:] or \
                   attrvalue[:1] == '"' == attrvalue[-1:]:
                  attrvalue = attrvalue[1:-1]
!             attrs.append((attrname.lower(), attrvalue))
              k = match.end(0)
          if rawdata[j] == '>':
              j = j+1
***************
*** 265,271 ****
          if not match:
              return -1
          j = match.start(0)
!         tag = string.lower(string.strip(rawdata[i+2:j]))
          if rawdata[j] == '>':
              j = j+1
          self.finish_endtag(tag)
--- 265,271 ----
          if not match:
              return -1
          j = match.start(0)
!         tag = rawdata[i+2:j].strip().lower()
          if rawdata[j] == '>':
              j = j+1
          self.finish_endtag(tag)
***************
*** 344,351 ****
      # Example -- handle character reference, no need to override
      def handle_charref(self, name):
          try:
!             n = string.atoi(name)
!         except string.atoi_error:
              self.unknown_charref(name)
              return
          if not 0 <= n <= 255:
--- 344,351 ----
      # Example -- handle character reference, no need to override
      def handle_charref(self, name):
          try:
!             n = int(name)
!         except ValueError:
              self.unknown_charref(name)
              return
          if not 0 <= n <= 255:
Index: Lib/smtplib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/smtplib.py,v
retrieving revision 1.23
diff -c -r1.23 smtplib.py
*** smtplib.py	2000/03/28 21:45:46	1.23
--- smtplib.py	2000/04/11 11:56:51
***************
*** 40,46 ****
  # This was modified from the Python 1.5 library HTTP lib.
  
  import socket
- import string
  import re
  import rfc822
  import types
--- 40,45 ----
***************
*** 204,214 ****
  
          """
          if not port:
!             i = string.find(host, ':')
              if i >= 0:
                  host, port = host[:i], host[i+1:]
!                 try: port = string.atoi(port)
!                 except string.atoi_error:
                      raise socket.error, "nonnumeric port"
          if not port: port = SMTP_PORT
          self.sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
--- 203,213 ----
  
          """
          if not port:
!             i = host.find(':')
              if i >= 0:
                  host, port = host[:i], host[i+1:]
!                 try: port = int(port)
!                 except ValueError:
                      raise socket.error, "nonnumeric port"
          if not port: port = SMTP_PORT
          self.sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
***************
*** 259,270 ****
                  self.close()
                  raise SMTPServerDisconnected("Connection unexpectedly closed")
              if self.debuglevel > 0: print 'reply:', `line`
!             resp.append(string.strip(line[4:]))
              code=line[:3]
              # Check that the error code is syntactically correct.
              # Don't attempt to read a continuation line if it is broken.
              try:
!                 errcode = string.atoi(code)
              except ValueError:
                  errcode = -1
                  break
--- 258,269 ----
                  self.close()
                  raise SMTPServerDisconnected("Connection unexpectedly closed")
              if self.debuglevel > 0: print 'reply:', `line`
!             resp.append(line[4:].strip())
              code=line[:3]
              # Check that the error code is syntactically correct.
              # Don't attempt to read a continuation line if it is broken.
              try:
!                 errcode = int(code)
              except ValueError:
                  errcode = -1
                  break
***************
*** 272,278 ****
              if line[3:4]!="-":
                  break
  
!         errmsg = string.join(resp,"\n")
          if self.debuglevel > 0: 
              print 'reply: retcode (%s); Msg: %s' % (errcode,errmsg)
          return errcode, errmsg
--- 271,277 ----
              if line[3:4]!="-":
                  break
  
!         errmsg = "\n".join(resp)
          if self.debuglevel > 0: 
              print 'reply: retcode (%s); Msg: %s' % (errcode,errmsg)
          return errcode, errmsg
***************
*** 288,294 ****
          Hostname to send for this command defaults to the FQDN of the local
          host.
          """
!         name=string.strip(name)
          if len(name)==0:
              name = socket.gethostname()
              try:
--- 287,293 ----
          Hostname to send for this command defaults to the FQDN of the local
          host.
          """
!         name=name.strip()
          if len(name)==0:
              name = socket.gethostname()
              try:
***************
*** 305,311 ****
          Hostname to send for this command defaults to the FQDN of the local
          host.
          """
!         name=string.strip(name)
          if len(name)==0:
              name = socket.gethostname()
              try:
--- 304,310 ----
          Hostname to send for this command defaults to the FQDN of the local
          host.
          """
!         name=name.strip()
          if len(name)==0:
              name = socket.gethostname()
              try:
***************
*** 324,342 ****
              return (code,msg)
          self.does_esmtp=1
          #parse the ehlo responce -ddm
!         resp=string.split(self.ehlo_resp,'\n')
          del resp[0]
          for each in resp:
              m=re.match(r'(?P<feature>[A-Za-z0-9][A-Za-z0-9\-]*)',each)
              if m:
!                 feature=string.lower(m.group("feature"))
!                 params=string.strip(m.string[m.end("feature"):])
                  self.esmtp_features[feature]=params
          return (code,msg)
  
      def has_extn(self, opt):
          """Does the server support a given SMTP service extension?"""
!         return self.esmtp_features.has_key(string.lower(opt))
  
      def help(self, args=''):
          """SMTP 'help' command.
--- 323,341 ----
              return (code,msg)
          self.does_esmtp=1
          #parse the ehlo responce -ddm
!         resp=self.ehlo_resp.split('\n')
          del resp[0]
          for each in resp:
              m=re.match(r'(?P<feature>[A-Za-z0-9][A-Za-z0-9\-]*)',each)
              if m:
!                 feature=m.group("feature").lower()
!                 params=m.string[m.end("feature"):].strip()
                  self.esmtp_features[feature]=params
          return (code,msg)
  
      def has_extn(self, opt):
          """Does the server support a given SMTP service extension?"""
!         return self.esmtp_features.has_key(opt.lower())
  
      def help(self, args=''):
          """SMTP 'help' command.
***************
*** 356,362 ****
          """SMTP 'mail' command -- begins mail xfer session."""
          optionlist = ''
          if options and self.does_esmtp:
!             optionlist = ' ' + string.join(options, ' ')
          self.putcmd("mail", "FROM:%s%s" % (quoteaddr(sender) ,optionlist))
          return self.getreply()
  
--- 355,361 ----
          """SMTP 'mail' command -- begins mail xfer session."""
          optionlist = ''
          if options and self.does_esmtp:
!             optionlist = ' ' + ' '.join(options)
          self.putcmd("mail", "FROM:%s%s" % (quoteaddr(sender) ,optionlist))
          return self.getreply()
  
***************
*** 364,370 ****
          """SMTP 'rcpt' command -- indicates 1 recipient for this mail."""
          optionlist = ''
          if options and self.does_esmtp:
!             optionlist = ' ' + string.join(options, ' ')
          self.putcmd("rcpt","TO:%s%s" % (quoteaddr(recip),optionlist))
          return self.getreply()
  
--- 363,369 ----
          """SMTP 'rcpt' command -- indicates 1 recipient for this mail."""
          optionlist = ''
          if options and self.does_esmtp:
!             optionlist = ' ' + ' '.join(options)
          self.putcmd("rcpt","TO:%s%s" % (quoteaddr(recip),optionlist))
          return self.getreply()
  
***************
*** 521,530 ****
  
      def prompt(prompt):
          sys.stdout.write(prompt + ": ")
!         return string.strip(sys.stdin.readline())
  
      fromaddr = prompt("From")
!     toaddrs  = string.splitfields(prompt("To"), ',')
      print "Enter message, end with ^D:"
      msg = ''
      while 1:
--- 520,529 ----
  
      def prompt(prompt):
          sys.stdout.write(prompt + ": ")
!         return sys.stdin.readline().strip()
  
      fromaddr = prompt("From")
!     toaddrs  = prompt("To").split(',')
      print "Enter message, end with ^D:"
      msg = ''
      while 1:
Index: Lib/sre_compile.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/sre_compile.py,v
retrieving revision 1.3
diff -c -r1.3 sre_compile.py
*** sre_compile.py	2000/04/10 17:10:48	1.3
--- sre_compile.py	2000/04/11 11:56:51
***************
*** 17,23 ****
  # FIXME: <fl> formalize (objectify?) and document the compiler code
  # format, so that other frontends can use this compiler
  
! import array, string, sys
  
  import _sre
  
--- 17,23 ----
  # FIXME: <fl> formalize (objectify?) and document the compiler code
  # format, so that other frontends can use this compiler
  
! import array, sys
  
  import _sre
  
***************
*** 49,55 ****
  
  def _lower(literal):
      # return _sre._lower(literal) # FIXME
!     return string.lower(literal)
  
  def _compile(code, pattern, flags):
      append = code.append
--- 49,55 ----
  
  def _lower(literal):
      # return _sre._lower(literal) # FIXME
!     return literal.lower()
  
  def _compile(code, pattern, flags):
      append = code.append
Index: Lib/sre_constants.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/sre_constants.py,v
retrieving revision 1.3
diff -c -r1.3 sre_constants.py
*** sre_constants.py	2000/04/10 17:10:48	1.3
--- sre_constants.py	2000/04/11 11:56:51
***************
*** 120,131 ****
  }
  
  if __name__ == "__main__":
-     import string
      items = CODES.items()
      items.sort(lambda a, b: cmp(a[1], b[1]))
      f = open("sre_constants.h", "w")
      f.write("/* generated by sre_constants.py */\n")
      for k, v in items:
! 	f.write("#define SRE_OP_" + string.upper(k) + " " + str(v) + "\n")
      f.close()
      print "done"
--- 120,130 ----
  }
  
  if __name__ == "__main__":
      items = CODES.items()
      items.sort(lambda a, b: cmp(a[1], b[1]))
      f = open("sre_constants.h", "w")
      f.write("/* generated by sre_constants.py */\n")
      for k, v in items:
! 	f.write("#define SRE_OP_" + k.upper() + " " + str(v) + "\n")
      f.close()
      print "done"
Index: Lib/tabnanny.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/tabnanny.py,v
retrieving revision 1.10
diff -c -r1.10 tabnanny.py
*** tabnanny.py	2000/02/23 15:32:19	1.10
--- tabnanny.py	2000/04/11 11:56:52
***************
*** 12,18 ****
  
  import os
  import sys
- import string
  import getopt
  import tokenize
  
--- 12,17 ----
***************
*** 236,247 ****
          return a
  
  def format_witnesses(w):
-     import string
      firsts = map(lambda tup: str(tup[0]), w)
      prefix = "at tab size"
      if len(w) > 1:
          prefix = prefix + "s"
!     return prefix + " " + string.join(firsts, ', ')
  
  # The collection of globals, the reset_globals() function, and the
  # tokeneater() function, depend on which version of tokenize is
--- 235,245 ----
          return a
  
  def format_witnesses(w):
      firsts = map(lambda tup: str(tup[0]), w)
      prefix = "at tab size"
      if len(w) > 1:
          prefix = prefix + "s"
!     return prefix + " " + ', '.join(firsts)
  
  # The collection of globals, the reset_globals() function, and the
  # tokeneater() function, depend on which version of tokenize is
Index: Lib/telnetlib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/telnetlib.py,v
retrieving revision 1.7
diff -c -r1.7 telnetlib.py
*** telnetlib.py	1998/12/23 23:04:17	1.7
--- telnetlib.py	2000/04/11 11:56:52
***************
*** 40,46 ****
  import sys
  import socket
  import select
- import string
  
  # Tunable parameters
  DEBUGLEVEL = 0
--- 40,45 ----
***************
*** 187,193 ****
  
          """
          if IAC in buffer:
!             buffer = string.replace(buffer, IAC, IAC+IAC)
          self.msg("send %s", `buffer`)
          self.sock.send(buffer)
  
--- 186,192 ----
  
          """
          if IAC in buffer:
!             buffer = buffer.replace(IAC, IAC+IAC)
          self.msg("send %s", `buffer`)
          self.sock.send(buffer)
  
***************
*** 201,207 ****
          """
          n = len(match)
          self.process_rawq()
!         i = string.find(self.cookedq, match)
          if i >= 0:
              i = i+n
              buf = self.cookedq[:i]
--- 200,206 ----
          """
          n = len(match)
          self.process_rawq()
!         i = self.cookedq.find(match)
          if i >= 0:
              i = i+n
              buf = self.cookedq[:i]
***************
*** 215,221 ****
              i = max(0, len(self.cookedq)-n)
              self.fill_rawq()
              self.process_rawq()
!             i = string.find(self.cookedq, match, i)
              if i >= 0:
                  i = i+n
                  buf = self.cookedq[:i]
--- 214,220 ----
              i = max(0, len(self.cookedq)-n)
              self.fill_rawq()
              self.process_rawq()
!             i = self.cookedq.find(match, i)
              if i >= 0:
                  i = i+n
                  buf = self.cookedq[:i]
Index: Lib/token.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/token.py,v
retrieving revision 1.7
diff -c -r1.7 token.py
*** token.py	2000/02/04 15:28:41	1.7
--- token.py	2000/04/11 11:56:52
***************
*** 71,77 ****
  
  def main():
      import re
-     import string
      import sys
      args = sys.argv[1:]
      inFileName = args and args[0] or "Include/token.h"
--- 71,76 ----
***************
*** 83,89 ****
      except IOError, err:
          sys.stdout.write("I/O error: %s\n" % str(err))
          sys.exit(1)
!     lines = string.splitfields(fp.read(), "\n")
      fp.close()
      prog = re.compile(
          "#define[ \t][ \t]*([A-Z][A-Z_]*)[ \t][ \t]*([0-9][0-9]*)",
--- 82,88 ----
      except IOError, err:
          sys.stdout.write("I/O error: %s\n" % str(err))
          sys.exit(1)
!     lines = fp.read().split("\n")
      fp.close()
      prog = re.compile(
          "#define[ \t][ \t]*([A-Z][A-Z_]*)[ \t][ \t]*([0-9][0-9]*)",
***************
*** 93,99 ****
          match = prog.match(line)
          if match:
              name, val = match.group(1, 2)
!             val = string.atoi(val)
              tokens[val] = name          # reverse so we can sort them...
      keys = tokens.keys()
      keys.sort()
--- 92,98 ----
          match = prog.match(line)
          if match:
              name, val = match.group(1, 2)
!             val = int(val)
              tokens[val] = name          # reverse so we can sort them...
      keys = tokens.keys()
      keys.sort()
***************
*** 103,109 ****
      except IOError, err:
          sys.stderr.write("I/O error: %s\n" % str(err))
          sys.exit(2)
!     format = string.splitfields(fp.read(), "\n")
      fp.close()
      try:
          start = format.index("#--start constants--") + 1
--- 102,108 ----
      except IOError, err:
          sys.stderr.write("I/O error: %s\n" % str(err))
          sys.exit(2)
!     format = fp.read().split("\n")
      fp.close()
      try:
          start = format.index("#--start constants--") + 1
***************
*** 120,126 ****
      except IOError, err:
          sys.stderr.write("I/O error: %s\n" % str(err))
          sys.exit(4)
!     fp.write(string.joinfields(format, "\n"))
      fp.close()
  
  
--- 119,125 ----
      except IOError, err:
          sys.stderr.write("I/O error: %s\n" % str(err))
          sys.exit(4)
!     fp.write("\n".join(format))
      fp.close()
  
  
Index: Lib/tokenize.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/tokenize.py,v
retrieving revision 1.12
diff -c -r1.12 tokenize.py
*** tokenize.py	1998/04/03 16:05:38	1.12
--- tokenize.py	2000/04/11 11:56:52
***************
*** 26,32 ****
  #     Imagnumber is new.  Expfloat is corrected to reject '0e4'.
  # Note: to quote a backslash in a regex, it must be doubled in a r'aw' string.
  
! def group(*choices): return '(' + string.join(choices, '|') + ')'
  def any(*choices): return apply(group, choices) + '*'
  def maybe(*choices): return apply(group, choices) + '?'
  
--- 26,32 ----
  #     Imagnumber is new.  Expfloat is corrected to reject '0e4'.
  # Note: to quote a backslash in a regex, it must be doubled in a r'aw' string.
  
! def group(*choices): return '(' + '|'.join(choices) + ')'
  def any(*choices): return apply(group, choices) + '*'
  def maybe(*choices): return apply(group, choices) + '?'
  
Index: Lib/traceback.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/traceback.py,v
retrieving revision 1.15
diff -c -r1.15 traceback.py
*** traceback.py	2000/04/10 16:29:29	1.15
--- traceback.py	2000/04/11 11:56:53
***************
*** 18,24 ****
  		_print(file,
  		       '  File "%s", line %d, in %s' % (filename,lineno,name))
  		if line:
! 			_print(file, '    %s' % string.strip(line))
  
  def format_list(extracted_list):
  	"""Given a list of tuples as returned by extract_tb() or
--- 18,24 ----
  		_print(file,
  		       '  File "%s", line %d, in %s' % (filename,lineno,name))
  		if line:
! 			_print(file, '    %s' % line.strip())
  
  def format_list(extracted_list):
  	"""Given a list of tuples as returned by extract_tb() or
***************
*** 31,37 ****
  	for filename, lineno, name, line in extracted_list:
  		item = '  File "%s", line %d, in %s\n' % (filename,lineno,name)
  		if line:
! 			item = item + '    %s\n' % string.strip(line)
  		list.append(item)
  	return list
  	
--- 31,37 ----
  	for filename, lineno, name, line in extracted_list:
  		item = '  File "%s", line %d, in %s\n' % (filename,lineno,name)
  		if line:
! 			item = item + '    %s\n' % line.strip()
  		list.append(item)
  	return list
  	
***************
*** 56,62 ****
  		_print(file,
  		       '  File "%s", line %d, in %s' % (filename,lineno,name))
  		line = linecache.getline(filename, lineno)
! 		if line: _print(file, '    ' + string.strip(line))
  		tb = tb.tb_next
  		n = n+1
  
--- 56,62 ----
  		_print(file,
  		       '  File "%s", line %d, in %s' % (filename,lineno,name))
  		line = linecache.getline(filename, lineno)
! 		if line: _print(file, '    ' + line.strip())
  		tb = tb.tb_next
  		n = n+1
  
***************
*** 85,91 ****
  		filename = co.co_filename
  		name = co.co_name
  		line = linecache.getline(filename, lineno)
! 		if line: line = string.strip(line)
  		else: line = None
  		list.append((filename, lineno, name, line))
  		tb = tb.tb_next
--- 85,91 ----
  		filename = co.co_filename
  		name = co.co_name
  		line = linecache.getline(filename, lineno)
! 		if line: line = line.strip()
  		else: line = None
  		list.append((filename, lineno, name, line))
  		tb = tb.tb_next
***************
*** 157,163 ****
  				while i < len(line) and \
  				      line[i] in string.whitespace:
  					i = i+1
! 				list.append('    %s\n' % string.strip(line))
  				s = '    '
  				for c in line[i:offset-1]:
  					if c in string.whitespace:
--- 157,163 ----
  				while i < len(line) and \
  				      line[i] in string.whitespace:
  					i = i+1
! 				list.append('    %s\n' % line.strip())
  				s = '    '
  				for c in line[i:offset-1]:
  					if c in string.whitespace:
***************
*** 236,242 ****
  		filename = co.co_filename
  		name = co.co_name
  		line = linecache.getline(filename, lineno)
! 		if line: line = string.strip(line)
  		else: line = None
  		list.append((filename, lineno, name, line))
  		f = f.f_back
--- 236,242 ----
  		filename = co.co_filename
  		name = co.co_name
  		line = linecache.getline(filename, lineno)
! 		if line: line = line.strip()
  		else: line = None
  		list.append((filename, lineno, name, line))
  		f = f.f_back
Index: Lib/urllib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/urllib.py,v
retrieving revision 1.91
diff -c -r1.91 urllib.py
*** urllib.py	2000/02/04 15:28:41	1.91
--- urllib.py	2000/04/11 11:56:53
***************
*** 150,156 ****
          self.type = type
          if '-' in name:
              # replace - with _
!             name = string.join(string.split(name, '-'), '_')
          if not hasattr(self, name):
              if data is None:
                  return self.open_unknown(fullurl)
--- 150,156 ----
          self.type = type
          if '-' in name:
              # replace - with _
!             name = name.replace ('-', '_')
          if not hasattr(self, name):
              if data is None:
                  return self.open_unknown(fullurl)
***************
*** 239,245 ****
              urltype, rest = splittype(selector)
              url = rest
              user_passwd = None
!             if string.lower(urltype) != 'http':
                  realhost = None
              else:
                  realhost, rest = splithost(rest)
--- 239,245 ----
              urltype, rest = splittype(selector)
              url = rest
              user_passwd = None
!             if urltype.lower() != 'http':
                  realhost = None
              else:
                  realhost, rest = splithost(rest)
***************
*** 251,257 ****
          if not host: raise IOError, ('http error', 'no host given')
          if user_passwd:
              import base64
!             auth = string.strip(base64.encodestring(user_passwd))
          else:
              auth = None
          h = httplib.HTTP(host)
--- 251,257 ----
          if not host: raise IOError, ('http error', 'no host given')
          if user_passwd:
              import base64
!             auth = base64.encodestring(user_passwd).strip()
          else:
              auth = None
          h = httplib.HTTP(host)
***************
*** 308,314 ****
              else:
                  host, selector = url
                  urltype, rest = splittype(selector)
!                 if string.lower(urltype) == 'https':
                      realhost, rest = splithost(rest)
                      user_passwd, realhost = splituser(realhost)
                      if user_passwd:
--- 308,314 ----
              else:
                  host, selector = url
                  urltype, rest = splittype(selector)
!                 if urltype.lower() == 'https':
                      realhost, rest = splithost(rest)
                      user_passwd, realhost = splituser(realhost)
                      if user_passwd:
***************
*** 317,323 ****
              if not host: raise IOError, ('https error', 'no host given')
              if user_passwd:
                  import base64
!                 auth = string.strip(base64.encodestring(user_passwd))
              else:
                  auth = None
              h = httplib.HTTPS(host, 0,
--- 317,323 ----
              if not host: raise IOError, ('https error', 'no host given')
              if user_passwd:
                  import base64
!                 auth = base64.encodestring(user_passwd).strip()
              else:
                  auth = None
              h = httplib.HTTPS(host, 0,
***************
*** 399,409 ****
              port = int(port)
          path, attrs = splitattr(path)
          path = unquote(path)
!         dirs = string.splitfields(path, '/')
          dirs, file = dirs[:-1], dirs[-1]
          if dirs and not dirs[0]: dirs = dirs[1:]
          if dirs and not dirs[0]: dirs[0] = '/'
!         key = (user, host, port, string.joinfields(dirs, '/'))
          # XXX thread unsafe!
          if len(self.ftpcache) > MAXFTPCACHE:
              # Prune the cache, rather arbitrarily
--- 399,409 ----
              port = int(port)
          path, attrs = splitattr(path)
          path = unquote(path)
!         dirs = path.split('/')
          dirs, file = dirs[:-1], dirs[-1]
          if dirs and not dirs[0]: dirs = dirs[1:]
          if dirs and not dirs[0]: dirs[0] = '/'
!         key = (user, host, port, '/'.join(dirs))
          # XXX thread unsafe!
          if len(self.ftpcache) > MAXFTPCACHE:
              # Prune the cache, rather arbitrarily
***************
*** 420,428 ****
              else: type = 'I'
              for attr in attrs:
                  attr, value = splitvalue(attr)
!                 if string.lower(attr) == 'type' and \
                     value in ('a', 'A', 'i', 'I', 'd', 'D'):
!                     type = string.upper(value)
              (fp, retrlen) = self.ftpcache[key].retrfile(file, type)
              if retrlen is not None and retrlen >= 0:
                  import mimetools, StringIO
--- 420,428 ----
              else: type = 'I'
              for attr in attrs:
                  attr, value = splitvalue(attr)
!                 if attr.lower() == 'type' and \
                     value in ('a', 'A', 'i', 'I', 'd', 'D'):
!                     type = value.upper()
              (fp, retrlen) = self.ftpcache[key].retrfile(file, type)
              if retrlen is not None and retrlen >= 0:
                  import mimetools, StringIO
***************
*** 445,456 ****
          # parameter := attribute "=" value
          import StringIO, mimetools, time
          try:
!             [type, data] = string.split(url, ',', 1)
          except ValueError:
              raise IOError, ('data error', 'bad data URL')
          if not type:
              type = 'text/plain;charset=US-ASCII'
!         semi = string.rfind(type, ';')
          if semi >= 0 and '=' not in type[semi:]:
              encoding = type[semi+1:]
              type = type[:semi]
--- 445,456 ----
          # parameter := attribute "=" value
          import StringIO, mimetools, time
          try:
!             [type, data] = url.split(',', 1)
          except ValueError:
              raise IOError, ('data error', 'bad data URL')
          if not type:
              type = 'text/plain;charset=US-ASCII'
!         semi = type.rfind(';')
          if semi >= 0 and '=' not in type[semi:]:
              encoding = type[semi+1:]
              type = type[:semi]
***************
*** 468,474 ****
          msg.append('Content-length: %d' % len(data))
          msg.append('')
          msg.append(data)
!         msg = string.join(msg, '\n')
          f = StringIO.StringIO(msg)
          headers = mimetools.Message(f, 0)
          f.fileno = None     # needed for addinfourl
--- 468,474 ----
          msg.append('Content-length: %d' % len(data))
          msg.append('')
          msg.append(data)
!         msg = '\n'.join(msg)
          f = StringIO.StringIO(msg)
          headers = mimetools.Message(f, 0)
          f.fileno = None     # needed for addinfourl
***************
*** 518,524 ****
              match = re.match('[ \t]*([^ \t]+)[ \t]+realm="([^"]*)"', stuff)
              if match:
                  scheme, realm = match.groups()
!                 if string.lower(scheme) == 'basic':
                     name = 'retry_' + self.type + '_basic_auth'
                     if data is None:
                         return getattr(self,name)(url, realm)
--- 518,524 ----
              match = re.match('[ \t]*([^ \t]+)[ \t]+realm="([^"]*)"', stuff)
              if match:
                  scheme, realm = match.groups()
!                 if scheme.lower() == 'basic':
                     name = 'retry_' + self.type + '_basic_auth'
                     if data is None:
                         return getattr(self,name)(url, realm)
***************
*** 527,533 ****
  
      def retry_http_basic_auth(self, url, realm, data=None):
          host, selector = splithost(url)
!         i = string.find(host, '@') + 1
          host = host[i:]
          user, passwd = self.get_user_passwd(host, realm, i)
          if not (user or passwd): return None
--- 527,533 ----
  
      def retry_http_basic_auth(self, url, realm, data=None):
          host, selector = splithost(url)
!         i = host.find('@') + 1
          host = host[i:]
          user, passwd = self.get_user_passwd(host, realm, i)
          if not (user or passwd): return None
***************
*** 540,546 ****
     
      def retry_https_basic_auth(self, url, realm, data=None):
              host, selector = splithost(url)
!             i = string.find(host, '@') + 1
              host = host[i:]
              user, passwd = self.get_user_passwd(host, realm, i)
              if not (user or passwd): return None
--- 540,546 ----
     
      def retry_https_basic_auth(self, url, realm, data=None):
              host, selector = splithost(url)
!             i = host.find('@') + 1
              host = host[i:]
              user, passwd = self.get_user_passwd(host, realm, i)
              if not (user or passwd): return None
***************
*** 549,555 ****
              return self.open_https(newurl)
  
      def get_user_passwd(self, host, realm, clear_cache = 0):
!         key = realm + '@' + string.lower(host)
          if self.auth_cache.has_key(key):
              if clear_cache:
                  del self.auth_cache[key]
--- 549,555 ----
              return self.open_https(newurl)
  
      def get_user_passwd(self, host, realm, clear_cache = 0):
!         key = realm + '@' + host.lower()
          if self.auth_cache.has_key(key):
              if clear_cache:
                  del self.auth_cache[key]
***************
*** 774,780 ****
              i = len(basepath)
          else:
              # else replace last component
!             i = string.rfind(basepath, '/')
          if i < 0:
              # basepath not absolute
              if host:
--- 774,780 ----
              i = len(basepath)
          else:
              # else replace last component
!             i = basepath.rfind('/')
          if i < 0:
              # basepath not absolute
              if host:
***************
*** 789,795 ****
          # Interpret ../ (important because of symlinks)
          while basepath and path[:3] == '../':
              path = path[3:]
!             i = string.rfind(basepath[:-1], '/')
              if i > 0:
                  basepath = basepath[:i+1]
              elif i == 0:
--- 789,795 ----
          # Interpret ../ (important because of symlinks)
          while basepath and path[:3] == '../':
              path = path[3:]
!             i = basepath[:-1].rfind('/')
              if i > 0:
                  basepath = basepath[:i+1]
              elif i == 0:
***************
*** 823,832 ****
  
  def unwrap(url):
      """unwrap('<URL:type://host/path>') --> 'type://host/path'."""
!     url = string.strip(url)
      if url[:1] == '<' and url[-1:] == '>':
!         url = string.strip(url[1:-1])
!     if url[:4] == 'URL:': url = string.strip(url[4:])
      return url
  
  _typeprog = None
--- 823,832 ----
  
  def unwrap(url):
      """unwrap('<URL:type://host/path>') --> 'type://host/path'."""
!     url = url.strip()
      if url[:1] == '<' and url[-1:] == '>':
!         url = url[1:-1].strip()
!     if url[:4] == 'URL:': url = url[4:].strip()
      return url
  
  _typeprog = None
***************
*** 907,915 ****
      if match:
          host, port = match.group(1, 2)
          try:
!             if not port: raise string.atoi_error, "no digits"
!             nport = string.atoi(port)
!         except string.atoi_error:
              nport = None
          return host, nport
      return host, defport
--- 907,915 ----
      if match:
          host, port = match.group(1, 2)
          try:
!             if not port: raise ValueError, "no digits"
!             nport = int(port)
!         except ValueError:
              nport = None
          return host, nport
      return host, defport
***************
*** 941,947 ****
  def splitattr(url):
      """splitattr('/path;attr1=value1;attr2=value2;...') ->
          '/path', ['attr1=value1', 'attr2=value2', ...]."""
!     words = string.splitfields(url, ';')
      return words[0], words[1:]
  
  _valueprog = None
--- 941,947 ----
  def splitattr(url):
      """splitattr('/path;attr1=value1;attr2=value2;...') ->
          '/path', ['attr1=value1', 'attr2=value2', ...]."""
!     words = url.split(';')
      return words[0], words[1:]
  
  _valueprog = None
***************
*** 965,972 ****
  def unquote(s):
      """unquote('abc%20def') -> 'abc def'."""
      mychr = chr
!     myatoi = string.atoi
!     list = string.split(s, '%')
      res = [list[0]]
      myappend = res.append
      del list[0]
--- 965,972 ----
  def unquote(s):
      """unquote('abc%20def') -> 'abc def'."""
      mychr = chr
!     myatoi = int
!     list = s.split('%')
      res = [list[0]]
      myappend = res.append
      del list[0]
***************
*** 979,990 ****
                  myappend('%' + item)
          else:
              myappend('%' + item)
!     return string.join(res, "")
  
  def unquote_plus(s):
      if '+' in s:
          # replace '+' with ' '
!         s = string.join(string.split(s, '+'), ' ')
      return unquote(s)
  
  always_safe = string.letters + string.digits + '_,.-'
--- 979,990 ----
                  myappend('%' + item)
          else:
              myappend('%' + item)
!     return "".join(res)
  
  def unquote_plus(s):
      if '+' in s:
          # replace '+' with ' '
! 	s = s.replace('+', ' ')
      return unquote(s)
  
  always_safe = string.letters + string.digits + '_,.-'
***************
*** 997,1012 ****
          c = res[i]
          if c not in safe:
              res[i] = '%%%02x' % ord(c)
!     return string.joinfields(res, '')
  
  def quote_plus(s, safe = '/'):
      # XXX Can speed this up an order of magnitude
      if ' ' in s:
          # replace ' ' with '+'
!         l = string.split(s, ' ')
          for i in range(len(l)):
              l[i] = quote(l[i], safe)
!         return string.join(l, '+')
      else:
          return quote(s, safe)
  
--- 997,1012 ----
          c = res[i]
          if c not in safe:
              res[i] = '%%%02x' % ord(c)
!     return ''.join(res)
  
  def quote_plus(s, safe = '/'):
      # XXX Can speed this up an order of magnitude
      if ' ' in s:
          # replace ' ' with '+'
!         l = s.split(' ')
          for i in range(len(l)):
              l[i] = quote(l[i], safe)
!         return '+'.join(l)
      else:
          return quote(s, safe)
  
***************
*** 1017,1023 ****
          k = quote_plus(str(k))
          v = quote_plus(str(v))
          l.append(k + '=' + v)
!     return string.join(l, '&')
  
  
  # Proxy handling
--- 1017,1023 ----
          k = quote_plus(str(k))
          v = quote_plus(str(v))
          l.append(k + '=' + v)
!     return '&'.join(l)
  
  
  # Proxy handling
***************
*** 1064,1070 ****
          """
          proxies = {}
          for name, value in os.environ.items():
!             name = string.lower(name)
              if value and name[-6:] == '_proxy':
                  proxies[name[:-6]] = value
          return proxies
--- 1064,1070 ----
          """
          proxies = {}
          for name, value in os.environ.items():
!             name = name.lower()
              if value and name[-6:] == '_proxy':
                  proxies[name[:-6]] = value
          return proxies
***************
*** 1119,1125 ****
              del fp
              if '\r' in data:
                  table = string.maketrans("", "")
!                 data = string.translate(data, table, "\r")
              print data
              fn, h = None, None
          print '-'*40
--- 1119,1125 ----
              del fp
              if '\r' in data:
                  table = string.maketrans("", "")
!                 data = data.translate(table, "\r")
              print data
              fn, h = None, None
          print '-'*40
Index: Lib/urllib2.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/urllib2.py,v
retrieving revision 1.3
diff -c -r1.3 urllib2.py
*** urllib2.py	2000/02/10 17:17:14	1.3
--- urllib2.py	2000/04/11 11:56:54
***************
*** 85,91 ****
  # gopher can return a socket.error
  # check digest against correct (i.e. non-apache) implementation
  
- import string
  import socket
  import UserDict
  import httplib
--- 85,90 ----
***************
*** 262,274 ****
                      self.handle_open[protocol] = [handler]
                  added = 1
                  continue
!             i = string.find(meth, '_')
!             j = string.find(meth[i+1:], '_') + i + 1
              if j != -1 and meth[i+1:j] == 'error':
                  proto = meth[:i]
                  kind = meth[j+1:]
                  try:
!                     kind = string.atoi(kind)
                  except ValueError:
                      pass
                  dict = self.handle_error.get(proto, {})
--- 261,273 ----
                      self.handle_open[protocol] = [handler]
                  added = 1
                  continue
!             i = meth.find('_')
!             j = meth[i+1:].find('_') + i + 1
              if j != -1 and meth[i+1:j] == 'error':
                  proto = meth[:i]
                  kind = meth[j+1:]
                  try:
!                     kind = int(kind)
                  except ValueError:
                      pass
                  dict = self.handle_error.get(proto, {})
***************
*** 593,599 ****
              mo = HTTPBasicAuthHandler.rx.match(authreq)
              if mo:
                  scheme, realm = mo.groups()
!                 if string.lower(scheme) == 'basic':
                      return self.retry_http_basic_auth(req, realm)
  
      def retry_http_basic_auth(self, req, realm):
--- 592,598 ----
              mo = HTTPBasicAuthHandler.rx.match(authreq)
              if mo:
                  scheme, realm = mo.groups()
!                 if scheme.lower() == 'basic':
                      return self.retry_http_basic_auth(req, realm)
  
      def retry_http_basic_auth(self, req, realm):
***************
*** 607,613 ****
          user,pw = self.passwd.find_user_password(realm, host)
          if pw:
              raw = "%s:%s" % (user, pw)
!             auth = string.strip(base64.encodestring(raw))
              req.add_header('Authorization', 'Basic %s' % auth)
              resp = self.parent.open(req)
              self.__current_realm = None
--- 606,612 ----
          user,pw = self.passwd.find_user_password(realm, host)
          if pw:
              raw = "%s:%s" % (user, pw)
!             auth = base64.encodestring(raw).strip()
              req.add_header('Authorization', 'Basic %s' % auth)
              resp = self.parent.open(req)
              self.__current_realm = None
***************
*** 632,643 ****
          # XXX could be mult. headers
          authreq = headers.get('www-authenticate', None)
          if authreq:
!             kind = string.split(authreq)[0]
              if kind == 'Digest':
                  return self.retry_http_digest_auth(req, authreq)
  
      def retry_http_digest_auth(self, req, auth):
!         token, challenge = string.split(auth, ' ', 1)
          chal = parse_keqv_list(parse_http_list(challenge))
          auth = self.get_authorization(req, chal)
          if auth:
--- 631,642 ----
          # XXX could be mult. headers
          authreq = headers.get('www-authenticate', None)
          if authreq:
!             kind = authreq.split()[0]
              if kind == 'Digest':
                  return self.retry_http_digest_auth(req, authreq)
  
      def retry_http_digest_auth(self, req, auth):
!         token, challenge = auth.split(' ', 1)
          chal = parse_keqv_list(parse_http_list(challenge))
          auth = self.get_authorization(req, chal)
          if auth:
***************
*** 717,723 ****
          hexrep.append(hex(n)[-1])
          n = ord(c) & 0xf
          hexrep.append(hex(n)[-1])
!     return string.join(hexrep, '')
          
          
  class HTTPHandler(BaseHandler):
--- 716,722 ----
          hexrep.append(hex(n)[-1])
          n = ord(c) & 0xf
          hexrep.append(hex(n)[-1])
!     return ''.join(hexrep)
          
          
  class HTTPHandler(BaseHandler):
***************
*** 768,774 ****
      """Parse list of key=value strings where keys are not duplicated."""
      parsed = {}
      for elt in l:
!         k, v = string.split(elt, '=', 1)
          if v[0] == '"' and v[-1] == '"':
              v = v[1:-1]
          parsed[k] = v
--- 767,773 ----
      """Parse list of key=value strings where keys are not duplicated."""
      parsed = {}
      for elt in l:
!         k, v = elt.split('=', 1)
          if v[0] == '"' and v[-1] == '"':
              v = v[1:-1]
          parsed[k] = v
***************
*** 790,797 ****
      start = 0
      while i < end:
          cur = s[i:]
!         c = string.find(cur, ',')
!         q = string.find(cur, '"')
          if c == -1:
              list.append(s[start:])
              break
--- 789,796 ----
      start = 0
      while i < end:
          cur = s[i:]
!         c = cur.find(',')
!         q = cur.find('"')
          if c == -1:
              list.append(s[start:])
              break
***************
*** 818,824 ****
              else:
                  inquote = 1
                  i = i + q + 1
!     return map(string.strip, list)
  
  class FileHandler(BaseHandler):
      # Use local file or FTP depending on form of URL
--- 817,823 ----
              else:
                  inquote = 1
                  i = i + q + 1
!     return map(lambda s: s.strip(), list)
  
  class FileHandler(BaseHandler):
      # Use local file or FTP depending on form of URL
***************
*** 865,871 ****
              port = ftplib.FTP_PORT
          path, attrs = splitattr(req.get_selector())
          path = unquote(path)
!         dirs = string.splitfields(path, '/')
          dirs, file = dirs[:-1], dirs[-1]
          if dirs and not dirs[0]:
              dirs = dirs[1:]
--- 864,870 ----
              port = ftplib.FTP_PORT
          path, attrs = splitattr(req.get_selector())
          path = unquote(path)
!         dirs = path.split('/')
          dirs, file = dirs[:-1], dirs[-1]
          if dirs and not dirs[0]:
              dirs = dirs[1:]
***************
*** 875,883 ****
              type = file and 'I' or 'D'
              for attr in attrs:
                  attr, value = splitattr(attr)
!                 if string.lower(attr) == 'type' and \
                     value in ('a', 'A', 'i', 'I', 'd', 'D'):
!                     type = string.upper(value)
              fp, retrlen = fw.retrfile(file, type)
              if retrlen is not None and retrlen >= 0:
                  sf = StringIO('Content-Length: %d\n' % retrlen)
--- 874,882 ----
              type = file and 'I' or 'D'
              for attr in attrs:
                  attr, value = splitattr(attr)
!                 if attr.lower() == 'type' and \
                     value in ('a', 'A', 'i', 'I', 'd', 'D'):
!                     type = value.upper()
              fp, retrlen = fw.retrfile(file, type)
              if retrlen is not None and retrlen >= 0:
                  sf = StringIO('Content-Length: %d\n' % retrlen)
Index: Lib/urlparse.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/urlparse.py,v
retrieving revision 1.23
diff -c -r1.23 urlparse.py
*** urlparse.py	2000/04/10 17:02:46	1.23
--- urlparse.py	2000/04/11 11:56:54
***************
*** 6,12 ****
  
  # Standard/builtin Python modules
  import string
- from string import join, split, rfind
  
  # A classification of schemes ('' means apply by default)
  uses_relative = ['ftp', 'http', 'gopher', 'nntp', 'wais', 'file',
--- 6,11 ----
***************
*** 54,82 ****
  		return cached
  	if len(_parse_cache) >= MAX_CACHE_SIZE:	# avoid runaway growth
  		clear_cache()
- 	find = string.find
  	netloc = path = params = query = fragment = ''
! 	i = find(url, ':')
  	if i > 0:
  		if url[:i] == 'http': # optimize the common case
! 			scheme = string.lower(url[:i])
  			url = url[i+1:]
  			if url[:2] == '//':
! 				i = find(url, '/', 2)
  				if i < 0:
  					i = len(url)
  				netloc = url[2:i]
  				url = url[i:]
  			if allow_fragments:
! 				i = string.rfind(url, '#')
  				if i >= 0:
  					fragment = url[i+1:]
  					url = url[:i]
! 			i = find(url, '?')
  			if i >= 0:
  				query = url[i+1:]
  				url = url[:i]
! 			i = find(url, ';')
  			if i >= 0:
  				params = url[i+1:]
  				url = url[:i]
--- 53,80 ----
  		return cached
  	if len(_parse_cache) >= MAX_CACHE_SIZE:	# avoid runaway growth
  		clear_cache()
  	netloc = path = params = query = fragment = ''
! 	i = url.find(':')
  	if i > 0:
  		if url[:i] == 'http': # optimize the common case
! 			scheme = url[:i].lower()
  			url = url[i+1:]
  			if url[:2] == '//':
! 				i = url.find('/', 2)
  				if i < 0:
  					i = len(url)
  				netloc = url[2:i]
  				url = url[i:]
  			if allow_fragments:
! 				i = url.rfind('#')
  				if i >= 0:
  					fragment = url[i+1:]
  					url = url[:i]
! 			i = url.find('?')
  			if i >= 0:
  				query = url[i+1:]
  				url = url[:i]
! 			i = url.find(';')
  			if i >= 0:
  				params = url[i+1:]
  				url = url[:i]
***************
*** 87,109 ****
  			if c not in scheme_chars:
  				break
  		else:
! 			scheme, url = string.lower(url[:i]), url[i+1:]
  	if scheme in uses_netloc:
  		if url[:2] == '//':
! 			i = find(url, '/', 2)
  			if i < 0:
  				i = len(url)
  			netloc, url = url[2:i], url[i:]
  	if allow_fragments and scheme in uses_fragment:
! 		i = string.rfind(url, '#')
  		if i >= 0:
  			url, fragment = url[:i], url[i+1:]
  	if scheme in uses_query:
! 		i = find(url, '?')
  		if i >= 0:
  			url, query = url[:i], url[i+1:]
  	if scheme in uses_params:
! 		i = find(url, ';')
  		if i >= 0:
  			url, params = url[:i], url[i+1:]
  	tuple = scheme, netloc, url, params, query, fragment
--- 85,107 ----
  			if c not in scheme_chars:
  				break
  		else:
! 			scheme, url = url[:i].lower(), url[i+1:]
  	if scheme in uses_netloc:
  		if url[:2] == '//':
! 			i = url.find('/', 2)
  			if i < 0:
  				i = len(url)
  			netloc, url = url[2:i], url[i:]
  	if allow_fragments and scheme in uses_fragment:
! 		i = url.rfind('#')
  		if i >= 0:
  			url, fragment = url[:i], url[i+1:]
  	if scheme in uses_query:
! 		i = url.find('?')
  		if i >= 0:
  			url, query = url[:i], url[i+1:]
  	if scheme in uses_params:
! 		i = url.find(';')
  		if i >= 0:
  			url, params = url[:i], url[i+1:]
  	tuple = scheme, netloc, url, params, query, fragment
***************
*** 151,157 ****
  	if not path:
  		return urlunparse((scheme, netloc, bpath,
  				   params, query or bquery, fragment))
! 	segments = split(bpath, '/')[:-1] + split(path, '/')
  	# XXX The stuff below is bogus in various ways...
  	if segments[-1] == '.':
  		segments[-1] = ''
--- 149,155 ----
  	if not path:
  		return urlunparse((scheme, netloc, bpath,
  				   params, query or bquery, fragment))
! 	segments = bpath.split('/')[:-1] + path.split('/')
  	# XXX The stuff below is bogus in various ways...
  	if segments[-1] == '.':
  		segments[-1] = ''
***************
*** 171,177 ****
  		segments[-1] = ''
  	elif len(segments) >= 2 and segments[-1] == '..':
  		segments[-2:] = ['']
! 	return urlunparse((scheme, netloc, join(segments, '/'),
  			   params, query, fragment))
  
  def urldefrag(url):
--- 169,175 ----
  		segments[-1] = ''
  	elif len(segments) >= 2 and segments[-1] == '..':
  		segments[-2:] = ['']
! 	return urlunparse((scheme, netloc, '/'.join(segments),
  			   params, query, fragment))
  
  def urldefrag(url):
***************
*** 236,242 ****
  	while 1:
  		line = fp.readline()
  		if not line: break
! 		words = string.split(line)
  		if not words:
  			continue
  		url = words[0]
--- 234,240 ----
  	while 1:
  		line = fp.readline()
  		if not line: break
! 		words = line.split()
  		if not words:
  			continue
  		url = words[0]
Index: Lib/uu.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/uu.py,v
retrieving revision 1.11
diff -c -r1.11 uu.py
*** uu.py	2000/02/04 15:28:42	1.11
--- uu.py	2000/04/11 11:56:54
***************
*** 32,38 ****
  
  import binascii
  import os
- import string
  import sys
  
  Error = 'uu.Error'
--- 32,37 ----
***************
*** 96,112 ****
              raise Error, 'No valid begin line found in input file'
          if hdr[:5] != 'begin':
              continue
!         hdrfields = string.split(hdr)
          if len(hdrfields) == 3 and hdrfields[0] == 'begin':
              try:
!                 string.atoi(hdrfields[1], 8)
                  break
              except ValueError:
                  pass
      if out_file == None:
          out_file = hdrfields[2]
      if mode == None:
!         mode = string.atoi(hdrfields[1], 8)
      #
      # Open the output file
      #
--- 95,111 ----
              raise Error, 'No valid begin line found in input file'
          if hdr[:5] != 'begin':
              continue
!         hdrfields = hdr.split()
          if len(hdrfields) == 3 and hdrfields[0] == 'begin':
              try:
!                 int(hdrfields[1], 8)
                  break
              except ValueError:
                  pass
      if out_file == None:
          out_file = hdrfields[2]
      if mode == None:
!         mode = int(hdrfields[1], 8)
      #
      # Open the output file
      #
Index: Lib/xmllib.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/xmllib.py,v
retrieving revision 1.17
diff -c -r1.17 xmllib.py
*** xmllib.py	2000/02/04 15:39:30	1.17
--- xmllib.py	2000/04/11 11:56:55
***************
*** 190,198 ****
              post = data[i:]
              if str[0] == '#':
                  if str[1] == 'x':
!                     str = chr(string.atoi(str[2:], 16))
                  else:
!                     str = chr(string.atoi(str[1:]))
                  data = pre + str + post
                  i = res.start(0)+len(str)
              elif all:
--- 190,198 ----
              post = data[i:]
              if str[0] == '#':
                  if str[1] == 'x':
!                     str = chr(int(str[2:], 16))
                  else:
!                     str = chr(int(str[1:]))
                  data = pre + str + post
                  i = res.start(0)+len(str)
              elif all:
***************
*** 221,227 ****
              if self.nomoretags:
                  data = rawdata[i:n]
                  self.handle_data(data)
!                 self.lineno = self.lineno + string.count(data, '\n')
                  i = n
                  break
              res = interesting.search(rawdata, i)
--- 221,227 ----
              if self.nomoretags:
                  data = rawdata[i:n]
                  self.handle_data(data)
!                 self.lineno = self.lineno + data.count('\n')
                  i = n
                  break
              res = interesting.search(rawdata, i)
***************
*** 239,245 ****
                  if not self.__accept_utf8 and illegal.search(data):
                      self.syntax_error('illegal character in content')
                  self.handle_data(data)
!                 self.lineno = self.lineno + string.count(data, '\n')
              i = j
              if i == n: break
              if rawdata[i] == '<':
--- 239,245 ----
                  if not self.__accept_utf8 and illegal.search(data):
                      self.syntax_error('illegal character in content')
                  self.handle_data(data)
!                 self.lineno = self.lineno + data.count('\n')
              i = j
              if i == n: break
              if rawdata[i] == '<':
***************
*** 247,283 ****
                      if self.literal:
                          data = rawdata[i]
                          self.handle_data(data)
!                         self.lineno = self.lineno + string.count(data, '\n')
                          i = i+1
                          continue
                      k = self.parse_starttag(i)
                      if k < 0: break
                      self.__seen_starttag = 1
!                     self.lineno = self.lineno + string.count(rawdata[i:k], '\n')
                      i = k
                      continue
                  if endtagopen.match(rawdata, i):
                      k = self.parse_endtag(i)
                      if k < 0: break
!                     self.lineno = self.lineno + string.count(rawdata[i:k], '\n')
                      i =  k
                      continue
                  if commentopen.match(rawdata, i):
                      if self.literal:
                          data = rawdata[i]
                          self.handle_data(data)
!                         self.lineno = self.lineno + string.count(data, '\n')
                          i = i+1
                          continue
                      k = self.parse_comment(i)
                      if k < 0: break
!                     self.lineno = self.lineno + string.count(rawdata[i:k], '\n')
                      i = k
                      continue
                  if cdataopen.match(rawdata, i):
                      k = self.parse_cdata(i)
                      if k < 0: break
!                     self.lineno = self.lineno + string.count(rawdata[i:i], '\n')
                      i = k
                      continue
                  res = xmldecl.match(rawdata, i)
--- 247,283 ----
                      if self.literal:
                          data = rawdata[i]
                          self.handle_data(data)
!                         self.lineno = self.lineno + data.count('\n')
                          i = i+1
                          continue
                      k = self.parse_starttag(i)
                      if k < 0: break
                      self.__seen_starttag = 1
!                     self.lineno = self.lineno + rawdata[i:k].count('\n')
                      i = k
                      continue
                  if endtagopen.match(rawdata, i):
                      k = self.parse_endtag(i)
                      if k < 0: break
!                     self.lineno = self.lineno + rawdata[i:k].count('\n')
                      i =  k
                      continue
                  if commentopen.match(rawdata, i):
                      if self.literal:
                          data = rawdata[i]
                          self.handle_data(data)
!                         self.lineno = self.lineno + data.count('\n')
                          i = i+1
                          continue
                      k = self.parse_comment(i)
                      if k < 0: break
!                     self.lineno = self.lineno + rawdata[i:k].count('\n')
                      i = k
                      continue
                  if cdataopen.match(rawdata, i):
                      k = self.parse_cdata(i)
                      if k < 0: break
!                     self.lineno = self.lineno + rawdata[i:i].count('\n')
                      i = k
                      continue
                  res = xmldecl.match(rawdata, i)
***************
*** 298,304 ****
                  if res:
                      k = self.parse_proc(i)
                      if k < 0: break
!                     self.lineno = self.lineno + string.count(rawdata[i:k], '\n')
                      i = k
                      continue
                  res = doctype.match(rawdata, i)
--- 298,304 ----
                  if res:
                      k = self.parse_proc(i)
                      if k < 0: break
!                     self.lineno = self.lineno + rawdata[i:k].count('\n')
                      i = k
                      continue
                  res = doctype.match(rawdata, i)
***************
*** 306,312 ****
                      if self.literal:
                          data = rawdata[i]
                          self.handle_data(data)
!                         self.lineno = self.lineno + string.count(data, '\n')
                          i = i+1
                          continue
                      if self.__seen_doctype:
--- 306,312 ----
                      if self.literal:
                          data = rawdata[i]
                          self.handle_data(data)
!                         self.lineno = self.lineno + data.count('\n')
                          i = i+1
                          continue
                      if self.__seen_doctype:
***************
*** 317,324 ****
                      if k < 0: break
                      self.__seen_doctype = res.group('name')
                      if self.__map_case:
!                         self.__seen_doctype = string.lower(self.__seen_doctype)
!                     self.lineno = self.lineno + string.count(rawdata[i:k], '\n')
                      i = k
                      continue
              elif rawdata[i] == '&':
--- 317,324 ----
                      if k < 0: break
                      self.__seen_doctype = res.group('name')
                      if self.__map_case:
!                         self.__seen_doctype = self.__seen_doctype.lower()
!                     self.lineno = self.lineno + rawdata[i:k].count('\n')
                      i = k
                      continue
              elif rawdata[i] == '&':
***************
*** 336,342 ****
                      if not self.stack:
                          self.syntax_error('data not in content')
                      self.handle_charref(res.group('char')[:-1])
!                     self.lineno = self.lineno + string.count(res.group(0), '\n')
                      continue
                  res = entityref.match(rawdata, i)
                  if res is not None:
--- 336,342 ----
                      if not self.stack:
                          self.syntax_error('data not in content')
                      self.handle_charref(res.group('char')[:-1])
!                     self.lineno = self.lineno + res.group(0).count('\n')
                      continue
                  res = entityref.match(rawdata, i)
                  if res is not None:
***************
*** 346,359 ****
                          i = i-1
                      name = res.group('name')
                      if self.__map_case:
!                         name = string.lower(name)
                      if self.entitydefs.has_key(name):
                          self.rawdata = rawdata = rawdata[:res.start(0)] + self.entitydefs[name] + rawdata[i:]
                          n = len(rawdata)
                          i = res.start(0)
                      else:
                          self.unknown_entityref(name)
!                     self.lineno = self.lineno + string.count(res.group(0), '\n')
                      continue
              elif rawdata[i] == ']':
                  if self.literal:
--- 346,359 ----
                          i = i-1
                      name = res.group('name')
                      if self.__map_case:
!                         name = name.lower()
                      if self.entitydefs.has_key(name):
                          self.rawdata = rawdata = rawdata[:res.start(0)] + self.entitydefs[name] + rawdata[i:]
                          n = len(rawdata)
                          i = res.start(0)
                      else:
                          self.unknown_entityref(name)
!                     self.lineno = self.lineno + res.group(0).count('\n')
                      continue
              elif rawdata[i] == ']':
                  if self.literal:
***************
*** 382,388 ****
              if not self.__accept_utf8 and illegal.search(data):
                  self.syntax_error('illegal character in content')
              self.handle_data(data)
!             self.lineno = self.lineno + string.count(data, '\n')
              self.rawdata = rawdata[i+1:]
              return self.goahead(end)
          self.rawdata = rawdata[i:]
--- 382,388 ----
              if not self.__accept_utf8 and illegal.search(data):
                  self.syntax_error('illegal character in content')
              self.handle_data(data)
!             self.lineno = self.lineno + data.count('\n')
              self.rawdata = rawdata[i+1:]
              return self.goahead(end)
          self.rawdata = rawdata[i:]
***************
*** 418,428 ****
          n = len(rawdata)
          name = res.group('name')
          if self.__map_case:
!             name = string.lower(name)
          pubid, syslit = res.group('pubid', 'syslit')
          if pubid is not None:
              pubid = pubid[1:-1]         # remove quotes
!             pubid = string.join(string.split(pubid)) # normalize
          if syslit is not None: syslit = syslit[1:-1] # remove quotes
          j = k = res.end(0)
          if k >= n:
--- 418,428 ----
          n = len(rawdata)
          name = res.group('name')
          if self.__map_case:
!             name = name.lower()
          pubid, syslit = res.group('pubid', 'syslit')
          if pubid is not None:
              pubid = pubid[1:-1]         # remove quotes
!             pubid = ' '.join(pubid.split()) # normalize
          if syslit is not None: syslit = syslit[1:-1] # remove quotes
          j = k = res.end(0)
          if k >= n:
***************
*** 492,498 ****
          k = res.end(0)
          name = res.group(0)
          if self.__map_case:
!             name = string.lower(name)
          if name == 'xml:namespace':
              self.syntax_error('old-fashioned namespace declaration')
              self.__use_namespaces = -1
--- 492,498 ----
          k = res.end(0)
          name = res.group(0)
          if self.__map_case:
!             name = name.lower()
          if name == 'xml:namespace':
              self.syntax_error('old-fashioned namespace declaration')
              self.__use_namespaces = -1
***************
*** 517,523 ****
                  self.syntax_error('xml:namespace prefix not unique')
              self.__namespaces[prefix] = attrdict['ns']
          else:
!             if string.lower(name) == 'xml':
                  self.syntax_error('illegal processing instruction target name')
              self.handle_proc(name, rawdata[k:j])
          return end.end(0)
--- 517,523 ----
                  self.syntax_error('xml:namespace prefix not unique')
              self.__namespaces[prefix] = attrdict['ns']
          else:
!             if name.lower() == 'xml':
                  self.syntax_error('illegal processing instruction target name')
              self.handle_proc(name, rawdata[k:j])
          return end.end(0)
***************
*** 533,539 ****
                  break
              attrname, attrvalue = res.group('name', 'value')
              if self.__map_case:
!                 attrname = string.lower(attrname)
              i = res.end(0)
              if attrvalue is None:
                  self.syntax_error("no value specified for attribute `%s'" % attrname)
--- 533,539 ----
                  break
              attrname, attrvalue = res.group('name', 'value')
              if self.__map_case:
!                 attrname = attrname.lower()
              i = res.end(0)
              if attrvalue is None:
                  self.syntax_error("no value specified for attribute `%s'" % attrname)
***************
*** 555,561 ****
                  self.syntax_error("`<' illegal in attribute value")
              if attrdict.has_key(attrname):
                  self.syntax_error("attribute `%s' specified twice" % attrname)
!             attrvalue = string.translate(attrvalue, attrtrans)
              attrdict[attrname] = self.translate_references(attrvalue)
          return attrdict, namespace, i
  
--- 555,561 ----
                  self.syntax_error("`<' illegal in attribute value")
              if attrdict.has_key(attrname):
                  self.syntax_error("attribute `%s' specified twice" % attrname)
!             attrvalue = attrvalue.translate(attrtrans)
              attrdict[attrname] = self.translate_references(attrvalue)
          return attrdict, namespace, i
  
***************
*** 572,578 ****
              return end.end(0)
          nstag = tagname = tag.group('tagname')
          if self.__map_case:
!             nstag = tagname = string.lower(nstag)
          if not self.__seen_starttag and self.__seen_doctype and \
             tagname != self.__seen_doctype:
              self.syntax_error('starttag does not match DOCTYPE')
--- 572,578 ----
              return end.end(0)
          nstag = tagname = tag.group('tagname')
          if self.__map_case:
!             nstag = tagname = nstag.lower()
          if not self.__seen_starttag and self.__seen_doctype and \
             tagname != self.__seen_doctype:
              self.syntax_error('starttag does not match DOCTYPE')
***************
*** 608,614 ****
                  if res is not None:
                      aprefix, key = res.group('prefix', 'local')
                      if self.__map_case:
!                         key = string.lower(key)
                      if aprefix is None:
                          aprefix = ''
                      ans = None
--- 608,614 ----
                  if res is not None:
                      aprefix, key = res.group('prefix', 'local')
                      if self.__map_case:
!                         key = key.lower()
                      if aprefix is None:
                          aprefix = ''
                      ans = None
***************
*** 657,663 ****
          else:
              tag = res.group(0)
              if self.__map_case:
!                 tag = string.lower(tag)
              if self.literal:
                  if not self.stack or tag != self.stack[-1][0]:
                      self.handle_data(rawdata[i])
--- 657,663 ----
          else:
              tag = res.group(0)
              if self.__map_case:
!                 tag = tag.lower()
              if self.literal:
                  if not self.stack or tag != self.stack[-1][0]:
                      self.handle_data(rawdata[i])
***************
*** 730,739 ****
      def handle_charref(self, name):
          try:
              if name[0] == 'x':
!                 n = string.atoi(name[1:], 16)
              else:
!                 n = string.atoi(name)
!         except string.atoi_error:
              self.unknown_charref(name)
              return
          if not 0 <= n <= 255:
--- 730,739 ----
      def handle_charref(self, name):
          try:
              if name[0] == 'x':
!                 n = int(name[1:], 16)
              else:
!                 n = int(name)
!         except ValueError:
              self.unknown_charref(name)
              return
          if not 0 <= n <= 255:


From skip@mojam.com (Skip Montanaro)  Tue Apr 11 13:21:59 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Tue, 11 Apr 2000 07:21:59 -0500 (CDT)
Subject: [Patches] Please review before applying
In-Reply-To: <200004111225.OAA31234@localhost.localdomain>
References: <200004111225.OAA31234@localhost.localdomain>
Message-ID: <14579.6375.601339.630742@beluga.mojam.com>

From the changes to Lib/BaseHTTPServer.py:
  
    Fred>       # The Python system version, truncated to its first component.
    Fred> !     sys_version = "Python/" + sys.version.split[0]


The change here should be to 

    sys.version.split()[0]

This suggests to me that the BaseHTTPRequestHandler class isn't tested by
'make test'.

-- 
Skip Montanaro | http://www.mojam.com/
skip@mojam.com | http://www.musi-cal.com/


From akuchlin@mems-exchange.org  Tue Apr 11 14:50:45 2000
From: akuchlin@mems-exchange.org (Andrew M. Kuchling)
Date: Tue, 11 Apr 2000 09:50:45 -0400 (EDT)
Subject: [Patches] add string precisions to PyErr_Format calls
In-Reply-To: <Pine.LNX.4.10.10004100529580.11297-100000@nebula.lyra.org>
References: <14577.47506.14149.790979@beluga.mojam.com>
 <Pine.LNX.4.10.10004100529580.11297-100000@nebula.lyra.org>
Message-ID: <14579.11701.733010.789688@amarok.cnri.reston.va.us>

Greg Stein writes:
>Wouldn't it be best to simply fix PyErr_Format so that we don't have to
>continue to worry about buffer overruns?

A while ago I suggested using nsprintf() in PyErr_Format, but that
means stealing the implementation from Apache for those platforms
where libc doesn't include nsprintf().  Haven't done it yet...

--amk


From gward@mems-exchange.org  Tue Apr 11 17:52:37 2000
From: gward@mems-exchange.org (Greg Ward)
Date: Tue, 11 Apr 2000 12:52:37 -0400
Subject: [Patches] fileinput.py argument handling (and suggestion)
In-Reply-To: <Pine.GSO.4.10.10004110842170.4465-100000@sundial>; from moshez@math.huji.ac.il on Tue, Apr 11, 2000 at 08:43:56AM +0200
References: <20000410151626.A1168@mems-exchange.org> <Pine.GSO.4.10.10004110842170.4465-100000@sundial>
Message-ID: <20000411125236.A1539@mems-exchange.org>

On 11 April 2000, Moshe Zadka said:
> I really like the text_file.py. Can't it just be in the standard library
> and have a__getitem__ method (or is there one already? I only looked at
> the doc you sent). This will at least solve the FAQ about reading a file
> line-by-line in a nice way.

No, there's no __getitem__.  I guess I could add one along the lines of
FileInput's; I didn't know about that class when I wrote my TextFile
class, and the idea of a strictly sequential __getitem__ didn't occur to
me independently.

The other problem, which I blithely assume is a problem with both the
standard FileInput and my TextFile, is performance.  Sometimes, you
just want to loop over all lines in a text file as quickly as possible,
doing something with each line.  Personally, I do those tasks using a
well-known and popular scripting language that was carefully optimized
for that very task.  Others in the Python community have expressed
distaste for that particular language, but the fact remains that P**l's
facilities for looping-over-all-lines-in-a-text-file-quickly are much
better than Python's.  You can write lovely high-level classes that
provide the same functionality as Perl's bag o' tricks, but the
performance will, presumably, stink.

[...considerable time passes...]

OK, I got curious: how bad *is* the performance of Python's regular file
input, the FileInput class, and my TextFile class?  My test data is the
entire Python 1.5.2 standard library, i.e. all 344 *.py files under
$prefix/lib/python1.5 (hmm, I guess that includes whatever I happen to
have in site-packages right now -- whatever).  The basic tasks are:

  1. read every line and discard it
  2. read all lines in one fell swoop and discard them
  3. read every line, stripping comments and whitespace and ignoring
     blanks... and then discard what's left (ie. we only exercise
     detecting blank lines)

I implemented each basic task in the most obvious and straightforward
way with Perl, and implemented task 1 and 3 in a couple of different
ways with Python.  Here are the results:

Script                 CPU time    Elapsed
readfile1.pl   (4 LoC)     0.27       0.44   "1 while <F>"
readfile1.py    6          3.72       4.37   "while 1: readline()"
readfile1a.py   4          7.34       8.24   FileInput (one per file)
readfile1b.py   3          7.20       8.07   FileInput (one for whole list)
readfile1c.py   5          8.50       9.53   TextFile (all options off)

readfile2.pl    4 LoC      0.45       0.63   "@lines = <F>"
readfile2a.pl   1          0.44       0.73   "@all_lines = <>"
readfile2.py    4          0.51       1.11   "lines = readlines()"

readfile3.pl    8 LoC      1.01       1.28   chomp, regex tricks to trim
readfile3.py    13         7.30       8.10   string.find, string.rstrip
readfile3a.py   6         11.09      12.24   TextFile (default options)

(Lines-of-code excludes comments, blanks, curly braces, and imports.)

Conclusion: FileInput and TextFile are indeed dogs, but TextFile isn't
enormously worse than FileInput when it's not doing any input-munging.
TextFile is considerably slower than rolling your own munge-every-line
code every time.  The fastest way to read a file in Python is with
".readlines()"; surprisingly, in Perl it's to read lines individually.

Does anybody care?  I do -- the reason I wrote TextFile was to process a
file that is currently ~450k (~20,000 lines); it currently takes ~30 sec
to process using CPython.  After this little benchmarking session, I'm
starting to wonder if there's a better way.  (What?!?  Write a C
extension to read a text file??!  ;-)

        Greg
-- 
Greg Ward - software developer                gward@mems-exchange.org
MEMS Exchange / CNRI                           voice: +1-703-262-5376
Reston, Virginia, USA                            fax: +1-703-262-5367


From guido@python.org  Tue Apr 11 18:31:50 2000
From: guido@python.org (Guido van Rossum)
Date: Tue, 11 Apr 2000 13:31:50 -0400
Subject: [Patches] Please review before applying
In-Reply-To: Your message of "Tue, 11 Apr 2000 14:25:36 +0200."
 <200004111225.OAA31234@localhost.localdomain>
References: <200004111225.OAA31234@localhost.localdomain>
Message-ID: <200004111731.NAA09194@eric.cnri.reston.va.us>

> Most of the changes are simple translations, i.e.:
> 
>     string.split (s, sep)	--> s.split (sep)
> 
> Some of the changes also take the context into account, see:
>     - cgi.py
>     - formatter.py
>     - imaplib.py
> 
> Both 'make test' and 'Lib/compileall.py Lib' produced no error-output
> so I'd expect no big problems.
> 
> The following migt go into the csv-log:
> ________________________________________________________________________
> This (mega-)patch updates all the modules in the standard-module (all
> Lib/*.py files) to use the new string-methods instead of the old string
> module.

Thanks, Fred!

I don't have time to review this before 1.6a2 goes out, but there will
be a third alpha (at least) and some betas.  I hope others will review
parts of your patch!

One concern: Greg Ward is trying to keep distutils compatible with
Python 1.5.1.  So those files should not be touched.

I'm not sure if there are similar concerns elsewhere...

--Guido van Rossum (home page: http://www.python.org/~guido/)


From bwarsaw@python.org  Tue Apr 11 23:31:32 2000
From: bwarsaw@python.org (Barry Warsaw)
Date: Tue, 11 Apr 2000 18:31:32 -0400 (EDT)
Subject: [Patches] posixmodule.c memory leak patch
Message-ID: <14579.42948.116873.135215@anthem.cnri.reston.va.us>

In the process of mem checking my funcattrs patches, I ran across a
memory leak in setup_confname_table() in posixmodule.c.  The cause is
clear: PyDict_SetItemString() borrows ownership of the value object
(i.e. it INCREFs it whether it succeeds or not).  Thus both the
individual int objects and the conf dictionary are all leaked.  This
patch fixes the memory leak (according to Purify) and simplifies the
logic in the meantime.

BTW, Purify reports a bunch of other memory problems which should be
nailed before 1.6 final goes out.  I'll see what I can do about
whacking them.

-Barry

-------------------- snip snip --------------------
Index: posixmodule.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/posixmodule.c,v
retrieving revision 2.129
diff -c -r2.129 posixmodule.c
*** posixmodule.c	2000/03/31 01:26:23	2.129
--- posixmodule.c	2000/04/11 22:21:57
***************
*** 4325,4351 ****
       PyObject *moddict;
  {
      PyObject *d = NULL;
  
      qsort(table, tablesize, sizeof(struct constdef), cmp_constdefs);
      d = PyDict_New();
!     if (d != NULL) {
!         PyObject *o;
!         size_t i = 0;
  
!         for (; i < tablesize; ++i) {
!             o = PyInt_FromLong(table[i].value);
!             if (o == NULL
!                 || PyDict_SetItemString(d, table[i].name, o) == -1) {
!                 Py_DECREF(d);
!                 d = NULL;
!                 return -1;
              }
!         }
!         if (PyDict_SetItemString(moddict, tablename, d) == -1)
!             return -1;
!         return 0;
      }
!     return -1;
  }
  
  /* Return -1 on failure, 0 on success. */
--- 4325,4350 ----
       PyObject *moddict;
  {
      PyObject *d = NULL;
+     size_t i;
+     int status;
  
      qsort(table, tablesize, sizeof(struct constdef), cmp_constdefs);
      d = PyDict_New();
!     if (d == NULL)
! 	    return -1;
  
!     for (i=0; i < tablesize; ++i) {
!             PyObject *o = PyInt_FromLong(table[i].value);
!             if (o == NULL || PyDict_SetItemString(d, table[i].name, o) == -1) {
! 		    Py_XDECREF(o);
! 		    Py_DECREF(d);
! 		    return -1;
              }
! 	    Py_DECREF(o);
      }
!     status = PyDict_SetItemString(moddict, tablename, d);
!     Py_DECREF(d);
!     return status;
  }
  
  /* Return -1 on failure, 0 on success. */


From bwarsaw@cnri.reston.va.us  Wed Apr 12 00:22:14 2000
From: bwarsaw@cnri.reston.va.us (Barry A. Warsaw)
Date: Tue, 11 Apr 2000 19:22:14 -0400 (EDT)
Subject: [Patches] Second round: arbitrary function and method attributes
Message-ID: <14579.45990.603625.434317@anthem.cnri.reston.va.us>

--HXjrLbAr5v
Content-Type: text/plain; charset=us-ascii
Content-Description: message body text
Content-Transfer-Encoding: 7bit


Here's the second go at adding arbitrary attribute support to function
and method objects.  Note that this time it's illegal (TypeError) to
set an attribute on a bound method object; getting an attribute on a
bound method object returns the value on the underlying function
object.  First the diffs, then the test case and test output.

Enjoy,
-Barry


--HXjrLbAr5v
Content-Type: text/plain
Content-Description: Diff -u to add arbitrary attrs to funcs and meths
Content-Disposition: inline;
	filename="methdiff.txt"
Content-Transfer-Encoding: 7bit

Index: Include/funcobject.h
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Include/funcobject.h,v
retrieving revision 2.16
diff -u -r2.16 funcobject.h
--- funcobject.h	1998/12/04 18:48:02	2.16
+++ funcobject.h	2000/04/07 21:30:40
@@ -44,6 +44,7 @@
 	PyObject *func_defaults;
 	PyObject *func_doc;
 	PyObject *func_name;
+	PyObject *func_dict;
 } PyFunctionObject;
 
 extern DL_IMPORT(PyTypeObject) PyFunction_Type;
Index: Objects/classobject.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/classobject.c,v
retrieving revision 2.84
diff -u -r2.84 classobject.c
--- classobject.c	2000/04/10 13:03:19	2.84
+++ classobject.c	2000/04/11 22:05:08
@@ -1550,28 +1550,75 @@
 	/* Dummies that are not handled by getattr() except for __members__ */
 	{"__doc__",	T_INT,		0},
 	{"__name__",	T_INT,		0},
+	{"__dict__",    T_INT,          0},
 	{NULL}	/* Sentinel */
 };
 
+static int
+instancemethod_setattro(im, name, v)
+	register PyMethodObject *im;
+	PyObject *name;
+	PyObject *v;
+{
+	char* sname = PyString_AsString(name);
+	if (sname == NULL)
+		return -1;
+	
+	if (PyEval_GetRestricted() ||
+	    strcmp(sname, "im_func") == 0 ||
+	    strcmp(sname, "im_self") == 0 ||
+	    strcmp(sname, "im_class") == 0)
+	{
+	       PyErr_Format(PyExc_TypeError, "read-only attribute: %s", sname);
+	       return -1;
+	}
+	if (im->im_self != NULL) {
+		PyErr_Format(PyExc_TypeError,
+			     "cannot set bound instance-method attribute: %s",
+			     sname);
+		return -1;
+	}
+	return PyObject_SetAttr(im->im_func, name, v);
+}
+
+
 static PyObject *
-instancemethod_getattr(im, name)
+instancemethod_getattro(im, name)
 	register PyMethodObject *im;
 	PyObject *name;
 {
-	char *sname = PyString_AsString(name);
+	PyObject *rtn;
+	char* sname = PyString_AsString(name);
+
+	if (sname == NULL)
+		return NULL;
+
 	if (sname[0] == '_') {
 		/* Inherit __name__ and __doc__ from the callable object
-		   implementing the method */
-	        if (strcmp(sname, "__name__") == 0 ||
-		    strcmp(sname, "__doc__") == 0)
+		   implementing the method.  Can't allow access to __dict__
+		   here because it should not be readable in restricted
+		   execution mode.
+		*/
+		if (strcmp(sname, "__name__") == 0 ||
+		    strcmp(sname, "__doc__") == 0) {
 			return PyObject_GetAttr(im->im_func, name);
+		}
 	}
 	if (PyEval_GetRestricted()) {
-		PyErr_SetString(PyExc_RuntimeError,
-	    "instance-method attributes not accessible in restricted mode");
+		PyErr_Format(PyExc_RuntimeError,
+	    "instance-method attributes not accessible in restricted mode: %s",
+			     sname);
 		return NULL;
+	}
+	if (sname[0] == '_' && strcmp(sname, "__dict__") == 0)
+		return PyObject_GetAttr(im->im_func, name);
+
+	rtn = PyMember_Get((char *)im, instancemethod_memberlist, sname);
+	if (rtn == NULL) {
+		PyErr_Clear();
+		rtn = PyObject_GetAttr(im->im_func, name);
 	}
-	return PyMember_Get((char *)im, instancemethod_memberlist, sname);
+	return rtn;
 }
 
 static void
@@ -1672,8 +1719,8 @@
 	(hashfunc)instancemethod_hash, /*tp_hash*/
 	0,			/*tp_call*/
 	0,			/*tp_str*/
-	(getattrofunc)instancemethod_getattr, /*tp_getattro*/
-	0,			/*tp_setattro*/
+	(getattrofunc)instancemethod_getattro, /*tp_getattro*/
+	(setattrofunc)instancemethod_setattro, /*tp_setattro*/
 };
 
 /* Clear out the free list */
Index: Objects/funcobject.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/funcobject.c,v
retrieving revision 2.18
diff -u -r2.18 funcobject.c
--- funcobject.c	1998/05/22 00:55:34	2.18
+++ funcobject.c	2000/04/11 22:06:12
@@ -62,6 +62,7 @@
 			doc = Py_None;
 		Py_INCREF(doc);
 		op->func_doc = doc;
+		op->func_dict = PyDict_New();
 	}
 	return (PyObject *)op;
 }
@@ -133,6 +134,8 @@
 	{"__name__",	T_OBJECT,	OFF(func_name),		READONLY},
 	{"func_defaults",T_OBJECT,	OFF(func_defaults)},
 	{"func_doc",	T_OBJECT,	OFF(func_doc)},
+	{"func_dict",   T_OBJECT,       OFF(func_dict)},
+	{"__dict__",    T_OBJECT,       OFF(func_dict)},
 	{"__doc__",	T_OBJECT,	OFF(func_doc)},
 	{NULL}	/* Sentinel */
 };
@@ -142,12 +145,21 @@
 	PyFunctionObject *op;
 	char *name;
 {
+	PyObject* rtn;
+
 	if (name[0] != '_' && PyEval_GetRestricted()) {
 		PyErr_SetString(PyExc_RuntimeError,
 		  "function attributes not accessible in restricted mode");
 		return NULL;
+	}
+	rtn = PyMember_Get((char *)op, func_memberlist, name);
+	if (rtn == NULL) {
+		PyErr_Clear();
+		rtn = PyMapping_GetItemString(op->func_dict, name);
+		if (rtn == NULL)
+			PyErr_SetString(PyExc_AttributeError, name);
 	}
-	return PyMember_Get((char *)op, func_memberlist, name);
+	return rtn;
 }
 
 static int
@@ -156,6 +168,8 @@
 	char *name;
 	PyObject *value;
 {
+	int rtn;
+	
 	if (PyEval_GetRestricted()) {
 		PyErr_SetString(PyExc_RuntimeError,
 		  "function attributes not settable in restricted mode");
@@ -178,8 +192,23 @@
 		}
 		if (value == Py_None)
 			value = NULL;
+	}
+	else if (strcmp(name, "func_dict") == 0 ||
+		 strcmp(name, "__dict__") == 0)
+	{
+		if (value == NULL || !PyMapping_Check(value)) {
+			PyErr_SetString(
+				PyExc_TypeError,
+				"must set func_dict to a mapping object");
+			return -1;
+		}
+	}
+	rtn = PyMember_Set((char *)op, func_memberlist, name, value);
+	if (rtn < 0) {
+		PyErr_Clear();
+		rtn = PyMapping_SetItemString(op->func_dict, name, value);
 	}
-	return PyMember_Set((char *)op, func_memberlist, name, value);
+	return rtn;
 }
 
 static void
@@ -191,6 +220,7 @@
 	Py_DECREF(op->func_name);
 	Py_XDECREF(op->func_defaults);
 	Py_XDECREF(op->func_doc);
+	Py_XDECREF(op->func_dict);
 	PyMem_DEL(op);
 }
 

--HXjrLbAr5v
Content-Type: text/plain
Content-Description: Test of func/meth attrs
Content-Disposition: inline;
	filename="test_funcattrs.py"
Content-Transfer-Encoding: 7bit

from test_support import verbose

class F:
    def a(self):
        pass

def b():
    pass

# setting attributes on functions
try:
    b.blah
except AttributeError:
    pass
else:
    print 'did not get expected AttributeError'

b.blah = 1
print b.blah == 1
print 'blah' in dir(b)

# setting attributes on unbound methods
try:
    F.a.blah
except AttributeError:
    pass
else:
    print 'did not get expected AttributeError'

F.a.blah = 1
print F.a.blah == 1
print 'blah' in dir(F.a)

# setting attributes on bound methods is illegal
f1 = F()
try:
    f1.a.snerp = 1
except TypeError:
    pass
else:
    print 'did not get expected TypeError'

# but accessing attributes on bound methods is fine
print f1.a.blah
print 'blah' in dir(f1.a)

f2 = F()
print f1.a.blah == f2.a.blah

F.a.wazoo = F
f1.a.wazoo is f2.a.wazoo

# try setting __dict__ illegally
try:
    F.a.__dict__ = (1, 2, 3)
except TypeError:
    pass
else:
    print 'did not get expected TypeError'

F.a.__dict__ = {'one': 111, 'two': 222, 'three': 333}
print f1.a.two == 222

from UserDict import UserDict
d = UserDict({'four': 444, 'five': 555})
F.a.__dict__ = d

try:
    f2.a.two
except AttributeError:
    pass
else:
    print 'did not get expected AttributeError'

print f2.a.four is f1.a.four is F.a.four

--HXjrLbAr5v
Content-Type: text/plain
Content-Description: Output of test of func/meth attrs
Content-Disposition: inline;
	filename="test_funcattrs"
Content-Transfer-Encoding: 7bit

test_funcattrs
1
1
1
1
1
1
1
1
1

--HXjrLbAr5v--


From Vladimir.Marangozov@inrialpes.fr  Wed Apr 12 05:22:09 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Wed, 12 Apr 2000 06:22:09 +0200 (CEST)
Subject: [Patches] regression test finalization
Message-ID: <200004120422.GAA06916@python.inrialpes.fr>

Feature-specific modules keep being imported until the end of
the test suite. This increases unnecessarily the overall memory
consumption. This patch suggests a best effort per-test finalization
by unloading the newly imported modules after every test.

If not convinced in the usefulness of the patch, I can provide figures.

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--
I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------  
*** Lib/test/regrtest.py-orig	Wed Apr 12 06:09:22 2000
--- Lib/test/regrtest.py	Wed Apr 12 06:08:38 2000
***************
*** 105,110 ****
--- 105,111 ----
      if single:
          tests = tests[:1]
      test_support.verbose = verbose      # Tell tests to be moderately quiet
+     save_modules = sys.modules.keys()
      for test in tests:
          if not quiet:
              print test
***************
*** 118,123 ****
--- 119,128 ----
                  print "test", test,
                  print "skipped -- an optional feature could not be imported"
              skipped.append(test)
+         # Unload the newly imported modules (best effort finalization)
+         for module in sys.modules.keys():
+             if module not in save_modules:
+                 test_support.unload(module)
      if good and not quiet:
          if not bad and not skipped and len(good) > 1:
              print "All",


From hniksic@iskon.hr  Wed Apr 12 08:18:50 2000
From: hniksic@iskon.hr (Hrvoje Niksic)
Date: 12 Apr 2000 09:18:50 +0200
Subject: [Patches] Re: Fix for readline() method of mmap objects
In-Reply-To: Hrvoje Niksic's message of "04 Apr 2000 20:11:32 +0200"
References: <dpd7o5u46j.fsf@mraz.iskon.hr>
Message-ID: <dpitxn4wit.fsf@mraz.iskon.hr>

Hi!

Has the readline() fix been applied?  I've received no response to my
email, nor can I find anything at the patch list archives.


From bwarsaw@cnri.reston.va.us  Wed Apr 12 18:24:39 2000
From: bwarsaw@cnri.reston.va.us (Barry A. Warsaw)
Date: Wed, 12 Apr 2000 13:24:39 -0400 (EDT)
Subject: [Patches] posixmodule.c memory leak patch
References: <14579.42948.116873.135215@anthem.cnri.reston.va.us>
Message-ID: <14580.45399.505489.166542@anthem.cnri.reston.va.us>

>>>>> "BW" == Barry Warsaw <bwarsaw@python.org> writes:

    BW> BTW, Purify reports a bunch of other memory problems which
    BW> should be nailed before 1.6 final goes out.  I'll see what I
    BW> can do about whacking them.

On further inspection, most of those are in system libraries, or in
pcre.  Nothing I can do about the former, and given sre, I'm not going
to worry about the latter.

-Barry


From mhammond@skippinet.com.au  Thu Apr 13 03:23:43 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Thu, 13 Apr 2000 12:23:43 +1000
Subject: [Patches] Patches to build process for Windows
Message-ID: <ECEPKNMJLHAPFFJHDOJBMENLCIAA.mhammond@skippinet.com.au>

Attached are minor build-process changes for Windows.  Notable
changes are:

* Temp directory for all projects are now specific to the project
(rather than common as before).  This avoids any conflicts with
debug symbols or common file names etc.
NOTE: You should manually delete your existing build directory after
applying this patch, as the MSVC "clean" command will now only clean
the new temporary directories - not the existing common temp
directory.

* Base address for all extension modules updated. PC\dllbase_nt.txt
also updated.  Erroneous "libpath" directory removed for all
projects.

* winsound module moved from a builtin module to an extension
module.  This was done primarily to avoid Python16.dll needing to
pull in winmm.dll.  Really dumb test added for winsound - but if
nothing else it ensures the module imports.

* PCbuild\.cvsignore updated to hide more junk!

Ive attached everything as complete files, rather than diffs (most
of them are either new files, or modified binary files)

Release info:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that
I have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce,
distribute, perform and/or display publicly, prepare derivative
versions, and otherwise use this contribution as part of the Python
software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize
others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and
its related documentation.  I further grant CNRI permission to use
my name and other identifying information provided to CNRI by me for
use in connection with the Python software and its related
documentation.

Mark.


begin 666 win32_build_patches.zip
M4$L#!!0``@`(`"E?C2AC)S `5 4``%,@```1````4$-B=6EL9"]B<V1D8BYD
M<W#M66M/VT@4_8[$?[AU]T-;E3B$"M%HLU*(DS1J7HI#V:Z0HK%GDLS&]EA^
M`/GW>\>/Q,X+"&3A`Q45\SQS[IWA''O\$3K<](0OQ@%H[)99PF4>Z$%(N8"^
M)_YE9@`-;C$X@2ZQ644Q?$H-!:M]8L[(A$'OSF%>Y<]O?QT??=R%UF0XC@2,
MPF7(+1JA?H6&\&P2P"_F^5PX<%XH%B7.ER^@]:#;&T)=:PVQ>GPDFX?507/X
MNU\'Y9H[9R7X=']Q_AFTN4-L;IZTN3.#-C<\XLT5*-X73XNEC;.J[?Z/ZHYY
MYW+>\5&MT:Q$X6*T\<RJY4X)AF:$D^.C#YVZKE>;=1A.N0_XXX@`"-P2BU.P
MR8R-,<("# 484<"!'.8F.0U][DR@VZG^K'_-0(4^PW$,ZO>N\ +H)"A@"MLF
M#@7YWPN=S(Q,,4(#=0SQ)A60@U+8//2W",$D#O@N,_EXCK1-X8SY),0=DOMP
M-V6.7,A9L,Q,-N9 D5;4)\G:!#<=,%N ,V5#RM;B#B8`MQC8/;%=BY4?3USB
M):=M4_J5S4A]X?O<D F;"FXR'\:X>#XTXFVCL;+:@%F,^$R!3P;^HC*XAX_=
MY^UX,>^70LOF8@/FKB/^(.J.R!^+&_^]7K()=U(=D0W]0:^/BUCBKL^\6K0O
M&G.90YEC<MRNXF*0;IHC.5&J3L)2R76VA4FL/@FFH!0*BESO0ZL!H/SQ"4_.
M9P4JE6T;&G.+@"ZK>AVN?#;J-&J9Q1?-47Y'<8!Y?M&07ABX83#2N ?*$CT[
MHN4$S+,9Y:A\N\8-B3=A"=*R:YW9;E)9/H4ES#H)/'(G`;/=$R]F<[.2X-;$
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MK=0PTP%8--VDLDBZ'QK^W,<_GC*N3L6=CYFQ+-P`8D[1[\JMLXOS-)HDD SW
M7$RO$LC:"8_.R W:;BY,J?QEI7A_RDXOBOA/V1XYE7J43P`>P3 H*P4U-G1W
M3G$^B]2DC \.0=0LLZ1?70X'U=HP395ZRSQ#X%.0ZE*C[ B'1<I>;^OU!]0]
M?2S86]M/MVM[BKU;V3>->HZNGSY#UZ,]^5]5G28:;&=TF*XIY7[RO@7[(!H_
M.HC&CYZA\:,7U?C1`QH_>M?XIRD="A5N"RO[S W>A1]Q\.60A%: J&6EH_^J
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MHKI#\VF(F](]BVN+CWS_`5!+`P04``(`" `I7XTH)(%H'$D$``!A#P``$ ``
M`%!#8G5I;&0O;6UA<"YD<W#M5VUOXC@0_EZI_V$VO0_M:DLHG*K=ZC@)2$!H
M>8D(O6Y7*T7&-N"K$T=Y:<N_OW%,>&FAN]>]2G?2H:+:GIEGGAG;3\()# 1-
M5*IF&3C\GDL5\P3\+&="@9>H/SG-H",DAW,8DI WK# DL84SC] [,N<P>HAX
MTOCMU]^/CTY>`NMR]",99]#*A60%Z ?HJ"0D&?S!DU2H""XKU:K&>?\>G!$,
M1Q-PG=X$I\='>GG2''<GMYX+UHV(ZC4X??QX>0;.,B*AH.=]$=U!7TP3DBPM
MJ#Y6+ZHU'=?N=!N:-9(V80Z?YO/CHW<#U_>;71<F"Y$"_D4J`P+W1 H&(;GC
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MCJVO*)$>R19@52IH:GM>@\H*?^3'1X-)OQ$*MIJ-_78CH6:,#'L=`.N74]RG
M,PL:C0-=-/R+;*VF[\)URH-!I[U%<+U<="HP3=BMH7 9Y5F<9X$C$K VZ-L>
MO2CC2<B90)5XR6]"DCE?(6U,SYF]3&J;3V4#\YP$;O1YQL/X/#%LONUN0F\>
MJ80'YL+K3%LYGC#M?*HVV,SLP DT'<?4@UL&=J2DFBNP!Q.P;^I@=[^ /:J!
M[>"!ZPWK-:L8#AVW==TUXP#7G=&-OYH-6NUR>.V/G7[?3 :#IA>X7[S1>*+-
MMPC;<<"F)8/=Y,YN\IX^5-]Z$94YPQNUFGMMZ^UY%9W!`[PAMYTFO,N6L133
M6A79/N@36\:](J1(A;<#;(E/#5K]!#9;!Y9.A^TMOUVO-:8IU8^,9YM;6$M&
M:\ON8K\W_(P0*+=WS^*-#>YX$G%9KU6P`OV025;#.1.K$9:4QDK)8H(*SN1\
M92'LGL2E6[K@LL11DF]&),]*^%PPL\JFM'3 (8U7DW5_F9386$(7^*"XZM4_
M7I;4_]6LM91?61?\HE/%C[6G#+!5GEU9%;MXGL5+5NC@.[?ONR_+9OF,>[5H
M7AP6S1+[9<G<Y_4S@GGQ$X+)--9;RR5;259H9.MK#Z6+/=6GX)_2I^[7P]IY
M@,D/B^C;D-PKI,'?%]+@M4(:?$=(@_^%U"A0<5V>"E',IMAN?I7R./L/J^O>
MVK9$-F"%S#ZM5XONT$'-W7Z/-C)A5C:OQ7M>7?>:-P)IT'R5)Y07/T<UI#^Z
M'K?=!M[0@6*YY&FA7V$QKA27RL5?1CM!9JED96;K=_V_`%!+`P04``(`" `I
M7XTHZ(%-<2H%``#C'@``$@```%!#8G5I;&0O<&%R<V5R+F1S<.U9:4_C2!#]
MCL1_J/'L!Q@-B0DC-!-M5@IQDHTFETA8EA62U>GN)+VQW98/(/]^JWTD-CFX
M`LQJ!H'HH_KUJZK.J[C]$3J">M*7XP`,?L,MZ7(/!D'(A(2^)__E-("&L#@<
M09?8O**YQ/.YIV&_3^B,3#CT;AWN57[_\L?^WL=M<$V.=B3@#,Y"8;$(]C,T
MI&>3`/[BGB^D`Z<%75<XGSZ!T8-N;PAUHS7$[OZ>&AY6SYO#JWX=M$OAG)3@
MX.[KZ2$8<X?8@AZUA3.#MAAYQ)MKH-_IQWII[:IJN_]G=<NZ4[5N?Z_6:%9B
M?]'=>&G5<J<$?1N%D_V]#YWZ8%!MUF$X%3[@KR,#('!#+,' )C,^1A<+,)0P
MBCP.E)F;1#7TA3.!;J?ZO?XY`Q7Z'.TXU.]<Z07025" 2MLF#@/UYX5.9D6F
M&:%!<0Q)F@I(0BNLM[V2(5#B@.]R*L9SY$VE,Q:3$'.D,G$[Y8[:R5G0S"P>
MS8$AKVA.L;4)IATP7H KU4!*UQ(.1@"3#/R.V*[%RT]@K@#3$[<N`]IZK+[T
M?3%2,9M*0;D/8]P^[QSQ-A&YO]TYMSCQN08'(_S'E'\/G[W#+8 Q\YW!9<.Q
M!G3;27\8=HOSCP6./[=G?"*<5%#40/^\U\=-+'G;YUXMRHW!7>XP[E"!*=,7
M1@-*3;50R4]*4\O-MB4E5I\$4] *!4UM^*'5`-!^.\#S<ZA!I;(QJS&[".FL
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M+3$JZ<A;XL1%&T=OU4%-`9ZW]GQ0JWATA0@.0]'".O=%_P9%MD!+C3;/GPUJ
M)Z7*R*>JN*W@1K,IS<5,?K#=ZGY'""P+LY7U\1S,N.=PZZ140+=4.?22YH2)
MI(7N^:Z45M3!2L.L23)#V UQ4S-_RJT41UI\V2)AD,*'@L6C;$13`VQ2-^DL
M@NZ'(W_NXP>EC+LS>>MC9"P+$T#H%&M<N77R]33U)G$DPSWGT_LZHI2\K.EW
MQ[RJJQ]MLV],J4O>13QD85#6"L6D3+MSA@`\TH8R?A\(XG$5B<'%V?"\6ANF
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ME"8O[BW>-?T'4$L#!!0``@`(`"E?C2AT$O+ 2@4``) ?```3````4$-B=6EL
M9"]P>65X<&%T+F1S<.U9:6_B2!#]'BG_H<;9#YG1!!PRBF;0LA+!P*#A4B [
MFU4DJ^ENH!?;;?F8P+_?:A]@AR,7LYE5$A&EC^K7KZJ+5W'["#J">M*7XP`,
M_H-;TN4>#(*0"0E]3_[#:0`-87$X@2ZQ>45S%WSNDD##@3ZA,S+AT+MUN%?Y
M_=,?AP='N_":'.U(P!E<A,)B$>Y':$C/)@'\R3U?2 ?."[JN<#Y\`*,'W=X0
MZD9KB-W#`S4\K%XVA]?].FC?A7-6@N/YY_/W8"P<8@MZTA;.#-IBY!%OH8$^
MUT_UTL95U7;_:W7'NG.U[O"@UFA6$H?1WWAMU7*G!)T;A9/#@W>=^F!0;=9A
M.!4^X,>1`1#X02S!P"8S/D8?"S"4,(I<#I29F\0U](4S@6ZG^JW^,0,5^ASM
M.-3GKO0"Z"0H0*5M$X>!^O5")[,BTXS0H#B&]* *R$(K;#:^EB%0XH#O<BK&
M"R1.I3,6DQ!/29W%[90[:BMGR3.S>+0`AL2B.477)GCP@!$#7*D&4KZ6<# $
M>,S Y\1V+5Y^#'6%N,RZ38>@;4;K2]\7(Q6VJ124^S!&`GGWB+>-RMI^E]SB
MQ.<:'(_P#U,NWI^ [W<AQMSWAY>-R ;470G_`-P=_C\4.?[^7O")<%)E40/]
MRUX?-['D;9][M>A\#.YRAW&'"CPV?6DTH-14"Y4.+7EJN>FVI,3JDV *6J&@
MJ1W?M1H`VF_'F$7O-:A4MA]MS"^"NJ@.ZG#E<[/3J&4(+(>C*)NQDWF.D4DO
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MXY[#K;-2`=U2!=%+FA,FDA:ZY[M26E$'2PVS)LD,83^(FYKY4VZE.-+BJQ8)
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M_8!%MU$[X-10FEMQ;_D^[%]02P,$% `"``@`U5V-*.9!"#ZX! ``TQH``!(`
M``!00V)U:6QD+W!Y=&AO;BYD<W#M66EOVS@0_1X@_V&J[H>D:"0G+H(V6"]@
M6[+7J _!<MK-HH!!D[3-C4P*.G+\^R5UV/(EQ\DF:1<!$H3'\,V;(?DH,N^A
MP[ O`C$.P:0WU!4>]<$)(\($V+[XA^(0&LRE< )=-*,5S;L/IX)KLFXC?(TF
M%'JWG/J5WS_]<7CPO@BN2:4="BF!6L1<$L-^A(;P9RB$;]0/F.!PKI=*"N?#
M!S![T.T-P#);`UD]/%#-@VJ_.;BR+="^,UX^@Z.[S^?'4!<\$))CU?-<AE'(
M%,'27>FT5-XXJ-JV_ZQN'W:NAAT>U!O-2A*M##89676]*9*1C:+)X<&[CN4X
MU:8%@RD+0/YP$0*"&^0R`C-T3<<R0!T&`D9QO*$R\]*<1@'C$^AVJE^MCSFH
M**#2CH)UYPD_A$Z*`EC,9H@34+]^Q',C<L48#8PQI).D2Q*:OMGV2D2 $8? 
MHYB-[R5O+/B832(_S@/<3BE7GOB<9F[PZ!Z(Y!7W*;8S)"<=9+Y CE0-&5V7
M<9D!.<5 []#,<^G%'LP58+;>-LV MAG+%D' 1BIG4\$P#6 LW2\'A_QM1%;=
M]:E+44 U.!K)/T3%MW/E'1?@)<3_*[1\,C9@%BSSW:@%D3\0-]FQ-3IA/),2
MU6#W>[;TX8I;F_KU>%Y,ZE%.*,=,3E=I;N1@/%0#E?!D++6EWK; R+51. 5-
MUS7E\%VK`:#]=B37SK$&E<K6&4W8Q4BUJF/!94"'G48]YW[>'"=XV&8C'_G+
M#&.37A1Z43@TF0_:`CUOT>(A]6>4,*E^178#Y$]HBK3H6F=63"K/1U_ K).0
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MF,%'C"2M9(0S`UG$7EJ9SUH0C8+[0*[ "YQ*B2%/DJD\,RY:Y<_G62!I#-O@
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MAN.%9W&R;+P3H2-(Y-(@E1X=[_UPN>=;PJ,^Q1]Y`BW-\_ZIB9,;GXQRP^H^
MWH*BFK+ED]3F+\S_`E!+`P04``(`" `I7XTH1?"H1K (``#0C@``% ```%!#
M8G5I;&0O<'ET:&]N,38N9'-P[5WO3^LV%_Z.Q/_@Y;X?[IUVVP(;VM Z"6AA
MU8!6M.SN3DB5X[BIWR9V9#M ]]?/3EI(:6N7'W?C:@<!31SG^/CQ\7-.3IST
M'3IG1 HE1AJUZ U-1$8EZNL\8@+UI/@_)1J=L(2BC^@"I[099%,]%GQG/S E
M/4PF.*:H>\NI;/[\_2_;6^]<`D^IJ8<UC=!1SI*H$/P=.A$RQ1K]3J5B@J/]
M6J-AY7S[+6IUT45W@-JMSL#L;F_9XL'AY>G@<Z^-@D^,[^VB]W<_[G] K2G'
M*2,?SQB?H#,62BRG`6K<-78:NRO/.CSK_7KH.&_?GK>]=7QRVISWV'2X//DP
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MZ'(^.&NF+%KNL3GPH%ZUO72BIUG"PMV&T=<<NRV,,[BX.@OF`IYW[F7_N"G)
MDB*F&-43X_6^;_R$ZM&]M'FEQ>/L$6C5^D?]X[W=9JB(=7Q+[11'YVK?'UDL
M/.M<_&9$&(<Q63J_/(8F5'*:[.W63#>MJY2SS3ABLRW3:94)D10[Q@=%23P[
M@J,;G,VKJ3%-YG)$0A^V<*[GXG,6E:512.85S";)9COW@Z#R4$V5F3('IO5(
MW"J#3)*8`<%D;+S?06?OQ_UY;V8=2>PD85P7@M;U:D'C^]8L^1X$C;L=NM.P
M/\%Z!2)+!HMZ6#G&<>,\T4;J06#^D6*T^U='@\O#X\%<PWH6A0=<<%H0:/NL
MW_:3Z-P?/YM"=]93Z%RVFT!7U7H)?>Z\@#X+]/]I\HQF5)=6Z"Y:XK'GL>AC
MV9U2MIM*JVT]9LRU[/H"*AV^@$J'KTJEPR=2Z1"H]&E,9AC*#!,]4#33;YI?
MC0WE^B"HU>=D,(QJ1D"PW(778.'%:Z.WQ<4N[OLWR!I;K%Z-KC>AZ%/=>!2-
M>BG4S]R.:+1P":L9_1]7I15]E5SNXNHWR>VW2I#)C-<6F-#T/4V=G/A%Z?\5
M.+;(-+P6R58QJS#MO^])7K67+W8E;SDW\D4=2N-%#N55TB>;IDR6N7SW"5Q^
M\60N7Y]I6=#DRVOQM>587#F4-YES^4\ZDZ_=36S@`"Y:AO^K"?F2><J2Q?SZ
MBBSYVBH/4?^:"HN7)YYJU09++?LBEX06]P^MJOWNU>5QNVFFR+F(\H2JZR$Q
M<!"5%KLU\HS\_Y.S7<^\,'NV$UX8O#:/%D'9'*G$WBFA@-1:I+JAO46EKG&H
MM,1$`T@K0.IA:2CN&A-B9$J R#'CL)1X"O/-CY-=,R(RP,B!4<@X5H0Q`&D]
M)X5,*PJLO1*AHH7K,-$FC =&\D4`83X:42F*/0!JO3T1>H,3`,AA223!2H$A
M>?T;2;$>`S/Y[0ELR4M*148`$'(AE&8,)IIGHMGERW<PW=8;T_%UN<(:T'$Y
MMAXC$YAK;HSZ6C(>=[J DH.1(D8TT)&+CJ)DR $<A^./ICP1.!K>,@XP.0B)
M2LD%7(WXS,G )"1$VB[2MH_S`6G[83*T!,[-3TRCS""4`#-M`I2F2@-2_JDG
M<0H4Y?5U(RG^HA U.2TIYY">]!I23#66,41-+O8V&!4O46!\) `HIRW-<KFP
M$L<'4S:5+!X#-;EQ2C5+(5YR8R0RL*(U&3B#3H;U&-9RN2TH2[ >"9D"3$Z8
M;LHW10%*#I3,U1OC# C)L2[08I1BN0,8.4)NEN*8PB)<UUQCJ7V"#Q!:[?I+
M=)Y[`Z[R>.'LB<&B=O#,]]&\6-B;-+T(UEJZ\D^,PRT$KS-,F-+V$4+ R&%)
M%B0P)3],@L<`DS>T2C$LMW Y-Q.;JS$\1^"Q(5@>OPE,T0\$`'(#!&;DBY%2
MJO$L:0`P.?R_P6DL(H@`O Y.1"K/('G@,Z>"F,"<7%F65-T0"<N:_!0^E11'
MQ=<1`$KKB6FJM-0BA]C;%3-Q>@L3SC?A(*?D\6[@U[P3S7Z%$-8"@F['/,N*
M#T#(@Y 6$\#(,=4R(N%5<!O )!2[`YQ\<60V'64`D-.0IG;&`40N&U(::X!H
MS85_V8A=7B')"W&V'S*'6U*N:%UB'D,JRL]KDL84'.2&.&40MSLQBD.6QF!,
M?J"$N4@&G+PXJ3$&E/PHL9AC>*;<'Q.HA!&("38P*)'SB-X!1BY;*MYX!<;D
M-R8M10;DY+NJ,S#E1*<T#2$Y[+,G"Q08E,^@IO"E*EZ0]-BN,@"$'/.MA A,
MR0\42^&.C/?FGA83RME?X.2<M"0QH2$F< O4%8+K/(/5AIO@-,T`)C],.6?%
MF^<M7 "4'R@P*6](,,UOB. W]+^Z;MP6S;^GLMSK26&M9GOK;U!+`P04``(`
M" #!78TH[@=3P. $```_'0``$P```%!#8G5I;&0O<'ET:&]N=RYD<W#MF6UO
MXC@0Q]]7ZG>8S=Z+=K7E<57MHN,D(,"AY4E K]?32I%Q#/A(["@/A7[[&R<$
M0GEH"VQW[ZY5JSKV^.^9\? SF/?0XM25GASYH+-[9DF'N=#W`Y-+Z+KR;T9]
MJ'&+P16TB<V*FO/@3Z28:=C1)71*Q@PZ,\'<XJ^??CL_>[]/K\[0COC,A'+ 
M+3/4_0@UZ=K$AS^8ZW$IX#J5R2B=#Q] [T"[,X"JWAC@X_F9ZAZ4>O7!7;<*
MVBT7^1Q<S#]?7T+)<2Q.B8\"&F3FF6PFN]6XU.S^7MHTOU;FYV>56KVX" ^C
MBZ:4+&=",)1A,#X_>]>J]ONE>A4&$^X!_@KI`X%[8G$3;#)E(XPH!0,)PS! 
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M?A5N/&:T:I6$#\ON,*U&DP]=XJZ[&9IT`M\)?$/G+F@K]:1%0_C,M9G)$6W[
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MV*I%`C^6#[@9]9I#&AM@DSJ+AV72O6#H/7A8]P5<W90S#W-/Z 0/I4(C__DZ
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M+[^)?^FUTV$?D0X]9-?V>TMJ5%=<`-'3\DN/?P!02P,$% `"``@`*E^-*&;$
MG5J#! ``-1 ``!(```!00V)U:6QD+W-E;&5C="YD<W#M5VUOXC@0_EZI_V$N
M>Q_:U9;PLJIVT7$2$$!H>1.AU^MII<C8AOI(["@O!?[]C?,"H12J;>]T7[:B
MK3TS?OS,C/.8?("AH($*U2("BS]Q5_D\`#N*F5 P"=3?G$;0%2Z'&Q@1CS>,
MD+MH,W ^(71%EAS&:\F#QF^??[^\^' .KL<QCD2<02L6+DM@/T%7!1Z)X \>
MA$))N"V5RQKGXT>PQC :SZ!C]6<XO;S0YEESVIL]3#I@W M9J\+5YLOM-5A;
M23Q!;P9"KF @Y@$)M@:4-^5*N:K7M;N]1LH;::<++3Z/EY<7OPP[MMWL=6#V
M*$+ CU01$'@BKF#@D15?(,D2S!3,$\Z1#O.SNL2AD$L8#9O?.I\*4''(,8Y#
M9^.K(()AA@)4>1Z1#/1O$,O"BL(P00-S`5FA2TC"*+T<^Z!BH$1"Z',J%EOD
M395<B&6,5=:U7#]RJ7>2.YJ%Q?,M,.25^#1;CV#C`"L%N%(;<KJND%@!;!/P
M#?%\E]=_@+D&S,_,8>V-EU$F*@S%7%?K40G*0UC@QH=ID> 4A><;37%*0F[ 
MU1S_,9W9Z^?F^@Q@ROS-<.D9;O&ED/G#I0V3Z7@"3==5ZPD/VDFN%O>Y9%Q2
M@24H[X)L2AV]4#^*.3GCP#M0E+@3$CV"42JAJSV9-*A;XAM^>3&<#1J>8-EL
M:K<;`4W'R+'?!3!^O<)^71O0:)RL99I#LE^K:7?@+N3.L-LND-R9DVHY:2$.
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MQ[^#[WD'RU)3SBV=[=Y8_@%02P,$% `"``@`*E^-*&Q"^-%Q! ``CQ ``!<`
M``!00V)U:6QD+W5N:6-O9&5D871A+F1S<.U7;6_B.!#^7JG_82Z]#^UJ2RB<
MJEUTG 0D(+2\1(1>MZN5(A,;\)'845Y:^/<[3@B$\M+=]O9T)QT"87MF'C\S
MMA\G%]#G;B@C.8W!8(_,DP$+P8X3RB58H?R+N3&TN<?@&@;$9W4M$=R5E%$2
M$PT'+>(NR(S!\$FPL/[[;W^<GUV<PNPP]",QH]!,N$=3[/?0EJ%/8OB3A1&7
M`FY+Y;+">?<.C"$,AF,PC>X8N^=G:GC<&'7&#Y8)VCT7U0I<+C_<7H&Q$L3G
M[G6/BP7T^"0DX4J#\K)\4ZZHN%:[4R^01^Y9M,$FR>S\[)>^:=N-C@GC.8\`
MOT+&0."1>)R"3Q9LBDQ+,)8P28G'RBU85RB)N)C!H-_X9+XO0"410S\&YC*0
M80S]-0JXTO>)H*!^82(*$85FB@;Z%(HE+R$3K70XX$$FX!(!4<!</ETA>5>*
M*9\E6&]5U:<Y$VHZL>%:")ZL@"*YU*8H^P27$+!F@)%J(.?L<8%EP 4#MB1^
MX+':C])7J#O[:'<IM,-XEHPB/E'%FTONL@BF2&$W01(>(W-PMA'S&(F8!I<3
M_*,JT9<WU-5+J%D.K\;,=GB3S;C(SY\:L$9#"QJ>)Y\L%K;2K T6,$&9<#D6
MH[QQLEW748'JM.XPU'9<>M(EGD7B.6BE$II:EE5WO1);LO.S_KA7]SE=]T9V
MJQZZ61N)=ML`VJ^7N(97&M3KITN;99-.VFS8)MQ%S.FW6P6ZF^&T;DY6DMV,
M4I=A$@=)[!@\!&V+7O3HBIB%/J,<U>64WYB$,[9&VIKVF9TF5>13VL+LD\!E
MOXZ9'UR'&9NO!Y>D.Q,R9$XF%&K"PE3/"+<_ENMTFJW'!30,(TL+%Q!T(3TY
MDZ#WQZ#?5T'O? 9]6 '=P%W8'50K6MH<&&;SKI.U'1PWAO?VNM=OMO+FG3TR
M>KVLT^\W+,?\; U'8V5^0-BV`;J;,]B=W-B=O*NVV->N<+V$XFE;]ZV6]O-Y
MI97![;PE5YS&7\2KP..32AG9/JF-F\>](B2="L\*Z!Y>.F[Y(^AT$Y@[';<W
M[5:U4I]$KKIJ]A8WM>:,-I;=P5YW\ DA4)\7>_&9#18L%,RK5DJ8@;J<PG5S
M1OFZA2E%@91>VD')I]YL;2'TD02Y6S1G7HXC/;9MD23.X1-.LU$Z<7,';+K!
MNK.I+_4\+"QQYWBSU+K5#[<Y]7\U:Z7O-:V\O#%N*F7\: <2`5TF<4TKZ<4K
M,%A1]&7I2:_AA1L7C"IS^ZXY'C5:XSQ]/:"3FI BDU^S9YO?)<'Y7?IJ`;XY
M+L Y]FGY/>3U%O&]>8/X4H7U#TDO7<N?GTG@ER[*('VN=<[?I76=+\=U^ B3
M[Q;DGT/RH"@[/R[*SFM%V7E!E)W_13G3LO34/)<TE",L-ZM%+(C_TTI],+M]
MP7;H6K*+B;]!P <&ZG?Q,3\3F6QD[ZG]P"/U*:^MYF;8MDQ"EZ5OV6H">W@W
M:IEU/.U]21./145)+*4GU,1WO)VHUT"IDA^%4T-YSEEO\Z+S#5!+`P04``(`
M" `J7XTH/*=.&DX$``#F#@``% ```%!#8G5I;&0O=VEN<V]U;F0N9'-P[5=M
M;^(X$/Y>J?]A+GL?VM4VH7"J=M'E)" !H>4E(NEUNUHI,HD!'XD=Y:6%?W_C
MA(304JKMKO:^'"K"'L\\\WAL/W;?P9AYL4C$(@6#/M! 1#0&.\U\)L"*Q3_4
M2Z'/`@I7,"$AU95'QA.1<5]!BT6\-5E2F#YR&NM__O'7^=F[4X #BGXDI3YT
M,Q;X.? 'Z(LX)"G\3>.$"0XW:J,A<=Z_!V,*DZD#IC%TL'M^)LU.9S9P[BT3
ME#O&6TVXV'R\N01CRTG(O*L1XVL8L7E,XJT"C4WCNM&4<;W^0"^9(_$BU*#S
M;'E^]MO8M.W.P 1GQ1+ /RY2(/! `N9#2-9T@315< 3,<]:I=(MVM<D2QI<P
M&7<^FQ]J4%E"T8^"N8E$G,)XAP*>"$."%.0WSG@MHM;,T4!;0%5L%6DHZG'O
M>Y&!1S@D$?788HO,/<$7;)EAI64]'U>4RUR\(EH+GF_!1V;YF.0;$EP\P&H!
M1DI#23A@'&N 2P5T0\(HH.WOXBXA]WOG< 64XTB62!(VES5;">;1!!:8_'!J
M)'Z)QO-4,QI0DE %+N;XX\OYO;Z#+D]"%NS?#%CLYRY=,EX>-6FP9E,+.D$@
M'BT:]_+Y&C2BW*?<8UB&1N5D>YXK`^7!!$4YL(^$1P*+I*M\H&=9NA>H=$//
MS\;.2 ^9O^O-[)X>>T4;N0W[`,KO%[A:EPKH^HDZ%NSS?-V.;<)M0MUQOU>C
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M\>2M\6PE\]&2435R:!P-)Y\1`M5V_2R^&(,UC3D-6DT59R#OF7C77/ILUY+;
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MI3IJ?BY,1#GP+DPEK:)7/:K_!5!+`P04``(`" `J7XTH^$B%PY8$``!-$ ``
M$ ```%!#8G5I;&0O>FQI8BYD<W#M5]MNVS@0?0^0?YA5]R$I&LN7(FB-]0*V
M)1M&?8/E;#9%`($F:9L;211T2>S]^AWJ8DOK2[<)BGUIX#CDS/#PS) \#-_!
M2-! AG(9@<&?N2-]'H 5Q4Q(F ;R+TXCZ F'PPV,B<M;VM^.6&C8FQ+Z1%8<
M)B\>#UJ_??S]\N+=.; ^QS@2<0:=6#@L`?T`/1FX)((_>! *Z<%MI5I5.._?
M@S&!\60.IC&88_?R0IGG[5E__C U0;L77J,.5YM/M]=@;#WB"GHS%-X3#,4B
M(,%6@^JF6JO6U;ANK]]2K)%T.LS@BWAU>?'+R+2L=M^$^5J$@!]/1D#@F3B"
M@4N>^!(I5F N89$PCE28G]4D#H6W@O&H_<7\4("*0XYQ',R-+X,(1AD*4.FZ
MQ&.@?H/8*XPH-!,TT)>0%+F"%+3*\<@'&0,E'H0^IV*Y1=94>DNQBK'"JHXO
M:^ZI>;P=R<+@Q188LDI\BJM+<-$`JP0X4AERLH[P,']<(N ;XOH.;_YGW@HN
MW2OEJFO'$:8R#,5"U6DM!>4A+''2<DHD.#5]>9H9=S@)N097"_S#5$[?WBW7
M)^%2UJ\&2_=MAZ^$EQ\G99C.)E-H.XY\F?*@F^1I<)][C'M48/K579!%J:T&
MJL.74M-*OJ&DQ)F2: U:I8*N[G3:HDZ%;_CEQ6@^;+F"9;V9U6T%-&TCPT$/
M0/OU"M?I6H-6ZT054_[);)VV9<)=R.U1KUL@N#,GE;+3(I1S2$(F<>3'D6V(
M`+0]>C%BX$4\<#D3J!+GXN8D6/$,:>\Z9':>5)%/90]S2 (7^B;BKG\3I&P>
MRXLP6'DRX'9ZX-5,A3G^Q;3WN=IBRW0%WD';,-)\<,E ]Z0C5Q+TT1ST^P;H
M_3]!G]1!-W###<:-NI8TQX;9N>NG;1OMQN3>PMX#!O<,T&F.6X8T]I!?10H[
M4-OE<>!1)V9X5K+^M+MKXD?E6:LU#F<N$"J0P-[7X:!C&\/A,4I)JK@C][R*
MJ.Y3M/5QOGH5B4ITW"F0%[47<X"WC$TFQP,`NH,7P\?J9]#9#B$/.NWO6-U&
MO;4(J;H5#M8O\>;4=IZR<3@8?T$(5-2G@_&I#YYXX'&G4:^H<XCW2) U5TQD
M+4PI]*5TD@Z*-'-6F8>P9^+G8>&:.SF.=/B^1>(HAX\%2ZUL0?,`;%(_Z^P*
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M&I$?&ZCQQ3=#*DNI9?\$./)O^E'W7I!3-$O&`>7)TUM!6I.[6==L845'DL4.
M#Y,JNDF[DIQ_$U^!I4&I*6>5]G;OFG\`4$L#!!0``@`(`"E?C2AG$9H*9P0`
M`-P/```3````4$-B=6EL9"]?<V]C:V5T+F1S<.U7;6_B1A#^'BG_8>KKA^1T
MP01.T1TJE0`#0L>;,&F:*I)EO -LL7<MOP3X]YWU"Y@C$%U2M1_:*%%V9V:?
M>69V]UGX``/N!#*4\P@,?$97^AB &<6,2Q@'\D]T(NAP%^$&AK:'=<T*I;/"
M2"/#V'96]@)AM!88U'_Y_.OEQ8=S>%VD.#M"!LV8NRS!_00=&7AV!+]A$'(I
MX*Y4+BN<CQ_!&,%P-(6VT9O2]/)"F:>-27?Z.&Z#]L!%M0)7FR]WUV!LA>UQ
MYZ;/Q0KZ?!;8P5:#\J9\6ZZH=:U.MYX1)][I2@-G\>+RXJ=!VS0;W39,ESP$
M^A4R`AN>;9<S\.P5SHEE":829@GI2(7Y66?BD(L%# >-;^U/!:@X1(I#:&]\
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M0D9QY,>19? `M#UZ,:(G(@P\9)S4XUS<U X6F"'M7<?,SI,J\BGM88Y)T';?
M1.CY-T'*YNEH*WH+(0.T4B%0R0IIOB/;^5JNLWFZ#Q^@81AI2;1QH OIRH4$
M?3 %_:$*>O=WT$<5T TZ>;UAM:(EPZ'1;MYWT[%%=F/T8-+LD8([!NA.CGL(
M:>PA_^ I;$\=FJ>><-R8T;W)YN/6<9I"]G,9DTKHV.W3%G&\5;3U73ZKE(F'
M),=]GZQK==IR@/>L39+3*0?=I2?A<_DKZ&R'D >=]C?-5K52GX6.>@R.MB?Q
MYM1VGD-COS?\1A"DHJNC]:D/5A@(=*N5$I6BGH\@&RX8ST944NA+Z283$F;F
M+C*/S9YM/P\+E^CF.-+%_<B.HQP^YBRULIF3!]#0\;/)KM%A/ NW(9WP&F5G
M<AU29UR7FFX[2WH2:KWJE[N\FJR0`O>#FOZ50M;J.GY?E!+MFE;>W.*M4:8?
M[72=3$G$8;ETR.*HII7T_(7SMXP0,+G@-7I+H\RA^F+>-Z>31FN:-T?WV:PF
MI$B5MMTWVZ^J;?Y,OEEK;T]K;8Y]7FE?BGJ/SMZ^0V>3#?D'5)9EFN@5=)$=
M"=[;Y/8$]GG-M?XFS;7>H;G6NS77>D5SK?\U]\>TB/2$M@)K(?K1?UR(+99)
M<;$G;Q3FH4&Z7/R4GHI(:CGXT/W"I^)3$7L=33%-&0<.)E^"%; YNI^TVG6Z
M\P/)8A?#IW2]E\Q*R;5NTW>Q@V6I*6>7SG;?*/X"4$L#!!0``@`(`"E?C2CM
M"/Z*) 4``(X>```0````4$-B=6EL9"]?<W)E+F1S<.U9:V_B.!3]7JG_X4YF
M/TQ'4TCIJ)I!RTJ4`(N&EPK=;E>5(F,;\)+841[3\N_W.B&0M$#?[:S:JA5^
M7!^?>Z\YMW$^0D=07P5J'(+%?W)'>=R'01@QH:#OJW\Y#:$A' [[T"4NKQAV
MX',#>WU"9V3"H7<IN5_Y_>L?NSL?MX$U.=J1D#,XCH3#8M OT%"^2T+XB_N!
M4!*."J:I<3Y_!JL'W=X0ZE9KB-W='3T\K)XTA^?].AAG0AZ6X-/5MZ,]L.:2
MN(+NMX6<05N,?.+/#3"OS .SM'95M=W_L[IEW9%>M[M3:S0KVEMT-EE8=;PI
M0<]&T61WYT.G/AA4FW483D4`^"M5" 1^$D<P<,F,C]'!`@P5C&)_0VWF+2(:
M!4).H-NI_JA_R4!%`4<[#O4K3_DA=!8H0)7K$LE __F1S*S(-&,T*(XA3E$!
M*1B%]9;G*@)*) 0>IV(\1]94R;&81)@?G87+*9=Z'[DDF5D\F@-#5O&<YNH2
M3#E@K !7ZH&4K",D^H\)!GY%7,_AY3OSUG#)25L7>V,]3E\%@1CI:$V5H#R 
M,6Z==XSXFTCD-SOA#B<!'O-/(_Q@VK/;3]S>1KB$]1.!90.Q!G+;V;X-=(O;
M=X5-OJ?'?")D*A]ZH'_2Z^,FCKKL<[\6Y\3B'I>,2RHP5>;2:$"IK1=JL4E(
M&KFYMJ+$Z9-P"D:A8.CM/K0:`,9OG_#,[!E0J6S(9<(LQCFN#NIP&G"[TZAE
MMEX.Q\&U$_?R[&*37A1Z46A;P@=CA9ZU:,F0^RYG`O5NF]V0^!.^0%I-W62V
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M=84YD\4,83^)EYH%4^ZD.,KAJQ:)PA0^$BP992.:&F"3>HO.,NA!- KF`7Y-
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M!!0``@`(`'B^C"C9*"Q'A 4``,DB```4````4$-B=6EL9"]?=&MI;G1E<BYD
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M6,P$F\SH&#=9`IW#*-IS(-S<)+*ASYP)J+W&E]:''%3H4_2CT)J[W N@EZ" 
MP6V;.":('R]T<BMRPP@-RF/(4E5"&E)IO?<W'H)!'/!=:K#Q`ID;W!FS28AY
M$MEXF%)'W,O)B.86CQ9@(K-H3O"U":8>,&: *X4A)6PQ!V. B08Z)[9KT=H/
M<1>0RY.W+@_2>CR-^SX;B<A-.3.H#V.D4-P@\3:1V7Y#.!L1'T\R0NQSIM[M
M`7Q++8J8AX7>R';S<[,=< O+_2#CY_B:3IB3EAAAT&[[&@;"X@\:]9I1CA3J
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MMYFY&FJ<6-+/@]NS8.%:;%21<2<<)/6NB]8'\<"D`$];>SMHUCUCA0B:H6RA
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MVN\C^RPQ_Y+8/_C<F"7&`^?A<5D]1EI6,X)JV>H.6OLJ9OK!X!F:*6_6S"7Z
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MP8RXKC@P_\\H"5-ZW..K[/WQOU!+`P04``(`" "I8(TH>7W1X18#``#N" ``
M$0```%!#+V1L;&)A<V5?;G0N='AT=99-;]0P$(;OD?(?1KT`4EM(TD*1$*AE
MBX0$`@DDQ EYX]G$U+$MVVF[_'K&R6Z<;+WI)77F\;SSX?%^5O!+J*H\A=67
M+\\<,(MP8BR>2:'ND)] [X1J@($S6(N-0 YKYA 8YQ:=.\^S7RTJ\"V&#4 X
MD)IQY*? %##OL3,^K':T"%Z#D:Q&$)Z^Y9EOF9]V`OB\@?E*P)BTR/@6A"(A
M2)N"PH>%@CPC,X<2:S]ZY9,8UZ__TG)PNQ&/O0D^KHVA$)67V]-@YS#/QF]#
MY/=HM^"D?A@W<J)1%'/-E >#=J-MQQ3);YA0#F@9UL0/H:VW.Q%CMAJMGR0J
MSWZVI)7KNN](00B/E#$I=<T\+JQQ^%1KDO1]ZUNM(,^"H-T_Y]]_KR@DB>XT
MV#7BGDC20NFV4+=,-0AZ`]IXT3$YEWZ^$S%W%H081DYYV"PD;ZB]W94^++R\
MN?YQFV>U[KH@@[XCN ?AZW:(*_R-TLZYE'#XG$&!Q:OPC*]7P^OSL^)%GOWP
MM"&S'&X?/2HG*+JOFO<46K0,4#GR<'*29[2P=IRO(?4LJ.+J:D[]\7="A1RE
MJ+>1N@ZO`T'-XE+V,)F-Q,U$_),B+6PR&XF/$_$@E,7F"/$Q$JN)^.-T?8<^
M3:PB<3LCJ)6.J;J-Q*>)H$J;H\2G&1&>T !?0[>Z?3EQ7TZJXVI6_.%ME]LM
M/AJ6C&)&K(IB(L83!FFBB$0Y$;T2M:9QQ#Q[2I21J.;5<+JG)D_YJ")Q,1)Y
M]HT.B(WQ0C>V[WZSJF1&I))8QJ24Y=Q]55KFDD09B6I)R'_)0I55)"XF@J9"
MNJ<GLY&X7#1#F#=)XC(2KY>JC#"8(EY'XLU$L'L)QQIN-#LD^+IEKCU"S$;'
MVZ4JE3HY>[.1N%X2#NO>"K\]).(`*&\.(K<T:!(^X@`HZXEXIQRU2+@:Z*ZC
M&:P_?'@_$3P2N/31].F^PDALX@"@FZ )XSQ%;":BBL>SZZ47S%JV?4I4L7>K
M8JFJ16D2/JHB$@?=7DMAUCH,C241N[VJ]H=MO&7H%POA`YS*P8!?1(F7K_;X
M?U!+`P04``(`" `J6HTH9D AC-,%```)#P``"P```%!#+V-O;F9I9RYCI5??
MCYM&$'ZN)?\/H\M+<O+=-:GRD%9](#9W1L7@`JX5596%86U6!18MBR]6=/][
M9W;A;!S\$-4Z<;O[S7XS.[]8'F[A[O8.IOIY^V._\6@JJJ/D^TS!^T^?WM_A
MXR-LCQ JGF2*EWM8Q"IC1:QXG60P9:6233$!JZ@5DVE<3,:C*&/@,922>5RF
M]?UX-![!E9^5YQ"0NAH"5C-Y8"F)+YDL>%US48(2T-1L`@D:-H%"I'R'_Y$8
M4EXKR;>-8J R7D,M=NHYEDR#7-7C42J2ID 3T5IDV@F)T!&J1E:B-F+/7&6B
M4;!C#) ";69XVKV,2\52/$LEQ8&G+$4-L<('[MJ*`]/6&#>5`EV#RU7%8@F\
MA!B/A#!GM=9 &\>CK5"9X?A^IQ9"Y=7IT)>DXU'=5)60.@*]4TU>E6CKRKA 
MO6)W)6 87Q,Q0%],UX[^)] =\N0B3P_C7 <DEAAEI^2*X^J!J"5R>(%#)L*6
M46Q2?>KTP-"\FC0B2=5L<YYP=:13J9B7M*X$AJ0+&JE#.\GHU\!UT:@KEO =
M3R;P++E2K(1*<M)\<I'.JG7&<Z;/@>XC)JXMS9&QD>@_.DXO-=!;)H4X6HT&
M$]-XA )%G**[#S'/XRU28@X0W0^Y9CQZ2XYY!VTH'$PA63*%KDDEJVO8J>K7
MAP=\WE='/&=Y+^1>'R.,G.D\<KPG6%C1W,:'$T[G,+6]*%@MP/)FH%T^PV77
M<G#%=6%M!8'E18X=PMJ)YN-18#]9P0PB'Z*Y$T+H/T8H8T_ \:;N:D;TM,]9
M+%W'GIWO]Q_'HX4=3.<XMSX[KA-]T4H?G<BSPY 8P//!_@L-@G!.+,,F8VJT
M-ON!,?FS#:YC?79M>,0ER_L"X=*>.I9+I#,GL*>1EO6]T/YSA9L1&H]FUL)Z
M(L/,EFZZGEM1Z*,9`01VN'*U_L? 7X#KAW0,6(5XWID56;@5NTC@XQ'0_O7<
M1B,#.H>%?]/(\3T21[51@-,)>/:3ZSS9WM0FI3Y)8R/S`Y1<A>T.['*!$Y).
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MX 61?]71/3%TU=OR=-.^&EW%K80>7\"ZE#M<3RX$Q$F#'O=A4]4M;B87`EG<
MH5E\`5%Y=R"-OX,Q%T\X3OH"5.DM3,.SK*/WWB::![8UZVW0':#;HB?#2??:
M#[J<ZN87:6<:0R=D9GT1Z@\M3L.>C;I3_(3YHYM#ET)Z0G*(M+VAA=K9N<E&
MRG2'3LK,2.IJJ_AP:A6=J=T[V;QOX2;7MVM>WNA+,-Y'=0^Y3RYRL%U'W?T^
M8]1?DK>DRG!V[^T+2EQ^+<5J@ >OND ?#?J#!B_I](VD;\/UL;YO+P8;<XR-
M^?SH\V\V!=XW-AM4XJU<MQ^O37>UN(*CCK/UGF4A&L5+EO?5_3R!GU_&HQ=L
MK_\!4$L#!!0``@`(`!=BC2A,;&0"H ```-(````9````;&EB+W1E<W0O=&5S
M=%]W:6YS;W5N9"YP>3V.30N"0!"&[PO['][LHB!B000=.W7NTEG<40=T9]L/
MQ'_?&M0<AH%YWH\CGF2X3[.D,&\(O+B9$"E$R( X$5:V09(U6,2D_!O$X\76
MR!H:K;3*"O'QCVFU`PRV\)T=J3RU;8USN^]\5C>MD.>'-W<B5W*-ZZ72RGFV
M$<5#' UISGTV29BH\P9!%L)7$K)W[ZD+;,<]9O#T3F3[[5#L?3Y02P,$% `"
M``@`-V*-*/!TJ-U%````20```!T```!L:6(O=&5S="]O=71P=70O=&5S=%]W
M:6YS;W5N9!W)80J ( P&T/^"=UBWZ18A^EF";;4YPML7_7OP!FQL3V,3YQ+#
M*A>J]SYIBM.!I(5,3M#_1HVS(EGC_2-5Q>W@/)<87E!+`0(4"Q0``@`(`"E?
MC2AC)S `5 4``%,@```1``````````$`( "V@0````!00V)U:6QD+V)S9&1B
M+F1S<%!+`0(4"Q0``@`(`"E?C2@D@6@<200``&$/```0``````````$`( "V
M@8,%``!00V)U:6QD+VUM87 N9'-P4$L!`A0+% `"``@`*5^-*.B!37$J!0``
MXQX``!(``````````0`@`+:!^@D``%!#8G5I;&0O<&%R<V5R+F1S<%!+`0(4
M"Q0``@`(`"E?C2AT$O+ 2@4``) ?```3``````````$`( "V@50/``!00V)U
M:6QD+W!Y97AP870N9'-P4$L!`A0+% `"``@`U5V-*.9!"#ZX! ``TQH``!(`
M`````````0`@`+:!SQ0``%!#8G5I;&0O<'ET:&]N+F1S<%!+`0(4"Q0``@`(
M`"E?C2A%\*A&L @``-".```4``````````$`( "V@;<9``!00V)U:6QD+W!Y
M=&AO;C$V+F1S<%!+`0(4"Q0``@`(`,%=C2CN!U/ X 0``#\=```3````````
M``$`( "V@9DB``!00V)U:6QD+W!Y=&AO;G<N9'-P4$L!`A0+% `"``@`*E^-
M*&;$G5J#! ``-1 ``!(``````````0`@`+:!JB<``%!#8G5I;&0O<V5L96-T
M+F1S<%!+`0(4"Q0``@`(`"I?C2AL0OC1<00``(\0```7``````````$`( "V
M@5TL``!00V)U:6QD+W5N:6-O9&5D871A+F1S<%!+`0(4"Q0``@`(`"I?C2@\
MITX:3@0``.8.```4``````````$`( "V@0,Q``!00V)U:6QD+W=I;G-O=6YD
M+F1S<%!+`0(4"Q0``@`(`"I?C2CX2(7#E@0``$T0```0``````````$`( "V
M@8,U``!00V)U:6QD+WIL:6(N9'-P4$L!`A0+% `"``@`*5^-*&<1F@IG! ``
MW \``!,``````````0`@`+:!1SH``%!#8G5I;&0O7W-O8VME="YD<W!02P$"
M% L4``(`" `I7XTH[0C^BB0%``".'@``$ `````````!`" `MH'?/@``4$-B
M=6EL9"]?<W)E+F1S<%!+`0(4"Q0``@`(`'B^C"C9*"Q'A 4``,DB```4````
M``````$`( "V@3%$``!00V)U:6QD+U]T:VEN=&5R+F1S<%!+`0(4"Q0``@`(
M`*E@C2AY?='A%@,``.X(```1``````````$`( "V@>=)``!00R]D;&QB87-E
M7VYT+G1X=%!+`0(4"Q0``@`(`"I:C2AF0"&,TP4```D/```+``````````$`
M( "V@2Q-``!00R]C;VYF:6<N8U!+`0(4"Q0``@`(`!=BC2A,;&0"H ```-(`
M```9``````````$`( "V@2A3``!L:6(O=&5S="]T97-T7W=I;G-O=6YD+G!Y
M4$L!`A0+% `"``@`-V*-*/!TJ-U%````20```!T``````````0`@`+:!_U,`
M`&QI8B]T97-T+V]U='!U="]T97-T7W=I;G-O=6YD4$L%!@`````2`!(`D00`
'`']4````````
`
end



From anthony@interlink.com.au  Thu Apr 13 04:25:25 2000
From: anthony@interlink.com.au (Anthony Baxter)
Date: Thu, 13 Apr 2000 13:25:25 +1000
Subject: [Patches] RTSP support for urlparse.py
Message-ID: <200004130325.NAA06753@mbuna.arbhome.com.au>

The following adds support for RTSP (RFC2326) URLs to the standard
urlparse.py module. 

(i'm using this mod in our unified messaging voice/fax/&c mail gateway)

Thanks,
Anthony


*** /opt/python152/lib/python1.5/urlparse.py	Tue May 25 22:26:51 1999
--- urlparse.py	Thu Apr 13 13:16:14 2000
***************
*** 8,27 ****
  # A classification of schemes ('' means apply by default)
  uses_relative = ['ftp', 'http', 'gopher', 'nntp', 'wais', 'file',
  		 'https', 'shttp',
! 		 'prospero', '']
  uses_netloc = ['ftp', 'http', 'gopher', 'nntp', 'telnet', 'wais',
  	       'file',
  	       'https', 'shttp', 'snews',
! 	       'prospero', '']
  non_hierarchical = ['gopher', 'hdl', 'mailto', 'news', 'telnet', 'wais',
  		    'snews',
  		    ]
  uses_params = ['ftp', 'hdl', 'prospero', 'http',
! 	       'https', 'shttp',
  	       '']
  uses_query = ['http', 'wais',
  	      'https', 'shttp',
! 	      'gopher',
  	      '']
  uses_fragment = ['ftp', 'hdl', 'http', 'gopher', 'news', 'nntp', 'wais',
  		 'https', 'shttp', 'snews',
--- 8,27 ----
  # A classification of schemes ('' means apply by default)
  uses_relative = ['ftp', 'http', 'gopher', 'nntp', 'wais', 'file',
  		 'https', 'shttp', 
! 		 'prospero', 'rtsp', '']
  uses_netloc = ['ftp', 'http', 'gopher', 'nntp', 'telnet', 'wais',
  	       'file',
  	       'https', 'shttp', 'snews',
! 	       'prospero', 'rtsp', '']
  non_hierarchical = ['gopher', 'hdl', 'mailto', 'news', 'telnet', 'wais',
  		    'snews',
  		    ]
  uses_params = ['ftp', 'hdl', 'prospero', 'http',
! 	       'https', 'shttp', 'rtsp',
  	       '']
  uses_query = ['http', 'wais',
  	      'https', 'shttp',
! 	      'gopher', 'rtsp',
  	      '']
  uses_fragment = ['ftp', 'hdl', 'http', 'gopher', 'news', 'nntp', 'wais',
  		 'https', 'shttp', 'snews', 




I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.




From mal@lemburg.com  Thu Apr 13 12:49:05 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Thu, 13 Apr 2000 13:49:05 +0200
Subject: [Patches] Unicode Patch Set 2000-04-13
Message-ID: <38F5B431.39FB76EA@lemburg.com>

This is a multi-part message in MIME format.
--------------8747A0C814C2A18EFA16F3DA
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

The attached patch set contains the following fixes:

Patch Set Contents:
-------------------

Lib/test/output/test_unicode:

Added test output for Unicode string concatenation test.

Python/compile.c:

Fixed problem with Unicode string concatenation:
u = (u"abc" u"abc") previously dumped core.

Lib/codecs.py:

Added more documentation. Clarified some existing comments.

Lib/test/test_unicode.py:

Added test for Unicode string concatenation.

Misc/unicode.txt:

Updated to version 1.4.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------8747A0C814C2A18EFA16F3DA
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-13.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-13.patch"

diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Lib/test/output/test_unicode Python+Unicode/Lib/test/output/test_unicode
--- CVS-Python/Lib/test/output/test_unicode	Wed Apr  5 22:11:19 2000
+++ Python+Unicode/Lib/test/output/test_unicode	Thu Apr 13 13:40:53 2000
@@ -4,3 +4,4 @@
 Testing Unicode formatting strings... done.
 Testing builtin codecs... done.
 Testing standard mapping codecs... 0-127... 128-255... done.
+Testing Unicode string concatenation... done.
Only in CVS-Python/Lib/test/output: test_zipfile
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Python/compile.c Python+Unicode/Python/compile.c
--- CVS-Python/Python/compile.c	Tue Apr 11 11:26:15 2000
+++ Python+Unicode/Python/compile.c	Wed Apr 12 18:05:29 2000
@@ -984,11 +984,32 @@
 	REQ(CHILD(n, 0), STRING);
 	if ((v = parsestr(STR(CHILD(n, 0)))) != NULL) {
 		/* String literal concatenation */
-		for (i = 1; i < NCH(n) && v != NULL; i++) {
-			PyString_ConcatAndDel(&v, parsestr(STR(CHILD(n, i))));
+		for (i = 1; i < NCH(n); i++) {
+		    PyObject *s;
+		    s = parsestr(STR(CHILD(n, i)));
+		    if (s == NULL)
+			goto onError;
+		    if (PyString_Check(v) && PyString_Check(s)) {
+			PyString_ConcatAndDel(&v, s);
+			if (v == NULL)
+			    goto onError;
+		    }
+		    else {
+			PyObject *temp;
+			temp = PyUnicode_Concat(v, s);
+			Py_DECREF(s);
+			if (temp == NULL)
+			    goto onError;
+			Py_DECREF(v);
+			v = temp;
+		    }
 		}
 	}
 	return v;
+
+ onError:
+	Py_XDECREF(v);
+	return NULL;
 }
 
 static void
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Lib/codecs.py Python+Unicode/Lib/codecs.py
--- CVS-Python/Lib/codecs.py	Thu Apr 13 11:11:33 2000
+++ Python+Unicode/Lib/codecs.py	Thu Apr 13 13:40:21 2000
@@ -231,10 +231,13 @@
         """ Read one line from the input stream and return the
             decoded data.
 
-            Note: Unlike the .readlines() method, line breaking must
-            be implemented by the underlying stream's .readline()
-            method -- there is currently no support for line breaking
-            using the codec decoder due to lack of line buffering.
+            Note: Unlike the .readlines() method, this method inherits
+            the line breaking knowledge from the underlying stream's
+            .readline() method -- there is currently no support for
+            line breaking using the codec decoder due to lack of line
+            buffering. Sublcasses should however, if possible, try to
+            implement this method using their own knowledge of line
+            breaking.
 
             size, if given, is passed as size argument to the stream's
             .readline() method.
@@ -288,6 +291,14 @@
 
 class StreamReaderWriter:
 
+    """ StreamReaderWriter instances allow wrapping streams which
+        work in both read and write modes.
+
+        The design is such that one can use the factory functions
+        returned by the codec.lookup() function to contruct the
+        instance.
+
+    """
     # Optional attributes set by the file wrappers below
     encoding = 'unknown'
 
@@ -346,6 +357,21 @@
 
 class StreamRecoder:
 
+    """ StreamRecoder instances provide a frontend - backend
+        view of encoding data.
+
+        They use the complete set of APIs returned by the
+        codecs.lookup() function to implement their task.
+
+        Data written to the stream is first decoded into an
+        intermediate format (which is dependent on the given codec
+        combination) and then written to the stream using an instance
+        of the provided Writer class.
+
+        In the other direction, data is read from the stream using a
+        Reader instance and then return encoded data to the caller.
+
+    """
     # Optional attributes set by the file wrappers below
     data_encoding = 'unknown'
     file_encoding = 'unknown'
@@ -452,6 +478,11 @@
         buffering has the same meaning as for the builtin open() API.
         It defaults to line buffered.
 
+        The returned wrapped file object provides an extra attribute
+        .encoding which allows querying the used encoding. This
+        attribute is only available if an encoding was specified as
+        parameter.
+
     """
     if encoding is not None and \
        'b' not in mode:
@@ -487,6 +518,11 @@
 
         data_encoding and file_encoding are added to the wrapped file
         object as attributes .data_encoding and .file_encoding resp.
+
+        The returned wrapped file object provides two extra attributes
+        .data_encoding and .file_encoding which reflect the given
+        parameters of the same name. The attributes can be used for
+        introspection by Python programs.
 
     """
     if file_encoding is None:
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Lib/test/test_unicode.py Python+Unicode/Lib/test/test_unicode.py
--- CVS-Python/Lib/test/test_unicode.py	Thu Apr 13 11:11:37 2000
+++ Python+Unicode/Lib/test/test_unicode.py	Wed Apr 12 18:38:30 2000
@@ -391,3 +391,11 @@
         print '*** codec for "%s" failed: %s' % (encoding, why)
 
 print 'done.'
+
+print 'Testing Unicode string concatenation...',
+assert (u"abc" u"def") == u"abcdef"
+assert ("abc" u"def") == u"abcdef"
+assert (u"abc" "def") == u"abcdef"
+assert (u"abc" u"def" "ghi") == u"abcdefghi"
+assert ("abc" "def" u"ghi") == u"abcdefghi"
+print 'done.'
diff -u -rP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.c -x *.h -x *.in -x output CVS-Python/Misc/unicode.txt Python+Unicode/Misc/unicode.txt
--- CVS-Python/Misc/unicode.txt	Tue Apr 11 11:25:53 2000
+++ Python+Unicode/Misc/unicode.txt	Thu Apr 13 13:40:51 2000
@@ -1,5 +1,5 @@
 =============================================================================
- Python Unicode Integration                            Proposal Version: 1.3
+ Python Unicode Integration                            Proposal Version: 1.4
 -----------------------------------------------------------------------------
 
 
@@ -162,6 +162,17 @@
 For the same reason, Unicode objects should return the same hash value
 as their UTF-8 equivalent strings.
 
+When compared using cmp() (or PyObject_Compare()) the implementation
+should mask TypeErrors raised during the conversion to remain in synch
+with the string behavior. All other errors such as ValueErrors raised
+during coercion of strings to Unicode should not be masked and passed
+through to the user.
+
+In containment tests ('a' in u'abc' and u'a' in 'abc') both sides
+should be coerced to Unicode before applying the test. Errors occuring
+during coercion (e.g. None in u'abc') should not be masked.
+
+
 Coercion:
 ---------
 
@@ -380,6 +391,13 @@
         data, consumed = self.encode(object,self.errors)
         self.stream.write(data)
         
+    def writelines(self, list):
+
+        """ Writes the concatenated list of strings to the stream
+            using .write().
+        """
+        self.write(''.join(list))
+        
     def reset(self):
 
         """ Flushes and resets the codec buffers used for keeping state.
@@ -463,6 +481,47 @@
             else:
                 return object
 
+    def readline(self, size=None):
+
+        """ Read one line from the input stream and return the
+            decoded data.
+
+            Note: Unlike the .readlines() method, this method inherits
+            the line breaking knowledge from the underlying stream's
+            .readline() method -- there is currently no support for
+            line breaking using the codec decoder due to lack of line
+            buffering. Sublcasses should however, if possible, try to
+            implement this method using their own knowledge of line
+            breaking.
+
+            size, if given, is passed as size argument to the stream's
+            .readline() method.
+            
+        """
+        if size is None:
+            line = self.stream.readline()
+        else:
+            line = self.stream.readline(size)
+        return self.decode(line)[0]
+
+    def readlines(self, sizehint=0):
+
+        """ Read all lines available on the input stream
+            and return them as list of lines.
+
+            Line breaks are implemented using the codec's decoder
+            method and are included in the list entries.
+            
+            sizehint, if given, is passed as size argument to the
+            stream's .read() method.
+
+        """
+        if sizehint is None:
+            data = self.stream.read()
+        else:
+            data = self.stream.read(sizehint)
+        return self.decode(data)[0].splitlines(1)
+
     def reset(self):
 
         """ Resets the codec buffers used for keeping state.
@@ -482,9 +541,6 @@
         """
         return getattr(self.stream,name)
 
-XXX What about .readline(), .readlines() ? These could be implemented
-    using .read() as generic functions instead of requiring their
-    implementation by all codecs. Also see Line Breaks.
 
 Stream codec implementors are free to combine the StreamWriter and
 StreamReader interfaces into one class. Even combining all these with
@@ -692,9 +748,10 @@
 are used as format strings, the following interpretations should be in
 effect:
 
-  '%s':                 '%s' does str(u) for Unicode objects embedded
-                        in Python strings, so the output will be
-                        u.encode(<default encoding>)
+  '%s':                 For Unicode objects this will cause coercion of the
+			whole format string to Unicode. Note that
+			you should use a Unicode format string to start
+			with for performance reasons.
 
 In case the format string is an Unicode object, all parameters are coerced
 to Unicode first and then put together and formatted according to the format
@@ -922,6 +979,9 @@
 	Introducing Unicode to ECMAScript --
 	http://www-4.ibm.com/software/developer/library/internationalization-support.html
 
+IANA Character Set Names:
+	ftp://ftp.isi.edu/in-notes/iana/assignments/character-sets
+
 Encodings:
 
     Overview:
@@ -944,6 +1004,12 @@
 
 History of this Proposal:
 -------------------------
+1.4: Added note about mixed type comparisons and contains tests.
+     Changed treating of Unicode objects in format strings (if used
+     with '%s' % u they will now cause the format string to be
+     coerced to Unicode, thus producing a Unicode object on return).
+     Added link to IANA charset names (thanks to Lars Marius Garshol).
+     Added new codec methods .readline(), .readlines() and .writelines().
 1.3: Added new "es" and "es#" parser markers
 1.2: Removed POD about codecs.open()
 1.1: Added note about comparisons and hash values. Added note about

--------------8747A0C814C2A18EFA16F3DA--




From fdrake@acm.org  Thu Apr 13 15:13:42 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Thu, 13 Apr 2000 10:13:42 -0400 (EDT)
Subject: [Patches] Unicode Patch Set 2000-04-13
In-Reply-To: <38F5B431.39FB76EA@lemburg.com>
References: <38F5B431.39FB76EA@lemburg.com>
Message-ID: <14581.54806.244258.685523@seahag.cnri.reston.va.us>

M.-A. Lemburg writes:
 > The attached patch set contains the following fixes:

  I've applied this to the CVS repository; thanks!


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From thomas.heller@ion-tof.com  Thu Apr 13 15:34:11 2000
From: thomas.heller@ion-tof.com (Thomas Heller)
Date: Thu, 13 Apr 2000 16:34:11 +0200
Subject: [Patches] Patch to ihooks.py
Message-ID: <012401bfa555$55b63170$4500a8c0@thomasnotebook>

ihooks.ModuleLoader does not implement reload(mod) correctly:
If mod has already been loaded by ModuleLoader, it has
been returned from a cache. Added an additional parameter
to import_it() to force reloading.


*** ihooks.old.py Thu Apr 13 16:26:06 2000
--- ihooks.py Thu Apr 13 16:27:18 2000
***************
*** 467,480 ****
                  if not submod:
                      raise ImportError, "No module named " + subname
  
!     def import_it(self, partname, fqname, parent):
          if not partname:
              raise ValueError, "Empty module name"
          try:
-             return self.modules[fqname]
-         except KeyError:
-             pass
-         try:
              path = parent and parent.__path__
          except AttributeError:
              return None
--- 467,481 ----
                  if not submod:
                      raise ImportError, "No module named " + subname
  
!     def import_it(self, partname, fqname, parent, force_load=0):
          if not partname:
              raise ValueError, "Empty module name"
+         if not force_load:
+             try:
+                 return self.modules[fqname]
+             except KeyError:
+                 pass
          try:
              path = parent and parent.__path__
          except AttributeError:
              return None
***************
*** 489,499 ****
      def reload(self, module):
          name = module.__name__
          if '.' not in name:
!             return self.import_it(name, name, None)
          i = string.rfind(name, '.')
          pname = name[:i]
          parent = self.modules[pname]
!         return self.import_it(name[i+1:], name, parent)
  
  
  default_importer = None
--- 490,500 ----
      def reload(self, module):
          name = module.__name__
          if '.' not in name:
!             return self.import_it(name, name, None, force_load=1)
          i = string.rfind(name, '.')
          pname = name[:i]
          parent = self.modules[pname]
!         return self.import_it(name[i+1:], name, parent, force_load=1)
  
  
  default_importer = None

Disclaimer:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.





From fdrake@acm.org  Thu Apr 13 16:10:32 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Thu, 13 Apr 2000 11:10:32 -0400 (EDT)
Subject: [Patches] RTSP support for urlparse.py
In-Reply-To: <200004130325.NAA06753@mbuna.arbhome.com.au>
References: <200004130325.NAA06753@mbuna.arbhome.com.au>
Message-ID: <14581.58216.675833.972193@seahag.cnri.reston.va.us>

Anthony Baxter writes:
 > The following adds support for RTSP (RFC2326) URLs to the standard
 > urlparse.py module. 

Anthony,
  Is there a reason you supported the rtsp: scheme and not the rtspu:
scheme?  I'm looking at section 3.2 of RFC 2326, and they look like
the same syntax to me.
  Thanks!


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From jeremy@cnri.reston.va.us  Thu Apr 13 23:06:39 2000
From: jeremy@cnri.reston.va.us (Jeremy Hylton)
Date: Thu, 13 Apr 2000 18:06:39 -0400 (EDT)
Subject: [Patches] Re: Comparison of cyclic objects (was RE: [Python-Dev] trashcan and PR#7)
In-Reply-To: <14582.14222.865019.806313@bitdiddle.cnri.reston.va.us>
References: <14580.48029.512656.911718@goon.cnri.reston.va.us>
 <000701bfa505$31008380$4d2d153f@tim>
 <14582.5791.148277.87450@walden>
 <14582.14222.865019.806313@bitdiddle.cnri.reston.va.us>
Message-ID: <14582.17647.662905.959786@bitdiddle.cnri.reston.va.us>

I did one more round of work on this idea, and I'm satisfied with the
results.  Most of the performance hit can be eliminated by doing
nothing until there are at least N recursive calls to
PyObject_Compare, where N is fairly large. (I picked 25000.)
Non-circular objects that are not deeply nested only pay for an
integer increment, a decrement, and a compare.

Background for patches-only readers: This patch appears to fix PR#7.

Comments and suggestions solicitied.  I think this is worth checking
in. 

Jeremy

Index: Include/object.h
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Include/object.h,v
retrieving revision 2.52
diff -r2.52 object.h
286a287,289
> /* tstate dict key for PyObject_Compare helper */
> extern PyObject *_PyCompareState_Key;
> 
Index: Python/pythonrun.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/pythonrun.c,v
retrieving revision 2.91
diff -r2.91 pythonrun.c
151a152,153
> 	_PyCompareState_Key = PyString_InternFromString("cmp_state");
> 
Index: Objects/object.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/object.c,v
retrieving revision 2.67
diff -r2.67 object.c
300a301,306
> PyObject *_PyCompareState_Key;
> 
> int _PyCompareState_nesting = 0;
> int _PyCompareState_flag = 0;
> #define NESTING_LIMIT 25000
> 
305a312,313
> 	int result;
> 
372c380
< 	if (vtp->tp_compare == NULL)
---
> 	if (vtp->tp_compare == NULL) {
374c382,440
< 	return (*vtp->tp_compare)(v, w);
---
> 	}
> 	++_PyCompareState_nesting;
> 	if (_PyCompareState_nesting > NESTING_LIMIT)
> 		_PyCompareState_flag = 1;
> 	if (_PyCompareState_flag &&
> 	    (vtp->tp_as_mapping || (vtp->tp_as_sequence &&
> 				   !PyString_Check(v))))
> 	{
> 		PyObject *tstate_dict, *cmp_dict, *pair;
> 
> 		tstate_dict = PyThreadState_GetDict();
> 		if (tstate_dict == NULL) {
> 			PyErr_BadInternalCall();
> 			return -1;
> 		} 
> 		cmp_dict = PyDict_GetItem(tstate_dict, _PyCompareState_Key);
> 		if (cmp_dict == NULL) {
> 			cmp_dict = PyDict_New();
> 			if (cmp_dict == NULL)
> 				return -1;
> 			PyDict_SetItem(tstate_dict,
> 				       _PyCompareState_Key,
> 				       cmp_dict);  
> 		}
> 
> 		pair = PyTuple_New(2);
> 		if (pair == NULL) {
> 			return -1;
> 		}
> 		if ((long)v <= (long)w) {
> 			PyTuple_SET_ITEM(pair, 0, PyInt_FromLong((long)v));
> 			PyTuple_SET_ITEM(pair, 1, PyInt_FromLong((long)w));
> 		} else {
> 			PyTuple_SET_ITEM(pair, 0, PyInt_FromLong((long)w));
> 			PyTuple_SET_ITEM(pair, 1, PyInt_FromLong((long)v));
> 		}
> 		if (PyDict_GetItem(cmp_dict, pair)) {
> 			/* already comparing these objects.  assume
> 			   they're equal until shown otherwise 
> 			*/
> 			Py_DECREF(pair);
> 			--_PyCompareState_nesting;
> 			if (_PyCompareState_nesting == 0)
> 				_PyCompareState_flag = 0;
> 			return 0;
> 		}
> 		if (PyDict_SetItem(cmp_dict, pair, pair) == -1) {
> 			return -1;
> 		}
> 		result = (*vtp->tp_compare)(v, w);
> 		PyDict_DelItem(cmp_dict, pair);
> 		Py_DECREF(pair);
> 	} else {
> 		result = (*vtp->tp_compare)(v, w);
> 	}
> 	--_PyCompareState_nesting;
> 	if (_PyCompareState_nesting == 0)
> 		_PyCompareState_flag = 0;
> 	return result;




From bwarsaw@cnri.reston.va.us  Fri Apr 14 00:14:49 2000
From: bwarsaw@cnri.reston.va.us (Barry A. Warsaw)
Date: Thu, 13 Apr 2000 19:14:49 -0400 (EDT)
Subject: [Patches] Re: Comparison of cyclic objects (was RE: [Python-Dev] trashcan and PR#7)
References: <14580.48029.512656.911718@goon.cnri.reston.va.us>
 <000701bfa505$31008380$4d2d153f@tim>
 <14582.5791.148277.87450@walden>
 <14582.14222.865019.806313@bitdiddle.cnri.reston.va.us>
 <14582.17647.662905.959786@bitdiddle.cnri.reston.va.us>
Message-ID: <14582.21737.387268.332139@anthem.cnri.reston.va.us>

    JH> Comments and suggestions solicitied.  I think this is worth
    JH> checking in.

Please regenerate with unified or context diffs!

-Barry


From jeremy@cnri.reston.va.us  Fri Apr 14 00:30:02 2000
From: jeremy@cnri.reston.va.us (Jeremy Hylton)
Date: Thu, 13 Apr 2000 19:30:02 -0400 (EDT)
Subject: [Patches] Re: Comparison of cyclic objects (was RE: [Python-Dev] trashcan and PR#7)
In-Reply-To: <14582.21737.387268.332139@anthem.cnri.reston.va.us>
References: <14580.48029.512656.911718@goon.cnri.reston.va.us>
 <000701bfa505$31008380$4d2d153f@tim>
 <14582.5791.148277.87450@walden>
 <14582.14222.865019.806313@bitdiddle.cnri.reston.va.us>
 <14582.17647.662905.959786@bitdiddle.cnri.reston.va.us>
 <14582.21737.387268.332139@anthem.cnri.reston.va.us>
Message-ID: <14582.22650.792191.474554@bitdiddle.cnri.reston.va.us>

Here it is contextified.  One small difference from the previous patch
is that NESTING_LIMIT is now only 1000.  I think this is sufficient to
cover commonly occuring nested containers.

Jeremy

Index: Include/object.h
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Include/object.h,v
retrieving revision 2.52
diff -c -r2.52 object.h
*** object.h	2000/03/21 16:14:47	2.52
--- object.h	2000/04/13 21:50:10
***************
*** 284,289 ****
--- 284,292 ----
  extern DL_IMPORT(int) Py_ReprEnter Py_PROTO((PyObject *));
  extern DL_IMPORT(void) Py_ReprLeave Py_PROTO((PyObject *));
  
+ /* tstate dict key for PyObject_Compare helper */
+ extern PyObject *_PyCompareState_Key;
+ 
  /* Flag bits for printing: */
  #define Py_PRINT_RAW	1	/* No string quotes etc. */
  
Index: Python/pythonrun.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/pythonrun.c,v
retrieving revision 2.91
diff -c -r2.91 pythonrun.c
*** pythonrun.c	2000/03/10 23:03:54	2.91
--- pythonrun.c	2000/04/13 21:50:25
***************
*** 149,154 ****
--- 149,156 ----
  	/* Init Unicode implementation; relies on the codec registry */
  	_PyUnicode_Init();
  
+ 	_PyCompareState_Key = PyString_InternFromString("cmp_state");
+ 
  	bimod = _PyBuiltin_Init_1();
  	if (bimod == NULL)
  		Py_FatalError("Py_Initialize: can't initialize __builtin__");
Index: Objects/object.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/object.c,v
retrieving revision 2.67
diff -c -r2.67 object.c
*** object.c	2000/04/10 13:42:33	2.67
--- object.c	2000/04/13 21:44:42
***************
*** 298,308 ****
--- 298,316 ----
  	return PyInt_FromLong(c);
  }
  
+ PyObject *_PyCompareState_Key;
+ 
+ int _PyCompareState_nesting = 0;
+ int _PyCompareState_flag = 0;
+ #define NESTING_LIMIT 1000
+ 
  int
  PyObject_Compare(v, w)
  	PyObject *v, *w;
  {
  	PyTypeObject *vtp, *wtp;
+ 	int result;
+ 
  	if (v == NULL || w == NULL) {
  		PyErr_BadInternalCall();
  		return -1;
***************
*** 369,377 ****
  		/* Numerical types compare smaller than all other types */
  		return strcmp(vname, wname);
  	}
! 	if (vtp->tp_compare == NULL)
  		return (v < w) ? -1 : 1;
! 	return (*vtp->tp_compare)(v, w);
  }
  
  long
--- 377,443 ----
  		/* Numerical types compare smaller than all other types */
  		return strcmp(vname, wname);
  	}
! 	if (vtp->tp_compare == NULL) {
  		return (v < w) ? -1 : 1;
! 	}
! 	++_PyCompareState_nesting;
! 	if (_PyCompareState_nesting > NESTING_LIMIT)
! 		_PyCompareState_flag = 1;
! 	if (_PyCompareState_flag &&
! 	    (vtp->tp_as_mapping || (vtp->tp_as_sequence &&
! 				   !PyString_Check(v))))
! 	{
! 		PyObject *tstate_dict, *cmp_dict, *pair;
! 
! 		tstate_dict = PyThreadState_GetDict();
! 		if (tstate_dict == NULL) {
! 			PyErr_BadInternalCall();
! 			return -1;
! 		} 
! 		cmp_dict = PyDict_GetItem(tstate_dict, _PyCompareState_Key);
! 		if (cmp_dict == NULL) {
! 			cmp_dict = PyDict_New();
! 			if (cmp_dict == NULL)
! 				return -1;
! 			PyDict_SetItem(tstate_dict,
! 				       _PyCompareState_Key,
! 				       cmp_dict);  
! 		}
! 
! 		pair = PyTuple_New(2);
! 		if (pair == NULL) {
! 			return -1;
! 		}
! 		if ((long)v <= (long)w) {
! 			PyTuple_SET_ITEM(pair, 0, PyInt_FromLong((long)v));
! 			PyTuple_SET_ITEM(pair, 1, PyInt_FromLong((long)w));
! 		} else {
! 			PyTuple_SET_ITEM(pair, 0, PyInt_FromLong((long)w));
! 			PyTuple_SET_ITEM(pair, 1, PyInt_FromLong((long)v));
! 		}
! 		if (PyDict_GetItem(cmp_dict, pair)) {
! 			/* already comparing these objects.  assume
! 			   they're equal until shown otherwise 
! 			*/
! 			Py_DECREF(pair);
! 			--_PyCompareState_nesting;
! 			if (_PyCompareState_nesting == 0)
! 				_PyCompareState_flag = 0;
! 			return 0;
! 		}
! 		if (PyDict_SetItem(cmp_dict, pair, pair) == -1) {
! 			return -1;
! 		}
! 		result = (*vtp->tp_compare)(v, w);
! 		PyDict_DelItem(cmp_dict, pair);
! 		Py_DECREF(pair);
! 	} else {
! 		result = (*vtp->tp_compare)(v, w);
! 	}
! 	--_PyCompareState_nesting;
! 	if (_PyCompareState_nesting == 0)
! 		_PyCompareState_flag = 0;
! 	return result;
  }
  
  long



From Anthony Baxter <anthony@interlink.com.au>  Fri Apr 14 01:55:51 2000
From: Anthony Baxter <anthony@interlink.com.au> (Anthony Baxter)
Date: Fri, 14 Apr 2000 10:55:51 +1000
Subject: [Patches] RTSP support for urlparse.py
In-Reply-To: Message from "Fred L. Drake, Jr." <fdrake@acm.org>
 of "Thu, 13 Apr 2000 11:10:32 -0400." <14581.58216.675833.972193@seahag.cnri.reston.va.us>
Message-ID: <200004140055.KAA11267@mbuna.arbhome.com.au>

>>> "Fred L. Drake, Jr." wrote
> Anthony,
>   Is there a reason you supported the rtsp: scheme and not the rtspu:
> scheme?  I'm looking at section 3.2 of RFC 2326, and they look like
> the same syntax to me.
>   Thanks!


Because I didn't need it, initially, and then, later, because I forgot
about it.

Feel free to put rtspu: in as well...

Anthony
-- 
Anthony Baxter     <anthony@interlink.com.au>   
It's never too late to have a happy childhood.



From thomas.heller@ion-tof.com  Fri Apr 14 12:38:55 2000
From: thomas.heller@ion-tof.com (Thomas Heller)
Date: Fri, 14 Apr 2000 13:38:55 +0200
Subject: [Patches] Patches for 1.5.2
Message-ID: <00fb01bfa606$0403dd80$4500a8c0@thomasnotebook>

Dear patchers,

does it make sense to submit patches for 1.5.2
additionally to patches against the CVS version?

Thomas Heller




From skip@mojam.com (Skip Montanaro)  Fri Apr 14 14:04:49 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Fri, 14 Apr 2000 08:04:49 -0500 (CDT)
Subject: [Patches] Patches for 1.5.2
In-Reply-To: <00fb01bfa606$0403dd80$4500a8c0@thomasnotebook>
References: <00fb01bfa606$0403dd80$4500a8c0@thomasnotebook>
Message-ID: <14583.6001.992507.138157@beluga.mojam.com>

    Thomas> does it make sense to submit patches for 1.5.2 additionally to
    Thomas> patches against the CVS version?

In general, no, I don't think it does.  The only place I can see a
1.5.2-only patch being worthwhile at this point is if it plugs a security
hole => if you don't apply the patch you open your system up more widely to
bad guys.

Skip



From fdrake@acm.org  Fri Apr 14 14:29:57 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Fri, 14 Apr 2000 09:29:57 -0400 (EDT)
Subject: [Patches] Patches for 1.5.2
In-Reply-To: <00fb01bfa606$0403dd80$4500a8c0@thomasnotebook>
References: <00fb01bfa606$0403dd80$4500a8c0@thomasnotebook>
Message-ID: <14583.7509.402900.368389@seahag.cnri.reston.va.us>

Thomas Heller writes:
 > does it make sense to submit patches for 1.5.2
 > additionally to patches against the CVS version?

  At the most, it would appear on python.org on the patches list, but
with 1.6 ~6 weeks away, I think dealing with it would cost us too much
of our concentration.  Our time is limited, and we still have paying
jobs.
  It also depends on the severity of the problem; at this late date,
it's hard to imagine it would be very important.
  Others may say differently.


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From fdrake@acm.org  Fri Apr 14 15:02:32 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Fri, 14 Apr 2000 10:02:32 -0400 (EDT)
Subject: [Patches] RTSP support for urlparse.py
In-Reply-To: <200004130325.NAA06753@mbuna.arbhome.com.au>
References: <200004130325.NAA06753@mbuna.arbhome.com.au>
Message-ID: <14583.9464.919175.424474@seahag.cnri.reston.va.us>

Anthony Baxter writes:
 > The following adds support for RTSP (RFC2326) URLs to the standard
 > urlparse.py module. 

Anthony,
  Patch applied, with the changes we discussed.  Thanks!


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From sz@zmail.ru  Fri Apr 14 15:35:52 2000
From: sz@zmail.ru (Zaur Shibzoukhov)
Date: Fri, 14 Apr 2000 18:35:52 +0400
Subject: [Patches] Proposal To Modifying Map Function in 1.6
Message-ID: <web-26748960@backend1.aha.ru>

Proposal To Modifying Map Function in 1.6
-----------------------------------------

map(func, List_1, List_2, ..., List_n) use the following rule now:

    if len(List_i) < max(len(List_1),...,len(List_n)) then
        List_i[Ni]=None, List_i[Ni+1]=None, ...
            Ni=len(List_i)

I propose to change this rule to another:

    if len(List_i) < max(len(List_1),...,len(List_n)) then			
        List_i[j]=List_i[j (mod Ni)], List_i[j]=List_i[j+1 (mod Ni)],
...       (*)
            Ni=len(List_i)

Why?
Main reason: I can't use map without lambda expression in the following
common case:
   map(func, List, arg ) (1) 
or
   map(func, arg, List)  (2)
where
   func -- function with two arguments, arg -- not sequense.
Now I must do
   map(lambda x,y=arg: func(x,y), List) (1')
or
   map(lambda y,x=arg: func(x,y), List) (2')

With rule (*) I can write:
   map(func, List, [arg])

After this change I also can repeat lists of smaller length in map's
arguments.
I think that it can improve functionality of map.

For example:

    Ls=[list_1, ..., list_N]
    map(append, Ls, [a,b]) ->
          [list_1+[a],list_2+[b],list_3+[a],,list_4+[b], ...]

    or

    Ls = [func_1, ..., func_N]
    map(apply, Ls, [args])
        instead of
    map(lambda f,args=args: apply(f, args), Ls)

    and others ...

This modification allow to prevent using lambda expression
in map calls in some cases.

Here is an easy "patch" to map (for source of Python 1.5.2)

static PyObject *
builtin_map(self, args)
        PyObject *self;
        PyObject *args;
{
        typedef struct {
                PyObject *seq;
                PySequenceMethods *sqf;
                int len;
        } sequence;

        PyObject *func, *result;
        sequence *seqs = NULL, *sqp;
        int n, len;
        register int i, j;

        n = PyTuple_Size(args);
        if (n < 2) {
                PyErr_SetString(PyExc_TypeError,
                                "map() requires at least two args");
                return NULL;
        }

        func = PyTuple_GetItem(args, 0);
        n--;

        if (func == Py_None && n == 1) {
                /* map(None, S) is the same as list(S). */
                return PySequence_List(PyTuple_GetItem(args, 1));
        }

        if ((seqs = PyMem_NEW(sequence, n)) == NULL) {
                PyErr_NoMemory();
                goto Fail_2;
        }

        for (len = 0, i = 0, sqp = seqs; i < n; ++i, ++sqp) {
                int curlen;
                PySequenceMethods *sqf;
        
                if ((sqp->seq = PyTuple_GetItem(args, i + 1)) == NULL)
                        goto Fail_2;

                sqp->sqf = sqf = sqp->seq->ob_type->tp_as_sequence;
                if (sqf == NULL ||
                    sqf->sq_length == NULL ||
                    sqf->sq_item == NULL)
                {
                        static char errmsg[] =
                            "argument %d to map() must be a sequence
object";
                        char errbuf[sizeof(errmsg) + 25];

                        sprintf(errbuf, errmsg, i+2);
                        PyErr_SetString(PyExc_TypeError, errbuf);
                        goto Fail_2;
                }

                if ((curlen = sqp->len =
(*sqp->sqf->sq_length)(sqp->seq)) < 0)
                        goto Fail_2;

                if (curlen > len)
                        len = curlen;
        }

        if ((result = (PyObject *) PyList_New(len)) == NULL)
                goto Fail_2;

        for (i = 0; ; ++i) {
                PyObject *alist, *item=NULL, *value;
                int any = 0;

                if (func == Py_None && n == 1)
                        alist = NULL;
                else {
                        if ((alist = PyTuple_New(n)) == NULL)
                                goto Fail_1;
                }

                for (j = 0, sqp = seqs; j < n; ++j, ++sqp) {
                         if (sqp->len < 0) {
                   -            Py_INCREF(Py_None);
                   -            item = Py_None;
                   +            item = (*sqp->sqf->sq_item)(sqp->seq, i
% - sqp->len);
                         }
                         else {
                                item = (*sqp->sqf->sq_item)(sqp->seq,
i);
                                if (item == NULL) {
                                       if (PyErr_ExceptionMatches(
                                               PyExc_IndexError))
                                       {
                                               PyErr_Clear();
                   -                           Py_INCREF(Py_None);
                   -                           item = Py_None;
                   +                           item =
(*sqp->sqf->sq_item)(sqp->seq, i % sqp->len);
                   -                           sqp->len = -1; 
                   +                           sqp->len = - sqp->len; 
                                       }
                                       else {
                                               goto Fail_0;
                                       }
                                }
                                else
                                        any = 1;
                        }
                        if (!alist)
                                break;
                        if (PyTuple_SetItem(alist, j, item) < 0) {
                                Py_DECREF(item);
                                goto Fail_0;
                        }
                        continue;

                Fail_0:
                        Py_XDECREF(alist);
                        goto Fail_1;
                }

                if (!alist)
                        alist = item;

                if (!any) {
                        Py_DECREF(alist);
                        break;
                }

                if (func == Py_None)
                        value = alist;
                else {
                        value = PyEval_CallObject(func, alist);
                        Py_DECREF(alist);
                        if (value == NULL)
                                goto Fail_1;
                }
                if (i >= len) {
                        int status = PyList_Append(result, value);
                        Py_DECREF(value);
                        if (status < 0)
                                goto Fail_1;
                }
                else {
                        if (PyList_SetItem(result, i, value) < 0)
                                goto Fail_1;
                }
        }

        if (i < len && PyList_SetSlice(result, i, len, NULL) < 0)
                goto Fail_1;

        PyMem_DEL(seqs);
        return result;

Fail_1:
        Py_DECREF(result);
Fail_2:
        if (seqs) PyMem_DEL(seqs);
        return NULL;
}


Zaur Shibzoukhov
sz@zmail.ru


        
-----------------------------------------------
Get Your e-mail at http://zmail.ru



From skip@mojam.com (Skip Montanaro)  Fri Apr 14 15:45:46 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Fri, 14 Apr 2000 09:45:46 -0500 (CDT)
Subject: [Patches] Proposal To Modifying Map Function in 1.6
In-Reply-To: <web-26748960@backend1.aha.ru>
References: <web-26748960@backend1.aha.ru>
Message-ID: <14583.12058.856533.820813@beluga.mojam.com>

    Zaur> map(func, List_1, List_2, ..., List_n) use the following rule now:

    Zaur>     if len(List_i) < max(len(List_1),...,len(List_n)) then
    Zaur>         List_i[Ni]=None, List_i[Ni+1]=None, ...
    Zaur>             Ni=len(List_i)

    Zaur> I propose to change this rule to another:

    Zaur>     if len(List_i) < max(len(List_1),...,len(List_n)) then			
    Zaur>         List_i[j]=List_i[j (mod Ni)], List_i[j]=List_i[j+1 (mod Ni)],
    Zaur> ...       (*)
    Zaur>             Ni=len(List_i)

An equally useful rule for a short list is to extend it with the last value
of the list:

    if len(List_i) < max(len(List_1),...,len(List_n)) then
        last = List_i[len(List_i)-1]
        List_i[Ni]=last, List_i[Ni+1]=last, ...
        Ni=len(List_i)

(I suggested this a couple years ago but was rebuffed... ;-)

Given the somewhat arbitrary nature of choosing any of these three rules,
you might as well stay with the status quo to avoid code breakage in the
code that relies on the current semantics.

-- 
Skip Montanaro | http://www.mojam.com/
skip@mojam.com | http://www.musi-cal.com/


From Pekka.Pessi@nokia.com  Fri Apr 14 16:07:03 2000
From: Pekka.Pessi@nokia.com (Pekka Pessi)
Date: Fri, 14 Apr 2000 18:07:03 +0300 (EEST)
Subject: [Patches] Making  modules callable when they have the attribute "__call__"
Message-ID: <14583.13335.672195.314564@rama.research.nokia.com>

	Hello,

	Attached you find patches that make modules callable when they have
	an attribute "__call__".  I think that this makes Python mode
	orthogonal. YMMV.

					Pekka Pessi
	
-----8<-------------8<-------------8<-------------8<-------------8<-----
*** Objects/object.c-virgin	Mon Apr 10 16:42:33 2000
--- Objects/object.c	Fri Apr 14 17:56:20 2000
***************
*** 601,607 ****
  	    PyCFunction_Check(x) ||
  	    PyClass_Check(x))
  		return 1;
! 	if (PyInstance_Check(x)) {
  		PyObject *call = PyObject_GetAttrString(x, "__call__");
  		if (call == NULL) {
  			PyErr_Clear();
--- 601,608 ----
  	    PyCFunction_Check(x) ||
  	    PyClass_Check(x))
  		return 1;
! 	if (PyInstance_Check(x) ||
! 	    PyModule_Check(x)) {
  		PyObject *call = PyObject_GetAttrString(x, "__call__");
  		if (call == NULL) {
  			PyErr_Clear();
*** Python/ceval.c-virgin	Mon Apr 10 15:45:10 2000
--- Python/ceval.c	Fri Apr 14 17:56:16 2000
***************
*** 2430,2436 ****
  	if (PyClass_Check(func)) {
  		return PyInstance_New(func, arg, kw);
  	}
! 	if (PyInstance_Check(func)) {
  	        PyObject *res, *call = PyObject_GetAttrString(func,"__call__");
  		if (call == NULL) {
  			PyErr_Clear();
--- 2430,2437 ----
  	if (PyClass_Check(func)) {
  		return PyInstance_New(func, arg, kw);
  	}
! 	if (PyInstance_Check(func) ||
! 	    PyModule_Check(func)) {
  	        PyObject *res, *call = PyObject_GetAttrString(func,"__call__");
  		if (call == NULL) {
  			PyErr_Clear();
-----8<-------------8<-------------8<-------------8<-------------8<-----



--
	I confirm that, to the best of my knowledge and belief, this
        contribution is free of any claims of third parties under
        copyright, patent or other rights or interests ("claims").  To
        the extent that I have any such claims, I hereby grant to CNRI a
        nonexclusive, irrevocable, royalty-free, worldwide license to
        reproduce, distribute, perform and/or display publicly, prepare
        derivative versions, and otherwise use this contribution as part
        of the Python software and its related documentation, or any
        derivative versions thereof, at no cost to CNRI or its licensed
        users, and to authorize others to do so.
	
        I acknowledge that CNRI may, at its sole discretion, decide
        whether or not to incorporate this contribution in the Python
        software and its related documentation.  I further grant CNRI
        permission to use my name and other identifying information
        provided to CNRI by me for use in connection with the Python
        software and its related documentation.


From mhammond@skippinet.com.au  Fri Apr 14 16:27:17 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Sat, 15 Apr 2000 01:27:17 +1000
Subject: [Patches] Making  modules callable when they have the attribute "__call__"
In-Reply-To: <14583.13335.672195.314564@rama.research.nokia.com>
Message-ID: <ECEPKNMJLHAPFFJHDOJBCEACCJAA.mhammond@skippinet.com.au>

PiAJQXR0YWNoZWQgeW91IGZpbmQgcGF0Y2hlcyB0aGF0IG1ha2UgbW9kdWxlcyANCj4gY2FsbGFi
bGUgd2hlbiB0aGV5IGhhdmUNCj4gCWFuIGF0dHJpYnV0ZSAiX19jYWxsX18iLiAgSSB0aGluayB0
aGF0IHRoaXMgbWFrZXMgDQo+IFB5dGhvbiBtb2RlDQo+IAlvcnRob2dvbmFsLiBZTU1WLg0KDQpX
b3VsZG50IGEgYmV0dGVyIHBhdGNoIGJlIHRvIGZpbGwgb3V0IHRoZSB0cF9jYWxsIHNsb3QgZm9y
IHRoZSBtb2R1bGU/DQoNCk1hcmsuDQo=



From skip@mojam.com (Skip Montanaro)  Fri Apr 14 16:29:07 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Fri, 14 Apr 2000 10:29:07 -0500 (CDT)
Subject: [Patches] Making  modules callable when they have the attribute "__call__"
In-Reply-To: <14583.13335.672195.314564@rama.research.nokia.com>
References: <14583.13335.672195.314564@rama.research.nokia.com>
Message-ID: <14583.14659.40548.466043@beluga.mojam.com>

    Pekka> 	Attached you find patches that make modules callable when
    Pekka> 	they have an attribute "__call__".  I think that this makes
    Pekka> 	Python mode orthogonal. YMMV.

Pekka,

Can you provide some motivation for making modules callable?  It's not
obvious to me that it buys you anything.

-- 
Skip Montanaro | http://www.mojam.com/
skip@mojam.com | http://www.musi-cal.com/


From Pekka.Pessi@nokia.com  Fri Apr 14 17:17:04 2000
From: Pekka.Pessi@nokia.com (Pekka Pessi)
Date: Fri, 14 Apr 2000 19:17:04 +0300 (EEST)
Subject: [Patches] Making  modules callable when they have the attribute "__call__"
In-Reply-To: <14583.14659.40548.466043@beluga.mojam.com>
References: <14583.13335.672195.314564@rama.research.nokia.com> <14583.14659.40548.466043@beluga.mojam.com>
 <14583.14659.40548.466043@beluga.mojam.com>
Message-ID: <14583.17536.445303.206607@rama.research.nokia.com>

Skip Montanaro <skip@mojam.com> writes:
>    Pekka> 	Attached you find patches that make modules callable when
>    Pekka> 	they have an attribute "__call__".  I think that this makes
>    Pekka> 	Python mode orthogonal. YMMV.

>Can you provide some motivation for making modules callable?  It's not
>obvious to me that it buys you anything.

	If we have a simple extension module, say, md5, that basically
	defines one class/object type, say, md5, it is much cleaner to
	define that class/object type also as __call__. Now the single
	object type is sometimes called "new", sometimes something else.

					Pekka


From Pekka.Pessi@nokia.com  Fri Apr 14 17:18:57 2000
From: Pekka.Pessi@nokia.com (Pekka Pessi)
Date: Fri, 14 Apr 2000 19:18:57 +0300 (EEST)
Subject: [Patches] Making  modules callable when they have the attribute "__call__"
In-Reply-To: <ECEPKNMJLHAPFFJHDOJBCEACCJAA.mhammond@skippinet.com.au>
References: <ECEPKNMJLHAPFFJHDOJBCEACCJAA.mhammond@skippinet.com.au>
 <ECEPKNMJLHAPFFJHDOJBCEACCJAA.mhammond@skippinet.com.au>
Message-ID: <14583.17649.904861.454848@rama.research.nokia.com>

Mark Hammond <mhammond@skippinet.com.au> writes:

>> 	Attached you find patches that make modules 
>> callable when they have
>> 	an attribute "__call__".  I think that this makes 
>> Python mode
>> 	orthogonal. YMMV.

>Wouldnt a better patch be to fill out the tp_call slot for the module?

	Nope. Unless you want to make a module without __call__ attribute
	also callable.

						Pekka


From fdrake@acm.org  Fri Apr 14 17:44:33 2000
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Fri, 14 Apr 2000 12:44:33 -0400 (EDT)
Subject: [Patches] Making  modules callable when they have the attribute "__call__"
In-Reply-To: <14583.17536.445303.206607@rama.research.nokia.com>
References: <14583.13335.672195.314564@rama.research.nokia.com>
 <14583.14659.40548.466043@beluga.mojam.com>
 <14583.17536.445303.206607@rama.research.nokia.com>
Message-ID: <14583.19185.200179.572807@seahag.cnri.reston.va.us>

Pekka Pessi writes:
 > 	If we have a simple extension module, say, md5, that basically
 > 	defines one class/object type, say, md5, it is much cleaner to
 > 	define that class/object type also as __call__. Now the single
 > 	object type is sometimes called "new", sometimes something else.

  So you're looking for:

	import md5
        hasher = md5('my data')
        digest = hasher.digest()

  Why not use:

	from md5 import md5
        hasher = md5('my data')
        digest = hasher.digest()

  The import is the only difference, and there's no reason to think
that the md5 module won't grow additional convenience functions in the
future.  So if it's callable, we could do this:

        import sha

        hasher = sha('my data')
        digest = hasher.digest()

        hasher = sha.new('my data')
        digest = hasher.digest()

        # new convenience function to hash data
        # and return the printable form
        print sha.hexdigest('more data')

  This seems confusing.  I don't know of other languages which make
modules into anything other than containers/namespaces.

  -0


  -Fred

--
Fred L. Drake, Jr.	  <fdrake at acm.org>
Corporation for National Research Initiatives



From Vladimir.Marangozov@inrialpes.fr  Fri Apr 14 18:15:09 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Fri, 14 Apr 2000 19:15:09 +0200 (CEST)
Subject: [Patches] Re: Comparison of cyclic objects (was RE: [Python-Dev] trashcan and PR#7)
In-Reply-To: <14582.22650.792191.474554@bitdiddle.cnri.reston.va.us> from "Jeremy Hylton" at Apr 13, 2000 07:30:02 PM
Message-ID: <200004141715.TAA02492@python.inrialpes.fr>

Jeremy Hylton wrote:
> 
> Here it is contextified.  One small difference from the previous patch
> is that NESTING_LIMIT is now only 1000.  I think this is sufficient to
> cover commonly occuring nested containers.
> 
> Jeremy
> 
> [patch omitted]

Nice.

I think you don't need the _PyCompareState_flag. Like in trashcan,
_PyCompareState_nesting is enough to enter the sections of the code
that depend on _PyCompareState_flag.

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252


From jeremy@cnri.reston.va.us  Fri Apr 14 20:18:53 2000
From: jeremy@cnri.reston.va.us (Jeremy Hylton)
Date: Fri, 14 Apr 2000 15:18:53 -0400 (EDT)
Subject: [Patches] Re: Comparison of cyclic objects (was RE: [Python-Dev] trashcan and PR#7)
In-Reply-To: <200004141715.TAA02492@python.inrialpes.fr>
References: <14582.22650.792191.474554@bitdiddle.cnri.reston.va.us>
 <200004141715.TAA02492@python.inrialpes.fr>
Message-ID: <14583.28445.105079.446201@bitdiddle.cnri.reston.va.us>

>>>>> "VM" == Vladimir Marangozov <Vladimir.Marangozov@inrialpes.fr> writes:

  VM> Jeremy Hylton wrote:
  >>  Here it is contextified.  One small difference from the previous
  >> patch is that NESTING_LIMIT is now only 1000.  I think this is
  >> sufficient to cover commonly occuring nested containers.
  >> 
  >> Jeremy
  >> 
  >> [patch omitted]

  VM> Nice.

  VM> I think you don't need the _PyCompareState_flag. Like in
  VM> trashcan, _PyCompareState_nesting is enough to enter the
  VM> sections of the code that depend on _PyCompareState_flag.

Right.  Thanks for the suggestion, and thanks to Barry & Fred for
theirs.  I've checked in the changes.

Jeremy



From cgw@fnal.gov  Fri Apr 14 21:42:19 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Fri, 14 Apr 2000 15:42:19 -0500 (CDT)
Subject: [Patches] Memory leaks in Python 1.6
Message-ID: <14583.33451.120821.695270@buffalo.fnal.gov>

I've been running a small script I wrote called "leaktest" (attached
below) which runs tests from the the Python/Lib/test directory in a
loop, while watching the memory use of the process.  I've found quite
a few memory leaks this way, and am working on fixing them.  I will be
submitting a bunch of patches to the patches list over the next
several days.




Script: leaktest.py

#!/usr/bin/env python

"""
Runs tests repeatedly, looking for memory leaks
If you have gnuplot installed you can also get nice plots of memory
usage vs. time

Usage:  leaktest [-max max] [-min min] [-check check] [testname] [testname]...

runs one or more tests repeatedly, looking for memory leaks
The tests can be specified as module name (test_gzip) or file name
(/path/to/test_gzip.py)

memory usage is checked every "check" iterations of the test.

at least "min" readings are taken, and at most "max" readings.  If no
memory leak is detected after "min", the test stops.

If "DO_LOG" is set to 1, memory usage will be logged to a file.
If "DO_PLOT" is set to 1, pretty pictures will be drawn.
If "VERBOSE" is set to 1, memory usage will be printed during the test
"""


import sys
import os
import time
import string

DO_LOG=1
DO_PLOT=1
VERBOSE=0

def open_file(name):
    if os.path.exists(name):
        os.system("cp -b %s %s.bak"%(name,name))
    f=open(name,'w')
    return f

def mem_usage(pid):
    p=os.popen('ps uwp %s'%pid)
    lines=p.readlines()
    status=p.close()
    if status or len(lines)!=2:
        return None
    return int(string.split(lines[1])[4])

def no_increase(l):
    for i in xrange(1,len(l)):
        if l[i]>l[i-1]:
            return 0
    return 1
        
def run_test(testname, min_readings=10, max_readings=50, check_interval=10):
    print "RUN %s min %d max %d check %d"%(testname,min_readings,
                                           max_readings, check_interval)
    if DO_LOG:
        log = open_file('%s.log'%testname)
    logdata=[]
    ret=1
    pid=os.getpid()

    stdout, stderr = sys.stdout, sys.stderr
    sys.stdout, sys.stderr = open_file('%s.out'%testname), open_file('%s.err'%testname)
    
    for loop_readings in xrange(max_readings):
        mem = mem_usage(pid)
        if not mem:
            break
        if VERBOSE:
            stdout.write('%s\n'%mem)
        logdata.append(mem)
        if DO_LOG:
            log.write('%s\n'%mem)
            log.flush()

        for n in xrange(check_interval):
            try:
                if loop_readings==0:
                    exec "import test.%s"%testname
                else:
                    exec('reload(test.%s)'%testname)
            except:
                ret=-1
                break

        if ret<0:
            break

        if len(logdata)>=min_readings and no_increase(logdata[-min_readings:]):
            ret=0
            break

    sys.stdout.close()
    sys.stderr.close()
    sys.stdout, sys.stderr = stdout, stderr
    
    if DO_LOG:
        log.close()

    return ret, logdata
            
def do_plot(testname):
    cmdfile=open('%s.gnuplot' % testname, 'w')
    template='set nokey\nset xlabel "Iterations"\nset ylabel "Memory Use (KB)"\n\
                              set title "%s\nplot "%s" with lines\npause 60\n'
    cmdfile.write(template % (testname, testname+'.log'))
    cmdfile.close()
    os.system("gnuplot %s.gnuplot &"%testname)
    
def main(args):
    check_interval=10
    min_readings=10
    max_readings=50
    while args:
        arg = args.pop(0)
        if arg=='-min':
            min_readings = int(args.pop(0))
            continue
        if arg=='-max':
            max_readings = int(args.pop(0))
            continue
        if arg=='-check':
            check_interval = int(args.pop(0))
            continue
        ## allow the test module names to be specified as full pathnames
        ## for ease in calling this from shellscripts
        arg = os.path.split(arg)[1]
        arg = os.path.splitext(arg)[0]
        status, data=run_test(arg, min_readings, max_readings, check_interval)
        if status<0:
            print "test", arg, "did not run"
        elif status==0:
            print arg, "no memory leak detected", data
        else:
            print arg, "leaks memory", data
            if DO_PLOT:
                do_plot(arg)
    
if __name__ == '__main__':
    main(sys.argv[1:])
    
       


From cgw@fnal.gov  Fri Apr 14 22:00:08 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Fri, 14 Apr 2000 16:00:08 -0500 (CDT)
Subject: [Patches] Fix for memory leak in test_cpickle
Message-ID: <14583.34520.183725.620111@buffalo.fnal.gov>

Problem description:
	
	Run the following script:
	
import test.test_cpickle
for x in xrange(1000000):
    reload(test.test_cpickle)

Watch Python's memory use go up up and away!


In the course of debugging this I also saw that cPickle is
inconsistent with pickle - if you attempt a pickle.load or pickle.dump
on a closed file, you get a ValueError, whereas the corresponding
cPickle operations give an IOError.  Since cPickle is advertised as
being compatible with pickle, I changed these exceptions to match.

Index: Modules/cPickle.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/cPickle.c,v
retrieving revision 2.39
diff -c -r2.39 cPickle.c
*** cPickle.c	2000/03/10 23:11:40	2.39
--- cPickle.c	2000/04/14 20:55:20
***************
*** 2151,2169 ****
        Py_INCREF(file);
      else
        file=Pdata_New();
! 
!     self->file = file;
  
!     UNLESS (self->memo = PyDict_New()) {
!        Py_XDECREF((PyObject *)self);
!        return NULL;
!     }
  
      if (PyFile_Check(file)) {
          self->fp = PyFile_AsFile(file);
  	if (self->fp == NULL) {
! 	    PyErr_SetString(PyExc_IOError, "output file closed");
! 	    return NULL;
  	}
          self->write_func = write_file;
      }
--- 2151,2168 ----
        Py_INCREF(file);
      else
        file=Pdata_New();
!     
!     UNLESS (self->file = file) 
!       goto err;
  
!     UNLESS (self->memo = PyDict_New()) 
!        goto err;
  
      if (PyFile_Check(file)) {
          self->fp = PyFile_AsFile(file);
  	if (self->fp == NULL) {
! 	    PyErr_SetString(PyExc_ValueError, "I/O operation on closed file");
! 	    goto err;
  	}
          self->write_func = write_file;
      }
***************
*** 4054,4063 ****
      self->safe_constructors = NULL;
      self->find_class = NULL;
  
!     UNLESS (self->memo = PyDict_New()) {
!        Py_XDECREF((PyObject *)self);
!        return NULL;
!     }
  
      Py_INCREF(f);
      self->file = f;
--- 4053,4060 ----
      self->safe_constructors = NULL;
      self->find_class = NULL;
  
!     UNLESS (self->memo = PyDict_New()) 
!        goto err;
  
      Py_INCREF(f);
      self->file = f;
***************
*** 4066,4073 ****
      if (PyFile_Check(f)) {
          self->fp = PyFile_AsFile(f);
  	if (self->fp == NULL) {
! 	    PyErr_SetString(PyExc_IOError, "input file closed");
! 	    return NULL;
  	}
          self->read_func = read_file;
          self->readline_func = readline_file;
--- 4063,4070 ----
      if (PyFile_Check(f)) {
          self->fp = PyFile_AsFile(f);
  	if (self->fp == NULL) {
! 	    PyErr_SetString(PyExc_ValueError, "I/O operation on closed file");
! 	    goto err;
  	}
          self->read_func = read_file;
          self->readline_func = readline_file;
Index: Lib/test/test_cpickle.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/test/test_cpickle.py,v
retrieving revision 1.4
diff -c -r1.4 test_cpickle.py
*** test_cpickle.py	1999/07/13 15:23:42	1.4
--- test_cpickle.py	2000/04/14 20:55:30
***************
*** 79,96 ****
      f.close()
      try:
          cPickle.dump(123, f)
!     except IOError:
          pass
      else:
!         print "dump to closed file should raise IOError"
      f = open(fn, "r")
      f.close()
      try:
          cPickle.load(f)
!     except IOError:
          pass
      else:
!         print "load from closed file should raise IOError"
      os.remove(fn)
  
      # Test specific bad cases
--- 79,96 ----
      f.close()
      try:
          cPickle.dump(123, f)
!     except ValueError:
          pass
      else:
!         print "dump to closed file should raise ValueError"
      f = open(fn, "r")
      f.close()
      try:
          cPickle.load(f)
!     except ValueError:
          pass
      else:
!         print "load from closed file should raise ValueError"
      os.remove(fn)
  
      # Test specific bad cases




Disclaimer:

                   I confirm that, to the best of my knowledge and belief, this
                   contribution is free of any claims of third parties under
                   copyright, patent or other rights or interests ("claims").  To
                   the extent that I have any such claims, I hereby grant to CNRI a
                   nonexclusive, irrevocable, royalty-free, worldwide license to
                   reproduce, distribute, perform and/or display publicly, prepare
                   derivative versions, and otherwise use this contribution as part
                   of the Python software and its related documentation, or any
                   derivative versions thereof, at no cost to CNRI or its licensed
                   users, and to authorize others to do so.

                   I acknowledge that CNRI may, at its sole discretion, decide
                   whether or not to incorporate this contribution in the Python
                   software and its related documentation.  I further grant CNRI
                   permission to use my name and other identifying information
                   provided to CNRI by me for use in connection with the Python
                   software and its related documentation.


From mhammond@skippinet.com.au  Sat Apr 15 02:49:16 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Sat, 15 Apr 2000 11:49:16 +1000
Subject: [Patches] Making  modules callable when they have the attribute "__call__"
In-Reply-To: <14583.17649.904861.454848@rama.research.nokia.com>
Message-ID: <ECEPKNMJLHAPFFJHDOJBOEAOCJAA.mhammond@skippinet.com.au>

> Mark Hammond <mhammond@skippinet.com.au> writes:
>
> >> 	Attached you find patches that make modules
> >> callable when they have
> >> 	an attribute "__call__".  I think that this makes
> >> Python mode
> >> 	orthogonal. YMMV.
>
> >Wouldnt a better patch be to fill out the tp_call slot
> for the module?
>
> 	Nope. Unless you want to make a module without
> __call__ attribute
> 	also callable.

I would have thought so, yes.  Sort-of.

You mention your motiviation as making Python more orthogonal - then
IMO, calling a module with no __call__ attribute should raise an
attribute error, just like attempting to call an instance with no
__call__ attribute.

Mark.




From Pekka.Pessi@nokia.com  Sat Apr 15 09:27:56 2000
From: Pekka.Pessi@nokia.com (Pekka Pessi)
Date: Sat, 15 Apr 2000 11:27:56 +0300 (EEST)
Subject: [Patches] Making  modules callable when they have the attribute "__call__"
In-Reply-To: <14583.19185.200179.572807@seahag.cnri.reston.va.us>
References: <14583.13335.672195.314564@rama.research.nokia.com>
 <14583.14659.40548.466043@beluga.mojam.com>
 <14583.17536.445303.206607@rama.research.nokia.com> <14583.19185.200179.572807@seahag.cnri.reston.va.us>
 <14583.14659.40548.466043@beluga.mojam.com>
 <14583.17536.445303.206607@rama.research.nokia.com>
 <14583.19185.200179.572807@seahag.cnri.reston.va.us>
Message-ID: <14584.10252.273754.530946@rama.research.nokia.com>

Fred L. Drake, Jr. <fdrake@acm.org> writes:
>Pekka Pessi writes:
> > 	If we have a simple extension module, say, md5, that basically
> > 	defines one class/object type, say, md5, it is much cleaner to
> > 	define that class/object type also as __call__. Now the single
> > 	object type is sometimes called "new", sometimes something else.

>  So you're looking for:

>	import md5
>        hasher = md5('my data')
>        digest = hasher.digest()

	Actually, I'm looking something like

import socket

s = socket(socket.AF_INET6, socket.SOCK_SECKPACKET)

	where the __call__ in the socket module is something like a factory.

...
>  This seems confusing.  I don't know of other languages which make
>modules into anything other than containers/namespaces.

	What is a Python instance but a container/namespace?

					Pekka


From gstein@lyra.org  Sat Apr 15 18:35:15 2000
From: gstein@lyra.org (Greg Stein)
Date: Sat, 15 Apr 2000 10:35:15 -0700 (PDT)
Subject: [Patches] Proposal To Modifying Map Function in 1.6
In-Reply-To: <14583.12058.856533.820813@beluga.mojam.com>
Message-ID: <Pine.LNX.4.10.10004151033100.13301-100000@nebula.lyra.org>

On Fri, 14 Apr 2000, Skip Montanaro wrote:
> 
> Zaur wrote:
> > ... patch for map() change ...
> 
> ... another view on map() ...
> 
> Given the somewhat arbitrary nature of choosing any of these three rules,
> you might as well stay with the status quo to avoid code breakage in the
> code that relies on the current semantics.

There would be serious breakage should the definition of map() change.

Zaur: one way to do what you'd like:

    map(func, List, [arg]*len(List))


Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/



From cgw@alum.mit.edu  Sat Apr 15 23:12:03 2000
From: cgw@alum.mit.edu (Charles G Waldman)
Date: Sat, 15 Apr 2000 17:12:03 -0500 (CDT)
Subject: [Patches] Fix for memory leak in _PyResize_Tuple
Message-ID: <14584.59699.721573.716990@sirius.net.home>


In the course of testing for memory leaks I found this gem in
test_extcall (there are actually several different memory leaks
triggered by test_extcall, this is only the first of several fixes,
which I am submitting as separate patches).



Test Case:  run the following code:

class Nothing:
    def __len__(self):
        return 5
    def __getitem__(self, i):
        if i < 3:
            return i
        else:
            raise IndexError, i

def g(a,*b,**c):
    return

for x in xrange(1000000):
    g(*Nothing())


and watch Python's memory use go up and up.


Diagnosis:

The analysis begins with the call to PySequence_Tuple at line 1641 in
ceval.c - the argument to g is seen to be a sequence but not a tuple,
so it needs to be converted from an abstract sequence to a concrete
tuple.  PySequence_Tuple starts off by creating a new tuple of length
5 (line 1122 in abstract.c).  Then at line 1149, since only 3 elements
were assigned, _PyTuple_Resize is called to make the 5-tuple into a
3-tuple.  When we're all done the 3-tuple is decrefed, but rather than
being freed it is placed on the free_tuples cache.

The basic problem is that the 3-tuples are being added to the cache
but never picked up again, since _PyTuple_Resize doesn't make use of
the free_tuples cache.  If you are resizing a 5-tuple to a 3-tuple and
there is already a 3-tuple in free_tuples[3], instead of using this
tuple, _PyTuple_Resize will realloc the 5-tuple to a 3-tuple.  It
would more efficient to use the existing 3-tuple and cache the
5-tuple.

By making _PyTuple_Resize aware of the free_tuples (just as
PyTuple_New), we not only save a few calls to realloc, but also
prevent this misbehavior whereby tuples are being added to the
free_tuples list but never properly "recycled".

Patch:
Index: Objects/tupleobject.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/tupleobject.c,v
retrieving revision 2.30
diff -c -r2.30 tupleobject.c
*** tupleobject.c	2000/03/13 16:01:29	2.30
--- tupleobject.c	2000/04/15 22:08:27
***************
*** 469,482 ****
  		Py_XDECREF(v->ob_item[i]);
  		v->ob_item[i] = NULL;
  	}
! 	sv = (PyTupleObject *)
! 		realloc((char *)v,
! 			sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
! 	*pv = (PyObject *) sv;
! 	if (sv == NULL) {
! 		PyMem_DEL(v);
! 		PyErr_NoMemory();
! 		return -1;
  	}
  	_Py_NewReference((PyObject *)sv);
  	for (i = sv->ob_size; i < newsize; i++)
--- 469,515 ----
  		Py_XDECREF(v->ob_item[i]);
  		v->ob_item[i] = NULL;
  	}
! #if MAXSAVESIZE > 0
! 	if (newsize == 0 && free_tuples[0]) {
! 		sv = free_tuples[0];
! 		sv->ob_size = 0;
! 		Py_INCREF(sv);
! #ifdef COUNT_ALLOCS
! 		tuple_zero_allocs++;
! #endif
! 		_Py_Dealloc((PyObject*)v);
! 		*pv = (PyObject*) sv;
! 		return 0;
! 	}
! 	if (0 < newsize && newsize < MAXSAVESIZE &&
! 	    (sv = free_tuples[newsize]) != NULL)
! 	{
! 		free_tuples[newsize] = (PyTupleObject *) sv->ob_item[0];
! #ifdef COUNT_ALLOCS
! 		fast_tuple_allocs++;
! #endif
! #ifdef Py_TRACE_REFS 
! 		sv->ob_type = &PyTuple_Type; 
! #endif 
! 		for (i = 0; i < newsize; ++i){
! 			sv->ob_item[i] = v->ob_item[i];
! 			v->ob_item[i] = NULL;
! 		}
! 		sv->ob_size = v->ob_size;
! 		_Py_Dealloc((PyObject*)v);
! 		*pv = (PyObject *) sv;
! 	} else 
! #endif		
! 	{
! 		sv = (PyTupleObject *)
! 			realloc((char *)v,
! 				sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
! 		*pv = (PyObject *) sv;
! 		if (sv == NULL) {
! 			PyMem_DEL(v);
! 			PyErr_NoMemory();
! 			return -1;
! 		}
  	}
  	_Py_NewReference((PyObject *)sv);
  	for (i = sv->ob_size; i < newsize; i++)





Disclaimer:

     I confirm that, to the best of my knowledge and belief, this
     contribution is free of any claims of third parties
     under copyright, patent or other rights or interests ("claims").  To
     the extent that I have any such claims, I hereby grant to CNRI a
     nonexclusive, irrevocable, royalty-free, worldwide license to
     reproduce, distribute, perform and/or display publicly, prepare
     derivative versions, and otherwise use this contribution as part
     of the Python software and its related documentation, or any
     derivative versions thereof, at no cost to CNRI or its licensed
     users, and to authorize others to do so.

     I acknowledge that CNRI may, at its sole discretion, decide
     whether or not to incorporate this contribution in the Python
     software and its related documentation.  I further grant CNRI
     permission to use my name and other identifying information
     provided to CNRI by me for use in connection with the Python
     software and its related documentation.


From jeremy@alum.mit.edu  Sun Apr 16 21:53:43 2000
From: jeremy@alum.mit.edu (Jeremy Hylton)
Date: Sun, 16 Apr 2000 16:53:43 -0400 (EDT)
Subject: [Patches] Fix for memory leak in _PyResize_Tuple
In-Reply-To: <14584.59699.721573.716990@sirius.net.home>
References: <14584.59699.721573.716990@sirius.net.home>
Message-ID: <14586.10327.527863.292916@bitdiddle.cnri.reston.va.us>

>>>>> "CGW" == Charles G Waldman <cgw@alum.mit.edu> writes:

  CGW> The basic problem is that the 3-tuples are being added to the
  CGW> cache but never picked up again, since _PyTuple_Resize doesn't
  CGW> make use of the free_tuples cache.  If you are resizing a
  CGW> 5-tuple to a 3-tuple and there is already a 3-tuple in
  CGW> free_tuples[3], instead of using this tuple, _PyTuple_Resize
  CGW> will realloc the 5-tuple to a 3-tuple.  It would more efficient
  CGW> to use the existing 3-tuple and cache the 5-tuple.

  CGW> By making _PyTuple_Resize aware of the free_tuples (just as
  CGW> PyTuple_New), we not only save a few calls to realloc, but also
  CGW> prevent this misbehavior whereby tuples are being added to the
  CGW> free_tuples list but never properly "recycled".

Question about this patch: Is it always more efficient to re-use a
tuple (and update all the ob_item pointers)?  Or are there cases where
it would be more efficient to realloc and avoid the cost of updating
the pointers?  

There's another problem with the current strategy for reusing tuples
that we might fix at the same time.  If a program creates a large
number of tuples, later releases them all, and never creates any more
tuples of the same size, that memory will live forever on the free
list.  If I have a program that creates a million 3-tuples at one
blow, that memory will never get freed.

It should be possible to fix this problem by keeping track of the size
of each freelist and never letting it exceed some maximum size.  If it
actually ends up being more efficient to realloc than copy, then this
patch would limit the amount of loss caused by oddball cases like
calling PyObject_Sequence on an instance that overstates its size.

Jeremy

Index: Objects/tupleobject.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/tupleobject.c,v
retrieving revision 2.30
diff -c -r2.30 tupleobject.c
*** tupleobject.c	2000/03/13 16:01:29	2.30
--- tupleobject.c	2000/04/16 20:51:36
***************
*** 42,47 ****
--- 42,49 ----
     tuple () of which at most one instance will be allocated.
  */
  static PyTupleObject *free_tuples[MAXSAVESIZE];
+ static int free_length[MAXSAVESIZE];
+ #define MAX_FREE_LENGTH 100
  #endif
  #ifdef COUNT_ALLOCS
  int fast_tuple_allocs;
***************
*** 71,76 ****
--- 73,79 ----
  	    (op = free_tuples[size]) != NULL)
  	{
  		free_tuples[size] = (PyTupleObject *) op->ob_item[0];
+ 		free_length[size]--;
  #ifdef COUNT_ALLOCS
  		fast_tuple_allocs++;
  #endif
***************
*** 178,186 ****
  		while (--i >= 0)
  			Py_XDECREF(op->ob_item[i]);
  #if MAXSAVESIZE > 0
! 		if (op->ob_size < MAXSAVESIZE) {
  			op->ob_item[0] = (PyObject *) free_tuples[op->ob_size];
  			free_tuples[op->ob_size] = op;
  			goto done; /* return */
  		}
  #endif
--- 181,195 ----
  		while (--i >= 0)
  			Py_XDECREF(op->ob_item[i]);
  #if MAXSAVESIZE > 0
! 		if ((op->ob_size < MAXSAVESIZE)
! 		    && (free_tuples[op->ob_size] != NULL
! 			&& free_length[op->ob_size] < MAX_FREE_LENGTH)) {
  			op->ob_item[0] = (PyObject *) free_tuples[op->ob_size];
  			free_tuples[op->ob_size] = op;
+ 			if (free_tuples[op->ob_size] == NULL)
+ 			    free_length[op->ob_size] = 1;
+ 			else
+ 			    free_length[op->ob_size]++;
  			goto done; /* return */
  		}
  #endif



From cgw@fnal.gov  Mon Apr 17 03:19:19 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Sun, 16 Apr 2000 21:19:19 -0500 (CDT)
Subject: [Patches] Fix for memory leak in _PyResize_Tuple
In-Reply-To: <14586.10327.527863.292916@bitdiddle.cnri.reston.va.us>
References: <14584.59699.721573.716990@sirius.net.home>
 <14586.10327.527863.292916@bitdiddle.cnri.reston.va.us>
Message-ID: <14586.29863.886669.766779@buffalo.fnal.gov>

Jeremy Hylton writes:
 > 
 > Question about this patch: Is it always more efficient to re-use a
 > tuple (and update all the ob_item pointers)?  Or are there cases where
 > it would be more efficient to realloc and avoid the cost of updating
 > the pointers?

According to Webster's, "efficient" has two meanings:
1: immediately effecting 
2: productive of desired effects; esp : productive without waste 


Realloc might be (?) more efficient in sense 1 (speed), but certainly
not in sense 2 (waste).  It just seems ecologically wrong to be
keeping this heap of used tuples around for recycling, then going to
the store and buying a new one when your old one doesn't fit any more,
instead of looking on the recycling heap first.  If speed is a great
concern, the updating of the ob_item pointers could be done with a
memcpy (I thought about doing it this way but the for-loop seemed
clearer).

 > There's another problem with the current strategy for reusing tuples
 > that we might fix at the same time.  If a program creates a large
 > number of tuples, later releases them all, and never creates any more
 > tuples of the same size, that memory will live forever on the free
 > list.  If I have a program that creates a million 3-tuples at one
 > blow, that memory will never get freed.

Making a limit on the size of the recycling heap is a different issue
and maybe that should be done as well, but I don't think that's a
reason not to fix _PyTuple_Resize.

I thought about adding this too, but chose not to for the following 4
reasons: (1) I have to choose an arbitrary value for the cut-off and I
don't like to impose arbitrary limits.  Do you want to keep 42 free
tuples?  137?  1024?  Somebody else come up with the number and I'll
put that in too, except that (2) I am not really 100% sure that the
behavior you describe is a problem.  Maybe the function that creates
the million 3-tuples will run again and need to create another million
3-tuples and having them all cached will be advantageous.  Or maybe
not.  But (4) I have been thinking about memory leaks and "ecology"
and it seems to me to be a desirable property that you be able to call
a function f() then call it again and again and not use up more memory
on each call.  (Of course I'm assuming you use the same parameters and
f isn't deliberately constructed to consume memory...)  To me,
continuing to leak a little bit of memory on each call is worse than
doing a huge alloc one time.... it's OK if the high-water mark goes up
way high and doesn't recede, as long as it doesn't keep inching up.

When I iterate the tests in the Python Lib/test package and plot the
interpreter's memory footprint vs. iteration number, I see two
distinct patterns - a monotone increasing ramp, or a "mesa".  If I see
the ramp, I know right away that there is leaky behavior, and I can
try to fix it.  (Right now I see this with the modules test_cpickle,
test_extcall, test_nis, and test_unicode, and I hope to fix all of
these leaks, time permitting).

With a limit on the number of tuples in the freelist, there would be a
third type of pattern - a ramp which flattens off when a hard-limit is
reached.  This seems like it would be harder to detect if you are
doing memory-leak testing.

 > It should be possible to fix this problem by keeping track of the size
 > of each freelist and never letting it exceed some maximum size.  If it
 > actually ends up being more efficient to realloc than copy, then this
 > patch would limit the amount of loss caused by oddball cases like
 > calling PyObject_Sequence on an instance that overstates its size.

Yes, it's an oddball case, but it came from Python's own test suite!
(I'm not devious enough to have come up with this one, I don't think).
Ideally, no matter what you throw at it, you should be able to get
Python to exhibit signs of insidious memory leaks, especially not
running the built-in test suite.  "Limiting the loss" is less elegant
than not allowing it to happen in the first place.  If the limit was
1024 I could run the "text_extcall" test 1000 times and think I saw a
memory leak.




From jeremy@cnri.reston.va.us  Mon Apr 17 04:31:39 2000
From: jeremy@cnri.reston.va.us (Jeremy Hylton)
Date: Sun, 16 Apr 2000 23:31:39 -0400 (EDT)
Subject: [Patches] Fix for memory leak in _PyResize_Tuple
In-Reply-To: <14586.29863.886669.766779@buffalo.fnal.gov>
References: <14584.59699.721573.716990@sirius.net.home>
 <14586.10327.527863.292916@bitdiddle.cnri.reston.va.us>
 <14586.29863.886669.766779@buffalo.fnal.gov>
Message-ID: <14586.34203.790440.916175@bitdiddle.cnri.reston.va.us>

>>>>> "CGW" == Charles G Waldman <cgw@fnal.gov> writes:

  CGW> Jeremy Hylton writes:
  >>  Question about this patch: Is it always more efficient to re-use
  >> a tuple (and update all the ob_item pointers)?  Or are there
  >> cases where it would be more efficient to realloc and avoid the
  >> cost of updating the pointers?

  CGW> According to Webster's, "efficient" has two meanings: 1:
  CGW> immediately effecting 2: productive of desired effects; esp :
  CGW> productive without waste

  CGW> Realloc might be (?) more efficient in sense 1 (speed), but
  CGW> certainly not in sense 2 (waste).  It just seems ecologically
  CGW> wrong to be keeping this heap of used tuples around for
  CGW> recycling, then going to the store and buying a new one when
  CGW> your old one doesn't fit any more, instead of looking on the
  CGW> recycling heap first.  If speed is a great concern, the
  CGW> updating of the ob_item pointers could be done with a memcpy (I
  CGW> thought about doing it this way but the for-loop seemed
  CGW> clearer).

The list of free tuples is, I assume, a speed optimization.  If speed
were not an issue, it would be simplest by far to use malloc and free
each time.  If the free list is a speed optimization, it seems
sensible to consider that issue.

I think your proposed patch looks reasonable, but I didn't know what
values to use for the cost model.  (How frequently does realloc move
the pointer?  How much does that cost?  How much does the pointer
update cost?)  Performance may not be a crucial issue here, since
_PyTuple_Resize is used three times in the Python core, but seems
worth considering.

  >> There's another problem with the current strategy for reusing
  >> tuples that we might fix at the same time.  If a program creates
  >> a large number of tuples, later releases them all, and never
  >> creates any more tuples of the same size, that memory will live
  >> forever on the free list.  If I have a program that creates a
  >> million 3-tuples at one blow, that memory will never get freed.

  CGW> Making a limit on the size of the recycling heap is a different
  CGW> issue and maybe that should be done as well, but I don't think
  CGW> that's a reason not to fix _PyTuple_Resize.

It's a related issue.  If the cost of the patch is much greater than
the cost of realloc (not saying that it is), then it may be worth
leaving realloc alone.

  CGW> I thought about adding this too, but chose not to for the
  CGW> following 4 reasons: (1) I have to choose an arbitrary value
  CGW> for the cut-off and I don't like to impose arbitrary limits.
  CGW> Do you want to keep 42 free tuples?  137?  1024?  Somebody else
  CGW> come up with the number and I'll put that in too, except that

Yes.  Arbitrary constants are worrisome, but I think not worrisome
enough to leave 2,000,000 unused 10-tuples on the free list
indefinitely :-).

  CGW> (2) I am not really 100% sure that the behavior you describe is
  CGW> a problem.  Maybe the function that creates the million
  CGW> 3-tuples will run again and need to create another million
  CGW> 3-tuples and having them all cached will be advantageous.  Or

I think it's a bug because it is a surprising effect.  If you create
and then free a lot of tuples, it seems unfortunate that a speed
optimization causes memory utilization to stay high.

  CGW> maybe not.  But (4) I have been thinking about memory leaks and
  CGW> "ecology" and it seems to me to be a desirable property that
  CGW> you be able to call a function f() then call it again and again
  CGW> and not use up more memory on each call.  (Of course I'm
  CGW> assuming you use the same parameters and f isn't deliberately
  CGW> constructed to consume memory...)  To me, continuing to leak a
  CGW> little bit of memory on each call is worse than doing a huge
  CGW> alloc one time.... it's OK if the high-water mark goes up way
  CGW> high and doesn't recede, as long as it doesn't keep inching up.

I'm not sure I'm convinced.  I don't think it matters who the tank
fills up (a little at a time or all at once), if the program doesn't
need the memory it should remain in use because of some detail of the
language implementation.

  CGW> When I iterate the tests in the Python Lib/test package and
  CGW> plot the interpreter's memory footprint vs. iteration number, I
  CGW> see two distinct patterns - a monotone increasing ramp, or a
  CGW> "mesa".  If I see the ramp, I know right away that there is
  CGW> leaky behavior, and I can try to fix it.  (Right now I see this
  CGW> with the modules test_cpickle, test_extcall, test_nis, and
  CGW> test_unicode, and I hope to fix all of these leaks, time
  CGW> permitting).

Great!

  >> It should be possible to fix this problem by keeping track of the
  >> size of each freelist and never letting it exceed some maximum
  >> size.  If it actually ends up being more efficient to realloc
  >> than copy, then this patch would limit the amount of loss caused
  >> by oddball cases like calling PyObject_Sequence on an instance
  >> that overstates its size.

  CGW> Yes, it's an oddball case, but it came from Python's own test
  CGW> suite!  (I'm not devious enough to have come up with this one,
  CGW> I don't think).

I'll take that as a compliment!  I was feeling devious when I wrote it
:-).

  CGW>                 Ideally, no matter what you throw at it, you
  CGW> should be able to get Python to exhibit signs of insidious
  CGW> memory leaks, especially not running the built-in test suite.

Yes.

  CGW> "Limiting the loss" is less elegant than not allowing it to
  CGW> happen in the first place.  If the limit was 1024 I could run
  CGW> the "text_extcall" test 1000 times and think I saw a memory
  CGW> leak.

I think "limiting the loss" is a less elegant solution to the problem
you discovered, but I think the second problem is worth solving, too.

Jeremy
 





From anthony@interlink.com.au  Mon Apr 17 06:37:17 2000
From: anthony@interlink.com.au (Anthony Baxter)
Date: Mon, 17 Apr 2000 15:37:17 +1000
Subject: [Patches] support for writing pre-formatted ULAW samples in sunau.
Message-ID: <200004170537.PAA03485@mbuna.arbhome.com.au>

The sunau module's write objects assume that the data that's coming
in is always in linear format. When taking some existing data that's in
ULAW, and writing it out as ULAW, you have to use audioop.ulaw2lin, then
let the sunau module convert it back - not exactly the most pleasant 
approach (I also find a loss of quality). 

As a workaround for this, I added a formatted_ flag to writeframes() and
writeframesraw() - if this is set, it doesn't do the audioop conversion.

Couple of questions:

  Is this appropriate to add to the library (I think so, but... ?)
  
  Is this the best way?

  Which other modules should be similarly modified, if this is appropriate?
  (aifc, wave)? I'm willing to fix these to match, if this is deemed an 
  appropriate change...

ta,
Anthony

*** /opt/python152/lib/python1.5/sunau.py	Tue May 25 22:26:50 1999
--- sunau.py	Mon Apr 17 15:29:56 2000
***************
*** 377,394 ****
  	def tell(self):
  		return self._nframeswritten
  
! 	def writeframesraw(self, data):
  		self._ensure_header_written()
  		nframes = len(data) / self._framesize
! 		if self._comptype == 'ULAW':
  			import audioop
  			data = audioop.lin2ulaw(data, self._sampwidth)
  		self._file.write(data)
  		self._nframeswritten = self._nframeswritten + nframes
  		self._datawritten = self._datawritten + len(data)
  
! 	def writeframes(self, data):
! 		self.writeframesraw(data)
  		if self._nframeswritten != self._nframes or \
  			  self._datalength != self._datawritten:
  			self._patchheader()
--- 377,394 ----
  	def tell(self):
  		return self._nframeswritten
  
! 	def writeframesraw(self, data, formatted_=0):
  		self._ensure_header_written()
  		nframes = len(data) / self._framesize
! 		if self._comptype == 'ULAW' and not formatted_:
  			import audioop
  			data = audioop.lin2ulaw(data, self._sampwidth)
  		self._file.write(data)
  		self._nframeswritten = self._nframeswritten + nframes
  		self._datawritten = self._datawritten + len(data)
  
! 	def writeframes(self, data, formatted_=0):
! 		self.writeframesraw(data, formatted_)
  		if self._nframeswritten != self._nframes or \
  			  self._datalength != self._datawritten:
  			self._patchheader()



I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.


--
Anthony Baxter     <anthony@interlink.com.au>
It's never too late to have a happy childhood.


From brian@garage.co.jp  Tue Apr 18 03:13:32 2000
From: brian@garage.co.jp (Brian Hooper)
Date: Tue, 18 Apr 2000 02:13:32 GMT
Subject: [Patches] trailing backslash breaks os.stat
Message-ID: <20000418021332.91759.qmail@hotmail.com>

This is a multi-part message in MIME format.

------=_NextPart_000_1bdd1a90_126466dd$11a6c204
Content-Type: text/plain; format=flowed

posixmodule's posix_do_stat simply calls the underlying OS's stat
function, so it's really the underlying OS's fault, but a trailing
backslash in the path on a call to os.stat on NT returns an error
rather than the file info; this seems inconsistent with stat on
other platforms, where a trailing forward slash is OK.  A dos
'dir \temp\' with a trailing backslash, for example, also works, but
stat on that same directory will fail.

A path which is too long will also blow up the library call.

The patch attempts to correct the above two issues.  The patch was
written by Don Bennett (dpb@infoseek.com), and only tweaked by me.

--Brian Hooper

----

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.


______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

------=_NextPart_000_1bdd1a90_126466dd$11a6c204
Content-Type: text/plain; name="posixmodule.patch.txt"; format=flowed
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment; filename="posixmodule.patch.txt"

*** Modules/posixmodule.c       Thu Apr 13 08:20:40 2000
--- Modules.new/posixmodule.c   Mon Apr 17 16:09:09 2000
***************
*** 549,556 ****
--- 549,577 ----
        struct stat st;
        char *path;
        int res;
+
+ #ifdef MS_WIN32
+       int pathlen;
+ #endif /* MS_WIN32 */
+
        if (!PyArg_ParseTuple(args, format, &path))
                return NULL;
+
+ #ifdef MS_WIN32
+       pathlen = strlen(path);
+       /* the library call can blow up if the file name is too long! */
+       if (pathlen > MAX_PATH) {
+           errno = ENAMETOOLONG;
+           return posix_error();
+       }
+
+       if (!((pathlen==3) && (path[1] == ':') && (path[2] == '\\'))) {
+           if ((pathlen > 0) && (path[pathlen-1] == '\\')) {
+               path[pathlen-1] = 0; /* nuke the trailing backslash */
+           }
+       }
+ #endif /* MS_WIN32 */
+
        Py_BEGIN_ALLOW_THREADS
        res = (*statfunc)(path, &st);
        Py_END_ALLOW_THREADS

------=_NextPart_000_1bdd1a90_126466dd$11a6c204--


From cgw@fnal.gov  Tue Apr 18 23:52:02 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Tue, 18 Apr 2000 17:52:02 -0500 (CDT)
Subject: [Patches] Revised patch for _PyResize_Tuple memory leak
Message-ID: <14588.59154.860973.528863@buffalo.fnal.gov>

This patch replaces my submission of Sun, 16 Apr and addresses Jeremy
Hylton's suggestions that we also limit the size of the free tuple
list.  I chose 2000 as the maximum number of tuples of any particular
size to save.

There was also a problem with the previous version of this patch
causing a core dump if Python was built with Py_TRACE_REFS.  This is
fixed in the below version of the patch, which uses tupledealloc
instead of _Py_Dealloc.

Disclaimer:
                   I confirm that, to the best of my knowledge and belief, this
                   contribution is free of any claims of third parties under
                   copyright, patent or other rights or interests ("claims").  To
                   the extent that I have any such claims, I hereby grant to CNRI a
                   nonexclusive, irrevocable, royalty-free, worldwide license to
                   reproduce, distribute, perform and/or display publicly, prepare
                   derivative versions, and otherwise use this contribution as part
                   of the Python software and its related documentation, or any
                   derivative versions thereof, at no cost to CNRI or its licensed
                   users, and to authorize others to do so.

                   I acknowledge that CNRI may, at its sole discretion, decide
                   whether or not to incorporate this contribution in the Python
                   software and its related documentation.  I further grant CNRI
                   permission to use my name and other identifying information
                   provided to CNRI by me for use in connection with the Python
                   software and its related documentation.

Index: Objects/tupleobject.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Objects/tupleobject.c,v
retrieving revision 2.30
diff -c -r2.30 tupleobject.c
*** tupleobject.c	2000/03/13 16:01:29	2.30
--- tupleobject.c	2000/04/18 20:36:10
***************
*** 33,47 ****
  
  #include "Python.h"
  
  #ifndef MAXSAVESIZE
! #define MAXSAVESIZE	20
  #endif
  
  #if MAXSAVESIZE > 0
! /* Entries 1 upto MAXSAVESIZE are free lists, entry 0 is the empty
     tuple () of which at most one instance will be allocated.
  */
  static PyTupleObject *free_tuples[MAXSAVESIZE];
  #endif
  #ifdef COUNT_ALLOCS
  int fast_tuple_allocs;
--- 33,52 ----
  
  #include "Python.h"
  
+ /* Speed optimization to avoid frequent malloc/free of small tuples */
  #ifndef MAXSAVESIZE
! #define MAXSAVESIZE	20  /* Largest tuple to save on free list */
  #endif
+ #ifndef MAXSAVEDTUPLES 
+ #define MAXSAVEDTUPLES  2000  /* Maximum number of tuples of each size to save */
+ #endif
  
  #if MAXSAVESIZE > 0
! /* Entries 1 up to MAXSAVESIZE are free lists, entry 0 is the empty
     tuple () of which at most one instance will be allocated.
  */
  static PyTupleObject *free_tuples[MAXSAVESIZE];
+ static int num_free_tuples[MAXSAVESIZE];
  #endif
  #ifdef COUNT_ALLOCS
  int fast_tuple_allocs;
***************
*** 71,76 ****
--- 76,82 ----
  	    (op = free_tuples[size]) != NULL)
  	{
  		free_tuples[size] = (PyTupleObject *) op->ob_item[0];
+ 		num_free_tuples[size]--;
  #ifdef COUNT_ALLOCS
  		fast_tuple_allocs++;
  #endif
***************
*** 104,109 ****
--- 110,116 ----
  #if MAXSAVESIZE > 0
  	if (size == 0) {
  		free_tuples[0] = op;
+ 		++num_free_tuples[0];
  		Py_INCREF(op);	/* extra INCREF so that this is never freed */
  	}
  #endif
***************
*** 171,186 ****
  	register PyTupleObject *op;
  {
  	register int i;
! 
  	Py_TRASHCAN_SAFE_BEGIN(op)
! 	if (op->ob_size > 0) {
! 		i = op->ob_size;
  		while (--i >= 0)
  			Py_XDECREF(op->ob_item[i]);
  #if MAXSAVESIZE > 0
! 		if (op->ob_size < MAXSAVESIZE) {
! 			op->ob_item[0] = (PyObject *) free_tuples[op->ob_size];
! 			free_tuples[op->ob_size] = op;
  			goto done; /* return */
  		}
  #endif
--- 178,194 ----
  	register PyTupleObject *op;
  {
  	register int i;
! 	register int len =  op->ob_size;
  	Py_TRASHCAN_SAFE_BEGIN(op)
! 	if (len > 0) {
! 		i = len;
  		while (--i >= 0)
  			Py_XDECREF(op->ob_item[i]);
  #if MAXSAVESIZE > 0
! 		if (len < MAXSAVESIZE && num_free_tuples[len] < MAXSAVEDTUPLES) {
! 			op->ob_item[0] = (PyObject *) free_tuples[len];
! 			num_free_tuples[len]++;
! 			free_tuples[len] = op;
  			goto done; /* return */
  		}
  #endif
***************
*** 469,482 ****
  		Py_XDECREF(v->ob_item[i]);
  		v->ob_item[i] = NULL;
  	}
! 	sv = (PyTupleObject *)
! 		realloc((char *)v,
! 			sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
! 	*pv = (PyObject *) sv;
! 	if (sv == NULL) {
! 		PyMem_DEL(v);
! 		PyErr_NoMemory();
! 		return -1;
  	}
  	_Py_NewReference((PyObject *)sv);
  	for (i = sv->ob_size; i < newsize; i++)
--- 477,525 ----
  		Py_XDECREF(v->ob_item[i]);
  		v->ob_item[i] = NULL;
  	}
! #if MAXSAVESIZE > 0
! 	if (newsize == 0 && free_tuples[0]) {
! 		num_free_tuples[0]--;
! 		sv = free_tuples[0];
! 		sv->ob_size = 0;
! 		Py_INCREF(sv);
! #ifdef COUNT_ALLOCS
! 		tuple_zero_allocs++;
! #endif
! 		tupledealloc(v);
! 		*pv = (PyObject*) sv;
! 		return 0;
! 	}
! 	if (0 < newsize && newsize < MAXSAVESIZE &&
! 	    (sv = free_tuples[newsize]) != NULL)
! 	{
! 		free_tuples[newsize] = (PyTupleObject *) sv->ob_item[0];
! 		num_free_tuples[newsize]--;
! #ifdef COUNT_ALLOCS
! 		fast_tuple_allocs++;
! #endif
! #ifdef Py_TRACE_REFS 
! 		sv->ob_type = &PyTuple_Type; 
! #endif 
! 		for (i = 0; i < newsize; ++i){
! 			sv->ob_item[i] = v->ob_item[i];
! 			v->ob_item[i] = NULL;
! 		}
! 		sv->ob_size = v->ob_size;
! 		tupledealloc(v);
! 		*pv = (PyObject *) sv;
! 	} else 
! #endif		
! 	{
! 		sv = (PyTupleObject *)
! 			realloc((char *)v,
! 				sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
! 		*pv = (PyObject *) sv;
! 		if (sv == NULL) {
! 			PyMem_DEL(v);
! 			PyErr_NoMemory();
! 			return -1;
! 		}
  	}
  	_Py_NewReference((PyObject *)sv);
  	for (i = sv->ob_size; i < newsize; i++)


From tismer@tismer.com  Wed Apr 19 12:35:28 2000
From: tismer@tismer.com (Christian Tismer)
Date: Wed, 19 Apr 2000 13:35:28 +0200
Subject: [Patches] Revised patch for _PyResize_Tuple memory leak
References: <14588.59154.860973.528863@buffalo.fnal.gov>
Message-ID: <38FD9A00.C283D205@tismer.com>


Charles G Waldman wrote:
> 
> This patch replaces my submission of Sun, 16 Apr and addresses Jeremy
> Hylton's suggestions that we also limit the size of the free tuple
> list.  I chose 2000 as the maximum number of tuples of any particular
> size to save.

I think this limit is great. It limits the cashed
data to about 540000 words.

K, L = 3+1, 3+20 ; print reduce(lambda x,y:x+y, range(K, L+1))*2000

This makes me think of integers which are cashed as well.
After some big range expressions, we have thousands of thousands
of integer slots available, eating memory.

In this case, things are more difficult, since integers are
alloted in pages. Not so bad, but we might end up
with many many pages where only a few ints are used.

Any ideas how to limit this acache as well?

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From mwh21@cam.ac.uk  Wed Apr 19 19:28:10 2000
From: mwh21@cam.ac.uk (Michael Hudson)
Date: Wed, 19 Apr 2000 19:28:10 +0100 (BST)
Subject: [Patches] docstrings for os.spawn*
Message-ID: <Pine.LNX.4.10.10004191921590.29959-100000@localhost.localdomain>

I have written some (rather poor, IMO) docstrings for the new-ish
os.spawn* functions - I hope these can either go into the reposistory or
provoke someone into writing better ones!

I've tried to ape the style of the other docstrings in the standard
library - are there any guidelines for how such things should be written?

Cheers,
M.

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

Index: os.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/os.py,v
retrieving revision 1.31
diff -u -r1.31 os.py
--- os.py	2000/04/04 19:50:04	1.31
+++ os.py	2000/04/19 18:27:30
@@ -333,17 +333,44 @@
                     raise error, "Not stopped, signaled or exited???"
 
     def spawnv(mode, file, args):
+        """spawnv(mode, file, args) -> integer
+
+Execute file with arguments from args in a subprocess.
+If mode == P_NOWAIT return the pid of the process.
+If mode == P_WAIT return the process's exit code if it exits normally;
+otherwise return - (the signal that killed it). """   
         return _spawnvef(mode, file, args, None, execv)
 
     def spawnve(mode, file, args, env):
+        """spawnve(mode, file, args, env) -> integer
+
+Execute file with arguments from args in a subprocess with the
+specified environment.
+If mode == P_NOWAIT return the pid of the process.
+If mode == P_WAIT return the process's exit code if it exits normally;
+otherwise return -SIG, where SIG is the signal that killed it.. """
         return _spawnvef(mode, file, args, env, execve)
 
     # Note: spawnvp[e] is't currently supported on Windows
 
     def spawnvp(mode, file, args):
+        """spawnvp(mode, file, args) -> integer
+
+Execute file (which is looked for along $PATH) with arguments from
+args in a subprocess.
+If mode == P_NOWAIT return the pid of the process.
+If mode == P_WAIT return the process's exit code if it exits normally;
+otherwise return -SIG, where SIG is the signal that killed it.. """
         return _spawnvef(mode, file, args, None, execvp)
 
     def spawnvpe(mode, file, args, env):
+        """spawnvpe(mode, file, args, env) -> integer
+
+Execute file (which is looked for along $PATH) with arguments from
+args in a subprocess with the supplied environment.
+If mode == P_NOWAIT return the pid of the process.
+If mode == P_WAIT return the process's exit code if it exits normally;
+otherwise return -SIG, where SIG is the signal that killed it.. """
         return _spawnvef(mode, file, args, env, execvpe)
 
 if _exists("spawnv"):
@@ -351,9 +378,22 @@
     # but can be easily implemented in Python
 
     def spawnl(mode, file, *args):
+        """spawnl(mode, file, *args) -> integer
+
+Execute file with arguments from args in a subprocess.
+If mode == P_NOWAIT return the pid of the process.
+If mode == P_WAIT return the process's exit code if it exits normally;
+otherwise return -SIG, where SIG is the signal that killed it. """
         return spawnv(mode, file, args)
 
     def spawnle(mode, file, *args):
+        """spawnle(mode, file, *args, env) -> integer
+
+Execute file with arguments from args in a subprocess with the
+supplied environment.
+If mode == P_NOWAIT return the pid of the process.
+If mode == P_WAIT return the process's exit code if it exits normally;
+otherwise return -SIG, where SIG is the signal that killed it. """
         env = args[-1]
         return spawnve(mode, file, args[:-1], env)
 
@@ -361,8 +401,24 @@
     # At the moment, Windows doesn't implement spawnvp[e],
     # so it won't have spawnlp[e] either.
     def spawnlp(mode, file, *args):
+        """spawnlp(mode, file, *args, env) -> integer
+
+Execute file (which is looked for along $PATH) with arguments from
+args in a subprocess with the supplied environment.
+If mode == P_NOWAIT return the pid of the process.
+If mode == P_WAIT return the process's exit code if it exits normally;
+otherwise return -SIG, where SIG is the signal that killed it. """
         return spawnvp(mode, file, args)
 
     def spawnlpe(mode, file, *args):
+        """spawnlpe(mode, file, *args, env) -> integer
+
+Execute file (which is looked for along $PATH) with arguments from
+args in a subprocess with the supplied environment.
+If mode == P_NOWAIT return the pid of the process.
+If mode == P_WAIT return the process's exit code if it exits normally;
+otherwise return -SIG, where SIG is the signal that killed it. """
         env = args[-1]
         return spawnvpe(mode, file, args[:-1], env)
+
+



From brian@garage.co.jp  Thu Apr 20 02:31:56 2000
From: brian@garage.co.jp (Brian Hooper)
Date: Thu, 20 Apr 2000 01:31:56 GMT
Subject: [Patches] Py_BuildValue Unicode support
Message-ID: <20000420013156.77083.qmail@hotmail.com>

This is a multi-part message in MIME format.

------=_NextPart_000_55bdc2c8_12bfb28d$6d08a67
Content-Type: text/plain; format=flowed

Hi there -

There might be a nicer way to do this, but I wanted to be able to
use Py_BuildValue to produce Unicode objects, so I added a few
lines to modsupport to do this.  Each Unicode object requires
two arguments, the raw Py_UNICODE string and the length as an
integer.  I think since all Python Unicode strings are 0 terminated,
it might be nice to make this into "U#" and also allow a plain "U"
to mean a (double) null-terminated Unicode string as input; however,
I'll leave that up to you all to determine (I would personally like
to see this option).

Bye!
Brian Hooper

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

------=_NextPart_000_55bdc2c8_12bfb28d$6d08a67
Content-Type: text/plain; name="modsupport.diff.txt"; format=flowed
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment; filename="modsupport.diff.txt"

diff -c Python/modsupport.c Python.new/modsupport.c
*** Python/modsupport.c Mon Apr 17 13:24:25 2000
--- Python.new/modsupport.c     Wed Apr 19 17:21:52 2000
***************
*** 295,301 ****
                case 'L':
                        return PyLong_FromLongLong((LONG_LONG)va_arg(*p_va, 
LONG
_LONG));
  #endif
!
                case 'f':
                case 'd':
                        return PyFloat_FromDouble(
--- 295,306 ----
                case 'L':
                        return PyLong_FromLongLong((LONG_LONG)va_arg(*p_va, 
LONG
_LONG));
  #endif
!               case 'U':
!               {
!                       Py_UNICODE *u = va_arg(*p_va, Py_UNICODE *);
!                       int ulen = va_arg(*p_va, int);
!                       return PyUnicode_FromUnicode(u, ulen);
!               }
                case 'f':
                case 'd':
                        return PyFloat_FromDouble(


------=_NextPart_000_55bdc2c8_12bfb28d$6d08a67--


From mhammond@skippinet.com.au  Fri Apr 21 12:04:01 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Fri, 21 Apr 2000 21:04:01 +1000
Subject: [Patches] Update to compiler for extended call syntax.
Message-ID: <ECEPKNMJLHAPFFJHDOJBEEFKCJAA.mhammond@skippinet.com.au>

Attached is a set of diffs for the .py compiler that adds support
for the new extended call syntax.

Jeremy - please pay close attention to the stack depth checking and
lineno changes I made - these tests all complete and I think they
are correct, but it is worth mentioning.

NOTE - many of these files have inconsistent indentation.  Time for
another visit from the tabnanny! :-)  I havent addressed the
whitespace at all.

compile.py:
On Windows, use 'nul' instead of '/dev/null'.

test.py:
Use double-quotes for the command-line, as Windows doesnt recognise
singles.


compiler/ast.py:
CallFunc node gets 2 new children to support extended call syntax -
"star_args" (for "*args") and "dstar_args" (for "**args")


compiler/pyassem.py
It appear that self.lnotab is supposed to be responsible for
tracking line numbers, but self.firstlineno was still hanging
around.  Removed self.firstlineno completely.  NOTE - I didnt
actually test that the generated code has the correct line numbers!!

Stack depth tracking appeared a little broken - the checks never
made it beyond the "self.patterns" check - thus, the custom methods
were never called!  Fixed this.

Added support for the new extended call syntax opcodes for depth
calculations.


compiler/pycodegen.py

Added support for the new extended call syntax opcodes.

compiler/transformer.py

Added support for the new extended call syntax.


tests/test_func.py

Added tests of the new extended call syntax.

Release info:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that
I have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce,
distribute, perform and/or display publicly, prepare derivative
versions, and otherwise use this contribution as part of the Python
software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize
others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and
its related documentation.  I further grant CNRI permission to use
my name and other identifying information provided to CNRI by me for
use in connection with the Python software and its related
documentation.

Mark.


begin 666 compiler_diffs.txt
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K<FEN="!T97-T-2@@,BP@*B@P+" Q+" R*2 I#0H@( T*("!T97-T*"D-"@``
`
end



From guido@python.org  Fri Apr 21 19:32:11 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 21 Apr 2000 14:32:11 -0400
Subject: [Patches] Py_BuildValue Unicode support
In-Reply-To: Your message of "Thu, 20 Apr 2000 01:31:56 GMT."
 <20000420013156.77083.qmail@hotmail.com>
References: <20000420013156.77083.qmail@hotmail.com>
Message-ID: <200004211832.OAA16628@eric.cnri.reston.va.us>

> There might be a nicer way to do this, but I wanted to be able to
> use Py_BuildValue to produce Unicode objects, so I added a few
> lines to modsupport to do this.  Each Unicode object requires
> two arguments, the raw Py_UNICODE string and the length as an
> integer.  I think since all Python Unicode strings are 0 terminated,
> it might be nice to make this into "U#" and also allow a plain "U"
> to mean a (double) null-terminated Unicode string as input; however,
> I'll leave that up to you all to determine (I would personally like
> to see this option).

Yes, what you coded here should be called U#.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Fri Apr 21 19:50:02 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 21 Apr 2000 14:50:02 -0400
Subject: [Patches] trailing backslash breaks os.stat
In-Reply-To: Your message of "Tue, 18 Apr 2000 02:13:32 GMT."
 <20000418021332.91759.qmail@hotmail.com>
References: <20000418021332.91759.qmail@hotmail.com>
Message-ID: <200004211850.OAA16684@eric.cnri.reston.va.us>

Brian,

Thanks for this patch.  As presented, it's got one flaw: it modifies
an incoming string object.  I've changed this by making a copy in a
buffer on the stack.

I'm also taking the liberty to swap the tests for trailing backslash
with the test for the drive root (and add a comment to the latter).
Here's the modified patch, which I'll check in in a bit:

Index: posixmodule.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/posixmodule.c,v
retrieving revision 2.130
diff -c -r2.130 posixmodule.c
*** posixmodule.c	2000/04/13 15:20:40	2.130
--- posixmodule.c	2000/04/21 18:48:42
***************
*** 549,556 ****
--- 549,584 ----
  	struct stat st;
  	char *path;
  	int res;
+ 
+ #ifdef MS_WIN32
+       int pathlen;
+       char pathcopy[MAX_PATH];
+ #endif /* MS_WIN32 */
+ 
  	if (!PyArg_ParseTuple(args, format, &path))
  		return NULL;
+ 
+ #ifdef MS_WIN32
+ 	pathlen = strlen(path);
+ 	/* the library call can blow up if the file name is too long! */
+ 	if (pathlen > MAX_PATH) {
+ 		errno = ENAMETOOLONG;
+ 		return posix_error();
+ 	}
+ 
+ 	if ((pathlen > 0) && (path[pathlen-1] == '\\' || path[pathlen-1] == '/')) {
+ 		/* exception for drive root */
+ 		if (!((pathlen == 3) &&
+ 		      (path[1] == ':') &&
+ 		      (path[2] == '\\' || path[2] == '/')))
+ 		{
+ 			strncpy(pathcopy, path, pathlen);
+ 			pathcopy[pathlen-1] = '\0'; /* nuke the trailing backslash */
+ 			path = pathcopy;
+ 		}
+ 	}
+ #endif /* MS_WIN32 */
+ 
  	Py_BEGIN_ALLOW_THREADS
  	res = (*statfunc)(path, &st);
  	Py_END_ALLOW_THREADS

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Fri Apr 21 20:15:20 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 21 Apr 2000 15:15:20 -0400
Subject: [Patches] support for writing pre-formatted ULAW samples in sunau.
In-Reply-To: Your message of "Mon, 17 Apr 2000 15:37:17 +1000."
 <200004170537.PAA03485@mbuna.arbhome.com.au>
References: <200004170537.PAA03485@mbuna.arbhome.com.au>
Message-ID: <200004211915.PAA16728@eric.cnri.reston.va.us>

> The sunau module's write objects assume that the data that's coming
> in is always in linear format. When taking some existing data that's in
> ULAW, and writing it out as ULAW, you have to use audioop.ulaw2lin, then
> let the sunau module convert it back - not exactly the most pleasant 
> approach (I also find a loss of quality). 

You're right, a quick test shows that ulaw2lin + lin2ulaw changes two
of the data bytes!  These are the extreme values 0 and 127; somehow 0
gets mapped to 2 instead of to 0 and 127 gets mapped to 255 instead of
to 127.

There's explicit code in the lin2ulaw code that traps a zero output
byte and replaces it with 2.  This is conditional code and called
"CCITT trap"; in addition, it is enabled with a comment "per MIL-STD".
I wonder what this is about?  Maybe it should be disabled?  It turns
the loudest negative value into the third-loudest negative value.
This doesn't make sense to me.

That explains the change of 0 -> 2.  The change from 127 to 255
shouldn't make a difference: both the 127 and the 255 ULAW value map
to 0 in linear space, and I understand that the reverse mapping has to
pick one.  It shouldn't sound any difference because of this.

> As a workaround for this, I added a formatted_ flag to writeframes() and
> writeframesraw() - if this is set, it doesn't do the audioop conversion.
> 
> Couple of questions:
> 
>   Is this appropriate to add to the library (I think so, but... ?)

I'm not sure.  When you're just copying ULAW data, you might be better
off bypassing the entire sunau module altogether; alternatively, you
might subclass the Au_write class to do what you want.

>   Is this the best way?

Not my preference.  Also note that you didn't bother to add a similar
feature to the readframes method.  My guess is that it's a hack, not a
well-thought-out feature.

>   Which other modules should be similarly modified, if this is appropriate?
>   (aifc, wave)? I'm willing to fix these to match, if this is deemed an 
>   appropriate change...

The afc.py module has similar ULAW support.  But I'm not thrilled by
this patch.

(BTW why did you name the argument with a trailing underscore?)

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Fri Apr 21 21:43:42 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 21 Apr 2000 16:43:42 -0400
Subject: [Patches] Fix for memory leak in _PyResize_Tuple
In-Reply-To: Your message of "Sun, 16 Apr 2000 23:31:39 EDT."
 <14586.34203.790440.916175@bitdiddle.cnri.reston.va.us>
References: <14584.59699.721573.716990@sirius.net.home> <14586.10327.527863.292916@bitdiddle.cnri.reston.va.us> <14586.29863.886669.766779@buffalo.fnal.gov>
 <14586.34203.790440.916175@bitdiddle.cnri.reston.va.us>
Message-ID: <200004212043.QAA17297@eric.cnri.reston.va.us>

Did this ever get resolved?  I personally am in Charles' camp: the
problem he fixes is more serious than the problem that Jeremy wants to
fix at the same time.

I thought about the latter problem when I coded the free list, long
ago.  I decided not to lose sleep over it: it seems more likely that
code which allocates a million small tuples and then releases them all
will do so again soon; or will exit soon.

But Charles' problem surprised me: that a simple loop like that can
leak memory was not at all expected.  So I propose to use Charles'
patch.

--Guido van Rossum (home page: http://www.python.org/~guido/)



From cgw@fnal.gov  Fri Apr 21 21:58:27 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Fri, 21 Apr 2000 15:58:27 -0500 (CDT)
Subject: [Patches] memory leaks triggered by "test_extcall"
Message-ID: <14592.49395.596576.170304@buffalo.fnal.gov>

Running "test_extcall" repeatedly results in memory leaks.

One of these can't be fixed (at least not easily!), it happens since
this code:

def saboteur(**kw):
    kw['x'] = locals()
d = {}
saboteur(a=1, **d)

creates a circular reference - d['x']['d']==d

The others are due to some missing decrefs in ceval.c, fixed by the
patch attached below.  

Note:  I originally wrote this without the "goto", just adding the
missing decref's where needed.  But I think the goto is justified in
keeping the executable code size of ceval as small as possible.

By the way, this section of ceval is not indented according to
ccpy "python" style, which makes editing it in Emacs somewhat
cumbersome.  I avoided re-indenting so as not to send a monster
patch.  What is the rule on re-indenting C sources?


Disclaimer:

                   I confirm that, to the best of my knowledge and belief, this
                   contribution is free of any claims of third parties under
                   copyright, patent or other rights or interests ("claims").  To
                   the extent that I have any such claims, I hereby grant to CNRI a
                   nonexclusive, irrevocable, royalty-free, worldwide license to
                   reproduce, distribute, perform and/or display publicly, prepare
                   derivative versions, and otherwise use this contribution as part
                   of the Python software and its related documentation, or any
                   derivative versions thereof, at no cost to CNRI or its licensed
                   users, and to authorize others to do so.

                   I acknowledge that CNRI may, at its sole discretion, decide
                   whether or not to incorporate this contribution in the Python
                   software and its related documentation.  I further grant CNRI
                   permission to use my name and other identifying information
                   provided to CNRI by me for use in connection with the Python
                   software and its related documentation.

Patch:

Index: Python/ceval.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/ceval.c,v
retrieving revision 2.175
diff -c -r2.175 ceval.c
*** ceval.c	2000/04/10 12:45:10	2.175
--- ceval.c	2000/04/21 20:52:04
***************
*** 1623,1630 ****
  			    if (!PyDict_Check(kwdict)) {
  				PyErr_SetString(PyExc_TypeError,
  					"** argument must be a dictionary");
! 				x = NULL;
! 				break;
  			    }
  			}
  			if (flags & 1) {
--- 1623,1629 ----
  			    if (!PyDict_Check(kwdict)) {
  				PyErr_SetString(PyExc_TypeError,
  					"** argument must be a dictionary");
! 				goto extcall_fail;
  			    }
  			}
  			if (flags & 1) {
***************
*** 1632,1670 ****
  			    if (!PySequence_Check(stararg)) {
  				PyErr_SetString(PyExc_TypeError,
  					"* argument must be a sequence");
! 				x = NULL;
! 				break;
  			    }
  			    /* Convert abstract sequence to concrete tuple */
  			    if (!PyTuple_Check(stararg)) {
  				PyObject *t = NULL;
  				t = PySequence_Tuple(stararg);
  				if (t == NULL) {
! 				    x = NULL;
! 				    break;
  				}
  				Py_DECREF(stararg);
  				stararg = t;
  			    }
  			    nstar = PyTuple_GET_SIZE(stararg);
  			    if (nstar < 0) {
! 				x = NULL;
! 				break;
  			    }
  			}
  			if (nk > 0) {
  			    if (kwdict == NULL) {
  				kwdict = PyDict_New();
  				if (kwdict == NULL) {
! 				    x = NULL;
! 				    break;
  				}
  			    }
  			    else {
  				    PyObject *d = PyDict_Copy(kwdict);
  				    if (d == NULL) {
! 					    x = NULL;
! 					    break;
  				    }
  				    Py_DECREF(kwdict);
  				    kwdict = d;
--- 1631,1664 ----
  			    if (!PySequence_Check(stararg)) {
  				PyErr_SetString(PyExc_TypeError,
  					"* argument must be a sequence");
! 				goto extcall_fail;
  			    }
  			    /* Convert abstract sequence to concrete tuple */
  			    if (!PyTuple_Check(stararg)) {
  				PyObject *t = NULL;
  				t = PySequence_Tuple(stararg);
  				if (t == NULL) {
! 				    goto extcall_fail;
  				}
  				Py_DECREF(stararg);
  				stararg = t;
  			    }
  			    nstar = PyTuple_GET_SIZE(stararg);
  			    if (nstar < 0) {
! 				goto extcall_fail;
  			    }
  			}
  			if (nk > 0) {
  			    if (kwdict == NULL) {
  				kwdict = PyDict_New();
  				if (kwdict == NULL) {
!     				    goto extcall_fail;
  				}
  			    }
  			    else {
  				    PyObject *d = PyDict_Copy(kwdict);
  				    if (d == NULL) {
! 					    goto extcall_fail;
  				    }
  				    Py_DECREF(kwdict);
  				    kwdict = d;
***************
*** 1680,1686 ****
  						 PyString_AsString(key));
  				    Py_DECREF(key);
  				    Py_DECREF(value);
! 				    break;
  				}
  				err = PyDict_SetItem(kwdict, key, value);
  				Py_DECREF(key);
--- 1674,1680 ----
  						 PyString_AsString(key));
  				    Py_DECREF(key);
  				    Py_DECREF(value);
! 				    goto extcall_fail;
  				}
  				err = PyDict_SetItem(kwdict, key, value);
  				Py_DECREF(key);
***************
*** 1689,1695 ****
  				    break;
  			    }
  			    if (err) {
! 				Py_DECREF(kwdict);
  				break;
  			    }
  			}
--- 1683,1693 ----
  				    break;
  			    }
  			    if (err) {
! 			      extcall_fail:
! 				Py_XDECREF(kwdict);
! 				Py_XDECREF(stararg);
! 				Py_DECREF(func);
! 				x=NULL;
  				break;
  			    }
  			}
***************
*** 2382,2387 ****
--- 2380,2386 ----
  	if (kw != NULL && !PyDict_Check(kw)) {
  		PyErr_SetString(PyExc_TypeError,
  				"keyword list must be a dictionary");
+ 		Py_DECREF(arg);
  		return NULL;
  	}
  


From tismer@tismer.com  Fri Apr 21 22:07:22 2000
From: tismer@tismer.com (Christian Tismer)
Date: Fri, 21 Apr 2000 23:07:22 +0200
Subject: [Patches] memory leaks triggered by "test_extcall"
References: <14592.49395.596576.170304@buffalo.fnal.gov>
Message-ID: <3900C30A.439B562C@tismer.com>


Charles G Waldman wrote:
> 
> Running "test_extcall" repeatedly results in memory leaks.
> 
> One of these can't be fixed (at least not easily!), it happens since
> this code:
> 
> def saboteur(**kw):
>     kw['x'] = locals()
> d = {}
> saboteur(a=1, **d)
> 
> creates a circular reference - d['x']['d']==d

Yes, but aren't we going to implant Neil's GC, soon?

ciao - chris

p.s.: Stackless will have it probably after Easter.
-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From guido@python.org  Fri Apr 21 22:09:38 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 21 Apr 2000 17:09:38 -0400
Subject: [Patches] memory leaks triggered by "test_extcall"
In-Reply-To: Your message of "Fri, 21 Apr 2000 23:07:22 +0200."
 <3900C30A.439B562C@tismer.com>
References: <14592.49395.596576.170304@buffalo.fnal.gov>
 <3900C30A.439B562C@tismer.com>
Message-ID: <200004212109.RAA17737@eric.cnri.reston.va.us>

> Charles G Waldman wrote:
> > 
> > Running "test_extcall" repeatedly results in memory leaks.
> > 
> > One of these can't be fixed (at least not easily!), it happens since
> > this code:
> > 
> > def saboteur(**kw):
> >     kw['x'] = locals()
> > d = {}
> > saboteur(a=1, **d)
> > 
> > creates a circular reference - d['x']['d']==d

Christian Tismer:

> Yes, but aren't we going to implant Neil's GC, soon?

Maybe -- I'm a little worried that it will destabilize 1.6.  Or am I
too worried?  Has anyone tried this yet in large amounts of production
code?

It *would* be nice, and Neil even mailed me the release form with a
wet signature.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From hylton@jagunet.com  Fri Apr 21 22:33:15 2000
From: hylton@jagunet.com (Jeremy Hylton)
Date: Fri, 21 Apr 2000 17:33:15 -0400 (EDT)
Subject: [Patches] Re: Update to compiler for extended call syntax.
In-Reply-To: <ECEPKNMJLHAPFFJHDOJBEEFKCJAA.mhammond@skippinet.com.au>
References: <ECEPKNMJLHAPFFJHDOJBEEFKCJAA.mhammond@skippinet.com.au>
Message-ID: <14592.51445.10558.56447@walden>

Thanks for the patches, Mark!  I won't get a chance to look at them
before I leave for the weekend, but I'll get to them first thing next
week. 

Jeremy



From guido@python.org  Fri Apr 21 22:37:20 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 21 Apr 2000 17:37:20 -0400
Subject: [Patches] Please review before applying
In-Reply-To: Your message of "Tue, 11 Apr 2000 14:25:36 +0200."
 <200004111225.OAA31234@localhost.localdomain>
References: <200004111225.OAA31234@localhost.localdomain>
Message-ID: <200004212137.RAA18210@eric.cnri.reston.va.us>

Wanted: reviews for this patch!  It seems usable but (as the author
admits) needs another careful eye before I can apply it.

--Guido van Rossum (home page: http://www.python.org/~guido/)

(Fred Gansevles wrote:)

> This (mega-)patch updates all the modules in the standard-module (all
> Lib/*.py files) to use the new string-methods instead of the old string
> module.

> Most of the changes are simple translations, i.e.:
> 
>     string.split (s, sep)	--> s.split (sep)
> 
> Some of the changes also take the context into account, see:
>     - cgi.py
>     - formatter.py
>     - imaplib.py
> 
> Both 'make test' and 'Lib/compileall.py Lib' produced no error-output
> so I'd expect no big problems.


From skip@mojam.com (Skip Montanaro)  Fri Apr 21 22:42:10 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Fri, 21 Apr 2000 16:42:10 -0500 (CDT)
Subject: [Patches] Please review before applying
In-Reply-To: <200004212137.RAA18210@eric.cnri.reston.va.us>
References: <200004111225.OAA31234@localhost.localdomain>
 <200004212137.RAA18210@eric.cnri.reston.va.us>
Message-ID: <14592.52018.694281.166186@beluga.mojam.com>

    Guido> Wanted: reviews for this patch!  It seems usable but (as the
    Guido> author admits) needs another careful eye before I can apply it.

I don't have time to review it, but I recall thinking when it went by that
distutils might (not "will", "might") be affected by this.  I know Greg is
trying to make sure distutils will still work with 1.5.1.  It's worth
spending a few seconds considering, especially if any of Fred's patches are
to the distutils subdirectory.

Skip


From tismer@tismer.com  Fri Apr 21 22:50:39 2000
From: tismer@tismer.com (Christian Tismer)
Date: Fri, 21 Apr 2000 23:50:39 +0200
Subject: [Patches] Please review before applying
References: <200004111225.OAA31234@localhost.localdomain> <200004212137.RAA18210@eric.cnri.reston.va.us>
Message-ID: <3900CD2F.55E35F16@tismer.com>


Guido van Rossum wrote:
> 
> Wanted: reviews for this patch!  It seems usable but (as the author
> admits) needs another careful eye before I can apply it.

I begun skimming over it and stumbled here:

--- 236,242 ----
          elif requestline[-1:] == '\n':
              requestline = requestline[:-1]
          self.requestline = requestline
!         words = requestline.split()
          if len(words) == 3:
              [command, path, version] = words
              if version[:5] != 'HTTP/':
***************

Shouldn't we be consequent and use the new startswith methods
as well, instead of if version[:5] != 'HTTP/' ?
I think it would be better to do this all in one big shot.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From guido@python.org  Fri Apr 21 22:46:01 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 21 Apr 2000 17:46:01 -0400
Subject: [Patches] Please review before applying
In-Reply-To: Your message of "Fri, 21 Apr 2000 23:50:39 +0200."
 <3900CD2F.55E35F16@tismer.com>
References: <200004111225.OAA31234@localhost.localdomain> <200004212137.RAA18210@eric.cnri.reston.va.us>
 <3900CD2F.55E35F16@tismer.com>
Message-ID: <200004212146.RAA18313@eric.cnri.reston.va.us>

> > Wanted: reviews for this patch!  It seems usable but (as the author
> > admits) needs another careful eye before I can apply it.
> 
> I begun skimming over it and stumbled here:
> 
> --- 236,242 ----
>           elif requestline[-1:] == '\n':
>               requestline = requestline[:-1]
>           self.requestline = requestline
> !         words = requestline.split()
>           if len(words) == 3:
>               [command, path, version] = words
>               if version[:5] != 'HTTP/':
> ***************
> 
> Shouldn't we be consequent and use the new startswith methods
> as well, instead of if version[:5] != 'HTTP/' ?
> I think it would be better to do this all in one big shot.

No, I think it's fine to do it in one fell swoop.  You can either make
many of this kind of changes to one module or make one type of change
to many modules; if you try both, you'll touch too much and start
making mistakes.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From gansevle@cs.utwente.nl  Fri Apr 21 22:50:05 2000
From: gansevle@cs.utwente.nl (Fred Gansevles)
Date: Fri, 21 Apr 2000 23:50:05 +0200
Subject: [Patches] Please review before applying
In-Reply-To: Your message of Fri, 21 Apr 2000 16:42:10 CDT
Message-ID: <200004212150.XAA32590@localhost.localdomain>

> 
>     Guido> Wanted: reviews for this patch!  It seems usable but (as the
>     Guido> author admits) needs another careful eye before I can apply it.
> 
> I don't have time to review it, but I recall thinking when it went by that
> distutils might (not "will", "might") be affected by this.  I know Greg is
> trying to make sure distutils will still work with 1.5.1.  It's worth
> spending a few seconds considering, especially if any of Fred's patches are
> to the distutils subdirectory.

I only patched the Lib/*.py files, no files from:

    Lib/distutils
    Lib/dos-8x3
    Lib/encodings
    Lib/lib-old
    Lib/lib-stdwin
    Lib/lib-tk
    Lib/plat-aix3
    Lib/plat-aix4
    Lib/plat-beos
    Lib/plat-freebsd2
    Lib/plat-freebsd3
    Lib/plat-generic
    Lib/plat-irix5
    Lib/plat-irix6
    Lib/plat-linux1
    Lib/plat-linux2
    Lib/plat-netbsd1
    Lib/plat-next3
    Lib/plat-sunos4
    Lib/plat-sunos5
    Lib/plat-win
    Lib/site-packages
    Lib/test

> 
> Skip
____________________________________________________________________________
 Fred Gansevles <mailto:Fred.Gansevles@cs.utwente.nl> Phone: +31 53 489 4613
        >>> Your one-stop-shop for Linux/WinNT/NetWare <<<
 Org.: Twente University, Fac. of CS, Box 217, 7500 AE Enschede, Netherlands
	    "Bill needs more time to learn Linux" - Steve B.


From gansevle@cs.utwente.nl  Fri Apr 21 22:55:43 2000
From: gansevle@cs.utwente.nl (Fred Gansevles)
Date: Fri, 21 Apr 2000 23:55:43 +0200
Subject: [Patches] Please review before applying
In-Reply-To: Your message of Fri, 21 Apr 2000 17:46:01 EDT
Message-ID: <200004212155.XAA32664@localhost.localdomain>

> > > Wanted: reviews for this patch!  It seems usable but (as the author
> > > admits) needs another careful eye before I can apply it.
> > 
> > I begun skimming over it and stumbled here:
> > 
> > --- 236,242 ----
> >           elif requestline[-1:] == '\n':
> >               requestline = requestline[:-1]
> >           self.requestline = requestline
> > !         words = requestline.split()
> >           if len(words) == 3:
> >               [command, path, version] = words
> >               if version[:5] != 'HTTP/':
> > ***************
> > 
> > Shouldn't we be consequent and use the new startswith methods
> > as well, instead of if version[:5] != 'HTTP/' ?
> > I think it would be better to do this all in one big shot.
> 
> No, I think it's fine to do it in one fell swoop.  You can either make
> many of this kind of changes to one module or make one type of change
> to many modules; if you try both, you'll touch too much and start
> making mistakes.
> 

Mainly this consideration made me ask for a deeper review (and it never
hurts to let someone look closely at a patch to find typo-s :-).
At a few points I've made some changes to the code, more than just a simple
translation.

> --Guido van Rossum (home page: http://www.python.org/~guido/)
____________________________________________________________________________
 Fred Gansevles <mailto:Fred.Gansevles@cs.utwente.nl> Phone: +31 53 489 4613
        >>> Your one-stop-shop for Linux/WinNT/NetWare <<<
 Org.: Twente University, Fac. of CS, Box 217, 7500 AE Enschede, Netherlands
	    "Bill needs more time to learn Linux" - Steve B.


From Jack.Jansen@oratrix.com  Sat Apr 22 00:07:24 2000
From: Jack.Jansen@oratrix.com (Jack Jansen)
Date: Sat, 22 Apr 2000 01:07:24 +0200
Subject: [Patches] Patch to myselect.h
Message-ID: <3900DF2D.AA9AC625@oratrix.nl>

This is a multi-part message in MIME format.
--------------5F9F63D9205AB44C4BEC667D
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Here's a patch to myselect.h. The GUSI 2.0 I/O library (which is used on the
Mac) doesn't use the
special header file for select anymore.
----
I confirm that, to the best of my knowledge and belief, this
                                                       contribution is free of
any claims of third parties under
                                                       copyright, patent or
other rights or interests ("claims").  To
                                                       the extent that I have
any such claims, I hereby grant to CNRI a
                                                       nonexclusive,
irrevocable, royalty-free, worldwide license to
                                                       reproduce, distribute,
perform and/or display publicly, prepare
                                                       derivative versions, and
otherwise use this contribution as part
                                                       of the Python software
and its related documentation, or any
                                                       derivative versions
thereof, at no cost to CNRI or its licensed
                                                       users, and to authorize
others to do so.

                                                       I acknowledge that CNRI
may, at its sole discretion, decide
                                                       whether or not to
incorporate this contribution in the Python
                                                       software and its related
documentation.  I further grant CNRI
                                                       permission to use my
name and other identifying information
                                                       provided to CNRI by me
for use in connection with the Python
                                                       software and its related documentation.
--------------5F9F63D9205AB44C4BEC667D
Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="522A6368";
 name="patch-myselect"
Content-Transfer-Encoding: 7bit
Content-Description: Unknown Document
Content-Disposition: inline;
 filename="patch-myselect"

Index: myselect.h
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Include/myselect.h,v
retrieving revision 2.9
diff -c -r2.9 myselect.h
*** myselect.h	1998/09/28 22:05:22	2.9
--- myselect.h	2000/04/21 23:03:40
***************
*** 51,57 ****
  
  #else /* !HAVE_SYS_SELECT_H */
  
! #ifdef USE_GUSI
  /* If we don't have sys/select the definition may be in unistd.h */
  #include <GUSI.h>
  #endif
--- 51,57 ----
  
  #else /* !HAVE_SYS_SELECT_H */
  
! #ifdef USE_GUSI1
  /* If we don't have sys/select the definition may be in unistd.h */
  #include <GUSI.h>
  #endif


--------------5F9F63D9205AB44C4BEC667D--



From Jack.Jansen@oratrix.com  Sat Apr 22 00:15:15 2000
From: Jack.Jansen@oratrix.com (Jack Jansen)
Date: Sat, 22 Apr 2000 01:15:15 +0200
Subject: [Patches] Patch to import.c and dynload_mac.c
Message-ID: <3900E103.61B9DBBE@oratrix.nl>

This is a multi-part message in MIME format.
--------------6363959758E67113C45E8563
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Here's patches to import.c and dynload_mac.c. The new version of the GUSI i/o
library on the Macintosh
has a few slightly different calls from the old one.
---
I confirm that, to the best of my knowledge and belief, this
                                                       contribution is free of
any claims of third parties under
                                                       copyright, patent or
other rights or interests ("claims").  To
                                                       the extent that I have
any such claims, I hereby grant to CNRI a
                                                       nonexclusive,
irrevocable, royalty-free, worldwide license to
                                                       reproduce, distribute,
perform and/or display publicly, prepare
                                                       derivative versions, and
otherwise use this contribution as part
                                                       of the Python software
and its related documentation, or any
                                                       derivative versions
thereof, at no cost to CNRI or its licensed
                                                       users, and to authorize
others to do so.

                                                       I acknowledge that CNRI
may, at its sole discretion, decide
                                                       whether or not to
incorporate this contribution in the Python
                                                       software and its related
documentation.  I further grant CNRI
                                                       permission to use my
name and other identifying information
                                                       provided to CNRI by me
for use in connection with the Python
                                                       software and its related documentation.
--------------6363959758E67113C45E8563
Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="522A6368";
 name="patch-import"
Content-Transfer-Encoding: 7bit
Content-Description: Unknown Document
Content-Disposition: inline;
 filename="patch-import"

Index: dynload_mac.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/dynload_mac.c,v
retrieving revision 2.3
diff -c -r2.3 dynload_mac.c
*** dynload_mac.c	2000/02/14 17:58:25	2.3
--- dynload_mac.c	2000/04/21 23:00:54
***************
*** 40,46 ****
  #define CFragConnectionID ConnectionID
  #define kLoadCFrag 0x01
  #endif
! #ifdef USE_GUSI
  #include "TFileSpec.h"		/* for Path2FSSpec() */
  #endif
  #include <Files.h>
--- 40,46 ----
  #define CFragConnectionID ConnectionID
  #define kLoadCFrag 0x01
  #endif
! #ifdef USE_GUSI1
  #include "TFileSpec.h"		/* for Path2FSSpec() */
  #endif
  #include <Files.h>
***************
*** 78,84 ****
  	Ptr mainAddr;
  	Str255 errMessage;
  	OSErr err;
! #ifndef USE_GUSI
  	Boolean isfolder, didsomething;
  #endif
  	char buf[512];
--- 78,84 ----
  	Ptr mainAddr;
  	Str255 errMessage;
  	OSErr err;
! #ifndef USE_GUSI1
  	Boolean isfolder, didsomething;
  #endif
  	char buf[512];
***************
*** 87,93 ****
  	CFragSymbolClass class;
  		
  	/* First resolve any aliases to find the real file */
! #ifdef USE_GUSI
  	err = Path2FSSpec(pathname, &libspec);
  #else
  	(void)FSMakeFSSpec(0, 0, Pstring(pathname), &libspec);
--- 87,93 ----
  	CFragSymbolClass class;
  		
  	/* First resolve any aliases to find the real file */
! #ifdef USE_GUSI1
  	err = Path2FSSpec(pathname, &libspec);
  #else
  	(void)FSMakeFSSpec(0, 0, Pstring(pathname), &libspec);
Index: import.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/import.c,v
retrieving revision 2.129
diff -c -r2.129 import.c
*** import.c	2000/02/29 13:59:29	2.129
--- import.c	2000/04/21 23:00:55
***************
*** 1087,1093 ****
  
  #ifdef macintosh
  #include <TextUtils.h>
! #ifdef USE_GUSI
  #include "TFileSpec.h"		/* for Path2FSSpec() */
  #endif
  static int
--- 1087,1093 ----
  
  #ifdef macintosh
  #include <TextUtils.h>
! #ifdef USE_GUSI1
  #include "TFileSpec.h"		/* for Path2FSSpec() */
  #endif
  static int
***************
*** 1095,1101 ****
  {
  	FSSpec fss;
  	OSErr err;
! #ifndef USE_GUSI
  	err = FSMakeFSSpec(0, 0, Pstring(buf), &fss);
  #else
  	/* GUSI's Path2FSSpec() resolves all possible aliases nicely on
--- 1095,1101 ----
  {
  	FSSpec fss;
  	OSErr err;
! #ifndef USE_GUSI1
  	err = FSMakeFSSpec(0, 0, Pstring(buf), &fss);
  #else
  	/* GUSI's Path2FSSpec() resolves all possible aliases nicely on

--------------6363959758E67113C45E8563--



From Jack.Jansen@oratrix.com  Sat Apr 22 00:16:09 2000
From: Jack.Jansen@oratrix.com (Jack Jansen)
Date: Sat, 22 Apr 2000 01:16:09 +0200
Subject: [Patches] Patch to thread.c
Message-ID: <3900E13A.B82AB94C@oratrix.nl>

This is a multi-part message in MIME format.
--------------E2C6AC962320E5BCAA1B5032
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Here's a patch to thread.c. Posix threads are now supported on the Macintosh too.
---
I confirm that, to the best of my knowledge and belief, this
                                                       contribution is free of
any claims of third parties under
                                                       copyright, patent or
other rights or interests ("claims").  To
                                                       the extent that I have
any such claims, I hereby grant to CNRI a
                                                       nonexclusive,
irrevocable, royalty-free, worldwide license to
                                                       reproduce, distribute,
perform and/or display publicly, prepare
                                                       derivative versions, and
otherwise use this contribution as part
                                                       of the Python software
and its related documentation, or any
                                                       derivative versions
thereof, at no cost to CNRI or its licensed
                                                       users, and to authorize
others to do so.

                                                       I acknowledge that CNRI
may, at its sole discretion, decide
                                                       whether or not to
incorporate this contribution in the Python
                                                       software and its related
documentation.  I further grant CNRI
                                                       permission to use my
name and other identifying information
                                                       provided to CNRI by me
for use in connection with the Python
                                                       software and its related documentation.
--------------E2C6AC962320E5BCAA1B5032
Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="522A6368";
 name="patch-thread"
Content-Transfer-Encoding: 7bit
Content-Description: Unknown Document
Content-Disposition: inline;
 filename="patch-thread"

Index: thread.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/thread.c,v
retrieving revision 2.27
diff -c -r2.27 thread.c
*** thread.c	1999/04/07 16:07:23	2.27
--- thread.c	2000/04/21 23:00:55
***************
*** 88,93 ****
--- 88,97 ----
  #define SUN_LWP
  #endif
  
+ #ifdef __MWERKS__
+ #define _POSIX_THREADS
+ #endif
+ 
  #endif /* _POSIX_THREADS */
  
  #ifdef __STDC__


--------------E2C6AC962320E5BCAA1B5032--



From Jack.Jansen@oratrix.com  Sat Apr 22 00:17:11 2000
From: Jack.Jansen@oratrix.com (Jack Jansen)
Date: Sat, 22 Apr 2000 01:17:11 +0200
Subject: [Patches] Patch to tools/begn/bgen/macsupport.py
Message-ID: <3900E178.4807A843@oratrix.nl>

This is a multi-part message in MIME format.
--------------5A9DE3FAC9FAC97D504AD2EE
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

A few new types needed by new API calls.
---
I confirm that, to the best of my knowledge and belief, this
                                                       contribution is free of
any claims of third parties under
                                                       copyright, patent or
other rights or interests ("claims").  To
                                                       the extent that I have
any such claims, I hereby grant to CNRI a
                                                       nonexclusive,
irrevocable, royalty-free, worldwide license to
                                                       reproduce, distribute,
perform and/or display publicly, prepare
                                                       derivative versions, and
otherwise use this contribution as part
                                                       of the Python software
and its related documentation, or any
                                                       derivative versions
thereof, at no cost to CNRI or its licensed
                                                       users, and to authorize
others to do so.

                                                       I acknowledge that CNRI
may, at its sole discretion, decide
                                                       whether or not to
incorporate this contribution in the Python
                                                       software and its related
documentation.  I further grant CNRI
                                                       permission to use my
name and other identifying information
                                                       provided to CNRI by me
for use in connection with the Python
                                                       software and its related documentation.
--------------5A9DE3FAC9FAC97D504AD2EE
Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="522A6368";
 name="patch-macsupport"
Content-Transfer-Encoding: 7bit
Content-Description: Unknown Document
Content-Disposition: inline;
 filename="patch-macsupport"

Index: macsupport.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Tools/bgen/bgen/macsupport.py,v
retrieving revision 1.16
diff -c -r1.16 macsupport.py
*** macsupport.py	2000/01/20 20:49:28	1.16
--- macsupport.py	2000/04/21 22:55:20
***************
*** 15,21 ****
  ScriptCode = Type("ScriptCode", "h")
  Size = Type("Size", "l")
  Style = Type("Style", "b")
! StyleParameter = Type("Style", "h")
  CharParameter = Type("CharParameter", "h")
  TextEncoding = Type("TextEncoding", "l")
  
--- 15,21 ----
  ScriptCode = Type("ScriptCode", "h")
  Size = Type("Size", "l")
  Style = Type("Style", "b")
! StyleParameter = Type("StyleParameter", "h")
  CharParameter = Type("CharParameter", "h")
  TextEncoding = Type("TextEncoding", "l")
  
***************
*** 31,37 ****
  Str255 = OpaqueArrayType("Str255", "PyMac_BuildStr255", "PyMac_GetStr255")
  
  # File System Specifications
! FSSpec = FSSpec_ptr = OpaqueType("FSSpec", "PyMac_BuildFSSpec", "PyMac_GetFSSpec")
  
  # OSType and ResType: 4-byte character strings
  def OSTypeType(typename):
--- 31,38 ----
  Str255 = OpaqueArrayType("Str255", "PyMac_BuildStr255", "PyMac_GetStr255")
  
  # File System Specifications
! FSSpec_ptr = OpaqueType("FSSpec", "PyMac_BuildFSSpec", "PyMac_GetFSSpec")
! FSSpec = OpaqueByValueType("FSSpec", "PyMac_BuildFSSpec", "PyMac_GetFSSpec")
  
  # OSType and ResType: 4-byte character strings
  def OSTypeType(typename):
***************
*** 85,90 ****
--- 86,92 ----
  # Various buffer types
  
  InBuffer = VarInputBufferType('char', 'long', 'l')		# (buf, len)
+ OptionalInBuffer = OptionalVarInputBufferType('char', 'long', 'l')		# (buf, len)
  
  InOutBuffer = HeapInputOutputBufferType('char', 'long', 'l')	# (inbuf, outbuf, len)
  VarInOutBuffer = VarHeapInputOutputBufferType('char', 'long', 'l') # (inbuf, outbuf, &len)
***************
*** 151,159 ****
  # This requires that the OSErr type (defined above) has a non-trivial
  # errorCheck method.
  class OSErrMixIn:
! 	"Mix-in class to treat OSErr return values special"
  	def makereturnvar(self):
! 		if self.returntype is OSErr:
  			return Variable(self.returntype, "_err", ErrorMode)
  		else:
  			return Variable(self.returntype, "_rv", OutMode)
--- 153,161 ----
  # This requires that the OSErr type (defined above) has a non-trivial
  # errorCheck method.
  class OSErrMixIn:
! 	"Mix-in class to treat OSErr/OSStatus return values special"
  	def makereturnvar(self):
! 		if self.returntype.__class__ == OSErrType:
  			return Variable(self.returntype, "_err", ErrorMode)
  		else:
  			return Variable(self.returntype, "_rv", OutMode)


--------------5A9DE3FAC9FAC97D504AD2EE--



From Jack.Jansen@oratrix.com  Sat Apr 22 00:19:08 2000
From: Jack.Jansen@oratrix.com (Jack Jansen)
Date: Sat, 22 Apr 2000 01:19:08 +0200
Subject: [Patches] Patch to socketmodule.c
Message-ID: <3900E1EC.A6CAA0C5@oratrix.nl>

This is a multi-part message in MIME format.
--------------2A0D8785DD8F008DF00A99BD
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

The GUSI 2.0 I/O library used on the Mac uses the socklen_t (unsigned int) for
most size parameters.
Apparently this is part of the UNIX 98 standard, but for now the code is
#ifdeffed with USE_GUSI.
---
I confirm that, to the best of my knowledge and belief, this
                                                       contribution is free of
any claims of third parties under
                                                       copyright, patent or
other rights or interests ("claims").  To
                                                       the extent that I have
any such claims, I hereby grant to CNRI a
                                                       nonexclusive,
irrevocable, royalty-free, worldwide license to
                                                       reproduce, distribute,
perform and/or display publicly, prepare
                                                       derivative versions, and
otherwise use this contribution as part
                                                       of the Python software
and its related documentation, or any
                                                       derivative versions
thereof, at no cost to CNRI or its licensed
                                                       users, and to authorize
others to do so.

                                                       I acknowledge that CNRI
may, at its sole discretion, decide
                                                       whether or not to
incorporate this contribution in the Python
                                                       software and its related
documentation.  I further grant CNRI
                                                       permission to use my
name and other identifying information
                                                       provided to CNRI by me
for use in connection with the Python
                                                       software and its related documentation.
--------------2A0D8785DD8F008DF00A99BD
Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="522A6368";
 name="patch-socketmodule"
Content-Transfer-Encoding: 7bit
Content-Description: Unknown Document
Content-Disposition: inline;
 filename="patch-socketmodule"

Index: socketmodule.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/socketmodule.c,v
retrieving revision 1.103
diff -c -r1.103 socketmodule.c
*** socketmodule.c	2000/04/21 20:33:00	1.103
--- socketmodule.c	2000/04/21 22:59:54
***************
*** 172,178 ****
  #ifdef __BEOS__
  #include <net/netdb.h>
  #else
! #ifndef macintosh
  #include <arpa/inet.h>
  #endif
  #endif
--- 172,178 ----
  #ifdef __BEOS__
  #include <net/netdb.h>
  #else
! #ifndef USE_GUSI1
  #include <arpa/inet.h>
  #endif
  #endif
***************
*** 192,202 ****
  #define O_NDELAY O_NONBLOCK	/* For QNX only? */
  #endif
  
! #ifdef USE_GUSI
  /* fdopen() isn't declared in stdio.h (sigh) */
  #include <GUSI.h>
  #endif
  
  #ifdef USE_SSL
  #include "rsa.h"
  #include "crypto.h"
--- 192,209 ----
  #define O_NDELAY O_NONBLOCK	/* For QNX only? */
  #endif
  
! #ifdef USE_GUSI1
  /* fdopen() isn't declared in stdio.h (sigh) */
  #include <GUSI.h>
  #endif
  
+ #ifndef USE_GUSI2
+ /* GUSI socket library seems to be the only one to define socklen_t.
+ ** And it defines it as an unsigned int too.
+ */
+ typedef int socklen_t;
+ #endif
+ 
  #ifdef USE_SSL
  #include "rsa.h"
  #include "crypto.h"
***************
*** 664,670 ****
  BUILD_FUNC_DEF_2(PySocketSock_accept,PySocketSockObject *,s, PyObject *,args)
  {
  	char addrbuf[256];
! 	int addrlen, newfd;
  	PyObject *sock = NULL;
  	PyObject *addr = NULL;
  	PyObject *res = NULL;
--- 671,678 ----
  BUILD_FUNC_DEF_2(PySocketSock_accept,PySocketSockObject *,s, PyObject *,args)
  {
  	char addrbuf[256];
! 	int newfd;
! 	socklen_t addrlen;
  	PyObject *sock = NULL;
  	PyObject *addr = NULL;
  	PyObject *res = NULL;
***************
*** 808,814 ****
  	int optname;
  	int res;
  	PyObject *buf;
! 	int buflen = 0;
  
  #ifdef __BEOS__
  /* We have incomplete socket support. */
--- 816,822 ----
  	int optname;
  	int res;
  	PyObject *buf;
! 	socklen_t buflen = 0;
  
  #ifdef __BEOS__
  /* We have incomplete socket support. */
***************
*** 822,828 ****
  	
  	if (buflen == 0) {
  		int flag = 0;
! 		int flagsize = sizeof flag;
  		res = getsockopt(s->sock_fd, level, optname,
  				 (ANY *)&flag, &flagsize);
  		if (res < 0)
--- 830,836 ----
  	
  	if (buflen == 0) {
  		int flag = 0;
! 		socklen_t flagsize = sizeof flag;
  		res = getsockopt(s->sock_fd, level, optname,
  				 (ANY *)&flag, &flagsize);
  		if (res < 0)
***************
*** 1010,1016 ****
  BUILD_FUNC_DEF_2(PySocketSock_getsockname,PySocketSockObject *,s, PyObject *,args)
  {
  	char addrbuf[256];
! 	int addrlen, res;
  	if (!PyArg_ParseTuple(args, ":getsockname"))
  		return NULL;
  	if (!getsockaddrlen(s, &addrlen))
--- 1018,1026 ----
  BUILD_FUNC_DEF_2(PySocketSock_getsockname,PySocketSockObject *,s, PyObject *,args)
  {
  	char addrbuf[256];
! 	int res;
! 	socklen_t addrlen;
! 
  	if (!PyArg_ParseTuple(args, ":getsockname"))
  		return NULL;
  	if (!getsockaddrlen(s, &addrlen))
***************
*** 1038,1044 ****
  BUILD_FUNC_DEF_2(PySocketSock_getpeername,PySocketSockObject *,s, PyObject *,args)
  {
  	char addrbuf[256];
! 	int addrlen, res;
  	if (!PyArg_ParseTuple(args, ":getpeername"))
  		return NULL;
  	if (!getsockaddrlen(s, &addrlen))
--- 1048,1056 ----
  BUILD_FUNC_DEF_2(PySocketSock_getpeername,PySocketSockObject *,s, PyObject *,args)
  {
  	char addrbuf[256];
! 	int res;
! 	socklen_t addrlen;
! 
  	if (!PyArg_ParseTuple(args, ":getpeername"))
  		return NULL;
  	if (!getsockaddrlen(s, &addrlen))
***************
*** 1177,1183 ****
  	PyObject *addr = NULL;
  	PyObject *ret = NULL;
  
! 	int addrlen, len, n, flags = 0;
  	if (!PyArg_ParseTuple(args, "i|i:recvfrom", &len, &flags))
  		return NULL;
  	if (!getsockaddrlen(s, &addrlen))
--- 1189,1196 ----
  	PyObject *addr = NULL;
  	PyObject *ret = NULL;
  
! 	int len, n, flags = 0;
! 	socklen_t addrlen;
  	if (!PyArg_ParseTuple(args, "i|i:recvfrom", &len, &flags))
  		return NULL;
  	if (!getsockaddrlen(s, &addrlen))
***************
*** 1882,1888 ****
  	if (!PyArg_ParseTuple(args, "s:inet_aton", &ip_addr)) {
  		return NULL;
  	}
! #ifdef macintosh
  	packed_addr = (long)inet_addr(ip_addr).s_addr;
  #else
  	packed_addr = inet_addr(ip_addr);
--- 1895,1901 ----
  	if (!PyArg_ParseTuple(args, "s:inet_aton", &ip_addr)) {
  		return NULL;
  	}
! #ifdef USE_GUSI1
  	packed_addr = (long)inet_addr(ip_addr).s_addr;
  #else
  	packed_addr = inet_addr(ip_addr);


--------------2A0D8785DD8F008DF00A99BD--



From Jack.Jansen@oratrix.com  Sat Apr 22 00:22:23 2000
From: Jack.Jansen@oratrix.com (Jack Jansen)
Date: Sat, 22 Apr 2000 01:22:23 +0200
Subject: [Patches] Patch to timemodule.c
Message-ID: <3900E2B0.AD10EE6D@oratrix.nl>

On the Macintosh the C library (gmtime, localtime, ctime) uses the ANSI-C
standard epoch of 1-Jan-1900,
but the I/O library (time, stat, etc) uses the Mac epoch of 1-Jan-1904. We
offset the values so the Python
code always sees the Mac standard.
---
I confirm that, to the best of my knowledge and belief, this
                                                       contribution is free of
any claims of third parties under
                                                       copyright, patent or
other rights or interests ("claims").  To
                                                       the extent that I have
any such claims, I hereby grant to CNRI a
                                                       nonexclusive,
irrevocable, royalty-free, worldwide license to
                                                       reproduce, distribute,
perform and/or display publicly, prepare
                                                       derivative versions, and
otherwise use this contribution as part
                                                       of the Python software
and its related documentation, or any
                                                       derivative versions
thereof, at no cost to CNRI or its licensed
                                                       users, and to authorize
others to do so.

                                                       I acknowledge that CNRI
may, at its sole discretion, decide
                                                       whether or not to
incorporate this contribution in the Python
                                                       software and its related
documentation.  I further grant CNRI
                                                       permission to use my
name and other identifying information
                                                       provided to CNRI by me
for use in connection with the Python
                                                       software and its related documentation.


From gstein@lyra.org  Sat Apr 22 02:06:27 2000
From: gstein@lyra.org (Greg Stein)
Date: Fri, 21 Apr 2000 18:06:27 -0700 (PDT)
Subject: [Patches] Patch to timemodule.c
In-Reply-To: <3900E2B0.AD10EE6D@oratrix.nl>
Message-ID: <Pine.LNX.4.10.10004211806180.24901-100000@nebula.lyra.org>

You forgot to append the patch.

-g


On Sat, 22 Apr 2000, Jack Jansen wrote:

> On the Macintosh the C library (gmtime, localtime, ctime) uses the ANSI-C
> standard epoch of 1-Jan-1900,
> but the I/O library (time, stat, etc) uses the Mac epoch of 1-Jan-1904. We
> offset the values so the Python
> code always sees the Mac standard.
> ---
> I confirm that, to the best of my knowledge and belief, this
>                                                        contribution is free of
> any claims of third parties under
>                                                        copyright, patent or
> other rights or interests ("claims").  To
>                                                        the extent that I have
> any such claims, I hereby grant to CNRI a
>                                                        nonexclusive,
> irrevocable, royalty-free, worldwide license to
>                                                        reproduce, distribute,
> perform and/or display publicly, prepare
>                                                        derivative versions, and
> otherwise use this contribution as part
>                                                        of the Python software
> and its related documentation, or any
>                                                        derivative versions
> thereof, at no cost to CNRI or its licensed
>                                                        users, and to authorize
> others to do so.
> 
>                                                        I acknowledge that CNRI
> may, at its sole discretion, decide
>                                                        whether or not to
> incorporate this contribution in the Python
>                                                        software and its related
> documentation.  I further grant CNRI
>                                                        permission to use my
> name and other identifying information
>                                                        provided to CNRI by me
> for use in connection with the Python
>                                                        software and its related documentation.
> 
> _______________________________________________
> Patches mailing list
> Patches@python.org
> http://www.python.org/mailman/listinfo/patches
> 

-- 
Greg Stein, http://www.lyra.org/



From mhammond@skippinet.com.au  Sat Apr 22 04:39:21 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Sat, 22 Apr 2000 13:39:21 +1000
Subject: [Patches] Correction for error in my last mmap.dsp patch.
Message-ID: <ECEPKNMJLHAPFFJHDOJBIEGKCJAA.mhammond@skippinet.com.au>

Ooops - I think I know how this happened.

This patch fixes a linker error on mmap.pyd - the "0x" before the
base address was missing...

Release info:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that
I have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce,
distribute, perform and/or display publicly, prepare derivative
versions, and otherwise use this contribution as part of the Python
software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize
others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and
its related documentation.  I further grant CNRI permission to use
my name and other identifying information provided to CNRI by me for
use in connection with the Python software and its related
documentation.

Mark.


begin 666 mmap.dsp
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:;F0@5&%R9V5T#0HC($5N9"!0<F]J96-T#0H`
`
end



From Fredrik Lundh" <effbot@telia.com  Sat Apr 22 11:17:30 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Sat, 22 Apr 2000 12:17:30 +0200
Subject: [Patches] unicode tweaks, part 1
Message-ID: <002901bfac44$13466f80$34aab5d4@hagrid>

This is a multi-part message in MIME format.

------=_NextPart_000_0023_01BFAC54.BBF6B080
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

this patch removes a few explicit UTF-8 references from the Python
internals, moving them into a single support function:

    PyUnicode_ToString -- converts a unicode string to an 8-bit string
    using the standard encoding (currently UTF-8, just as before).  to
    convert the other way, use PyUnicode_FromObject.

</F>

I confirm that, to the best of my knowledge and belief, this =
contribution
is free of any claims of third parties under copyright, patent or
other rights or interests ("claims").  To the extent that I have
any such claims, I hereby grant to CNRI a nonexclusive, irrevocable,
royalty-free, worldwide license to reproduce, distribute, perform and/or
display publicly, prepare derivative versions, and otherwise use this
contribution as part of the Python software and its related =
documentation,
or any derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
        =20
I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.

------=_NextPart_000_0023_01BFAC54.BBF6B080
Content-Type: application/octet-stream;
	name="unicode-patch-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename="unicode-patch-1"

--- Python/bak/getargs.c	Tue Mar 28 22:09:47 2000
+++ Python/getargs.c	Sat Apr 22 11:47:31 2000
@@ -589,7 +589,8 @@
 			        if (PyString_Check(arg))
 				    *p = PyString_AS_STRING(arg);
 				else if (PyUnicode_Check(arg)) {
-				    arg = PyUnicode_AsUTF8String(arg);
+                                    /* XXX: this leaks, doesn't it? */
+				    arg = PyUnicode_ToString(arg);
 				    if (arg == NULL)
 					return "unicode conversion error";
 				    *p = PyString_AS_STRING(arg);
@@ -634,7 +635,8 @@
 				else if (PyString_Check(arg))
 				  *p = PyString_AsString(arg);
 				else if (PyUnicode_Check(arg)) {
-				  arg = PyUnicode_AsUTF8String(arg);
+                                  /* XXX: this leaks, doesn't it? */
+				  arg = PyUnicode_ToString(arg);
 				  if (arg == NULL)
 				      return "unicode conversion error";
 				  *p = PyString_AS_STRING(arg);
--- Objects/bak/unicodeobject.c	Mon Apr 17 17:57:09 2000
+++ Objects/unicodeobject.c	Sat Apr 22 12:06:03 2000
@@ -367,9 +367,16 @@
 	Py_INCREF(unicode_empty);
 	return (PyObject *)unicode_empty;
     }
+    /* default encoding is UTF-8 */
     return PyUnicode_DecodeUTF8(s, len, "strict");
 }
 
+PyObject *PyUnicode_ToString(register PyObject *obj)
+{
+    /* default encoding is UTF-8 */
+    return PyUnicode_AsUTF8String(obj);
+}
+
 PyObject *PyUnicode_Decode(const char *s,
 			   int size,
 			   const char *encoding,
@@ -2012,7 +2019,7 @@
 	    continue;
 	}
 	if (0 < ch && ch < 256) {
-	    *output++ = ch;
+	    *output++ = (char) ch;
 	    continue;
 	}
 	/* All other characters are considered invalid */

------=_NextPart_000_0023_01BFAC54.BBF6B080--



From Fredrik Lundh" <effbot@telia.com  Sat Apr 22 12:45:33 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Sat, 22 Apr 2000 13:45:33 +0200
Subject: [Patches] unicode tweaks, part 2
References: <002901bfac44$13466f80$34aab5d4@hagrid>
Message-ID: <003301bfac50$469fc460$34aab5d4@hagrid>

this patch must be installed on top of part 1.

changes:

    PyUnicode_ToString is changed to PyUnicode_AsString
    (turned out that the right thing here was to change the
    code to match the name, not vice versa)

    fixed reference leak in getargs (AsUTF8 returns a unique
    reference, not a shared reference)

    internal changes to unicodeobject to make it easier to
    change the default encoding

coming up next: replacing AsUTF8 with ToUTF8 (this includes
plugging a few more leaks...)

</F>

I confirm that, to the best of my knowledge and belief, this =
contribution
is free of any claims of third parties under copyright, patent or
other rights or interests ("claims").  To the extent that I have
any such claims, I hereby grant to CNRI a nonexclusive, irrevocable,
royalty-free, worldwide license to reproduce, distribute, perform and/or
display publicly, prepare derivative versions, and otherwise use this
contribution as part of the Python software and its related =
documentation,
or any derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
        =20
I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.



From Fredrik Lundh" <effbot@telia.com  Sat Apr 22 12:46:41 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Sat, 22 Apr 2000 13:46:41 +0200
Subject: [Patches] unicode tweaks, part 2
Message-ID: <004301bfac50$8842bee0$34aab5d4@hagrid>

This is a multi-part message in MIME format.

------=_NextPart_000_003D_01BFAC61.30B90420
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

(forgot the patch ;-)

this patch must be installed on top of part 1.

changes:

    PyUnicode_ToString is changed to PyUnicode_AsString
    (turned out that the right thing here was to change the
    code to match the name, not vice versa)

    fixed reference leak in getargs (AsUTF8 returns a unique
    reference, not a shared reference)

    internal changes to unicodeobject to make it easier to
    change the default encoding

coming up next: replacing AsUTF8 with ToUTF8 (this includes
plugging a few more leaks...)

</F>

I confirm that, to the best of my knowledge and belief, this =
contribution
is free of any claims of third parties under copyright, patent or
other rights or interests ("claims").  To the extent that I have
any such claims, I hereby grant to CNRI a nonexclusive, irrevocable,
royalty-free, worldwide license to reproduce, distribute, perform and/or
display publicly, prepare derivative versions, and otherwise use this
contribution as part of the Python software and its related =
documentation,
or any derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
        =20
I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.

------=_NextPart_000_003D_01BFAC61.30B90420
Content-Type: application/octet-stream;
	name="unicode-patch-2"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename="unicode-patch-2"

--- Include/bak/unicodeobject.h	Sat Apr 22 13:38:39 2000
+++ Include/unicodeobject.h	Sat Apr 22 13:24:41 2000
@@ -182,7 +182,7 @@
     int length;			/* Length of raw Unicode data in buffer */
     Py_UNICODE *str;		/* Raw Unicode buffer */
     long hash;			/* Hash value; -1 if not set */
-    PyObject *utf8str;		/* UTF-8 encoded version as Python string,
+    PyObject *utf8str;		/* 8-bit encoded version as Python string,
 				   or NULL */
 } PyUnicodeObject;
 
@@ -279,11 +279,16 @@
     register PyObject *obj 	/* Object */
     );
 
-/* Convert Unicode object to 8-bit string using the default 8-bit
-   encoding (currently UTF-8).  Returns NULL if conversion fails,
-   caller must decref */
+/* Return a shared reference to a String object holding an 8-bit
+   encoding of the Unicode object (currently UTF-8).
 
-extern DL_IMPORT(PyObject*) PyUnicode_ToString(
+   Note that the reference is owned by the Unicode object, and is
+   guaranteed to live at least as long as the Unicode object itself.
+   The caller must not decref this reference.
+
+   This function returns NULL if the conversion fails. */
+
+extern DL_IMPORT(PyObject*) PyUnicode_AsString(
     register PyObject *obj 	/* Object */
     );
 
--- Python/bak/getargs.c	Sat Apr 22 11:47:31 2000
+++ Python/getargs.c	Sat Apr 22 13:03:17 2000
@@ -589,8 +589,7 @@
 			        if (PyString_Check(arg))
 				    *p = PyString_AS_STRING(arg);
 				else if (PyUnicode_Check(arg)) {
-                                    /* XXX: this leaks, doesn't it? */
-				    arg = PyUnicode_ToString(arg);
+				    arg = PyUnicode_AsString(arg);
 				    if (arg == NULL)
 					return "unicode conversion error";
 				    *p = PyString_AS_STRING(arg);
@@ -635,8 +634,7 @@
 				else if (PyString_Check(arg))
 				  *p = PyString_AsString(arg);
 				else if (PyUnicode_Check(arg)) {
-                                  /* XXX: this leaks, doesn't it? */
-				  arg = PyUnicode_ToString(arg);
+				  arg = PyUnicode_AsString(arg);
 				  if (arg == NULL)
 				      return "unicode conversion error";
 				  *p = PyString_AS_STRING(arg);
--- Objects/bak/unicodeobject.c	Sat Apr 22 12:06:03 2000
+++ Objects/unicodeobject.c	Sat Apr 22 13:24:21 2000
@@ -371,10 +371,48 @@
     return PyUnicode_DecodeUTF8(s, len, "strict");
 }
 
-PyObject *PyUnicode_ToString(register PyObject *obj)
+/* Return a Python string holding an 8-bit encoded value of the
+   Unicode object. 
+
+   The resulting string is cached in the Unicode object for subsequent
+   usage by this function. The cached version is needed to implement
+   the character buffer interface.
+
+   The refcount of the string is *not* incremented.
+
+*/
+
+static PyObject *as_string(register PyUnicodeObject *self)
 {
+    PyObject *v = self->utf8str;
+
+    if (!v) {
+        
+        /* default encoding is UTF-8 */
+        v = PyUnicode_EncodeUTF8(
+            PyUnicode_AS_UNICODE(self),
+            PyUnicode_GET_SIZE(self),
+            NULL); /* strict */
+
+        /* XXX: note that if we change this to an encoding that may
+           fail, the encode function will be called over and over
+           again.  in that case, utf8str should probably be set to None */
+
+        self->utf8str = v;
+        
+    }
+
+    return v;
+}
+
+PyObject *PyUnicode_AsString(register PyObject *obj)
+{
+    if (!PyUnicode_Check(obj)) {
+        PyErr_BadArgument();
+        return NULL;
+    }
     /* default encoding is UTF-8 */
-    return PyUnicode_AsUTF8String(obj);
+    return as_string((PyUnicodeObject *) obj);
 }
 
 PyObject *PyUnicode_Decode(const char *s,
@@ -709,33 +747,6 @@
     return NULL;
 }
 
-/* Return a Python string holding the UTF-8 encoded value of the
-   Unicode object. 
-
-   The resulting string is cached in the Unicode object for subsequent
-   usage by this function. The cached version is needed to implement
-   the character buffer interface.
-
-   The refcount of the string is *not* incremented.
-
-*/
-
-static
-PyObject *utf8_string(PyUnicodeObject *self,
-		      const char *errors)
-{
-    PyObject *v = self->utf8str;
-
-    if (v)
-        return v;
-    v = PyUnicode_EncodeUTF8(PyUnicode_AS_UNICODE(self),
-			     PyUnicode_GET_SIZE(self),
-			     errors);
-    if (v && errors == NULL)
-        self->utf8str = v;
-    return v;
-}
-
 PyObject *PyUnicode_AsUTF8String(PyObject *unicode)
 {
     PyObject *str;
@@ -744,7 +755,9 @@
         PyErr_BadArgument();
         return NULL;
     }
-    str = utf8_string((PyUnicodeObject *)unicode, NULL);
+    /* XXX: this breaks down if we change the default encoding
+       to something that isn't UTF-8 */
+    str = as_string((PyUnicodeObject *)unicode);
     if (str == NULL)
         return NULL;
     Py_INCREF(str);
@@ -3179,21 +3192,21 @@
 unicode_hash(PyUnicodeObject *self)
 {
     long hash;
-    PyObject *utf8;
+    PyObject *str;
 
-    /* Since Unicode objects compare equal to their UTF-8 string
-       counterparts, they should also use the UTF-8 strings as basis
+    /* Since Unicode objects compare equal to their 8-bit string
+       counterparts, they should also use the 8-bit strings as basis
        for their hash value. This is needed to assure that strings and
        Unicode objects behave in the same way as dictionary
        keys. Unfortunately, this costs some performance and also some
-       memory if the cached UTF-8 representation is not used later
+       memory if the cached 8-bit representation is not used later
        on. */
     if (self->hash != -1)
 	return self->hash;
-    utf8 = utf8_string(self, NULL);
-    if (utf8 == NULL)
+    str = as_string(self);
+    if (str == NULL)
 	return -1;
-    hash = PyObject_Hash(utf8);
+    hash = PyObject_Hash(str);
     if (hash == -1)
 	return -1;
     self->hash = hash;
@@ -3812,7 +3825,9 @@
 static
 PyObject *unicode_str(PyUnicodeObject *self)
 {
-    return PyUnicode_AsUTF8String((PyObject *)self);
+    PyObject* str = as_string(self);
+    Py_XINCREF(str);
+    return str;
 }
 
 static char strip__doc__[] =
@@ -4094,7 +4109,7 @@
 			"accessing non-existent unicode segment");
         return -1;
     }
-    str = utf8_string(self, NULL);
+    str = as_string(self);
     if (str == NULL)
 	return -1;
     *ptr = (void *) PyString_AS_STRING(str);

------=_NextPart_000_003D_01BFAC61.30B90420--



From Fredrik Lundh" <effbot@telia.com  Sat Apr 22 13:44:16 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Sat, 22 Apr 2000 14:44:16 +0200
Subject: [Patches] unicode tweaks, part 3
Message-ID: <006901bfac58$79fd34c0$34aab5d4@hagrid>

This is a multi-part message in MIME format.

------=_NextPart_000_0067_01BFAC69.3C3F7B00
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

this patch must be installed on top of parts 1 and 2.

changes:

    PyUnicode_AsEncodedString and PyUnicode_AsUTF8String
    both use as_string directly, to save some code.

    fixed reference leak in _tkinter.c (AsUTF8 increments the
    reference count, despite what the name implies)

(I think most PyUnicode_As functions should be renamed to
PyUnicode_To, but I'll leave that for another patch)

this is the last part in this round.  the forth part changes
as_string to use 8-bit unicode (and puts utf8_string back
in again), but I won't post that one just yet.

</F>

I confirm that, to the best of my knowledge and belief, this =
contribution
is free of any claims of third parties under copyright, patent or
other rights or interests ("claims").  To the extent that I have
any such claims, I hereby grant to CNRI a nonexclusive, irrevocable,
royalty-free, worldwide license to reproduce, distribute, perform and/or
display publicly, prepare derivative versions, and otherwise use this
contribution as part of the Python software and its related =
documentation,
or any derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
        =20
I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.

------=_NextPart_000_0067_01BFAC69.3C3F7B00
Content-Type: application/octet-stream;
	name="unicode-patch-3"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename="unicode-patch-3"

*** Objects/bak/unicodeobject.c	Sat Apr 22 13:24:21 2000
--- Objects/unicodeobject.c	Sat Apr 22 14:22:43 2000
***************
*** 389,398 ****
      if (!v) {
          
          /* default encoding is UTF-8 */
!         v = PyUnicode_EncodeUTF8(
!             PyUnicode_AS_UNICODE(self),
!             PyUnicode_GET_SIZE(self),
!             NULL); /* strict */
  
          /* XXX: note that if we change this to an encoding that may
             fail, the encode function will be called over and over
--- 389,397 ----
      if (!v) {
          
          /* default encoding is UTF-8 */
!         v = PyUnicode_EncodeUTF8(PyUnicode_AS_UNICODE(self),
!                                  PyUnicode_GET_SIZE(self),
!                                  NULL); /* strict */
  
          /* XXX: note that if we change this to an encoding that may
             fail, the encode function will be called over and over
***************
*** 474,484 ****
          PyErr_BadArgument();
          goto onError;
      }
!     /* Shortcut for the default encoding UTF-8 */
      if ((encoding == NULL || 
  	 (strcmp(encoding, "utf-8") == 0)) &&
! 	errors == NULL)
!         return PyUnicode_AsUTF8String(unicode);
  
      /* Encode via the codec registry */
      v = PyCodec_Encode(unicode, encoding, errors);
--- 473,488 ----
          PyErr_BadArgument();
          goto onError;
      }
! 
!     /* Shortcut for the built-in encoding UTF-8 */
      if ((encoding == NULL || 
  	 (strcmp(encoding, "utf-8") == 0)) &&
! 	errors == NULL) {
!         /* default encoding is UTF-8 */
!         v = as_string((PyUnicodeObject*) unicode);
!         Py_XINCREF(v);
!         return v;
!     }
  
      /* Encode via the codec registry */
      v = PyCodec_Encode(unicode, encoding, errors);
***************
*** 755,766 ****
          PyErr_BadArgument();
          return NULL;
      }
!     /* XXX: this breaks down if we change the default encoding
!        to something that isn't UTF-8 */
      str = as_string((PyUnicodeObject *)unicode);
!     if (str == NULL)
!         return NULL;
!     Py_INCREF(str);
      return str;
  }
  
--- 759,767 ----
          PyErr_BadArgument();
          return NULL;
      }
!     /* default encoding is UTF-8 */
      str = as_string((PyUnicodeObject *)unicode);
!     Py_XINCREF(str);
      return str;
  }
  
*** Modules/bak/_tkinter.c	Fri Mar 31 12:08:53 2000
--- Modules/_tkinter.c	Sat Apr 22 14:15:39 2000
***************
*** 550,560 ****
  		return result;
  	}
  	else if (PyUnicode_Check(value)) {
! 		PyObject* utf8 = PyUnicode_AsUTF8String (value);
  		if (!utf8)
  			return 0;
! 		return Tcl_NewStringObj (PyString_AS_STRING (utf8),
! 					 PyString_GET_SIZE (utf8));
  	}
  	else {
  		PyObject *v = PyObject_Str(value);
--- 550,562 ----
  		return result;
  	}
  	else if (PyUnicode_Check(value)) {
! 		PyObject* utf8 = PyUnicode_AsUTF8String(value);
  		if (!utf8)
  			return 0;
! 		result = Tcl_NewStringObj(PyString_AS_STRING(utf8),
!                                           PyString_GET_SIZE(utf8));
!                 Py_DECREF(utf8);
!                 return result;
  	}
  	else {
  		PyObject *v = PyObject_Str(value);

------=_NextPart_000_0067_01BFAC69.3C3F7B00--



From Fredrik Lundh" <effbot@telia.com  Sat Apr 22 13:56:42 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Sat, 22 Apr 2000 14:56:42 +0200
Subject: [Patches] Re: unicode tweaks
References: <006901bfac58$79fd34c0$34aab5d4@hagrid>
Message-ID: <009201bfac5a$36d62060$34aab5d4@hagrid>

> this patch must be installed on top of parts 1 and 2.

just figured out why my cvs diff script didn't work.  if you
want a "composite patch" instead, drop me a line.

</F>



From Fredrik Lundh" <effbot@telia.com  Sat Apr 22 15:03:54 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Sat, 22 Apr 2000 16:03:54 +0200
Subject: [Patches] urllib: POST with arbitrary mime type
Message-ID: <00c201bfac63$b03d2440$34aab5d4@hagrid>

This is a multi-part message in MIME format.

------=_NextPart_000_00BC_01BFAC74.5C77C480
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

as currently implemented, the urllib module allows you to
post with a single content type (x-www-form-urlencoded).

this patch modifies urllib to accept either a string or a
(subclass of) the urllib.RequestBody class.  usage:

    import urllib

    body =3D urllib.RequestBody(mydata, "text/xml")
    result =3D urllib.urlopen(host, body)

or:

    body =3D urllib.RequestBody(name=3D"fredrik", =
email=3D"effbot@telia.com")
    result =3D urllib.urlopen(host, body)

</F>

I confirm that, to the best of my knowledge and belief, this =
contribution
is free of any claims of third parties under copyright, patent or
other rights or interests ("claims").  To the extent that I have
any such claims, I hereby grant to CNRI a nonexclusive, irrevocable,
royalty-free, worldwide license to reproduce, distribute, perform and/or
display publicly, prepare derivative versions, and otherwise use this
contribution as part of the Python software and its related =
documentation,
or any derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
        =20
I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.


------=_NextPart_000_00BC_01BFAC74.5C77C480
Content-Type: application/octet-stream;
	name="urllib-patch-1"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment;
	filename="urllib-patch-1"

Index: Lib/urllib.py
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
RCS file: /projects/cvsroot/python/dist/src/Lib/urllib.py,v
retrieving revision 1.91
diff -d -c -r1.91 urllib.py
*** urllib.py	2000/02/04 15:28:41	1.91
--- urllib.py	2000/04/22 14:03:03
***************
*** 69,75 ****
--- 69,93 ----
      if _urlopener:
          _urlopener.cleanup()
 =20
+ class RequestBody:
+     """Class to hold a POST request body"""
+     def __init__(self, body, =
type=3D'application/x-www-form-urlencoded'):
+         self.body =3D body
+         self.type =3D type
+     def gettype(self):
+         return self.type
+     def getbody(self):
+         return self.body
 =20
+ class Form(RequestBody):
+     """Class to hold a POST form request"""
+     def __init__(self, dict, **extra):
+ 	dict =3D dict.copy()
+ 	dict.update(extra)
+ 	RequestBody.__init__(self, dict)
+     def getbody(self):
+ 	return urlencode(self.body)
+=20
  ftpcache =3D {}
  class URLopener:
      """Class to open URLs.
***************
*** 256,264 ****
              auth =3D None
          h =3D httplib.HTTP(host)
          if data is not None:
              h.putrequest('POST', selector)
!             h.putheader('Content-type', =
'application/x-www-form-urlencoded')
!             h.putheader('Content-length', '%d' % len(data))
          else:
              h.putrequest('GET', selector)
          if auth: h.putheader('Authorization', 'Basic %s' % auth)
--- 274,285 ----
              auth =3D None
          h =3D httplib.HTTP(host)
          if data is not None:
+             if type(data) is type(''):
+                 data =3D RequestBody(data)
+             body =3D data.getbody()
              h.putrequest('POST', selector)
!             h.putheader('Content-type', data.gettype())
!             h.putheader('Content-length', str(len(body)))
          else:
              h.putrequest('GET', selector)
          if auth: h.putheader('Authorization', 'Basic %s' % auth)
***************
*** 266,272 ****
          for args in self.addheaders: apply(h.putheader, args)
          h.endheaders()
          if data is not None:
!             h.send(data + '\r\n')
          errcode, errmsg, headers =3D h.getreply()
          fp =3D h.getfile()
          if errcode =3D=3D 200:
--- 287,293 ----
          for args in self.addheaders: apply(h.putheader, args)
          h.endheaders()
          if data is not None:
!             h.send(body + '\r\n')
          errcode, errmsg, headers =3D h.getreply()
          fp =3D h.getfile()
          if errcode =3D=3D 200:

------=_NextPart_000_00BC_01BFAC74.5C77C480--



From Fredrik Lundh" <effbot@telia.com  Sat Apr 22 15:07:01 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Sat, 22 Apr 2000 16:07:01 +0200
Subject: [Patches] Re: urllib: POST with arbitrary mime type
References: <00c201bfac63$b03d2440$34aab5d4@hagrid>
Message-ID: <00da01bfac64$090fca00$34aab5d4@hagrid>

bzzt.  the second example was supposed to be:

    body =3D urllib.Form(name=3D"fredrik", email=3D"effbot@telia.com")
    result =3D urllib.urlopen(host, body)

</F>



From mal@lemburg.com  Sat Apr 22 16:04:16 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Sat, 22 Apr 2000 17:04:16 +0200
Subject: [Patches] Py_BuildValue Unicode support
References: <20000420013156.77083.qmail@hotmail.com> <200004211832.OAA16628@eric.cnri.reston.va.us>
Message-ID: <3901BF70.DE9C8F03@lemburg.com>

Guido van Rossum wrote:
> 
> > There might be a nicer way to do this, but I wanted to be able to
> > use Py_BuildValue to produce Unicode objects, so I added a few
> > lines to modsupport to do this.  Each Unicode object requires
> > two arguments, the raw Py_UNICODE string and the length as an
> > integer.  I think since all Python Unicode strings are 0 terminated,
> > it might be nice to make this into "U#" and also allow a plain "U"
> > to mean a (double) null-terminated Unicode string as input; however,
> > I'll leave that up to you all to determine (I would personally like
> > to see this option).
> 
> Yes, what you coded here should be called U#.

Perhaps the naming should be "u#" and "u" rather than "U#" and 
"U" !? This would be in sync with "s#" and "s" for 8-bit strings.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From tismer@tismer.com  Sat Apr 22 17:49:23 2000
From: tismer@tismer.com (Christian Tismer)
Date: Sat, 22 Apr 2000 18:49:23 +0200
Subject: [Patches] Review (was: Please review before applying)
References: <200004111225.OAA31234@localhost.localdomain> <14579.6375.601339.630742@beluga.mojam.com>
Message-ID: <3901D813.D2148C44@tismer.com>

So,
4 hours later, here's my review. (pffffft)

Skip Montanaro wrote:
> 
> >From the changes to Lib/BaseHTTPServer.py:
> 
>     Fred>       # The Python system version, truncated to its first component.
>     Fred> !     sys_version = "Python/" + sys.version.split[0]
> 
> The change here should be to
> 
>     sys.version.split()[0]
> 
> This suggests to me that the BaseHTTPRequestHandler class isn't tested by
> 'make test'.

Found this one, and then saw your message.

Markup of this document:

--- starts a section with a comment, but no change request
+++ also comment, with mild change request
*** starts a section with a real bug.

The latter also suggests to update the test suite!

More bugs or just comments:

---/+++ cgi.py, line 704
you are changing
    plist = map(string.strip, string.splitfields(line, ';'))
into
    plist = map(lambda s: s.strip(), line.split(';'))

I don't the extra lambda-ness and the extra frames generated here
are an advantage.

Guido, while dropping the string module, wouldn't it make sense
to provide a way to get at the "unbound methods" of strings,
like so? :

    plist = map("".string.strip, line.split(';'))
or
    plist = map("".__class__.strip, line.split(';'))

or something alike?

Or is this just a case where string definately should be used?

--- cgi.py, Line 711
minor cosmetic glitch:

             name = p[:i].strip().lower ()
should read
             name = p[:i].strip().lower()

*** dospath.py, line 18
This is a real bug.
I think the tests do not test normcase().

     return s.replace(s, "/", "\\").lower()
should read
     return s.replace("/", "\\").lower()
according to the rest of the style of the source.

+++ formatter.py
This module still uses string. This contradicts Fred's claim of
"""
This (mega-)patch updates all the modules in the standard-module (all
Lib/*.py files) to use the new string-methods instead of the old string
module.
"""

The claim suggests that string is always no longer imported.
I suggest to add a note that the string module is still used
for some constants in some modules.

--- formatter.py, line 171
minor cosmetic glitch:
         data = ' '.join (data.split())
should read
         data = ' '.join(data.split())
according to the rest of the style of the source.

--- gzip.py, line 140
minor cosmetic glitch:
         self.write(' '.join (lines))
should read
         self.write(' '.join(lines))
according to the rest of the style of the source.

--- imaplib.py, line 1001
cosmetic glitch with space after join (no source shown here)

--- macpath.py, line 10
cosmetic glitch with space after lower (no source shown here)

--- multifile.py, nturlpath.py, re.py
uses string, see comment for formatter.py

--- rfc822.py, line 465
cosmetic glitch with space after replace (no source shown here)

--- sgmllib.py, traceback.py, urllib.py
uses string, see comment for formatter.py

--- urllib2.py, line 820
uses lambdaness, see cgi.py

--- urlparse.py, xmllib.py
uses string, see comment for formatter.py


Summary:
--------

I cannot find more than two real bugs.
I suggest to correct these, and perhaps to think
of changing the extra lambdas back to use string.

After that, I'm +1 on applying this patch.

Congrats for the huge work - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From gstein@lyra.org  Sat Apr 22 20:16:02 2000
From: gstein@lyra.org (Greg Stein)
Date: Sat, 22 Apr 2000 12:16:02 -0700 (PDT)
Subject: [Patches] Review (was: Please review before applying)
In-Reply-To: <3901D813.D2148C44@tismer.com>
Message-ID: <Pine.LNX.4.10.10004221212190.22005-100000@nebula.lyra.org>

On Sat, 22 Apr 2000, Christian Tismer wrote:
>...
> ---/+++ cgi.py, line 704
> you are changing
>     plist = map(string.strip, string.splitfields(line, ';'))
> into
>     plist = map(lambda s: s.strip(), line.split(';'))
> 
> I don't the extra lambda-ness and the extra frames generated here
> are an advantage.

No... use the lambda. string.strip is a function that is equal to that
lambda. Therefore, it is better to avoid the string module altogether and
to use the lambda.


Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/



From gstein@lyra.org  Sat Apr 22 20:32:39 2000
From: gstein@lyra.org (Greg Stein)
Date: Sat, 22 Apr 2000 12:32:39 -0700 (PDT)
Subject: [Patches] Please review before applying
In-Reply-To: <200004212146.RAA18313@eric.cnri.reston.va.us>
Message-ID: <Pine.LNX.4.10.10004221230390.22005-100000@nebula.lyra.org>

On Fri, 21 Apr 2000, Guido van Rossum wrote:
>...
> > Shouldn't we be consequent and use the new startswith methods
> > as well, instead of if version[:5] != 'HTTP/' ?
> > I think it would be better to do this all in one big shot.
> 
> No, I think it's fine to do it in one fell swoop.  You can either make
> many of this kind of changes to one module or make one type of change
> to many modules; if you try both, you'll touch too much and start
> making mistakes.

I'm with Guido on this one. If you are making a *large* set of patches,
then it is best to stick with a single type of change. Fred's patch is
generally a simple translation: string.foo(s, ...)  => s.foo(...)

Any more than that, and your brain has to keep changing gears.

A second round of patching can easily go through and make use of the new
.startswith().

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/



From tismer@tismer.com  Sat Apr 22 21:25:39 2000
From: tismer@tismer.com (Christian Tismer)
Date: Sat, 22 Apr 2000 22:25:39 +0200
Subject: [Patches] Please review before applying
References: <Pine.LNX.4.10.10004221230390.22005-100000@nebula.lyra.org>
Message-ID: <39020AC3.B5762381@tismer.com>


Greg Stein wrote:
> 
> On Fri, 21 Apr 2000, Guido van Rossum wrote:
> >...
> > > Shouldn't we be consequent and use the new startswith methods
> > > as well, instead of if version[:5] != 'HTTP/' ?
> > > I think it would be better to do this all in one big shot.
> >
> > No, I think it's fine to do it in one fell swoop.  You can either make
> > many of this kind of changes to one module or make one type of change
> > to many modules; if you try both, you'll touch too much and start
> > making mistakes.
> 
> I'm with Guido on this one. If you are making a *large* set of patches,
> then it is best to stick with a single type of change. Fred's patch is
> generally a simple translation: string.foo(s, ...)  => s.foo(...)
> 
> Any more than that, and your brain has to keep changing gears.
> 
> A second round of patching can easily go through and make use of the new
> .startswith().

Yeah. And you bet how my brain exploded already with
this "small" one. :-)

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From tismer@tismer.com  Sat Apr 22 21:44:37 2000
From: tismer@tismer.com (Christian Tismer)
Date: Sat, 22 Apr 2000 22:44:37 +0200
Subject: [Patches] Trashcan at its best
Message-ID: <39020F35.14D02423@tismer.com>

This is a multi-part message in MIME format.
--------------B49D064F71B53523C347932D
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

This is a patch for a very much improved
version of trashcan - secure destruction of
deeply nested objects.

The affected files are object.c and object.h only.

Improvements:
- does no longer need any extra memory
- has no relationship to tstate
- works in debug mode
- can easily be modified for free threading (hi Greg:)

Side effects:
Trashcan does change the order of object destruction.
Prevending that would be quite an immense effort, as
my attempts have shown. This version works always
the same, with debug mode or not. The slightly
changed destruction order should therefore be no problem.

Algorithm:
While the old idea of delaying the destruction of some
obejcts at a certain recursion level was kept, we now
no longer aloocate an object to hold these objects.
The delayed objects are instead chained together
via their ob_type field. The type is encoded via
ob_refcnt. When it comes to the destruction of the
chain of waiting objects, the topmost object is popped
off the chain and revived with type and refcount 1,
then it gets a normal Py_DECREF.

I am confident that this solution is near optimum
for minimizing side effects and code bloat.

ciao - chris

<blurb>
I confirm that, to the best of my knowledge and belief, this contribution
is free of any claims of third parties under copyright, patent or
other rights or interests ("claims").  To the extent that I have
any such claims, I hereby grant to CNRI a nonexclusive, irrevocable,
royalty-free, worldwide license to reproduce, distribute, perform and/or
display publicly, prepare derivative versions, and otherwise use this
contribution as part of the Python software and its related documentation,
or any derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.
         
I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.
</blurb>
-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com
--------------B49D064F71B53523C347932D
Content-Type: text/plain; charset=us-ascii;
 name="object.h.diff"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="object.h.diff"

*** python/dist/src/Include/object.h	Wed Apr 19 20:52:02 2000
--- python/spc/src/Include/object.h	Sat Apr 22 19:27:52 2000
***************
*** 536,541 ****
--- 536,546 ----
    Also, we could do an exact stack measure then.
    Unfortunately, deallocations also take place when
    the thread state is undefined.
+ 
+   CT 2k0422 complete rewrite.
+   There is no need to allocate new objects.
+   Everything is done vialob_refcnt and ob_type now.
+   Adding support for free-threading should be easy, too.
  */
  
  #define PyTrash_UNWIND_LEVEL 50
***************
*** 551,561 ****
  			_PyTrash_deposit_object((PyObject*)op);\
  		--_PyTrash_delete_nesting; \
  		if (_PyTrash_delete_later && _PyTrash_delete_nesting <= 0) \
! 			_PyTrash_destroy_list(); \
  	} \
  
  extern DL_IMPORT(void) _PyTrash_deposit_object Py_PROTO((PyObject*));
! extern DL_IMPORT(void) _PyTrash_destroy_list Py_PROTO(());
  
  extern DL_IMPORT(int) _PyTrash_delete_nesting;
  extern DL_IMPORT(PyObject *) _PyTrash_delete_later;
--- 556,566 ----
  			_PyTrash_deposit_object((PyObject*)op);\
  		--_PyTrash_delete_nesting; \
  		if (_PyTrash_delete_later && _PyTrash_delete_nesting <= 0) \
! 			_PyTrash_destroy_chain(); \
  	} \
  
  extern DL_IMPORT(void) _PyTrash_deposit_object Py_PROTO((PyObject*));
! extern DL_IMPORT(void) _PyTrash_destroy_chain Py_PROTO(());
  
  extern DL_IMPORT(int) _PyTrash_delete_nesting;
  extern DL_IMPORT(PyObject *) _PyTrash_delete_later;
***************
*** 564,569 ****
--- 569,575 ----
  
  #define xxPy_TRASHCAN_SAFE_BEGIN(op) 
  #define xxPy_TRASHCAN_SAFE_END(op) ;
+ 
  #ifdef __cplusplus
  }
  #endif

--------------B49D064F71B53523C347932D
Content-Type: text/plain; charset=us-ascii;
 name="object.c.diff"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="object.c.diff"

*** python/dist/src/Objects/object.c	Wed Apr 19 20:52:12 2000
--- python/spc/src/Objects/object.c	Sat Apr 22 21:35:04 2000
***************
*** 33,38 ****
--- 33,43 ----
  
  #include "Python.h"
  
+ /* just for trashcan: */
+ #include "compile.h"
+ #include "frameobject.h"
+ #include "traceback.h"
+ 
  #if defined( Py_TRACE_REFS ) || defined( Py_REF_DEBUG )
  DL_IMPORT(long) _Py_RefTotal;
  #endif
***************
*** 822,828 ****
  {
  	destructor dealloc = op->ob_type->tp_dealloc;
  	_Py_ForgetReference(op);
! 	op->ob_type = NULL;
  	(*dealloc)(op);
  }
  
--- 707,714 ----
  {
  	destructor dealloc = op->ob_type->tp_dealloc;
  	_Py_ForgetReference(op);
! 	if (_PyTrash_delete_nesting < PyTrash_UNWIND_LEVEL-1)
! 		op->ob_type = NULL;
  	(*dealloc)(op);
  }
  
***************
*** 1045,1081 ****
  
    CT 2k0325
    added better safe than sorry check for threadstate
  */
  
  int _PyTrash_delete_nesting = 0;
  PyObject * _PyTrash_delete_later = NULL;
  
  void
  _PyTrash_deposit_object(op)
  	PyObject *op;
  {
! 	PyObject *error_type, *error_value, *error_traceback;
! 
! 	if (PyThreadState_GET() != NULL)
! 	    PyErr_Fetch(&error_type, &error_value, &error_traceback);
  
! 	if (!_PyTrash_delete_later)
! 		_PyTrash_delete_later = PyList_New(0);
! 	if (_PyTrash_delete_later)
! 		PyList_Append(_PyTrash_delete_later, (PyObject *)op);
  
! 	if (PyThreadState_GET() != NULL)
! 	    PyErr_Restore(error_type, error_value, error_traceback);
  }
  
  void
! _PyTrash_destroy_list()
  {
  	while (_PyTrash_delete_later) {
  		PyObject *shredder = _PyTrash_delete_later;
! 		_PyTrash_delete_later = NULL;
  		++_PyTrash_delete_nesting;
  		Py_DECREF(shredder);
  		--_PyTrash_delete_nesting;
  	}
  }
--- 931,1007 ----
  
    CT 2k0325
    added better safe than sorry check for threadstate
+ 
+   CT 2k0422
+   complete rewrite. We now build a chain via ob_type
+   and save the limited number of types in ob_refcnt.
+   This is perfect since we don't need any memory.
+   A patch for free-threading would need just a lock.
  */
  
+ #define Py_TRASHCAN_TUPLE       1
+ #define Py_TRASHCAN_LIST        2
+ #define Py_TRASHCAN_DICT        3
+ #define Py_TRASHCAN_FRAME       4
+ #define Py_TRASHCAN_TRACEBACK   5
+ /* extend here if other objects want protection */
+ 
  int _PyTrash_delete_nesting = 0;
+ 
  PyObject * _PyTrash_delete_later = NULL;
  
  void
  _PyTrash_deposit_object(op)
  	PyObject *op;
  {
! 	int typecode;
! 	PyObject *hold = _PyTrash_delete_later;
  
! 	if (PyTuple_Check(op))
! 		typecode = Py_TRASHCAN_TUPLE;
! 	else if (PyList_Check(op))
! 		typecode = Py_TRASHCAN_LIST;
! 	else if (PyDict_Check(op))
! 		typecode = Py_TRASHCAN_DICT;
! 	else if (PyFrame_Check(op))
! 		typecode = Py_TRASHCAN_FRAME;
! 	else if (PyTraceBack_Check(op))
! 		typecode = Py_TRASHCAN_TRACEBACK;
! 	op->ob_refcnt = typecode;
  
! 	op->ob_type = (PyTypeObject*)_PyTrash_delete_later;
! 	_PyTrash_delete_later = op;
  }
  
  void
! _PyTrash_destroy_chain()
  {
  	while (_PyTrash_delete_later) {
  		PyObject *shredder = _PyTrash_delete_later;
! 		_PyTrash_delete_later = (PyObject*) shredder->ob_type;
! 
! 		switch (shredder->ob_refcnt) {
! 		case Py_TRASHCAN_TUPLE:
! 			shredder->ob_type = &PyTuple_Type;
! 			break;
! 		case Py_TRASHCAN_LIST:
! 			shredder->ob_type = &PyList_Type;
! 			break;
! 		case Py_TRASHCAN_DICT:
! 			shredder->ob_type = &PyDict_Type;
! 			break;
! 		case Py_TRASHCAN_FRAME:
! 			shredder->ob_type = &PyFrame_Type;
! 			break;
! 		case Py_TRASHCAN_TRACEBACK:
! 			shredder->ob_type = &PyTraceBack_Type;
! 			break;
! 		}
! 		_Py_NewReference(shredder);
! 
  		++_PyTrash_delete_nesting;
  		Py_DECREF(shredder);
  		--_PyTrash_delete_nesting;
  	}
  }
+ 

--------------B49D064F71B53523C347932D--



From brian@garage.co.jp  Sat Apr 22 23:31:23 2000
From: brian@garage.co.jp (Brian Hooper)
Date: Sat, 22 Apr 2000 22:31:23 GMT
Subject: [Patches] Py_BuildValue Unicode support
Message-ID: <20000422223123.17511.qmail@hotmail.com>

I just did "U" because I seem to remember seeing big "U" in
PyArg_ParseTuple.

Do you like the idea of having a version that can be used without
length as long as the input Unicode data is null-terminated?  From
the playing around that I have been doing up to this point, I think
this would be convenient to have.

--Brian

>From: "M.-A. Lemburg" <mal@lemburg.com>
>To: Guido van Rossum <guido@python.org>
>CC: brian@garage.co.jp, patches@python.org
>Subject: Re: [Patches] Py_BuildValue Unicode support
>Date: Sat, 22 Apr 2000 17:04:16 +0200
>
>Guido van Rossum wrote:
> >
> > > There might be a nicer way to do this, but I wanted to be able to
> > > use Py_BuildValue to produce Unicode objects, so I added a few
> > > lines to modsupport to do this.  Each Unicode object requires
> > > two arguments, the raw Py_UNICODE string and the length as an
> > > integer.  I think since all Python Unicode strings are 0 terminated,
> > > it might be nice to make this into "U#" and also allow a plain "U"
> > > to mean a (double) null-terminated Unicode string as input; however,
> > > I'll leave that up to you all to determine (I would personally like
> > > to see this option).
> >
> > Yes, what you coded here should be called U#.
>
>Perhaps the naming should be "u#" and "u" rather than "U#" and
>"U" !? This would be in sync with "s#" and "s" for 8-bit strings.
>
>--
>Marc-Andre Lemburg
>______________________________________________________________________
>Business:                                      http://www.lemburg.com/
>Python Pages:                           http://www.lemburg.com/python/
>
>

________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com



From mhammond@skippinet.com.au  Sun Apr 23 03:34:59 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Sun, 23 Apr 2000 12:34:59 +1000
Subject: [Patches] RE: Trashcan at its best
In-Reply-To: <39020F35.14D02423@tismer.com>
Message-ID: <ECEPKNMJLHAPFFJHDOJBAEHGCJAA.mhammond@skippinet.com.au>

> This is a patch for a very much improved
> version of trashcan - secure destruction of
> deeply nested objects.

Cool.

> The delayed objects are instead chained together
> via their ob_type field. The type is encoded via
> ob_refcnt. When it comes to the destruction of the
> chain of waiting objects, the topmost object is popped
> off the chain and revived with type and refcount 1,
> then it gets a normal Py_DECREF.

Eeek!!!

Erm - Im glad Im not the one having the final say on this :-) =20

FYI - I did _not_ review the patch, but did apply it.  It appears to =
leave my machine running fine, including the deeply recursive code that =
blew this up in the first place...  Should this status change, I'll let =
you know ;-)

Mark.



From gward@mems-exchange.org  Sun Apr 23 04:13:53 2000
From: gward@mems-exchange.org (Greg Ward)
Date: Sat, 22 Apr 2000 23:13:53 -0400
Subject: [Patches] Review (was: Please review before applying)
In-Reply-To: <Pine.LNX.4.10.10004221212190.22005-100000@nebula.lyra.org>; from gstein@lyra.org on Sat, Apr 22, 2000 at 12:16:02PM -0700
References: <3901D813.D2148C44@tismer.com> <Pine.LNX.4.10.10004221212190.22005-100000@nebula.lyra.org>
Message-ID: <20000422231352.A9588@mems-exchange.org>

On 22 April 2000, Greg Stein said:
> On Sat, 22 Apr 2000, Christian Tismer wrote:
> >...
> > ---/+++ cgi.py, line 704
> > you are changing
> >     plist = map(string.strip, string.splitfields(line, ';'))
> > into
> >     plist = map(lambda s: s.strip(), line.split(';'))
> > 
> > I don't the extra lambda-ness and the extra frames generated here
> > are an advantage.
> 
> No... use the lambda. string.strip is a function that is equal to that
> lambda. Therefore, it is better to avoid the string module altogether and
> to use the lambda.

No way!  Christian is *totally* right here.  I'll let Christian be the
expert on the cost of creating a new stack frame or a new function
object -- he should know -- but two non-propellor-head arguments are
completely compelling:

  * string.split is implemented in C, "lambda s: s.string()" is not

  * understanding "map (string.split, ...)" requires approximately
    zero brainpower; understanding "map (lambda s: ...)" requires
    a good second or so of bouncing around the expression trying
    to figure out what gets plugged in where.

        Greg
-- 
Greg Ward - software developer                gward@mems-exchange.org
MEMS Exchange / CNRI                           voice: +1-703-262-5376
Reston, Virginia, USA                            fax: +1-703-262-5367


From mhammond@skippinet.com.au  Sun Apr 23 05:57:38 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Sun, 23 Apr 2000 14:57:38 +1000
Subject: [Patches] Change to reference-dump procedure at shutdown
Message-ID: <ECEPKNMJLHAPFFJHDOJBGEHHCJAA.mhammond@skippinet.com.au>

This patch only affects debug builds with Py_TRACE_REFS defined.  This =
is the default for debug builds on Windows.

At shutdown, Python offers to dump the remaining references.  This =
upsets a few programs, such as debug COM servers, or any program that =
uses 'sys.executable' to launch a child-process.  The process never =
terminates, as it is waiting to find out if you want the references =
dumped.

Further, simply starting and closing the Python interpreter on Windows =
results in over 2000 references - its unlikely people want to print them =
that often :-)

The patch below only offers to dump the references if the environment =
variable PYTHONDUMPREFS is set.  Otherwise, it still prints the number =
of references, but immediately terminates.

If this patch is acceptable and it is felt this variable should be =
documented, let me know, and I will try and hunt down all the relevant =
places - but IMO, anyone who really needs these references dumped is =
likely to see this code :-)


diff -c -r2.92 pythonrun.c
*** pythonrun.c	2000/04/14 19:13:24	2.92
--- pythonrun.c	2000/04/23 04:49:05
***************
*** 242,248 ****
  #endif
 =20
  #ifdef Py_TRACE_REFS
! 	if (_Py_AskYesNo("Print left references?")) {
  		_Py_PrintReferences(stderr);
  	}
  #endif /* Py_TRACE_REFS */
--- 242,249 ----
  #endif
 =20
  #ifdef Py_TRACE_REFS
! 	if (getenv("PYTHONDUMPREFS") &&=20
! 	    _Py_AskYesNo("Print left references?")) {
  		_Py_PrintReferences(stderr);
  	}
  #endif /* Py_TRACE_REFS */

Release info:

I confirm that, to the best of my knowledge and belief, this =
contribution is free of any claims of third parties under copyright, =
patent or other rights or interests ("claims").  To the extent that I =
have any such claims, I hereby grant to CNRI a nonexclusive, =
irrevocable, royalty-free, worldwide license to reproduce, distribute, =
perform and/or display publicly, prepare derivative versions, and =
otherwise use this contribution as part of the Python software and its =
related documentation, or any derivative versions thereof, at no cost to =
CNRI or its licensed users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or =
not to incorporate this contribution in the Python software and its =
related documentation.  I further grant CNRI permission to use my name =
and other identifying information provided to CNRI by me for use in =
connection with the Python software and its related documentation.

Mark.



From Moshe Zadka <mzadka@geocities.com>  Sun Apr 23 06:11:46 2000
From: Moshe Zadka <mzadka@geocities.com> (Moshe Zadka)
Date: Sun, 23 Apr 2000 08:11:46 +0300 (IDT)
Subject: [Patches] Review (was: Please review before applying)
In-Reply-To: <20000422231352.A9588@mems-exchange.org>
Message-ID: <Pine.GSO.4.10.10004230810030.15546-100000@sundial>

On Sat, 22 Apr 2000, Greg Ward wrote:

> No way!  Christian is *totally* right here.  I'll let Christian be the
> expert on the cost of creating a new stack frame or a new function
> object -- he should know -- but two non-propellor-head arguments are
> completely compelling:
> 
>   * string.split is implemented in C, "lambda s: s.string()" is not

Ummm....Isn't string.split implemented in Python, and just calls the 
.split method of it's first argument? oldstring.split is implemented
in C, but is deprecated.

>   * understanding "map (string.split, ...)" requires approximately
>     zero brainpower; understanding "map (lambda s: ...)" requires
>     a good second or so of bouncing around the expression trying
>     to figure out what gets plugged in where.

That's probably a better excuse <wink>

anything-to-get-rid-of-lambdas-ly y'rs, Z.
--
Moshe Zadka <mzadka@geocities.com>. 
http://www.oreilly.com/news/prescod_0300.html
http://www.linux.org.il -- we put the penguin in .com



From skip@mojam.com (Skip Montanaro)  Sun Apr 23 13:23:50 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Sun, 23 Apr 2000 07:23:50 -0500 (CDT)
Subject: [Patches] Review (was: Please review before applying)
In-Reply-To: <20000422231352.A9588@mems-exchange.org>
References: <3901D813.D2148C44@tismer.com>
 <Pine.LNX.4.10.10004221212190.22005-100000@nebula.lyra.org>
 <20000422231352.A9588@mems-exchange.org>
Message-ID: <14594.60246.161584.33770@beluga.mojam.com>

    Greg>   * string.split is implemented in C, "lambda s: s.string()" is not

If you look at string.py now you'll see that string.split actually calls a
Python function (essentially the same function as executing the lambda
statement generates), unlike the old days where string would have imported
split from the strop extension module.  In short, calling string.* generally
will incur an extra function call than in 1.5.

    Greg>   * understanding "map (string.split, ...)" requires approximately
    Greg>     zero brainpower; understanding "map (lambda s: ...)" requires
    Greg>     a good second or so of bouncing around the expression trying
    Greg>     to figure out what gets plugged in where.

Since this particular code is in cgi.parse_header, which isn't going to be a
performance bottleneck (leave that to forking the CGI script or doing the
script's real work), if clarity is in question let me suggest that the map
be replaced with a for loop, something like:

    plist = line.split(';')
    for i in range(len(plist)):
	plist[i] = plist[i].split()

-- 
Skip Montanaro | http://www.mojam.com/
skip@mojam.com | http://www.musi-cal.com/


From tismer@tismer.com  Sun Apr 23 15:43:03 2000
From: tismer@tismer.com (Christian Tismer)
Date: Sun, 23 Apr 2000 16:43:03 +0200
Subject: [Patches] Re: Trashcan at its best
References: <ECEPKNMJLHAPFFJHDOJBAEHGCJAA.mhammond@skippinet.com.au>
Message-ID: <39030BF7.D353146E@tismer.com>


Mark Hammond wrote:
> 
> > This is a patch for a very much improved
> > version of trashcan - secure destruction of
> > deeply nested objects.
> 
> Cool.
> 
> > The delayed objects are instead chained together
> > via their ob_type field. The type is encoded via
> > ob_refcnt. When it comes to the destruction of the
> > chain of waiting objects, the topmost object is popped
> > off the chain and revived with type and refcount 1,
> > then it gets a normal Py_DECREF.
> 
> Eeek!!!
> 
> Erm - Im glad Im not the one having the final say on this :-)

Oh, this isn't so different from the first version:
It appended dead objects to a list, therefore reviving them.
Now this is done explicitly, by _PyNew_Reference, that's
just fine.
Or do you mean that refcount abuse? It is just an integer,
and I switch with a case. No, I think this belongs to
some of my cleanest code snippets.

> FYI - I did _not_ review the patch, but did apply it.  It appears to leave my machine running fine, including the deeply recursive code that blew this up in the first place...  Should this status change, I'll let you know ;-)

I knew that it would work. I did an extra debug build for
you and tested things hard until I sent it out.
Especially the right condition on when to clear the ob_type
field was not so obvious to find in the first place.

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From guido@python.org  Sun Apr 23 17:02:42 2000
From: guido@python.org (Guido van Rossum)
Date: Sun, 23 Apr 2000 12:02:42 -0400
Subject: [Patches] Py_BuildValue Unicode support
In-Reply-To: Your message of "Sat, 22 Apr 2000 22:31:23 GMT."
 <20000422223123.17511.qmail@hotmail.com>
References: <20000422223123.17511.qmail@hotmail.com>
Message-ID: <200004231602.MAA19202@eric.cnri.reston.va.us>

> I just did "U" because I seem to remember seeing big "U" in
> PyArg_ParseTuple.

The difference there is that it takes a Unicode *object*, like O and S
-- not a unicode data pointer, like s#.

> Do you like the idea of having a version that can be used without
> length as long as the input Unicode data is null-terminated?  From
> the playing around that I have been doing up to this point, I think
> this would be convenient to have.

Sure.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From gstein@lyra.org  Mon Apr 24 04:49:45 2000
From: gstein@lyra.org (Greg Stein)
Date: Sun, 23 Apr 2000 20:49:45 -0700 (PDT)
Subject: [Patches] Review (was: Please review before applying)
In-Reply-To: <14594.60246.161584.33770@beluga.mojam.com>
Message-ID: <Pine.LNX.4.10.10004232044010.18820-100000@nebula.lyra.org>

On Sun, 23 Apr 2000, Skip Montanaro wrote:
>     Greg>   * string.split is implemented in C, "lambda s: s.string()" is not
> 
> If you look at string.py now you'll see that string.split actually calls a
> Python function (essentially the same function as executing the lambda
> statement generates), unlike the old days where string would have imported
> split from the strop extension module.  In short, calling string.* generally
> will incur an extra function call than in 1.5.

Exactly. This is why I recommended avoiding string.*

(note the lambda in question is to call s.strip())

>     Greg>   * understanding "map (string.split, ...)" requires approximately
>     Greg>     zero brainpower; understanding "map (lambda s: ...)" requires
>     Greg>     a good second or so of bouncing around the expression trying
>     Greg>     to figure out what gets plugged in where.
> 
> Since this particular code is in cgi.parse_header, which isn't going to be a
> performance bottleneck (leave that to forking the CGI script or doing the
> script's real work), if clarity is in question let me suggest that the map
> be replaced with a for loop, something like:
> 
>     plist = line.split(';')
>     for i in range(len(plist)):
> 	plist[i] = plist[i].split()

And it should generally be faster, too, since there aren't function
invocations in the main loop.


The main point is to avoid string.*.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/




From tismer@trixie.triqs.com  Mon Apr 24 13:49:11 2000
From: tismer@trixie.triqs.com (Christian Tismer)
Date: Mon, 24 Apr 2000 14:49:11 +0200
Subject: [Patches] Review (was: Please review before applying)
References: <Pine.LNX.4.10.10004232044010.18820-100000@nebula.lyra.org>
Message-ID: <390442C7.F30179D9@trixie.triqs.com>


Greg Stein wrote:
...
> The main point is to avoid string.*.

Agreed. Also replacing map by a loop might not even be slower.
What remains as open question: Several modules need access
to string constants, and they therefore still have to import
string.
Is there an elegant solution to this?
That's why i asked for some way to access "".__class__ or
whatever, to get into some common namespace with the constants.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From guido@python.org  Mon Apr 24 14:50:57 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 24 Apr 2000 09:50:57 -0400
Subject: [Patches] Change to reference-dump procedure at shutdown
In-Reply-To: Your message of "Sun, 23 Apr 2000 14:57:38 +1000."
 <ECEPKNMJLHAPFFJHDOJBGEHHCJAA.mhammond@skippinet.com.au>
References: <ECEPKNMJLHAPFFJHDOJBGEHHCJAA.mhammond@skippinet.com.au>
Message-ID: <200004241350.JAA20387@eric.cnri.reston.va.us>

> This patch only affects debug builds with Py_TRACE_REFS defined.  This is the default for debug builds on Windows.
> 
> At shutdown, Python offers to dump the remaining references.  This upsets a few programs, such as debug COM servers, or any program that uses 'sys.executable' to launch a child-process.  The process never terminates, as it is waiting to find out if you want the references dumped.
> 
> Further, simply starting and closing the Python interpreter on Windows results in over 2000 references - its unlikely people want to print them that often :-)
> 
> The patch below only offers to dump the references if the environment variable PYTHONDUMPREFS is set.  Otherwise, it still prints the number of references, but immediately terminates.
> 
> If this patch is acceptable and it is felt this variable should be documented, let me know, and I will try and hunt down all the relevant places - but IMO, anyone who really needs these references dumped is likely to see this code :-)

Mark, shouldn't this be enclosed in an #ifdef MS_WINDOWS bracket?

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Mon Apr 24 15:40:36 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 24 Apr 2000 10:40:36 -0400
Subject: [Patches] urllib: POST with arbitrary mime type
In-Reply-To: Your message of "Sat, 22 Apr 2000 16:03:54 +0200."
 <00c201bfac63$b03d2440$34aab5d4@hagrid>
References: <00c201bfac63$b03d2440$34aab5d4@hagrid>
Message-ID: <200004241440.KAA23325@eric.cnri.reston.va.us>

One suggestion -- rather than testing for type(data)==type(''),
wouldn't it be better to test for isinstance(data, RequestBody)?
Also, it seems that your Form class is broken.  It passes a dict as
the body argument the the base class constructor, but it is later used
as a string.

I suggest that you test this one more extensively.  Perhaps add some
testcases to test_urllib.py?

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Mon Apr 24 15:42:39 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 24 Apr 2000 10:42:39 -0400
Subject: [Patches] unicode tweaks, part 3
In-Reply-To: Your message of "Sat, 22 Apr 2000 14:44:16 +0200."
 <006901bfac58$79fd34c0$34aab5d4@hagrid>
References: <006901bfac58$79fd34c0$34aab5d4@hagrid>
Message-ID: <200004241442.KAA23592@eric.cnri.reston.va.us>

In private mail, Marc-Andre mentioned that he thinks Fredrik's patches
aren't quite ready for prime time.  I suggest that the two hash this
out on python-dev before I accept the patch.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From guido@python.org  Mon Apr 24 15:53:13 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 24 Apr 2000 10:53:13 -0400
Subject: [Patches] Patch to socketmodule.c
In-Reply-To: Your message of "Sat, 22 Apr 2000 01:19:08 +0200."
 <3900E1EC.A6CAA0C5@oratrix.nl>
References: <3900E1EC.A6CAA0C5@oratrix.nl>
Message-ID: <200004241453.KAA25772@eric.cnri.reston.va.us>

> Subject: [Patches] Patch to socketmodule.c
> The GUSI 2.0 I/O library used on the Mac uses the socklen_t (unsigned int) for
> most size parameters.
> Apparently this is part of the UNIX 98 standard, but for now the code is
> #ifdeffed with USE_GUSI.

Alas, it breaks on Solaris (SunOS 5.7).  I guess I'll have to add a
test for it to configure.in.  Sigh.  I'll check in the correct fix.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From Fredrik Lundh" <effbot@telia.com  Mon Apr 24 18:41:17 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Mon, 24 Apr 2000 19:41:17 +0200
Subject: [Patches] urllib: POST with arbitrary mime type
References: <00c201bfac63$b03d2440$34aab5d4@hagrid>  <200004241440.KAA23325@eric.cnri.reston.va.us>
Message-ID: <00e301bfae14$4d21c160$34aab5d4@hagrid>

Guido wrote:
> One suggestion -- rather than testing for type(data)=3D=3Dtype(''),
> wouldn't it be better to test for isinstance(data, RequestBody)?

that would make it impossible to drop in an object implementing
the same protocol (gettype and getbody) -- and the old code
only worked for strings anyway...

but sure, I can change this.  how about:

    if isinstance(data, RequestBody):
        body =3D data
    else:
        body =3D RequestBody(str(data))

> Also, it seems that your Form class is broken.  It passes a dict as
> the body argument the the base class constructor, but it is later used
> as a string.

note that it also overrides the getbody() function.

but alright, I probably don't need to optimize for the case where
httplib.HOST fails to connect ;-)

btw, what about making the class even more flexible?  how about
something like:

    class RequestBody:
        ...
        def gettype(self):
            return "application/x-www-form-urlencoded"
        def getrequest(self):
            return "POST"
        def sendheader(self, http):
            http.putrequest(self.getrequest())
            http.putheader("Content-type", self.gettype())
            http.putheader("Content-length", str(self.bytes))
        def sendbody(self, http) -> send
            http.send(self.data)

> I suggest that you test this one more extensively.  Perhaps add some
> testcases to test_urllib.py?

gotta create that one first, right?

</F>



From guido@python.org  Mon Apr 24 18:42:30 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 24 Apr 2000 13:42:30 -0400
Subject: [Patches] urllib: POST with arbitrary mime type
In-Reply-To: Your message of "Mon, 24 Apr 2000 19:41:17 +0200."
 <00e301bfae14$4d21c160$34aab5d4@hagrid>
References: <00c201bfac63$b03d2440$34aab5d4@hagrid> <200004241440.KAA23325@eric.cnri.reston.va.us>
 <00e301bfae14$4d21c160$34aab5d4@hagrid>
Message-ID: <200004241742.NAA29359@eric.cnri.reston.va.us>

Yes.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From cgw@fnal.gov  Mon Apr 24 20:38:20 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Mon, 24 Apr 2000 14:38:20 -0500 (CDT)
Subject: [Patches] why not just apply Garbage Collection patch?
Message-ID: <14596.41644.182242.918286@buffalo.fnal.gov>

Neil's GC is an excellent piece of work, and I hope it makes it into
1.6.  I spent a lot of time this Easter weekend reading the code and
playing around with it.

I'll do what I can to test it on large volumes of production code  (I
have plenty of machines at my disposal...)

Unfortunately, with my recent changes to tupleobject.c and Christian's
recent changes to the Trashcan, the gc patch no longer applies
cleanly, and after hand-patching, it doesn't seem to work (Python
hangs forever on startup, after printing a lone

  gc: collectable <Object 0x8189fb4>

message.

When I get some time, I'll try to find out why this is happening.

In the meanwhile, I propose that Neil's patch be applied to the main
CVS sources.  Since you have to enable it on the configure line with
"--with-cycle-gc" it seems harmless enough to just check the patches
in.  Then we don't have the problem of the GC stuff getting
out-of-sync with all the other changes that are happening.
Furthermore, the earlier it gets into the alpha process, the more
people will be encouraged to play around and test it out.  If it turns
out to be a disaster (I'm not expecting this!) we can just back it all
out, right?  Everything is already protected with "#ifdef
WITH_CYCLE_GC" anyhow, right?

Just my $0.002,

     	-C


From guido@python.org  Mon Apr 24 21:13:08 2000
From: guido@python.org (Guido van Rossum)
Date: Mon, 24 Apr 2000 16:13:08 -0400
Subject: [Patches] why not just apply Garbage Collection patch?
In-Reply-To: Your message of "Mon, 24 Apr 2000 14:38:20 CDT."
 <14596.41644.182242.918286@buffalo.fnal.gov>
References: <14596.41644.182242.918286@buffalo.fnal.gov>
Message-ID: <200004242013.QAA29938@eric.cnri.reston.va.us>

> Neil's GC is an excellent piece of work, and I hope it makes it into
> 1.6.  I spent a lot of time this Easter weekend reading the code and
> playing around with it.
> 
> I'll do what I can to test it on large volumes of production code  (I
> have plenty of machines at my disposal...)

Thanks.  I haven't looked at it lately.  What do you think of binary
compatibility?  Could an extension built before this was patched in
get in trouble because of changed object lay-out?

> Unfortunately, with my recent changes to tupleobject.c and Christian's
> recent changes to the Trashcan, the gc patch no longer applies
> cleanly, and after hand-patching, it doesn't seem to work (Python
> hangs forever on startup, after printing a lone
> 
>   gc: collectable <Object 0x8189fb4>
> 
> message.
> 
> When I get some time, I'll try to find out why this is happening.

Maybe Neil or Christian can lend a hand?  The trashcan stuff is
definitely hackish. :-)

> In the meanwhile, I propose that Neil's patch be applied to the main
> CVS sources.  Since you have to enable it on the configure line with
> "--with-cycle-gc" it seems harmless enough to just check the patches
> in.  Then we don't have the problem of the GC stuff getting
> out-of-sync with all the other changes that are happening.
> Furthermore, the earlier it gets into the alpha process, the more
> people will be encouraged to play around and test it out.  If it turns
> out to be a disaster (I'm not expecting this!) we can just back it all
> out, right?  Everything is already protected with "#ifdef
> WITH_CYCLE_GC" anyhow, right?

I'd rather wait until you or Neil submits a new version of the patch
that (1) applies cleanly to the current CVS and (2) actually runs.
Plus, I owe Vladimir Marangozov a green light to generate a new
version of his memory management patches; that will also interfere
with the GC patches (and make a few things easier, I believe).
Vladimir's patches should preferably be applied before Neil's.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From mhammond@skippinet.com.au  Tue Apr 25 01:33:22 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Tue, 25 Apr 2000 10:33:22 +1000
Subject: [Patches] Change to reference-dump procedure at shutdown
In-Reply-To: <200004241350.JAA20387@eric.cnri.reston.va.us>
Message-ID: <ECEPKNMJLHAPFFJHDOJBMEIOCJAA.mhammond@skippinet.com.au>

> > If this patch is acceptable and it is felt this=20
> variable should be documented, let me know, and I will=20
> try and hunt down all the relevant places - but IMO,=20
> anyone who really needs these references dumped is likely=20
> to see this code :-)
>=20
> Mark, shouldn't this be enclosed in an #ifdef MS_WINDOWS bracket?

Quite possibly.  However, I was assuming that given there are over 2000 =
references outstanding when just starting and stopping the interpreter, =
the behaviour made sense for _all_ platforms, not just Windows.

Of course, just getting it for Windows still suits me fine...

Mark.



From mhammond@skippinet.com.au  Tue Apr 25 02:38:40 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Tue, 25 Apr 2000 11:38:40 +1000
Subject: [Patches] MSVC project change to include winsound as dependency
Message-ID: <ECEPKNMJLHAPFFJHDOJBGEJBCJAA.mhammond@skippinet.com.au>

This is a multi-part message in MIME format.

------=_NextPart_000_005D_01BFAEAA.CDBF1A70
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

This updates the MSVC workspace to ensure the winsound project depends =
on the core DLL.

Ive attached the file complete, rather than a diff, as it is checked in =
as binary.

Release info:

I confirm that, to the best of my knowledge and belief, this =
contribution is free of any claims of third parties under copyright, =
patent or other rights or interests ("claims").  To the extent that I =
have any such claims, I hereby grant to CNRI a nonexclusive, =
irrevocable, royalty-free, worldwide license to reproduce, distribute, =
perform and/or display publicly, prepare derivative versions, and =
otherwise use this contribution as part of the Python software and its =
related documentation, or any derivative versions thereof, at no cost to =
CNRI or its licensed users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or =
not to incorporate this contribution in the Python software and its =
related documentation.  I further grant CNRI permission to use my name =
and other identifying information provided to CNRI by me for use in =
connection with the Python software and its related documentation.

Mark.

------=_NextPart_000_005D_01BFAEAA.CDBF1A70
Content-Type: application/octet-stream;
	name="pcbuild.dsw"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment;
	filename="pcbuild.dsw"

Microsoft Developer Studio Workspace File, Format Version 6.00
# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE!

#########################################################################=
######

Project: "_socket"=3D".\_socket.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "_sre"=3D".\_sre.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "_tkinter"=3D".\_tkinter.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "bsddb"=3D".\bsddb.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "mmap"=3D".\mmap.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "parser"=3D".\parser.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "pyexpat"=3D".\pyexpat.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "python"=3D".\python.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "python16"=3D".\python16.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
}}}

#########################################################################=
######

Project: "pythonw"=3D".\pythonw.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "select"=3D".\select.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "unicodedata"=3D".\unicodedata.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "winreg"=3D".\winreg.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "winsound"=3D".\winsound.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Project: "zlib"=3D".\zlib.dsp" - Package Owner=3D<4>

Package=3D<5>
{{{
}}}

Package=3D<4>
{{{
    Begin Project Dependency
    Project_Dep_Name python16
    End Project Dependency
}}}

#########################################################################=
######

Global:

Package=3D<5>
{{{
}}}

Package=3D<3>
{{{
}}}

#########################################################################=
######


------=_NextPart_000_005D_01BFAEAA.CDBF1A70--



From tim_one@email.msn.com  Tue Apr 25 06:43:54 2000
From: tim_one@email.msn.com (Tim Peters)
Date: Tue, 25 Apr 2000 01:43:54 -0400
Subject: [Patches] memory leaks triggered by "test_extcall"
In-Reply-To: <200004212109.RAA17737@eric.cnri.reston.va.us>
Message-ID: <000501bfae79$40004aa0$152d153f@tim>

[Christian Tismer]
> Yes, but aren't we going to implant Neil's GC, soon?

[Guido]
> Maybe -- I'm a little worried that it will destabilize 1.6.  Or am I
> too worried?

IIRC, it was all #ifdef'ed -- leave it off by default, if need be, but get
it out there.  People will play with it if it's easy to enable, and you'll
be worried for just as long as people *don't* play with it <wink>.




From mwh21@cam.ac.uk  Tue Apr 25 09:00:45 2000
From: mwh21@cam.ac.uk (Michael Hudson)
Date: Tue, 25 Apr 2000 09:00:45 +0100 (BST)
Subject: [Patches] more docstrings for os
Message-ID: <Pine.LNX.4.10.10004250857350.20527-100000@localhost.localdomain>

I think that after this patch, all objects in the os module (with names
that don't start with "_") that can have docstrings, do, on Linux at
least.

Also fix a nit in one of my spawn* docstrings.

Cheers,
M.

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

Index: os.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/os.py,v
retrieving revision 1.32
diff -u -r1.32 os.py
--- os.py	2000/04/21 18:35:36	1.32
+++ os.py	2000/04/25 07:57:25
@@ -183,23 +183,51 @@
     environ = {}
 
 def execl(file, *args):
+    """execl(file, *args)
+
+    Execute the executable file with argument list args, replacing the
+    current process. """
     execv(file, args)
 
 def execle(file, *args):
+    """execle(file, *args, env)
+
+    Execute the executable file with argument list args and
+    environment env, replacing the current process. """
     env = args[-1]
     execve(file, args[:-1], env)
 
 def execlp(file, *args):
+    """execlp(file, *args)
+
+    Execute the executable file (which is searched for along $PATH)
+    with argument list args, replacing the current process. """
     execvp(file, args)
 
 def execlpe(file, *args):
+    """execlpe(file, *args, env)
+
+    Execute the executable file (which is searched for along $PATH)
+    with argument list args and environment env, replacing the current
+    process. """    
     env = args[-1]
     execvpe(file, args[:-1], env)
 
 def execvp(file, args):
+    """execp(file, args)
+
+    Execute the executable file (which is searched for along $PATH)
+    with argument list args, replacing the current process.
+    args may be a list or tupe of strings. """
     _execvpe(file, args)
 
 def execvpe(file, args, env):
+    """execv(file, args, env)
+
+    Execute the executable file (which is searched for along $PATH)
+    with argument list args and environment env , replacing the
+    current process.
+    args may be a list or tupe of strings. """    
     _execvpe(file, args, env)
 
 _notfound = None
@@ -338,7 +366,7 @@
 Execute file with arguments from args in a subprocess.
 If mode == P_NOWAIT return the pid of the process.
 If mode == P_WAIT return the process's exit code if it exits normally;
-otherwise return - (the signal that killed it). """   
+otherwise return -SIG, where SIG is the signal that killed it. """   
         return _spawnvef(mode, file, args, None, execv)
 
     def spawnve(mode, file, args, env):



From cgw@alum.mit.edu  Tue Apr 25 15:51:04 2000
From: cgw@alum.mit.edu (Charles G Waldman)
Date: Tue, 25 Apr 2000 09:51:04 -0500 (CDT)
Subject: [Patches] Slightly cleaned-up cycle-gc patch
Message-ID: <14597.45272.956559.551023@sirius.net.home>
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From cgw@alum.mit.edu  Tue Apr 25 16:01:48 2000
From: cgw@alum.mit.edu (Charles G Waldman)
Date: Tue, 25 Apr 2000 10:01:48 -0500 (CDT)
Subject: [Patches] fix for pcremodule.c (DECREF, don't DEL)
Message-ID: <14597.45916.428921.176160@sirius.net.home>

Doing a PyObject_New then PyMem_DEL causes havoc if you are trying to
use Py_TRACE_REFS.

Disclaimer: http://www.python.org/patches/bugrelease.html


Index: Modules/pcremodule.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/pcremodule.c,v
retrieving revision 2.17
diff -c -r2.17 pcremodule.c
*** pcremodule.c	2000/02/29 13:59:23	2.17
--- pcremodule.c	2000/04/25 14:58:59
***************
*** 204,210 ****
  				 &error, &erroroffset, dictionary);
  	if (rv->regex==NULL) 
  	{
! 		PyMem_DEL(rv);
  		if (!PyErr_Occurred())
  		{
  			PyObject *errval = Py_BuildValue("si", error, erroroffset);
--- 204,210 ----
  				 &error, &erroroffset, dictionary);
  	if (rv->regex==NULL) 
  	{
! 		Py_DECREF(rv);
  		if (!PyErr_Occurred())
  		{
  			PyObject *errval = Py_BuildValue("si", error, erroroffset);
***************
*** 217,223 ****
  	if (rv->regex_extra==NULL && error!=NULL) 
  	{
  		PyObject *errval = Py_BuildValue("si", error, 0);
! 		PyMem_DEL(rv);
  		PyErr_SetObject(ErrorObject, errval);
  		Py_XDECREF(errval);
  		return NULL;
--- 217,223 ----
  	if (rv->regex_extra==NULL && error!=NULL) 
  	{
  		PyObject *errval = Py_BuildValue("si", error, 0);
! 		Py_DECREF(rv);
  		PyErr_SetObject(ErrorObject, errval);
  		Py_XDECREF(errval);
  		return NULL;
***************
*** 228,234 ****
  		PyObject *errval = Py_BuildValue("si", error, rv->num_groups);
  		PyErr_SetObject(ErrorObject, errval);
  		Py_XDECREF(errval);
! 		PyMem_DEL(rv);
  		return NULL;
  	}
  	return (PyObject *)rv;
--- 228,234 ----
  		PyObject *errval = Py_BuildValue("si", error, rv->num_groups);
  		PyErr_SetObject(ErrorObject, errval);
  		Py_XDECREF(errval);
! 		Py_DECREF(rv);
  		return NULL;
  	}
  	return (PyObject *)rv;


From tismer@tismer.com  Tue Apr 25 17:07:00 2000
From: tismer@tismer.com (Christian Tismer)
Date: Tue, 25 Apr 2000 18:07:00 +0200
Subject: [Patches] Slightly cleaned-up cycle-gc patch
References: <14597.45272.956559.551023@sirius.net.home>
Message-ID: <3905C2A4.D9007891@tismer.com>


Charles G Waldman wrote:
> 
> The problems I was having getting Neil's cyclic-gc stuff to work with
> the latest CVS Python had nothing to do with Trashcan - sorry,
> Christian!  

Tss tss - my poor little trashcan pet, now it's happy again :-)

ciao - chris

p.s.: Oh, did you look at the end of your patch?
There appears to be some binary crap. Maybe your
python16.dsp file is broekn?

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From cgw@fnal.gov  Tue Apr 25 17:02:09 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Tue, 25 Apr 2000 11:02:09 -0500 (CDT)
Subject: [Patches] Sightly clean-up cycle-gc patch
Message-ID: <14597.49537.611625.343576@buffalo.fnal.gov>
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From cgw@fnal.gov  Tue Apr 25 17:14:06 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Tue, 25 Apr 2000 11:14:06 -0500 (CDT)
Subject: [Patches] Cycle-GC patch, one more time
In-Reply-To: <20000425160217.CB7F61CE52@dinsdale.python.org>
References: <20000425160217.CB7F61CE52@dinsdale.python.org>
Message-ID: <14597.50254.400745.494894@buffalo.fnal.gov>

I don't know why these messages keep getting base-64 encoded.  I
certainly didn't intend to send them that way!

The first message I sent had the wrong file attached.  Please
disregard.  Sorry for the confusion.


The problems I was having getting Neil's cyclic-gc stuff to work with
the latest CVS Python had nothing to do with Trashcan - sorry,
Christian!  It was my own changes to tupleobject.c.  It turned out to
be a little tricky to get the GC info handled correctly during
_PyTuple_Resize.  Below is a version of Neil's patch that applies
cleanly to current CVS and seems to work correctly.  All credit is due
to Neil, I just did a tiny amount of cleanup work in a few places.

Differences:  

  (1) The above-mentioned _PyTuple_Resize fix

  (2) A few problems were exposed by building with Py_TRACE_REFS
defined.  Specifially, PyObject_New and PyObject_NewVar already do a
_Py_NewReference internally, but in a few places PyObject_New is
used to create a new object which is then _Py_NewReference'd.  If
Py_TRACE_REFS is not set, this is merely redundant; with
Py_TRACE_REFS set it causes serious problems because the object is
added to the reference chain twice.



Here's a patch against Neil's patch.

Please disregard the earlier messages on this topic.


                   I confirm that, to the best of my knowledge and belief, this
                   contribution is free of any claims of third parties under
                   copyright, patent or other rights or interests ("claims").  To
                   the extent that I have any such claims, I hereby grant to CNRI a
                   nonexclusive, irrevocable, royalty-free, worldwide license to
                   reproduce, distribute, perform and/or display publicly, prepare
                   derivative versions, and otherwise use this contribution as part
                   of the Python software and its related documentation, or any
                   derivative versions thereof, at no cost to CNRI or its licensed
                   users, and to authorize others to do so.

                   I acknowledge that CNRI may, at its sole discretion, decide
                   whether or not to incorporate this contribution in the Python
                   software and its related documentation.  I further grant CNRI
                   permission to use my name and other identifying information
                   provided to CNRI by me for use in connection with the Python
                   software and its related documentation.

--- gc-cycle.diff.neil	Mon Apr 24 20:54:44 2000
+++ gc-cycle.diff.cgw	Tue Apr 25 10:50:41 2000
@@ -857,7 +857,7 @@
 +}
 +
 --- Python-cvs/Objects/classobject.c	Tue Feb 29 14:55:15 2000
-+++ Python-gc/Objects/classobject.c	Sat Apr  8 02:01:35 2000
++++ Python-gc/Objects/classobject.c	Mon Apr 24 21:19:18 2000
 @@ -110,7 +110,7 @@
  		}
  		Py_INCREF(bases);
@@ -947,16 +947,22 @@
  	if (inst->in_dict == NULL) {
  		Py_DECREF(inst);
  		return NULL;
-@@ -498,6 +528,9 @@
+@@ -498,11 +528,14 @@
  	PyObject *error_type, *error_value, *error_traceback;
  	PyObject *del;
  	static PyObject *delstr;
++	extern long _Py_RefTotal;
 +#ifdef WITH_CYCLE_GC
 +	PyGC_Remove((PyObject *)inst);
 +#endif
  	/* Call the __del__ method if it exists.  First temporarily
  	   revive the object and save the current exception, if any. */
  #ifdef Py_TRACE_REFS
+ 	/* much too complicated if Py_TRACE_REFS defined */
+-	extern long _Py_RefTotal;
+ 	inst->ob_type = &PyInstance_Type;
+ 	_Py_NewReference((PyObject *)inst);
+ 	_Py_RefTotal--;		/* compensate for increment in NEWREF */
 @@ -550,6 +583,9 @@
  #ifdef COUNT_ALLOCS
  		inst->ob_type->tp_free--;
@@ -967,7 +973,11 @@
  		return; /* __del__ added a reference; don't delete now */
  	}
  #ifdef Py_TRACE_REFS
-@@ -561,7 +597,7 @@
+@@ -557,11 +593,10 @@
+ 	inst->ob_type->tp_free--;	/* compensate for increment in UNREF */
+ #endif
+ 	_Py_ForgetReference((PyObject *)inst);
+-	inst->ob_type = NULL;
  #endif /* Py_TRACE_REFS */
  	Py_DECREF(inst->in_class);
  	Py_XDECREF(inst->in_dict);
@@ -976,7 +986,7 @@
  }
  
  static PyObject *
-@@ -837,6 +873,16 @@
+@@ -840,6 +875,16 @@
  	return outcome;
  }
  
@@ -993,7 +1003,7 @@
  static PyObject *getitemstr, *setitemstr, *delitemstr, *lenstr;
  
  static int
-@@ -1460,7 +1506,11 @@
+@@ -1463,7 +1508,11 @@
  	(getattrofunc)instance_getattr, /*tp_getattro*/
  	(setattrofunc)instance_setattr, /*tp_setattro*/
          0, /* tp_as_buffer */
@@ -1006,7 +1016,7 @@
  };
  
  
-@@ -1490,7 +1540,7 @@
+@@ -1493,7 +1542,7 @@
  		_Py_NewReference((PyObject *)im);
  	}
  	else {
@@ -1182,8 +1192,8 @@
 +#endif
  };
 --- Python-cvs/Objects/listobject.c	Fri Mar 24 22:32:30 2000
-+++ Python-gc/Objects/listobject.c	Sat Apr  8 02:02:36 2000
-@@ -61,34 +61,38 @@
++++ Python-gc/Objects/listobject.c	Mon Apr 24 21:14:42 2000
+@@ -61,34 +61,37 @@
  	int i;
  	PyListObject *op;
  	size_t nbytes;
@@ -1231,14 +1241,14 @@
 +	op->ob_item = item;
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
- 	_Py_NewReference((PyObject *)op);
+-	_Py_NewReference((PyObject *)op);
 +#ifdef WITH_CYCLE_GC
 +	PyGC_Insert((PyObject *)op);
 +#endif
  	return (PyObject *) op;
  }
  
-@@ -216,6 +220,9 @@
+@@ -216,6 +219,9 @@
  {
  	int i;
  	Py_TRASHCAN_SAFE_BEGIN(op)
@@ -1248,7 +1258,7 @@
  	if (op->ob_item != NULL) {
  		/* Do it backwards, for Christian Tismer.
  		   There's a simple test case where somehow this reduces
-@@ -227,7 +234,7 @@
+@@ -227,7 +233,7 @@
  		}
  		free((ANY *)op->ob_item);
  	}
@@ -1257,7 +1267,7 @@
  	Py_TRASHCAN_SAFE_END(op)
  }
  
-@@ -1399,6 +1406,29 @@
+@@ -1399,6 +1405,29 @@
  	return NULL;
  }
  
@@ -1287,7 +1297,7 @@
  static char append_doc[] =
  "L.append(object) -- append object to end";
  static char extend_doc[] =
-@@ -1464,6 +1494,18 @@
+@@ -1464,6 +1493,18 @@
  	0,		/*tp_as_number*/
  	&list_as_sequence,	/*tp_as_sequence*/
  	0,		/*tp_as_mapping*/
@@ -1306,7 +1316,7 @@
  };
  
  
-@@ -1532,5 +1574,16 @@
+@@ -1532,5 +1573,16 @@
  	0,		/*tp_as_number*/
  	&immutable_list_as_sequence,	/*tp_as_sequence*/
  	0,		/*tp_as_mapping*/
@@ -1440,8 +1450,18 @@
  
  int
 --- Python-cvs/Objects/tupleobject.c	Fri Mar 24 22:32:30 2000
-+++ Python-gc/Objects/tupleobject.c	Sat Apr  8 02:11:45 2000
-@@ -91,12 +91,10 @@
++++ Python-gc/Objects/tupleobject.c	Mon Apr 24 21:21:59 2000
+@@ -82,8 +82,8 @@
+ #endif
+ #ifdef Py_TRACE_REFS
+ 		op->ob_type = &PyTuple_Type;
+-		op->ob_size = size;
+ #endif
++		_Py_NewReference((PyObject *)op);
+ 	}
+ 	else
+ #endif
+@@ -97,23 +97,23 @@
  			return PyErr_NoMemory();
  		}
  		;
@@ -1455,7 +1475,12 @@
  	}
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
-@@ -107,6 +105,9 @@
+-	_Py_NewReference((PyObject *)op);
++
+ #if MAXSAVESIZE > 0
+ 	if (size == 0) {
+ 		free_tuples[0] = op;
+-		++num_free_tuples[0];
  		Py_INCREF(op);	/* extra INCREF so that this is never freed */
  	}
  #endif
@@ -1465,17 +1490,17 @@
  	return (PyObject *) op;
  }
  
-@@ -173,6 +174,9 @@
+@@ -180,6 +180,9 @@
  	register int i;
- 
+ 	register int len =  op->ob_size;
  	Py_TRASHCAN_SAFE_BEGIN(op)
 +#ifdef WITH_CYCLE_GC
 +	PyGC_Remove((PyObject *)op);
 +#endif
- 	if (op->ob_size > 0) {
- 		i = op->ob_size;
+ 	if (len > 0) {
+ 		i = len;
  		while (--i >= 0)
-@@ -185,7 +189,7 @@
+@@ -193,7 +196,7 @@
  		}
  #endif
  	}
@@ -1484,7 +1509,7 @@
  done:
  	Py_TRASHCAN_SAFE_END(op)
  }
-@@ -393,6 +397,22 @@
+@@ -401,6 +404,22 @@
  	return (PyObject *) np;
  }
  
@@ -1507,7 +1532,7 @@
  static PySequenceMethods tuple_as_sequence = {
  	(inquiry)tuplelength, /*sq_length*/
  	(binaryfunc)tupleconcat, /*sq_concat*/
-@@ -419,6 +439,16 @@
+@@ -427,6 +446,16 @@
  	&tuple_as_sequence,	/*tp_as_sequence*/
  	0,		/*tp_as_mapping*/
  	(hashfunc)tuplehash, /*tp_hash*/
@@ -1524,43 +1549,49 @@
  };
  
  /* The following function breaks the notion that tuples are immutable:
-@@ -441,6 +471,10 @@
+@@ -449,6 +478,9 @@
  	int i;
  	int sizediff;
  
 +#ifdef WITH_CYCLE_GC
 +	PyGCInfo *g = PY_GCINFO(*pv); 
-+	PyGC_Remove(*pv);
 +#endif
  	v = (PyTupleObject *) *pv;
  	if (v == NULL || !PyTuple_Check(v) || v->ob_refcnt != 1) {
  		*pv = 0;
-@@ -469,16 +503,29 @@
- 		Py_XDECREF(v->ob_item[i]);
- 		v->ob_item[i] = NULL;
- 	}
--	sv = (PyTupleObject *)
--		realloc((char *)v,
--			sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
+@@ -479,7 +511,6 @@
+ 	}
+ #if MAXSAVESIZE > 0
+ 	if (newsize == 0 && free_tuples[0]) {
+-		num_free_tuples[0]--;
+ 		sv = free_tuples[0];
+ 		sv->ob_size = 0;
+ 		Py_INCREF(sv);
+@@ -511,9 +542,22 @@
+ 	} else 
+ #endif		
+ 	{
 +#ifdef WITH_CYCLE_GC
-+	g = (PyGCInfo *) realloc((char *)g, sizeof(*g) + 
++		PyGC_Remove((PyObject *)v);
++
++		g = (PyGCInfo *) realloc((char *)g, sizeof(*g) + 
 +				sizeof(PyTupleObject) + 
 +				newsize * sizeof(PyObject *));
-+	if (g == NULL) {
-+		sv == NULL;
-+	} else {
-+		sv = (PyTupleObject *) PY_GCOBJ(g);
-+	}
++		if (g == NULL) {
++			sv == NULL;
++		} else {
++			sv = (PyTupleObject *) PY_GCOBJ(g);
++		}
 +#else
-+	sv = (PyTupleObject *) realloc((char *)v, sizeof(PyTupleObject) +
-+				newsize * sizeof(PyObject *));
-+#endif
- 	*pv = (PyObject *) sv;
- 	if (sv == NULL) {
--		PyMem_DEL(v);
-+		PyObject_Del((PyObject *)v);
- 		PyErr_NoMemory();
- 		return -1;
+ 		sv = (PyTupleObject *)
+ 			realloc((char *)v,
+ 				sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
++#endif
+ 		*pv = (PyObject *) sv;
+ 		if (sv == NULL) {
+ 			PyMem_DEL(v);
+@@ -522,6 +566,9 @@
+ 		}
  	}
  	_Py_NewReference((PyObject *)sv);
 +#ifdef WITH_CYCLE_GC
@@ -1569,12 +1600,16 @@
  	for (i = sv->ob_size; i < newsize; i++)
  		sv->ob_item[i] = NULL;
  	if (last_is_sticky && sizediff > 0) {
-@@ -508,7 +555,7 @@
+@@ -551,7 +598,11 @@
  		while (p) {
  			q = p;
  			p = (PyTupleObject *)(p->ob_item[0]);
 -			PyMem_DEL(q);
++#ifdef Py_TRACE_REFS
++			q->ob_type =  &PyTuple_Type;
++#endif
 +			PyObject_Del((PyObject *)q);
++			
  		}
  	}
  #endif


 > Send Patches mailing list submissions to
 > 	patches@python.org
 > 
 > To subscribe or unsubscribe via the World Wide Web, visit
 > 	http://www.python.org/mailman/listinfo/patches
 > or, via email, send a message with subject or body 'help' to
 > 	patches-request@python.org
 > 
 > You can reach the person managing the list at
 > 	patches-admin@python.org
 > 
 > When replying, please edit your Subject line so it is more specific
 > than "Re: Contents of Patches digest..."
 > 
 > 
 > Today's Topics:
 > 
 >   1. Re: Slightly cleaned-up cycle-gc patch (Christian Tismer)
 >   2. Sightly clean-up cycle-gc patch (Charles G Waldman)
 > 
 > --__--__--
 > 
 > Message: 1
 > Date: Tue, 25 Apr 2000 18:07:00 +0200
 > From: Christian Tismer <tismer@tismer.com>
 > To: Charles G Waldman <cgw@alum.mit.edu>
 > Cc: patches@python.org
 > Subject: Re: [Patches] Slightly cleaned-up cycle-gc patch
 > 
 > 
 > 
 > Charles G Waldman wrote:
 > > 
 > > The problems I was having getting Neil's cyclic-gc stuff to work with
 > > the latest CVS Python had nothing to do with Trashcan - sorry,
 > > Christian!  
 > 
 > Tss tss - my poor little trashcan pet, now it's happy again :-)
 > 
 > ciao - chris
 > 
 > p.s.: Oh, did you look at the end of your patch?
 > There appears to be some binary crap. Maybe your
 > python16.dsp file is broekn?
 > 
 > -- 
 > Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
 > Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
 > Kaunstr. 26                  :    *Starship* http://starship.python.net
 > 14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
 > PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
 >      where do you want to jump today?   http://www.stackless.com
 > 
 > 
 > --__--__--
 > 
 > Message: 2
 > From: Charles G Waldman <cgw@fnal.gov>
 > Date: Tue, 25 Apr 2000 11:02:09 -0500 (CDT)
 > To: patches@python.org
 > Subject: [Patches] Sightly clean-up cycle-gc patch
 > 
 > DQogSSBkb24ndCBrbm93IHdoYXQgaGFwcGVuZWQgd2l0aCB0aGF0IGxhc3QgbWVzc2FnZSAt
 > IGl0IHNvbWVob3cgZ290DQpiYXNlLTY0IGVuY29kZWQgd2l0aG91dCBtZSBhc2tpbmcgZm9y
 > IHRoYXQuICBTb3JyeSEgIFR1cm5zIG91dCB0aGF0DQpub3Qgb25seSB3YXMgdGhlIG1lc3Nh
 > Z2UgdW5yZWFkYWJsZSwgYnV0IEkgYXR0YWNoZWQgdGhlIHdyb25nIGZpbGUuDQpEb3VibGUt
 > b29wcyENCg0KDQpUaGUgcHJvYmxlbXMgSSB3YXMgaGF2aW5nIGdldHRpbmcgTmVpbCdzIGN5
 > Y2xpYy1nYyBzdHVmZiB0byB3b3JrIHdpdGgNCnRoZSBsYXRlc3QgQ1ZTIFB5dGhvbiBoYWQg
 > bm90aGluZyB0byBkbyB3aXRoIFRyYXNoY2FuIC0gc29ycnksDQpDaHJpc3RpYW4hICBJdCB3
 > YXMgbXkgb3duIGNoYW5nZXMgdG8gdHVwbGVvYmplY3QuYy4gIEl0IHR1cm5lZCBvdXQgdG8N
 > CmJlIGEgbGl0dGxlIHRyaWNreSB0byBnZXQgdGhlIEdDIGluZm8gaGFuZGxlZCBjb3JyZWN0
 > bHkgZHVyaW5nDQpfUHlUdXBsZV9SZXNpemUuICBCZWxvdyBpcyBhIHZlcnNpb24gb2YgTmVp
 > bCdzIHBhdGNoIHRoYXQgYXBwbGllcw0KY2xlYW5seSB0byBjdXJyZW50IENWUyBhbmQgc2Vl
 > bXMgdG8gd29yayBjb3JyZWN0bHkuICBBbGwgY3JlZGl0IGlzIGR1ZQ0KdG8gTmVpbCwgSSBq
 > dXN0IGRpZCBhIHRpbnkgYW1vdW50IG9mIGNsZWFudXAgd29yayBpbiBhIGZldyBwbGFjZXMu
 > DQooSSB3YXMgdGhpbmtpbmcgb2YgcG9zdGluZyB0aGlzIGFzIGEgZGlmZiB0byBOZWlsJ3Mg
 > ZGlmZiwgYnV0IHdhc24ndA0Kc3VyZSBob3cgdGhhdCB3b3VsZCBnbyBvdmVyLi4uICBJZiB0
 > aGlzIGdldHMgY2hlY2tlZCBpbiBpdCB3aWxsIGJlIGENCmxvdCBlYXNpZXIgdG8gd29yayBv
 > biBpdCBhbmQgc2VuZCBzbWFsbCBwYXRjaGVzLCByYXRoZXIgdGhhbiBoYXZpbmcgdG8NCnJl
 > Z2VuZXJhdGUgdGhlIHdob2xlIHRoaW5nIGVhY2ggdGltZSkuDQoNCg0KDQpEaWZmZXJlbmNl
 > czogIA0KDQogICgxKSBUaGUgYWJvdmUtbWVudGlvbmVkIF9QeVR1cGxlX1Jlc2l6ZSBmaXgN
 > Cg0KICAoMikgQSBmZXcgcHJvYmxlbXMgd2VyZSBleHBvc2VkIGJ5IGJ1aWxkaW5nIHdpdGgg
 > UHlfVFJBQ0VfUkVGUw0KICBkZWZpbmVkLiAgU3BlY2lmaWFsbHksIFB5T2JqZWN0X05ldyBh
 > bmQgUHlPYmplY3RfTmV3VmFyIGFscmVhZHkgZG8gYQ0KICBfUHlfTmV3UmVmZXJlbmNlIGlu
 > dGVybmFsbHksIGJ1dCBpbiBhIGZldyBwbGFjZXMgKGUuZy4gUHlMaXN0X05ldyksDQogIFB5
 > T2JqZWN0X05ldyB3YXMgYmVpbmcgdXNlZCB0byBjcmVhdGUgYSBuZXcgb2JqZWN0IHdoaWNo
 > IGlzIHRoZW4NCiAgX1B5X05ld1JlZmVyZW5jZSdkLiAgSWYgUHlfVFJBQ0VfUkVGUyBpcyBu
 > b3Qgc2V0LCB0aGlzIGlzIG1lcmVseQ0KICByZWR1bmRhbnQ7IHdpdGggUHlfVFJBQ0VfUkVG
 > UyBzZXQgaXQgY2F1c2VzIHNlcmlvdXMgcHJvYmxlbXMgYmVjYXVzZQ0KICB0aGUgb2JqZWN0
 > IGlzIGFkZGVkIHRvIHRoZSByZWZlcmVuY2UgY2hhaW4gdHdpY2UuDQoNCg0KDQogICAgICAg
 > ICAgICAgICAgICAgSSBjb25maXJtIHRoYXQsIHRvIHRoZSBiZXN0IG9mIG15IGtub3dsZWRn
 > ZSBhbmQgYmVsaWVmLCB0aGlzDQogICAgICAgICAgICAgICAgICAgY29udHJpYnV0aW9uIGlz
 > IGZyZWUgb2YgYW55IGNsYWltcyBvZiB0aGlyZCBwYXJ0aWVzIHVuZGVyDQogICAgICAgICAg
 > ICAgICAgICAgY29weXJpZ2h0LCBwYXRlbnQgb3Igb3RoZXIgcmlnaHRzIG9yIGludGVyZXN0
 > cyAoImNsYWltcyIpLiAgVG8NCiAgICAgICAgICAgICAgICAgICB0aGUgZXh0ZW50IHRoYXQg
 > SSBoYXZlIGFueSBzdWNoIGNsYWltcywgSSBoZXJlYnkgZ3JhbnQgdG8gQ05SSSBhDQogICAg
 > ICAgICAgICAgICAgICAgbm9uZXhjbHVzaXZlLCBpcnJldm9jYWJsZSwgcm95YWx0eS1mcmVl
 > LCB3b3JsZHdpZGUgbGljZW5zZSB0bw0KICAgICAgICAgICAgICAgICAgIHJlcHJvZHVjZSwg
 > ZGlzdHJpYnV0ZSwgcGVyZm9ybSBhbmQvb3IgZGlzcGxheSBwdWJsaWNseSwgcHJlcGFyZQ0K
 > ICAgICAgICAgICAgICAgICAgIGRlcml2YXRpdmUgdmVyc2lvbnMsIGFuZCBvdGhlcndpc2Ug
 > dXNlIHRoaXMgY29udHJpYnV0aW9uIGFzIHBhcnQNCiAgICAgICAgICAgICAgICAgICBvZiB0
 > aGUgUHl0aG9uIHNvZnR3YXJlIGFuZCBpdHMgcmVsYXRlZCBkb2N1bWVudGF0aW9uLCBvciBh
 > bnkNCiAgICAgICAgICAgICAgICAgICBkZXJpdmF0aXZlIHZlcnNpb25zIHRoZXJlb2YsIGF0
 > IG5vIGNvc3QgdG8gQ05SSSBvciBpdHMgbGljZW5zZWQNCiAgICAgICAgICAgICAgICAgICB1
 > c2VycywgYW5kIHRvIGF1dGhvcml6ZSBvdGhlcnMgdG8gZG8gc28uDQoNCiAgICAgICAgICAg
 > ICAgICAgICBJIGFja25vd2xlZGdlIHRoYXQgQ05SSSBtYXksIGF0IGl0cyBzb2xlIGRpc2Ny
 > ZXRpb24sIGRlY2lkZQ0KICAgICAgICAgICAgICAgICAgIHdoZXRoZXIgb3Igbm90IHRvIGlu
 > Y29ycG9yYXRlIHRoaXMgY29udHJpYnV0aW9uIGluIHRoZSBQeXRob24NCiAgICAgICAgICAg
 > ICAgICAgICBzb2Z0d2FyZSBhbmQgaXRzIHJlbGF0ZWQgZG9jdW1lbnRhdGlvbi4gIEkgZnVy
 > dGhlciBncmFudCBDTlJJDQogICAgICAgICAgICAgICAgICAgcGVybWlzc2lvbiB0byB1c2Ug
 > bXkgbmFtZSBhbmQgb3RoZXIgaWRlbnRpZnlpbmcgaW5mb3JtYXRpb24NCiAgICAgICAgICAg
 > ICAgICAgICBwcm92aWRlZCB0byBDTlJJIGJ5IG1lIGZvciB1c2UgaW4gY29ubmVjdGlvbiB3
 > aXRoIHRoZSBQeXRob24NCiAgICAgICAgICAgICAgICAgICBzb2Z0d2FyZSBhbmQgaXRzIHJl
 > bGF0ZWQgZG9jdW1lbnRhdGlvbi4NCg0KLS0tIFB5dGhvbi1jdnMvSW5jbHVkZS9vYmplY3Qu
 > aAlGcmkgTWFyIDI0IDIyOjMyOjE2IDIwMDANCisrKyBQeXRob24tZ2MvSW5jbHVkZS9vYmpl
 > Y3QuaAlTYXQgQXByICA4IDAxOjM4OjI1IDIwMDANCkBAIC0xNDUsNiArMTQ1LDEwIEBADQog
 > dHlwZWRlZiBpbnQgKCpnZXRzZWdjb3VudHByb2MpIFB5X1BST1RPKChQeU9iamVjdCAqLCBp
 > bnQgKikpOw0KIHR5cGVkZWYgaW50ICgqZ2V0Y2hhcmJ1ZmZlcnByb2MpIFB5X1BST1RPKChQ
 > eU9iamVjdCAqLCBpbnQsIGNvbnN0IGNoYXIgKiopKTsNCiB0eXBlZGVmIGludCAoKm9iam9i
 > anByb2MpIFB5X1BST1RPKChQeU9iamVjdCAqLCBQeU9iamVjdCAqKSk7DQorI2lmZGVmIFdJ
 > VEhfQ1lDTEVfR0MNCit0eXBlZGVmIGludCAoKnZpc2l0cHJvYykgUHlfUFJPVE8oKFB5T2Jq
 > ZWN0ICosIHZvaWQgKikpOw0KK3R5cGVkZWYgaW50ICgqcmVjdXJzZXByb2MpIFB5X1BST1RP
 > KChQeU9iamVjdCAqLCB2aXNpdHByb2MsIHZvaWQgKikpOw0KKyNlbmRpZg0KIA0KIHR5cGVk
 > ZWYgc3RydWN0IHsNCiAJYmluYXJ5ZnVuYyBuYl9hZGQ7DQpAQCAtMjQzLDkgKzI0NywxOCBA
 > QA0KIA0KIAljaGFyICp0cF9kb2M7IC8qIERvY3VtZW50YXRpb24gc3RyaW5nICovDQogDQot
 > CS8qIE1vcmUgc3BhcmVzICovDQorI2lmZGVmIFdJVEhfQ1lDTEVfR0MNCisJLyogY2FsbCBm
 > dW5jdGlvbiBmb3IgYWxsIGFjY2Vzc2libGUgb2JqZWN0cyAqLw0KKwlyZWN1cnNlcHJvYyB0
 > cF9yZWN1cnNlOw0KKwkNCisJLyogZGVsZXRlIHJlZmVyZW5jZXMgdG8gY29udGFpbmVkIG9i
 > amVjdHMgKi8NCisJaW5xdWlyeSB0cF9jbGVhcjsNCisjZWxzZQ0KIAlsb25nIHRwX3h4eDU7
 > DQogCWxvbmcgdHBfeHh4NjsNCisjZW5kaWYNCisNCisJLyogTW9yZSBzcGFyZXMgKi8NCiAJ
 > bG9uZyB0cF94eHg3Ow0KIAlsb25nIHRwX3h4eDg7DQogDQpAQCAtMzE1LDYgKzMyOCw5IEBA
 > DQogDQogLyogUHlTZXF1ZW5jZU1ldGhvZHMgY29udGFpbnMgc3FfY29udGFpbnMgKi8NCiAj
 > ZGVmaW5lIFB5X1RQRkxBR1NfSEFWRV9TRVFVRU5DRV9JTiAoMUw8PDEpDQorDQorLyogT2Jq
 > ZWN0cyB3aXRoIGEgR0MgaW5mbyBwcmVmaXggKGFsbG9jYXRlZCAqYmVmb3JlKiB0aGUgb2Jq
 > ZWN0IGl0c2VsZiEpICovDQorI2RlZmluZSBQeV9UUEZMQUdTX0hBVkVfR0NJTkZPICgxTDw8
 > MikNCiANCiAjZGVmaW5lIFB5X1RQRkxBR1NfREVGQVVMVCAgKFB5X1RQRkxBR1NfSEFWRV9H
 > RVRDSEFSQlVGRkVSIHwgXA0KICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgUHlfVFBG
 > TEFHU19IQVZFX1NFUVVFTkNFX0lOKQ0KLS0tIFB5dGhvbi1jdnMvSW5jbHVkZS9vYmppbXBs
 > LmgJVGh1IE1hciAgMiAxNzowMjoyMyAyMDAwDQorKysgUHl0aG9uLWdjL0luY2x1ZGUvb2Jq
 > aW1wbC5oCVNhdCBBcHIgIDggMDM6MTg6MjIgMjAwMA0KQEAgLTM4LDggKzM4LDQyIEBADQog
 > LyoNCiBBZGRpdGlvbmFsIG1hY3JvcyBmb3IgbW9kdWxlcyB0aGF0IGltcGxlbWVudCBuZXcg
 > b2JqZWN0IHR5cGVzLg0KIFlvdSBtdXN0IGZpcnN0IGluY2x1ZGUgIm9iamVjdC5oIi4NCisq
 > Lw0KKw0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDIA0KKw0KKy8qIFN0cnVjdHVyZSAqcHJlZml4
 > ZWQqIHRvIGNvbnRhaW5lciBvYmplY3RzIHBhcnRpY2lwYXRpbmcgaW4gR0MgKi8gDQordHlw
 > ZWRlZiBzdHJ1Y3QgX2djaW5mbyB7DQorCXN0cnVjdCBfZ2NpbmZvICpnY19uZXh0Ow0KKwlz
 > dHJ1Y3QgX2djaW5mbyAqZ2NfcHJldjsNCisJaW50IGdjX3JlZnM7DQorfSBQeUdDSW5mbzsN
 > CisNCisvKiBUZXN0IGlmIGEgdHlwZSBoYXMgR0MgaW5mbyAqLw0KKyNkZWZpbmUgUFlfVFlQ
 > RUlTR0ModCkgUHlUeXBlX0hhc0ZlYXR1cmUoKHQpLCBQeV9UUEZMQUdTX0hBVkVfR0NJTkZP
 > KQ0KKw0KKy8qIFRlc3QgaWYgYW4gb2JqZWN0IGhhcyBHQyBpbmZvICovDQorI2RlZmluZSBQ
 > WV9JU0dDKG8pIFBZX1RZUEVJU0dDKChvKS0+b2JfdHlwZSkNCisNCisvKiBHZXQgYW4gb2Jq
 > ZWN0J3MgR0MgaW5mbyAtLSBOVUxMIGlmIHRoZSBvYmplY3QgaGFzIG5vbmUgKi8NCisjZGVm
 > aW5lIFBZX0dDSU5GTyhvKSAoUFlfSVNHQyhvKSA/ICgoUHlHQ0luZm8gKikobyktMSkgOiBO
 > VUxMKQ0KKw0KKy8qIFVuc2FmZSB2ZXJzaW9uIG9mIFBZX0dDSU5GTygpIC0tIG9ubHkgY2Fs
 > bCBpZiBQWV9JU0dDKHApIGlzIHRydWUgKi8NCisjZGVmaW5lIFBZX0dDSU5GT19VTlNBRkUo
 > bykgKChQeUdDSW5mbyAqKShvKS0xKQ0KKw0KKy8qIEdldCB0aGUgb2JqZWN0IGdpdmVuIHRo
 > ZSBQeUdDSW5mbyAqLw0KKyNkZWZpbmUgUFlfR0NPQkooZykgKChQeU9iamVjdCAqKSgoZykr
 > MSkpDQogDQotUHlPYmplY3RfTkVXKHR5cGUsIHR5cGVvYmopIGFsbG9jYXRlcyBtZW1vcnkg
 > Zm9yIGEgbmV3IG9iamVjdCBvZiB0aGUgZ2l2ZW4NCisvKiBBZGQgdGhlIG9iamVjdCBpbnRv
 > IHRoZSBjb250YWluZXIgc2V0ICovDQorZXh0ZXJuIERMX0lNUE9SVCh2b2lkKSBQeUdDX0lu
 > c2VydCBQeV9QUk9UTygoUHlPYmplY3QgKikpOw0KKw0KKy8qIFJlbW92ZSB0aGUgb2JqZWN0
 > IGZyb20gdGhlIGNvbnRhaW5lciBzZXQgKi8NCitleHRlcm4gRExfSU1QT1JUKHZvaWQpIFB5
 > R0NfUmVtb3ZlIFB5X1BST1RPKChQeU9iamVjdCAqKSk7DQorDQorI2VuZGlmDQorDQorLyoN
 > CitQeU9iamVjdF9OZXcodHlwZSwgdHlwZW9iaikgYWxsb2NhdGVzIG1lbW9yeSBmb3IgYSBu
 > ZXcgb2JqZWN0IG9mIHRoZSBnaXZlbg0KIHR5cGU7IGhlcmUgJ3R5cGUnIG11c3QgYmUgdGhl
 > IEMgc3RydWN0dXJlIHR5cGUgdXNlZCB0byByZXByZXNlbnQgdGhlDQogb2JqZWN0IGFuZCAn
 > dHlwZW9iaicgdGhlIGFkZHJlc3Mgb2YgdGhlIGNvcnJlc3BvbmRpbmcgdHlwZSBvYmplY3Qu
 > DQogUmVmZXJlbmNlIGNvdW50IGFuZCB0eXBlIHBvaW50ZXIgYXJlIGZpbGxlZCBpbjsgdGhl
 > IHJlc3Qgb2YgdGhlIGJ5dGVzIG9mDQpAQCAtNDcsMzAgKzgxLDI5IEBADQogVGhlIHNpemUg
 > b2YgdGhlIG9iamVjdCBpcyBhY3R1YWxseSBkZXRlcm1pbmVkIGJ5IHRoZSB0cF9iYXNpY3Np
 > emUgZmllbGQNCiBvZiB0aGUgdHlwZSBvYmplY3QuDQogDQotUHlPYmplY3RfTkVXX1ZBUih0
 > eXBlLCB0eXBlb2JqLCBuKSBpcyBzaW1pbGFyIGJ1dCBhbGxvY2F0ZXMgYSB2YXJpYWJsZS1z
 > aXplDQorUHlPYmplY3RfTmV3VmFyKHR5cGUsIHR5cGVvYmosIG4pIGlzIHNpbWlsYXIgYnV0
 > IGFsbG9jYXRlcyBhIHZhcmlhYmxlLXNpemUNCiBvYmplY3Qgd2l0aCBuIGV4dHJhIGl0ZW1z
 > LiAgVGhlIHNpemUgaXMgY29tcHV0ZWQgYXMgdHBfYmFzaWNzaXplIHBsdXMNCiBuICogdHBf
 > aXRlbXNpemUuICBUaGlzIGZpbGxzIGluIHRoZSBvYl9zaXplIGZpZWxkIGFzIHdlbGwuDQog
 > Ki8NCiANCi0jaWZuZGVmIE1TX0NPUkVETEwNCiBleHRlcm4gRExfSU1QT1JUKFB5T2JqZWN0
 > ICopIF9QeU9iamVjdF9OZXcgUHlfUFJPVE8oKFB5VHlwZU9iamVjdCAqKSk7DQogZXh0ZXJu
 > IERMX0lNUE9SVChQeVZhck9iamVjdCAqKSBfUHlPYmplY3RfTmV3VmFyIFB5X1BST1RPKChQ
 > eVR5cGVPYmplY3QgKiwgaW50KSk7DQorZXh0ZXJuIERMX0lNUE9SVCh2b2lkKSBQeU9iamVj
 > dF9EZWwgUHlfUFJPVE8oKFB5T2JqZWN0ICopKTsNCiANCi0jZGVmaW5lIFB5T2JqZWN0X05F
 > Vyh0eXBlLCB0eXBlb2JqKSAoKHR5cGUgKikgX1B5T2JqZWN0X05ldyh0eXBlb2JqKSkNCi0j
 > ZGVmaW5lIFB5T2JqZWN0X05FV19WQVIodHlwZSwgdHlwZW9iaiwgbikgKCh0eXBlICopIF9Q
 > eU9iamVjdF9OZXdWYXIodHlwZW9iaiwgbikpDQotDQotI2Vsc2UNCi0vKiBGb3IgYW4gTVMt
 > V2luZG93cyBETEwsIHdlIGNoYW5nZSB0aGUgd2F5IGFuIG9iamVjdCBpcyBjcmVhdGVkLCBz
 > byB0aGF0IHRoZQ0KLSAgIGV4dGVuc2lvbiBtb2R1bGUncyBtYWxsb2MgaXMgdXNlZCwgcmF0
 > aGVyIHRoYW4gdGhlIGNvcmUgRExMIG1hbGxvYywgYXMgdGhlcmUgaXMNCi0gICBubyBndWFy
 > YW50ZWUgdGhleSB3aWxsIHVzZSB0aGUgc2FtZSBoZWFwDQotKi8NCi1leHRlcm4gRExfSU1Q
 > T1JUKFB5T2JqZWN0ICopIF9QeU9iamVjdF9OZXcgUHlfUFJPVE8oKFB5VHlwZU9iamVjdCAq
 > LCBQeU9iamVjdCAqKSk7DQotZXh0ZXJuIERMX0lNUE9SVChQeVZhck9iamVjdCAqKSBfUHlP
 > YmplY3RfTmV3VmFyIFB5X1BST1RPKChQeVR5cGVPYmplY3QgKiwgaW50LCBQeVZhck9iamVj
 > dCAqKSk7DQotDQotI2RlZmluZSBQeU9iamVjdF9ORVcodHlwZSwgdHlwZW9iaikgKCh0eXBl
 > ICopIF9QeU9iamVjdF9OZXcodHlwZW9iaiwoUHlPYmplY3QgKiltYWxsb2MoKHR5cGVvYmop
 > LT50cF9iYXNpY3NpemUpKSkNCi0jZGVmaW5lIFB5T2JqZWN0X05FV19WQVIodHlwZSwgdHlw
 > ZW9iaiwgbikgKCh0eXBlICopIF9QeU9iamVjdF9OZXdWYXIodHlwZW9iaiwgbiwgKFB5VmFy
 > T2JqZWN0ICopbWFsbG9jKCh0eXBlb2JqKS0+dHBfYmFzaWNzaXplICsgbiAqICh0eXBlb2Jq
 > KS0+dHBfaXRlbXNpemUpKSkNCi0NCi0jZW5kaWYgLyogTVNfQ09SRURMTCAqLw0KKy8qIEZ1
 > bmN0aW9ucyAqLw0KKyNkZWZpbmUgUHlPYmplY3RfTmV3KHR5cGUsIHR5cGVvYmopIFwNCisJ
 > CSgodHlwZSAqKSBfUHlPYmplY3RfTmV3KHR5cGVvYmopKQ0KKyNkZWZpbmUgUHlPYmplY3Rf
 > TmV3VmFyKHR5cGUsIHR5cGVvYmosIG4pIFwNCisgICAgICAgICAgICAgICAgKCh0eXBlICop
 > IF9QeU9iamVjdF9OZXdWYXIoKHR5cGVvYmopLCAobikpKQ0KKw0KKw0KK2V4dGVybiBETF9J
 > TVBPUlQoUHlPYmplY3QgKikgX1B5T2JqZWN0X0Zyb21UeXBlIFB5X1BST1RPKChQeVR5cGVP
 > YmplY3QgKiwgUHlPYmplY3QgKikpOw0KK2V4dGVybiBETF9JTVBPUlQoUHlWYXJPYmplY3Qg
 > KikgX1B5T2JqZWN0X1ZhckZyb21UeXBlIFB5X1BST1RPKChQeVR5cGVPYmplY3QgKiwgaW50
 > LCBQeVZhck9iamVjdCAqKSk7DQorDQorLyogVGhlc2UgbWFjcm9zIG1heSBub3QgdXNlIHRo
 > ZSBzYW1lIG1hbGxvYyBhcyBQeXRob24uIFRoZXkgY2Fubm90IGJlIHVzZWQNCisgKiBvYmpl
 > Y3RzIHRoYXQgYXJlIGludm9sdmVkIGluIGdhcmJhZ2UgY29sbGVjdGlvbiAqLw0KKyNkZWZp
 > bmUgUHlPYmplY3RfTkVXKHR5cGUsIHR5cGVvYmopICgodHlwZSAqKSBfUHlPYmplY3RfRnJv
 > bVR5cGUodHlwZW9iaiwoUHlPYmplY3QgKiltYWxsb2MoKHR5cGVvYmopLT50cF9iYXNpY3Np
 > emUpKSkNCisjZGVmaW5lIFB5T2JqZWN0X05FV19WQVIodHlwZSwgdHlwZW9iaiwgbikgKCh0
 > eXBlICopIF9QeU9iamVjdF9WYXJGcm9tVHlwZSh0eXBlb2JqLCBuLCAoUHlWYXJPYmplY3Qg
 > KiltYWxsb2MoKHR5cGVvYmopLT50cF9iYXNpY3NpemUgKyBuICogKHR5cGVvYmopLT50cF9p
 > dGVtc2l6ZSkpKQ0KIA0KICNpZmRlZiBfX2NwbHVzcGx1cw0KIH0NCi0tLSBQeXRob24tY3Zz
 > L01vZHVsZXMvY1BpY2tsZS5jCUZyaSBNYXIgMjQgMjI6MzI6MjUgMjAwMA0KKysrIFB5dGhv
 > bi1nYy9Nb2R1bGVzL2NQaWNrbGUuYwlTYXQgQXByICA4IDAyOjEzOjExIDIwMDANCkBAIC0y
 > ODg2LDcgKzI4ODYsNyBAQA0KICAgICAgICAgICAgICAgUHlJbnN0YW5jZU9iamVjdCAqaW5z
 > dDsNCiANCiAgICAgICAgICAgICAgIFB5RXJyX0NsZWFyKCk7DQotICAgICAgICAgICAgICBV
 > TkxFU1MgKGluc3Q9UHlPYmplY3RfTkVXKFB5SW5zdGFuY2VPYmplY3QsICZQeUluc3RhbmNl
 > X1R5cGUpKQ0KKyAgICAgICAgICAgICAgVU5MRVNTIChpbnN0PVB5T2JqZWN0X05ldyhQeUlu
 > c3RhbmNlT2JqZWN0LCAmUHlJbnN0YW5jZV9UeXBlKSkNCiAgICAgICAgICAgICAgICAgZ290
 > byBlcnI7DQogICAgICAgICAgICAgICBpbnN0LT5pbl9jbGFzcz0oUHlDbGFzc09iamVjdCop
 > Y2xzOw0KICAgICAgICAgICAgICAgUHlfSU5DUkVGKGNscyk7DQpAQCAtMjg5NCw3ICsyODk0
 > LDkgQEANCiAgICAgICAgICAgICAgICAgUHlfREVDUkVGKGluc3QpOw0KICAgICAgICAgICAg
 > ICAgICBnb3RvIGVycjsNCiAgICAgICAgICAgICAgIH0NCi0NCisjaWZkZWYgV0lUSF9DWUNM
 > RV9HQw0KKyAgICAgICAgICAgICAgUHlHQ19JbnNlcnQoKFB5T2JqZWN0ICopaW5zdCk7DQor
 > I2VuZGlmDQogICAgICAgICAgICAgICByZXR1cm4gKFB5T2JqZWN0ICopaW5zdDsNCiAgICAg
 > ICAgICAgICB9DQogICAgICAgICAgIFB5X0RFQ1JFRihfX2dldGluaXRhcmdzX18pOw0KLS0t
 > IFB5dGhvbi1jdnMvTW9kdWxlcy9uZXdtb2R1bGUuYwlUdWUgRmViIDI5IDE0OjU1OjA5IDIw
 > MDANCisrKyBQeXRob24tZ2MvTW9kdWxlcy9uZXdtb2R1bGUuYwlTYXQgQXByICA4IDAyOjEz
 > OjI2IDIwMDANCkBAIC00OSwxMyArNDksMTYgQEANCiAJCQkgICAgICAmUHlDbGFzc19UeXBl
 > LCAma2xhc3MsDQogCQkJICAgICAgJlB5RGljdF9UeXBlLCAmZGljdCkpDQogCQlyZXR1cm4g
 > TlVMTDsNCi0JaW5zdCA9IFB5T2JqZWN0X05FVyhQeUluc3RhbmNlT2JqZWN0LCAmUHlJbnN0
 > YW5jZV9UeXBlKTsNCisJaW5zdCA9IFB5T2JqZWN0X05ldyhQeUluc3RhbmNlT2JqZWN0LCAm
 > UHlJbnN0YW5jZV9UeXBlKTsNCiAJaWYgKGluc3QgPT0gTlVMTCkNCiAJCXJldHVybiBOVUxM
 > Ow0KIAlQeV9JTkNSRUYoa2xhc3MpOw0KIAlQeV9JTkNSRUYoZGljdCk7DQogCWluc3QtPmlu
 > X2NsYXNzID0gKFB5Q2xhc3NPYmplY3QgKilrbGFzczsNCiAJaW5zdC0+aW5fZGljdCA9IGRp
 > Y3Q7DQorI2lmZGVmIFdJVEhfQ1lDTEVfR0MNCisJUHlHQ19JbnNlcnQoKFB5T2JqZWN0ICop
 > aW5zdCk7DQorI2VuZGlmDQogCXJldHVybiAoUHlPYmplY3QgKilpbnN0Ow0KIH0NCiANCi0t
 > LSBQeXRob24tY3ZzL01vZHVsZXMvZ2Ntb2R1bGUuYw0KKysrIFB5dGhvbi1nYy9Nb2R1bGVz
 > L2djbW9kdWxlLmMJU2F0IEFwciAgOCAwMzoyNTo1NyAyMDAwDQpAQCAtMCwwICsxLDY3NyBA
 > QA0KKy8qDQorIA0KKyAgUmVmZXJlbmNlIEN5Y2xlIEdhcmJhZ2UgQ29sbGVjdGlvbg0KKyAg
 > PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQ0KKw0KKyAgTmVpbCBTY2hlbWVu
 > YXVlciA8bmFzY2hlbWVAZW5tZS51Y2FsZ2FyeS5jYT4NCisNCisgIEJhc2VkIG9uIGEgcG9z
 > dCBvbiB0aGUgcHl0aG9uLWRldiBsaXN0LiAgSWRlYXMgZnJvbSBHdWlkbyB2YW4gUm9zc3Vt
 > LA0KKyAgRXJpYyBUaWVkZW1hbm4sIGFuZCB2YXJpb3VzIG90aGVycy4NCisNCisgIGh0dHA6
 > Ly93d3cuZW5tZS5jYWxnYXJ5LmNhL35uYXNjaGVtZS9weXRob24vZ2MuaHRtbA0KKyAgaHR0
 > cDovL3d3dy5weXRob24ub3JnL3BpcGVybWFpbC9weXRob24tZGV2LzIwMDAtTWFyY2gvMDAz
 > ODY5Lmh0bWwNCisgIGh0dHA6Ly93d3cucHl0aG9uLm9yZy9waXBlcm1haWwvcHl0aG9uLWRl
 > di8yMDAwLU1hcmNoLzAwNDAxMC5odG1sDQorICBodHRwOi8vd3d3LnB5dGhvbi5vcmcvcGlw
 > ZXJtYWlsL3B5dGhvbi1kZXYvMjAwMC1NYXJjaC8wMDQwMjIuaHRtbA0KKw0KKyAgRm9yIGEg
 > aGlnaGxldmVsIHZpZXcgb2YgdGhlIGNvbGxlY3Rpb24gcHJvY2VzcywgcmVhZCB0aGUgY29s
 > bGVjdA0KKyAgZnVuY3Rpb24uDQorDQorICBUT0RPOg0KKwl1c2UgYSBkaWZmZXJlbnQgaW50
 > ZXJmYWNlIGZvciBzZXRfZGVidWcoKSAoa2V5d29yZHMpPw0KKwlpbmxpbmUgUHlHQ19JbnNl
 > cnQgYW5kIFB5R0NfUmVtb3ZlIGNhbGxzPw0KKwl0dW5lIHBhcmFtZXRlcnMNCisNCisqLw0K
 > Kw0KKyNpbmNsdWRlICJQeXRob24uaCINCisNCisjaWZuZGVmIFdJVEhfQ1lDTEVfR0MNCisj
 > ZXJyb3IgIllvdSBtdXN0IGRlZmluZSBXSVRIX0NZQ0xFX0dDIHRvIGluY2x1ZGUgdGhpcyBt
 > b2R1bGUiDQorI2VuZGlmDQorDQorLyogbWFnaWMgZ2NfcmVmcyB2YWx1ZSAqLw0KKyNkZWZp
 > bmUgR0NfTU9WRUQgLTENCisNCisNCisvKioqIEdsb2JhbCBHQyBzdGF0ZSAqKiovDQorDQor
 > LyogbGlua2VkIGxpc3RzIG9mIGNvbnRhaW5lciBvYmplY3RzICovDQorc3RhdGljIFB5R0NJ
 > bmZvIGdlbmVyYXRpb24wID0geyZnZW5lcmF0aW9uMCwgJmdlbmVyYXRpb24wLCAwfTsNCitz
 > dGF0aWMgUHlHQ0luZm8gZ2VuZXJhdGlvbjEgPSB7JmdlbmVyYXRpb24xLCAmZ2VuZXJhdGlv
 > bjEsIDB9Ow0KK3N0YXRpYyBQeUdDSW5mbyBnZW5lcmF0aW9uMiA9IHsmZ2VuZXJhdGlvbjIs
 > ICZnZW5lcmF0aW9uMiwgMH07DQorc3RhdGljIGludCBnZW5lcmF0aW9uID0gMDsgLyogY3Vy
 > cmVudCBnZW5lcmF0aW9uIGJlaW5nIGNvbGxlY3RlZCAqLw0KKw0KKy8qIGNvbGxlY3Rpb24g
 > ZnJlcXVlbmNpZXMsIFhYWCB0dW5lIHRoZXNlICovDQorc3RhdGljIGludCB0aHJlc2hvbGQw
 > ID0gMTAwOyAvKiBuZXQgbmV3IGNvbnRhaW5lcnMgYmVmb3JlIGNvbGxlY3Rpb24gKi8NCitz
 > dGF0aWMgaW50IHRocmVzaG9sZDEgPSAxMDsgIC8qIGdlbmVyYXRpb24wIGNvbGxlY3Rpb25z
 > IGJlZm9yZSBjb2xsZWN0aW5nIDEgKi8NCitzdGF0aWMgaW50IHRocmVzaG9sZDIgPSAxMDsg
 > IC8qIGdlbmVyYXRpb24xIGNvbGxlY3Rpb25zIGJlZm9yZSBjb2xsZWN0aW5nIDIgKi8NCisN
 > CisvKiBuZXQgbmV3IG9iamVjdHMgYWxsb2NhdGVkIHNpbmNlIGxhc3QgY29sbGVjdGlvbiAq
 > Lw0KK3N0YXRpYyBpbnQgYWxsb2NhdGVkID0gMDsNCisNCisvKiBzZXQgZm9yIGRlYnVnZ2lu
 > ZyBpbmZvcm1hdGlvbiAqLw0KKyNkZWZpbmUgREVCVUdfU1RBVFMJCSgxPDwwKSAvKiBwcmlu
 > dCBjb2xsZWN0aW9uIHN0YXRpc3RpY3MgKi8NCisjZGVmaW5lIERFQlVHX0NPTExFQ1RBQkxF
 > CSgxPDwxKSAvKiBwcmludCBjb2xsZWN0YWJsZSBvYmplY3RzICovDQorI2RlZmluZSBERUJV
 > R19VTkNPTExFQ1RBQkxFCSgxPDwyKSAvKiBwcmludCB1bmNvbGxlY3RhYmxlIG9iamVjdHMg
 > Ki8NCisjZGVmaW5lIERFQlVHX0lOU1RBTkNFUwkJKDE8PDMpIC8qIHByaW50IGluc3RhbmNl
 > cyAqLw0KKyNkZWZpbmUgREVCVUdfT0JKRUNUUwkJKDE8PDQpIC8qIHByaW50IG90aGVyIG9i
 > amVjdHMgKi8NCisjZGVmaW5lIERFQlVHX0xFQUsJCURFQlVHX0NPTExFQ1RBQkxFIHwgXA0K
 > KwkJCQlERUJVR19VTkNPTExFQ1RBQkxFIHwgXA0KKwkJCQlERUJVR19JTlNUQU5DRVMgfCBc
 > DQorCQkJCURFQlVHX09CSkVDVFMNCitzdGF0aWMgaW50IGRlYnVnID0gREVCVUdfTEVBSzsN
 > CisNCisvKiBsaXN0IG9mIHVuY29sbGVjdGFibGUgb2JqZWN0cyAqLw0KK3N0YXRpYyBQeU9i
 > amVjdCAqZ2FyYmFnZSA9IE5VTEw7DQorDQorDQorLyoqKiBsaXN0IGZ1bmN0aW9ucyAoc2hv
 > dWxkIHByb2JhYmx5IGJlIG1hY3Jvcywgc3R1cGlkIGNvbXBpbGVycykgKioqLw0KKw0KK3N0
 > YXRpYyB2b2lkDQorTElTVF9JTklUKFB5R0NJbmZvICpsaXN0KQ0KK3sNCisJbGlzdC0+Z2Nf
 > cHJldiA9IGxpc3Q7DQorCWxpc3QtPmdjX25leHQgPSBsaXN0Ow0KK30NCisNCitzdGF0aWMg
 > dm9pZA0KK0xJU1RfQVBQRU5EKFB5R0NJbmZvICpub2RlLCBQeUdDSW5mbyAqbGlzdCkNCit7
 > DQorCW5vZGUtPmdjX25leHQgPSBsaXN0Ow0KKwlub2RlLT5nY19wcmV2ID0gbGlzdC0+Z2Nf
 > cHJldjsNCisJbm9kZS0+Z2NfcHJldi0+Z2NfbmV4dCA9IG5vZGU7DQorCWxpc3QtPmdjX3By
 > ZXYgPSBub2RlOw0KK30NCisNCitzdGF0aWMgdm9pZA0KK0xJU1RfUkVNT1ZFKFB5R0NJbmZv
 > ICpub2RlKQ0KK3sNCisJbm9kZS0+Z2NfcHJldi0+Z2NfbmV4dCA9IG5vZGUtPmdjX25leHQ7
 > DQorCW5vZGUtPmdjX25leHQtPmdjX3ByZXYgPSBub2RlLT5nY19wcmV2Ow0KK30NCisNCitz
 > dGF0aWMgdm9pZCANCitMSVNUX01PVkUoUHlHQ0luZm8gKmZyb20sIFB5R0NJbmZvICp0bykN
 > Cit7DQorCWlmIChmcm9tLT5nY19uZXh0ID09IGZyb20pIHsNCisJCS8qIGVtcHR5IGZyb20g
 > bGlzdCAqLw0KKwkJTElTVF9JTklUKHRvKTsNCisJfSBlbHNlIHsNCisJCXRvLT5nY19uZXh0
 > ID0gZnJvbS0+Z2NfbmV4dDsNCisJCXRvLT5nY19uZXh0LT5nY19wcmV2ID0gdG87DQorCQl0
 > by0+Z2NfcHJldiA9IGZyb20tPmdjX3ByZXY7DQorCQl0by0+Z2NfcHJldi0+Z2NfbmV4dCA9
 > IHRvOw0KKwl9DQorCUxJU1RfSU5JVChmcm9tKTsNCit9DQorDQorLyogYXBwZW5kIGEgbGlz
 > dCBvbnRvIGFub3RoZXIgbGlzdCAqLw0KK3N0YXRpYyB2b2lkDQorTElTVF9MQVBQRU5EKFB5
 > R0NJbmZvICpmcm9tLCBQeUdDSW5mbyAqdG8pDQorew0KKwlQeUdDSW5mbyAqdGFpbDsNCisJ
 > aWYgKGZyb20tPmdjX25leHQgIT0gZnJvbSkgew0KKwkJdGFpbCA9IHRvLT5nY19wcmV2Ow0K
 > KwkJdGFpbC0+Z2NfbmV4dCA9IGZyb20tPmdjX25leHQ7DQorCQl0YWlsLT5nY19uZXh0LT5n
 > Y19wcmV2ID0gdGFpbDsNCisJCXRvLT5nY19wcmV2ID0gZnJvbS0+Z2NfcHJldjsNCisJCXRv
 > LT5nY19wcmV2LT5nY19uZXh0ID0gdG87DQorCX0NCisJTElTVF9JTklUKGZyb20pOw0KK30N
 > CisNCitzdGF0aWMgbG9uZw0KK0xJU1RfU0laRShQeUdDSW5mbyAqbGlzdCkNCit7DQorCVB5
 > R0NJbmZvICpnYzsNCisJbG9uZyBuID0gMDsNCisJZm9yIChnYyA9IGxpc3QtPmdjX25leHQ7
 > IGdjICE9IGxpc3Q7IGdjID0gZ2MtPmdjX25leHQpIHsNCisJCW4rKzsNCisJfQ0KKwlyZXR1
 > cm4gbjsNCit9DQorDQorDQorI2lmZGVmIERFQlVHDQorc3RhdGljIHZvaWQNCitsaXN0X3By
 > aW50KFB5R0NJbmZvICpsaXN0KSB7DQorCVB5R0NJbmZvICpvcCA9IGxpc3Q7DQorCWRvIHsN
 > CisJCXByaW50ZigiJXggJXggJXggJWRcbiIsIChsb25nKW9wLCAobG9uZylvcC0+Z2NfcHJl
 > diwgDQorCQkJCShsb25nKW9wLT5nY19uZXh0LCAobG9uZylvcC0+Z2NfcmVmcyk7DQorDQor
 > CQlvcCA9IG9wLT5nY19uZXh0Ow0KKwl9IHdoaWxlIChvcCAhPSBsaXN0KTsNCit9DQorI2Vu
 > ZGlmDQorDQorLyoqKiBlbmQgb2YgbGlzdCBzdHVmZiAqKiovDQorDQorDQorLyogU2V0IGFs
 > bCBnY19yZWZzID0gb2JfcmVmY250ICovDQorc3RhdGljIHZvaWQNCit1cGRhdGVfcmVmcyhQ
 > eUdDSW5mbyAqY29udGFpbmVycykNCit7DQorCVB5R0NJbmZvICpnYyA9IGNvbnRhaW5lcnMt
 > PmdjX25leHQ7DQorCWZvciAoOyBnYyAhPSBjb250YWluZXJzOyBnYz1nYy0+Z2NfbmV4dCkg
 > ew0KKwkJZ2MtPmdjX3JlZnMgPSBQWV9HQ09CSihnYyktPm9iX3JlZmNudDsNCisJfQ0KK30N
 > CisNCitzdGF0aWMgaW50DQordmlzaXRfZGVjcmVmKFB5T2JqZWN0ICpvcCwgdm9pZCAqZGF0
 > YSkNCit7DQorCWlmIChvcCAmJiBQWV9JU0dDKG9wKSkgew0KKwkJUFlfR0NJTkZPX1VOU0FG
 > RShvcCktPmdjX3JlZnMtLTsNCisJfQ0KKwlyZXR1cm4gMTsNCit9DQorDQorLyogU3VidHJh
 > Y3QgaW50ZXJuYWwgcmVmZXJlbmNlcyBmcm9tIGdjX3JlZnMgKi8NCitzdGF0aWMgdm9pZA0K
 > K3N1YnRyYWN0X3JlZnMoUHlHQ0luZm8gKmNvbnRhaW5lcnMpDQorew0KKwlyZWN1cnNlcHJv
 > YyByZWN1cnNlOw0KKwlQeUdDSW5mbyAqZ2MgPSBjb250YWluZXJzLT5nY19uZXh0Ow0KKwlm
 > b3IgKDsgZ2MgIT0gY29udGFpbmVyczsgZ2M9Z2MtPmdjX25leHQpIHsNCisJCXJlY3Vyc2Ug
 > PSBQWV9HQ09CSihnYyktPm9iX3R5cGUtPnRwX3JlY3Vyc2U7DQorCQkodm9pZCkgcmVjdXJz
 > ZShQWV9HQ09CSihnYyksDQorCQkJICAgICAgICh2aXNpdHByb2MpdmlzaXRfZGVjcmVmLA0K
 > KwkJCSAgICAgICBOVUxMKTsNCisJfQ0KK30NCisNCisvKiBBcHBlbmQgb2JqZWN0cyB3aXRo
 > IGdjX3JlZnMgPiAwIHRvIHJvb3RzIGxpc3QgKi8NCitzdGF0aWMgdm9pZA0KK21vdmVfcm9v
 > dHMoUHlHQ0luZm8gKmNvbnRhaW5lcnMsIFB5R0NJbmZvICpyb290cykNCit7DQorCVB5R0NJ
 > bmZvICpuZXh0Ow0KKwlQeUdDSW5mbyAqZ2MgPSBjb250YWluZXJzLT5nY19uZXh0Ow0KKwl3
 > aGlsZSAoZ2MgIT0gY29udGFpbmVycykgew0KKwkJbmV4dCA9IGdjLT5nY19uZXh0Ow0KKwkJ
 > aWYgKGdjLT5nY19yZWZzID4gMCkgew0KKwkJCUxJU1RfUkVNT1ZFKGdjKTsNCisJCQlMSVNU
 > X0FQUEVORChnYywgcm9vdHMpOw0KKwkJCWdjLT5nY19yZWZzID0gR0NfTU9WRUQ7DQorCQl9
 > DQorCQlnYyA9IG5leHQ7DQorCX0NCit9DQorDQorc3RhdGljIGludA0KK3Zpc2l0X3JlYWNo
 > YWJsZShQeU9iamVjdCAqb3AsIFB5R0NJbmZvICpyb290cykNCit7DQorCVB5R0NJbmZvICpn
 > YyA9IFBZX0dDSU5GTyhvcCk7DQorCWlmIChnYyAmJiBnYy0+Z2NfcmVmcyAhPSBHQ19NT1ZF
 > RCkgew0KKwkJTElTVF9SRU1PVkUoZ2MpOw0KKwkJTElTVF9BUFBFTkQoZ2MsIHJvb3RzKTsN
 > CisJCWdjLT5nY19yZWZzID0gR0NfTU9WRUQ7DQorCX0NCisJcmV0dXJuIDE7DQorfQ0KKw0K
 > Ky8qIE1vdmUgb2JqZWN0cyByZWZlcmVuY2VkIGZyb20gcmVhY2hhYmxlIHRvIHJlYWNoYWJs
 > ZSBzZXQuICovDQorc3RhdGljIHZvaWQNCittb3ZlX3Jvb3RfcmVhY2hhYmxlKFB5R0NJbmZv
 > ICpyZWFjaGFibGUpDQorew0KKwlyZWN1cnNlcHJvYyByZWN1cnNlOw0KKwlQeUdDSW5mbyAq
 > Z2MgPSByZWFjaGFibGUtPmdjX25leHQ7DQorCWZvciAoOyBnYyAhPSByZWFjaGFibGU7IGdj
 > PWdjLT5nY19uZXh0KSB7DQorCQkvKiBjYXJlZnVsLCByZWFjaGFibGUgbGlzdCBpcyBncm93
 > aW5nIGhlcmUgKi8NCisJCVB5T2JqZWN0ICpvcCA9IFBZX0dDT0JKKGdjKTsNCisJCXJlY3Vy
 > c2UgPSBvcC0+b2JfdHlwZS0+dHBfcmVjdXJzZTsNCisJCSh2b2lkKSByZWN1cnNlKG9wLA0K
 > KwkJCSAgICAgICAodmlzaXRwcm9jKXZpc2l0X3JlYWNoYWJsZSwNCisJCQkgICAgICAgKHZv
 > aWQgKilyZWFjaGFibGUpOw0KKwl9DQorfQ0KKw0KKy8qIG1vdmUgYWxsIG9iamVjdHMgd2l0
 > aCBmaW5hbGl6ZXJzIChpbnN0YW5jZXMgd2l0aCBfX2RlbF9fKSAqLw0KK3N0YXRpYyB2b2lk
 > DQorbW92ZV9maW5hbGl6ZXJzKFB5R0NJbmZvICp1bnJlYWNoYWJsZSwgUHlHQ0luZm8gKmZp
 > bmFsaXplcnMpDQorew0KKwlQeUdDSW5mbyAqbmV4dDsNCisJUHlHQ0luZm8gKmdjID0gdW5y
 > ZWFjaGFibGUtPmdjX25leHQ7DQorCXN0YXRpYyBQeU9iamVjdCAqZGVsc3RyOw0KKwlpZiAo
 > ZGVsc3RyID09IE5VTEwpIHsNCisJCWRlbHN0ciA9IFB5U3RyaW5nX0ludGVybkZyb21TdHJp
 > bmcoIl9fZGVsX18iKTsNCisJfQ0KKwlmb3IgKDsgZ2MgIT0gdW5yZWFjaGFibGU7IGdjPW5l
 > eHQpIHsNCisJCVB5T2JqZWN0ICpvcCA9IFBZX0dDT0JKKGdjKTsNCisJCW5leHQgPSBnYy0+
 > Z2NfbmV4dDsNCisJCWlmIChQeUluc3RhbmNlX0NoZWNrKG9wKSAmJiBQeU9iamVjdF9IYXNB
 > dHRyKG9wLCBkZWxzdHIpKSB7DQorCQkJTElTVF9SRU1PVkUoZ2MpOw0KKwkJCUxJU1RfQVBQ
 > RU5EKGdjLCBmaW5hbGl6ZXJzKTsNCisJCX0NCisJfQ0KK30NCisNCisNCisvKiBjYWxsZWQg
 > YnkgdHBfcmVjdXJzZSAqLw0KK3N0YXRpYyBpbnQNCit2aXNpdF9maW5hbGl6ZXJfcmVhY2hh
 > YmxlKFB5T2JqZWN0ICpvcCwgUHlHQ0luZm8gKmZpbmFsaXplcnMpDQorew0KKwlQeUdDSW5m
 > byAqZ2MgPSBQWV9HQ0lORk8ob3ApOw0KKwlpZiAoZ2MgJiYgZ2MtPmdjX3JlZnMgIT0gR0Nf
 > TU9WRUQpIHsNCisJCUxJU1RfUkVNT1ZFKGdjKTsNCisJCUxJU1RfQVBQRU5EKGdjLCBmaW5h
 > bGl6ZXJzKTsNCisJCWdjLT5nY19yZWZzID0gR0NfTU9WRUQ7DQorCX0NCisJcmV0dXJuIDE7
 > DQorfQ0KKw0KKy8qIE1vdmUgb2JqZWN0cyByZWZlcmVuY2VkIGZyb20gcm9vdHMgdG8gcm9v
 > dHMgKi8NCitzdGF0aWMgdm9pZA0KK21vdmVfZmluYWxpemVyX3JlYWNoYWJsZShQeUdDSW5m
 > byAqZmluYWxpemVycykNCit7DQorCXJlY3Vyc2Vwcm9jIHJlY3Vyc2U7DQorCVB5R0NJbmZv
 > ICpnYyA9IGZpbmFsaXplcnMtPmdjX25leHQ7DQorCWZvciAoOyBnYyAhPSBmaW5hbGl6ZXJz
 > OyBnYz1nYy0+Z2NfbmV4dCkgew0KKwkJLyogY2FyZWZ1bCwgZmluYWxpemVycyBsaXN0IGlz
 > IGdyb3dpbmcgaGVyZSAqLw0KKwkJcmVjdXJzZSA9IFBZX0dDT0JKKGdjKS0+b2JfdHlwZS0+
 > dHBfcmVjdXJzZTsNCisJCSh2b2lkKSByZWN1cnNlKFBZX0dDT0JKKGdjKSwgDQorCQkJICAg
 > ICAgICh2aXNpdHByb2MpdmlzaXRfZmluYWxpemVyX3JlYWNoYWJsZSwNCisJCQkgICAgICAg
 > KHZvaWQgKilmaW5hbGl6ZXJzKTsNCisJfQ0KK30NCisNCitzdGF0aWMgdm9pZA0KK2RlYnVn
 > X2luc3RhbmNlKFB5T2JqZWN0ICpvdXRwdXQsIGNoYXIgKm1zZywgUHlJbnN0YW5jZU9iamVj
 > dCAqaW5zdCkNCit7DQorCWNoYXIgYnVmWzIwMF07DQorCWNoYXIgKmNuYW1lOw0KKwkvKiBi
 > ZSBjYXJlZnVsIG5vdCB0byBjcmVhdGUgbmV3IGRpY3Rpb25hcmllcyAqLw0KKwlQeU9iamVj
 > dCAqY2xhc3NuYW1lID0gaW5zdC0+aW5fY2xhc3MtPmNsX25hbWU7DQorCWlmIChjbGFzc25h
 > bWUgIT0gTlVMTCAmJiBQeVN0cmluZ19DaGVjayhjbGFzc25hbWUpKQ0KKwkJY25hbWUgPSBQ
 > eVN0cmluZ19Bc1N0cmluZyhjbGFzc25hbWUpOw0KKwllbHNlDQorCQljbmFtZSA9ICI/IjsN
 > CisJc3ByaW50ZihidWYsICJnYzogJXM8JS4xMDBzIGluc3RhbmNlIGF0ICVseD5cbiIsIA0K
 > KwkJCW1zZywgY25hbWUsIChsb25nKWluc3QpOw0KKwlQeUZpbGVfV3JpdGVTdHJpbmcoYnVm
 > LCBvdXRwdXQpOw0KK30NCisNCitzdGF0aWMgdm9pZA0KK2RlYnVnX2N5Y2xlKFB5T2JqZWN0
 > ICpvdXRwdXQsIGNoYXIgKm1zZywgUHlPYmplY3QgKm9wKQ0KK3sNCisJaWYgKChkZWJ1ZyAm
 > IERFQlVHX0lOU1RBTkNFUykgJiYgUHlJbnN0YW5jZV9DaGVjayhvcCkpIHsNCisJCWRlYnVn
 > X2luc3RhbmNlKG91dHB1dCwgbXNnLCAoUHlJbnN0YW5jZU9iamVjdCAqKW9wKTsNCisJfSBl
 > bHNlIGlmIChkZWJ1ZyAmIERFQlVHX09CSkVDVFMpIHsNCisJCWNoYXIgYnVmWzIwMF07DQor
 > CQlzcHJpbnRmKGJ1ZiwgImdjOiAlczxPYmplY3QgMHgleD5cbiIsIG1zZywgKGxvbmcpb3Ap
 > Ow0KKwkJUHlGaWxlX1dyaXRlU3RyaW5nKGJ1Ziwgb3V0cHV0KTsNCisJfQ0KK30NCisNCisv
 > KiBIYW5kbGUgdW5jb2xsZWN0YWJsZSBnYXJiYWdlIChjeWNsZXMgd2l0aCBmaW5hbGl6ZXJz
 > KS4gKi8NCitzdGF0aWMgdm9pZA0KK2hhbmRsZV9maW5hbGl6ZXJzKFB5R0NJbmZvICpmaW5h
 > bGl6ZXJzLCBQeUdDSW5mbyAqb2xkKQ0KK3sNCisJUHlHQ0luZm8gKmdjOw0KKwlpZiAoZ2Fy
 > YmFnZSA9PSBOVUxMKSB7DQorCQlnYXJiYWdlID0gUHlMaXN0X05ldygwKTsNCisJfQ0KKwlm
 > b3IgKGdjID0gZmluYWxpemVycy0+Z2NfbmV4dDsgZ2MgIT0gZmluYWxpemVyczsNCisJCQln
 > YyA9IGZpbmFsaXplcnMtPmdjX25leHQpIHsNCisJCVB5T2JqZWN0ICpvcCA9IFBZX0dDT0JK
 > KGdjKTsNCisJCS8qIEFkZCBhbGwgaW5zdGFuY2VzIHRvIGEgUHl0aG9uIGFjY2Vzc2libGUg
 > bGlzdCBvZiBnYXJiYWdlICovDQorCQlpZiAoUHlJbnN0YW5jZV9DaGVjayhvcCkpIHsNCisJ
 > CQlQeUxpc3RfQXBwZW5kKGdhcmJhZ2UsIG9wKTsNCisJCX0NCisJCS8qIFdlIGFzc3VtZSB0
 > aGF0IGFsbCBvYmplY3RzIGluIGZpbmFsaXplcnMgYXJlIHJlYWNoYWJsZSBmcm9tDQorCQkg
 > KiBpbnN0YW5jZXMuICBPbmNlIHdlIGFkZCB0aGUgaW5zdGFuY2VzIHRvIHRoZSBnYXJiYWdl
 > IGxpc3QNCisJCSAqIGV2ZXJ5dGhpbmcgaXMgcmVhY2hhYmxlIGZyb20gUHl0aG9uIGFnYWlu
 > LiAqLw0KKwkJTElTVF9SRU1PVkUoZ2MpOw0KKwkJTElTVF9BUFBFTkQoZ2MsIG9sZCk7DQor
 > CX0NCit9DQorDQorLyogQnJlYWsgcmVmZXJlbmNlIGN5Y2xlcyBieSBjbGVhcmluZyB0aGUg
 > Y29udGFpbmVycyBpbnZvbHZlZC4gIFRoaXMgaXMNCisgKiB0cmlja3kgYnVzaW5lc3MgYXMg
 > dGhlIGxpc3RzIGNhbiBiZSBjaGFuZ2luZyBhbmQgd2UgZG9uJ3Qga25vdyB3aGljaA0KKyAq
 > IG9iamVjdHMgbWF5IGJlIGZyZWVkLiAgSXQgaXMgcG9zc2libGUgSSBzY3Jld2VkIHNvbWV0
 > aGluZyB1cCBoZXJlLiAqLw0KK3N0YXRpYyB2b2lkDQorZGVsZXRlX2dhcmJhZ2UoUHlHQ0lu
 > Zm8gKnVucmVhY2hhYmxlLCBQeUdDSW5mbyAqb2xkKQ0KK3sNCisJaW5xdWlyeSBjbGVhcjsN
 > CisNCisJd2hpbGUgKHVucmVhY2hhYmxlLT5nY19uZXh0ICE9IHVucmVhY2hhYmxlKSB7DQor
 > CQlQeUdDSW5mbyAqZ2MgPSB1bnJlYWNoYWJsZS0+Z2NfbmV4dDsNCisJCVB5T2JqZWN0ICpv
 > cCA9IFBZX0dDT0JKKGdjKTsNCisJCWlmICgoY2xlYXIgPSBvcC0+b2JfdHlwZS0+dHBfY2xl
 > YXIpICE9IE5VTEwpIHsNCisJCQlQeV9JTkNSRUYob3ApOw0KKwkJCWNsZWFyKChQeU9iamVj
 > dCAqKW9wKTsNCisJCQlQeV9ERUNSRUYob3ApOw0KKwkJfQ0KKwkJLyogb25seSB0cnkgdG8g
 > Y2FsbCB0cF9jbGVhciBvbmNlIGZvciBlYWNoIG9iamVjdCAqLw0KKwkJaWYgKHVucmVhY2hh
 > YmxlLT5nY19uZXh0ID09IGdjKSB7DQorCQkJLyogc3RpbGwgYWxpdmUsIG1vdmUgaXQsIGl0
 > IG1heSBkaWUgbGF0ZXIgKi8NCisJCQlMSVNUX1JFTU9WRShnYyk7DQorCQkJTElTVF9BUFBF
 > TkQoZ2MsIG9sZCk7DQorCQl9DQorCX0NCit9DQorDQorLyogVGhpcyBpcyB0aGUgbWFpbiBm
 > dW5jdGlvbi4gIFJlYWQgdGhpcyB0byB1bmRlcnN0YW5kIGhvdyB0aGUNCisgKiBjb2xsZWN0
 > aW9uIHByb2Nlc3Mgd29ya3MuICovDQorc3RhdGljIGxvbmcNCitjb2xsZWN0KFB5R0NJbmZv
 > ICp5b3VuZywgUHlHQ0luZm8gKm9sZCkNCit7DQorCWxvbmcgbiA9IDA7DQorCWxvbmcgbSA9
 > IDA7DQorCVB5R0NJbmZvIHJlYWNoYWJsZTsNCisJUHlHQ0luZm8gdW5yZWFjaGFibGU7DQor
 > CVB5R0NJbmZvIGZpbmFsaXplcnM7DQorCVB5R0NJbmZvICpnYzsNCisJUHlPYmplY3QgKm91
 > dHB1dCA9IE5VTEw7DQorDQorCWlmIChkZWJ1Zykgew0KKwkJb3V0cHV0ID0gUHlTeXNfR2V0
 > T2JqZWN0KCJzdGRlcnIiKTsNCisJfQ0KKwlpZiAoZGVidWcgJiBERUJVR19TVEFUUykgew0K
 > KwkJY2hhciBidWZbMTAwXTsNCisJCXNwcmludGYoYnVmLCAiZ2M6IGNvbGxlY3RpbmcgZ2Vu
 > ZXJhdGlvbiAlZC4uLlxuIiwgZ2VuZXJhdGlvbik7DQorCQlQeUZpbGVfV3JpdGVTdHJpbmco
 > YnVmLG91dHB1dCk7DQorCQlzcHJpbnRmKGJ1ZiwgImdjOiBvYmplY3RzIGluIGVhY2ggZ2Vu
 > ZXJhdGlvbjogJWQgJWQgJWRcbiIsDQorCQkJTElTVF9TSVpFKCZnZW5lcmF0aW9uMCksDQor
 > CQkJTElTVF9TSVpFKCZnZW5lcmF0aW9uMSksDQorCQkJTElTVF9TSVpFKCZnZW5lcmF0aW9u
 > MikpOw0KKwkJUHlGaWxlX1dyaXRlU3RyaW5nKGJ1ZixvdXRwdXQpOw0KKwl9DQorDQorCS8q
 > IFVzaW5nIG9iX3JlZmNudCBhbmQgZ2NfcmVmcywgY2FsY3VsYXRlIHdoaWNoIG9iamVjdHMg
 > aW4gdGhlDQorCSAqIGNvbnRhaW5lciBzZXQgYXJlIHJlYWNoYWJsZSBmcm9tIG91dHNpZGUg
 > dGhlIHNldCAoaWUuIGhhdmUgYQ0KKwkgKiByZWZjb3VudCBncmVhdGVyIHRoYW4gMCB3aGVu
 > IGFsbCB0aGUgcmVmZXJlbmNlcyB3aXRoaW4gdGhlDQorCSAqIHNldCBhcmUgdGFrZW4gaW50
 > byBhY2NvdW50ICovDQorCXVwZGF0ZV9yZWZzKHlvdW5nKTsNCisJc3VidHJhY3RfcmVmcyh5
 > b3VuZyk7DQorDQorCS8qIE1vdmUgZXZlcnl0aGluZyByZWFjaGFibGUgZnJvbSBvdXRzaWRl
 > IHRoZSBzZXQgaW50byB0aGUNCisJICogcmVhY2hhYmxlIHNldCAoaWUuIGdjX3JlZnMgPiAw
 > KS4gIE5leHQsIG1vdmUgZXZlcnl0aGluZw0KKwkgKiByZWFjaGFibGUgZnJvbSBvYmplY3Rz
 > IGluIHRoZSByZWFjaGFibGUgc2V0LiAqLw0KKwlMSVNUX0lOSVQoJnJlYWNoYWJsZSk7DQor
 > CW1vdmVfcm9vdHMoeW91bmcsICZyZWFjaGFibGUpOw0KKwltb3ZlX3Jvb3RfcmVhY2hhYmxl
 > KCZyZWFjaGFibGUpOw0KKw0KKwkvKiBtb3ZlIHVucmVhY2hhYmxlIG9iamVjdHMgdG8gYSB0
 > ZW1wb3JhcnkgbGlzdCwgbmV3IG9iamVjdHMgY2FuIGJlDQorCSAqIGFsbG9jYXRlZCBhZnRl
 > ciB0aGlzIHBvaW50ICovDQorCUxJU1RfSU5JVCgmdW5yZWFjaGFibGUpOw0KKwlMSVNUX01P
 > VkUoeW91bmcsICZ1bnJlYWNoYWJsZSk7DQorCS8vTElTVF9JTklUKHlvdW5nKTsNCisNCisJ
 > LyogbW92ZSByZWFjaGFibGUgb2JqZWN0cyB0byBuZXh0IGdlbmVyYXRpb24gKi8NCisJTElT
 > VF9MQVBQRU5EKCZyZWFjaGFibGUsIG9sZCk7DQorDQorCS8qIE1vdmUgb2JqZWN0cyByZWFj
 > aGFibGUgZnJvbSBmaW5hbGl6ZXJzLCB3ZSBjYW4ndCBzYWZlbHkgZGVsZXRlDQorCSAqIHRo
 > ZW0uICBQeXRob24gcHJvZ3JhbW1lcnMgc2hvdWxkIHRha2UgY2FyZSBub3QgdG8gY3JlYXRl
 > IHN1Y2gNCisJICogdGhpbmdzLiAgRm9yIFB5dGhvbiBmaW5hbGl6ZXJzIG1lYW5zIGluc3Rh
 > bmNlIG9iamVjdHMgd2l0aA0KKwkgKiBfX2RlbF9fIG1ldGhvZHMuICovDQorCUxJU1RfSU5J
 > VCgmZmluYWxpemVycyk7DQorCW1vdmVfZmluYWxpemVycygmdW5yZWFjaGFibGUsICZmaW5h
 > bGl6ZXJzKTsNCisJbW92ZV9maW5hbGl6ZXJfcmVhY2hhYmxlKCZmaW5hbGl6ZXJzKTsNCisN
 > CisJLyogQ29sbGVjdCBzdGF0aXN0aWNzIG9uIGNvbGxlY3RhYmxlIG9iamVjdHMgZm91bmQg
 > YW5kIHByaW50DQorCSAqIGRlYnVnZ2luZyBpbmZvcm1hdGlvbi4gKi8NCisJZm9yIChnYyA9
 > IHVucmVhY2hhYmxlLmdjX25leHQ7IGdjICE9ICZ1bnJlYWNoYWJsZTsNCisJCQlnYyA9IGdj
 > LT5nY19uZXh0KSB7DQorCQltKys7DQorCQlpZiAob3V0cHV0ICE9IE5VTEwgJiYgKGRlYnVn
 > ICYgREVCVUdfQ09MTEVDVEFCTEUpKSB7DQorCQkJZGVidWdfY3ljbGUob3V0cHV0LCAiY29s
 > bGVjdGFibGUgIiwgUFlfR0NPQkooZ2MpKTsNCisJCX0NCisJfQ0KKwkvKiBjYWxsIHRwX2Ns
 > ZWFyIG9uIG9iamVjdHMgaW4gdGhlIGNvbGxlY3RhYmxlIHNldC4gIFRoaXMgd2lsbCBjYXVz
 > ZQ0KKwkgKiB0aGUgcmVmZXJlbmNlIGN5Y2xlcyB0byBiZSBicm9rZW4uIEl0IG1heSBhbHNv
 > IGNhdXNlIHNvbWUgb2JqZWN0cyBpbg0KKwkgKiBmaW5hbGl6ZXJzIHRvIGJlIGZyZWVkICov
 > DQorCWRlbGV0ZV9nYXJiYWdlKCZ1bnJlYWNoYWJsZSwgb2xkKTsNCisNCisJLyogQ29sbGVj
 > dCBzdGF0aXN0aWNzIG9uIHVuY29sbGVjdGFibGUgb2JqZWN0cyBmb3VuZCBhbmQgcHJpbnQN
 > CisJICogZGVidWdnaW5nIGluZm9ybWF0aW9uLiAqLw0KKwlmb3IgKGdjID0gZmluYWxpemVy
 > cy5nY19uZXh0OyBnYyAhPSAmZmluYWxpemVyczsNCisJCQlnYyA9IGdjLT5nY19uZXh0KSB7
 > DQorCQluKys7DQorCQlpZiAob3V0cHV0ICE9IE5VTEwgJiYgKGRlYnVnICYgREVCVUdfVU5D
 > T0xMRUNUQUJMRSkpIHsNCisJCQlkZWJ1Z19jeWNsZShvdXRwdXQsICJ1bmNvbGxlY3RhYmxl
 > ICIsIFBZX0dDT0JKKGdjKSk7DQorCQl9DQorCX0NCisJaWYgKG91dHB1dCAhPSBOVUxMICYm
 > IChkZWJ1ZyAmIERFQlVHX1NUQVRTKSkgew0KKwkJaWYgKG0gPT0gMCAmJiBuID09IDApIHsN
 > CisJCQlQeUZpbGVfV3JpdGVTdHJpbmcoImdjOiBkb25lLlxuIiwgb3V0cHV0KTsNCisJCX0g
 > ZWxzZSB7DQorCQkJY2hhciBidWZbMjAwXTsNCisJCQlzcHJpbnRmKGJ1ZiwNCisJCQkJImdj
 > OiBkb25lLCAlZCB1bnJlYWNoYWJsZSwgJWQgdW5jb2xsZWN0YWJsZS5cbiIsDQorCQkJCW4r
 > bSwgbik7DQorCQkJUHlGaWxlX1dyaXRlU3RyaW5nKGJ1Ziwgb3V0cHV0KTsNCisJCX0NCisJ
 > fQ0KKw0KKwkvKiBBcHBlbmQgaW5zdGFuY2VzIGluIHRoZSB1bmNvbGxlY3RhYmxlIHNldCB0
 > byBhIFB5dGhvbg0KKwkgKiByZWFjaGFibGUgbGlzdCBvZiBnYXJiYWdlLiAgVGhlIHByb2dy
 > YW1tZXIgaGFzIHRvIGRlYWwgd2l0aA0KKwkgKiB0aGlzIGlmIHRoZXkgaW5zaXN0IG9uIGNy
 > ZWF0aW5nIHRoaXMgdHlwZSBvZiBzdHJ1Y3R1cmUuICovDQorCWhhbmRsZV9maW5hbGl6ZXJz
 > KCZmaW5hbGl6ZXJzLCBvbGQpOw0KKw0KKwlhbGxvY2F0ZWQgPSAwOw0KKwlQeUVycl9DbGVh
 > cigpOyAvKiBpbiBjYXNlIHdyaXRpbmcgdG8gc3lzLnN0ZGVyciBmYWlsZWQgKi8NCisJcmV0
 > dXJuIG4rbTsNCit9DQorDQorc3RhdGljIGxvbmcNCitjb2xsZWN0X2dlbmVyYXRpb25zKHZv
 > aWQpDQorew0KKwlzdGF0aWMgbG9uZyBjb2xsZWN0aW9uczAgPSAwOw0KKwlzdGF0aWMgbG9u
 > ZyBjb2xsZWN0aW9uczEgPSAwOw0KKwlsb25nIG47DQorDQorDQorCWlmIChjb2xsZWN0aW9u
 > czEgPiB0aHJlc2hvbGQyKSB7DQorCQlnZW5lcmF0aW9uID0gMjsNCisJCUxJU1RfTEFQUEVO
 > RCgmZ2VuZXJhdGlvbjAsICZnZW5lcmF0aW9uMik7DQorCQlMSVNUX0xBUFBFTkQoJmdlbmVy
 > YXRpb24xLCAmZ2VuZXJhdGlvbjIpOw0KKwkJaWYgKGdlbmVyYXRpb24yLmdjX25leHQgIT0g
 > JmdlbmVyYXRpb24yKSB7DQorCQkJbiA9IGNvbGxlY3QoJmdlbmVyYXRpb24yLCAmZ2VuZXJh
 > dGlvbjIpOw0KKwkJfQ0KKwkJY29sbGVjdGlvbnMxID0gMDsNCisJfSBlbHNlIGlmIChjb2xs
 > ZWN0aW9uczAgPiB0aHJlc2hvbGQxKSB7DQorCQlnZW5lcmF0aW9uID0gMTsNCisJCWNvbGxl
 > Y3Rpb25zMSsrOw0KKwkJTElTVF9MQVBQRU5EKCZnZW5lcmF0aW9uMCwgJmdlbmVyYXRpb24x
 > KTsNCisJCWlmIChnZW5lcmF0aW9uMS5nY19uZXh0ICE9ICZnZW5lcmF0aW9uMSkgew0KKwkJ
 > CW4gPSBjb2xsZWN0KCZnZW5lcmF0aW9uMSwgJmdlbmVyYXRpb24yKTsNCisJCX0NCisJCWNv
 > bGxlY3Rpb25zMCA9IDA7DQorCX0gZWxzZSB7DQorCQlnZW5lcmF0aW9uID0gMDsNCisJCWNv
 > bGxlY3Rpb25zMCsrOw0KKwkJaWYgKGdlbmVyYXRpb24wLmdjX25leHQgIT0gJmdlbmVyYXRp
 > b24wKSB7DQorCQkJbiA9IGNvbGxlY3QoJmdlbmVyYXRpb24wLCAmZ2VuZXJhdGlvbjEpOw0K
 > KwkJfQ0KKwl9DQorCXJldHVybiBuOw0KK30NCisNCit2b2lkDQorUHlHQ19JbnNlcnQoUHlP
 > YmplY3QgKm9wKQ0KK3sNCisJLyogY29sbGVjdGlvbiBsb2NrIHNpbmNlIGNvbGxlY3Rpbmcg
 > bWF5IGNhdXNlIGFsbG9jYXRpb25zICovDQorCXN0YXRpYyBpbnQgY29sbGVjdGluZyA9IDA7
 > DQorDQorI2lmZGVmIERFQlVHDQorCS8qZnByaW50ZihzdGRlcnIsICJJbnNlcnQgMHgleFxu
 > Iiwgb3ApOyovDQorI2VuZGlmDQorCWlmICh0aHJlc2hvbGQwICYmIGFsbG9jYXRlZCA+IHRo
 > cmVzaG9sZDAgJiYgIWNvbGxlY3RpbmcpIHsNCisJCWNvbGxlY3RpbmcrKzsNCisJCWNvbGxl
 > Y3RfZ2VuZXJhdGlvbnMoKTsNCisJCWNvbGxlY3RpbmctLTsNCisJfQ0KKwlhbGxvY2F0ZWQr
 > KzsNCisJTElTVF9BUFBFTkQoUFlfR0NJTkZPKG9wKSwgJmdlbmVyYXRpb24wKTsNCit9DQor
 > DQordm9pZA0KK1B5R0NfUmVtb3ZlKFB5T2JqZWN0ICpvcCkNCit7DQorI2lmZGVmIERFQlVH
 > DQorCS8qZnByaW50ZihzdGRlcnIsICJSZW1vdmUgMHgleFxuIiwgb3ApOyovDQorI2VuZGlm
 > DQorCWlmIChhbGxvY2F0ZWQgPiAwKSB7DQorCQlhbGxvY2F0ZWQtLTsNCisJfQ0KKwlMSVNU
 > X1JFTU9WRShQWV9HQ0lORk8ob3ApKTsNCit9DQorDQorDQorc3RhdGljIGNoYXIgY29sbGVj
 > dF9fZG9jX19bXSA9DQorImNvbGxlY3QoKSAtPiBuXG4iDQorIlxuIg0KKyJSdW4gYSBmdWxs
 > IGNvbGxlY3Rpb24uICBUaGUgbnVtYmVyIG9mIHVucmVhY2hhYmxlIG9iamVjdHMgaXMgcmV0
 > dXJuZWQuXG4iDQorOw0KKw0KK3N0YXRpYyBQeU9iamVjdCAqDQorUHlfY29sbGVjdChzZWxm
 > LCBhcmdzKQ0KKwlQeU9iamVjdCAqc2VsZjsNCisJUHlPYmplY3QgKmFyZ3M7DQorew0KKwls
 > b25nIG47DQorDQorCWlmKCFQeUFyZ19QYXJzZVR1cGxlKGFyZ3MsICIiKSkJLyogY2hlY2sg
 > bm8gYXJncyAqLw0KKwkJcmV0dXJuIE5VTEw7DQorDQorCWdlbmVyYXRpb24gPSAyOw0KKwlM
 > SVNUX0xBUFBFTkQoJmdlbmVyYXRpb24wLCAmZ2VuZXJhdGlvbjIpOw0KKwlMSVNUX0xBUFBF
 > TkQoJmdlbmVyYXRpb24xLCAmZ2VuZXJhdGlvbjIpOw0KKwluID0gY29sbGVjdCgmZ2VuZXJh
 > dGlvbjIsICZnZW5lcmF0aW9uMik7DQorDQorCXJldHVybiBQeV9CdWlsZFZhbHVlKCJpIiwg
 > bik7DQorfQ0KKw0KK3N0YXRpYyBjaGFyIHNldF9kZWJ1Z19fZG9jX19bXSA9IA0KKyJzZXRf
 > ZGVidWcoZmxhZ3MpIC0+IE5vbmVcbiINCisiXG4iDQorIlNldCB0aGUgZ2FyYmFnZSBjb2xs
 > ZWN0aW9uIGRlYnVnZ2luZyBmbGFncy4gRGVidWdnaW5nIGluZm9ybWF0aW9uIGlzXG4iDQor
 > IndyaXR0ZW4gdG8gc3lzLnN0ZGVyci5cbiINCisiXG4iDQorImZsYWdzIGlzIGFuIGludGVn
 > ZXIgYW5kIGNhbiBoYXZlIHRoZSBmb2xsb3dpbmcgYml0cyB0dXJuZWQgb246XG4iDQorIlxu
 > Ig0KKyIgIERFQlVHX1NUQVRTIC0gUHJpbnQgc3RhdGlzdGljcyBkdXJpbmcgY29sbGVjdGlv
 > bi5cbiINCisiICBERUJVR19DT0xMRUNUQUJMRSAtIFByaW50IGNvbGxlY3RhYmxlIG9iamVj
 > dHMgZm91bmQuXG4iDQorIiAgREVCVUdfVU5DT0xMRUNUQUJMRSAtIFByaW50IHVucmVhY2hh
 > YmxlIGJ1dCB1bmNvbGxlY3RhYmxlIG9iamVjdHMgZm91bmQuXG4iDQorIiAgREVCVUdfSU5T
 > VEFOQ0VTIC0gUHJpbnQgaW5zdGFuY2Ugb2JqZWN0cy5cbiINCisiICBERUJVR19PQkpFQ1RT
 > IC0gUHJpbnQgb2JqZWN0cyBvdGhlciB0aGFuIGluc3RhbmNlcy5cbiINCisiICBERUJVR19M
 > RUFLIC0gRGVidWcgbGVha2luZyBwcm9ncmFtcyAoZXZlcnl0aGluZyBidXQgU1RBVFMpLlxu
 > Ig0KKzsNCisNCitzdGF0aWMgUHlPYmplY3QgKg0KK1B5X3NldF9kZWJ1ZyhzZWxmLCBhcmdz
 > KQ0KKwlQeU9iamVjdCAqc2VsZjsNCisJUHlPYmplY3QgKmFyZ3M7DQorew0KKwlpZiAoIVB5
 > QXJnX1BhcnNlVHVwbGUoYXJncywgImwiLCAmZGVidWcpKQ0KKwkJcmV0dXJuIE5VTEw7DQor
 > DQorCVB5X0lOQ1JFRihQeV9Ob25lKTsNCisJcmV0dXJuIFB5X05vbmU7DQorfQ0KKw0KK3N0
 > YXRpYyBjaGFyIGdldF9kZWJ1Z19fZG9jX19bXSA9IA0KKyJnZXRfZGVidWcoKSAtPiBmbGFn
 > c1xuIg0KKyJcbiINCisiR2V0IHRoZSBnYXJiYWdlIGNvbGxlY3Rpb24gZGVidWdnaW5nIGZs
 > YWdzLlxuIg0KKzsNCisNCitzdGF0aWMgUHlPYmplY3QgKg0KK1B5X2dldF9kZWJ1ZyhzZWxm
 > LCBhcmdzKQ0KKwlQeU9iamVjdCAqc2VsZjsNCisJUHlPYmplY3QgKmFyZ3M7DQorew0KKwlp
 > ZighUHlBcmdfUGFyc2VUdXBsZShhcmdzLCAiIikpCS8qIG5vIGFyZ3MgKi8NCisJCXJldHVy
 > biBOVUxMOw0KKw0KKwlyZXR1cm4gUHlfQnVpbGRWYWx1ZSgiaSIsIGRlYnVnKTsNCit9DQor
 > DQorc3RhdGljIGNoYXIgc2V0X3RocmVzaF9fZG9jX19bXSA9DQorInNldF90aHJlc2hvbGQo
 > dGhyZXNob2xkMCwgW3RocmVob2xkMSwgdGhyZXNob2xkMl0pIC0+IE5vbmVcbiINCisiXG4i
 > DQorIlNldHMgdGhlIGNvbGxlY3Rpb24gdGhyZXNob2xkcy4gIFNldHRpbmcgdGhyZXNob2xk
 > MCB0byB6ZXJvIGRpc2FibGVzXG4iDQorImNvbGxlY3Rpb24uXG4iDQorOw0KKw0KK3N0YXRp
 > YyBQeU9iamVjdCAqDQorUHlfc2V0X3RocmVzaChzZWxmLCBhcmdzKQ0KKwlQeU9iamVjdCAq
 > c2VsZjsNCisJUHlPYmplY3QgKmFyZ3M7DQorew0KKwlpZiAoIVB5QXJnX1BhcnNlVHVwbGUo
 > YXJncywgIml8aWkiLCAmdGhyZXNob2xkMCwgDQorCQkJCSZ0aHJlc2hvbGQxLCAmdGhyZXNo
 > b2xkMikpDQorCQlyZXR1cm4gTlVMTDsNCisNCisJUHlfSU5DUkVGKFB5X05vbmUpOw0KKwly
 > ZXR1cm4gUHlfTm9uZTsNCit9DQorDQorc3RhdGljIGNoYXIgZ2V0X3RocmVzaF9fZG9jX19b
 > XSA9DQorImdldF90cmVzaG9sZCgpIC0+ICh0aHJlc2hvbGQwLCB0aHJlc2hvbGQxLCB0aHJl
 > c2hvbGQyKVxuIg0KKyJcbiINCisiUmV0dXJuIHRoZSBjdXJyZW50IGNvbGxlY3Rpb24gdGhy
 > ZXNob2xkc1xuIg0KKzsNCisNCitzdGF0aWMgUHlPYmplY3QgKg0KK1B5X2dldF90aHJlc2go
 > c2VsZiwgYXJncykNCisJUHlPYmplY3QgKnNlbGY7DQorCVB5T2JqZWN0ICphcmdzOw0KK3sN
 > CisJaWYoIVB5QXJnX1BhcnNlVHVwbGUoYXJncywgIiIpKQkvKiBubyBhcmdzICovDQorCQly
 > ZXR1cm4gTlVMTDsNCisNCisJcmV0dXJuIFB5X0J1aWxkVmFsdWUoIihpaWkpIiwgdGhyZXNo
 > b2xkMCwgdGhyZXNob2xkMSwgdGhyZXNob2xkMik7DQorfQ0KKw0KKw0KK3N0YXRpYyBjaGFy
 > IGdjX19kb2NfXyBbXSA9DQorIlRoaXMgbW9kdWxlIHByb3ZpZGVzIGFjY2VzcyB0byB0aGUg
 > Z2FyYmFnZSBjb2xsZWN0b3IgZm9yIHJlZmVyZW5jZSBjeWNsZXMuXG4iDQorIlxuIg0KKyJj
 > b2xsZWN0KCkgLS0gRG8gYSBmdWxsIGNvbGxlY3Rpb24gcmlnaHQgbm93LlxuIg0KKyJzZXRf
 > ZGVidWcoKSAtLSBTZXQgZGVidWdnaW5nIGZsYWdzLlxuIg0KKyJnZXRfZGVidWcoKSAtLSBH
 > ZXQgZGVidWdnaW5nIGZsYWdzLlxuIg0KKyJzZXRfdGhyZXNob2xkKCkgLS0gU2V0IHRoZSBj
 > b2xsZWN0aW9uIHRocmVzaG9sZHMuXG4iDQorImdldF90aHJlc2hvbGQoKSAtLSBSZXR1cm4g
 > dGhlIGN1cnJlbnQgdGhlIGNvbGxlY3Rpb24gdGhyZXNob2xkcy5cbiINCis7DQorDQorc3Rh
 > dGljIFB5TWV0aG9kRGVmIEdjTWV0aG9kc1tdID0gew0KKwl7InNldF9kZWJ1ZyIsCQlQeV9z
 > ZXRfZGVidWcsICBNRVRIX1ZBUkFSR1MsIHNldF9kZWJ1Z19fZG9jX199LA0KKwl7ImdldF9k
 > ZWJ1ZyIsCQlQeV9nZXRfZGVidWcsICBNRVRIX1ZBUkFSR1MsIGdldF9kZWJ1Z19fZG9jX199
 > LA0KKwl7InNldF90aHJlc2hvbGQiLAlQeV9zZXRfdGhyZXNoLCBNRVRIX1ZBUkFSR1MsIHNl
 > dF90aHJlc2hfX2RvY19ffSwNCisJeyJnZXRfdGhyZXNob2xkIiwJUHlfZ2V0X3RocmVzaCwg
 > TUVUSF9WQVJBUkdTLCBnZXRfdGhyZXNoX19kb2NfX30sDQorCXsiY29sbGVjdCIsCQlQeV9j
 > b2xsZWN0LCAgICBNRVRIX1ZBUkFSR1MsIGNvbGxlY3RfX2RvY19ffSwNCisJe05VTEwsCU5V
 > TEx9CQkvKiBTZW50aW5lbCAqLw0KK307DQorDQordm9pZA0KK2luaXRnYyh2b2lkKQ0KK3sN
 > CisJUHlPYmplY3QgKm07DQorCVB5T2JqZWN0ICpkOw0KKw0KKwltID0gUHlfSW5pdE1vZHVs
 > ZTQoImdjIiwNCisJCQkgICAgICBHY01ldGhvZHMsDQorCQkJICAgICAgZ2NfX2RvY19fLA0K
 > KwkJCSAgICAgIE5VTEwsDQorCQkJICAgICAgUFlUSE9OX0FQSV9WRVJTSU9OKTsNCisJZCA9
 > IFB5TW9kdWxlX0dldERpY3QobSk7DQorCWlmIChnYXJiYWdlID09IE5VTEwpIHsNCisJCWdh
 > cmJhZ2UgPSBQeUxpc3RfTmV3KDApOw0KKwl9DQorCVB5RGljdF9TZXRJdGVtU3RyaW5nKGQs
 > ICJnYXJiYWdlIiwgZ2FyYmFnZSk7DQorCVB5RGljdF9TZXRJdGVtU3RyaW5nKGQsICJERUJV
 > R19TVEFUUyIsDQorCQkJUHlJbnRfRnJvbUxvbmcoREVCVUdfU1RBVFMpKTsNCisJUHlEaWN0
 > X1NldEl0ZW1TdHJpbmcoZCwgIkRFQlVHX0NPTExFQ1RBQkxFIiwNCisJCQlQeUludF9Gcm9t
 > TG9uZyhERUJVR19DT0xMRUNUQUJMRSkpOw0KKwlQeURpY3RfU2V0SXRlbVN0cmluZyhkLCAi
 > REVCVUdfVU5DT0xMRUNUQUJMRSIsDQorCQkJUHlJbnRfRnJvbUxvbmcoREVCVUdfVU5DT0xM
 > RUNUQUJMRSkpOw0KKwlQeURpY3RfU2V0SXRlbVN0cmluZyhkLCAiREVCVUdfSU5TVEFOQ0VT
 > IiwNCisJCQlQeUludF9Gcm9tTG9uZyhERUJVR19JTlNUQU5DRVMpKTsNCisJUHlEaWN0X1Nl
 > dEl0ZW1TdHJpbmcoZCwgIkRFQlVHX09CSkVDVFMiLA0KKwkJCVB5SW50X0Zyb21Mb25nKERF
 > QlVHX09CSkVDVFMpKTsNCisJUHlEaWN0X1NldEl0ZW1TdHJpbmcoZCwgIkRFQlVHX0xFQUsi
 > LA0KKwkJCVB5SW50X0Zyb21Mb25nKERFQlVHX0xFQUspKTsNCit9DQorDQotLS0gUHl0aG9u
 > LWN2cy9PYmplY3RzL2NsYXNzb2JqZWN0LmMJVHVlIEZlYiAyOSAxNDo1NToxNSAyMDAwDQor
 > KysgUHl0aG9uLWdjL09iamVjdHMvY2xhc3NvYmplY3QuYwlNb24gQXByIDI0IDIxOjE5OjE4
 > IDIwMDANCkBAIC0xMTAsNyArMTEwLDcgQEANCiAJCX0NCiAJCVB5X0lOQ1JFRihiYXNlcyk7
 > DQogCX0NCi0Jb3AgPSBQeU9iamVjdF9ORVcoUHlDbGFzc09iamVjdCwgJlB5Q2xhc3NfVHlw
 > ZSk7DQorCW9wID0gUHlPYmplY3RfTmV3KFB5Q2xhc3NPYmplY3QsICZQeUNsYXNzX1R5cGUp
 > Ow0KIAlpZiAob3AgPT0gTlVMTCkgew0KIAkJUHlfREVDUkVGKGJhc2VzKTsNCiAJCXJldHVy
 > biBOVUxMOw0KQEAgLTEzMSw2ICsxMzEsOSBAQA0KIAlQeV9YSU5DUkVGKG9wLT5jbF9nZXRh
 > dHRyKTsNCiAJUHlfWElOQ1JFRihvcC0+Y2xfc2V0YXR0cik7DQogCVB5X1hJTkNSRUYob3At
 > PmNsX2RlbGF0dHIpOw0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorCVB5R0NfSW5zZXJ0KChQ
 > eU9iamVjdCAqKW9wKTsNCisjZW5kaWYNCiAJcmV0dXJuIChQeU9iamVjdCAqKSBvcDsNCiB9
 > DQogDQpAQCAtMTQwLDEzICsxNDMsMTYgQEANCiBjbGFzc19kZWFsbG9jKG9wKQ0KIAlQeUNs
 > YXNzT2JqZWN0ICpvcDsNCiB7DQorI2lmZGVmIFdJVEhfQ1lDTEVfR0MNCisJUHlHQ19SZW1v
 > dmUoKFB5T2JqZWN0ICopb3ApOw0KKyNlbmRpZg0KIAlQeV9ERUNSRUYob3AtPmNsX2Jhc2Vz
 > KTsNCiAJUHlfREVDUkVGKG9wLT5jbF9kaWN0KTsNCiAJUHlfWERFQ1JFRihvcC0+Y2xfbmFt
 > ZSk7DQogCVB5X1hERUNSRUYob3AtPmNsX2dldGF0dHIpOw0KIAlQeV9YREVDUkVGKG9wLT5j
 > bF9zZXRhdHRyKTsNCiAJUHlfWERFQ1JFRihvcC0+Y2xfZGVsYXR0cik7DQotCWZyZWUoKEFO
 > WSAqKW9wKTsNCisJUHlPYmplY3RfRGVsKChQeU9iamVjdCAqKW9wKTsNCiB9DQogDQogc3Rh
 > dGljIFB5T2JqZWN0ICoNCkBAIC0zODYsNiArMzkyLDIxIEBADQogCXJldHVybiByZXM7DQog
 > fQ0KIA0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorc3RhdGljIGludA0KK2NsYXNzX3JlY3Vy
 > c2UoUHlDbGFzc09iamVjdCAqbywgdmlzaXRwcm9jIHZpc2l0LCB2b2lkICpjbG9zdXJlKQ0K
 > K3sNCisJaWYgKG8tPmNsX2Jhc2VzKSB2aXNpdChvLT5jbF9iYXNlcywgY2xvc3VyZSk7DQor
 > CWlmIChvLT5jbF9kaWN0KSB2aXNpdChvLT5jbF9kaWN0LCBjbG9zdXJlKTsNCisJaWYgKG8t
 > PmNsX25hbWUpIHZpc2l0KG8tPmNsX25hbWUsIGNsb3N1cmUpOw0KKwlpZiAoby0+Y2xfZ2V0
 > YXR0cikgdmlzaXQoby0+Y2xfZ2V0YXR0ciwgY2xvc3VyZSk7DQorCWlmIChvLT5jbF9zZXRh
 > dHRyKSB2aXNpdChvLT5jbF9zZXRhdHRyLCBjbG9zdXJlKTsNCisJaWYgKG8tPmNsX2RlbGF0
 > dHIpIHZpc2l0KG8tPmNsX2RlbGF0dHIsIGNsb3N1cmUpOw0KKwlyZXR1cm4gMTsNCit9DQor
 > I2VuZGlmIC8qIFdJVEhfQ1lDTEVfR0MgKi8NCisNCisNCiBQeVR5cGVPYmplY3QgUHlDbGFz
 > c19UeXBlID0gew0KIAlQeU9iamVjdF9IRUFEX0lOSVQoJlB5VHlwZV9UeXBlKQ0KIAkwLA0K
 > QEAgLTQwNiw2ICs0MjcsMTIgQEANCiAJKHJlcHJmdW5jKWNsYXNzX3N0ciwgLyp0cF9zdHIq
 > Lw0KIAkoZ2V0YXR0cm9mdW5jKWNsYXNzX2dldGF0dHIsIC8qdHBfZ2V0YXR0cm8qLw0KIAko
 > c2V0YXR0cm9mdW5jKWNsYXNzX3NldGF0dHIsIC8qdHBfc2V0YXR0cm8qLw0KKyNpZmRlZiBX
 > SVRIX0NZQ0xFX0dDDQorCTAsCQkvKiB0cF9hc19idWZmZXIgKi8NCisJUHlfVFBGTEFHU19E
 > RUZBVUxUIHwgUHlfVFBGTEFHU19IQVZFX0dDSU5GTywgLyp0cF9mbGFncyovDQorCTAsCQkv
 > KiB0cF9kb2MgKi8NCisJKHJlY3Vyc2Vwcm9jKWNsYXNzX3JlY3Vyc2UsCS8qIHRwX3JlY3Vy
 > c2UgKi8NCisjZW5kaWYNCiB9Ow0KIA0KIGludA0KQEAgLTQ0NCwxMiArNDcxLDE1IEBADQog
 > CQlQeUVycl9CYWRJbnRlcm5hbENhbGwoKTsNCiAJCXJldHVybiBOVUxMOw0KIAl9DQotCWlu
 > c3QgPSBQeU9iamVjdF9ORVcoUHlJbnN0YW5jZU9iamVjdCwgJlB5SW5zdGFuY2VfVHlwZSk7
 > DQorCWluc3QgPSBQeU9iamVjdF9OZXcoUHlJbnN0YW5jZU9iamVjdCwgJlB5SW5zdGFuY2Vf
 > VHlwZSk7DQogCWlmIChpbnN0ID09IE5VTEwpDQogCQlyZXR1cm4gTlVMTDsNCiAJUHlfSU5D
 > UkVGKGNsYXNzKTsNCiAJaW5zdC0+aW5fY2xhc3MgPSAoUHlDbGFzc09iamVjdCAqKWNsYXNz
 > Ow0KIAlpbnN0LT5pbl9kaWN0ID0gUHlEaWN0X05ldygpOw0KKyNpZmRlZiBXSVRIX0NZQ0xF
 > X0dDDQorCVB5R0NfSW5zZXJ0KChQeU9iamVjdCAqKWluc3QpOw0KKyNlbmRpZg0KIAlpZiAo
 > aW5zdC0+aW5fZGljdCA9PSBOVUxMKSB7DQogCQlQeV9ERUNSRUYoaW5zdCk7DQogCQlyZXR1
 > cm4gTlVMTDsNCkBAIC00OTgsMTEgKzUyOCwxNCBAQA0KIAlQeU9iamVjdCAqZXJyb3JfdHlw
 > ZSwgKmVycm9yX3ZhbHVlLCAqZXJyb3JfdHJhY2ViYWNrOw0KIAlQeU9iamVjdCAqZGVsOw0K
 > IAlzdGF0aWMgUHlPYmplY3QgKmRlbHN0cjsNCisJZXh0ZXJuIGxvbmcgX1B5X1JlZlRvdGFs
 > Ow0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorCVB5R0NfUmVtb3ZlKChQeU9iamVjdCAqKWlu
 > c3QpOw0KKyNlbmRpZg0KIAkvKiBDYWxsIHRoZSBfX2RlbF9fIG1ldGhvZCBpZiBpdCBleGlz
 > dHMuICBGaXJzdCB0ZW1wb3JhcmlseQ0KIAkgICByZXZpdmUgdGhlIG9iamVjdCBhbmQgc2F2
 > ZSB0aGUgY3VycmVudCBleGNlcHRpb24sIGlmIGFueS4gKi8NCiAjaWZkZWYgUHlfVFJBQ0Vf
 > UkVGUw0KIAkvKiBtdWNoIHRvbyBjb21wbGljYXRlZCBpZiBQeV9UUkFDRV9SRUZTIGRlZmlu
 > ZWQgKi8NCi0JZXh0ZXJuIGxvbmcgX1B5X1JlZlRvdGFsOw0KIAlpbnN0LT5vYl90eXBlID0g
 > JlB5SW5zdGFuY2VfVHlwZTsNCiAJX1B5X05ld1JlZmVyZW5jZSgoUHlPYmplY3QgKilpbnN0
 > KTsNCiAJX1B5X1JlZlRvdGFsLS07CQkvKiBjb21wZW5zYXRlIGZvciBpbmNyZW1lbnQgaW4g
 > TkVXUkVGICovDQpAQCAtNTUwLDYgKzU4Myw5IEBADQogI2lmZGVmIENPVU5UX0FMTE9DUw0K
 > IAkJaW5zdC0+b2JfdHlwZS0+dHBfZnJlZS0tOw0KICNlbmRpZg0KKyNpZmRlZiBXSVRIX0NZ
 > Q0xFX0dDDQorCQlQeUdDX0luc2VydCgoUHlPYmplY3QgKilpbnN0KTsNCisjZW5kaWYNCiAJ
 > CXJldHVybjsgLyogX19kZWxfXyBhZGRlZCBhIHJlZmVyZW5jZTsgZG9uJ3QgZGVsZXRlIG5v
 > dyAqLw0KIAl9DQogI2lmZGVmIFB5X1RSQUNFX1JFRlMNCkBAIC01NTcsMTEgKzU5MywxMCBA
 > QA0KIAlpbnN0LT5vYl90eXBlLT50cF9mcmVlLS07CS8qIGNvbXBlbnNhdGUgZm9yIGluY3Jl
 > bWVudCBpbiBVTlJFRiAqLw0KICNlbmRpZg0KIAlfUHlfRm9yZ2V0UmVmZXJlbmNlKChQeU9i
 > amVjdCAqKWluc3QpOw0KLQlpbnN0LT5vYl90eXBlID0gTlVMTDsNCiAjZW5kaWYgLyogUHlf
 > VFJBQ0VfUkVGUyAqLw0KIAlQeV9ERUNSRUYoaW5zdC0+aW5fY2xhc3MpOw0KIAlQeV9YREVD
 > UkVGKGluc3QtPmluX2RpY3QpOw0KLQlmcmVlKChBTlkgKilpbnN0KTsNCisJUHlPYmplY3Rf
 > RGVsKChQeU9iamVjdCAqKWluc3QpOw0KIH0NCiANCiBzdGF0aWMgUHlPYmplY3QgKg0KQEAg
 > LTg0MCw2ICs4NzUsMTYgQEANCiAJcmV0dXJuIG91dGNvbWU7DQogfQ0KIA0KKyNpZmRlZiBX
 > SVRIX0NZQ0xFX0dDDQorc3RhdGljIGludA0KK2luc3RhbmNlX3JlY3Vyc2UoUHlJbnN0YW5j
 > ZU9iamVjdCAqbywgdmlzaXRwcm9jIHZpc2l0LCB2b2lkICpjbG9zdXJlKQ0KK3sNCisJaWYg
 > KG8tPmluX2NsYXNzKSB2aXNpdCgoUHlPYmplY3QgKikoby0+aW5fY2xhc3MpLCBjbG9zdXJl
 > KTsNCisJaWYgKG8tPmluX2RpY3QpIHZpc2l0KG8tPmluX2RpY3QsIGNsb3N1cmUpOw0KKwly
 > ZXR1cm4gMTsNCit9DQorI2VuZGlmIC8qIFdJVEhfQ1lDTEVfR0MgKi8NCisNCiBzdGF0aWMg
 > UHlPYmplY3QgKmdldGl0ZW1zdHIsICpzZXRpdGVtc3RyLCAqZGVsaXRlbXN0ciwgKmxlbnN0
 > cjsNCiANCiBzdGF0aWMgaW50DQpAQCAtMTQ2Myw3ICsxNTA4LDExIEBADQogCShnZXRhdHRy
 > b2Z1bmMpaW5zdGFuY2VfZ2V0YXR0ciwgLyp0cF9nZXRhdHRybyovDQogCShzZXRhdHRyb2Z1
 > bmMpaW5zdGFuY2Vfc2V0YXR0ciwgLyp0cF9zZXRhdHRybyovDQogICAgICAgICAwLCAvKiB0
 > cF9hc19idWZmZXIgKi8NCi0JUHlfVFBGTEFHU19ERUZBVUxULCAvKnRwX2ZsYWdzICovDQor
 > CVB5X1RQRkxBR1NfREVGQVVMVCB8IFB5X1RQRkxBR1NfSEFWRV9HQ0lORk8sIC8qdHBfZmxh
 > Z3MqLw0KKwkwLAkJLyogdHBfZG9jICovDQorI2lmZGVmIFdJVEhfQ1lDTEVfR0MNCisJKHJl
 > Y3Vyc2Vwcm9jKWluc3RhbmNlX3JlY3Vyc2UsCS8qIHRwX3JlY3Vyc2UgKi8NCisjZW5kaWYN
 > CiB9Ow0KIA0KIA0KQEAgLTE0OTMsNyArMTU0Miw3IEBADQogCQlfUHlfTmV3UmVmZXJlbmNl
 > KChQeU9iamVjdCAqKWltKTsNCiAJfQ0KIAllbHNlIHsNCi0JCWltID0gUHlPYmplY3RfTkVX
 > KFB5TWV0aG9kT2JqZWN0LCAmUHlNZXRob2RfVHlwZSk7DQorCQlpbSA9IFB5T2JqZWN0X05l
 > dyhQeU1ldGhvZE9iamVjdCwgJlB5TWV0aG9kX1R5cGUpOw0KIAkJaWYgKGltID09IE5VTEwp
 > DQogCQkJcmV0dXJuIE5VTEw7DQogCX0NCi0tLSBQeXRob24tY3ZzL09iamVjdHMvZGljdG9i
 > amVjdC5jCVRodSBNYXIgMzAgMjI6MDI6NTIgMjAwMA0KKysrIFB5dGhvbi1nYy9PYmplY3Rz
 > L2RpY3RvYmplY3QuYwlTYXQgQXByICA4IDAyOjA5OjU5IDIwMDANCkBAIC0xMjEsNyArMTIx
 > LDcgQEANCiAJCWlmIChkdW1teSA9PSBOVUxMKQ0KIAkJCXJldHVybiBOVUxMOw0KIAl9DQot
 > CW1wID0gUHlPYmplY3RfTkVXKGRpY3RvYmplY3QsICZQeURpY3RfVHlwZSk7DQorCW1wID0g
 > UHlPYmplY3RfTmV3KGRpY3RvYmplY3QsICZQeURpY3RfVHlwZSk7DQogCWlmIChtcCA9PSBO
 > VUxMKQ0KIAkJcmV0dXJuIE5VTEw7DQogCW1wLT5tYV9zaXplID0gMDsNCkBAIC0xMjksNiAr
 > MTI5LDkgQEANCiAJbXAtPm1hX3RhYmxlID0gTlVMTDsNCiAJbXAtPm1hX2ZpbGwgPSAwOw0K
 > IAltcC0+bWFfdXNlZCA9IDA7DQorI2lmZGVmIFdJVEhfQ1lDTEVfR0MNCisJUHlHQ19JbnNl
 > cnQoKFB5T2JqZWN0ICopbXApOw0KKyNlbmRpZg0KIAlyZXR1cm4gKFB5T2JqZWN0ICopbXA7
 > DQogfQ0KIA0KQEAgLTQ4MCw2ICs0ODMsOSBAQA0KIAlyZWdpc3RlciBpbnQgaTsNCiAJcmVn
 > aXN0ZXIgZGljdGVudHJ5ICplcDsNCiAJUHlfVFJBU0hDQU5fU0FGRV9CRUdJTihtcCkNCisj
 > aWZkZWYgV0lUSF9DWUNMRV9HQw0KKwlQeUdDX1JlbW92ZSgoUHlPYmplY3QgKiltcCk7DQor
 > I2VuZGlmDQogCWZvciAoaSA9IDAsIGVwID0gbXAtPm1hX3RhYmxlOyBpIDwgbXAtPm1hX3Np
 > emU7IGkrKywgZXArKykgew0KIAkJaWYgKGVwLT5tZV9rZXkgIT0gTlVMTCkgew0KIAkJCVB5
 > X0RFQ1JFRihlcC0+bWVfa2V5KTsNCkBAIC00ODksNyArNDk1LDcgQEANCiAJCX0NCiAJfQ0K
 > IAlQeU1lbV9YREVMKG1wLT5tYV90YWJsZSk7DQotCVB5TWVtX0RFTChtcCk7DQorCVB5T2Jq
 > ZWN0X0RlbCgoUHlPYmplY3QgKiltcCk7DQogCVB5X1RSQVNIQ0FOX1NBRkVfRU5EKG1wKQ0K
 > IH0NCiANCkBAIC0xMDM2LDYgKzEwNDIsMzMgQEANCiAJcmV0dXJuIFB5X05vbmU7DQogfQ0K
 > IA0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorc3RhdGljIGludA0KK2RpY3RfcmVjdXJzZShQ
 > eU9iamVjdCAqb3AsIHZpc2l0cHJvYyB2aXNpdCwgdm9pZCAqY2xvc3VyZSkNCit7DQorCWlu
 > dCBpID0gMDsNCisJUHlPYmplY3QgKnBrOw0KKwlQeU9iamVjdCAqcHY7DQorDQorCXdoaWxl
 > IChQeURpY3RfTmV4dChvcCwgJmksICZwaywgJnB2KSkgew0KKwkJaWYgKCF2aXNpdChwaywg
 > Y2xvc3VyZSkpIHsNCisJCQlyZXR1cm4gMDsNCisJCX0NCisJCWlmICghdmlzaXQocHYsIGNs
 > b3N1cmUpKSB7DQorCQkJcmV0dXJuIDA7DQorCQl9DQorCX0NCisJcmV0dXJuIDE7DQorfQ0K
 > Kw0KK3N0YXRpYyBpbnQNCitkaWN0X2djX2NsZWFyKFB5T2JqZWN0ICpvcCkNCit7DQorCVB5
 > RGljdF9DbGVhcihvcCk7DQorCXJldHVybiAwOw0KK30NCisjZW5kaWYgLyogV0lUSF9DWUNM
 > RV9HQyAqLw0KKw0KIHN0YXRpYyBQeU1ldGhvZERlZiBtYXBwX21ldGhvZHNbXSA9IHsNCiAJ
 > eyJoYXNfa2V5IiwJKFB5Q0Z1bmN0aW9uKWRpY3RfaGFzX2tleSwgICAgICBNRVRIX1ZBUkFS
 > R1N9LA0KIAl7ImtleXMiLAkoUHlDRnVuY3Rpb24pZGljdF9rZXlzfSwNCkBAIC0xMDcxLDYg
 > KzExMDQsMTggQEANCiAJMCwJCQkvKnRwX2FzX251bWJlciovDQogCTAsCQkJLyp0cF9hc19z
 > ZXF1ZW5jZSovDQogCSZkaWN0X2FzX21hcHBpbmcsCS8qdHBfYXNfbWFwcGluZyovDQorI2lm
 > ZGVmIFdJVEhfQ1lDTEVfR0MNCisJMCwJCS8qIHRwX2hhc2ggKi8NCisJMCwJCS8qIHRwX2Nh
 > bGwgKi8NCisJMCwJCS8qIHRwX3N0ciAqLw0KKwkwLAkJLyogdHBfZ2V0YXR0cm8gKi8NCisJ
 > MCwJCS8qIHRwX3NldGF0dHJvICovDQorCTAsCQkvKiB0cF9hc19idWZmZXIgKi8NCisJUHlf
 > VFBGTEFHU19ERUZBVUxUIHwgUHlfVFBGTEFHU19IQVZFX0dDSU5GTywgLyp0cF9mbGFncyov
 > DQorCTAsCQkvKiB0cF9kb2MgKi8NCisJKHJlY3Vyc2Vwcm9jKWRpY3RfcmVjdXJzZSwJLyog
 > dHBfcmVjdXJzZSAqLw0KKwkoaW5xdWlyeSlkaWN0X2djX2NsZWFyLAkJLyogdHBfY2xlYXIg
 > Ki8NCisjZW5kaWYNCiB9Ow0KIA0KIC8qIEZvciBiYWNrd2FyZCBjb21wYXRpYmlsaXR5IHdp
 > dGggb2xkIGRpY3Rpb25hcnkgaW50ZXJmYWNlICovDQotLS0gUHl0aG9uLWN2cy9PYmplY3Rz
 > L2Z1bmNvYmplY3QuYwlUaHUgTWF5IDIxIDE4OjU1OjM0IDE5OTgNCisrKyBQeXRob24tZ2Mv
 > T2JqZWN0cy9mdW5jb2JqZWN0LmMJU2F0IEFwciAgOCAwMjowOTo1MCAyMDAwDQpAQCAtNDAs
 > NyArNDAsNyBAQA0KIAlQeU9iamVjdCAqY29kZTsNCiAJUHlPYmplY3QgKmdsb2JhbHM7DQog
 > ew0KLQlQeUZ1bmN0aW9uT2JqZWN0ICpvcCA9IFB5T2JqZWN0X05FVyhQeUZ1bmN0aW9uT2Jq
 > ZWN0LA0KKwlQeUZ1bmN0aW9uT2JqZWN0ICpvcCA9IFB5T2JqZWN0X05ldyhQeUZ1bmN0aW9u
 > T2JqZWN0LA0KIAkJCQkJICAgICZQeUZ1bmN0aW9uX1R5cGUpOw0KIAlpZiAob3AgIT0gTlVM
 > TCkgew0KIAkJUHlPYmplY3QgKmRvYzsNCkBAIC02Myw2ICs2Myw5IEBADQogCQlQeV9JTkNS
 > RUYoZG9jKTsNCiAJCW9wLT5mdW5jX2RvYyA9IGRvYzsNCiAJfQ0KKyNpZmRlZiBXSVRIX0NZ
 > Q0xFX0dDDQorCVB5R0NfSW5zZXJ0KChQeU9iamVjdCAqKW9wKTsNCisjZW5kaWYNCiAJcmV0
 > dXJuIChQeU9iamVjdCAqKW9wOw0KIH0NCiANCkBAIC0xODYsMTIgKzE4OSwxNSBAQA0KIGZ1
 > bmNfZGVhbGxvYyhvcCkNCiAJUHlGdW5jdGlvbk9iamVjdCAqb3A7DQogew0KKyNpZmRlZiBX
 > SVRIX0NZQ0xFX0dDDQorCVB5R0NfUmVtb3ZlKChQeU9iamVjdCAqKW9wKTsNCisjZW5kaWYN
 > CiAJUHlfREVDUkVGKG9wLT5mdW5jX2NvZGUpOw0KIAlQeV9ERUNSRUYob3AtPmZ1bmNfZ2xv
 > YmFscyk7DQogCVB5X0RFQ1JFRihvcC0+ZnVuY19uYW1lKTsNCiAJUHlfWERFQ1JFRihvcC0+
 > ZnVuY19kZWZhdWx0cyk7DQogCVB5X1hERUNSRUYob3AtPmZ1bmNfZG9jKTsNCi0JUHlNZW1f
 > REVMKG9wKTsNCisJUHlPYmplY3RfRGVsKChQeU9iamVjdCAqKW9wKTsNCiB9DQogDQogc3Rh
 > dGljIFB5T2JqZWN0Kg0KQEAgLTIzOSw2ICsyNDUsMTkgQEANCiAJcmV0dXJuIGg7DQogfQ0K
 > IA0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorc3RhdGljIGludA0KK2Z1bmNfcmVjdXJzZShQ
 > eUZ1bmN0aW9uT2JqZWN0ICpmLCB2aXNpdHByb2MgdmlzaXQsIHZvaWQgKmNsb3N1cmUpDQor
 > ew0KKwlpZiAoZi0+ZnVuY19jb2RlKSB2aXNpdChmLT5mdW5jX2NvZGUsIGNsb3N1cmUpOw0K
 > KwlpZiAoZi0+ZnVuY19nbG9iYWxzKSB2aXNpdChmLT5mdW5jX2dsb2JhbHMsIGNsb3N1cmUp
 > Ow0KKwlpZiAoZi0+ZnVuY19kZWZhdWx0cykgdmlzaXQoZi0+ZnVuY19kZWZhdWx0cywgY2xv
 > c3VyZSk7DQorCWlmIChmLT5mdW5jX2RvYykgdmlzaXQoZi0+ZnVuY19kb2MsIGNsb3N1cmUp
 > Ow0KKwlpZiAoZi0+ZnVuY19uYW1lKSB2aXNpdChmLT5mdW5jX25hbWUsIGNsb3N1cmUpOw0K
 > KwlyZXR1cm4gMTsNCit9DQorI2VuZGlmIC8qIFdJVEhfQ1lDTEVfR0MgKi8NCisNCiBQeVR5
 > cGVPYmplY3QgUHlGdW5jdGlvbl9UeXBlID0gew0KIAlQeU9iamVjdF9IRUFEX0lOSVQoJlB5
 > VHlwZV9UeXBlKQ0KIAkwLA0KQEAgLTI1NSw0ICsyNzQsMTQgQEANCiAJMCwJCS8qdHBfYXNf
 > c2VxdWVuY2UqLw0KIAkwLAkJLyp0cF9hc19tYXBwaW5nKi8NCiAJKGhhc2hmdW5jKWZ1bmNf
 > aGFzaCwgLyp0cF9oYXNoKi8NCisjaWZkZWYgV0lUSF9DWUNMRV9HQw0KKwkwLAkJLyp0cF9j
 > YWxsKi8NCisJMCwJCS8qdHBfc3RyKi8NCisJMCwJCS8qdHBfZ2V0YXR0cm8qLw0KKwkwLAkJ
 > Lyp0cF9zZXRhdHRybyovDQorCTAsCQkvKiB0cF9hc19idWZmZXIgKi8NCisJUHlfVFBGTEFH
 > U19ERUZBVUxUIHwgUHlfVFBGTEFHU19IQVZFX0dDSU5GTywgLyp0cF9mbGFncyovDQorCTAs
 > CQkvKiB0cF9kb2MgKi8NCisJKHJlY3Vyc2Vwcm9jKWZ1bmNfcmVjdXJzZSwJLyogdHBfcmVj
 > dXJzZSAqLw0KKyNlbmRpZg0KIH07DQotLS0gUHl0aG9uLWN2cy9PYmplY3RzL2xpc3RvYmpl
 > Y3QuYwlGcmkgTWFyIDI0IDIyOjMyOjMwIDIwMDANCisrKyBQeXRob24tZ2MvT2JqZWN0cy9s
 > aXN0b2JqZWN0LmMJTW9uIEFwciAyNCAyMToxNDo0MiAyMDAwDQpAQCAtNjEsMzQgKzYxLDM3
 > IEBADQogCWludCBpOw0KIAlQeUxpc3RPYmplY3QgKm9wOw0KIAlzaXplX3QgbmJ5dGVzOw0K
 > KwlQeU9iamVjdCAqKml0ZW07DQorDQogCWlmIChzaXplIDwgMCkgew0KIAkJUHlFcnJfQmFk
 > SW50ZXJuYWxDYWxsKCk7DQogCQlyZXR1cm4gTlVMTDsNCiAJfQ0KLQluYnl0ZXMgPSBzaXpl
 > ICogc2l6ZW9mKFB5T2JqZWN0ICopOw0KLQkvKiBDaGVjayBmb3Igb3ZlcmZsb3cgKi8NCi0J
 > aWYgKG5ieXRlcyAvIHNpemVvZihQeU9iamVjdCAqKSAhPSAoc2l6ZV90KXNpemUpIHsNCi0J
 > CXJldHVybiBQeUVycl9Ob01lbW9yeSgpOw0KLQl9DQotCW9wID0gKFB5TGlzdE9iamVjdCAq
 > KSBtYWxsb2Moc2l6ZW9mKFB5TGlzdE9iamVjdCkpOw0KLQlpZiAob3AgPT0gTlVMTCkgew0K
 > LQkJcmV0dXJuIFB5RXJyX05vTWVtb3J5KCk7DQotCX0NCi0JaWYgKHNpemUgPD0gMCkgew0K
 > LQkJb3AtPm9iX2l0ZW0gPSBOVUxMOw0KLQl9DQorCWlmIChzaXplID09IDApDQorCQlpdGVt
 > ID0gTlVMTDsNCiAJZWxzZSB7DQotCQlvcC0+b2JfaXRlbSA9IChQeU9iamVjdCAqKikgbWFs
 > bG9jKG5ieXRlcyk7DQotCQlpZiAob3AtPm9iX2l0ZW0gPT0gTlVMTCkgew0KLQkJCWZyZWUo
 > KEFOWSAqKW9wKTsNCisJCW5ieXRlcyA9IHNpemUgKiBzaXplb2YoUHlPYmplY3QgKik7DQor
 > CQkvKiBDaGVjayBmb3Igb3ZlcmZsb3cgKi8NCisJCWlmIChuYnl0ZXMgLyBzaXplb2YoUHlP
 > YmplY3QgKikgIT0gKHNpemVfdClzaXplKSB7DQogCQkJcmV0dXJuIFB5RXJyX05vTWVtb3J5
 > KCk7DQogCQl9DQorCQlpdGVtID0gKFB5T2JqZWN0ICoqKSBtYWxsb2MobmJ5dGVzKTsNCisJ
 > CWlmIChpdGVtID09IE5VTEwpDQorCQkJcmV0dXJuIFB5RXJyX05vTWVtb3J5KCk7DQorCX0N
 > CisJb3AgPSBQeU9iamVjdF9OZXcoUHlMaXN0T2JqZWN0LCAmUHlMaXN0X1R5cGUpOw0KKwlp
 > ZiAob3AgPT0gTlVMTCkgew0KKwkJaWYgKGl0ZW0gIT0gTlVMTCkNCisJCQlmcmVlKChBTlkg
 > KilpdGVtKTsNCisJCXJldHVybiBQeUVycl9Ob01lbW9yeSgpOw0KIAl9DQotCW9wLT5vYl90
 > eXBlID0gJlB5TGlzdF9UeXBlOw0KIAlvcC0+b2Jfc2l6ZSA9IHNpemU7DQorCW9wLT5vYl9p
 > dGVtID0gaXRlbTsNCiAJZm9yIChpID0gMDsgaSA8IHNpemU7IGkrKykNCiAJCW9wLT5vYl9p
 > dGVtW2ldID0gTlVMTDsNCi0JX1B5X05ld1JlZmVyZW5jZSgoUHlPYmplY3QgKilvcCk7DQor
 > I2lmZGVmIFdJVEhfQ1lDTEVfR0MNCisJUHlHQ19JbnNlcnQoKFB5T2JqZWN0ICopb3ApOw0K
 > KyNlbmRpZg0KIAlyZXR1cm4gKFB5T2JqZWN0ICopIG9wOw0KIH0NCiANCkBAIC0yMTYsNiAr
 > MjE5LDkgQEANCiB7DQogCWludCBpOw0KIAlQeV9UUkFTSENBTl9TQUZFX0JFR0lOKG9wKQ0K
 > KyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorCVB5R0NfUmVtb3ZlKChQeU9iamVjdCAqKW9wKTsN
 > CisjZW5kaWYNCiAJaWYgKG9wLT5vYl9pdGVtICE9IE5VTEwpIHsNCiAJCS8qIERvIGl0IGJh
 > Y2t3YXJkcywgZm9yIENocmlzdGlhbiBUaXNtZXIuDQogCQkgICBUaGVyZSdzIGEgc2ltcGxl
 > IHRlc3QgY2FzZSB3aGVyZSBzb21laG93IHRoaXMgcmVkdWNlcw0KQEAgLTIyNyw3ICsyMzMs
 > NyBAQA0KIAkJfQ0KIAkJZnJlZSgoQU5ZICopb3AtPm9iX2l0ZW0pOw0KIAl9DQotCWZyZWUo
 > KEFOWSAqKW9wKTsNCisJUHlPYmplY3RfRGVsKChQeU9iamVjdCAqKW9wKTsNCiAJUHlfVFJB
 > U0hDQU5fU0FGRV9FTkQob3ApDQogfQ0KIA0KQEAgLTEzOTksNiArMTQwNSwyOSBAQA0KIAly
 > ZXR1cm4gTlVMTDsNCiB9DQogDQorI2lmZGVmIFdJVEhfQ1lDTEVfR0MNCitzdGF0aWMgaW50
 > DQorbGlzdF9yZWN1cnNlKFB5TGlzdE9iamVjdCAqbywgdmlzaXRwcm9jIHZpc2l0LCB2b2lk
 > ICpjbG9zdXJlKQ0KK3sNCisJaW50IGk7DQorCVB5T2JqZWN0ICp4Ow0KKw0KKwlmb3IgKGkg
 > PSBvLT5vYl9zaXplOyAtLWkgPj0gMDsgKSB7DQorCQl4ID0gby0+b2JfaXRlbVtpXTsNCisJ
 > CWlmICh4ICE9IE5VTEwgJiYgIXZpc2l0KHgsIGNsb3N1cmUpKQ0KKwkJCXJldHVybiAwOw0K
 > Kwl9DQorCXJldHVybiAxOw0KK30NCisNCitzdGF0aWMgaW50DQorbGlzdF9jbGVhcihQeUxp
 > c3RPYmplY3QgKmxwKQ0KK3sNCisJKHZvaWQpIFB5TGlzdF9TZXRTbGljZSgoUHlPYmplY3Qg
 > KilscCwgMCwgbHAtPm9iX3NpemUsIDApOw0KKwlyZXR1cm4gMDsNCit9DQorI2VuZGlmIC8q
 > IFdJVEhfQ1lDTEVfR0MgKi8NCisNCiBzdGF0aWMgY2hhciBhcHBlbmRfZG9jW10gPQ0KICJM
 > LmFwcGVuZChvYmplY3QpIC0tIGFwcGVuZCBvYmplY3QgdG8gZW5kIjsNCiBzdGF0aWMgY2hh
 > ciBleHRlbmRfZG9jW10gPQ0KQEAgLTE0NjQsNiArMTQ5MywxOCBAQA0KIAkwLAkJLyp0cF9h
 > c19udW1iZXIqLw0KIAkmbGlzdF9hc19zZXF1ZW5jZSwJLyp0cF9hc19zZXF1ZW5jZSovDQog
 > CTAsCQkvKnRwX2FzX21hcHBpbmcqLw0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorCTAsCQkv
 > KiB0cF9oYXNoICovDQorCTAsCQkvKiB0cF9jYWxsICovDQorCTAsCQkvKiB0cF9zdHIgKi8N
 > CisJMCwJCS8qIHRwX2dldGF0dHJvICovDQorCTAsCQkvKiB0cF9zZXRhdHRybyAqLw0KKwkw
 > LAkJLyogdHBfYXNfYnVmZmVyICovDQorCVB5X1RQRkxBR1NfREVGQVVMVCB8IFB5X1RQRkxB
 > R1NfSEFWRV9HQ0lORk8sCS8qIHRwX2ZsYWdzICovDQorCTAsCQkvKiB0cF9kb2MgKi8NCisJ
 > KHJlY3Vyc2Vwcm9jKWxpc3RfcmVjdXJzZSwJLyogdHBfcmVjdXJzZSAqLw0KKwkoaW5xdWly
 > eSlsaXN0X2NsZWFyLAkvKiB0cF9jbGVhciAqLw0KKyNlbmRpZg0KIH07DQogDQogDQpAQCAt
 > MTUzMiw1ICsxNTczLDE2IEBADQogCTAsCQkvKnRwX2FzX251bWJlciovDQogCSZpbW11dGFi
 > bGVfbGlzdF9hc19zZXF1ZW5jZSwJLyp0cF9hc19zZXF1ZW5jZSovDQogCTAsCQkvKnRwX2Fz
 > X21hcHBpbmcqLw0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorCTAsCQkvKiB0cF9oYXNoICov
 > DQorCTAsCQkvKiB0cF9jYWxsICovDQorCTAsCQkvKiB0cF9zdHIgKi8NCisJMCwJCS8qIHRw
 > X2dldGF0dHJvICovDQorCTAsCQkvKiB0cF9zZXRhdHRybyAqLw0KKwkwLAkJLyogdHBfYXNf
 > YnVmZmVyICovDQorCVB5X1RQRkxBR1NfREVGQVVMVCwJLyogdHBfZmxhZ3MgKi8NCisJMCwJ
 > CS8qIHRwX2RvYyAqLw0KKwkocmVjdXJzZXByb2MpbGlzdF9yZWN1cnNlLAkvKiB0cF9yZWN1
 > cnNlICovDQorI2VuZGlmDQogfTsNCiANCi0tLSBQeXRob24tY3ZzL09iamVjdHMvb2JqZWN0
 > LmMJVHVlIE1hciAyOCAwODoxMTowOSAyMDAwDQorKysgUHl0aG9uLWdjL09iamVjdHMvb2Jq
 > ZWN0LmMJU2F0IEFwciAgOCAwMjoxMDowNiAyMDAwDQpAQCAtMTA3LDIwICsxMDcsMTIgQEAN
 > CiB9DQogI2VuZGlmDQogDQotI2lmbmRlZiBNU19DT1JFRExMDQotUHlPYmplY3QgKg0KLV9Q
 > eU9iamVjdF9OZXcodHApDQotCVB5VHlwZU9iamVjdCAqdHA7DQotI2Vsc2UNCisNCiBQeU9i
 > amVjdCAqDQotX1B5T2JqZWN0X05ldyh0cCxvcCkNCitfUHlPYmplY3RfRnJvbVR5cGUodHAs
 > b3ApDQogCVB5VHlwZU9iamVjdCAqdHA7DQogCVB5T2JqZWN0ICpvcDsNCi0jZW5kaWYNCiB7
 > DQotI2lmbmRlZiBNU19DT1JFRExMDQotCVB5T2JqZWN0ICpvcCA9IChQeU9iamVjdCAqKSBt
 > YWxsb2ModHAtPnRwX2Jhc2ljc2l6ZSk7DQotI2VuZGlmDQogCWlmIChvcCA9PSBOVUxMKQ0K
 > IAkJcmV0dXJuIFB5RXJyX05vTWVtb3J5KCk7DQogCW9wLT5vYl90eXBlID0gdHA7DQpAQCAt
 > MTI4LDI5ICsxMjAsNzggQEANCiAJcmV0dXJuIG9wOw0KIH0NCiANCi0jaWZuZGVmIE1TX0NP
 > UkVETEwNCi1QeVZhck9iamVjdCAqDQotX1B5T2JqZWN0X05ld1Zhcih0cCwgc2l6ZSkNCi0J
 > UHlUeXBlT2JqZWN0ICp0cDsNCi0JaW50IHNpemU7DQotI2Vsc2UNCiBQeVZhck9iamVjdCAq
 > DQotX1B5T2JqZWN0X05ld1Zhcih0cCwgc2l6ZSwgb3ApDQorX1B5T2JqZWN0X1ZhckZyb21U
 > eXBlKHRwLCBzaXplLCBvcCkNCiAJUHlUeXBlT2JqZWN0ICp0cDsNCiAJaW50IHNpemU7DQog
 > CVB5VmFyT2JqZWN0ICpvcDsNCi0jZW5kaWYNCiB7DQotI2lmbmRlZiBNU19DT1JFRExMDQot
 > CVB5VmFyT2JqZWN0ICpvcCA9IChQeVZhck9iamVjdCAqKQ0KLQkJbWFsbG9jKHRwLT50cF9i
 > YXNpY3NpemUgKyBzaXplICogdHAtPnRwX2l0ZW1zaXplKTsNCi0jZW5kaWYNCiAJaWYgKG9w
 > ID09IE5VTEwpDQogCQlyZXR1cm4gKFB5VmFyT2JqZWN0ICopUHlFcnJfTm9NZW1vcnkoKTsN
 > CiAJb3AtPm9iX3R5cGUgPSB0cDsNCiAJb3AtPm9iX3NpemUgPSBzaXplOw0KIAlfUHlfTmV3
 > UmVmZXJlbmNlKChQeU9iamVjdCAqKW9wKTsNCiAJcmV0dXJuIG9wOw0KK30NCisNCitQeU9i
 > amVjdCAqDQorX1B5T2JqZWN0X05ldyh0cCkNCisJUHlUeXBlT2JqZWN0ICp0cDsNCit7DQor
 > CVB5T2JqZWN0ICpvcDsNCisjaWZkZWYgV0lUSF9DWUNMRV9HQw0KKwlpZiAoUFlfVFlQRUlT
 > R0ModHApKSB7DQorCQlQeUdDSW5mbyAqZzsNCisJCWcgPSAoUHlHQ0luZm8gKikgbWFsbG9j
 > KHNpemVvZigqZykgKyB0cC0+dHBfYmFzaWNzaXplKTsNCisJCWlmIChnID09IE5VTEwpIHsN
 > CisJCQlvcCA9IE5VTEw7DQorCQl9IGVsc2Ugew0KKwkJCW9wID0gUFlfR0NPQkooZyk7DQor
 > CQl9DQorCX0gZWxzZQ0KKyNlbmRpZiAvKiBXSVRIX0NZQ0xFX0dDICovDQorCXsNCisJCW9w
 > ID0gKFB5T2JqZWN0ICopIG1hbGxvYyh0cC0+dHBfYmFzaWNzaXplKTsNCisJfQ0KKwlyZXR1
 > cm4gX1B5T2JqZWN0X0Zyb21UeXBlKHRwLCBvcCk7DQorfQ0KKw0KK1B5VmFyT2JqZWN0ICoN
 > CitfUHlPYmplY3RfTmV3VmFyKHRwLCBzaXplKQ0KKwlQeVR5cGVPYmplY3QgKnRwOw0KKwlp
 > bnQgc2l6ZTsNCit7DQorCVB5VmFyT2JqZWN0ICpvcDsNCisjaWZkZWYgV0lUSF9DWUNMRV9H
 > Qw0KKwlpZiAoUFlfVFlQRUlTR0ModHApKSB7DQorCQlQeUdDSW5mbyAqZzsNCisJCWcgPSAo
 > UHlHQ0luZm8gKikgbWFsbG9jKHNpemVvZigqZykgKyANCisJCQkJCXRwLT50cF9iYXNpY3Np
 > emUgKw0KKwkJCQkJc2l6ZSAqIHRwLT50cF9pdGVtc2l6ZSk7DQorCQlpZiAoZyA9PSBOVUxM
 > KSB7DQorCQkJb3AgPSBOVUxMOw0KKwkJfSBlbHNlIHsNCisJCQlvcCA9IChQeVZhck9iamVj
 > dCAqKVBZX0dDT0JKKGcpOw0KKwkJfQ0KKwl9IGVsc2UNCisjZW5kaWYgLyogV0lUSF9DWUNM
 > RV9HQyAqLw0KKwl7DQorCQlvcCA9IChQeVZhck9iamVjdCAqKQ0KKwkJCW1hbGxvYyh0cC0+
 > dHBfYmFzaWNzaXplICsgc2l6ZSAqIHRwLT50cF9pdGVtc2l6ZSk7DQorCX0NCisJcmV0dXJu
 > IF9QeU9iamVjdF9WYXJGcm9tVHlwZSh0cCwgc2l6ZSwgb3ApOw0KK30NCisNCit2b2lkDQor
 > UHlPYmplY3RfRGVsKG9wKQ0KKwlQeU9iamVjdCAqb3A7DQorew0KKyNpZmRlZiBXSVRIX0NZ
 > Q0xFX0dDDQorCWlmIChQWV9JU0dDKG9wKSkgew0KKwkJb3AgPSAoUHlPYmplY3QgKilQWV9H
 > Q0lORk9fVU5TQUZFKG9wKTsNCisJfQ0KKyNlbmRpZg0KKwlmcmVlKG9wKTsNCiB9DQogDQog
 > aW50DQotLS0gUHl0aG9uLWN2cy9PYmplY3RzL3R1cGxlb2JqZWN0LmMJRnJpIE1hciAyNCAy
 > MjozMjozMCAyMDAwDQorKysgUHl0aG9uLWdjL09iamVjdHMvdHVwbGVvYmplY3QuYwlNb24g
 > QXByIDI0IDIxOjIxOjU5IDIwMDANCkBAIC04Miw4ICs4Miw4IEBADQogI2VuZGlmDQogI2lm
 > ZGVmIFB5X1RSQUNFX1JFRlMNCiAJCW9wLT5vYl90eXBlID0gJlB5VHVwbGVfVHlwZTsNCi0J
 > CW9wLT5vYl9zaXplID0gc2l6ZTsNCiAjZW5kaWYNCisJCV9QeV9OZXdSZWZlcmVuY2UoKFB5
 > T2JqZWN0ICopb3ApOw0KIAl9DQogCWVsc2UNCiAjZW5kaWYNCkBAIC05NywyMyArOTcsMjMg
 > QEANCiAJCQlyZXR1cm4gUHlFcnJfTm9NZW1vcnkoKTsNCiAJCX0NCiAJCTsNCi0JCW9wID0g
 > KFB5VHVwbGVPYmplY3QgKikgbWFsbG9jKG5ieXRlcyk7DQorCQlvcCA9IFB5T2JqZWN0X05l
 > d1ZhcihQeVR1cGxlT2JqZWN0LCAmUHlUdXBsZV9UeXBlLCBzaXplKTsNCiAJCWlmIChvcCA9
 > PSBOVUxMKQ0KIAkJCXJldHVybiBQeUVycl9Ob01lbW9yeSgpOw0KIA0KLQkJb3AtPm9iX3R5
 > cGUgPSAmUHlUdXBsZV9UeXBlOw0KLQkJb3AtPm9iX3NpemUgPSBzaXplOw0KIAl9DQogCWZv
 > ciAoaSA9IDA7IGkgPCBzaXplOyBpKyspDQogCQlvcC0+b2JfaXRlbVtpXSA9IE5VTEw7DQot
 > CV9QeV9OZXdSZWZlcmVuY2UoKFB5T2JqZWN0ICopb3ApOw0KKw0KICNpZiBNQVhTQVZFU0la
 > RSA+IDANCiAJaWYgKHNpemUgPT0gMCkgew0KIAkJZnJlZV90dXBsZXNbMF0gPSBvcDsNCi0J
 > CSsrbnVtX2ZyZWVfdHVwbGVzWzBdOw0KIAkJUHlfSU5DUkVGKG9wKTsJLyogZXh0cmEgSU5D
 > UkVGIHNvIHRoYXQgdGhpcyBpcyBuZXZlciBmcmVlZCAqLw0KIAl9DQogI2VuZGlmDQorI2lm
 > ZGVmIFdJVEhfQ1lDTEVfR0MNCisJUHlHQ19JbnNlcnQoKFB5T2JqZWN0ICopb3ApOw0KKyNl
 > bmRpZg0KIAlyZXR1cm4gKFB5T2JqZWN0ICopIG9wOw0KIH0NCiANCkBAIC0xODAsNiArMTgw
 > LDkgQEANCiAJcmVnaXN0ZXIgaW50IGk7DQogCXJlZ2lzdGVyIGludCBsZW4gPSAgb3AtPm9i
 > X3NpemU7DQogCVB5X1RSQVNIQ0FOX1NBRkVfQkVHSU4ob3ApDQorI2lmZGVmIFdJVEhfQ1lD
 > TEVfR0MNCisJUHlHQ19SZW1vdmUoKFB5T2JqZWN0ICopb3ApOw0KKyNlbmRpZg0KIAlpZiAo
 > bGVuID4gMCkgew0KIAkJaSA9IGxlbjsNCiAJCXdoaWxlICgtLWkgPj0gMCkNCkBAIC0xOTMs
 > NyArMTk2LDcgQEANCiAJCX0NCiAjZW5kaWYNCiAJfQ0KLQlmcmVlKChBTlkgKilvcCk7DQor
 > CVB5T2JqZWN0X0RlbCgoUHlPYmplY3QgKilvcCk7DQogZG9uZToNCiAJUHlfVFJBU0hDQU5f
 > U0FGRV9FTkQob3ApDQogfQ0KQEAgLTQwMSw2ICs0MDQsMjIgQEANCiAJcmV0dXJuIChQeU9i
 > amVjdCAqKSBucDsNCiB9DQogDQorI2lmZGVmIFdJVEhfQ1lDTEVfR0MNCitzdGF0aWMgaW50
 > DQordHVwbGVyZWN1cnNlKFB5VHVwbGVPYmplY3QgKm8sIHZpc2l0cHJvYyB2aXNpdCwgdm9p
 > ZCAqY2xvc3VyZSkNCit7DQorCWludCBpOw0KKwlQeU9iamVjdCAqeDsNCisNCisJZm9yIChp
 > ID0gby0+b2Jfc2l6ZTsgLS1pID49IDA7ICkgew0KKwkJeCA9IG8tPm9iX2l0ZW1baV07DQor
 > CQlpZiAoeCAhPSBOVUxMICYmICF2aXNpdCh4LCBjbG9zdXJlKSkNCisJCQlyZXR1cm4gMDsN
 > CisJfQ0KKwlyZXR1cm4gMTsNCit9DQorI2VuZGlmIC8qIFdJVEhfQ1lDTEVfR0MgKi8NCisN
 > CiBzdGF0aWMgUHlTZXF1ZW5jZU1ldGhvZHMgdHVwbGVfYXNfc2VxdWVuY2UgPSB7DQogCShp
 > bnF1aXJ5KXR1cGxlbGVuZ3RoLCAvKnNxX2xlbmd0aCovDQogCShiaW5hcnlmdW5jKXR1cGxl
 > Y29uY2F0LCAvKnNxX2NvbmNhdCovDQpAQCAtNDI3LDYgKzQ0NiwxNiBAQA0KIAkmdHVwbGVf
 > YXNfc2VxdWVuY2UsCS8qdHBfYXNfc2VxdWVuY2UqLw0KIAkwLAkJLyp0cF9hc19tYXBwaW5n
 > Ki8NCiAJKGhhc2hmdW5jKXR1cGxlaGFzaCwgLyp0cF9oYXNoKi8NCisjaWZkZWYgV0lUSF9D
 > WUNMRV9HQw0KKwkwLAkJLyogdHBfY2FsbCAqLw0KKwkwLAkJLyogdHBfc3RyICovDQorCTAs
 > CQkvKiB0cF9nZXRhdHRybyAqLw0KKwkwLAkJLyogdHBfc2V0YXR0cm8gKi8NCisJMCwJCS8q
 > IHRwX2FzX2J1ZmZlciAqLw0KKwlQeV9UUEZMQUdTX0RFRkFVTFQgfCBQeV9UUEZMQUdTX0hB
 > VkVfR0NJTkZPLAkvKiB0cF9mbGFncyAqLw0KKwkwLAkJLyogdHBfZG9jICovDQorCShyZWN1
 > cnNlcHJvYyl0dXBsZXJlY3Vyc2UsCS8qIHRwX3JlY3Vyc2UgKi8NCisjZW5kaWYNCiB9Ow0K
 > IA0KIC8qIFRoZSBmb2xsb3dpbmcgZnVuY3Rpb24gYnJlYWtzIHRoZSBub3Rpb24gdGhhdCB0
 > dXBsZXMgYXJlIGltbXV0YWJsZToNCkBAIC00NDksNiArNDc4LDkgQEANCiAJaW50IGk7DQog
 > CWludCBzaXplZGlmZjsNCiANCisjaWZkZWYgV0lUSF9DWUNMRV9HQw0KKwlQeUdDSW5mbyAq
 > ZyA9IFBZX0dDSU5GTygqcHYpOyANCisjZW5kaWYNCiAJdiA9IChQeVR1cGxlT2JqZWN0ICop
 > ICpwdjsNCiAJaWYgKHYgPT0gTlVMTCB8fCAhUHlUdXBsZV9DaGVjayh2KSB8fCB2LT5vYl9y
 > ZWZjbnQgIT0gMSkgew0KIAkJKnB2ID0gMDsNCkBAIC00NzksNyArNTExLDYgQEANCiAJfQ0K
 > ICNpZiBNQVhTQVZFU0laRSA+IDANCiAJaWYgKG5ld3NpemUgPT0gMCAmJiBmcmVlX3R1cGxl
 > c1swXSkgew0KLQkJbnVtX2ZyZWVfdHVwbGVzWzBdLS07DQogCQlzdiA9IGZyZWVfdHVwbGVz
 > WzBdOw0KIAkJc3YtPm9iX3NpemUgPSAwOw0KIAkJUHlfSU5DUkVGKHN2KTsNCkBAIC01MTEs
 > OSArNTQyLDIyIEBADQogCX0gZWxzZSANCiAjZW5kaWYJCQ0KIAl7DQorI2lmZGVmIFdJVEhf
 > Q1lDTEVfR0MNCisJCVB5R0NfUmVtb3ZlKChQeU9iamVjdCAqKXYpOw0KKw0KKwkJZyA9IChQ
 > eUdDSW5mbyAqKSByZWFsbG9jKChjaGFyICopZywgc2l6ZW9mKCpnKSArIA0KKwkJCQlzaXpl
 > b2YoUHlUdXBsZU9iamVjdCkgKyANCisJCQkJbmV3c2l6ZSAqIHNpemVvZihQeU9iamVjdCAq
 > KSk7DQorCQlpZiAoZyA9PSBOVUxMKSB7DQorCQkJc3YgPT0gTlVMTDsNCisJCX0gZWxzZSB7
 > DQorCQkJc3YgPSAoUHlUdXBsZU9iamVjdCAqKSBQWV9HQ09CSihnKTsNCisJCX0NCisjZWxz
 > ZQ0KIAkJc3YgPSAoUHlUdXBsZU9iamVjdCAqKQ0KIAkJCXJlYWxsb2MoKGNoYXIgKil2LA0K
 > IAkJCQlzaXplb2YoUHlUdXBsZU9iamVjdCkgKyBuZXdzaXplICogc2l6ZW9mKFB5T2JqZWN0
 > ICopKTsNCisjZW5kaWYNCiAJCSpwdiA9IChQeU9iamVjdCAqKSBzdjsNCiAJCWlmIChzdiA9
 > PSBOVUxMKSB7DQogCQkJUHlNZW1fREVMKHYpOw0KQEAgLTUyMiw2ICs1NjYsOSBAQA0KIAkJ
 > fQ0KIAl9DQogCV9QeV9OZXdSZWZlcmVuY2UoKFB5T2JqZWN0ICopc3YpOw0KKyNpZmRlZiBX
 > SVRIX0NZQ0xFX0dDDQorCVB5R0NfSW5zZXJ0KChQeU9iamVjdCAqKXN2KTsNCisjZW5kaWYN
 > CiAJZm9yIChpID0gc3YtPm9iX3NpemU7IGkgPCBuZXdzaXplOyBpKyspDQogCQlzdi0+b2Jf
 > aXRlbVtpXSA9IE5VTEw7DQogCWlmIChsYXN0X2lzX3N0aWNreSAmJiBzaXplZGlmZiA+IDAp
 > IHsNCkBAIC01NTEsNyArNTk4LDExIEBADQogCQl3aGlsZSAocCkgew0KIAkJCXEgPSBwOw0K
 > IAkJCXAgPSAoUHlUdXBsZU9iamVjdCAqKShwLT5vYl9pdGVtWzBdKTsNCi0JCQlQeU1lbV9E
 > RUwocSk7DQorI2lmZGVmIFB5X1RSQUNFX1JFRlMNCisJCQlxLT5vYl90eXBlID0gICZQeVR1
 > cGxlX1R5cGU7DQorI2VuZGlmDQorCQkJUHlPYmplY3RfRGVsKChQeU9iamVjdCAqKXEpOw0K
 > KwkJCQ0KIAkJfQ0KIAl9DQogI2VuZGlmDQotLS0gUHl0aG9uLWN2cy9QQy9jb25maWcuYwlG
 > cmkgQXByICA3IDE1OjI3OjA0IDIwMDANCisrKyBQeXRob24tZ2MvUEMvY29uZmlnLmMJU2F0
 > IEFwciAgOCAwMzoxMjo1OCAyMDAwDQpAQCAtNDEsNiArNDEsOSBAQA0KIGV4dGVybiB2b2lk
 > IGluaXRiaW5hc2NpaSgpOw0KIGV4dGVybiB2b2lkIGluaXRjbWF0aCgpOw0KIGV4dGVybiB2
 > b2lkIGluaXRlcnJubygpOw0KKyNpZmRlZiBXSVRIX0NZQ0xFX0dDDQorZXh0ZXJuIHZvaWQg
 > aW5pdGdjKCk7DQorI2VuZGlmDQogZXh0ZXJuIHZvaWQgaW5pdGltYWdlb3AoKTsNCiBleHRl
 > cm4gdm9pZCBpbml0bWF0aCgpOw0KIGV4dGVybiB2b2lkIGluaXRtZDUoKTsNCkBAIC04MCw2
 > ICs4Myw5IEBADQogICAgICAgICB7ImJpbmFzY2lpIiwgaW5pdGJpbmFzY2lpfSwNCiAgICAg
 > ICAgIHsiY21hdGgiLCBpbml0Y21hdGh9LA0KICAgICAgICAgeyJlcnJubyIsIGluaXRlcnJu
 > b30sDQorI2lmZGVmIFdJVEhfQ1lDTEVfR0MNCisgICAgICAgIHsiZ2MiLCBpbml0Z2N9LA0K
 > KyNlbmRpZg0KICAgICAgICAgeyJpbWFnZW9wIiwgaW5pdGltYWdlb3B9LA0KICAgICAgICAg
 > eyJtYXRoIiwgaW5pdG1hdGh9LA0KICAgICAgICAgeyJtZDUiLCBpbml0bWQ1fSwNCi0tLSBQ
 > eXRob24tY3ZzL1BDL2NvbmZpZy5oCVRodSBNYXIgMzAgMjI6MDI6NTMgMjAwMA0KKysrIFB5
 > dGhvbi1nYy9QQy9jb25maWcuaAlTYXQgQXByICA4IDAyOjI0OjAwIDIwMDANCkBAIC0zODEs
 > NiArMzgxLDkgQEANCiAvKiBEZWZpbmUgaWYgeW91IHdhbnQgdG8gdXNlIHRoZSBHTlUgcmVh
 > ZGxpbmUgbGlicmFyeSAqLw0KIC8qICNkZWZpbmUgV0lUSF9SRUFETElORSAxICovDQogDQor
 > LyogRGVmaW5lIGlmIHlvdSB3YW50IGN5Y2xlIGdhcmJhZ2UgY29sbGVjdGlvbiAqLw0KKyNk
 > ZWZpbmUgV0lUSF9DWUNMRV9HQyAxDQorDQogLyogRGVmaW5lIGlmIHlvdSBoYXZlIGNsb2Nr
 > LiAgKi8NCiAvKiAjZGVmaW5lIEhBVkVfQ0xPQ0sgKi8NCiANCi0tLSBQeXRob24tY3ZzL01h
 > a2VmaWxlLmluCVN1biBBcHIgIDIgMjI6NTY6MjUgMjAwMA0KKysrIFB5dGhvbi1nYy9NYWtl
 > ZmlsZS5pbglTYXQgQXByICA4IDAxOjM4OjI1IDIwMDANCkBAIC00MDEsNyArNDAxLDcgQEAN
 > CiAJCSQoSU5TVEFMTF9EQVRBKSBNb2R1bGVzL01ha2VmaWxlICQoTElCUEwpL01ha2VmaWxl
 > DQogCQkkKElOU1RBTExfREFUQSkgTW9kdWxlcy9TZXR1cCAkKExJQlBMKS9TZXR1cA0KIAkJ
 > JChJTlNUQUxMX0RBVEEpIE1vZHVsZXMvU2V0dXAubG9jYWwgJChMSUJQTCkvU2V0dXAubG9j
 > YWwNCi0JCSQoSU5TVEFMTF9EQVRBKSBNb2R1bGVzL1NldHVwLnRocmVhZCAkKExJQlBMKS9T
 > ZXR1cC50aHJlYWQNCisJCSQoSU5TVEFMTF9EQVRBKSBNb2R1bGVzL1NldHVwLmF1dG8gJChM
 > SUJQTCkvU2V0dXAuYXV0bw0KIAkJJChJTlNUQUxMX1BST0dSQU0pICQoc3JjZGlyKS9Nb2R1
 > bGVzL21ha2VzZXR1cCAkKExJQlBMKS9tYWtlc2V0dXANCiAJCSQoSU5TVEFMTF9QUk9HUkFN
 > KSAkKHNyY2RpcikvaW5zdGFsbC1zaCAkKExJQlBMKS9pbnN0YWxsLXNoDQogCQkkKElOU1RB
 > TExfREFUQSkgJChzcmNkaXIpL01pc2MvTWFrZWZpbGUucHJlLmluICQoTElCUEwpL01ha2Vm
 > aWxlLnByZS5pbg0KLS0tIFB5dGhvbi1jdnMvY29uZmlnLmguaW4JRnJpIE1hciAyNCAyMjoz
 > MTo0MiAyMDAwDQorKysgUHl0aG9uLWdjL2NvbmZpZy5oLmluCVNhdCBBcHIgIDggMDE6Mzg6
 > MjUgMjAwMA0KQEAgLTIwNCw2ICsyMDQsOSBAQA0KICAgIChzaGFyZWQgbGlicmFyeSBwbHVz
 > IGFjY2Vzc29yeSBmaWxlcykuICovDQogI3VuZGVmIFdJVEhfTkVYVF9GUkFNRVdPUksNCiAN
 > CisvKiBEZWZpbmUgaWYgeW91IHdhbnQgY3ljbGUgZ2FyYmFnZSBjb2xsZWN0aW9uICovDQor
 > I3VuZGVmIFdJVEhfQ1lDTEVfR0MNCisNCiAvKiBUaGUgbnVtYmVyIG9mIGJ5dGVzIGluIGFu
 > IG9mZl90LiAqLw0KICN1bmRlZiBTSVpFT0ZfT0ZGX1QNCiANCi0tLSBQeXRob24tY3ZzL2Nv
 > bmZpZ3VyZS5pbglTdW4gQXByICAyIDIyOjU2OjI1IDIwMDANCisrKyBQeXRob24tZ2MvY29u
 > ZmlndXJlLmluCVNhdCBBcHIgIDggMDE6Mzg6MjUgMjAwMA0KQEAgLTEwNjksMTAgKzEwNjks
 > MjEgQEANCiBmaV0sDQogW0FDX01TR19SRVNVTFQobm8pXSkNCiANCitBQ19TVUJTVChVU0Vf
 > R0NfTU9EVUxFKQ0KK1VTRV9HQ19NT0RVTEU9IiMiDQorDQorQUNfTVNHX0NIRUNLSU5HKGZv
 > ciAtLXdpdGgtY3ljbGUtZ2MpDQorQUNfQVJHX1dJVEgoY3ljbGUtZ2MsIFstLXdpdGgtY3lj
 > bGUtZ2MgICAgICAgICAgIGVuYWJsZSBnYXJiYWdlIGNvbGxlY3Rpb25dLCBbDQorQUNfTVNH
 > X1JFU1VMVCgkd2l0aHZhbCkNCitBQ19ERUZJTkUoV0lUSF9DWUNMRV9HQykNCitVU0VfR0Nf
 > TU9EVUxFPQ0KK10sDQorQUNfTVNHX1JFU1VMVChubykpDQorDQogIyBnZW5lcmF0ZSBvdXRw
 > dXQgZmlsZXMNCiBBQ19PVVRQVVQoTWFrZWZpbGUgXA0KICBPYmplY3RzL01ha2VmaWxlIFwN
 > CiAgUGFyc2VyL01ha2VmaWxlIFwNCiAgUHl0aG9uL01ha2VmaWxlIFwNCiAgTW9kdWxlcy9N
 > YWtlZmlsZS5wcmUgXA0KLSBNb2R1bGVzL1NldHVwLnRocmVhZCkNCisgTW9kdWxlcy9TZXR1
 > cC5hdXRvKQ0KLS0tIFB5dGhvbi1jdnMvTWlzYy9NYWtlZmlsZS5wcmUuaW4JV2VkIERlYyAg
 > OSAxMDowNTozMyAxOTk4DQorKysgUHl0aG9uLWdjL01pc2MvTWFrZWZpbGUucHJlLmluCVNh
 > dCBBcHIgIDggMDE6Mzg6MjUgMjAwMA0KQEAgLTE2OCw3ICsxNjgsNyBAQA0KIE1BS0VGSUxF
 > PQkkKExJQlBMKS9NYWtlZmlsZQ0KIENPTkZJR0M9CSQoTElCUEwpL2NvbmZpZy5jDQogQ09O
 > RklHQ0lOPQkkKExJQlBMKS9jb25maWcuYy5pbg0KLVNFVFVQPQkJJChMSUJQTCkvU2V0dXAu
 > dGhyZWFkICQoTElCUEwpL1NldHVwLmxvY2FsICQoTElCUEwpL1NldHVwDQorU0VUVVA9CQkk
 > KExJQlBMKS9TZXR1cC5hdXRvICQoTElCUEwpL1NldHVwLmxvY2FsICQoTElCUEwpL1NldHVw
 > DQogDQogU1lTTElCUz0JJChMSUJNKSAkKExJQkMpDQogDQotLS0gUHl0aG9uLWN2cy9Nb2R1
 > bGVzL01ha2VmaWxlLnByZS5pbglGcmkgTWFyIDI0IDIyOjMyOjI1IDIwMDANCisrKyBQeXRo
 > b24tZ2MvTW9kdWxlcy9NYWtlZmlsZS5wcmUuaW4JU2F0IEFwciAgOCAwMTozODoyNSAyMDAw
 > DQpAQCAtMTQ3LDEwICsxNDcsMTAgQEANCiAjIGdldHMgcmVtYWRlIGZyb20gc2NyYXRjaDsg
 > dGhpcyBlbnN1cmVzIHRvIHJlbW92ZSBtb2R1bGVzIHRoYXQgYXJlIG5vDQogIyBsb25nZXIg
 > cGVydGluZW50IChidXQgdGhhdCB3ZXJlIGluIGEgcHJldmlvdXMgY29uZmlndXJhdGlvbiku
 > DQogY29uZmlnLmMgTWFrZWZpbGU6IE1ha2VmaWxlLnByZSBjb25maWcuYy5pbiAkKE1BS0VT
 > RVRVUCkNCi1jb25maWcuYyBNYWtlZmlsZTogU2V0dXAudGhyZWFkIFNldHVwIFNldHVwLmxv
 > Y2FsDQorY29uZmlnLmMgTWFrZWZpbGU6IFNldHVwLmF1dG8gU2V0dXAgU2V0dXAubG9jYWwN
 > CiBjb25maWcuYyBNYWtlZmlsZToNCiAJCS1ybSAtZiAkKExJQlJBUlkpDQotCQkkKFNIRUxM
 > KSAkKE1BS0VTRVRVUCkgU2V0dXAudGhyZWFkIFNldHVwLmxvY2FsIFNldHVwDQorCQkkKFNI
 > RUxMKSAkKE1BS0VTRVRVUCkgU2V0dXAuYXV0byBTZXR1cC5sb2NhbCBTZXR1cA0KIA0KIGhh
 > c3NpZ25hbDoNCiAJCS1ybSAtZiBoYXNzaWduYWwNCi0tLSBQeXRob24tY3ZzL01vZHVsZXMv
 > U2V0dXAuaW4JU3VuIEFwciAgMiAyMjo1NjozNyAyMDAwDQorKysgUHl0aG9uLWdjL01vZHVs
 > ZXMvU2V0dXAuaW4JU2F0IEFwciAgOCAwMTozODoyNSAyMDAwDQpAQCAtMTAwLDcgKzEwMCw3
 > IEBADQogR0xIQUNLPS1EY2xlYXI9X19HTGNsZWFyDQogI2dsIGdsbW9kdWxlLmMgY2dlbnN1
 > cHBvcnQuYyAtSSQoc3JjZGlyKSAkKEdMSEFDSykgLWxnbCAtbFgxMQ0KIA0KLSMgVGhlIHRo
 > cmVhZCBtb2R1bGUgaXMgbm93IGF1dG9tYXRpY2FsbHkgZW5hYmxlZCwgc2VlIFNldHVwLnRo
 > cmVhZC4NCisjIFRoZSB0aHJlYWQgbW9kdWxlIGlzIG5vdyBhdXRvbWF0aWNhbGx5IGVuYWJs
 > ZWQsIHNlZSBTZXR1cC5hdXRvLg0KIA0KICMgUHVyZSBtb2R1bGUuICBDYW5ub3QgYmUgbGlu
 > a2VkIGR5bmFtaWNhbGx5Lg0KICMgLURXSVRIX1FVQU5USUZZLCAtRFdJVEhfUFVSSUZZLCBv
 > ciAtRFdJVEhfQUxMX1BVUkUNCi0tLSBQeXRob24tY3ZzL01vZHVsZXMvU2V0dXAuYXV0by5p
 > bg0KKysrIFB5dGhvbi1nYy9Nb2R1bGVzL1NldHVwLmF1dG8uaW4JU2F0IEFwciAgOCAwMToz
 > ODoyNSAyMDAwDQpAQCAtMCwwICsxLDE2IEBADQorIyBUaGlzIGZpbGUgaXMgdHJhbnNtb2dy
 > aWZpZWQgaW50byBTZXR1cC5hdXRvIGJ5IGNvbmZpZy5zdGF0dXMuICBJdHMNCisjIHB1cnBv
 > c2UgaXMgdG8gYXV0b21hdGljYWxseSBlbmFibGUgbW9kdWxlcyBnaXZlbiB3aGVuIGNlcnRh
 > aW4NCisjIGFyZ3VtZW50cyBhcmUgZ2l2ZW4gdG8gdGhlIGNvbmZpZ3VyZSBzY3JpcHQuICBJ
 > dCByZXBsYWNlcyB0aGUgb2xkDQorIyBTZXR1cC50aHJlYWQuaW4gZmlsZS4NCisNCisjIElu
 > Y2x1ZGUgdGhlIHRocmVhZCBtb2R1bGUgd2hlbiB0aGUgLS13aXRoLXRocmVhZCBhcmd1bWVu
 > dCBpcyBnaXZlbiB0bw0KKyMgdGhlIGNvbmZpZ3VyZSBzY3JpcHQuDQorIw0KKyMgKk5PVEUq
 > OiBpZiB0aGUgY29uZmlndXJlIHNjcmlwdCBkZWNpZGVzIGl0IGNhbid0IHN1cHBvcnQgdGhy
 > ZWFkcywgdGhlDQorIyB0aHJlYWQgbW9kdWxlIHdpbGwgc3RpbGwgYmUgZW5hYmxlZCBhbmQg
 > Y2F1c2UgY29tcGlsZSBlcnJvcnMuICBUaGUNCisjIHNvbHV0aW9uIGlzIG5vdCB0byB1c2Ug
 > LS13aXRoLXRocmVhZCBvbiBwbGF0Zm9ybXMgdGhhdCBkb24ndCBzdXBwb3J0DQorIyB0aHJl
 > YWRzLg0KK0BVU0VfVEhSRUFEX01PRFVMRUB0aHJlYWQgdGhyZWFkbW9kdWxlLmMNCisNCisj
 > IEdhcmJhZ2UgY29sbGVjdGlvbiBlbmFibGVkIHdpdGggLS13aXRoLWN5Y2xlLWdjDQorQFVT
 > RV9HQ19NT0RVTEVAZ2MgZ2Ntb2R1bGUuYw0KLS0tIFB5dGhvbi1jdnMvTGliL3Rlc3QvdGVz
 > dF9nYy5weQ0KKysrIFB5dGhvbi1nYy9MaWIvdGVzdC90ZXN0X2djLnB5CVNhdCBBcHIgIDgg
 > MDM6MjU6MjUgMjAwMA0KQEAgLTAsMCArMSw5MCBAQA0KK2ltcG9ydCBnYw0KKw0KK2RlZiB0
 > ZXN0X2xpc3QoKToNCisgICAgbCA9IFtdDQorICAgIGwuYXBwZW5kKGwpDQorICAgIHByaW50
 > ICdsaXN0IDB4JXgnICUgaWQobCkNCisgICAgZ2MuY29sbGVjdCgpDQorICAgIGRlbCBsDQor
 > ICAgIGFzc2VydCBnYy5jb2xsZWN0KCkgPT0gMQ0KKw0KK2RlZiB0ZXN0X2RpY3QoKToNCisg
 > ICAgZCA9IHt9DQorICAgIGRbMV0gPSBkDQorICAgIHByaW50ICdkaWN0IDB4JXgnICUgaWQo
 > ZCkNCisgICAgZ2MuY29sbGVjdCgpDQorICAgIGRlbCBkDQorICAgIGFzc2VydCBnYy5jb2xs
 > ZWN0KCkgPT0gMQ0KKw0KK2RlZiB0ZXN0X3R1cGxlKCk6DQorICAgIGwgPSBbXQ0KKyAgICB0
 > ID0gKGwsKQ0KKyAgICBsLmFwcGVuZCh0KQ0KKyAgICBwcmludCAnbGlzdCAweCV4JyAlIGlk
 > KGwpDQorICAgIHByaW50ICd0dXBsZSAweCV4JyAlIGlkKHQpDQorICAgIGdjLmNvbGxlY3Qo
 > KQ0KKyAgICBkZWwgdA0KKyAgICBkZWwgbA0KKyAgICBhc3NlcnQgZ2MuY29sbGVjdCgpID09
 > IDINCisNCitkZWYgdGVzdF9jbGFzcygpOg0KKyAgICBjbGFzcyBBOg0KKyAgICAgICAgcGFz
 > cw0KKyAgICBBLmEgPSBBDQorICAgIHByaW50ICdjbGFzcyAweCV4JyAlIGlkKEEpDQorICAg
 > IGdjLmNvbGxlY3QoKQ0KKyAgICBkZWwgQQ0KKyAgICBhc3NlcnQgZ2MuY29sbGVjdCgpID09
 > IDINCisNCitkZWYgdGVzdF9pbnN0YW5jZSgpOg0KKyAgICBjbGFzcyBBOg0KKyAgICAgICAg
 > cGFzcw0KKyAgICBhID0gQSgpDQorICAgIGEuYSA9IGENCisgICAgcHJpbnQgcmVwcihhKQ0K
 > KyAgICBnYy5jb2xsZWN0KCkNCisgICAgZGVsIGENCisgICAgYXNzZXJ0IGdjLmNvbGxlY3Qo
 > KSA9PSAyDQorDQorZGVmIHRlc3RfZmluYWxpemVyKCk6DQorICAgIGNsYXNzIEE6DQorICAg
 > ICAgICBkZWYgX19kZWxfXyhzZWxmKTogcGFzcw0KKyAgICBjbGFzcyBCOg0KKyAgICAgICAg
 > cGFzcw0KKyAgICBhID0gQSgpDQorICAgIGEuYSA9IGENCisgICAgaWRfYSA9IGlkKGEpDQor
 > ICAgIGIgPSBCKCkNCisgICAgYi5iID0gYg0KKyAgICBwcmludCAnYScsIHJlcHIoYSkNCisg
 > ICAgcHJpbnQgJ2InLCByZXByKGIpDQorICAgIGdjLmNvbGxlY3QoKQ0KKyAgICBnYy5nYXJi
 > YWdlWzpdID0gW10NCisgICAgZGVsIGENCisgICAgZGVsIGINCisgICAgYXNzZXJ0IGdjLmNv
 > bGxlY3QoKSA9PSA0DQorICAgIGFzc2VydCBpZChnYy5nYXJiYWdlWzBdKSA9PSBpZF9hDQor
 > DQorZGVmIHRlc3RfZnVuY3Rpb24oKToNCisgICAgZCA9IHt9DQorICAgIGV4ZWMoImRlZiBm
 > KCk6IHBhc3NcbiIpIGluIGQNCisgICAgcHJpbnQgJ2RpY3QgMHgleCcgJSBpZChkKQ0KKyAg
 > ICBwcmludCAnZnVuYyAweCV4JyAlIGlkKGRbJ2YnXSkNCisgICAgZ2MuY29sbGVjdCgpDQor
 > ICAgIGRlbCBkDQorICAgIGFzc2VydCBnYy5jb2xsZWN0KCkgPT0gMg0KKw0KKw0KK2RlZiB0
 > ZXN0X2FsbCgpOg0KKyAgICBkZWJ1ZyA9IGdjLmdldF9kZWJ1ZygpDQorICAgIGdjLnNldF9k
 > ZWJ1ZyhnYy5ERUJVR19MRUFLIHwgZ2MuREVCVUdfU1RBVFMpDQorICAgIHRlc3RfbGlzdCgp
 > DQorICAgIHRlc3RfZGljdCgpDQorICAgIHRlc3RfdHVwbGUoKQ0KKyAgICB0ZXN0X2NsYXNz
 > KCkNCisgICAgdGVzdF9pbnN0YW5jZSgpDQorICAgIHRlc3RfZmluYWxpemVyKCkNCisgICAg
 > dGVzdF9mdW5jdGlvbigpDQorICAgIGdjLnNldF9kZWJ1ZyhkZWJ1ZykNCisNCit0ZXN0X2Fs
 > bCgpDQotLS0gUHl0aG9uLWN2cy9QQ2J1aWxkL3B5dGhvbjE2LmRzcAlGcmkgQXByICA3IDE1
 > OjI3OjA0IDIwMDANCisrKyBQeXRob24tZ2MvUENidWlsZC9weXRob24xNi5kc3AJU2F0IEFw
 > ciAgOCAwMjoyMjo0OCAyMDAwDQpAQCAtNjQ1LDYgKzY0NSwyMSBAQA0KICMgRW5kIFNvdXJj
 > ZSBGaWxlDQ0KICMgQmVnaW4gU291cmNlIEZpbGUNDQogDQ0KK1NPVVJDRT0uLlxNb2R1bGVz
 > XGdjbW9kdWxlLmMNDQorDQ0KKyFJRiAgIiQoQ0ZHKSIgPT0gInB5dGhvbjE2IC0gV2luMzIg
 > UmVsZWFzZSINDQorDQ0KKyFFTFNFSUYgICIkKENGRykiID09ICJweXRob24xNiAtIFdpbjMy
 > IERlYnVnIg0NCisNDQorIUVMU0VJRiAgIiQoQ0ZHKSIgPT0gInB5dGhvbjE2IC0gV2luMzIg
 > QWxwaGEgRGVidWciDQ0KKw0NCishRUxTRUlGICAiJChDRkcpIiA9PSAicHl0aG9uMTYgLSBX
 > aW4zMiBBbHBoYSBSZWxlYXNlIg0NCisNDQorIUVORElGIA0NCisNDQorIyBFbmQgU291cmNl
 > IEZpbGUNDQorIyBCZWdpbiBTb3VyY2UgRmlsZQ0NCisNDQogU09VUkNFPS4uXFB5dGhvblxn
 > ZXRhcmdzLmMNDQogDQ0KICFJRiAgIiQoQ0ZHKSIgPT0gInB5dGhvbjE2IC0gV2luMzIgUmVs
 > ZWFzZSIN
 > 
 > 
 > 
 > 
 > --__--__--
 > 
 > _______________________________________________
 > Patches mailing list
 > Patches@python.org
 > http://www.python.org/mailman/listinfo/patches
 > 
 > 
 > End of Patches Digest_______________________________________________
 > Patches mailing list
 > Patches@python.org
 > http://www.python.org/mailman/listinfo/patches


From nascheme@enme.ucalgary.ca  Tue Apr 25 17:58:29 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 25 Apr 2000 10:58:29 -0600
Subject: [Patches] why not just apply Garbage Collection patch?
In-Reply-To: <14596.41644.182242.918286@buffalo.fnal.gov>; from cgw@fnal.gov on Mon, Apr 24, 2000 at 02:38:20PM -0500
References: <14596.41644.182242.918286@buffalo.fnal.gov>
Message-ID: <20000425105829.A27743@acs.ucalgary.ca>

On Mon, Apr 24, 2000 at 02:38:20PM -0500, Charles G Waldman wrote:
> Unfortunately, with my recent changes to tupleobject.c and Christian's
> recent changes to the Trashcan, the gc patch no longer applies
> cleanly

I'll fix that today.

> Everything is already protected with "#ifdef
> WITH_CYCLE_GC" anyhow, right?

Unfortunately not.  I had to change the way objects are allocated
in order to make GC Python binary compatible with GC-disabled
Python.  PyObject_NEW* cannot be macros which expand to malloc()
(which they do currently on Windows).  GC Python must allocate
extra memory for container objects.

I think Vladimir's malloc patch will fix this problem.  Either
his patch should be applied first or at least the part that makes
this change.

    Neil

-- 
"God, root, what is difference?" - Pitr
"God is more forgiving." - Dave Aronson


From nascheme@enme.ucalgary.ca  Tue Apr 25 18:05:06 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 25 Apr 2000 11:05:06 -0600
Subject: [Patches] why not just apply Garbage Collection patch?
In-Reply-To: <200004242013.QAA29938@eric.cnri.reston.va.us>; from guido@python.org on Mon, Apr 24, 2000 at 04:13:08PM -0400
References: <14596.41644.182242.918286@buffalo.fnal.gov> <200004242013.QAA29938@eric.cnri.reston.va.us>
Message-ID: <20000425110506.B27743@acs.ucalgary.ca>

On Mon, Apr 24, 2000 at 04:13:08PM -0400, Guido van Rossum wrote:
> Thanks.  I haven't looked at it lately.  What do you think of binary
> compatibility?  Could an extension built before this was patched in
> get in trouble because of changed object lay-out?

Yes.  The way PyObject_NEW works on Windows is a problem.

> Plus, I owe Vladimir Marangozov a green light to generate a new
> version of his memory management patches; that will also interfere
> with the GC patches (and make a few things easier, I believe).
> Vladimir's patches should preferably be applied before Neil's.

Yes.  Vladimir can then get the blame for breaking already
compiled extensions. :)

    Neil

-- 
"C makes it easy to shoot yourself in the foot;
C++ makes it harder, but when you do it blows your whole leg off."
                -- Bjarne Stroustrup, Inventor of C++


From cgw@fnal.gov  Tue Apr 25 20:17:26 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Tue, 25 Apr 2000 14:17:26 -0500
Subject: [Patches] Trashcan vs. GC vs. TRACE_REFS
Message-ID: <200004251917.OAA20467@buffalo.fnal.gov>

Hi Christian et al.

Well, I said there was nothing wrong with Trashcan, but I spoke a
little too soon ;-)

With the current gc-cycle patch and Py_TRACE_REFS set, we get a core
dump.

The culprit is this little piece of code in object.c

void
_Py_Dealloc(op)
	PyObject *op;
{
	destructor dealloc = op->ob_type->tp_dealloc;
	_Py_ForgetReference(op);
***	if (_PyTrash_delete_nesting < PyTrash_UNWIND_LEVEL-1)
***	    op->ob_type = NULL; 
	(*dealloc)(op);
}


I'd really love to understand what those lines marked "***" are for.

If you set ob_type to NULL before doing a dealloc, you get into
trouble, because the new GC dealloc functions need to check ob_type to
decide if the object being deleted has GC info.  By setting type to
NULL you are breaking this.

But what really seems odd to me is that the _Py_Dealloc function above
is only used when Py_TRACE_REFS is set.  Normally the following macro
definition is used instead (see object.h)

ifndef Py_TRACE_REFS
...
#define _Py_Dealloc(op) (*(op)->ob_type->tp_dealloc)((PyObject *)(op))
...
#endif /* !Py_TRACE_REFS */

So, what is the point of worrying about _PyTrash_UNWIND_LEVEL only in
a function which is only used in special builds?  If this is really
important, I'd think you'd be doing it in the _Py_Dealloc macro as
well.  But, of course, you can't be clobbering ob_type at all, because
the gc dealloc functions *need* that ob_type.

As far as I can tell, building with Py_TRACE_REFS is not too widely
done.  But I'm finding it incredibly useful for tracking down those
reference leaks which still remain even with Neil's GC patches in
place.




From tismer@tismer.com  Tue Apr 25 20:35:26 2000
From: tismer@tismer.com (Christian Tismer)
Date: Tue, 25 Apr 2000 21:35:26 +0200
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
References: <200004251917.OAA20467@buffalo.fnal.gov>
Message-ID: <3905F37E.3DB3AA9B@tismer.com>


Charles G Waldman wrote:
> 
> Hi Christian et al.
> 
> Well, I said there was nothing wrong with Trashcan, but I spoke a
> little too soon ;-)
> 
> With the current gc-cycle patch and Py_TRACE_REFS set, we get a core
> dump.

Ok, that may be. I just inserted what's needed to make
MHammond happy when debugging.
Maybe we need some more.
(But I'm glad that it doesn't crash in the release build!)

((ha, even more glad: It isn't my error at all. hee hee))

> The culprit is this little piece of code in object.c
> 
> void
> _Py_Dealloc(op)
>         PyObject *op;
> {
>         destructor dealloc = op->ob_type->tp_dealloc;
>         _Py_ForgetReference(op);
> ***     if (_PyTrash_delete_nesting < PyTrash_UNWIND_LEVEL-1)
> ***         op->ob_type = NULL;
>         (*dealloc)(op);
> }
> 
> I'd really love to understand what those lines marked "***" are for.

This is so: Guido wipes the type field, in order to provoke
a core dump when you are using a dead object.
The code looked before like so:

> void
> _Py_Dealloc(op)
>         PyObject *op;
> {
>         destructor dealloc = op->ob_type->tp_dealloc;
>         _Py_ForgetReference(op);
>         op->ob_type = NULL;
>         (*dealloc)(op);
> }

That means, Neil's code would crash anyway in this mode.
What I do here is:
I allow to wipe the type field if and only if I'm sure that
I don't need it. This is on final destruction, which will happen
only if _PyTrash_delete_nesting < PyTrash_UNWIND_LEVEL-1 .

> If you set ob_type to NULL before doing a dealloc, you get into
> trouble, because the new GC dealloc functions need to check ob_type to
> decide if the object being deleted has GC info.  By setting type to
> NULL you are breaking this.

Agreed, but it's not me.
I guess we should remove this piece of code for Neil,
let's embrace it in a #ifdef.

> But what really seems odd to me is that the _Py_Dealloc function above
> is only used when Py_TRACE_REFS is set.  Normally the following macro
> definition is used instead (see object.h)
> 
> ifndef Py_TRACE_REFS
> ...
> #define _Py_Dealloc(op) (*(op)->ob_type->tp_dealloc)((PyObject *)(op))
> ...
> #endif /* !Py_TRACE_REFS */
> 
> So, what is the point of worrying about _PyTrash_UNWIND_LEVEL only in
> a function which is only used in special builds?  If this is really
> important, I'd think you'd be doing it in the _Py_Dealloc macro as
> well.  But, of course, you can't be clobbering ob_type at all, because
> the gc dealloc functions *need* that ob_type.

It is only there to allow Guido (or whoever wrote this debug
piece) to allow to clear the type if it doesn't hurt *me*.
When I have to delay destruction, I must have the type alive.
Neil probably needs it always?

> As far as I can tell, building with Py_TRACE_REFS is not too widely
> done.  But I'm finding it incredibly useful for tracking down those
> reference leaks which still remain even with Neil's GC patches in
> place.

Ok, all clear now?
Well, my lovely little trashcan pet is really strong.
Brave pet, brave pet :-))

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From tismer@tismer.com  Tue Apr 25 20:47:40 2000
From: tismer@tismer.com (Christian Tismer)
Date: Tue, 25 Apr 2000 21:47:40 +0200
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
References: <200004251917.OAA20467@buffalo.fnal.gov> <3905F37E.3DB3AA9B@tismer.com>
Message-ID: <3905F65C.873B0B26@tismer.com>

One more on this:
If ob_type has to be cleared when tracing refs,
and if Neil's code is touching the object the last time,
shouldn't he then clear the type, right before the thing
gets doomed into the everblacks?

The situation is a bit hairy:
We cannot do it after the dealloc is done. I tried that before,
but I crashed, since after the dealloc the memory might already
be in use again, as a side effect of the debug code.

That means: We need to decide here wether the thingy will go
through Neil's code, and then also not clear the type.
Quit bad, this needs another if..else if.. cascade to check
the type.

Guido? How urgent do you think is to wipe the
type code? It seems to become a hair with GC.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From brian@garage.co.jp  Tue Apr 25 21:37:31 2000
From: brian@garage.co.jp (Brian Hooper)
Date: Tue, 25 Apr 2000 20:37:31 GMT
Subject: [Patches] slightly expanded Unicode supported Py_BuildValue
Message-ID: <20000425203731.80051.qmail@hotmail.com>

This is a multi-part message in MIME format.

------=_NextPart_000_119635c0_742cd712$5a06bf89
Content-Type: text/plain; format=flowed

Hi again,

Here's a patch which changes modsupport to add 'U' and 'U#',
to support building Unicode objects from a null-terminated
Py_UNICODE *, and a Py_UNICODE * with length, respectively.

If this is acceptable, then I'll submit a patch for the
corresponding documentation also.

--Brian


I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.
________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com

------=_NextPart_000_119635c0_742cd712$5a06bf89
Content-Type: application/x-unknown-content-type-Excel.DIF; name="Modsupport2.diff.dif"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="Modsupport2.diff.dif"
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------=_NextPart_000_119635c0_742cd712$5a06bf89--


From cgw@fnal.gov  Tue Apr 25 22:32:17 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Tue, 25 Apr 2000 16:32:17 -0500 (CDT)
Subject: [Patches] tiny patch for socketmodule.c
Message-ID: <14598.3809.919559.931801@buffalo.fnal.gov>

This patch does two things:

1)  Adds MSG_DONTWAIT if defined (I needed this)
2)  Spells "coreectly" correctly ;-)

Disclaimer:  http://www.python.org/patches/bugrelease.html

Index: Modules/socketmodule.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/socketmodule.c,v
retrieving revision 1.104
diff -c -r1.104 socketmodule.c
*** socketmodule.c	2000/04/24 15:16:03	1.104
--- socketmodule.c	2000/04/25 21:30:55
***************
*** 524,530 ****
  	}
  
  #ifdef __BEOS__
! 	/* XXX: BeOS version of accept() doesn't set family coreectly */
  	addr->sa_family = AF_INET;
  #endif
  
--- 524,530 ----
  	}
  
  #ifdef __BEOS__
! 	/* XXX: BeOS version of accept() doesn't set family correctly */
  	addr->sa_family = AF_INET;
  #endif
  
***************
*** 2525,2530 ****
--- 2525,2533 ----
  #endif
  #ifdef	MSG_DONTROUTE
  	insint(d, "MSG_DONTROUTE", MSG_DONTROUTE);
+ #endif
+ #ifdef	MSG_DONTWAIT
+ 	insint(d, "MSG_DONWAIT", MSG_DONTWAIT);
  #endif
  #ifdef	MSG_EOR
  	insint(d, "MSG_EOR", MSG_EOR);


From guido@python.org  Tue Apr 25 22:57:03 2000
From: guido@python.org (Guido van Rossum)
Date: Tue, 25 Apr 2000 17:57:03 -0400
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: Your message of "Tue, 25 Apr 2000 21:47:40 +0200."
 <3905F65C.873B0B26@tismer.com>
References: <200004251917.OAA20467@buffalo.fnal.gov> <3905F37E.3DB3AA9B@tismer.com>
 <3905F65C.873B0B26@tismer.com>
Message-ID: <200004252157.RAA02365@eric.cnri.reston.va.us>

> Guido? How urgent do you think is to wipe the
> type code? It seems to become a hair with GC.

I'll deal with GC later.  I think it's an important debugging feature
since it traps any use of the object after it has been deallocated
immediately.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From skip@mojam.com (Skip Montanaro)  Tue Apr 25 23:02:36 2000
From: skip@mojam.com (Skip Montanaro) (Skip Montanaro)
Date: Tue, 25 Apr 2000 17:02:36 -0500 (CDT)
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: <200004252157.RAA02365@eric.cnri.reston.va.us>
References: <200004251917.OAA20467@buffalo.fnal.gov>
 <3905F37E.3DB3AA9B@tismer.com>
 <3905F65C.873B0B26@tismer.com>
 <200004252157.RAA02365@eric.cnri.reston.va.us>
Message-ID: <14598.5628.536367.666435@beluga.mojam.com>

    >> Guido? How urgent do you think is to wipe the type code? It seems to
    >> become a hair with GC.

    Guido> I'll deal with GC later.  I think it's an important debugging
    Guido> feature since it traps any use of the object after it has been
    Guido> deallocated immediately.

Isn't this a problem because the ob_type field is effectively being used for
two competing purposes (I haven't been following *real* closely for fear of
someone slamming on the brakes...)?  Is there some way to split those two
functions into two separate fields, at least when Py_TRACE_REFS is defined?

-- 
Skip Montanaro | http://www.mojam.com/
skip@mojam.com | http://www.musi-cal.com/


From gstein@lyra.org  Tue Apr 25 23:14:29 2000
From: gstein@lyra.org (Greg Stein)
Date: Tue, 25 Apr 2000 15:14:29 -0700 (PDT)
Subject: [Patches] tiny patch for socketmodule.c
In-Reply-To: <14598.3809.919559.931801@buffalo.fnal.gov>
Message-ID: <Pine.LNX.4.10.10004251514080.25776-100000@nebula.lyra.org>

On Tue, 25 Apr 2000, Charles G Waldman wrote:
>...
> *** 2525,2530 ****
> --- 2525,2533 ----
>   #endif
>   #ifdef	MSG_DONTROUTE
>   	insint(d, "MSG_DONTROUTE", MSG_DONTROUTE);
> + #endif
> + #ifdef	MSG_DONTWAIT
> + 	insint(d, "MSG_DONWAIT", MSG_DONTWAIT);
>   #endif
>   #ifdef	MSG_EOR
>   	insint(d, "MSG_EOR", MSG_EOR);

You may want to spell the Python symbol correctly...

:-)

-- 
Greg Stein, http://www.lyra.org/



From cgw@fnal.gov  Tue Apr 25 23:09:28 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Tue, 25 Apr 2000 17:09:28 -0500 (CDT)
Subject: [Patches] tiny patch for socketmodule.c
In-Reply-To: <Pine.LNX.4.10.10004251514080.25776-100000@nebula.lyra.org>
References: <14598.3809.919559.931801@buffalo.fnal.gov>
 <Pine.LNX.4.10.10004251514080.25776-100000@nebula.lyra.org>
Message-ID: <14598.6040.389251.942676@buffalo.fnal.gov>

Greg Stein writes:
 > 
 > You may want to spell the Python symbol correctly...
 > 
 > :-)

Geez, I feel stupid!!!



From mal@lemburg.com  Tue Apr 25 23:16:02 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Wed, 26 Apr 2000 00:16:02 +0200
Subject: [Patches] slightly expanded Unicode supported Py_BuildValue
References: <20000425203731.80051.qmail@hotmail.com>
Message-ID: <39061922.F627F85D@lemburg.com>

Brian Hooper wrote:
> 
> Hi again,
> 
> Here's a patch which changes modsupport to add 'U' and 'U#',
> to support building Unicode objects from a null-terminated
> Py_UNICODE *, and a Py_UNICODE * with length, respectively.

If you would change the 'U' (capital U) to 'u' (lowercase u)
the patch would be perfect :-)
 
BTW, why do you support converting a NULL pointer to None ?
I think this would probably only mask programming errors
instead of doing any good.

> If this is acceptable, then I'll submit a patch for the
> corresponding documentation also.
>---
>                            Name: Modsupport2.diff.dif
>    Modsupport2.diff.dif    Type: application/x-unknown-content-type-Excel.DIF
>                        Encoding: base64

Your mailer seems to have gotten the content type wrong;
I'd suggest to use .patch as extension for the file.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/



From nascheme@enme.ucalgary.ca  Wed Apr 26 02:03:10 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 25 Apr 2000 19:03:10 -0600
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: <200004252157.RAA02365@eric.cnri.reston.va.us>; from guido@python.org on Tue, Apr 25, 2000 at 05:57:03PM -0400
References: <200004251917.OAA20467@buffalo.fnal.gov> <3905F37E.3DB3AA9B@tismer.com> <3905F65C.873B0B26@tismer.com> <200004252157.RAA02365@eric.cnri.reston.va.us>
Message-ID: <20000425190310.B25863@acs.ucalgary.ca>

On Tue, Apr 25, 2000 at 05:57:03PM -0400, Guido van Rossum wrote:
> I think it's an important debugging feature since it traps any
> use of the object after it has been deallocated immediately.

That feature works with GC now.  I have changed my patch to
minimize the impact on the rest of the Python code.  Almost all
the details are now hidden in the gc module.  I no longer need
Vladimir's malloc cleanup to be applied first.  The patch is
here:

    http://www.enme.ucalgary.ca/~nascheme/python/gc-cycle.diff

Guido, when you talk about not breaking extension modules what
exactly do you mean?  With my current patch, extension types that
are involved with GC must match the interpreter (GC vs.  no-GC).
Other extensions do not have to be recompiled (even ones compiled
before applying the patch).

Is this sufficient?  If not, it is possible to fix it so extensions
that implement GC will be okay too.


    Neil


From mhammond@skippinet.com.au  Wed Apr 26 02:39:00 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Wed, 26 Apr 2000 11:39:00 +1000
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: <20000425190310.B25863@acs.ucalgary.ca>
Message-ID: <ECEPKNMJLHAPFFJHDOJBEEKHCJAA.mhammond@skippinet.com.au>

> With my current patch, extension types that
> are involved with GC must match the interpreter (GC vs.  no-GC).

What exactly does "involved in" mean?  (Obviously, I havent looked at =
the patches).  Must an extension module actually "opt in" for GC, or do =
a module's inherent properties (eg, implements a new extension type) =
force it to become involved?

> Is this sufficient?  If not, it is possible to fix it so=20
> extensions
> that implement GC will be okay too.

That would certainly be ideal, especially on Windows where people often =
dont have the tools to build all their own binaries.  It would certainly =
help people on Windows experiment with this feature if they did not have =
to rebuild their world (or even parts of their world) to use it.

But I guess its importance (to me) will be determined by the answer to =
my first question...

Mark.



From nascheme@enme.ucalgary.ca  Wed Apr 26 02:52:20 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 25 Apr 2000 19:52:20 -0600
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: <ECEPKNMJLHAPFFJHDOJBEEKHCJAA.mhammond@skippinet.com.au>; from mhammond@skippinet.com.au on Wed, Apr 26, 2000 at 11:39:00AM +1000
References: <20000425190310.B25863@acs.ucalgary.ca> <ECEPKNMJLHAPFFJHDOJBEEKHCJAA.mhammond@skippinet.com.au>
Message-ID: <20000425195220.A26364@acs.ucalgary.ca>

On Wed, Apr 26, 2000 at 11:39:00AM +1000, Mark Hammond wrote:
> > With my current patch, extension types that
> > are involved with GC must match the interpreter (GC vs.  no-GC).
> 
> What exactly does "involved in" mean?

Extension types that can create reference cycles and want to
support GC.  There are currently none of these unless someone
quickly implemented one.

> > Is this sufficient? If not, it is possible to fix it so
> > extensions that implement GC will be okay too.
> 
> That would certainly be ideal, especially on Windows where
> people often dont have the tools to build all their own
> binaries.  It would certainly help people on Windows experiment
> with this feature if they did not have to rebuild their world
> (or even parts of their world) to use it.

To achieve this, GC-disabled Python would have to be a little
slower than current Python (some extra function calls).  With my
current patch they should be very nearly the same speed.

    Neil

-- 
Real Life? I played that game. The plot sucks but the graphics are
awesome.


From tismer@tismer.com  Wed Apr 26 01:56:32 2000
From: tismer@tismer.com (Christian Tismer)
Date: Wed, 26 Apr 2000 02:56:32 +0200
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
References: <200004251917.OAA20467@buffalo.fnal.gov> <3905F37E.3DB3AA9B@tismer.com> <3905F65C.873B0B26@tismer.com> <200004252157.RAA02365@eric.cnri.reston.va.us> <20000425190310.B25863@acs.ucalgary.ca>
Message-ID: <39063EC0.23174B81@tismer.com>


Neil Schemenauer wrote:
<schnapps - hick*>
> Guido, when you talk about not breaking extension modules what
> exactly do you mean?

I guess he is meaning this:
Does an extension have to be recompiled for your GC.

> With my current patch, extension types that
> are involved with GC must match the interpreter (GC vs.  no-GC).
> Other extensions do not have to be recompiled (even ones compiled
> before applying the patch).

That's important -- I can replace python16.dll, and the win32
extensions will not break -- ok?

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From mhammond@skippinet.com.au  Wed Apr 26 02:56:59 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Wed, 26 Apr 2000 11:56:59 +1000
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: <20000425195220.A26364@acs.ucalgary.ca>
Message-ID: <ECEPKNMJLHAPFFJHDOJBGEKICJAA.mhammond@skippinet.com.au>

> On Wed, Apr 26, 2000 at 11:39:00AM +1000, Mark Hammond wrote:
> > > With my current patch, extension types that
> > > are involved with GC must match the interpreter (GC=20
> vs.  no-GC).
> >=20
> > What exactly does "involved in" mean?
>=20
> Extension types that can create reference cycles and want to
> support GC.  There are currently none of these unless someone
> quickly implemented one.

If you are saying that there are no (known) extension types that could =
possibly be affected, and that any written in the future must match =
Python's GC mode - then it sounds perfect, and a bridge to cross when we =
come to it (by which time of course the patch will prove so wonderful =
that it is made standard :-)

That's-one-long-sentence-ly,

Mark.



From tismer@tismer.com  Wed Apr 26 02:00:42 2000
From: tismer@tismer.com (Christian Tismer)
Date: Wed, 26 Apr 2000 03:00:42 +0200
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
References: <200004251917.OAA20467@buffalo.fnal.gov> <3905F37E.3DB3AA9B@tismer.com>
 <3905F65C.873B0B26@tismer.com> <200004252157.RAA02365@eric.cnri.reston.va.us>
Message-ID: <39063FBA.592A42F@tismer.com>


Guido van Rossum wrote:
> 
> > Guido? How urgent do you think is to wipe the
> > type code? It seems to become a hair with GC.
> 
> I'll deal with GC later.  I think it's an important debugging feature
> since it traps any use of the object after it has been deallocated
> immediately.

hehee, how good that this isn't my problem
since I'mnot in charge...

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From nascheme@enme.ucalgary.ca  Wed Apr 26 03:11:31 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 25 Apr 2000 20:11:31 -0600
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: <ECEPKNMJLHAPFFJHDOJBGEKICJAA.mhammond@skippinet.com.au>; from mhammond@skippinet.com.au on Wed, Apr 26, 2000 at 11:56:59AM +1000
References: <20000425195220.A26364@acs.ucalgary.ca> <ECEPKNMJLHAPFFJHDOJBGEKICJAA.mhammond@skippinet.com.au>
Message-ID: <20000425201131.A26520@acs.ucalgary.ca>

On Wed, Apr 26, 2000 at 11:56:59AM +1000, Mark Hammond wrote:
> If you are saying that there are no (known) extension types
> that could possibly be affected, and that any written in the
> future must match Python's GC mode - then it sounds perfect,
> and a bridge to cross when we come to it (by which time of
> course the patch will prove so wonderful that it is made
> standard :-)

That's what I'm saying.  Types that want to "be involved in GC"
must:

    Use PyGC_NEW* instead of PyObject_NEW*.

    Use PyGC_Del to free the object structure.

    Call PyGC_Insert and PyGC_Remove appropriately.  Objects
    should only be included in the GC set when they are valid.
    
    Implement tp_recurse.

    Add the feature bit Py_TPFLAGS_HAVE_GCINFO.

Types which don't "op-in" for GC are safe.

> That's-one-long-sentence-ly,

Especially when its on one line. :)


    Neil


From guido@python.org  Wed Apr 26 03:15:08 2000
From: guido@python.org (Guido van Rossum)
Date: Tue, 25 Apr 2000 22:15:08 -0400
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: Your message of "Wed, 26 Apr 2000 02:56:32 +0200."
 <39063EC0.23174B81@tismer.com>
References: <200004251917.OAA20467@buffalo.fnal.gov> <3905F37E.3DB3AA9B@tismer.com> <3905F65C.873B0B26@tismer.com> <200004252157.RAA02365@eric.cnri.reston.va.us> <20000425190310.B25863@acs.ucalgary.ca>
 <39063EC0.23174B81@tismer.com>
Message-ID: <200004260215.WAA06150@eric.cnri.reston.va.us>

> > With my current patch, extension types that
> > are involved with GC must match the interpreter (GC vs.  no-GC).
> > Other extensions do not have to be recompiled (even ones compiled
> > before applying the patch).
> 
> That's important -- I can replace python16.dll, and the win32
> extensions will not break -- ok?

One concern (I also haven't seen Neil's recent patches).  If an
extension uses MACROS to access tuple or list items, then if the GC
code changes the header size for those objects those extensions are in
deep doodoo.

My own version avoided this by allocating the special fields *before*
the standard header (playing pointer tricks) but there are problems
with that approach too...

--Guido van Rossum (home page: http://www.python.org/~guido/)


From nascheme@enme.ucalgary.ca  Wed Apr 26 03:18:10 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 25 Apr 2000 20:18:10 -0600
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: <200004260215.WAA06150@eric.cnri.reston.va.us>; from guido@python.org on Tue, Apr 25, 2000 at 10:15:08PM -0400
References: <200004251917.OAA20467@buffalo.fnal.gov> <3905F37E.3DB3AA9B@tismer.com> <3905F65C.873B0B26@tismer.com> <200004252157.RAA02365@eric.cnri.reston.va.us> <20000425190310.B25863@acs.ucalgary.ca> <39063EC0.23174B81@tismer.com> <200004260215.WAA06150@eric.cnri.reston.va.us>
Message-ID: <20000425201810.A26682@acs.ucalgary.ca>

On Tue, Apr 25, 2000 at 10:15:08PM -0400, Guido van Rossum wrote:
> My own version avoided this by allocating the special fields *before*
> the standard header (playing pointer tricks) but there are problems
> with that approach too...

I use the same (ugly) trick.  What were your problems?

    Neil

-- 
"C makes it easy to shoot yourself in the foot;
C++ makes it harder, but when you do it blows your whole leg off."
                -- Bjarne Stroustrup, Inventor of C++


From guido@python.org  Wed Apr 26 04:12:27 2000
From: guido@python.org (Guido van Rossum)
Date: Tue, 25 Apr 2000 23:12:27 -0400
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: Your message of "Tue, 25 Apr 2000 20:18:10 MDT."
 <20000425201810.A26682@acs.ucalgary.ca>
References: <200004251917.OAA20467@buffalo.fnal.gov> <3905F37E.3DB3AA9B@tismer.com> <3905F65C.873B0B26@tismer.com> <200004252157.RAA02365@eric.cnri.reston.va.us> <20000425190310.B25863@acs.ucalgary.ca> <39063EC0.23174B81@tismer.com> <200004260215.WAA06150@eric.cnri.reston.va.us>
 <20000425201810.A26682@acs.ucalgary.ca>
Message-ID: <200004260312.XAA06255@eric.cnri.reston.va.us>

[me]
> > My own version avoided this by allocating the special fields *before*
> > the standard header (playing pointer tricks) but there are problems
> > with that approach too...

[Neil]
> I use the same (ugly) trick.  What were your problems?

Mostly just to get all the details right (e.g. I never got
_PyResize_Tuple done, and Charles seems to have had problems
integrating that with his own patches to that piece of code...).

Since you've got it done, and since it's the only way to be binary
compatible, it's best to keep it this way.

--Guido van Rossum (home page: http://www.python.org/~guido/)


From mwh21@cam.ac.uk  Wed Apr 26 11:10:13 2000
From: mwh21@cam.ac.uk (Michael Hudson)
Date: Wed, 26 Apr 2000 11:10:13 +0100 (BST)
Subject: [Patches] Better error handling for posix.execv
Message-ID: <Pine.LNX.4.10.10004261058180.20527-100000@localhost.localdomain>

This patch changes posixmodule.c:execv to 

a) check for zero length args (does this to execve, too), raising
   ValueError.
b) raises more rational exceptions for various flavours of duff arguments.
   I *hate*

TypeError: "illegal argument type for built-in operation"

   It has to be one of the most frustrating error messages ever.

Also, isn't the docstring for spawnv just wrong? (about replacing the
current process).  Another patch, maybe.

Cheers,
Michael

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

Index: posixmodule.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/posixmodule.c,v
retrieving revision 2.131
diff -u -r2.131 posixmodule.c
--- posixmodule.c	2000/04/21 18:54:45	2.131
+++ posixmodule.c	2000/04/26 10:07:42
@@ -1343,8 +1343,13 @@
 		getitem = PyTuple_GetItem;
 	}
 	else {
- badarg:
-		PyErr_BadArgument();
+		PyErr_SetString(PyExc_TypeError, "argv must be tuple or list");
+		return NULL;
+	}
+
+	if (argc == 0) {
+		PyErr_SetString(PyExc_ValueError, 
+				"empty argument list");
 		return NULL;
 	}
 
@@ -1354,7 +1359,10 @@
 	for (i = 0; i < argc; i++) {
 		if (!PyArg_Parse((*getitem)(argv, i), "s", &argvlist[i])) {
 			PyMem_DEL(argvlist);
-			goto badarg;
+			PyErr_SetString(PyExc_TypeError, 
+					"all arguments must be strings");
+			return NULL;
+			
 		}
 	}
 	argvlist[argc] = NULL;
@@ -1413,6 +1421,12 @@
 	}
 	if (!PyMapping_Check(env)) {
 		PyErr_SetString(PyExc_TypeError, "env must be mapping object");
+		return NULL;
+	}
+
+	if (argc == 0) {
+		PyErr_SetString(PyExc_ValueError, 
+				"empty argument list");
 		return NULL;
 	}
 



From mwh21@cam.ac.uk  Wed Apr 26 11:45:20 2000
From: mwh21@cam.ac.uk (Michael Hudson)
Date: Wed, 26 Apr 2000 11:45:20 +0100 (BST)
Subject: [Patches] accompaniment to patch to posixmodule.c
Message-ID: <Pine.LNX.4.10.10004261141350.20527-100000@localhost.localdomain>

Thanks heavens for test suites; my last patch broke os.exec[vl]p*, which
broke popen3 ... in some circumstances.

Anyway, this patch sorts that out.

Cheers,
M.

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under
copyright, patent or other rights or interests ("claims").  To
the extent that I have any such claims, I hereby grant to CNRI a
nonexclusive, irrevocable, royalty-free, worldwide license to
reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part
of the Python software and its related documentation, or any
derivative versions thereof, at no cost to CNRI or its licensed
users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide
whether or not to incorporate this contribution in the Python
software and its related documentation.  I further grant CNRI
permission to use my name and other identifying information
provided to CNRI by me for use in connection with the Python
software and its related documentation.

Index: os.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/os.py,v
retrieving revision 1.33
diff -u -r1.33 os.py
--- os.py	2000/04/25 10:53:22	1.33
+++ os.py	2000/04/26 10:44:54
@@ -252,7 +252,7 @@
     if not _notfound:
         import tempfile
         # Exec a file that is guaranteed not to exist
-        try: execv(tempfile.mktemp(), ())
+        try: execv(tempfile.mktemp(), ('blah',))
         except error, _notfound: pass
     exc, arg = error, _notfound
     for dir in PATH:



From tismer@tismer.com  Wed Apr 26 12:52:32 2000
From: tismer@tismer.com (Christian Tismer)
Date: Wed, 26 Apr 2000 13:52:32 +0200
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
References: <200004251917.OAA20467@buffalo.fnal.gov>
 <3905F37E.3DB3AA9B@tismer.com>
 <3905F65C.873B0B26@tismer.com>
 <200004252157.RAA02365@eric.cnri.reston.va.us> <14598.5628.536367.666435@beluga.mojam.com>
Message-ID: <3906D880.4324BB52@tismer.com>

Skip Montanaro wrote:
...
> Isn't this a problem because the ob_type field is effectively being used for
> two competing purposes (I haven't been following *real* closely for fear of
> someone slamming on the brakes...)?  Is there some way to split those two
> functions into two separate fields, at least when Py_TRACE_REFS is defined?

Aahh - that sounds good.

Proposal: (given Py_TRACE_REFS)
- allow ob_type to be wiped as it is done today
- add ob_typebak and save the type there
- in Neil's code refer to ob_typebak

yes these were in fact competing purposes - ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer@appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com


From gstein@lyra.org  Wed Apr 26 13:38:11 2000
From: gstein@lyra.org (Greg Stein)
Date: Wed, 26 Apr 2000 05:38:11 -0700 (PDT)
Subject: [Patches] Sightly clean-up cycle-gc patch
In-Reply-To: <14597.49537.611625.343576@buffalo.fnal.gov>
Message-ID: <Pine.LNX.4.10.10004260519550.25776-100000@nebula.lyra.org>

On Tue, 25 Apr 2000, Charles G Waldman wrote:
>...
> --- Python-cvs/Include/object.h	Fri Mar 24 22:32:16 2000
> +++ Python-gc/Include/object.h	Sat Apr  8 01:38:25 2000
> @@ -145,6 +145,10 @@
>  typedef int (*getsegcountproc) Py_PROTO((PyObject *, int *));
>  typedef int (*getcharbufferproc) Py_PROTO((PyObject *, int, const char **));
>  typedef int (*objobjproc) Py_PROTO((PyObject *, PyObject *));
> +#ifdef WITH_CYCLE_GC
> +typedef int (*visitproc) Py_PROTO((PyObject *, void *));
> +typedef int (*recurseproc) Py_PROTO((PyObject *, visitproc, void *));
> +#endif
>  
>  typedef struct {
>  	binaryfunc nb_add;
> @@ -243,9 +247,18 @@
>  
>  	char *tp_doc; /* Documentation string */
>  
> -	/* More spares */
> +#ifdef WITH_CYCLE_GC
> +	/* call function for all accessible objects */
> +	recurseproc tp_recurse;
> +	
> +	/* delete references to contained objects */
> +	inquiry tp_clear;
> +#else
>  	long tp_xxx5;
>  	long tp_xxx6;
> +#endif

Just rename those spares. I don't see that we could truly use them for
something else -- if somebody defined WITH_CYCLE_GC, then there would be a
conflict.

This also implies the typedefs further above would not be protected by the
macro.

>...
> --- Python-cvs/Include/objimpl.h	Thu Mar  2 17:02:23 2000
> +++ Python-gc/Include/objimpl.h	Sat Apr  8 03:18:22 2000
> @@ -38,8 +38,42 @@
>  /*
>  Additional macros for modules that implement new object types.
>  You must first include "object.h".
> +*/
> +
> +#ifdef WITH_CYCLE_GC 
> +
> +/* Structure *prefixed* to container objects participating in GC */ 
> +typedef struct _gcinfo {
> +	struct _gcinfo *gc_next;
> +	struct _gcinfo *gc_prev;
> +	int gc_refs;
> +} PyGCInfo;
> +
> +/* Test if a type has GC info */
> +#define PY_TYPEISGC(t) PyType_HasFeature((t), Py_TPFLAGS_HAVE_GCINFO)
> +
> +/* Test if an object has GC info */
> +#define PY_ISGC(o) PY_TYPEISGC((o)->ob_type)
> +
> +/* Get an object's GC info -- NULL if the object has none */
> +#define PY_GCINFO(o) (PY_ISGC(o) ? ((PyGCInfo *)(o)-1) : NULL)
> +
> +/* Unsafe version of PY_GCINFO() -- only call if PY_ISGC(p) is true */
> +#define PY_GCINFO_UNSAFE(o) ((PyGCInfo *)(o)-1)
> +
> +/* Get the object given the PyGCInfo */
> +#define PY_GCOBJ(g) ((PyObject *)((g)+1))

IMO, if it doesn't change the Python functionality, then don't protect
this stuff with WITH_CYCLE_GC.

These macros should also be named Py_* rather than PY_*. I'm not away of
anything using the PY_ prefix.

> -PyObject_NEW(type, typeobj) allocates memory for a new object of the given
> +/* Add the object into the container set */
> +extern DL_IMPORT(void) PyGC_Insert Py_PROTO((PyObject *));
> +
> +/* Remove the object from the container set */
> +extern DL_IMPORT(void) PyGC_Remove Py_PROTO((PyObject *));

These functions probably need to be protected. Seems like I recall a
linker that will look for them if an "extern" is found anywhere in the
program (even if no call is made to it).

>...
> --- Python-cvs/Modules/newmodule.c	Tue Feb 29 14:55:09 2000
> +++ Python-gc/Modules/newmodule.c	Sat Apr  8 02:13:26 2000
> @@ -49,13 +49,16 @@
>  			      &PyClass_Type, &klass,
>  			      &PyDict_Type, &dict))
>  		return NULL;
> -	inst = PyObject_NEW(PyInstanceObject, &PyInstance_Type);
> +	inst = PyObject_New(PyInstanceObject, &PyInstance_Type);
>  	if (inst == NULL)
>  		return NULL;
>  	Py_INCREF(klass);
>  	Py_INCREF(dict);
>  	inst->in_class = (PyClassObject *)klass;
>  	inst->in_dict = dict;
> +#ifdef WITH_CYCLE_GC
> +	PyGC_Insert((PyObject *)inst);
> +#endif

This would be MUCH cleaner if PyGC_Insert() evaluated to a no-op in the
non-GC case.

>...
> +/*** list functions (should probably be macros, stupid compilers) ***/
> +
> +static void
> +LIST_INIT(PyGCInfo *list)

But since they are not macros, the upper case is "wierd." I'd recommend
making them lower case, like the rest of Python functions.

>...
> @@ -498,11 +528,14 @@
>  	PyObject *error_type, *error_value, *error_traceback;
>  	PyObject *del;
>  	static PyObject *delstr;
> +	extern long _Py_RefTotal;
> +#ifdef WITH_CYCLE_GC
> +	PyGC_Remove((PyObject *)inst);
> +#endif

Same on PyGC_Remove() ... make it a no-op macro when WITH_CYCLE_GC is not
defined.

>...
> -		im = PyObject_NEW(PyMethodObject, &PyMethod_Type);
> +		im = PyObject_New(PyMethodObject, &PyMethod_Type);

Can these changes be broken out into a separate patch? I'm guessing that
this can be done *regardless* of the GC work, and that it would be an
improvement to things. If that is the case, then let's see a separate
patch for JUST this and get it applied ASAP. Could reduce Vladimir's patch
size, too.

I'm a big fan of avoiding comingling of patch concepts...

>...
> @@ -489,7 +495,7 @@
>  		}
>  	}
>  	PyMem_XDEL(mp->ma_table);
> -	PyMem_DEL(mp);
> +	PyObject_Del((PyObject *)mp);
>  	Py_TRASHCAN_SAFE_END(mp)
>  }

Same logic applies here. Let's see this change in a basic patch that the
GC patches can THEN build upon.

By getting small pieces, each individually reviewable, each individually
applicable, then we have a larger hope of a successful review of these
patches. As it is, I'm only doing a cursory review over this stuff because
the dumb thing is so big...

>...
> --- Python-cvs/Objects/object.c	Tue Mar 28 08:11:09 2000
> +++ Python-gc/Objects/object.c	Sat Apr  8 02:10:06 2000
> @@ -107,20 +107,12 @@
>  }
>  #endif
>  
> -#ifndef MS_COREDLL
> -PyObject *
> -_PyObject_New(tp)
> -	PyTypeObject *tp;
> -#else
> +
>  PyObject *
> -_PyObject_New(tp,op)
> +_PyObject_FromType(tp,op)
>  	PyTypeObject *tp;
>  	PyObject *op;
> -#endif
>  {
> -#ifndef MS_COREDLL
> -	PyObject *op = (PyObject *) malloc(tp->tp_basicsize);
> -#endif
>  	if (op == NULL)
>  		return PyErr_NoMemory();
>  	op->ob_type = tp;
> @@ -128,29 +120,78 @@
>  	return op;
>  }
>  
> -#ifndef MS_COREDLL
> -PyVarObject *
> -_PyObject_NewVar(tp, size)
> -	PyTypeObject *tp;
> -	int size;
> -#else
>  PyVarObject *
> -_PyObject_NewVar(tp, size, op)
> +_PyObject_VarFromType(tp, size, op)
>  	PyTypeObject *tp;
>  	int size;
>  	PyVarObject *op;
> -#endif
>  {
> -#ifndef MS_COREDLL
> -	PyVarObject *op = (PyVarObject *)
> -		malloc(tp->tp_basicsize + size * tp->tp_itemsize);
> -#endif

Umm.... what is this stuff with removing MS_COREDLL? Is that right? Is
this part of the PyObject_New thing? Should it go into the separate patch?

>...
>  #if MAXSAVESIZE > 0
>  	if (size == 0) {
>  		free_tuples[0] = op;
> -		++num_free_tuples[0];

Woah! Did you start with an old version of this file? Maybe you should do
a "cvs update" before generating your diff, to make sure that CVS changes
are integrated into your local source.

>...
>  #if MAXSAVESIZE > 0
>  	if (newsize == 0 && free_tuples[0]) {
> -		num_free_tuples[0]--;
>  		sv = free_tuples[0];
>  		sv->ob_size = 0;

Another one here.


Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/





From cgw@fnal.gov  Wed Apr 26 15:07:11 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Wed, 26 Apr 2000 09:07:11 -0500 (CDT)
Subject: [Patches] Sightly clean-up cycle-gc patch
In-Reply-To: <Pine.LNX.4.10.10004260519550.25776-100000@nebula.lyra.org>
References: <14597.49537.611625.343576@buffalo.fnal.gov>
 <Pine.LNX.4.10.10004260519550.25776-100000@nebula.lyra.org>
Message-ID: <14598.63503.871276.183817@buffalo.fnal.gov>

Greg Stein writes:
 > 
 > >...
 > >  #if MAXSAVESIZE > 0
 > >  	if (size == 0) {
 > >  		free_tuples[0] = op;
 > > -		++num_free_tuples[0];
 > 
 > Woah! Did you start with an old version of this file? Maybe you should do
 > a "cvs update" before generating your diff, to make sure that CVS changes
 > are integrated into your local source.
 > >  	if (newsize == 0 && free_tuples[0]) {
 > > -		num_free_tuples[0]--;
 > >  		sv = free_tuples[0];
 > >  		sv->ob_size = 0;
 > 
 > Another one here.


The changes above were intentional, when I was generating my update of
Neil's patch I picked up this tiny change to tupleobject.c that I had
made in my local workspace.  I should have submitted just this change
as a separate patch, to avoid confusion.  Sorry.  But I realized that
the code that was handling num_free_tuples[0] was both (A) wrong and
(B) unneccessary.  For the free_tuples list, size=0 is a special case,
there is a unique empty tuple and although we keep handing out
references to it we never remove it from the free list.  So
num_free_tuples[0]-- is not correct.  In fact we don't need to keep
track of num_free_tuples[0] at all.

My apologies for mixing this change in with something unrelated
(Neil's GC).  

I agree with the other points made in your posting, just wanted to
clear up this small detail.

 > Cheers,
 > -g

Your watchful and observant eye is much appreciated!


From nascheme@enme.ucalgary.ca  Wed Apr 26 16:51:37 2000
From: nascheme@enme.ucalgary.ca (Neil Schemenauer)
Date: Wed, 26 Apr 2000 09:51:37 -0600
Subject: [Patches] Sightly clean-up cycle-gc patch
In-Reply-To: <Pine.LNX.4.10.10004260519550.25776-100000@nebula.lyra.org>; from gstein@lyra.org on Wed, Apr 26, 2000 at 05:38:11AM -0700
References: <14597.49537.611625.343576@buffalo.fnal.gov> <Pine.LNX.4.10.10004260519550.25776-100000@nebula.lyra.org>
Message-ID: <20000426095137.B13009@acs.ucalgary.ca>

Thanks for the comments Greg.  I had already made most of the
changes you suggested.  As per your suggestions, I have renamed
the macros from PY_* to Py_* and lowercased the list function
names.  The latest version is attached.


    Neil


Index: 0.3/Makefile.in
*** 0.3/Makefile.in Tue, 25 Apr 2000 17:33:19 -0600 nas (python/0_Makefile.i 1.1 644)
--- gc.6(w)/Makefile.in Tue, 25 Apr 2000 17:59:00 -0600 nas (python/0_Makefile.i 1.1.1.1 644)
***************
*** 401,407 ****
  		$(INSTALL_DATA) Modules/Makefile $(LIBPL)/Makefile
  		$(INSTALL_DATA) Modules/Setup $(LIBPL)/Setup
  		$(INSTALL_DATA) Modules/Setup.local $(LIBPL)/Setup.local
! 		$(INSTALL_DATA) Modules/Setup.thread $(LIBPL)/Setup.thread
  		$(INSTALL_PROGRAM) $(srcdir)/Modules/makesetup $(LIBPL)/makesetup
  		$(INSTALL_PROGRAM) $(srcdir)/install-sh $(LIBPL)/install-sh
  		$(INSTALL_DATA) $(srcdir)/Misc/Makefile.pre.in $(LIBPL)/Makefile.pre.in
--- 401,407 ----
  		$(INSTALL_DATA) Modules/Makefile $(LIBPL)/Makefile
  		$(INSTALL_DATA) Modules/Setup $(LIBPL)/Setup
  		$(INSTALL_DATA) Modules/Setup.local $(LIBPL)/Setup.local
! 		$(INSTALL_DATA) Modules/Setup.auto $(LIBPL)/Setup.auto
  		$(INSTALL_PROGRAM) $(srcdir)/Modules/makesetup $(LIBPL)/makesetup
  		$(INSTALL_PROGRAM) $(srcdir)/install-sh $(LIBPL)/install-sh
  		$(INSTALL_DATA) $(srcdir)/Misc/Makefile.pre.in $(LIBPL)/Makefile.pre.in
Index: 0.3/config.h.in
*** 0.3/config.h.in Tue, 25 Apr 2000 17:33:19 -0600 nas (python/3_config.h.i 1.1 644)
--- gc.6(w)/config.h.in Tue, 25 Apr 2000 17:59:00 -0600 nas (python/3_config.h.i 1.1.1.1 644)
***************
*** 207,212 ****
--- 207,215 ----
     (shared library plus accessory files). */
  #undef WITH_NEXT_FRAMEWORK
  
+ /* Define if you want cycle garbage collection */
+ #undef WITH_CYCLE_GC
+ 
  /* The number of bytes in an off_t. */
  #undef SIZEOF_OFF_T
  
Index: 0.3/configure.in
*** 0.3/configure.in Tue, 25 Apr 2000 17:33:19 -0600 nas (python/5_configure. 1.1 644)
--- gc.6(w)/configure.in Tue, 25 Apr 2000 17:59:00 -0600 nas (python/5_configure. 1.1.1.1 644)
***************
*** 1070,1079 ****
  fi],
  [AC_MSG_RESULT(no)])
  
  # generate output files
  AC_OUTPUT(Makefile \
   Objects/Makefile \
   Parser/Makefile \
   Python/Makefile \
   Modules/Makefile.pre \
!  Modules/Setup.thread)
--- 1070,1090 ----
  fi],
  [AC_MSG_RESULT(no)])
  
+ AC_SUBST(USE_GC_MODULE)
+ USE_GC_MODULE="#"
+ 
+ AC_MSG_CHECKING(for --with-cycle-gc)
+ AC_ARG_WITH(cycle-gc, [--with-cycle-gc           enable garbage collection], [
+ AC_MSG_RESULT($withval)
+ AC_DEFINE(WITH_CYCLE_GC)
+ USE_GC_MODULE=
+ ],
+ AC_MSG_RESULT(no))
+ 
  # generate output files
  AC_OUTPUT(Makefile \
   Objects/Makefile \
   Parser/Makefile \
   Python/Makefile \
   Modules/Makefile.pre \
!  Modules/Setup.auto)
Index: 0.3/Include/object.h
*** 0.3/Include/object.h Tue, 25 Apr 2000 17:33:19 -0600 nas (python/o/18_object.h 1.1 644)
--- gc.6(w)/Include/object.h Tue, 25 Apr 2000 17:59:00 -0600 nas (python/o/18_object.h 1.1.1.1 644)
***************
*** 145,150 ****
--- 145,152 ----
  typedef int (*getsegcountproc) Py_PROTO((PyObject *, int *));
  typedef int (*getcharbufferproc) Py_PROTO((PyObject *, int, const char **));
  typedef int (*objobjproc) Py_PROTO((PyObject *, PyObject *));
+ typedef int (*visitproc) Py_PROTO((PyObject *, void *));
+ typedef int (*recurseproc) Py_PROTO((PyObject *, visitproc, void *));
  
  typedef struct {
  	binaryfunc nb_add;
***************
*** 243,251 ****
  
  	char *tp_doc; /* Documentation string */
  
  	/* More spares */
- 	long tp_xxx5;
- 	long tp_xxx6;
  	long tp_xxx7;
  	long tp_xxx8;
  
--- 245,257 ----
  
  	char *tp_doc; /* Documentation string */
  
+ 	/* call function for all accessible objects */
+ 	recurseproc tp_recurse;
+ 	
+ 	/* delete references to contained objects */
+ 	inquiry tp_clear;
+ 
  	/* More spares */
  	long tp_xxx7;
  	long tp_xxx8;
  
***************
*** 318,323 ****
--- 324,332 ----
  
  /* PySequenceMethods contains sq_contains */
  #define Py_TPFLAGS_HAVE_SEQUENCE_IN (1L<<1)
+ 
+ /* Objects with a GC info prefix (allocated *before* the object itself!) */
+ #define Py_TPFLAGS_HAVE_GCINFO (1L<<2)
  
  #define Py_TPFLAGS_DEFAULT  (Py_TPFLAGS_HAVE_GETCHARBUFFER | \
                               Py_TPFLAGS_HAVE_SEQUENCE_IN)
Index: 0.3/Include/objimpl.h
*** 0.3/Include/objimpl.h Tue, 25 Apr 2000 17:33:19 -0600 nas (python/o/19_objimpl.h 1.1 644)
--- gc.6(w)/Include/objimpl.h Tue, 25 Apr 2000 20:38:07 -0600 nas (python/o/19_objimpl.h 1.1.1.2 644)
***************
*** 72,77 ****
--- 72,94 ----
  
  #endif /* MS_COREDLL */
  
+ /*** Garbage Collection Support ***/
+ 
+ /* Add the object into the container set */
+ extern DL_IMPORT(void) PyGC_Insert Py_PROTO((PyObject *));
+ 
+ /* Remove the object from the container set */
+ extern DL_IMPORT(void) PyGC_Remove Py_PROTO((PyObject *));
+ 
+ /* GC objects are prefixed with PyGCInfo */
+ extern DL_IMPORT(PyObject *) _PyGC_New Py_PROTO((PyTypeObject *));
+ extern DL_IMPORT(PyVarObject *) _PyGC_NewVar Py_PROTO((PyTypeObject *, int));
+ extern DL_IMPORT(PyObject *) PyGC_Realloc Py_PROTO((PyObject *, int));
+ extern DL_IMPORT(void) PyGC_Del Py_PROTO((PyObject *));
+ 
+ #define PyGC_NEW(type, typeobj) ((type *) _PyGC_New(typeobj))
+ #define PyGC_NEW_VAR(type, typeobj, n) ((type *) _PyGC_NewVar(typeobj, n))
+ 
  #ifdef __cplusplus
  }
  #endif
Index: 0.3/Misc/Makefile.pre.in
*** 0.3/Misc/Makefile.pre.in Tue, 25 Apr 2000 17:33:19 -0600 nas (python/B/45_Makefile.p 1.1 644)
--- gc.6(w)/Misc/Makefile.pre.in Tue, 25 Apr 2000 17:59:00 -0600 nas (python/B/45_Makefile.p 1.2 644)
***************
*** 168,174 ****
  MAKEFILE=	$(LIBPL)/Makefile
  CONFIGC=	$(LIBPL)/config.c
  CONFIGCIN=	$(LIBPL)/config.c.in
! SETUP=		$(LIBPL)/Setup.thread $(LIBPL)/Setup.local $(LIBPL)/Setup
  
  SYSLIBS=	$(LIBM) $(LIBC)
  
--- 168,174 ----
  MAKEFILE=	$(LIBPL)/Makefile
  CONFIGC=	$(LIBPL)/config.c
  CONFIGCIN=	$(LIBPL)/config.c.in
! SETUP=		$(LIBPL)/Setup.auto $(LIBPL)/Setup.local $(LIBPL)/Setup
  
  SYSLIBS=	$(LIBM) $(LIBC)
  
Index: 0.3/Modules/Makefile.pre.in
*** 0.3/Modules/Makefile.pre.in Tue, 25 Apr 2000 17:33:19 -0600 nas (python/C/17_Makefile.p 1.1 644)
--- gc.6(w)/Modules/Makefile.pre.in Tue, 25 Apr 2000 17:59:00 -0600 nas (python/C/17_Makefile.p 1.1.1.1 644)
***************
*** 147,156 ****
  # gets remade from scratch; this ensures to remove modules that are no
  # longer pertinent (but that were in a previous configuration).
  config.c Makefile: Makefile.pre config.c.in $(MAKESETUP)
! config.c Makefile: Setup.thread Setup Setup.local
  config.c Makefile:
  		-rm -f $(LIBRARY)
! 		$(SHELL) $(MAKESETUP) Setup.thread Setup.local Setup
  
  hassignal:
  		-rm -f hassignal
--- 147,156 ----
  # gets remade from scratch; this ensures to remove modules that are no
  # longer pertinent (but that were in a previous configuration).
  config.c Makefile: Makefile.pre config.c.in $(MAKESETUP)
! config.c Makefile: Setup.auto Setup Setup.local
  config.c Makefile:
  		-rm -f $(LIBRARY)
! 		$(SHELL) $(MAKESETUP) Setup.auto Setup.local Setup
  
  hassignal:
  		-rm -f hassignal
Index: 0.3/Modules/Setup.in
*** 0.3/Modules/Setup.in Tue, 25 Apr 2000 17:33:19 -0600 nas (python/C/18_Setup.in 1.1 644)
--- gc.6(w)/Modules/Setup.in Tue, 25 Apr 2000 17:59:00 -0600 nas (python/C/18_Setup.in 1.1.1.1 644)
***************
*** 100,106 ****
  GLHACK=-Dclear=__GLclear
  #gl glmodule.c cgensupport.c -I$(srcdir) $(GLHACK) -lgl -lX11
  
! # The thread module is now automatically enabled, see Setup.thread.
  
  # Pure module.  Cannot be linked dynamically.
  # -DWITH_QUANTIFY, -DWITH_PURIFY, or -DWITH_ALL_PURE
--- 100,106 ----
  GLHACK=-Dclear=__GLclear
  #gl glmodule.c cgensupport.c -I$(srcdir) $(GLHACK) -lgl -lX11
  
! # The thread module is now automatically enabled, see Setup.auto.
  
  # Pure module.  Cannot be linked dynamically.
  # -DWITH_QUANTIFY, -DWITH_PURIFY, or -DWITH_ALL_PURE
Index: 0.3/Modules/cPickle.c
*** 0.3/Modules/cPickle.c Tue, 25 Apr 2000 17:33:19 -0600 nas (python/C/27_cPickle.c 1.1 644)
--- gc.6(w)/Modules/cPickle.c Tue, 25 Apr 2000 17:59:00 -0600 nas (python/C/27_cPickle.c 1.1.1.1 644)
***************
*** 2885,2891 ****
                PyInstanceObject *inst;
  
                PyErr_Clear();
!               UNLESS (inst=PyObject_NEW(PyInstanceObject, &PyInstance_Type))
                  goto err;
                inst->in_class=(PyClassObject*)cls;
                Py_INCREF(cls);
--- 2885,2891 ----
                PyInstanceObject *inst;
  
                PyErr_Clear();
!               UNLESS (inst=PyGC_NEW(PyInstanceObject, &PyInstance_Type))
                  goto err;
                inst->in_class=(PyClassObject*)cls;
                Py_INCREF(cls);
***************
*** 2893,2899 ****
                  Py_DECREF(inst);
                  goto err;
                }
! 
                return (PyObject *)inst;
              }
            Py_DECREF(__getinitargs__);
--- 2893,2899 ----
                  Py_DECREF(inst);
                  goto err;
                }
!               PyGC_Insert((PyObject *)inst);
                return (PyObject *)inst;
              }
            Py_DECREF(__getinitargs__);
Index: 0.3/Modules/newmodule.c
*** 0.3/Modules/newmodule.c Tue, 25 Apr 2000 17:33:19 -0600 nas (python/D/13_newmodule. 1.1 644)
--- gc.6(w)/Modules/newmodule.c Tue, 25 Apr 2000 17:59:00 -0600 nas (python/D/13_newmodule. 1.1.1.1 644)
***************
*** 49,61 ****
  			      &PyClass_Type, &klass,
  			      &PyDict_Type, &dict))
  		return NULL;
! 	inst = PyObject_NEW(PyInstanceObject, &PyInstance_Type);
  	if (inst == NULL)
  		return NULL;
  	Py_INCREF(klass);
  	Py_INCREF(dict);
  	inst->in_class = (PyClassObject *)klass;
  	inst->in_dict = dict;
  	return (PyObject *)inst;
  }
  
--- 49,62 ----
  			      &PyClass_Type, &klass,
  			      &PyDict_Type, &dict))
  		return NULL;
! 	inst = PyGC_NEW(PyInstanceObject, &PyInstance_Type);
  	if (inst == NULL)
  		return NULL;
  	Py_INCREF(klass);
  	Py_INCREF(dict);
  	inst->in_class = (PyClassObject *)klass;
  	inst->in_dict = dict;
+ 	PyGC_Insert((PyObject *)inst);
  	return (PyObject *)inst;
  }
  
Index: 0.3/Objects/classobject.c
*** 0.3/Objects/classobject.c Tue, 25 Apr 2000 17:33:19 -0600 nas (python/E/16_classobjec 1.1 644)
--- gc.6(w)/Objects/classobject.c Tue, 25 Apr 2000 20:38:07 -0600 nas (python/E/16_classobjec 1.1.1.2 644)
***************
*** 110,116 ****
  		}
  		Py_INCREF(bases);
  	}
! 	op = PyObject_NEW(PyClassObject, &PyClass_Type);
  	if (op == NULL) {
  		Py_DECREF(bases);
  		return NULL;
--- 110,116 ----
  		}
  		Py_INCREF(bases);
  	}
! 	op = PyGC_NEW(PyClassObject, &PyClass_Type);
  	if (op == NULL) {
  		Py_DECREF(bases);
  		return NULL;
***************
*** 131,136 ****
--- 131,137 ----
  	Py_XINCREF(op->cl_getattr);
  	Py_XINCREF(op->cl_setattr);
  	Py_XINCREF(op->cl_delattr);
+ 	PyGC_Insert((PyObject *)op);
  	return (PyObject *) op;
  }
  
***************
*** 140,152 ****
  class_dealloc(op)
  	PyClassObject *op;
  {
  	Py_DECREF(op->cl_bases);
  	Py_DECREF(op->cl_dict);
  	Py_XDECREF(op->cl_name);
  	Py_XDECREF(op->cl_getattr);
  	Py_XDECREF(op->cl_setattr);
  	Py_XDECREF(op->cl_delattr);
! 	free((ANY *)op);
  }
  
  static PyObject *
--- 141,154 ----
  class_dealloc(op)
  	PyClassObject *op;
  {
+ 	PyGC_Remove((PyObject *)op);
  	Py_DECREF(op->cl_bases);
  	Py_DECREF(op->cl_dict);
  	Py_XDECREF(op->cl_name);
  	Py_XDECREF(op->cl_getattr);
  	Py_XDECREF(op->cl_setattr);
  	Py_XDECREF(op->cl_delattr);
! 	PyGC_Del((PyObject *)op);
  }
  
  static PyObject *
***************
*** 386,391 ****
--- 388,408 ----
  	return res;
  }
  
+ #ifdef WITH_CYCLE_GC
+ static int
+ class_recurse(PyClassObject *o, visitproc visit, void *closure)
+ {
+ 	if (o->cl_bases) visit(o->cl_bases, closure);
+ 	if (o->cl_dict) visit(o->cl_dict, closure);
+ 	if (o->cl_name) visit(o->cl_name, closure);
+ 	if (o->cl_getattr) visit(o->cl_getattr, closure);
+ 	if (o->cl_setattr) visit(o->cl_setattr, closure);
+ 	if (o->cl_delattr) visit(o->cl_delattr, closure);
+ 	return 1;
+ }
+ #endif /* WITH_CYCLE_GC */
+ 
+ 
  PyTypeObject PyClass_Type = {
  	PyObject_HEAD_INIT(&PyType_Type)
  	0,
***************
*** 406,411 ****
--- 423,434 ----
  	(reprfunc)class_str, /*tp_str*/
  	(getattrofunc)class_getattr, /*tp_getattro*/
  	(setattrofunc)class_setattr, /*tp_setattro*/
+ #ifdef WITH_CYCLE_GC
+ 	0,		/* tp_as_buffer */
+ 	Py_TPFLAGS_DEFAULT | Py_TPFLAGS_HAVE_GCINFO, /*tp_flags*/
+ 	0,		/* tp_doc */
+ 	(recurseproc)class_recurse,	/* tp_recurse */
+ #endif
  };
  
  int
***************
*** 444,455 ****
  		PyErr_BadInternalCall();
  		return NULL;
  	}
! 	inst = PyObject_NEW(PyInstanceObject, &PyInstance_Type);
  	if (inst == NULL)
  		return NULL;
  	Py_INCREF(class);
  	inst->in_class = (PyClassObject *)class;
  	inst->in_dict = PyDict_New();
  	if (inst->in_dict == NULL) {
  		Py_DECREF(inst);
  		return NULL;
--- 467,479 ----
  		PyErr_BadInternalCall();
  		return NULL;
  	}
! 	inst = PyGC_NEW(PyInstanceObject, &PyInstance_Type);
  	if (inst == NULL)
  		return NULL;
  	Py_INCREF(class);
  	inst->in_class = (PyClassObject *)class;
  	inst->in_dict = PyDict_New();
+ 	PyGC_Insert((PyObject *)inst);
  	if (inst->in_dict == NULL) {
  		Py_DECREF(inst);
  		return NULL;
***************
*** 498,508 ****
  	PyObject *error_type, *error_value, *error_traceback;
  	PyObject *del;
  	static PyObject *delstr;
  	/* Call the __del__ method if it exists.  First temporarily
  	   revive the object and save the current exception, if any. */
  #ifdef Py_TRACE_REFS
  	/* much too complicated if Py_TRACE_REFS defined */
- 	extern long _Py_RefTotal;
  	inst->ob_type = &PyInstance_Type;
  	_Py_NewReference((PyObject *)inst);
  	_Py_RefTotal--;		/* compensate for increment in NEWREF */
--- 522,533 ----
  	PyObject *error_type, *error_value, *error_traceback;
  	PyObject *del;
  	static PyObject *delstr;
+ 	extern long _Py_RefTotal;
+ 	PyGC_Remove((PyObject *)inst);
  	/* Call the __del__ method if it exists.  First temporarily
  	   revive the object and save the current exception, if any. */
  #ifdef Py_TRACE_REFS
  	/* much too complicated if Py_TRACE_REFS defined */
  	inst->ob_type = &PyInstance_Type;
  	_Py_NewReference((PyObject *)inst);
  	_Py_RefTotal--;		/* compensate for increment in NEWREF */
***************
*** 550,555 ****
--- 575,581 ----
  #ifdef COUNT_ALLOCS
  		inst->ob_type->tp_free--;
  #endif
+ 		PyGC_Insert((PyObject *)inst);
  		return; /* __del__ added a reference; don't delete now */
  	}
  #ifdef Py_TRACE_REFS
***************
*** 561,567 ****
  #endif /* Py_TRACE_REFS */
  	Py_DECREF(inst->in_class);
  	Py_XDECREF(inst->in_dict);
! 	free((ANY *)inst);
  }
  
  static PyObject *
--- 587,593 ----
  #endif /* Py_TRACE_REFS */
  	Py_DECREF(inst->in_class);
  	Py_XDECREF(inst->in_dict);
! 	PyGC_Del((PyObject *)inst);
  }
  
  static PyObject *
***************
*** 840,845 ****
--- 866,881 ----
  	return outcome;
  }
  
+ #ifdef WITH_CYCLE_GC
+ static int
+ instance_recurse(PyInstanceObject *o, visitproc visit, void *closure)
+ {
+ 	if (o->in_class) visit((PyObject *)(o->in_class), closure);
+ 	if (o->in_dict) visit(o->in_dict, closure);
+ 	return 1;
+ }
+ #endif /* WITH_CYCLE_GC */
+ 
  static PyObject *getitemstr, *setitemstr, *delitemstr, *lenstr;
  
  static int
***************
*** 1463,1469 ****
  	(getattrofunc)instance_getattr, /*tp_getattro*/
  	(setattrofunc)instance_setattr, /*tp_setattro*/
          0, /* tp_as_buffer */
! 	Py_TPFLAGS_DEFAULT, /*tp_flags */
  };
  
  
--- 1499,1511 ----
  	(getattrofunc)instance_getattr, /*tp_getattro*/
  	(setattrofunc)instance_setattr, /*tp_setattro*/
          0, /* tp_as_buffer */
! #ifndef WITH_CYCLE_GC
! 	Py_TPFLAGS_DEFAULT, /*tp_flags*/
! #else
! 	Py_TPFLAGS_DEFAULT | Py_TPFLAGS_HAVE_GCINFO, /*tp_flags*/
! 	0,		/* tp_doc */
! 	(recurseproc)instance_recurse,	/* tp_recurse */
! #endif
  };
  
  
Index: 0.3/Objects/dictobject.c
*** 0.3/Objects/dictobject.c Tue, 25 Apr 2000 17:33:19 -0600 nas (python/E/19_dictobject 1.1 644)
--- gc.6(w)/Objects/dictobject.c Tue, 25 Apr 2000 17:59:00 -0600 nas (python/E/19_dictobject 1.1.1.1 644)
***************
*** 121,127 ****
  		if (dummy == NULL)
  			return NULL;
  	}
! 	mp = PyObject_NEW(dictobject, &PyDict_Type);
  	if (mp == NULL)
  		return NULL;
  	mp->ma_size = 0;
--- 121,127 ----
  		if (dummy == NULL)
  			return NULL;
  	}
! 	mp = PyGC_NEW(dictobject, &PyDict_Type);
  	if (mp == NULL)
  		return NULL;
  	mp->ma_size = 0;
***************
*** 129,134 ****
--- 129,135 ----
  	mp->ma_table = NULL;
  	mp->ma_fill = 0;
  	mp->ma_used = 0;
+ 	PyGC_Insert((PyObject *)mp);
  	return (PyObject *)mp;
  }
  
***************
*** 479,484 ****
--- 480,486 ----
  {
  	register int i;
  	register dictentry *ep;
+ 	PyGC_Remove((PyObject *)mp);
  	Py_TRASHCAN_SAFE_BEGIN(mp)
  	for (i = 0, ep = mp->ma_table; i < mp->ma_size; i++, ep++) {
  		if (ep->me_key != NULL) {
***************
*** 489,495 ****
  		}
  	}
  	PyMem_XDEL(mp->ma_table);
! 	PyMem_DEL(mp);
  	Py_TRASHCAN_SAFE_END(mp)
  }
  
--- 491,497 ----
  		}
  	}
  	PyMem_XDEL(mp->ma_table);
! 	PyGC_Del((PyObject *)mp);
  	Py_TRASHCAN_SAFE_END(mp)
  }
  
***************
*** 1036,1041 ****
--- 1038,1070 ----
  	return Py_None;
  }
  
+ #ifdef WITH_CYCLE_GC
+ static int
+ dict_recurse(PyObject *op, visitproc visit, void *closure)
+ {
+ 	int i = 0;
+ 	PyObject *pk;
+ 	PyObject *pv;
+ 
+ 	while (PyDict_Next(op, &i, &pk, &pv)) {
+ 		if (!visit(pk, closure)) {
+ 			return 0;
+ 		}
+ 		if (!visit(pv, closure)) {
+ 			return 0;
+ 		}
+ 	}
+ 	return 1;
+ }
+ 
+ static int
+ dict_gc_clear(PyObject *op)
+ {
+ 	PyDict_Clear(op);
+ 	return 0;
+ }
+ #endif /* WITH_CYCLE_GC */
+ 
  static PyMethodDef mapp_methods[] = {
  	{"has_key",	(PyCFunction)dict_has_key,      METH_VARARGS},
  	{"keys",	(PyCFunction)dict_keys},
***************
*** 1071,1076 ****
--- 1100,1117 ----
  	0,			/*tp_as_number*/
  	0,			/*tp_as_sequence*/
  	&dict_as_mapping,	/*tp_as_mapping*/
+ #ifdef WITH_CYCLE_GC
+ 	0,		/* tp_hash */
+ 	0,		/* tp_call */
+ 	0,		/* tp_str */
+ 	0,		/* tp_getattro */
+ 	0,		/* tp_setattro */
+ 	0,		/* tp_as_buffer */
+ 	Py_TPFLAGS_DEFAULT | Py_TPFLAGS_HAVE_GCINFO, /*tp_flags*/
+ 	0,		/* tp_doc */
+ 	(recurseproc)dict_recurse,	/* tp_recurse */
+ 	(inquiry)dict_gc_clear,		/* tp_clear */
+ #endif
  };
  
  /* For backward compatibility with old dictionary interface */
Index: 0.3/Objects/funcobject.c
*** 0.3/Objects/funcobject.c Tue, 25 Apr 2000 17:33:19 -0600 nas (python/E/23_funcobject 1.1 644)
--- gc.6(w)/Objects/funcobject.c Tue, 25 Apr 2000 17:59:00 -0600 nas (python/E/23_funcobject 1.2 644)
***************
*** 40,47 ****
  	PyObject *code;
  	PyObject *globals;
  {
! 	PyFunctionObject *op = PyObject_NEW(PyFunctionObject,
! 					    &PyFunction_Type);
  	if (op != NULL) {
  		PyObject *doc;
  		PyObject *consts;
--- 40,46 ----
  	PyObject *code;
  	PyObject *globals;
  {
! 	PyFunctionObject *op = PyGC_NEW(PyFunctionObject, &PyFunction_Type);
  	if (op != NULL) {
  		PyObject *doc;
  		PyObject *consts;
***************
*** 63,68 ****
--- 62,68 ----
  		Py_INCREF(doc);
  		op->func_doc = doc;
  	}
+ 	PyGC_Insert((PyObject *)op);
  	return (PyObject *)op;
  }
  
***************
*** 186,197 ****
  func_dealloc(op)
  	PyFunctionObject *op;
  {
  	Py_DECREF(op->func_code);
  	Py_DECREF(op->func_globals);
  	Py_DECREF(op->func_name);
  	Py_XDECREF(op->func_defaults);
  	Py_XDECREF(op->func_doc);
! 	PyMem_DEL(op);
  }
  
  static PyObject*
--- 186,198 ----
  func_dealloc(op)
  	PyFunctionObject *op;
  {
+ 	PyGC_Remove((PyObject *)op);
  	Py_DECREF(op->func_code);
  	Py_DECREF(op->func_globals);
  	Py_DECREF(op->func_name);
  	Py_XDECREF(op->func_defaults);
  	Py_XDECREF(op->func_doc);
! 	PyGC_Del((PyObject *)op);
  }
  
  static PyObject*
***************
*** 239,244 ****
--- 240,258 ----
  	return h;
  }
  
+ #ifdef WITH_CYCLE_GC
+ static int
+ func_recurse(PyFunctionObject *f, visitproc visit, void *closure)
+ {
+ 	if (f->func_code) visit(f->func_code, closure);
+ 	if (f->func_globals) visit(f->func_globals, closure);
+ 	if (f->func_defaults) visit(f->func_defaults, closure);
+ 	if (f->func_doc) visit(f->func_doc, closure);
+ 	if (f->func_name) visit(f->func_name, closure);
+ 	return 1;
+ }
+ #endif /* WITH_CYCLE_GC */
+ 
  PyTypeObject PyFunction_Type = {
  	PyObject_HEAD_INIT(&PyType_Type)
  	0,
***************
*** 255,258 ****
--- 269,282 ----
  	0,		/*tp_as_sequence*/
  	0,		/*tp_as_mapping*/
  	(hashfunc)func_hash, /*tp_hash*/
+ #ifdef WITH_CYCLE_GC
+ 	0,		/*tp_call*/
+ 	0,		/*tp_str*/
+ 	0,		/*tp_getattro*/
+ 	0,		/*tp_setattro*/
+ 	0,		/* tp_as_buffer */
+ 	Py_TPFLAGS_DEFAULT | Py_TPFLAGS_HAVE_GCINFO, /*tp_flags*/
+ 	0,		/* tp_doc */
+ 	(recurseproc)func_recurse,	/* tp_recurse */
+ #endif
  };
Index: 0.3/Objects/listobject.c
*** 0.3/Objects/listobject.c Tue, 25 Apr 2000 17:33:19 -0600 nas (python/E/25_listobject 1.1 644)
--- gc.6(w)/Objects/listobject.c Tue, 25 Apr 2000 17:59:00 -0600 nas (python/E/25_listobject 1.1.1.1 644)
***************
*** 61,94 ****
  	int i;
  	PyListObject *op;
  	size_t nbytes;
  	if (size < 0) {
  		PyErr_BadInternalCall();
  		return NULL;
  	}
! 	nbytes = size * sizeof(PyObject *);
! 	/* Check for overflow */
! 	if (nbytes / sizeof(PyObject *) != (size_t)size) {
! 		return PyErr_NoMemory();
! 	}
! 	op = (PyListObject *) malloc(sizeof(PyListObject));
! 	if (op == NULL) {
! 		return PyErr_NoMemory();
! 	}
! 	if (size <= 0) {
! 		op->ob_item = NULL;
! 	}
  	else {
! 		op->ob_item = (PyObject **) malloc(nbytes);
! 		if (op->ob_item == NULL) {
! 			free((ANY *)op);
  			return PyErr_NoMemory();
  		}
  	}
- 	op->ob_type = &PyList_Type;
  	op->ob_size = size;
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
! 	_Py_NewReference((PyObject *)op);
  	return (PyObject *) op;
  }
  
--- 61,95 ----
  	int i;
  	PyListObject *op;
  	size_t nbytes;
+ 	PyObject **item;
+ 
  	if (size < 0) {
  		PyErr_BadInternalCall();
  		return NULL;
  	}
! 	if (size == 0)
! 		item = NULL;
  	else {
! 		nbytes = size * sizeof(PyObject *);
! 		/* Check for overflow */
! 		if (nbytes / sizeof(PyObject *) != (size_t)size) {
  			return PyErr_NoMemory();
  		}
+ 		item = (PyObject **) malloc(nbytes);
+ 		if (item == NULL)
+ 			return PyErr_NoMemory();
+ 	}
+ 	op = PyGC_NEW(PyListObject, &PyList_Type);
+ 	if (op == NULL) {
+ 		if (item != NULL)
+ 			free((ANY *)item);
+ 		return PyErr_NoMemory();
  	}
  	op->ob_size = size;
+ 	op->ob_item = item;
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
! 	PyGC_Insert((PyObject *)op);
  	return (PyObject *) op;
  }
  
***************
*** 215,220 ****
--- 216,222 ----
  	PyListObject *op;
  {
  	int i;
+ 	PyGC_Remove((PyObject *)op);
  	Py_TRASHCAN_SAFE_BEGIN(op)
  	if (op->ob_item != NULL) {
  		/* Do it backwards, for Christian Tismer.
***************
*** 227,233 ****
  		}
  		free((ANY *)op->ob_item);
  	}
! 	free((ANY *)op);
  	Py_TRASHCAN_SAFE_END(op)
  }
  
--- 229,235 ----
  		}
  		free((ANY *)op->ob_item);
  	}
! 	PyGC_Del((PyObject *)op);
  	Py_TRASHCAN_SAFE_END(op)
  }
  
***************
*** 1399,1404 ****
--- 1401,1429 ----
  	return NULL;
  }
  
+ #ifdef WITH_CYCLE_GC
+ static int
+ list_recurse(PyListObject *o, visitproc visit, void *closure)
+ {
+ 	int i;
+ 	PyObject *x;
+ 
+ 	for (i = o->ob_size; --i >= 0; ) {
+ 		x = o->ob_item[i];
+ 		if (x != NULL && !visit(x, closure))
+ 			return 0;
+ 	}
+ 	return 1;
+ }
+ 
+ static int
+ list_clear(PyListObject *lp)
+ {
+ 	(void) PyList_SetSlice((PyObject *)lp, 0, lp->ob_size, 0);
+ 	return 0;
+ }
+ #endif /* WITH_CYCLE_GC */
+ 
  static char append_doc[] =
  "L.append(object) -- append object to end";
  static char extend_doc[] =
***************
*** 1464,1469 ****
--- 1489,1506 ----
  	0,		/*tp_as_number*/
  	&list_as_sequence,	/*tp_as_sequence*/
  	0,		/*tp_as_mapping*/
+ #ifdef WITH_CYCLE_GC
+ 	0,		/* tp_hash */
+ 	0,		/* tp_call */
+ 	0,		/* tp_str */
+ 	0,		/* tp_getattro */
+ 	0,		/* tp_setattro */
+ 	0,		/* tp_as_buffer */
+ 	Py_TPFLAGS_DEFAULT | Py_TPFLAGS_HAVE_GCINFO,	/* tp_flags */
+ 	0,		/* tp_doc */
+ 	(recurseproc)list_recurse,	/* tp_recurse */
+ 	(inquiry)list_clear,	/* tp_clear */
+ #endif
  };
  
  
***************
*** 1532,1536 ****
--- 1569,1584 ----
  	0,		/*tp_as_number*/
  	&immutable_list_as_sequence,	/*tp_as_sequence*/
  	0,		/*tp_as_mapping*/
+ #ifdef WITH_CYCLE_GC
+ 	0,		/* tp_hash */
+ 	0,		/* tp_call */
+ 	0,		/* tp_str */
+ 	0,		/* tp_getattro */
+ 	0,		/* tp_setattro */
+ 	0,		/* tp_as_buffer */
+ 	Py_TPFLAGS_DEFAULT,	/* tp_flags */
+ 	0,		/* tp_doc */
+ 	(recurseproc)list_recurse,	/* tp_recurse */
+ #endif
  };
  
Index: 0.3/Objects/tupleobject.c
*** 0.3/Objects/tupleobject.c Tue, 25 Apr 2000 17:33:19 -0600 nas (python/E/33_tupleobjec 1.1 644)
--- gc.6(w)/Objects/tupleobject.c Tue, 25 Apr 2000 17:59:00 -0600 nas (python/E/33_tupleobjec 1.1.1.1 644)
***************
*** 84,89 ****
--- 84,90 ----
  		op->ob_type = &PyTuple_Type;
  		op->ob_size = size;
  #endif
+ 		_Py_NewReference((PyObject *)op);
  	}
  	else
  #endif
***************
*** 97,112 ****
  			return PyErr_NoMemory();
  		}
  		;
! 		op = (PyTupleObject *) malloc(nbytes);
  		if (op == NULL)
! 			return PyErr_NoMemory();
! 
! 		op->ob_type = &PyTuple_Type;
! 		op->ob_size = size;
  	}
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
! 	_Py_NewReference((PyObject *)op);
  #if MAXSAVESIZE > 0
  	if (size == 0) {
  		free_tuples[0] = op;
--- 98,110 ----
  			return PyErr_NoMemory();
  		}
  		;
! 		op = PyGC_NEW_VAR(PyTupleObject, &PyTuple_Type, size);
  		if (op == NULL)
! 			return NULL;
  	}
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
! 
  #if MAXSAVESIZE > 0
  	if (size == 0) {
  		free_tuples[0] = op;
***************
*** 114,119 ****
--- 112,118 ----
  		Py_INCREF(op);	/* extra INCREF so that this is never freed */
  	}
  #endif
+ 	PyGC_Insert((PyObject *)op);
  	return (PyObject *) op;
  }
  
***************
*** 179,184 ****
--- 178,184 ----
  {
  	register int i;
  	register int len =  op->ob_size;
+ 	PyGC_Remove((PyObject *)op);
  	Py_TRASHCAN_SAFE_BEGIN(op)
  	if (len > 0) {
  		i = len;
***************
*** 193,199 ****
  		}
  #endif
  	}
! 	free((ANY *)op);
  done:
  	Py_TRASHCAN_SAFE_END(op)
  }
--- 193,199 ----
  		}
  #endif
  	}
! 	PyGC_Del((PyObject *)op);
  done:
  	Py_TRASHCAN_SAFE_END(op)
  }
***************
*** 401,406 ****
--- 401,422 ----
  	return (PyObject *) np;
  }
  
+ #ifdef WITH_CYCLE_GC
+ static int
+ tuplerecurse(PyTupleObject *o, visitproc visit, void *closure)
+ {
+ 	int i;
+ 	PyObject *x;
+ 
+ 	for (i = o->ob_size; --i >= 0; ) {
+ 		x = o->ob_item[i];
+ 		if (x != NULL && !visit(x, closure))
+ 			return 0;
+ 	}
+ 	return 1;
+ }
+ #endif /* WITH_CYCLE_GC */
+ 
  static PySequenceMethods tuple_as_sequence = {
  	(inquiry)tuplelength, /*sq_length*/
  	(binaryfunc)tupleconcat, /*sq_concat*/
***************
*** 427,432 ****
--- 443,458 ----
  	&tuple_as_sequence,	/*tp_as_sequence*/
  	0,		/*tp_as_mapping*/
  	(hashfunc)tuplehash, /*tp_hash*/
+ #ifdef WITH_CYCLE_GC
+ 	0,		/* tp_call */
+ 	0,		/* tp_str */
+ 	0,		/* tp_getattro */
+ 	0,		/* tp_setattro */
+ 	0,		/* tp_as_buffer */
+ 	Py_TPFLAGS_DEFAULT | Py_TPFLAGS_HAVE_GCINFO,	/* tp_flags */
+ 	0,		/* tp_doc */
+ 	(recurseproc)tuplerecurse,	/* tp_recurse */
+ #endif
  };
  
  /* The following function breaks the notion that tuples are immutable:
***************
*** 511,519 ****
  	} else 
  #endif		
  	{
! 		sv = (PyTupleObject *)
! 			realloc((char *)v,
! 				sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
  		*pv = (PyObject *) sv;
  		if (sv == NULL) {
  			PyMem_DEL(v);
--- 537,545 ----
  	} else 
  #endif		
  	{
! 		PyGC_Remove((PyObject *)v);
! 		sv = (PyTupleObject *)PyGC_Realloc((PyObject *)v,
! 			sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
  		*pv = (PyObject *) sv;
  		if (sv == NULL) {
  			PyMem_DEL(v);
***************
*** 522,527 ****
--- 548,554 ----
  		}
  	}
  	_Py_NewReference((PyObject *)sv);
+ 	PyGC_Insert((PyObject *)sv);
  	for (i = sv->ob_size; i < newsize; i++)
  		sv->ob_item[i] = NULL;
  	if (last_is_sticky && sizediff > 0) {
***************
*** 551,557 ****
  		while (p) {
  			q = p;
  			p = (PyTupleObject *)(p->ob_item[0]);
! 			PyMem_DEL(q);
  		}
  	}
  #endif
--- 578,584 ----
  		while (p) {
  			q = p;
  			p = (PyTupleObject *)(p->ob_item[0]);
! 			PyGC_Del((PyObject *)q);
  		}
  	}
  #endif
Index: 0.3/PC/config.c
*** 0.3/PC/config.c Tue, 25 Apr 2000 17:33:19 -0600 nas (python/E/39_config.c 1.1 644)
--- gc.6(w)/PC/config.c Tue, 25 Apr 2000 17:59:00 -0600 nas (python/E/39_config.c 1.1.1.1 644)
***************
*** 41,46 ****
--- 41,49 ----
  extern void initbinascii();
  extern void initcmath();
  extern void initerrno();
+ #ifdef WITH_CYCLE_GC
+ extern void initgc();
+ #endif
  extern void initimageop();
  extern void initmath();
  extern void initmd5();
***************
*** 79,84 ****
--- 82,90 ----
          {"binascii", initbinascii},
          {"cmath", initcmath},
          {"errno", initerrno},
+ #ifdef WITH_CYCLE_GC
+         {"gc", initgc},
+ #endif
          {"imageop", initimageop},
          {"math", initmath},
          {"md5", initmd5},
Index: 0.3/PC/config.h
*** 0.3/PC/config.h Tue, 25 Apr 2000 17:33:19 -0600 nas (python/E/40_config.h 1.1 644)
--- gc.6(w)/PC/config.h Tue, 25 Apr 2000 17:59:00 -0600 nas (python/E/40_config.h 1.1.1.1 644)
***************
*** 384,389 ****
--- 384,392 ----
  /* Define if you want to use the GNU readline library */
  /* #define WITH_READLINE 1 */
  
+ /* Define if you want cycle garbage collection */
+ #define WITH_CYCLE_GC 1
+ 
  /* Define if you have clock.  */
  /* #define HAVE_CLOCK */
  
Index: 0.3/PCbuild/python16.dsp
*** 0.3/PCbuild/python16.dsp Tue, 25 Apr 2000 17:33:19 -0600 nas (python/G/1_python16.d 1.1 644)
--- gc.6(w)/PCbuild/python16.dsp Tue, 25 Apr 2000 17:59:00 -0600 nas (python/G/1_python16.d 1.2 644)
***************
*** 645,650 ****
--- 645,665 ----
  # End Source File
  # Begin Source File
  
+ SOURCE=..\Modules\gcmodule.c
+ 
+ !IF  "$(CFG)" == "python16 - Win32 Release"
+ 
+ !ELSEIF  "$(CFG)" == "python16 - Win32 Debug"
+ 
+ !ELSEIF  "$(CFG)" == "python16 - Win32 Alpha Debug"
+ 
+ !ELSEIF  "$(CFG)" == "python16 - Win32 Alpha Release"
+ 
+ !ENDIF 
+ 
+ # End Source File
+ # Begin Source File
+ 
  SOURCE=..\Python\getargs.c
  
  !IF  "$(CFG)" == "python16 - Win32 Release"
Index: 0.3/Modules/gcmodule.c
*** 0.3/Modules/gcmodule.c Wed, 26 Apr 2000 09:48:07 -0600 nas ()
--- gc.6(w)/Modules/gcmodule.c Wed, 26 Apr 2000 09:39:29 -0600 nas (python/M/10_gcmodule.c 1.3 644)
***************
*** 0 ****
--- 1,771 ----
+ /*
+  
+   Reference Cycle Garbage Collection
+   ==================================
+ 
+   Neil Schemenauer <nascheme@enme.ucalgary.ca>
+ 
+   Based on a post on the python-dev list.  Ideas from Guido van Rossum,
+   Eric Tiedemann, and various others.
+ 
+   http://www.enme.calgary.ca/~nascheme/python/gc.html
+   http://www.python.org/pipermail/python-dev/2000-March/003869.html
+   http://www.python.org/pipermail/python-dev/2000-March/004010.html
+   http://www.python.org/pipermail/python-dev/2000-March/004022.html
+ 
+   For a highlevel view of the collection process, read the collect
+   function.
+ 
+   TODO:
+ 	use a different interface for set_debug() (keywords)?
+ 	inline PyGC_Insert and PyGC_Remove calls?
+ 	tune parameters
+ 
+ */
+ 
+ #include "Python.h"
+ 
+ #define DEBUG
+ 
+ #ifndef WITH_CYCLE_GC
+ #error "You must define WITH_CYCLE_GC to include this module"
+ #endif
+ 
+ /* Structure *prefixed* to container objects participating in GC */ 
+ typedef struct _gcinfo {
+ 	struct _gcinfo *gc_next;
+ 	struct _gcinfo *gc_prev;
+ 	int gc_refs;
+ } PyGCInfo;
+ 
+ /* Test if a type has GC info */
+ #define Py_TYPEISGC(t) PyType_HasFeature((t), Py_TPFLAGS_HAVE_GCINFO)
+ 
+ /* Test if an object has GC info */
+ #define Py_ISGC(o) Py_TYPEISGC((o)->ob_type)
+ 
+ /* Get an object's GC info -- NULL if the object has none */
+ #define Py_GCINFO(o) (Py_ISGC(o) ? ((PyGCInfo *)(o)-1) : NULL)
+ 
+ /* Unsafe version of Py_GCINFO() -- only call if Py_ISGC(p) is true */
+ #define Py_GCINFO_UNSAFE(o) ((PyGCInfo *)(o)-1)
+ 
+ /* Get the object given the PyGCInfo */
+ #define Py_GCOBJ(g) ((PyObject *)((g)+1))
+ 
+ /* magic gc_refs value */
+ #define GC_MOVED -1
+ 
+ /*** Global GC state ***/
+ 
+ /* linked lists of container objects */
+ static PyGCInfo generation0 = {&generation0, &generation0, 0};
+ static PyGCInfo generation1 = {&generation1, &generation1, 0};
+ static PyGCInfo generation2 = {&generation2, &generation2, 0};
+ static int generation = 0; /* current generation being collected */
+ 
+ /* collection frequencies, XXX tune these */
+ static int threshold0 = 100; /* net new containers before collection */
+ static int threshold1 = 10;  /* generation0 collections before collecting 1 */
+ static int threshold2 = 10;  /* generation1 collections before collecting 2 */
+ 
+ /* net new objects allocated since last collection */
+ static int allocated = 0;
+ 
+ /* set for debugging information */
+ #define DEBUG_STATS		(1<<0) /* print collection statistics */
+ #define DEBUG_COLLECTABLE	(1<<1) /* print collectable objects */
+ #define DEBUG_UNCOLLECTABLE	(1<<2) /* print uncollectable objects */
+ #define DEBUG_INSTANCES		(1<<3) /* print instances */
+ #define DEBUG_OBJECTS		(1<<4) /* print other objects */
+ #define DEBUG_LEAK		DEBUG_COLLECTABLE | \
+ 				DEBUG_UNCOLLECTABLE | \
+ 				DEBUG_INSTANCES | \
+ 				DEBUG_OBJECTS
+ static int debug = DEBUG_LEAK;
+ 
+ /* list of uncollectable objects */
+ static PyObject *garbage = NULL;
+ 
+ 
+ /*** list functions ***/
+ 
+ static void
+ list_init(PyGCInfo *list)
+ {
+ 	list->gc_prev = list;
+ 	list->gc_next = list;
+ }
+ 
+ static void
+ list_append(PyGCInfo *node, PyGCInfo *list)
+ {
+ 	node->gc_next = list;
+ 	node->gc_prev = list->gc_prev;
+ 	node->gc_prev->gc_next = node;
+ 	list->gc_prev = node;
+ }
+ 
+ static void
+ list_remove(PyGCInfo *node)
+ {
+ 	node->gc_prev->gc_next = node->gc_next;
+ 	node->gc_next->gc_prev = node->gc_prev;
+ #ifdef DEBUG
+ 	node->gc_next = NULL;
+ 	node->gc_prev = NULL;
+ #endif
+ }
+ 
+ static void 
+ list_move(PyGCInfo *from, PyGCInfo *to)
+ {
+ 	if (from->gc_next == from) {
+ 		/* empty from list */
+ 		list_init(to);
+ 	} else {
+ 		to->gc_next = from->gc_next;
+ 		to->gc_next->gc_prev = to;
+ 		to->gc_prev = from->gc_prev;
+ 		to->gc_prev->gc_next = to;
+ 	}
+ 	list_init(from);
+ }
+ 
+ /* append a list onto another list */
+ static void
+ list_lappend(PyGCInfo *from, PyGCInfo *to)
+ {
+ 	PyGCInfo *tail;
+ 	if (from->gc_next != from) {
+ 		tail = to->gc_prev;
+ 		tail->gc_next = from->gc_next;
+ 		tail->gc_next->gc_prev = tail;
+ 		to->gc_prev = from->gc_prev;
+ 		to->gc_prev->gc_next = to;
+ 	}
+ 	list_init(from);
+ }
+ 
+ static long
+ list_size(PyGCInfo *list)
+ {
+ 	PyGCInfo *gc;
+ 	long n = 0;
+ 	for (gc = list->gc_next; gc != list; gc = gc->gc_next) {
+ 		n++;
+ 	}
+ 	return n;
+ }
+ 
+ 
+ #ifdef DEBUG
+ static void
+ list_print(PyGCInfo *list) {
+ 	PyGCInfo *op = list;
+ 	do {
+ 		printf("%x %x %x %d\n", (long)op, (long)op->gc_prev, 
+ 				(long)op->gc_next, (long)op->gc_refs);
+ 
+ 		op = op->gc_next;
+ 	} while (op != list);
+ }
+ #endif
+ 
+ /*** end of list stuff ***/
+ 
+ 
+ /* Set all gc_refs = ob_refcnt */
+ static void
+ update_refs(PyGCInfo *containers)
+ {
+ 	PyGCInfo *gc = containers->gc_next;
+ 	for (; gc != containers; gc=gc->gc_next) {
+ 		gc->gc_refs = Py_GCOBJ(gc)->ob_refcnt;
+ 	}
+ }
+ 
+ static int
+ visit_decref(PyObject *op, void *data)
+ {
+ 	if (op && Py_ISGC(op)) {
+ 		Py_GCINFO_UNSAFE(op)->gc_refs--;
+ 	}
+ 	return 1;
+ }
+ 
+ /* Subtract internal references from gc_refs */
+ static void
+ subtract_refs(PyGCInfo *containers)
+ {
+ 	recurseproc recurse;
+ 	PyGCInfo *gc = containers->gc_next;
+ 	for (; gc != containers; gc=gc->gc_next) {
+ 		recurse = Py_GCOBJ(gc)->ob_type->tp_recurse;
+ 		(void) recurse(Py_GCOBJ(gc),
+ 			       (visitproc)visit_decref,
+ 			       NULL);
+ 	}
+ }
+ 
+ /* Append objects with gc_refs > 0 to roots list */
+ static void
+ move_roots(PyGCInfo *containers, PyGCInfo *roots)
+ {
+ 	PyGCInfo *next;
+ 	PyGCInfo *gc = containers->gc_next;
+ 	while (gc != containers) {
+ 		next = gc->gc_next;
+ 		if (gc->gc_refs > 0) {
+ 			list_remove(gc);
+ 			list_append(gc, roots);
+ 			gc->gc_refs = GC_MOVED;
+ 		}
+ 		gc = next;
+ 	}
+ }
+ 
+ static int
+ visit_reachable(PyObject *op, PyGCInfo *roots)
+ {
+ 	PyGCInfo *gc = Py_GCINFO(op);
+ 	if (gc && gc->gc_refs != GC_MOVED) {
+ 		list_remove(gc);
+ 		list_append(gc, roots);
+ 		gc->gc_refs = GC_MOVED;
+ 	}
+ 	return 1;
+ }
+ 
+ /* Move objects referenced from reachable to reachable set. */
+ static void
+ move_root_reachable(PyGCInfo *reachable)
+ {
+ 	recurseproc recurse;
+ 	PyGCInfo *gc = reachable->gc_next;
+ 	for (; gc != reachable; gc=gc->gc_next) {
+ 		/* careful, reachable list is growing here */
+ 		PyObject *op = Py_GCOBJ(gc);
+ 		recurse = op->ob_type->tp_recurse;
+ 		(void) recurse(op,
+ 			       (visitproc)visit_reachable,
+ 			       (void *)reachable);
+ 	}
+ }
+ 
+ /* move all objects with finalizers (instances with __del__) */
+ static void
+ move_finalizers(PyGCInfo *unreachable, PyGCInfo *finalizers)
+ {
+ 	PyGCInfo *next;
+ 	PyGCInfo *gc = unreachable->gc_next;
+ 	static PyObject *delstr;
+ 	if (delstr == NULL) {
+ 		delstr = PyString_InternFromString("__del__");
+ 	}
+ 	for (; gc != unreachable; gc=next) {
+ 		PyObject *op = Py_GCOBJ(gc);
+ 		next = gc->gc_next;
+ 		if (PyInstance_Check(op) && PyObject_HasAttr(op, delstr)) {
+ 			list_remove(gc);
+ 			list_append(gc, finalizers);
+ 		}
+ 	}
+ }
+ 
+ 
+ /* called by tp_recurse */
+ static int
+ visit_finalizer_reachable(PyObject *op, PyGCInfo *finalizers)
+ {
+ 	PyGCInfo *gc = Py_GCINFO(op);
+ 	if (gc && gc->gc_refs != GC_MOVED) {
+ 		list_remove(gc);
+ 		list_append(gc, finalizers);
+ 		gc->gc_refs = GC_MOVED;
+ 	}
+ 	return 1;
+ }
+ 
+ /* Move objects referenced from roots to roots */
+ static void
+ move_finalizer_reachable(PyGCInfo *finalizers)
+ {
+ 	recurseproc recurse;
+ 	PyGCInfo *gc = finalizers->gc_next;
+ 	for (; gc != finalizers; gc=gc->gc_next) {
+ 		/* careful, finalizers list is growing here */
+ 		recurse = Py_GCOBJ(gc)->ob_type->tp_recurse;
+ 		(void) recurse(Py_GCOBJ(gc), 
+ 			       (visitproc)visit_finalizer_reachable,
+ 			       (void *)finalizers);
+ 	}
+ }
+ 
+ static void
+ debug_instance(PyObject *output, char *msg, PyInstanceObject *inst)
+ {
+ 	char buf[200];
+ 	char *cname;
+ 	/* be careful not to create new dictionaries */
+ 	PyObject *classname = inst->in_class->cl_name;
+ 	if (classname != NULL && PyString_Check(classname))
+ 		cname = PyString_AsString(classname);
+ 	else
+ 		cname = "?";
+ 	sprintf(buf, "gc: %s<%.100s instance at %lx>\n", 
+ 			msg, cname, (long)inst);
+ 	PyFile_WriteString(buf, output);
+ }
+ 
+ static void
+ debug_cycle(PyObject *output, char *msg, PyObject *op)
+ {
+ 	if ((debug & DEBUG_INSTANCES) && PyInstance_Check(op)) {
+ 		debug_instance(output, msg, (PyInstanceObject *)op);
+ 	} else if (debug & DEBUG_OBJECTS) {
+ 		char buf[200];
+ 		sprintf(buf, "gc: %s<%s 0x%x>\n",
+ 				msg,
+ 				op->ob_type->tp_name,
+ 				(long)op);
+ 		PyFile_WriteString(buf, output);
+ 	}
+ }
+ 
+ /* Handle uncollectable garbage (cycles with finalizers). */
+ static void
+ handle_finalizers(PyGCInfo *finalizers, PyGCInfo *old)
+ {
+ 	PyGCInfo *gc;
+ 	if (garbage == NULL) {
+ 		garbage = PyList_New(0);
+ 	}
+ 	for (gc = finalizers->gc_next; gc != finalizers;
+ 			gc = finalizers->gc_next) {
+ 		PyObject *op = Py_GCOBJ(gc);
+ 		/* Add all instances to a Python accessible list of garbage */
+ 		if (PyInstance_Check(op)) {
+ 			PyList_Append(garbage, op);
+ 		}
+ 		/* We assume that all objects in finalizers are reachable from
+ 		 * instances.  Once we add the instances to the garbage list
+ 		 * everything is reachable from Python again. */
+ 		list_remove(gc);
+ 		list_append(gc, old);
+ 	}
+ }
+ 
+ /* Break reference cycles by clearing the containers involved.  This is
+  * tricky business as the lists can be changing and we don't know which
+  * objects may be freed.  It is possible I screwed something up here. */
+ static void
+ delete_garbage(PyGCInfo *unreachable, PyGCInfo *old)
+ {
+ 	inquiry clear;
+ 
+ 	while (unreachable->gc_next != unreachable) {
+ 		PyGCInfo *gc = unreachable->gc_next;
+ 		PyObject *op = Py_GCOBJ(gc);
+ 		/*
+ 		PyList_Append(garbage, op);
+ 		*/
+ 		if ((clear = op->ob_type->tp_clear) != NULL) {
+ 			Py_INCREF(op);
+ 			clear((PyObject *)op);
+ 			Py_DECREF(op);
+ 		}
+ 		/* only try to call tp_clear once for each object */
+ 		if (unreachable->gc_next == gc) {
+ 			/* still alive, move it, it may die later */
+ 			list_remove(gc);
+ 			list_append(gc, old);
+ 		}
+ 	}
+ }
+ 
+ /* This is the main function.  Read this to understand how the
+  * collection process works. */
+ static long
+ collect(PyGCInfo *young, PyGCInfo *old)
+ {
+ 	long n = 0;
+ 	long m = 0;
+ 	PyGCInfo reachable;
+ 	PyGCInfo unreachable;
+ 	PyGCInfo finalizers;
+ 	PyGCInfo *gc;
+ 	PyObject *output = NULL;
+ 
+ 	if (debug) {
+ 		output = PySys_GetObject("stderr");
+ 	}
+ 	if (debug & DEBUG_STATS) {
+ 		char buf[100];
+ 		sprintf(buf, "gc: collecting generation %d...\n", generation);
+ 		PyFile_WriteString(buf,output);
+ 		sprintf(buf, "gc: objects in each generation: %d %d %d\n",
+ 			list_size(&generation0),
+ 			list_size(&generation1),
+ 			list_size(&generation2));
+ 		PyFile_WriteString(buf,output);
+ 	}
+ 
+ 	/* Using ob_refcnt and gc_refs, calculate which objects in the
+ 	 * container set are reachable from outside the set (ie. have a
+ 	 * refcount greater than 0 when all the references within the
+ 	 * set are taken into account */
+ 	update_refs(young);
+ 	subtract_refs(young);
+ 
+ 	/* Move everything reachable from outside the set into the
+ 	 * reachable set (ie. gc_refs > 0).  Next, move everything
+ 	 * reachable from objects in the reachable set. */
+ 	list_init(&reachable);
+ 	move_roots(young, &reachable);
+ 	move_root_reachable(&reachable);
+ 
+ 	/* move unreachable objects to a temporary list, new objects can be
+ 	 * allocated after this point */
+ 	list_init(&unreachable);
+ 	list_move(young, &unreachable);
+ 
+ 	/* move reachable objects to next generation */
+ 	list_lappend(&reachable, old);
+ 
+ 	/* Move objects reachable from finalizers, we can't safely delete
+ 	 * them.  Python programmers should take care not to create such
+ 	 * things.  For Python finalizers means instance objects with
+ 	 * __del__ methods. */
+ 	list_init(&finalizers);
+ 	move_finalizers(&unreachable, &finalizers);
+ 	move_finalizer_reachable(&finalizers);
+ 
+ 	/* Collect statistics on collectable objects found and print
+ 	 * debugging information. */
+ 	for (gc = unreachable.gc_next; gc != &unreachable;
+ 			gc = gc->gc_next) {
+ 		m++;
+ 		if (output != NULL && (debug & DEBUG_COLLECTABLE)) {
+ 			debug_cycle(output, "collectable ", Py_GCOBJ(gc));
+ 		}
+ 	}
+ 	/* call tp_clear on objects in the collectable set.  This will cause
+ 	 * the reference cycles to be broken. It may also cause some objects in
+ 	 * finalizers to be freed */
+ 	delete_garbage(&unreachable, old);
+ 
+ 	/* Collect statistics on uncollectable objects found and print
+ 	 * debugging information. */
+ 	for (gc = finalizers.gc_next; gc != &finalizers;
+ 			gc = gc->gc_next) {
+ 		n++;
+ 		if (output != NULL && (debug & DEBUG_UNCOLLECTABLE)) {
+ 			debug_cycle(output, "uncollectable ", Py_GCOBJ(gc));
+ 		}
+ 	}
+ 	if (output != NULL && (debug & DEBUG_STATS)) {
+ 		if (m == 0 && n == 0) {
+ 			PyFile_WriteString("gc: done.\n", output);
+ 		} else {
+ 			char buf[200];
+ 			sprintf(buf,
+ 				"gc: done, %d unreachable, %d uncollectable.\n",
+ 				n+m, n);
+ 			PyFile_WriteString(buf, output);
+ 		}
+ 	}
+ 
+ 	/* Append instances in the uncollectable set to a Python
+ 	 * reachable list of garbage.  The programmer has to deal with
+ 	 * this if they insist on creating this type of structure. */
+ 	handle_finalizers(&finalizers, old);
+ 
+ 	allocated = 0;
+ 	PyErr_Clear(); /* in case writing to sys.stderr failed */
+ 	return n+m;
+ }
+ 
+ static long
+ collect_generations(void)
+ {
+ 	static long collections0 = 0;
+ 	static long collections1 = 0;
+ 	long n;
+ 
+ 
+ 	if (collections1 > threshold2) {
+ 		generation = 2;
+ 		list_lappend(&generation0, &generation2);
+ 		list_lappend(&generation1, &generation2);
+ 		if (generation2.gc_next != &generation2) {
+ 			n = collect(&generation2, &generation2);
+ 		}
+ 		collections1 = 0;
+ 	} else if (collections0 > threshold1) {
+ 		generation = 1;
+ 		collections1++;
+ 		list_lappend(&generation0, &generation1);
+ 		if (generation1.gc_next != &generation1) {
+ 			n = collect(&generation1, &generation2);
+ 		}
+ 		collections0 = 0;
+ 	} else {
+ 		generation = 0;
+ 		collections0++;
+ 		if (generation0.gc_next != &generation0) {
+ 			n = collect(&generation0, &generation1);
+ 		}
+ 	}
+ 	return n;
+ }
+ 
+ void
+ PyGC_Insert(PyObject *op)
+ {
+ 	/* collection lock since collecting may cause allocations */
+ 	static int collecting = 0;
+ 
+ #ifdef DEBUG
+ 	/*fprintf(stderr, "Insert 0x%x\n", op);*/
+ 	if (!Py_ISGC(op)) {
+ 		abort();
+ 	}
+ #endif
+ 	if (threshold0 && allocated > threshold0 && !collecting) {
+ 		collecting++;
+ 		collect_generations();
+ 		collecting--;
+ 	}
+ 	allocated++;
+ 	list_append(Py_GCINFO(op), &generation0);
+ }
+ 
+ void
+ PyGC_Remove(PyObject *op)
+ {
+ 	PyGCInfo *g = Py_GCINFO_UNSAFE(op);
+ #ifdef DEBUG
+ 	/*fprintf(stderr, "Remove 0x%x 0x%x\n", g, op);*/
+ #endif
+ 	if (allocated > 0) {
+ 		allocated--;
+ 	}
+ 	list_remove(g);
+ }
+ 
+ PyObject *
+ _PyGC_New(PyTypeObject *tp)
+ {
+ 	PyObject *op;
+ 	PyGCInfo *g = malloc(sizeof(*g) + tp->tp_basicsize);
+ 	if (g == NULL) {
+ 		return PyErr_NoMemory();
+ 	}
+ #ifdef DEBUG
+ 	g->gc_next = NULL;
+ 	g->gc_prev = NULL;
+ #endif
+ 	op = Py_GCOBJ(g);
+ 	op->ob_type = tp;
+ 	_Py_NewReference(op);
+ 	return op;
+ }
+ 
+ PyVarObject *
+ _PyGC_NewVar(PyTypeObject *tp, int size)
+ {
+ 	PyVarObject *op;
+ 	PyGCInfo *g = 
+ 	g = (PyGCInfo *) malloc(sizeof(*g) + 
+ 				tp->tp_basicsize +
+ 				size * tp->tp_itemsize);
+ 	if (g == NULL) {
+ 		return (PyVarObject *)PyErr_NoMemory();
+ 	}
+ #ifdef DEBUG
+ 	g->gc_next = NULL;
+ 	g->gc_prev = NULL;
+ #endif
+ 	op = (PyVarObject *)Py_GCOBJ(g);
+ 	op->ob_type = tp;
+ 	op->ob_size = size;
+ 	_Py_NewReference((PyObject *)op);
+ 	return op;
+ }
+ 
+ PyObject *
+ PyGC_Realloc(PyObject *op, int size)
+ {
+ 	PyGCInfo *g = Py_GCINFO(op);
+ 	g = realloc((char *)g, sizeof(*g) + size);
+ 	if (g == NULL) {
+ 		return NULL;
+ 	} else {
+ 		return Py_GCOBJ(g);
+ 	}
+ }
+ 
+ void
+ PyGC_Del(PyObject *op)
+ {
+ 	PyGCInfo *g = Py_GCINFO_UNSAFE(op);
+ 	free(g);
+ }
+ 
+ static char collect__doc__[] =
+ "collect() -> n\n"
+ "\n"
+ "Run a full collection.  The number of unreachable objects is returned.\n"
+ ;
+ 
+ static PyObject *
+ Py_collect(self, args)
+ 	PyObject *self;
+ 	PyObject *args;
+ {
+ 	long n;
+ 
+ 	if(!PyArg_ParseTuple(args, ""))	/* check no args */
+ 		return NULL;
+ 
+ 	generation = 2;
+ 	list_lappend(&generation0, &generation2);
+ 	list_lappend(&generation1, &generation2);
+ 	n = collect(&generation2, &generation2);
+ 
+ 	return Py_BuildValue("i", n);
+ }
+ 
+ static char set_debug__doc__[] = 
+ "set_debug(flags) -> None\n"
+ "\n"
+ "Set the garbage collection debugging flags. Debugging information is\n"
+ "written to sys.stderr.\n"
+ "\n"
+ "flags is an integer and can have the following bits turned on:\n"
+ "\n"
+ "  DEBUG_STATS - Print statistics during collection.\n"
+ "  DEBUG_COLLECTABLE - Print collectable objects found.\n"
+ "  DEBUG_UNCOLLECTABLE - Print unreachable but uncollectable objects found.\n"
+ "  DEBUG_INSTANCES - Print instance objects.\n"
+ "  DEBUG_OBJECTS - Print objects other than instances.\n"
+ "  DEBUG_LEAK - Debug leaking programs (everything but STATS).\n"
+ ;
+ 
+ static PyObject *
+ Py_set_debug(self, args)
+ 	PyObject *self;
+ 	PyObject *args;
+ {
+ 	if (!PyArg_ParseTuple(args, "l", &debug))
+ 		return NULL;
+ 
+ 	Py_INCREF(Py_None);
+ 	return Py_None;
+ }
+ 
+ static char get_debug__doc__[] = 
+ "get_debug() -> flags\n"
+ "\n"
+ "Get the garbage collection debugging flags.\n"
+ ;
+ 
+ static PyObject *
+ Py_get_debug(self, args)
+ 	PyObject *self;
+ 	PyObject *args;
+ {
+ 	if(!PyArg_ParseTuple(args, ""))	/* no args */
+ 		return NULL;
+ 
+ 	return Py_BuildValue("i", debug);
+ }
+ 
+ static char set_thresh__doc__[] =
+ "set_threshold(threshold0, [threhold1, threshold2]) -> None\n"
+ "\n"
+ "Sets the collection thresholds.  Setting threshold0 to zero disables\n"
+ "collection.\n"
+ ;
+ 
+ static PyObject *
+ Py_set_thresh(self, args)
+ 	PyObject *self;
+ 	PyObject *args;
+ {
+ 	if (!PyArg_ParseTuple(args, "i|ii", &threshold0, 
+ 				&threshold1, &threshold2))
+ 		return NULL;
+ 
+ 	Py_INCREF(Py_None);
+ 	return Py_None;
+ }
+ 
+ static char get_thresh__doc__[] =
+ "get_threshold() -> (threshold0, threshold1, threshold2)\n"
+ "\n"
+ "Return the current collection thresholds\n"
+ ;
+ 
+ static PyObject *
+ Py_get_thresh(self, args)
+ 	PyObject *self;
+ 	PyObject *args;
+ {
+ 	if(!PyArg_ParseTuple(args, ""))	/* no args */
+ 		return NULL;
+ 
+ 	return Py_BuildValue("(iii)", threshold0, threshold1, threshold2);
+ }
+ 
+ 
+ static char gc__doc__ [] =
+ "This module provides access to the garbage collector for reference cycles.\n"
+ "\n"
+ "collect() -- Do a full collection right now.\n"
+ "set_debug() -- Set debugging flags.\n"
+ "get_debug() -- Get debugging flags.\n"
+ "set_threshold() -- Set the collection thresholds.\n"
+ "get_threshold() -- Return the current the collection thresholds.\n"
+ ;
+ 
+ static PyMethodDef GcMethods[] = {
+ 	{"set_debug",		Py_set_debug,  METH_VARARGS, set_debug__doc__},
+ 	{"get_debug",		Py_get_debug,  METH_VARARGS, get_debug__doc__},
+ 	{"set_threshold",	Py_set_thresh, METH_VARARGS, set_thresh__doc__},
+ 	{"get_threshold",	Py_get_thresh, METH_VARARGS, get_thresh__doc__},
+ 	{"collect",		Py_collect,    METH_VARARGS, collect__doc__},
+ 	{NULL,	NULL}		/* Sentinel */
+ };
+ 
+ void
+ initgc(void)
+ {
+ 	PyObject *m;
+ 	PyObject *d;
+ 
+ 	m = Py_InitModule4("gc",
+ 			      GcMethods,
+ 			      gc__doc__,
+ 			      NULL,
+ 			      PYTHON_API_VERSION);
+ 	d = PyModule_GetDict(m);
+ 	if (garbage == NULL) {
+ 		garbage = PyList_New(0);
+ 	}
+ 	PyDict_SetItemString(d, "garbage", garbage);
+ 	PyDict_SetItemString(d, "DEBUG_STATS",
+ 			PyInt_FromLong(DEBUG_STATS));
+ 	PyDict_SetItemString(d, "DEBUG_COLLECTABLE",
+ 			PyInt_FromLong(DEBUG_COLLECTABLE));
+ 	PyDict_SetItemString(d, "DEBUG_UNCOLLECTABLE",
+ 			PyInt_FromLong(DEBUG_UNCOLLECTABLE));
+ 	PyDict_SetItemString(d, "DEBUG_INSTANCES",
+ 			PyInt_FromLong(DEBUG_INSTANCES));
+ 	PyDict_SetItemString(d, "DEBUG_OBJECTS",
+ 			PyInt_FromLong(DEBUG_OBJECTS));
+ 	PyDict_SetItemString(d, "DEBUG_LEAK",
+ 			PyInt_FromLong(DEBUG_LEAK));
+ }
+ 
Index: 0.3/Modules/Setup.auto.in
*** 0.3/Modules/Setup.auto.in Wed, 26 Apr 2000 09:48:07 -0600 nas ()
--- gc.6(w)/Modules/Setup.auto.in Tue, 25 Apr 2000 17:59:00 -0600 nas (python/M/11_Setup.auto 1.1 644)
***************
*** 0 ****
--- 1,16 ----
+ # This file is transmogrified into Setup.auto by config.status.  Its
+ # purpose is to automatically enable modules given when certain
+ # arguments are given to the configure script.  It replaces the old
+ # Setup.thread.in file.
+ 
+ # Include the thread module when the --with-thread argument is given to
+ # the configure script.
+ #
+ # *NOTE*: if the configure script decides it can't support threads, the
+ # thread module will still be enabled and cause compile errors.  The
+ # solution is not to use --with-thread on platforms that don't support
+ # threads.
+ @USE_THREAD_MODULE@thread threadmodule.c
+ 
+ # Garbage collection enabled with --with-cycle-gc
+ @USE_GC_MODULE@gc gcmodule.c
Index: 0.3/Lib/test/test_gc.py
*** 0.3/Lib/test/test_gc.py Wed, 26 Apr 2000 09:48:07 -0600 nas ()
--- gc.6(w)/Lib/test/test_gc.py Tue, 25 Apr 2000 17:59:00 -0600 nas (python/M/12_test_gc.py 1.1 644)
***************
*** 0 ****
--- 1,90 ----
+ import gc
+ 
+ def test_list():
+     l = []
+     l.append(l)
+     print 'list 0x%x' % id(l)
+     gc.collect()
+     del l
+     assert gc.collect() == 1
+ 
+ def test_dict():
+     d = {}
+     d[1] = d
+     print 'dict 0x%x' % id(d)
+     gc.collect()
+     del d
+     assert gc.collect() == 1
+ 
+ def test_tuple():
+     l = []
+     t = (l,)
+     l.append(t)
+     print 'list 0x%x' % id(l)
+     print 'tuple 0x%x' % id(t)
+     gc.collect()
+     del t
+     del l
+     assert gc.collect() == 2
+ 
+ def test_class():
+     class A:
+         pass
+     A.a = A
+     print 'class 0x%x' % id(A)
+     gc.collect()
+     del A
+     assert gc.collect() > 0
+ 
+ def test_instance():
+     class A:
+         pass
+     a = A()
+     a.a = a
+     print repr(a)
+     gc.collect()
+     del a
+     assert gc.collect() > 0
+ 
+ def test_finalizer():
+     class A:
+         def __del__(self): pass
+     class B:
+         pass
+     a = A()
+     a.a = a
+     id_a = id(a)
+     b = B()
+     b.b = b
+     print 'a', repr(a)
+     print 'b', repr(b)
+     gc.collect()
+     gc.garbage[:] = []
+     del a
+     del b
+     assert gc.collect() > 0
+     assert id(gc.garbage[0]) == id_a
+ 
+ def test_function():
+     d = {}
+     exec("def f(): pass\n") in d
+     print 'dict 0x%x' % id(d)
+     print 'func 0x%x' % id(d['f'])
+     gc.collect()
+     del d
+     assert gc.collect() == 2
+ 
+ 
+ def test_all():
+     debug = gc.get_debug()
+     gc.set_debug(gc.DEBUG_LEAK | gc.DEBUG_STATS)
+     test_list()
+     test_dict()
+     test_tuple()
+     test_class()
+     test_instance()
+     test_finalizer()
+     test_function()
+     gc.set_debug(debug)
+ 
+ test_all()


From cgw@fnal.gov  Wed Apr 26 16:51:35 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Wed, 26 Apr 2000 10:51:35 -0500 (CDT)
Subject: [Patches] GC puzzle
In-Reply-To: <20000425204246.A26980@acs.ucalgary.ca>
References: <14596.41644.182242.918286@buffalo.fnal.gov>
 <20000425105829.A27743@acs.ucalgary.ca>
 <14597.55168.966170.542056@buffalo.fnal.gov>
 <20000425113959.A28093@acs.ucalgary.ca>
 <14598.12230.933027.669754@buffalo.fnal.gov>
 <20000425184352.A25863@acs.ucalgary.ca>
 <14598.21582.383802.338822@buffalo.fnal.gov>
 <20000425204246.A26980@acs.ucalgary.ca>
Message-ID: <14599.4231.551732.269169@buffalo.fnal.gov>

I'm testing the GC patches with Py_TRACE_REFS enabled and have come
across something odd:

Python 1.6a2 (#3, Apr 26 2000, 10:35:57)  [GCC 2.95.2 19991024 (release)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import gc
[6007 refs]
>>> 0
0
[6011 refs]
>>> d1={}
[6015 refs]
>>> d2={}
[6019 refs]
>>> d1[0]=d2
[6021 refs]
>>> d2[0]=d1
[6023 refs]
>>> del d1
[6022 refs]
>>> del d2
[6021 refs]
>>> gc.collect()
gc: collectable <dictionary 0x81b4ac4>
gc: collectable <dictionary 0x81c4ee4>
2
[6017 refs]   ### note - before I created d1 and d2 the total refs was
              ### 6011 - are the extra 6 references internal to the
	      ### GVC module ?

>>> d1={}     ### Now I've created a new object and the total refs has
              ### not gone up
[6017 refs]
>>> d2={}     ### Now I've created another, and the total refs has
              ### gone up by 2!
[6019 refs]
>>> d1[0]=d2
[6021 refs]
>>> d2[0]=d1
[6023 refs]
>>> del d1
[6022 refs]
>>> del d2
[6021 refs]
>>> gc.collect()
gc: collectable <dictionary 0x81b4ac4>
gc: collectable <dictionary 0x81c4ee4>
2
[6017 refs]        ## After this round of GC, the total refs has gone
>>>                ## back to the value before d1 and d2 were created, 
                   ## as I'd expect...


From cgw@fnal.gov  Wed Apr 26 16:57:41 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Wed, 26 Apr 2000 10:57:41 -0500 (CDT)
Subject: [Patches] Re: Trashcan vs. GC vs. TRACE_REFS
In-Reply-To: <20000426155023.A25E41CD20@dinsdale.python.org>
References: <20000426155023.A25E41CD20@dinsdale.python.org>
Message-ID: <14599.4597.464952.855298@buffalo.fnal.gov>

Guido van Rossum wrote:

 > Mostly just to get all the details right (e.g. I never got
 > _PyResize_Tuple done, and Charles seems to have had problems
 > integrating that with his own patches to that piece of code...).

Well, just because I got tripped up on this at first, don't use that
as an argument against Neil's approach!  It was a bit subtle, but
sometimes things just are that way.

And, in the meanwhile, Neil has written PyGC_Realloc which makes the
realloc'ing of GC-objects much simpler...


From cgw@fnal.gov  Wed Apr 26 17:02:05 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Wed, 26 Apr 2000 11:02:05 -0500 (CDT)
Subject: [Patches] odd comment in tupleobject.h
Message-ID: <14599.4861.472431.654856@buffalo.fnal.gov>

According to tupleobject.h:

/*
Another generally useful object type is an tuple of object pointers.
This is a mutable type: the tuple items can be changed (but not their
number).  Out-of-range indices or non-tuple objects are ignored.
*/

Mutable?  Really?  

Or does this just mean that tuples are mutable from inside C code?


From Fredrik Lundh" <effbot@telia.com  Wed Apr 26 17:57:17 2000
From: Fredrik Lundh" <effbot@telia.com (Fredrik Lundh)
Date: Wed, 26 Apr 2000 18:57:17 +0200
Subject: [Patches] odd comment in tupleobject.h
References: <14599.4861.472431.654856@buffalo.fnal.gov>
Message-ID: <003701bfafa0$833e7580$34aab5d4@hagrid>

Charles G Waldman wrote:
> /*
> Another generally useful object type is an tuple of object pointers.
> This is a mutable type: the tuple items can be changed (but not their
> number).  Out-of-range indices or non-tuple objects are ignored.
> */
>=20
> Mutable?  Really? =20
>=20
> Or does this just mean that tuples are mutable from inside C code?

or that a tuple containing mutable objects is in fact mutable?

>>> a =3D ([], {})
>>> a[0].append(1)
>>> a[1]["spam"] =3D "egg"
>>> a
([1], {'spam': 'egg'})
>>>

</F>



From guido@python.org  Wed Apr 26 18:07:20 2000
From: guido@python.org (Guido van Rossum)
Date: Wed, 26 Apr 2000 13:07:20 -0400
Subject: [Patches] odd comment in tupleobject.h
In-Reply-To: Your message of "Wed, 26 Apr 2000 11:02:05 CDT."
 <14599.4861.472431.654856@buffalo.fnal.gov>
References: <14599.4861.472431.654856@buffalo.fnal.gov>
Message-ID: <200004261707.NAA06596@eric.cnri.reston.va.us>

> According to tupleobject.h:
> 
> /*
> Another generally useful object type is an tuple of object pointers.
> This is a mutable type: the tuple items can be changed (but not their
> number).  Out-of-range indices or non-tuple objects are ignored.
> */
> 
> Mutable?  Really?  
> 
> Or does this just mean that tuples are mutable from inside C code?

The latter.  It's a very old comment...

--Guido van Rossum (home page: http://www.python.org/~guido/)


From NadavH@envision.co.il  Wed Apr 26 20:44:54 2000
From: NadavH@envision.co.il (Nadav Horesh)
Date: Wed, 26 Apr 2000 21:44:54 +0200
Subject: [Patches] acosh and asinh functions
Message-ID: <39074736.B1C7D90F@envision.co.il>

This is a multi-part message in MIME format.
--------------30186664822EB8C794609C59
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

The computation of the inverse hyperbolic functions 'acosh' and 'asinh'
is erroneous -- it has both low precision and discontinuities.

The patch (using diff) is followed:
====================================================================
60,61c60,63
<  return c_log(c_sum(x,c_prod(c_i,
<       c_sqrt(c_diff(c_1,c_prod(x,x))))));
---
>   Py_complex z;
>   z = c_sqrt(c_half);
>   c_temp = c_log(c_prod(z, c_sum( c_sqrt(c_sum(x,c_1)),
c_sqrt(c_diff(x,c_1)))));
>   return c_sum(c_temp,c_temp);
89,90c91
<  z = c_diff(c_sqrt(z),x);
<  return c_neg(c_log(z));
---
>   return c_log(c_sum(c_sqrt(z),x));
338c339
<   PyErr_SetFromErrno(PyExc_ValueError);
---
>   PyErr_SetFromErrno(PyExc_ValueError);
414c415
<
---
>
422c423
< }
---
> }
\ No newline at end of file
=================================================================

I confirm that, to the best of my knowledge and belief, this
                  contribution is free of any claims of third parties
under
                  copyright, patent or other rights or interests
("claims").  To
                  the extent that I have any such claims, I hereby grant
to CNRI a
                  nonexclusive, irrevocable, royalty-free, worldwide
license to
                  reproduce, distribute, perform and/or display
publicly, prepare
                  derivative versions, and otherwise use this
contribution as part
                  of the Python software and its related documentation,
or any
                  derivative versions thereof, at no cost to CNRI or its
licensed
                  users, and to authorize others to do so.

                  I acknowledge that CNRI may, at its sole discretion,
decide
                  whether or not to incorporate this contribution in the
Python
                  software and its related documentation.  I further
grant CNRI
                  permission to use my name and other identifying
information
                  provided to CNRI by me for use in connection with the
Python
                  software and its related documentation.

--------------30186664822EB8C794609C59
Content-Type: text/x-vcard; charset=us-ascii;
 name="NadavH.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Nadav Horesh
Content-Disposition: attachment;
 filename="NadavH.vcf"

begin:vcard 
n:Horesh;Nadav
tel;fax:03 9244335
tel;home:03 9314679
tel;work:03 9244339
x-mozilla-html:TRUE
adr:;;;;;;
version:2.1
email;internet:NadavH@Envision.com
fn:Nadav Horesh
end:vcard

--------------30186664822EB8C794609C59--



From Jack.Jansen@oratrix.com  Wed Apr 26 21:39:08 2000
From: Jack.Jansen@oratrix.com (Jack Jansen)
Date: Wed, 26 Apr 2000 22:39:08 +0200
Subject: [Patches] Patch to timemodule, this time with patch:-)
Message-ID: <390753EC.9A47E229@oratrix.nl>

This is a multi-part message in MIME format.
--------------85E38E06D9657C4E356D2881
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Here it is again (I hope:-).

This patch is a workaround for Macintosh, where the GUSI I/O library (time,
stat, etc) use the MacOS epoch of 1-Jan-1904 and the MSL C library (ctime,
localtime, etc) uses the (apparently ANSI standard) epoch of 1-Jan-1900. Python
programs see the MacOS epoch and we convert values when needed.
--------------85E38E06D9657C4E356D2881
Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="522A6368";
 name="patch-timemodule"
Content-Transfer-Encoding: 7bit
Content-Description: Unknown Document
Content-Disposition: inline;
 filename="patch-timemodule"


Index: timemodule.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/timemodule.c,v
retrieving revision 2.81
diff -c -r2.81 timemodule.c
*** timemodule.c	2000/03/24 20:35:20	2.81
--- timemodule.c	2000/04/21 22:58:16
***************
*** 41,46 ****
--- 41,54 ----
  
  #ifdef macintosh
  #include <time.h>
+ #include <OSUtils.h>
+ #ifdef USE_GUSI2
+ /* GUSI, the I/O library which has the time() function and such uses the
+ ** Mac epoch of 1904. MSL, the C library which has localtime() and so uses
+ ** the ANSI epoch of 1900.
+ */
+ #define GUSI_TO_MSL_EPOCH (4*365*24*60*60)
+ #endif /* USE_GUSI2 */
  #else
  #include <sys/types.h>
  #endif
***************
*** 271,276 ****
--- 279,287 ----
  {
  	struct tm *p;
  	errno = 0;
+ #if defined(macintosh) && defined(USE_GUSI2)
+ 	when = when + GUSI_TO_MSL_EPOCH;
+ #endif
  	p = function(&when);
  	if (p == NULL) {
  #ifdef EINVAL
***************
*** 480,485 ****
--- 491,499 ----
  	if (!PyArg_Parse(args, "d", &dt))
  		return NULL;
  	tt = (time_t)dt;
+ #if defined(macintosh) && defined(USE_GUSI2)
+ 	tt = tt + GUSI_TO_MSL_EPOCH;
+ #endif
  	p = ctime(&tt);
  	if (p == NULL) {
  		PyErr_SetString(PyExc_ValueError, "unconvertible time");
***************
*** 517,522 ****
--- 531,539 ----
                                  "mktime argument out of range");
  		return NULL;
  	}
+ #if defined(macintosh) && defined(USE_GUSI2)
+ 	tt = tt - GUSI_TO_MSL_EPOCH;
+ #endif
  	return PyFloat_FromDouble((double)tt);
  }

--------------85E38E06D9657C4E356D2881--



From mal@lemburg.com  Thu Apr 27 17:17:34 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Thu, 27 Apr 2000 18:17:34 +0200
Subject: [Patches] Unicode Patch Set 2000-04-27
Message-ID: <3908681E.993B2FE0@lemburg.com>

This is a multi-part message in MIME format.
--------------473E418C86175FAFC293B440
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Patch Set Contents:
-------------------

Modules/_tkinter.c:

Fixes a memory leak found by Fredrik Lundh.

Objects/funcobject.c:

Doc strings can now be given as Unicode strings.

Objects/unicodeobject.c:

Fixed a reference leak in the allocator.

Renamed utf8_string to _PyUnicode_AsUTF8String() and made
it external for use by other parts of the interpreter.

Python/getargs.c:

Fixed a memory leak found by Fredrik Lundh. Instead
of PyUnicode_AsUTF8String() we now use _PyUnicode_AsUTF8String()
which returns the string object without incremented refcount
(and assures that the so obtained object remains alive until
the Unicode object is garbage collected).

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------473E418C86175FAFC293B440
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-27.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-27.patch"

Only in CVS-Python/Lib/test/output: test_winsound
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Modules/_tkinter.c Python+Unicode/Modules/_tkinter.c
--- CVS-Python/Modules/_tkinter.c	Fri Mar 31 05:29:39 2000
+++ Python+Unicode/Modules/_tkinter.c	Thu Apr 27 17:37:24 2000
@@ -553,8 +553,10 @@
 		PyObject* utf8 = PyUnicode_AsUTF8String (value);
 		if (!utf8)
 			return 0;
-		return Tcl_NewStringObj (PyString_AS_STRING (utf8),
+		result = Tcl_NewStringObj (PyString_AS_STRING (utf8),
 					 PyString_GET_SIZE (utf8));
+		Py_DECREF(utf8);
+		return result;
 	}
 	else {
 		PyObject *v = PyObject_Str(value);
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Objects/funcobject.c Python+Unicode/Objects/funcobject.c
--- CVS-Python/Objects/funcobject.c	Fri May 22 02:55:34 1998
+++ Python+Unicode/Objects/funcobject.c	Thu Apr 13 18:07:29 2000
@@ -55,7 +55,7 @@
 		consts = ((PyCodeObject *)code)->co_consts;
 		if (PyTuple_Size(consts) >= 1) {
 			doc = PyTuple_GetItem(consts, 0);
-			if (!PyString_Check(doc))
+			if (!PyString_Check(doc) && !PyUnicode_Check(doc))
 				doc = Py_None;
 		}
 		else
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Objects/unicodeobject.c Python+Unicode/Objects/unicodeobject.c
--- CVS-Python/Objects/unicodeobject.c	Thu Apr 13 11:11:39 2000
+++ Python+Unicode/Objects/unicodeobject.c	Thu Apr 27 17:51:23 2000
@@ -208,8 +208,7 @@
 	    if ((unicode->length < length) &&
 		_PyUnicode_Resize(unicode, length)) {
 		free(unicode->str);
-		PyMem_DEL(unicode);
-		return NULL;
+		goto onError;
 	    }
 	}
 	else
@@ -222,8 +221,10 @@
 	unicode->str = PyMem_NEW(Py_UNICODE, length + 1);
     }
 
-    if (!unicode->str) 
+    if (!unicode->str) {
+	PyErr_NoMemory();
 	goto onError;
+    }
     unicode->str[length] = 0;
     unicode->length = length;
     unicode->hash = -1;
@@ -233,7 +234,6 @@
  onError:
     _Py_ForgetReference((PyObject *)unicode);
     PyMem_DEL(unicode);
-    PyErr_NoMemory();
     return NULL;
 }
 
@@ -707,25 +707,27 @@
 
    The resulting string is cached in the Unicode object for subsequent
    usage by this function. The cached version is needed to implement
-   the character buffer interface.
+   the character buffer interface and will live (at least) as long as
+   the Unicode object itself.
 
    The refcount of the string is *not* incremented.
 
+   *** Exported for internal use by the interpreter only !!! ***
+
 */
 
-static
-PyObject *utf8_string(PyUnicodeObject *self,
+PyObject *_PyUnicode_AsUTF8String(PyObject *unicode,
 		      const char *errors)
 {
-    PyObject *v = self->utf8str;
+    PyObject *v = ((PyUnicodeObject *)unicode)->utf8str;
 
     if (v)
         return v;
-    v = PyUnicode_EncodeUTF8(PyUnicode_AS_UNICODE(self),
-			     PyUnicode_GET_SIZE(self),
+    v = PyUnicode_EncodeUTF8(PyUnicode_AS_UNICODE(unicode),
+			     PyUnicode_GET_SIZE(unicode),
 			     errors);
     if (v && errors == NULL)
-        self->utf8str = v;
+        ((PyUnicodeObject *)unicode)->utf8str = v;
     return v;
 }
 
@@ -737,7 +739,7 @@
         PyErr_BadArgument();
         return NULL;
     }
-    str = utf8_string((PyUnicodeObject *)unicode, NULL);
+    str = _PyUnicode_AsUTF8String(unicode, NULL);
     if (str == NULL)
         return NULL;
     Py_INCREF(str);
@@ -3183,7 +3185,7 @@
        on. */
     if (self->hash != -1)
 	return self->hash;
-    utf8 = utf8_string(self, NULL);
+    utf8 = _PyUnicode_AsUTF8String((PyObject *)self, NULL);
     if (utf8 == NULL)
 	return -1;
     hash = PyObject_Hash(utf8);
@@ -4087,7 +4089,7 @@
 			"accessing non-existent unicode segment");
         return -1;
     }
-    str = utf8_string(self, NULL);
+    str = _PyUnicode_AsUTF8String((PyObject *)self, NULL);
     if (str == NULL)
 	return -1;
     *ptr = (void *) PyString_AS_STRING(str);
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Python/getargs.c Python+Unicode/Python/getargs.c
--- CVS-Python/Python/getargs.c	Tue Mar 28 22:29:59 2000
+++ Python+Unicode/Python/getargs.c	Thu Apr 27 17:51:08 2000
@@ -444,6 +444,11 @@
 }
 
 
+/* Internal API needed by convertsimple1(): */
+extern 
+PyObject *_PyUnicode_AsUTF8String(PyObject *unicode,
+				  const char *errors);
+
 /* Convert a non-tuple argument.  Return NULL if conversion went OK,
    or a string representing the expected type if the conversion failed.
    When failing, an exception may or may not have been raised.
@@ -589,7 +594,7 @@
 			        if (PyString_Check(arg))
 				    *p = PyString_AS_STRING(arg);
 				else if (PyUnicode_Check(arg)) {
-				    arg = PyUnicode_AsUTF8String(arg);
+				    arg = _PyUnicode_AsUTF8String(arg, NULL);
 				    if (arg == NULL)
 					return "unicode conversion error";
 				    *p = PyString_AS_STRING(arg);
@@ -634,7 +639,7 @@
 				else if (PyString_Check(arg))
 				  *p = PyString_AsString(arg);
 				else if (PyUnicode_Check(arg)) {
-				  arg = PyUnicode_AsUTF8String(arg);
+				  arg = _PyUnicode_AsUTF8String(arg, NULL);
 				  if (arg == NULL)
 				      return "unicode conversion error";
 				  *p = PyString_AS_STRING(arg);
Only in CVS-Python/Lib/test: test_winsound.py

--------------473E418C86175FAFC293B440--




From mhammond@skippinet.com.au  Fri Apr 28 00:15:10 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Fri, 28 Apr 2000 09:15:10 +1000
Subject: [Patches] Change to reference-dump procedure at shutdown
In-Reply-To: <ECEPKNMJLHAPFFJHDOJBMEIOCJAA.mhammond@skippinet.com.au>
Message-ID: <ECEPKNMJLHAPFFJHDOJBKEMPCJAA.mhammond@skippinet.com.au>

I havent seen a response to this yet.  Would you like me to submit
again with the MS_WINDOWS #defines?

Mark.

> > > If this patch is acceptable and it is felt this
> > variable should be documented, let me know, and I will
> > try and hunt down all the relevant places - but IMO,
> > anyone who really needs these references dumped is likely
> > to see this code :-)
> >
> > Mark, shouldn't this be enclosed in an #ifdef
> MS_WINDOWS bracket?
>
> Quite possibly.  However, I was assuming that given there
> are over 2000 references outstanding when just starting
> and stopping the interpreter, the behaviour made sense
> for _all_ platforms, not just Windows.
>
> Of course, just getting it for Windows still suits me fine...



From guido@python.org  Fri Apr 28 00:17:04 2000
From: guido@python.org (Guido van Rossum)
Date: Thu, 27 Apr 2000 19:17:04 -0400
Subject: [Patches] Change to reference-dump procedure at shutdown
In-Reply-To: Your message of "Fri, 28 Apr 2000 09:15:10 +1000."
 <ECEPKNMJLHAPFFJHDOJBKEMPCJAA.mhammond@skippinet.com.au>
References: <ECEPKNMJLHAPFFJHDOJBKEMPCJAA.mhammond@skippinet.com.au>
Message-ID: <200004272317.TAA15007@eric.cnri.reston.va.us>

> I havent seen a response to this yet.  Would you like me to submit
> again with the MS_WINDOWS #defines?

Yes, please do.  My time is overbooked. :-(

--Guido van Rossum (home page: http://www.python.org/~guido/)


From cgw@fnal.gov  Fri Apr 28 00:29:28 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Thu, 27 Apr 2000 18:29:28 -0500 (CDT)
Subject: [Patches] nice speedup for compile.c
Message-ID: <14600.52568.898364.667776@buffalo.fnal.gov>

This implements an idea suggested in one of the /*XXX*/ comments in
compile.c

With this patch in place, the time required to import a 5000-line file
(with lots of constants in it) is reduced from 15 seconds to under 3
seconds on my system.


                   I confirm that, to the best of my knowledge and belief, this
                   contribution is free of any claims of third parties under
                   copyright, patent or other rights or interests ("claims").  To
                   the extent that I have any such claims, I hereby grant to CNRI a
                   nonexclusive, irrevocable, royalty-free, worldwide license to
                   reproduce, distribute, perform and/or display publicly, prepare
                   derivative versions, and otherwise use this contribution as part
                   of the Python software and its related documentation, or any
                   derivative versions thereof, at no cost to CNRI or its licensed
                   users, and to authorize others to do so.

                   I acknowledge that CNRI may, at its sole discretion, decide
                   whether or not to incorporate this contribution in the Python
                   software and its related documentation.  I further grant CNRI
                   permission to use my name and other identifying information
                   provided to CNRI by me for use in connection with the Python
                   software and its related documentation.

Index: Python/compile.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/compile.c,v
retrieving revision 2.105
diff -c -r2.105 compile.c
*** compile.c	2000/04/13 14:10:44	2.105
--- compile.c	2000/04/27 23:26:20
***************
*** 300,307 ****
  
  struct compiling {
  	PyObject *c_code;		/* string */
! 	PyObject *c_consts;	/* list of objects */
! 	PyObject *c_names;	/* list of strings (names) */
  	PyObject *c_globals;	/* dictionary (value=None) */
  	PyObject *c_locals;	/* dictionary (value=localID) */
  	PyObject *c_varnames;	/* list (inverse of c_locals) */
--- 300,309 ----
  
  struct compiling {
  	PyObject *c_code;		/* string */
! 	PyObject *c_consts;	/* dictionary of objects */
! 	PyObject *c_const_list; /* inverse of c_consts */
! 	PyObject *c_names;	/* dictionary of strings (names) */
! 	PyObject *c_name_list;  /* inverse of c_names */
  	PyObject *c_globals;	/* dictionary (value=None) */
  	PyObject *c_locals;	/* dictionary (value=localID) */
  	PyObject *c_varnames;	/* list (inverse of c_locals) */
***************
*** 403,409 ****
  static void com_addoparg Py_PROTO((struct compiling *, int, int));
  static void com_addfwref Py_PROTO((struct compiling *, int, int *));
  static void com_backpatch Py_PROTO((struct compiling *, int));
! static int com_add Py_PROTO((struct compiling *, PyObject *, PyObject *));
  static int com_addconst Py_PROTO((struct compiling *, PyObject *));
  static int com_addname Py_PROTO((struct compiling *, PyObject *));
  static void com_addopname Py_PROTO((struct compiling *, int, node *));
--- 405,411 ----
  static void com_addoparg Py_PROTO((struct compiling *, int, int));
  static void com_addfwref Py_PROTO((struct compiling *, int, int *));
  static void com_backpatch Py_PROTO((struct compiling *, int));
! static int com_add Py_PROTO((struct compiling *, PyObject *, PyObject *, PyObject *));
  static int com_addconst Py_PROTO((struct compiling *, PyObject *));
  static int com_addname Py_PROTO((struct compiling *, PyObject *));
  static void com_addopname Py_PROTO((struct compiling *, int, node *));
***************
*** 421,442 ****
  	struct compiling *c;
  	char *filename;
  {
  	if ((c->c_code = PyString_FromStringAndSize((char *)NULL,
  						    1000)) == NULL)
! 		goto fail_3;
! 	if ((c->c_consts = PyList_New(0)) == NULL)
! 		goto fail_2;
! 	if ((c->c_names = PyList_New(0)) == NULL)
! 		goto fail_1;
  	if ((c->c_globals = PyDict_New()) == NULL)
! 		goto fail_0;
  	if ((c->c_locals = PyDict_New()) == NULL)
! 		goto fail_00;
  	if ((c->c_varnames = PyList_New(0)) == NULL)
! 		goto fail_000;
  	if ((c->c_lnotab = PyString_FromStringAndSize((char *)NULL,
  						      1000)) == NULL)
! 		goto fail_0000;
  	c->c_nlocals = 0;
  	c->c_argcount = 0;
  	c->c_flags = 0;
--- 423,449 ----
  	struct compiling *c;
  	char *filename;
  {
+ 	memset((void *)c, '\0', sizeof(struct compiling));
  	if ((c->c_code = PyString_FromStringAndSize((char *)NULL,
  						    1000)) == NULL)
! 		goto fail;
! 	if ((c->c_consts = PyDict_New()) == NULL)
! 		goto fail;
! 	if ((c->c_const_list = PyList_New(0)) == NULL)
! 		goto fail;
! 	if ((c->c_names = PyDict_New()) == NULL)
! 		goto fail;
! 	if ((c->c_name_list = PyList_New(0)) == NULL)
! 		goto fail;
  	if ((c->c_globals = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_locals = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_varnames = PyList_New(0)) == NULL)
! 		goto fail;
  	if ((c->c_lnotab = PyString_FromStringAndSize((char *)NULL,
  						      1000)) == NULL)
! 		goto fail;
  	c->c_nlocals = 0;
  	c->c_argcount = 0;
  	c->c_flags = 0;
***************
*** 458,476 ****
  	c-> c_lnotab_next = 0;
  	return 1;
  	
!   fail_0000:
!   	Py_DECREF(c->c_varnames);
!   fail_000:
!   	Py_DECREF(c->c_locals);
!   fail_00:
!   	Py_DECREF(c->c_globals);
!   fail_0:
!   	Py_DECREF(c->c_names);
!   fail_1:
! 	Py_DECREF(c->c_consts);
!   fail_2:
! 	Py_DECREF(c->c_code);
!   fail_3:
   	return 0;
  }
  
--- 465,472 ----
  	c-> c_lnotab_next = 0;
  	return 1;
  	
!   fail:
! 	com_free(c);
   	return 0;
  }
  
***************
*** 480,486 ****
--- 476,484 ----
  {
  	Py_XDECREF(c->c_code);
  	Py_XDECREF(c->c_consts);
+ 	Py_XDECREF(c->c_const_list);
  	Py_XDECREF(c->c_names);
+ 	Py_XDECREF(c->c_name_list);
  	Py_XDECREF(c->c_globals);
  	Py_XDECREF(c->c_locals);
  	Py_XDECREF(c->c_varnames);
***************
*** 665,688 ****
  /* Handle literals and names uniformly */
  
  static int
! com_add(c, list, v)
  	struct compiling *c;
  	PyObject *list;
  	PyObject *v;
  {
! 	int n = PyList_Size(list);
! 	int i;
! 	/* XXX This is quadratic in the number of names per compilation unit.
! 	   XXX Should use a dictionary. */
! 	for (i = n; --i >= 0; ) {
! 		PyObject *w = PyList_GetItem(list, i);
! 		if (v->ob_type == w->ob_type && PyObject_Compare(v, w) == 0)
! 			return i;
  	}
! 	/* Check for error from PyObject_Compare */
! 	if (PyErr_Occurred() || PyList_Append(list, v) != 0)
! 		c->c_errors++;
  	return n;
  }
  
  static int
--- 663,706 ----
  /* Handle literals and names uniformly */
  
  static int
! com_add(c, dict, list, v)
  	struct compiling *c;
+ 	PyObject *dict;
  	PyObject *list;
  	PyObject *v;
  {
! 	PyObject *w, *t=NULL, *np=NULL;
! 	int n;
! 	
! 	t = PyTuple_New(2);
! 	if (t == NULL)
! 	    goto fail;
! 	PyTuple_SetItem(t, 0, v);
! 	Py_INCREF(v);
! 	PyTuple_SetItem(t, 1, (PyObject *)v->ob_type);
! 	Py_INCREF(v->ob_type);
! 	w = PyDict_GetItem(dict, t);
! 	if (PyErr_Occurred()) /* check for error from Py_ObjectCompare */
! 	    goto fail;
! 	if (w){
! 		n = PyInt_AsLong(w);
! 	} else {
! 		n = PyList_Size(list);
! 		if ((np = PyInt_FromLong((long)n)) == NULL)
! 		    goto fail;
! 		if (PyList_Append(list, v) != 0)
! 		    goto fail;
! 		if (PyDict_SetItem(dict, t, np) != 0)
! 		    goto fail;
! 		Py_DECREF(np);
  	}
! 	Py_DECREF(t);
  	return n;
+   fail:
+ 	Py_XDECREF(np);
+ 	Py_XDECREF(t);
+ 	c->c_errors++;
+ 	return 0;
  }
  
  static int
***************
*** 690,696 ****
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_consts, v);
  }
  
  static int
--- 708,714 ----
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_consts, c->c_const_list, v);
  }
  
  static int
***************
*** 698,704 ****
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_names, v);
  }
  
  #ifdef PRIVATE_NAME_MANGLING
--- 716,722 ----
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_names, c->c_name_list, v);
  }
  
  #ifdef PRIVATE_NAME_MANGLING
***************
*** 3373,3379 ****
  #define NEXTOP()	(*next_instr++)
  #define NEXTARG()	(next_instr += 2, (next_instr[-1]<<8) + next_instr[-2])
  #define GETITEM(v, i)	(PyList_GetItem((v), (i)))
! #define GETNAMEOBJ(i)	(GETITEM(c->c_names, (i)))
  	
  	PyErr_Fetch(&error_type, &error_value, &error_traceback);
  
--- 3391,3397 ----
  #define NEXTOP()	(*next_instr++)
  #define NEXTARG()	(next_instr += 2, (next_instr[-1]<<8) + next_instr[-2])
  #define GETITEM(v, i)	(PyList_GetItem((v), (i)))
! #define GETNAMEOBJ(i)	(GETITEM(c->c_name_list, (i)))
  	
  	PyErr_Fetch(&error_type, &error_value, &error_traceback);
  
***************
*** 3480,3487 ****
  	co = NULL;
  	if (sc.c_errors == 0) {
  		PyObject *consts, *names, *varnames, *filename, *name;
! 		consts = PyList_AsTuple(sc.c_consts);
! 		names = PyList_AsTuple(sc.c_names);
  		varnames = PyList_AsTuple(sc.c_varnames);
  		filename = PyString_InternFromString(sc.c_filename);
  		name = PyString_InternFromString(sc.c_name);
--- 3498,3505 ----
  	co = NULL;
  	if (sc.c_errors == 0) {
  		PyObject *consts, *names, *varnames, *filename, *name;
! 		consts = PyList_AsTuple(sc.c_const_list);
! 		names = PyList_AsTuple(sc.c_name_list);
  		varnames = PyList_AsTuple(sc.c_varnames);
  		filename = PyString_InternFromString(sc.c_filename);
  		name = PyString_InternFromString(sc.c_name);


From mhammond@skippinet.com.au  Fri Apr 28 00:33:21 2000
From: mhammond@skippinet.com.au (Mark Hammond)
Date: Fri, 28 Apr 2000 09:33:21 +1000
Subject: [Patches] Modified patch for refcount dumping procedure.
Message-ID: <ECEPKNMJLHAPFFJHDOJBOENACJAA.mhammond@skippinet.com.au>

This is a modification to my recent patch.  The new behaviour is now
only for Windows.

Checkin message:
 For Windows debug builds, we now only offer to dump remaining
object references if the environment variable PYTHONDUMPREFS exists.
The default behaviour caused problems for background or otherwise
invisible processes that use the debug build of Python.


diff -c -r2.92 pythonrun.c
*** pythonrun.c	2000/04/14 19:13:24	2.92
--- pythonrun.c	2000/04/27 23:29:26
***************
*** 242,248 ****
  #endif

  #ifdef Py_TRACE_REFS
! 	if (_Py_AskYesNo("Print left references?")) {
  		_Py_PrintReferences(stderr);
  	}
  #endif /* Py_TRACE_REFS */
--- 242,252 ----
  #endif

  #ifdef Py_TRACE_REFS
! 	if (
! #ifdef MS_WINDOWS /* Only ask on Windows if env var set */
! 	    getenv("PYTHONDUMPREFS") &&
! #endif /* MS_WINDOWS */
! 	    _Py_AskYesNo("Print left references?")) {
  		_Py_PrintReferences(stderr);
  	}
  #endif /* Py_TRACE_REFS */

Release info:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that
I have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce,
distribute, perform and/or display publicly, prepare derivative
versions, and otherwise use this contribution as part of the Python
software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize
others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether
or not to incorporate this contribution in the Python software and
its related documentation.  I further grant CNRI permission to use
my name and other identifying information provided to CNRI by me for
use in connection with the Python software and its related
documentation.

Mark.



From mal@lemburg.com  Fri Apr 28 10:19:45 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Fri, 28 Apr 2000 11:19:45 +0200
Subject: [Patches] nice speedup for compile.c
References: <14600.52568.898364.667776@buffalo.fnal.gov>
Message-ID: <390957B1.76F6F199@lemburg.com>

I haven't tried the patch, but from looking at it, I guess
it would be better to leave the old names (c_names and c_consts)
together with their previous semantics in place and
add new c_consts_dict and c_names_dict objects
to the end of the compiling structure.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From sjoerd@oratrix.nl  Fri Apr 28 11:35:54 2000
From: sjoerd@oratrix.nl (Sjoerd Mullender)
Date: Fri, 28 Apr 2000 12:35:54 +0200
Subject: [Patches] freeze patch
Message-ID: <20000428103555.B79E8301CF9@bireme.oratrix.nl>

cmp.py is obsolete...

Index: bkfile.py
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Tools/freeze/bkfile.py,v
retrieving revision 1.1
diff -u -r1.1 bkfile.py
--- bkfile.py	1998/08/25 14:06:51	1.1
+++ bkfile.py	2000/04/28 10:32:51
@@ -35,11 +35,8 @@
 		self.__file.close()
 		if self.__backup is None:
 			return
-		import cmp
-		# don't use cmp.cmp because of NFS bugs :-( and
-		# anyway, the stat mtime values differ so do_cmp will
-		# most likely be called anyway
-		if cmp.do_cmp(self.__backup, self.__filename):
+		import filecmp
+		if filecmp.cmp(self.__backup, self.__filename, shallow = 0):
 			import os
 			os.unlink(self.__filename)
 			os.rename(self.__backup, self.__filename)

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that I
have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce, distribute,
perform and/or display publicly, prepare derivative versions, and
otherwise use this contribution as part of the Python software and its
related documentation, or any derivative versions thereof, at no cost
to CNRI or its licensed users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or
not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.

-- Sjoerd Mullender <sjoerd.mullender@oratrix.com>


From take.vos@travelunie.nl  Fri Apr 28 14:23:32 2000
From: take.vos@travelunie.nl (Take Vos)
Date: Fri, 28 Apr 2000 15:23:32 +0200 (CEST)
Subject: [Patches] port to NCR MP-RAS (sys V unix)
Message-ID: <00Apr28.152457cest.116002@ds1.travelunie.nl>

This message is in MIME format
--_=XFMail.1.4.4.Linux:20000428152332:724=_
Content-Type: text/plain; charset=us-ascii


what it does:
        port to NCR MP-RAS (Sys V unix)

changes:
        Makefile.in
        Use the dgux style compiling, to create a shared library
        from libpython.a, as NCR MP-RAS dlopen() doesn't have
        access to the symbols in the executable, but it has
        access to the symbols which are loaded in from a shared library.
        I guess dgux, is very similar. unix_sv is the uname -s name of
        NCR MP-RAS.

        I also copied this libpython.so to the Modules directory, so
        it will be linked into the python executable without the leading
        ../ which confuses the runtime loader to find the library from the
        LD_LIBRARY_PATH (or the defaults path).
        These changes are also found in Modules/Makefile.pre.in

        configure
        I had to edit configure, I guess there has never been ported GNU
        software to MP-RAS :-), and I probably had to edit configure.in
        but I don't have the development tools for configure.
        Anyway, I copied the special dgux stuff for MP-RAS. Added the -Hnocopyr
        to the OPT variable, so the compiler won't print the anoying copyright
        line.
        Also when checking for -OPT:Olimit=0 and -Olimit 1500, I let it be
        cached to "no" state, becuase the compiler accepts this options but
        says, "Don't know this optimilization option, turning off
        optimilization"
        also shared libraries are build with ld -G.

        That was it and that was that.

        Sorry that I didn't patch it against the CVS tree, but I am under
        high time pressure.
        


The disclamer:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright, patent or
other rights or interests ("claims"). To the extent that I have any such
claims, I hereby grant to CNRI a nonexclusive, irrevocable, royalty-free,
worldwide license to reproduce, distribute, perform and/or display publicly,
prepare derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others to
do so. I acknowledge that CNRI may, at its sole discretion, decide whether or
not to incorporate this contribution in the Python software and its related
documentation. I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with the
Python software and its related documentation.
--
Take Vos <takev@binary-magic.com>

--_=XFMail.1.4.4.Linux:20000428152332:724=_
Content-Disposition: attachment; filename="Python-1.5.2-mpras.patch"
Content-Transfer-Encoding: base64
Content-Description: Python-1.5.2-mpras.patch
Content-Type: application/octet-stream;
 name=Python-1.5.2-mpras.patch; SizeOnDisk=4943
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--_=XFMail.1.4.4.Linux:20000428152332:724=_--
End of MIME message


From guido@python.org  Fri Apr 28 15:11:40 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 28 Apr 2000 10:11:40 -0400
Subject: [Patches] nice speedup for compile.c
In-Reply-To: Your message of "Fri, 28 Apr 2000 11:19:45 +0200."
 <390957B1.76F6F199@lemburg.com>
References: <14600.52568.898364.667776@buffalo.fnal.gov>
 <390957B1.76F6F199@lemburg.com>
Message-ID: <200004281411.KAA16128@eric.cnri.reston.va.us>

> I haven't tried the patch, but from looking at it, I guess
> it would be better to leave the old names (c_names and c_consts)
> together with their previous semantics in place and
> add new c_consts_dict and c_names_dict objects
> to the end of the compiling structure.

I concur.  Charles, would you mind resubmitting?

--Guido van Rossum (home page: http://www.python.org/~guido/)


From cgw@fnal.gov  Fri Apr 28 16:59:03 2000
From: cgw@fnal.gov (Charles G Waldman)
Date: Fri, 28 Apr 2000 10:59:03 -0500 (CDT)
Subject: [Patches] Resubmit speedup patch for compile.c
In-Reply-To: <200004281411.KAA16128@eric.cnri.reston.va.us>
References: <14600.52568.898364.667776@buffalo.fnal.gov>
 <390957B1.76F6F199@lemburg.com>
 <200004281411.KAA16128@eric.cnri.reston.va.us>
Message-ID: <14601.46407.32803.275261@buffalo.fnal.gov>

 mal wrote:
 > I haven't tried the patch, but from looking at it, I guess
 > it would be better to leave the old names (c_names and c_consts)
 > together with their previous semantics in place and
 > add new c_consts_dict and c_names_dict objects
 > to the end of the compiling structure.

Below find the reworked patch.  I don't completely agree with the
objections, but, whatever makes you and Guido happy is good with me.

I can see the objections to changing the semantics of existing
variables, however I don't understand why the new members need to be
added at the _end_ of the compiling structure.  To me this makes the
code less readable because it puts related items further away in the
source code.  The compiling structure is only used internally, so I
don't quite see why rearranging it is a problem.  But, whatever, I
don't want to argue.

Disclaimer:

                   I confirm that, to the best of my knowledge and belief, this
                   contribution is free of any claims of third parties under
                   copyright, patent or other rights or interests ("claims").  To
                   the extent that I have any such claims, I hereby grant to CNRI a
                   nonexclusive, irrevocable, royalty-free, worldwide license to
                   reproduce, distribute, perform and/or display publicly, prepare
                   derivative versions, and otherwise use this contribution as part
                   of the Python software and its related documentation, or any
                   derivative versions thereof, at no cost to CNRI or its licensed
                   users, and to authorize others to do so.

                   I acknowledge that CNRI may, at its sole discretion, decide
                   whether or not to incorporate this contribution in the Python
                   software and its related documentation.  I further grant CNRI
                   permission to use my name and other identifying information
                   provided to CNRI by me for use in connection with the Python
                   software and its related documentation.


Commit notice:

Follow a suggestion in an /*XXX*/ comment to speed up compilation by
using supplemental dictionaries to keep track of names and constants,
eliminating quadratic behavior.  With this patch in place, the time to
import a 5000-line file with lots of constants is reduced from 20
seconds to under 3 on my system.

Patch:
Index: Python/compile.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/compile.c,v
retrieving revision 2.105
diff -c -r2.105 compile.c
*** compile.c	2000/04/13 14:10:44	2.105
--- compile.c	2000/04/28 15:57:48
***************
*** 327,332 ****
--- 327,334 ----
  #ifdef PRIVATE_NAME_MANGLING
  	char *c_private;	/* for private name mangling */
  #endif
+ 	PyObject *c_const_dict; /* inverse of c_consts */
+ 	PyObject *c_name_dict;  /* inverse of c_names */
  };
  
  
***************
*** 403,409 ****
  static void com_addoparg Py_PROTO((struct compiling *, int, int));
  static void com_addfwref Py_PROTO((struct compiling *, int, int *));
  static void com_backpatch Py_PROTO((struct compiling *, int));
! static int com_add Py_PROTO((struct compiling *, PyObject *, PyObject *));
  static int com_addconst Py_PROTO((struct compiling *, PyObject *));
  static int com_addname Py_PROTO((struct compiling *, PyObject *));
  static void com_addopname Py_PROTO((struct compiling *, int, node *));
--- 405,411 ----
  static void com_addoparg Py_PROTO((struct compiling *, int, int));
  static void com_addfwref Py_PROTO((struct compiling *, int, int *));
  static void com_backpatch Py_PROTO((struct compiling *, int));
! static int com_add Py_PROTO((struct compiling *, PyObject *, PyObject *, PyObject *));
  static int com_addconst Py_PROTO((struct compiling *, PyObject *));
  static int com_addname Py_PROTO((struct compiling *, PyObject *));
  static void com_addopname Py_PROTO((struct compiling *, int, node *));
***************
*** 421,442 ****
  	struct compiling *c;
  	char *filename;
  {
  	if ((c->c_code = PyString_FromStringAndSize((char *)NULL,
  						    1000)) == NULL)
! 		goto fail_3;
  	if ((c->c_consts = PyList_New(0)) == NULL)
! 		goto fail_2;
  	if ((c->c_names = PyList_New(0)) == NULL)
! 		goto fail_1;
  	if ((c->c_globals = PyDict_New()) == NULL)
! 		goto fail_0;
  	if ((c->c_locals = PyDict_New()) == NULL)
! 		goto fail_00;
  	if ((c->c_varnames = PyList_New(0)) == NULL)
! 		goto fail_000;
  	if ((c->c_lnotab = PyString_FromStringAndSize((char *)NULL,
  						      1000)) == NULL)
! 		goto fail_0000;
  	c->c_nlocals = 0;
  	c->c_argcount = 0;
  	c->c_flags = 0;
--- 423,449 ----
  	struct compiling *c;
  	char *filename;
  {
+ 	memset((void *)c, '\0', sizeof(struct compiling));
  	if ((c->c_code = PyString_FromStringAndSize((char *)NULL,
  						    1000)) == NULL)
! 		goto fail;
  	if ((c->c_consts = PyList_New(0)) == NULL)
! 		goto fail;
! 	if ((c->c_const_dict = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_names = PyList_New(0)) == NULL)
! 		goto fail;
! 	if ((c->c_name_dict = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_globals = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_locals = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_varnames = PyList_New(0)) == NULL)
! 		goto fail;
  	if ((c->c_lnotab = PyString_FromStringAndSize((char *)NULL,
  						      1000)) == NULL)
! 		goto fail;
  	c->c_nlocals = 0;
  	c->c_argcount = 0;
  	c->c_flags = 0;
***************
*** 458,476 ****
  	c-> c_lnotab_next = 0;
  	return 1;
  	
!   fail_0000:
!   	Py_DECREF(c->c_varnames);
!   fail_000:
!   	Py_DECREF(c->c_locals);
!   fail_00:
!   	Py_DECREF(c->c_globals);
!   fail_0:
!   	Py_DECREF(c->c_names);
!   fail_1:
! 	Py_DECREF(c->c_consts);
!   fail_2:
! 	Py_DECREF(c->c_code);
!   fail_3:
   	return 0;
  }
  
--- 465,472 ----
  	c-> c_lnotab_next = 0;
  	return 1;
  	
!   fail:
! 	com_free(c);
   	return 0;
  }
  
***************
*** 480,486 ****
--- 476,484 ----
  {
  	Py_XDECREF(c->c_code);
  	Py_XDECREF(c->c_consts);
+ 	Py_XDECREF(c->c_const_dict);
  	Py_XDECREF(c->c_names);
+ 	Py_XDECREF(c->c_name_dict);
  	Py_XDECREF(c->c_globals);
  	Py_XDECREF(c->c_locals);
  	Py_XDECREF(c->c_varnames);
***************
*** 665,688 ****
  /* Handle literals and names uniformly */
  
  static int
! com_add(c, list, v)
  	struct compiling *c;
  	PyObject *list;
  	PyObject *v;
  {
! 	int n = PyList_Size(list);
! 	int i;
! 	/* XXX This is quadratic in the number of names per compilation unit.
! 	   XXX Should use a dictionary. */
! 	for (i = n; --i >= 0; ) {
! 		PyObject *w = PyList_GetItem(list, i);
! 		if (v->ob_type == w->ob_type && PyObject_Compare(v, w) == 0)
! 			return i;
! 	}
! 	/* Check for error from PyObject_Compare */
! 	if (PyErr_Occurred() || PyList_Append(list, v) != 0)
! 		c->c_errors++;
  	return n;
  }
  
  static int
--- 663,708 ----
  /* Handle literals and names uniformly */
  
  static int
! com_add(c, list, dict, v)
  	struct compiling *c;
  	PyObject *list;
+ 	PyObject *dict;
  	PyObject *v;
  {
! 	PyObject *w, *t=NULL, *np=NULL;
! 	int n;
! 	
! 	t = PyTuple_New(2);
! 	if (t == NULL)
! 	    goto fail;
! 	PyTuple_SetItem(t, 0, v);
! 	Py_INCREF(v);
! 	/*need to keep track of both value and type, else e.g. 1 and 1L 
! 	  are treated as the same constant */
! 	PyTuple_SetItem(t, 1, (PyObject *)v->ob_type); 
! 	Py_INCREF(v->ob_type);
! 	w = PyDict_GetItem(dict, t);
! 	if (PyErr_Occurred()) /* check for error from Py_ObjectCompare */
! 	    goto fail;
! 	if (w){
! 		n = PyInt_AsLong(w);
! 	} else {
! 		n = PyList_Size(list);
! 		if ((np = PyInt_FromLong((long)n)) == NULL)
! 		    goto fail;
! 		if (PyList_Append(list, v) != 0)
! 		    goto fail;
! 		if (PyDict_SetItem(dict, t, np) != 0)
! 		    goto fail;
! 		Py_DECREF(np);
! 	}
! 	Py_DECREF(t);
  	return n;
+   fail:
+ 	Py_XDECREF(np);
+ 	Py_XDECREF(t);
+ 	c->c_errors++;
+ 	return 0;
  }
  
  static int
***************
*** 690,696 ****
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_consts, v);
  }
  
  static int
--- 710,716 ----
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_consts, c->c_const_dict, v);
  }
  
  static int
***************
*** 698,704 ****
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_names, v);
  }
  
  #ifdef PRIVATE_NAME_MANGLING
--- 718,724 ----
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_names, c->c_name_dict, v);
  }
  
  #ifdef PRIVATE_NAME_MANGLING


From guido@python.org  Fri Apr 28 17:41:40 2000
From: guido@python.org (Guido van Rossum)
Date: Fri, 28 Apr 2000 12:41:40 -0400
Subject: [Patches] Resubmit speedup patch for compile.c
In-Reply-To: Your message of "Fri, 28 Apr 2000 10:59:03 CDT."
 <14601.46407.32803.275261@buffalo.fnal.gov>
References: <14600.52568.898364.667776@buffalo.fnal.gov> <390957B1.76F6F199@lemburg.com> <200004281411.KAA16128@eric.cnri.reston.va.us>
 <14601.46407.32803.275261@buffalo.fnal.gov>
Message-ID: <200004281641.MAA17321@eric.cnri.reston.va.us>

> Below find the reworked patch.  I don't completely agree with the
> objections, but, whatever makes you and Guido happy is good with me.
> 
> I can see the objections to changing the semantics of existing
> variables, however I don't understand why the new members need to be
> added at the _end_ of the compiling structure.  To me this makes the
> code less readable because it puts related items further away in the
> source code.  The compiling structure is only used internally, so I
> don't quite see why rearranging it is a problem.  But, whatever, I
> don't want to argue.

I do want to argue! :-)  You're right, those new members don't need to
be banned to the end.

I've also replaced your code to create a tuple with a call to
Py_BuildValue() -- much less code and doesn't seem to be any slower.

Here's the patch that will go in now.

Also thanks for changing the bizarre error handling in com_init() into
something sane!

Index: compile.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Python/compile.c,v
retrieving revision 2.105
diff -c -r2.105 compile.c
*** compile.c	2000/04/13 14:10:44	2.105
--- compile.c	2000/04/28 16:39:16
***************
*** 301,307 ****
--- 301,309 ----
  struct compiling {
  	PyObject *c_code;		/* string */
  	PyObject *c_consts;	/* list of objects */
+ 	PyObject *c_const_dict; /* inverse of c_consts */
  	PyObject *c_names;	/* list of strings (names) */
+ 	PyObject *c_name_dict;  /* inverse of c_names */
  	PyObject *c_globals;	/* dictionary (value=None) */
  	PyObject *c_locals;	/* dictionary (value=localID) */
  	PyObject *c_varnames;	/* list (inverse of c_locals) */
***************
*** 403,409 ****
  static void com_addoparg Py_PROTO((struct compiling *, int, int));
  static void com_addfwref Py_PROTO((struct compiling *, int, int *));
  static void com_backpatch Py_PROTO((struct compiling *, int));
! static int com_add Py_PROTO((struct compiling *, PyObject *, PyObject *));
  static int com_addconst Py_PROTO((struct compiling *, PyObject *));
  static int com_addname Py_PROTO((struct compiling *, PyObject *));
  static void com_addopname Py_PROTO((struct compiling *, int, node *));
--- 405,411 ----
  static void com_addoparg Py_PROTO((struct compiling *, int, int));
  static void com_addfwref Py_PROTO((struct compiling *, int, int *));
  static void com_backpatch Py_PROTO((struct compiling *, int));
! static int com_add Py_PROTO((struct compiling *, PyObject *, PyObject *, PyObject *));
  static int com_addconst Py_PROTO((struct compiling *, PyObject *));
  static int com_addname Py_PROTO((struct compiling *, PyObject *));
  static void com_addopname Py_PROTO((struct compiling *, int, node *));
***************
*** 421,442 ****
  	struct compiling *c;
  	char *filename;
  {
  	if ((c->c_code = PyString_FromStringAndSize((char *)NULL,
  						    1000)) == NULL)
! 		goto fail_3;
  	if ((c->c_consts = PyList_New(0)) == NULL)
! 		goto fail_2;
  	if ((c->c_names = PyList_New(0)) == NULL)
! 		goto fail_1;
  	if ((c->c_globals = PyDict_New()) == NULL)
! 		goto fail_0;
  	if ((c->c_locals = PyDict_New()) == NULL)
! 		goto fail_00;
  	if ((c->c_varnames = PyList_New(0)) == NULL)
! 		goto fail_000;
  	if ((c->c_lnotab = PyString_FromStringAndSize((char *)NULL,
  						      1000)) == NULL)
! 		goto fail_0000;
  	c->c_nlocals = 0;
  	c->c_argcount = 0;
  	c->c_flags = 0;
--- 423,449 ----
  	struct compiling *c;
  	char *filename;
  {
+ 	memset((void *)c, '\0', sizeof(struct compiling));
  	if ((c->c_code = PyString_FromStringAndSize((char *)NULL,
  						    1000)) == NULL)
! 		goto fail;
  	if ((c->c_consts = PyList_New(0)) == NULL)
! 		goto fail;
! 	if ((c->c_const_dict = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_names = PyList_New(0)) == NULL)
! 		goto fail;
! 	if ((c->c_name_dict = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_globals = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_locals = PyDict_New()) == NULL)
! 		goto fail;
  	if ((c->c_varnames = PyList_New(0)) == NULL)
! 		goto fail;
  	if ((c->c_lnotab = PyString_FromStringAndSize((char *)NULL,
  						      1000)) == NULL)
! 		goto fail;
  	c->c_nlocals = 0;
  	c->c_argcount = 0;
  	c->c_flags = 0;
***************
*** 458,476 ****
  	c-> c_lnotab_next = 0;
  	return 1;
  	
!   fail_0000:
!   	Py_DECREF(c->c_varnames);
!   fail_000:
!   	Py_DECREF(c->c_locals);
!   fail_00:
!   	Py_DECREF(c->c_globals);
!   fail_0:
!   	Py_DECREF(c->c_names);
!   fail_1:
! 	Py_DECREF(c->c_consts);
!   fail_2:
! 	Py_DECREF(c->c_code);
!   fail_3:
   	return 0;
  }
  
--- 465,472 ----
  	c-> c_lnotab_next = 0;
  	return 1;
  	
!   fail:
! 	com_free(c);
   	return 0;
  }
  
***************
*** 480,486 ****
--- 476,484 ----
  {
  	Py_XDECREF(c->c_code);
  	Py_XDECREF(c->c_consts);
+ 	Py_XDECREF(c->c_const_dict);
  	Py_XDECREF(c->c_names);
+ 	Py_XDECREF(c->c_name_dict);
  	Py_XDECREF(c->c_globals);
  	Py_XDECREF(c->c_locals);
  	Py_XDECREF(c->c_varnames);
***************
*** 665,688 ****
  /* Handle literals and names uniformly */
  
  static int
! com_add(c, list, v)
  	struct compiling *c;
  	PyObject *list;
  	PyObject *v;
  {
! 	int n = PyList_Size(list);
! 	int i;
! 	/* XXX This is quadratic in the number of names per compilation unit.
! 	   XXX Should use a dictionary. */
! 	for (i = n; --i >= 0; ) {
! 		PyObject *w = PyList_GetItem(list, i);
! 		if (v->ob_type == w->ob_type && PyObject_Compare(v, w) == 0)
! 			return i;
! 	}
! 	/* Check for error from PyObject_Compare */
! 	if (PyErr_Occurred() || PyList_Append(list, v) != 0)
! 		c->c_errors++;
  	return n;
  }
  
  static int
--- 663,701 ----
  /* Handle literals and names uniformly */
  
  static int
! com_add(c, list, dict, v)
  	struct compiling *c;
  	PyObject *list;
+ 	PyObject *dict;
  	PyObject *v;
  {
! 	PyObject *w, *t, *np=NULL;
! 	long n;
! 
! 	t = Py_BuildValue("(OO)", v, v->ob_type);
! 	if (t == NULL)
! 	    goto fail;
! 	w = PyDict_GetItem(dict, t);
! 	if (w != NULL) {
! 		n = PyInt_AsLong(w);
! 	} else {
! 		n = PyList_Size(list);
! 		np = PyInt_FromLong(n);
! 		if (np == NULL)
! 		    goto fail;
! 		if (PyList_Append(list, v) != 0)
! 		    goto fail;
! 		if (PyDict_SetItem(dict, t, np) != 0)
! 		    goto fail;
! 		Py_DECREF(np);
! 	}
! 	Py_DECREF(t);
  	return n;
+   fail:
+ 	Py_XDECREF(np);
+ 	Py_XDECREF(t);
+ 	c->c_errors++;
+ 	return 0;
  }
  
  static int
***************
*** 690,696 ****
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_consts, v);
  }
  
  static int
--- 703,709 ----
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_consts, c->c_const_dict, v);
  }
  
  static int
***************
*** 698,704 ****
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_names, v);
  }
  
  #ifdef PRIVATE_NAME_MANGLING
--- 711,717 ----
  	struct compiling *c;
  	PyObject *v;
  {
! 	return com_add(c, c->c_names, c->c_name_dict, v);
  }
  
  #ifdef PRIVATE_NAME_MANGLING

--Guido van Rossum (home page: http://www.python.org/~guido/)


From mal@lemburg.com  Fri Apr 28 20:49:24 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Fri, 28 Apr 2000 21:49:24 +0200
Subject: [Patches] Adding -U command line option
Message-ID: <3909EB44.CC4E58A2@lemburg.com>

This is a multi-part message in MIME format.
--------------9D1A897A1EADDA003E5CBA4B
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

The attached patch adds a new experimental option -U to the
Python interpreter. With the option enabled the Python compiler
interprets all "..." strings as u"..." (same with r"..." and
ur"...").

This provides a way to "look into the future" when someday
all strings turn into Unicode and also simplifies testing the
Unicode/string integration.

Patch Set Contents:
-------------------

Include/pydebug.h:

Added Py_UnicodeFlag for use by the -U command line option.

Modules/main.c:

Added -U command line option.

With the option enabled the Python compiler interprets all "..."
strings as u"..." (same with r"..." and ur"...").

Python/compile.c:

Support for the new -U command line option option:
with the option enabled the Python compiler
interprets all "..." strings as u"..." (same with r"..." and
ur"...").

Python/pythonrun.c:

Added Py_UnicodeFlag for use by the -U command line option.

_____________________________________________________________________
Disclaimer:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that I
have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce, distribute,
perform and/or display publicly, prepare derivative versions, and
otherwise use this contribution as part of the Python software and its
related documentation, or any derivative versions thereof, at no cost
to CNRI or its licensed users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or
not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------9D1A897A1EADDA003E5CBA4B
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-28.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-28.patch"

diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Include/pydebug.h Python+Unicode/Include/pydebug.h
--- CVS-Python/Include/pydebug.h	Fri Dec  4 19:48:15 1998
+++ Python+Unicode/Include/pydebug.h	Fri Apr 28 20:42:35 2000
@@ -43,6 +43,7 @@
 extern DL_IMPORT(int) Py_UseClassExceptionsFlag;
 extern DL_IMPORT(int) Py_FrozenFlag;
 extern DL_IMPORT(int) Py_TabcheckFlag;
+extern DL_IMPORT(int) Py_UnicodeFlag;
 
 DL_IMPORT(void) Py_FatalError	Py_PROTO((char *));
 
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Modules/main.c Python+Unicode/Modules/main.c
--- CVS-Python/Modules/main.c	Mon Apr 19 19:54:19 1999
+++ Python+Unicode/Modules/main.c	Fri Apr 28 20:47:36 2000
@@ -75,6 +75,7 @@
 ";
 static char *usage_mid = "\
 -u     : unbuffered binary stdout and stderr (also PYTHONUNBUFFERED=x)\n\
+-U     : experimental: turns '...' into u'...'\n\
 -v     : verbose (trace import statements) (also PYTHONVERBOSE=x)\n\
 -x     : skip first line of source, allowing use of non-Unix forms of #!cmd\n\
 -X     : disable class based built-in exceptions\n\
@@ -119,7 +120,7 @@
 	if ((p = getenv("PYTHONUNBUFFERED")) && *p != '\0')
 		unbuffered = 1;
 
-	while ((c = getopt(argc, argv, "c:diOStuvxX")) != EOF) {
+	while ((c = getopt(argc, argv, "c:diOStuUvxX")) != EOF) {
 		if (c == 'c') {
 			/* -c is the last option; following arguments
 			   that look like options are left for the
@@ -170,6 +171,10 @@
 
 		case 'X':
 			Py_UseClassExceptionsFlag = 0;
+			break;
+
+		case 'U':
+			Py_UnicodeFlag++;
 			break;
 
 		/* This space reserved for other options */
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Python/compile.c Python+Unicode/Python/compile.c
--- CVS-Python/Python/compile.c	Thu Apr 13 18:10:59 2000
+++ Python+Unicode/Python/compile.c	Fri Apr 28 20:41:23 2000
@@ -907,7 +907,7 @@
 			return NULL;
 		}
 	}
-	if (unicode) {
+	if (unicode || Py_UnicodeFlag) {
 		if (rawmode)
 			return PyUnicode_DecodeRawUnicodeEscape(
 				s, len, NULL);
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Python/pythonrun.c Python+Unicode/Python/pythonrun.c
--- CVS-Python/Python/pythonrun.c	Tue Apr 25 12:34:21 2000
+++ Python+Unicode/Python/pythonrun.c	Fri Apr 28 20:34:42 2000
@@ -88,6 +88,7 @@
 int Py_NoSiteFlag; /* Suppress 'import site' */
 int Py_UseClassExceptionsFlag = 1; /* Needed by bltinmodule.c */
 int Py_FrozenFlag; /* Needed by getpath.c */
+int Py_UnicodeFlag = 0; /* Needed by compile.c */
 
 static int initialized = 0;
 

--------------9D1A897A1EADDA003E5CBA4B--




From bwarsaw@python.org  Sat Apr 29 01:20:20 2000
From: bwarsaw@python.org (Barry A. Warsaw)
Date: Fri, 28 Apr 2000 20:20:20 -0400 (EDT)
Subject: [Patches] utime(path, NULL)
Message-ID: <14602.10948.514287.329941@anthem.cnri.reston.va.us>

On some OSes (Solaris and Linux at least), the second argument to
utime() can be NULL, which sets the atime and mtime of the file to the
current time.  Here is a patch that gives this functionality if second
argument to os.utime() is omitted.  E.g.

import os
os.utime(path)

==

#include <utime.h>
utime(path, NULL)

One minor bogosity: because of the way default arguments work for
longs, os.utime(path) is equivalent to os.utime(path, (-1, -1)).  Is
this a big deal?  Should other magic be used instead?

-Barry

Index: libos.tex
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Doc/lib/libos.tex,v
retrieving revision 1.38
diff -c -r1.38 libos.tex
*** libos.tex	2000/04/03 20:13:53	1.38
--- libos.tex	2000/04/29 00:15:42
***************
*** 704,710 ****
  
  \begin{funcdesc}{utime}{path, (atime, mtime)}
  Set the access and modified time of the file to the given values.
! (The second argument is a tuple of two items.)
  Availability: Macintosh, \UNIX{}, Windows.
  \end{funcdesc}
  
--- 704,711 ----
  
  \begin{funcdesc}{utime}{path, (atime, mtime)}
  Set the access and modified time of the file to the given values.
! The second argument is a tuple of two items; if it is omitted, then
! these times are set to the current time.
  Availability: Macintosh, \UNIX{}, Windows.
  \end{funcdesc}
  
Index: posixmodule.c
===================================================================
RCS file: /projects/cvsroot/python/dist/src/Modules/posixmodule.c,v
retrieving revision 2.132
diff -c -r2.132 posixmodule.c
*** posixmodule.c	2000/04/26 20:34:28	2.132
--- posixmodule.c	2000/04/29 00:10:52
***************
*** 1258,1264 ****
  	PyObject *args;
  {
  	char *path;
! 	long atime, mtime;
  	int res;
  
  /* XXX should define struct utimbuf instead, above */
--- 1258,1264 ----
  	PyObject *args;
  {
  	char *path;
! 	long atime=-1, mtime=-1;
  	int res;
  
  /* XXX should define struct utimbuf instead, above */
***************
*** 1274,1286 ****
  #define UTIME_ARG buf
  #endif /* HAVE_UTIME_H */
  
! 	if (!PyArg_ParseTuple(args, "s(ll):utime", &path, &atime, &mtime))
  		return NULL;
! 	ATIME = atime;
! 	MTIME = mtime;
! 	Py_BEGIN_ALLOW_THREADS
! 	res = utime(path, UTIME_ARG);
! 	Py_END_ALLOW_THREADS
  	if (res < 0)
  		return posix_error_with_filename(path);
  	Py_INCREF(Py_None);
--- 1274,1294 ----
  #define UTIME_ARG buf
  #endif /* HAVE_UTIME_H */
  
! 	if (!PyArg_ParseTuple(args, "s|(ll):utime", &path, &atime, &mtime))
  		return NULL;
! 	if (atime == -1 && mtime == -1) {
! 		/* optional time values not given */
! 		Py_BEGIN_ALLOW_THREADS
! 		res = utime(path, NULL);
! 		Py_END_ALLOW_THREADS
! 	}
! 	else {
! 		ATIME = atime;
! 		MTIME = mtime;
! 		Py_BEGIN_ALLOW_THREADS
! 		res = utime(path, UTIME_ARG);
! 		Py_END_ALLOW_THREADS
! 	}
  	if (res < 0)
  		return posix_error_with_filename(path);
  	Py_INCREF(Py_None);


From brian@garage.co.jp  Sat Apr 29 04:13:55 2000
From: brian@garage.co.jp (Brian Hooper)
Date: Sat, 29 Apr 2000 03:13:55 GMT
Subject: [Patches] slightly expanded Unicode supported Py_BuildValue
Message-ID: <20000429031355.84728.qmail@hotmail.com>

Hi again,

> > Here's a patch which changes modsupport to add 'U' and 'U#',
> > to support building Unicode objects from a null-terminated
> > Py_UNICODE *, and a Py_UNICODE * with length, respectively.
>
>If you would change the 'U' (capital U) to 'u' (lowercase u)
>the patch would be perfect :-)

After using the API with the lower-case 'u' a little bit, I feel
I should mention that it's actually a little confusing to have
the argument to PyArgs_ParseTuple be 'U' but the argument to
Py_BuildValue be 'u' - this in comparison with 's', which is
the same for both PyArgs_ParseTuple and Py_BuildValue, even though
one signifies a Python string object and the other signifies
regular char * data which gets turned into a Python string.

Doesn't this case seem like it ought to be parallel?  I think after actually 
trying it out it seems a little unwieldy and confusing.  If 'u' is 
preferable to 'U', why not change the PyArgs_ParseTuple argument to 'u' 
also?  I think making PyArgs_ParseTuple and Py_BuildValue as complimentary 
as possible makes them easiest to use and understand.

Just my 2 cents,
--Brian

________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com



From robin@jessikat.demon.co.uk  Sat Apr 29 09:40:45 2000
From: robin@jessikat.demon.co.uk (Robin Becker)
Date: Sat, 29 Apr 2000 09:40:45 +0100
Subject: [Patches] win32 module case bugfix to allcaps83
Message-ID: <SOcFULANAqC5EwuY@jessikat.demon.co.uk>

The following patch seems to fix a module case bug in 1.6a2 caused by wrong return values in
routine allcaps83 in import.c

*** import.c.orig       Sat Apr 29 09:12:18 2000
--- import.c    Sat Apr 29 09:14:34 2000
***************
*** 1035,1046 ****
        char *end = strchr(s, '\0');
        if (dot != NULL) {
                if (dot-s > 8)
!                       return 1; /* More than 8 before '.' */
                if (end-dot > 4)
!                       return 1; /* More than 3 after '.' */
                end = strchr(dot+1, '.');
                if (end != NULL)
!                       return 1; /* More than one dot  */
        }
        else if (end-s > 8)
                return 1; /* More than 8 and no dot */
--- 1035,1046 ----
        char *end = strchr(s, '\0');
        if (dot != NULL) {
                if (dot-s > 8)
!                       return 0; /* More than 8 before '.' */
                if (end-dot > 4)
!                       return 0; /* More than 3 after '.' */
                end = strchr(dot+1, '.');
                if (end != NULL)
!                       return 0; /* More than one dot  */
        }
        else if (end-s > 8)
                return 1; /* More than 8 and no dot */


                     I confirm that, to the best of my knowledge and belief, this
                     contribution is free of any claims of third parties under
                     copyright, patent or other rights or interests ("claims").  To
                     the extent that I have any such claims, I hereby grant to CNRI a
                     nonexclusive, irrevocable, royalty-free, worldwide license to
                     reproduce, distribute, perform and/or display publicly, prepare
                     derivative versions, and otherwise use this contribution as part
                     of the Python software and its related documentation, or any
                     derivative versions thereof, at no cost to CNRI or its licensed
                     users, and to authorize others to do so.

                     I acknowledge that CNRI may, at its sole discretion, decide
                     whether or not to incorporate this contribution in the Python
                     software and its related documentation.  I further grant CNRI
                     permission to use my name and other identifying information
                     provided to CNRI by me for use in connection with the Python
                     software and its related documentation.
-- 
Robin Becker


From mal@lemburg.com  Sat Apr 29 14:01:01 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Sat, 29 Apr 2000 15:01:01 +0200
Subject: [Patches] slightly expanded Unicode supported Py_BuildValue
References: <20000429031355.84728.qmail@hotmail.com>
Message-ID: <390ADD0D.9722FA7D@lemburg.com>

Brian Hooper wrote:
> 
> Hi again,
> 
> > > Here's a patch which changes modsupport to add 'U' and 'U#',
> > > to support building Unicode objects from a null-terminated
> > > Py_UNICODE *, and a Py_UNICODE * with length, respectively.
> >
> >If you would change the 'U' (capital U) to 'u' (lowercase u)
> >the patch would be perfect :-)
> 
> After using the API with the lower-case 'u' a little bit, I feel
> I should mention that it's actually a little confusing to have
> the argument to PyArgs_ParseTuple be 'U' but the argument to
> Py_BuildValue be 'u' - this in comparison with 's', which is
> the same for both PyArgs_ParseTuple and Py_BuildValue, even though
> one signifies a Python string object and the other signifies
> regular char * data which gets turned into a Python string.
> 
> Doesn't this case seem like it ought to be parallel?  I think after actually
> trying it out it seems a little unwieldy and confusing.  If 'u' is
> preferable to 'U', why not change the PyArgs_ParseTuple argument to 'u'
> also?  I think making PyArgs_ParseTuple and Py_BuildValue as complimentary
> as possible makes them easiest to use and understand.

With the case switch, your patch gets in sync with what
"s" does for PyArgs_ParseTuple() and Py_BuildValue(): the
first reads string objects as char* while the second creates
string objects from char*.

The "U" in PyArgs_ParseTuple() does two things: it first
checks the object for being a Unicode object and then
returns a PyObject*. There also is a "S" marker which
does the same for string objects.

One could argue to also have a set of "u" markers for
PyArgs_ParseTuple(): these should then return Py_UNICODE*
instead of char*. Perhaps this is what you are thinking
about ? It would make a nice complement to your recent patches.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From mal@lemburg.com  Sat Apr 29 14:34:47 2000
From: mal@lemburg.com (M.-A. Lemburg)
Date: Sat, 29 Apr 2000 15:34:47 +0200
Subject: [Patches] Adding -U command line option
References: <3909EB44.CC4E58A2@lemburg.com>
Message-ID: <390AE4F7.84D6B996@lemburg.com>

This is a multi-part message in MIME format.
--------------44FAD210BAEBD7EFCE0EE2E0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

The attached patch adds a new experimental option -U to the
Python interpreter. With the option enabled the Python compiler
interprets all "..." strings as u"..." (same with r"..." and
ur"...").

This provides a way to "look into the future" when someday
all strings turn into Unicode and also simplifies testing the
Unicode/string integration.

The new version of the patch fixes a problem when trying to
use byte compiled PYC files compiled with -U option and without
this option side-by-side: when using -U, the magic number for
PYC files gets incremented by one so that a Python process
using this option will only use precompiled PYC files created
by Python processes also using -U.

This allow people willing to test drive the complete switch
to Unicode without harming their standard Python installation.


Patch Set Contents:
-------------------

Include/pydebug.h:

Added Py_UnicodeFlag for use by the -U command line option.

Modules/main.c:

Added -U command line option.

With the option enabled the Python compiler interprets all "..."
strings as u"..." (same with r"..." and ur"...").

Python/compile.c:

Support for the new -U command line option option:
with the option enabled the Python compiler
interprets all "..." strings as u"..." (same with r"..." and
ur"...").

Python/pythonrun.c:

Added Py_UnicodeFlag for use by the -U command line option.

Python/import.c:

Changed all references to the MAGIC constant to use a global
pyc_magic instead. This global is initially set to MAGIC, but can be
changed by the _PyImport_Init() function to provide for
special features implemented in the compiler which are settable
using command line switches and affect the way PYC files are
generated.

Currently this change is only done for the -U flag.

_____________________________________________________________________
Disclaimer:

I confirm that, to the best of my knowledge and belief, this
contribution is free of any claims of third parties under copyright,
patent or other rights or interests ("claims").  To the extent that I
have any such claims, I hereby grant to CNRI a nonexclusive,
irrevocable, royalty-free, worldwide license to reproduce, distribute,
perform and/or display publicly, prepare derivative versions, and
otherwise use this contribution as part of the Python software and its
related documentation, or any derivative versions thereof, at no cost
to CNRI or its licensed users, and to authorize others to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or
not to incorporate this contribution in the Python software and its
related documentation.  I further grant CNRI permission to use my name
and other identifying information provided to CNRI by me for use in
connection with the Python software and its related documentation.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/
--------------44FAD210BAEBD7EFCE0EE2E0
Content-Type: text/plain; charset=us-ascii;
 name="Unicode-Implementation-2000-04-29.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="Unicode-Implementation-2000-04-29.patch"

diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Include/pydebug.h Python+Unicode/Include/pydebug.h
--- CVS-Python/Include/pydebug.h	Fri Dec  4 19:48:15 1998
+++ Python+Unicode/Include/pydebug.h	Fri Apr 28 20:42:35 2000
@@ -43,6 +43,7 @@
 extern DL_IMPORT(int) Py_UseClassExceptionsFlag;
 extern DL_IMPORT(int) Py_FrozenFlag;
 extern DL_IMPORT(int) Py_TabcheckFlag;
+extern DL_IMPORT(int) Py_UnicodeFlag;
 
 DL_IMPORT(void) Py_FatalError	Py_PROTO((char *));
 
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Modules/main.c Python+Unicode/Modules/main.c
--- CVS-Python/Modules/main.c	Mon Apr 19 19:54:19 1999
+++ Python+Unicode/Modules/main.c	Fri Apr 28 20:47:36 2000
@@ -75,6 +75,7 @@
 ";
 static char *usage_mid = "\
 -u     : unbuffered binary stdout and stderr (also PYTHONUNBUFFERED=x)\n\
+-U     : experimental: turns '...' into u'...'\n\
 -v     : verbose (trace import statements) (also PYTHONVERBOSE=x)\n\
 -x     : skip first line of source, allowing use of non-Unix forms of #!cmd\n\
 -X     : disable class based built-in exceptions\n\
@@ -119,7 +120,7 @@
 	if ((p = getenv("PYTHONUNBUFFERED")) && *p != '\0')
 		unbuffered = 1;
 
-	while ((c = getopt(argc, argv, "c:diOStuvxX")) != EOF) {
+	while ((c = getopt(argc, argv, "c:diOStuUvxX")) != EOF) {
 		if (c == 'c') {
 			/* -c is the last option; following arguments
 			   that look like options are left for the
@@ -170,6 +171,10 @@
 
 		case 'X':
 			Py_UseClassExceptionsFlag = 0;
+			break;
+
+		case 'U':
+			Py_UnicodeFlag++;
 			break;
 
 		/* This space reserved for other options */
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Python/compile.c Python+Unicode/Python/compile.c
--- CVS-Python/Python/compile.c	Thu Apr 13 18:10:59 2000
+++ Python+Unicode/Python/compile.c	Fri Apr 28 20:41:23 2000
@@ -907,7 +907,7 @@
 			return NULL;
 		}
 	}
-	if (unicode) {
+	if (unicode || Py_UnicodeFlag) {
 		if (rawmode)
 			return PyUnicode_DecodeRawUnicodeEscape(
 				s, len, NULL);
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Python/import.c Python+Unicode/Python/import.c
--- CVS-Python/Python/import.c	Tue Apr 25 12:34:21 2000
+++ Python+Unicode/Python/import.c	Sat Apr 29 15:18:01 2000
@@ -84,7 +84,12 @@
 /* XXX Perhaps the magic number should be frozen and a version field
    added to the .pyc file header? */
 /* New way to come up with the magic number: (YEAR-1995), MONTH, DAY */
-#define MAGIC (20121 | ((long)'\r'<<16) | ((long)'\n'<<24))
+#define MAGIC (50428 | ((long)'\r'<<16) | ((long)'\n'<<24))
+
+/* Magic word as global; note that _PyImport_Init() can change the
+   value of this global to accomodate for alterations of how the
+   compiler works which are enabled by command line switches. */
+static long pyc_magic = MAGIC;
 
 /* See _PyImport_FixupExtension() below */
 static PyObject *extensions = NULL;
@@ -135,6 +140,13 @@
 				filetab->suffix = ".pyo";
 		}
 	}
+
+	if (Py_UnicodeFlag) {
+		/* Fix the pyc_magic so that byte compiled code created
+		   using the all-Unicode method doesn't interfere with
+		   code created in normal operation mode. */
+		pyc_magic = MAGIC + 1;
+	}
 }
 
 void
@@ -366,7 +378,7 @@
 long
 PyImport_GetMagicNumber()
 {
-	return MAGIC;
+	return pyc_magic;
 }
 
 
@@ -572,7 +584,7 @@
 	if (fp == NULL)
 		return NULL;
 	magic = PyMarshal_ReadLongFromFile(fp);
-	if (magic != MAGIC) {
+	if (magic != pyc_magic) {
 		if (Py_VerboseFlag)
 			PySys_WriteStderr("# %s has bad magic\n", cpathname);
 		fclose(fp);
@@ -627,7 +639,7 @@
 	PyObject *m;
 
 	magic = PyMarshal_ReadLongFromFile(fp);
-	if (magic != MAGIC) {
+	if (magic != pyc_magic) {
 		PyErr_Format(PyExc_ImportError,
 			     "Bad magic number in %.200s", cpathname);
 		return NULL;
@@ -685,7 +697,7 @@
 				"# can't create %s\n", cpathname);
 		return;
 	}
-	PyMarshal_WriteLongToFile(MAGIC, fp);
+	PyMarshal_WriteLongToFile(pyc_magic, fp);
 	/* First write a 0 for mtime */
 	PyMarshal_WriteLongToFile(0L, fp);
 	PyMarshal_WriteObjectToFile((PyObject *)co, fp);
@@ -1953,10 +1965,10 @@
 
 	if (!PyArg_ParseTuple(args, ":get_magic"))
 		return NULL;
-	buf[0] = (char) ((MAGIC >>  0) & 0xff);
-	buf[1] = (char) ((MAGIC >>  8) & 0xff);
-	buf[2] = (char) ((MAGIC >> 16) & 0xff);
-	buf[3] = (char) ((MAGIC >> 24) & 0xff);
+	buf[0] = (char) ((pyc_magic >>  0) & 0xff);
+	buf[1] = (char) ((pyc_magic >>  8) & 0xff);
+	buf[2] = (char) ((pyc_magic >> 16) & 0xff);
+	buf[3] = (char) ((pyc_magic >> 24) & 0xff);
 
 	return PyString_FromStringAndSize(buf, 4);
 }
diff -u -rbP -x *.o -x *.pyc -x Makefile -x *~ -x *.so -x add2lib -x pgen -x buildno -x config.* -x libpython* -x python -x Setup -x Setup.local -x Setup.thread -x hassignal -x Makefile.pre -x configure -x *.bak -x *.s -x DEADJOE -x *.rej -x *.orig -x Demo -x CVS -x Doc -x *.orig -x .#* -x distutils -x PC -x PCbuild -x *.py -x ACKS -x *.txt -x README CVS-Python/Python/pythonrun.c Python+Unicode/Python/pythonrun.c
--- CVS-Python/Python/pythonrun.c	Tue Apr 25 12:34:21 2000
+++ Python+Unicode/Python/pythonrun.c	Fri Apr 28 20:34:42 2000
@@ -88,6 +88,7 @@
 int Py_NoSiteFlag; /* Suppress 'import site' */
 int Py_UseClassExceptionsFlag = 1; /* Needed by bltinmodule.c */
 int Py_FrozenFlag; /* Needed by getpath.c */
+int Py_UnicodeFlag = 0; /* Needed by compile.c */
 
 static int initialized = 0;
 

--------------44FAD210BAEBD7EFCE0EE2E0--




From brian@garage.co.jp  Sat Apr 29 19:19:13 2000
From: brian@garage.co.jp (Brian Hooper)
Date: Sat, 29 Apr 2000 18:19:13 GMT
Subject: [Patches] slightly expanded Unicode supported Py_BuildValue
Message-ID: <20000429181913.28204.qmail@hotmail.com>

Hi Marc-Andre,

Thanks for your crystal clear explanation - I understand now!
'U' and 'u' are analogous to 'S' and 's'.

OK, I'm going to make everything symmetric, then :-)

--Brian

>Brian Hooper wrote:
> >
> > Hi again,
> >
> > > > Here's a patch which changes modsupport to add 'U' and 'U#',
> > > > to support building Unicode objects from a null-terminated
> > > > Py_UNICODE *, and a Py_UNICODE * with length, respectively.
> > >
> > >If you would change the 'U' (capital U) to 'u' (lowercase u)
> > >the patch would be perfect :-)
> >
> > After using the API with the lower-case 'u' a little bit, I feel
> > I should mention that it's actually a little confusing to have
> > the argument to PyArgs_ParseTuple be 'U' but the argument to
> > Py_BuildValue be 'u' - this in comparison with 's', which is
> > the same for both PyArgs_ParseTuple and Py_BuildValue, even though
> > one signifies a Python string object and the other signifies
> > regular char * data which gets turned into a Python string.
> >
> > Doesn't this case seem like it ought to be parallel?  I think after 
>actually
> > trying it out it seems a little unwieldy and confusing.  If 'u' is
> > preferable to 'U', why not change the PyArgs_ParseTuple argument to 'u'
> > also?  I think making PyArgs_ParseTuple and Py_BuildValue as 
>complimentary
> > as possible makes them easiest to use and understand.
>
>With the case switch, your patch gets in sync with what
>"s" does for PyArgs_ParseTuple() and Py_BuildValue(): the
>first reads string objects as char* while the second creates
>string objects from char*.
>
>The "U" in PyArgs_ParseTuple() does two things: it first
>checks the object for being a Unicode object and then
>returns a PyObject*. There also is a "S" marker which
>does the same for string objects.
>
>One could argue to also have a set of "u" markers for
>PyArgs_ParseTuple(): these should then return Py_UNICODE*
>instead of char*. Perhaps this is what you are thinking
>about ? It would make a nice complement to your recent patches.
>
>--
>Marc-Andre Lemburg
>______________________________________________________________________
>Business:                                      http://www.lemburg.com/
>Python Pages:                           http://www.lemburg.com/python/
>
>

________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com



From Vladimir.Marangozov@inrialpes.fr  Sun Apr 30 20:39:24 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sun, 30 Apr 2000 21:39:24 +0200 (CEST)
Subject: [Patches] PyMem [1/8] - Memory API: Summary
Message-ID: <200004301939.VAA14990@python.inrialpes.fr>


Second (and hopefully, final) round of patches relative to the malloc
cleanup and the renewed memory interfaces. It affects *lots* of files
in the distribution -- more than enough for May 1st ;-)

I hereby send a patch per directory to easy their review, except the
two key include files, mymalloc.h and objimpl.h which are to be saved
as is (better than a context diff for Include/* anyway).

Interface summary
=================

Each memory interface exports both functions and macros. Functions
preserve binary compatibility of the code across different releases
while MACROS trade it for speed.


   Public API       Functions                      MACROS
   --------------   ----------------------------   --------------------

1) Core Memory Allocator (see mymalloc.h)

 - raw mem malloc   PyMem_Malloc/Realloc/Free      PyMem_MALLOC/REALLOC/FREE
 - type-oriented    PyMem_New/Resize/Del           PyMem_NEW/RESIZE/DEL

2) Core Object Memory Allocator & Facilities (see objimpl.h)

 - object malloc    PyObject_Malloc/Realloc/Free   PyObject_MALLOC/REALLOC/FREE
 - initialization   PyObject_Init/InitVar          PyObject_INIT/INIT_VAR
 - constr./destr.   PyObject_New/NewVar/Del        PyObject_NEW/NEW_VAR/DEL


All APIs operate on the Python heap(s). Objects that are subject to custom
allocation strategies (custom = non Python = std malloc, C++ new, 3rd party)
are out of bounds -- we can't do much about them from Python except providing
handy macros related to their structure. The extension-type developer is
responsible for their allocation/deallocation with the malloc of choice.
See the comments in objimpl.h

I tend to be slightly verbose :-) so if you think the comments in the source
can be shortened (or improved) for clarity, you're welcome..

Needless to say, once checked in, a complete overhaul of the Python memory
interfaces would be in place. I encourage you to take some time reviewing
your favorite files in the distribution and see how and why they have
changed. More specific comments follow for each part of the patch suite.

--
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252


From Vladimir.Marangozov@inrialpes.fr  Sun Apr 30 20:40:33 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sun, 30 Apr 2000 21:40:33 +0200 (CEST)
Subject: [Patches] PyMem [2/8] - Include/mymalloc.h
Message-ID: <200004301940.VAA15018@python.inrialpes.fr>


[ Include/mymalloc.h ]

The Python core malloc. These interfaces have been discussed extensively
already -- I have nothing to add here. For now, the default Python malloc
is standard C malloc so I don't expect any code breakage.

--
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--

Disclaimer:

I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------
#ifndef Py_MYMALLOC_H
#define Py_MYMALLOC_H
/***********************************************************
Copyright 1991-1995 by Stichting Mathematisch Centrum, Amsterdam,
The Netherlands.

                        All Rights Reserved

Permission to use, copy, modify, and distribute this software and its
documentation for any purpose and without fee is hereby granted,
provided that the above copyright notice appear in all copies and that
both that copyright notice and this permission notice appear in
supporting documentation, and that the names of Stichting Mathematisch
Centrum or CWI or Corporation for National Research Initiatives or
CNRI not be used in advertising or publicity pertaining to
distribution of the software without specific, written prior
permission.

While CWI is the initial source for this software, a modified version
is made available by the Corporation for National Research Initiatives
(CNRI) at the Internet address ftp://ftp.python.org.

STICHTING MATHEMATISCH CENTRUM AND CNRI DISCLAIM ALL WARRANTIES WITH
REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL STICHTING MATHEMATISCH
CENTRUM OR CNRI BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
PERFORMANCE OF THIS SOFTWARE.

******************************************************************/

/* Lowest-level memory allocation interface */

#ifdef macintosh
#define ANY void
#endif

#ifdef __STDC__
#define ANY void
#endif

#ifdef __TURBOC__
#define ANY void
#endif

#ifdef __GNUC__
#define ANY void
#endif

#ifndef ANY
#define ANY char
#endif

#ifdef HAVE_STDLIB_H
#include <stdlib.h>
#endif

#include "myproto.h"

#ifdef __cplusplus
/* Move this down here since some C++ #include's don't like to be included
   inside an extern "C" */
extern "C" {
#endif

#ifdef SYMANTEC__CFM68K__
#pragma lib_export on
#endif

#ifndef DL_IMPORT       /* declarations for DLL import */
#define DL_IMPORT(RTYPE) RTYPE
#endif

#ifndef NULL
#define NULL ((ANY *)0)
#endif

#ifdef MALLOC_ZERO_RETURNS_NULL
/* XXX Always allocate one extra byte, since some malloc's return NULL
   XXX for malloc(0) or realloc(p, 0). */
#define _PyMem_EXTRA 1
#else
#define _PyMem_EXTRA 0
#endif

/*
 * Core memory allocator
 * =====================
 */

/* To make sure the interpreter is user-malloc friendly, all memory
   APIs are implemented on top of this one.

   The PyCore_* macros can be defined to make the interpreter use a
   custom allocator. Note that they are for internal use only. Both
   the core and extension modules should use the PyMem_* API. */

#ifndef PyCore_MALLOC_FUNC
#undef PyCore_REALLOC_FUNC
#undef PyCore_FREE_FUNC
#define PyCore_MALLOC_FUNC      malloc
#define PyCore_REALLOC_FUNC     realloc
#define PyCore_FREE_FUNC        free
#endif

#ifndef PyCore_MALLOC_PROTO
#undef PyCore_REALLOC_PROTO
#undef PyCore_FREE_PROTO
#define PyCore_MALLOC_PROTO     Py_PROTO((size_t))
#define PyCore_REALLOC_PROTO    Py_PROTO((ANY *, size_t))
#define PyCore_FREE_PROTO       Py_PROTO((ANY *))
#endif

#ifdef NEED_TO_DECLARE_MALLOC_AND_FRIEND
extern ANY *PyCore_MALLOC_FUNC PyCore_MALLOC_PROTO;
extern ANY *PyCore_REALLOC_FUNC PyCore_REALLOC_PROTO;
extern void PyCore_FREE_FUNC PyCore_FREE_PROTO;
#endif

#ifndef PyCore_MALLOC
#undef PyCore_REALLOC
#undef PyCore_FREE
#define PyCore_MALLOC(n)        PyCore_MALLOC_FUNC(n)
#define PyCore_REALLOC(p, n)    PyCore_REALLOC_FUNC((p), (n))
#define PyCore_FREE(p)          PyCore_FREE_FUNC(p)
#endif

/* BEWARE:

   Each interface exports both functions and macros. Extension modules
   should normally use the functions for ensuring binary compatibility
   of the user's code across Python versions. Subsequently, if the
   Python runtime switches to its own malloc (different from standard
   malloc), no recompilation is required for the extensions.

   The macro versions trade compatibility for speed. They can be used
   whenever there is a performance problem, but their use implies
   recompilation of the code for each new Python release. The Python
   core uses the macros because it *is* compiled on every upgrade.
   This might not be the case with 3rd party extensions in a custom
   setup (for example, a customer does not always have access to the
   source of 3rd party deliverables). You have been warned! */

/*
 * Raw memory interface
 * ====================
 */

/* Functions */

/* Function wrappers around PyCore_MALLOC and friends; useful if you
   need to be sure that you are using the same memory allocator as
   Python.  Note that the wrappers make sure that allocating 0 bytes
   returns a non-NULL pointer, even if the underlying malloc
   doesn't. Returned pointers must be checked for NULL explicitly.
   No action is performed on failure. */
extern DL_IMPORT(ANY *) PyMem_Malloc Py_PROTO((size_t));
extern DL_IMPORT(ANY *) PyMem_Realloc Py_PROTO((ANY *, size_t));
extern DL_IMPORT(void) PyMem_Free Py_PROTO((ANY *));

/* Starting from Python 1.6, the wrappers Py_{Malloc,Realloc,Free} are
   no longer supported. They used to call PyErr_NoMemory() on failure.
   The following are provided for backwards source code compatibility.
   Their use is deprecated. */
#define Py_Malloc(n)     /* deprecated Py_Malloc  */ PyMem_Malloc(n)
#define Py_Realloc(p, n) /* deprecated Py_Realloc */ PyMem_Realloc((p), (n))
#define Py_Free(p)       /* deprecated Py_Free    */ PyMem_Free(p)

/* Macros */
#define PyMem_MALLOC(n)         PyCore_MALLOC(n)
#define PyMem_REALLOC(p, n)     PyCore_REALLOC((ANY *)(p), (n))
#define PyMem_FREE(p)           PyCore_FREE((ANY *)(p))

/*
 * Type-oriented memory interface
 * ==============================
 */

/* Functions */
#define PyMem_New(type, n) \
	( (type *) PyMem_Malloc((n) * sizeof(type)) )
#define PyMem_Resize(p, type, n) \
	( (p) = (type *) PyMem_Realloc((n) * sizeof(type)) )
#define PyMem_Del(p) PyMem_Free(p)

/* Macros */
#define PyMem_NEW(type, n) \
	( (type *) PyMem_MALLOC(_PyMem_EXTRA + (n) * sizeof(type)) )
#define PyMem_RESIZE(p, type, n) \
	if ((p) == NULL) \
		(p) = (type *)(PyMem_MALLOC( \
				    _PyMem_EXTRA + (n) * sizeof(type))); \
	else \
		(p) = (type *)(PyMem_REALLOC((p), \
				    _PyMem_EXTRA + (n) * sizeof(type)))
#define PyMem_DEL(p) PyMem_FREE(p)

/* PyMem_XDEL is deprecated. To avoid the call when p is NULL,
   it is recommended to write the test explicitly in the code.
   Note that according to ANSI C, free(NULL) has no effect. */

#define PyMem_XDEL(p) /* deprecated PyMem_XDEL */ PyMem_DEL(p)

#ifdef __cplusplus
}
#endif

#endif /* !Py_MYMALLOC_H */


From Vladimir.Marangozov@inrialpes.fr  Sun Apr 30 20:41:26 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sun, 30 Apr 2000 21:41:26 +0200 (CEST)
Subject: [Patches] PyMem [3/8] - Include/objimpl.h
Message-ID: <200004301941.VAA15041@python.inrialpes.fr>


[ Include/objimpl.h ]

This is news for everybody.

This is the Python object allocator for implementing object contructors
and destructors. It may differ from the core malloc completely. Quite
useful for "real" GC, among others (not that GC over RC is unreal!)
By default, it's implemented on top (in terms) of the core malloc.

All extension (non-core) types should use the function version of the
API: PyObject_New/NewVar/Del, *not* the macro version. This ensures that
Python keeps control on *all* "normal" objects, except the "foreigners"
that are allocated with custom mallocs.

For custom mallocs, we still require the initialization of the object
through PyObject_Init/InitVar after its allocation! Thus, we can have
the potential to customize and/or change the internal object header
representation (the PyObject struct).

The object malloc API (PyObject_Malloc/Realloc/Free) is exported because
sometimes people want to customize their object-specific allocators, so
PyObject_New gets inlined (in practice, this is done in more or less
random ways). To avoid this randomness, we also provide some facilities
for customizing the object constructors (_PyObject_SIZE/VAR_SIZE and
PyObject_Init/InitVar). We're also willing to provide guidance on how to
achieve this customization with the interfaces defined below.

All this may seem unnecessary at first sight, but take a good look at the
incoming Objects/* and Modules/* patches. The code speaks for itself.

--
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--

Disclaimer:

I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------
#ifndef Py_OBJIMPL_H
#define Py_OBJIMPL_H
#ifdef __cplusplus
extern "C" {
#endif

/***********************************************************
Copyright 1991-1995 by Stichting Mathematisch Centrum, Amsterdam,
The Netherlands.

                        All Rights Reserved

Permission to use, copy, modify, and distribute this software and its
documentation for any purpose and without fee is hereby granted,
provided that the above copyright notice appear in all copies and that
both that copyright notice and this permission notice appear in
supporting documentation, and that the names of Stichting Mathematisch
Centrum or CWI or Corporation for National Research Initiatives or
CNRI not be used in advertising or publicity pertaining to
distribution of the software without specific, written prior
permission.

While CWI is the initial source for this software, a modified version
is made available by the Corporation for National Research Initiatives
(CNRI) at the Internet address ftp://ftp.python.org.

STICHTING MATHEMATISCH CENTRUM AND CNRI DISCLAIM ALL WARRANTIES WITH
REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL STICHTING MATHEMATISCH
CENTRUM OR CNRI BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
PERFORMANCE OF THIS SOFTWARE.

******************************************************************/

#include "mymalloc.h"

/*
Functions and macros for modules that implement new object types.
You must first include "object.h".

 - PyObject_New(type, typeobj) allocates memory for a new object of
   the given type; here 'type' must be the C structure type used to
   represent the object and 'typeobj' the address of the corresponding
   type object.  Reference count and type pointer are filled in; the
   rest of the bytes of the object are *undefined*!  The resulting
   expression type is 'type *'.  The size of the object is actually
   determined by the tp_basicsize field of the type object.

 - PyObject_NewVar(type, typeobj, n) is similar but allocates a
   variable-size object with n extra items.  The size is computed as
   tp_basicsize plus n * tp_itemsize.  This fills in the ob_size field
   as well.

 - PyObject_Del(op) releases the memory allocated for an object.

 - PyObject_Init(op, typeobj) and PyObject_InitVar(op, typeobj, n) are
   similar to PyObject_{New, NewVar} except that they don't allocate
   the memory needed for an object. Instead of the 'type' parameter,
   they accept the pointer of a new object (allocated by an arbitrary
   allocator) and initialize its object header fields.

Note that objects created with PyObject_{New, NewVar} are allocated
within the Python heap by an object allocator, the latter being
implemented (by default) on top of the Python raw memory
allocator. This ensures that Python keeps control on the user's
objects regarding their memory management; for instance, they may be
subject to automatic garbage collection.

In case a specific form of memory management is needed, implying that
the objects would not reside in the Python heap (for example standard
malloc heap(s) are mandatory, use of shared memory, C++ local storage
or operator new), you must first allocate the object with your custom
allocator, then pass its pointer to PyObject_{Init, InitVar} for
filling in its Python-specific fields: reference count, type pointer,
possibly others. You should be aware that Python has very limited
control over these objects because they don't cooperate with the
Python memory manager. Such objects may not be eligible for automatic
garbage collection and you have to make sure that they are released
accordingly whenever their destructor gets called (cf. the specific
form of memory management you're using).

Unless you have specific memory management requirements, it is
recommended to use PyObject_{New, NewVar, Del}. */

/* 
 * Core object memory allocator
 * ============================
 */

/* The purpose of the object allocator is to make make the distinction
   between "object memory" and the rest within the Python heap.
   
   Object memory is the one allocated by PyObject_{New, NewVar}, i.e.
   the one that holds the object's representation defined by its C
   type structure, *excluding* any object-specific memory buffers that
   might be referenced by the structure (for type structures that have
   pointer fields).  By default, the object memory allocator is
   implemented on top of the raw memory allocator.

   The PyCore_* macros can be defined to make the interpreter use a
   custom object memory allocator. They are reserved for internal
   memory management purposes exclusively. Both the core and extension
   modules should use the PyObject_* API. */

#ifndef PyCore_OBJECT_MALLOC_FUNC
#undef PyCore_OBJECT_REALLOC_FUNC
#undef PyCore_OBJECT_FREE_FUNC
#define PyCore_OBJECT_MALLOC_FUNC    PyCore_MALLOC_FUNC
#define PyCore_OBJECT_REALLOC_FUNC   PyCore_REALLOC_FUNC
#define PyCore_OBJECT_FREE_FUNC      PyCore_FREE_FUNC
#endif

#ifndef PyCore_OBJECT_MALLOC_PROTO
#undef PyCore_OBJECT_REALLOC_PROTO
#undef PyCore_OBJECT_FREE_PROTO
#define PyCore_OBJECT_MALLOC_PROTO   PyCore_MALLOC_PROTO
#define PyCore_OBJECT_REALLOC_PROTO  PyCore_REALLOC_PROTO
#define PyCore_OBJECT_FREE_PROTO     PyCore_FREE_PROTO
#endif

#ifdef NEED_TO_DECLARE_OBJECT_MALLOC_AND_FRIEND
extern ANY *PyCore_OBJECT_MALLOC_FUNC PyCore_OBJECT_MALLOC_PROTO;
extern ANY *PyCore_OBJECT_REALLOC_FUNC PyCore_OBJECT_REALLOC_PROTO;
extern void PyCore_OBJECT_FREE_FUNC PyCore_OBJECT_FREE_PROTO;
#endif

#ifndef PyCore_OBJECT_MALLOC
#undef PyCore_OBJECT_REALLOC
#undef PyCore_OBJECT_FREE
#define PyCore_OBJECT_MALLOC(n)      PyCore_OBJECT_MALLOC_FUNC(n)
#define PyCore_OBJECT_REALLOC(p, n)  PyCore_OBJECT_REALLOC_FUNC((p), (n))
#define PyCore_OBJECT_FREE(p)        PyCore_OBJECT_FREE_FUNC(p)
#endif

/*
 * Raw object memory interface
 * ===========================
 */

/* The use of this API should be avoided, unless a builtin object
   constructor inlines PyObject_{New, NewVar}, either because the
   latter functions cannot allocate the exact amount of needed memory,
   either for speed. This situation is exceptional, but occurs for
   some object constructors (PyBuffer_New, PyList_New...).  Inlining
   PyObject_{New, NewVar} for objects that are supposed to belong to
   the Python heap is discouraged. If you really have to, make sure
   the object is initialized with PyObject_{Init, InitVar}. Do *not*
   inline PyObject_{Init, InitVar} for user-extension types or you
   might seriously interfere with Python's memory management. */

/* Functions */

/* Wrappers around PyCore_OBJECT_MALLOC and friends; useful if you
   need to be sure that you are using the same object memory allocator
   as Python. These wrappers *do not* make sure that allocating 0
   bytes returns a non-NULL pointer. Returned pointers must be checked
   for NULL explicitly; no action is performed on failure. */
extern DL_IMPORT(ANY *) PyObject_Malloc Py_PROTO((size_t));
extern DL_IMPORT(ANY *) PyObject_Realloc Py_PROTO((ANY *, size_t));
extern DL_IMPORT(void) PyObject_Free Py_PROTO((ANY *));

/* Macros */
#define PyObject_MALLOC(n)           PyCore_OBJECT_MALLOC(n)
#define PyObject_REALLOC(op, n)      PyCore_OBJECT_REALLOC((ANY *)(op), (n))
#define PyObject_FREE(op)            PyCore_OBJECT_FREE((ANY *)(op))

/*
 * Generic object allocator interface
 * ==================================
 */

/* Functions */
extern DL_IMPORT(PyObject *) PyObject_Init Py_PROTO((PyObject *, PyTypeObject *));
extern DL_IMPORT(PyVarObject *) PyObject_InitVar Py_PROTO((PyVarObject *, PyTypeObject *, int));
extern DL_IMPORT(PyObject *) _PyObject_New Py_PROTO((PyTypeObject *));
extern DL_IMPORT(PyVarObject *) _PyObject_NewVar Py_PROTO((PyTypeObject *, int));
extern DL_IMPORT(void) _PyObject_Del Py_PROTO((PyObject *));

#define PyObject_New(type, typeobj) \
		( (type *) _PyObject_New(typeobj) )
#define PyObject_NewVar(type, typeobj, n) \
		( (type *) _PyObject_NewVar((typeobj), (n)) )
#define PyObject_Del(op) _PyObject_Del((PyObject *)(op))

/* Macros trading binary compatibility for speed. See also mymalloc.h.
   Note that these macros expect non-NULL object pointers.*/
#define PyObject_INIT(op, typeobj) \
	( _Py_NewReference((PyObject *)(op)), (op)->ob_type = (typeobj), (op) )
#define PyObject_INIT_VAR(op, typeobj, size) \
	( PyObject_INIT((op), (typeobj)) -> ob_size = (size), (op) )

#define _PyObject_SIZE(typeobj) ( (typeobj)->tp_basicsize )
#define _PyObject_VAR_SIZE(typeobj, n) \
	( (typeobj)->tp_basicsize + (n) * (typeobj)->tp_itemsize )

#define PyObject_NEW(type, typeobj) \
( (type *) PyObject_Init( \
	(PyObject *) PyObject_MALLOC( _PyObject_SIZE(typeobj) ), (typeobj)) )
#define PyObject_NEW_VAR(type, typeobj, n) \
( (type *) PyObject_InitVar( \
	(PyVarObject *) PyObject_MALLOC( _PyObject_VAR_SIZE((typeobj),(n)) ),\
	(typeobj), (n)) )
#define PyObject_DEL(op) PyObject_FREE(op)

/* This example code implements an object constructor with a custom
   allocator, where PyObject_New is inlined, and shows the important
   distinction between two steps (at least):
       1) the actual allocation of the object storage;
       2) the initialization of the Python specific fields
          in this storage with PyObject_{Init, InitVar}.

   PyObject *
   YourObject_New(...)
   {
       PyObject *op;

       op = (PyObject *) Your_Allocator(_PyObject_SIZE(YourTypeStruct));
       if (op == NULL)
           return PyErr_NoMemory();

       op = PyObject_Init(op, &YourTypeStruct);
       if (op == NULL)
           return NULL;

       op->ob_field = value;
       ...
       return op;
   }

   Note that in C++, the use of the new operator usually implies that
   the 1st step is performed automatically for you, so in a C++ class
   constructor you would start directly with PyObject_Init/InitVar. */

#ifdef __cplusplus
}
#endif
#endif /* !Py_OBJIMPL_H */


From Vladimir.Marangozov@inrialpes.fr  Sun Apr 30 20:42:38 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sun, 30 Apr 2000 21:42:38 +0200 (CEST)
Subject: [Patches] PyMem [4/8] - Parser/*
Message-ID: <200004301942.VAA15064@python.inrialpes.fr>


[ Parser/* ]

Nothing new compared to the first round of patches posted previously.
Just mixed mallocs medicine.

--
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--

Disclaimer:

I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------
diff -cr PyCVS/Parser/myreadline.c PyMem/Parser/myreadline.c
*** PyCVS/Parser/myreadline.c	Sat Apr 29 00:24:10 2000
--- PyMem/Parser/myreadline.c	Sat Apr 29 00:36:20 2000
***************
*** 89,95 ****
  	int n;
  	char *p;
  	n = 100;
! 	if ((p = malloc(n)) == NULL)
  		return NULL;
  	fflush(stdout);
  	if (prompt)
--- 89,95 ----
  	int n;
  	char *p;
  	n = 100;
! 	if ((p = PyMem_MALLOC(n)) == NULL)
  		return NULL;
  	fflush(stdout);
  	if (prompt)
***************
*** 99,105 ****
  	case 0: /* Normal case */
  		break;
  	case 1: /* Interrupt */
! 		free(p);
  		return NULL;
  	case -1: /* EOF */
  	case -2: /* Error */
--- 99,105 ----
  	case 0: /* Normal case */
  		break;
  	case 1: /* Interrupt */
! 		PyMem_FREE(p);
  		return NULL;
  	case -1: /* EOF */
  	case -2: /* Error */
***************
*** 117,135 ****
  	n = strlen(p);
  	while (n > 0 && p[n-1] != '\n') {
  		int incr = n+2;
! 		p = realloc(p, n + incr);
  		if (p == NULL)
  			return NULL;
  		if (my_fgets(p+n, incr, stdin) != 0)
  			break;
  		n += strlen(p+n);
  	}
! 	return realloc(p, n+1);
  }
  
  
  /* By initializing this function pointer, systems embedding Python can
!    override the readline function. */
  
  char *(*PyOS_ReadlineFunctionPointer) Py_PROTO((char *));
  
--- 117,137 ----
  	n = strlen(p);
  	while (n > 0 && p[n-1] != '\n') {
  		int incr = n+2;
! 		p = PyMem_REALLOC(p, n + incr);
  		if (p == NULL)
  			return NULL;
  		if (my_fgets(p+n, incr, stdin) != 0)
  			break;
  		n += strlen(p+n);
  	}
! 	return PyMem_REALLOC(p, n+1);
  }
  
  
  /* By initializing this function pointer, systems embedding Python can
!    override the readline function.
! 
!    Note: Python expects in return a buffer allocated with PyMem_Malloc. */
  
  char *(*PyOS_ReadlineFunctionPointer) Py_PROTO((char *));
  
diff -cr PyCVS/Parser/parsetok.c PyMem/Parser/parsetok.c
*** PyCVS/Parser/parsetok.c	Sat Apr 29 00:24:10 2000
--- PyMem/Parser/parsetok.c	Sat Apr 29 00:38:47 2000
***************
*** 192,198 ****
  		err_ret->offset = tok->cur - tok->buf;
  		if (tok->buf != NULL) {
  			int len = tok->inp - tok->buf;
! 			err_ret->text = malloc(len + 1);
  			if (err_ret->text != NULL) {
  				if (len > 0)
  					strncpy(err_ret->text, tok->buf, len);
--- 192,198 ----
  		err_ret->offset = tok->cur - tok->buf;
  		if (tok->buf != NULL) {
  			int len = tok->inp - tok->buf;
! 			err_ret->text = PyMem_NEW(char, len + 1);
  			if (err_ret->text != NULL) {
  				if (len > 0)
  					strncpy(err_ret->text, tok->buf, len);
diff -cr PyCVS/Parser/pgenmain.c PyMem/Parser/pgenmain.c
*** PyCVS/Parser/pgenmain.c	Sat Apr 29 00:24:10 2000
--- PyMem/Parser/pgenmain.c	Sat Apr 29 00:43:20 2000
***************
*** 139,145 ****
  					putc(' ', stderr);
  			}
  			fprintf(stderr, "^\n");
! 			free(err.text);
  		}
  		Py_Exit(1);
  	}
--- 139,145 ----
  					putc(' ', stderr);
  			}
  			fprintf(stderr, "^\n");
! 			PyMem_DEL(err.text);
  		}
  		Py_Exit(1);
  	}
***************
*** 196,202 ****
  	char *prompt;
  {
  	int n = 1000;
! 	char *p = malloc(n);
  	char *q;
  	if (p == NULL)
  		return NULL;
--- 196,202 ----
  	char *prompt;
  {
  	int n = 1000;
! 	char *p = PyMem_MALLOC(n);
  	char *q;
  	if (p == NULL)
  		return NULL;
***************
*** 209,215 ****
  	n = strlen(p);
  	if (n > 0 && p[n-1] != '\n')
  		p[n-1] = '\n';
! 	return realloc(p, n+1);
  }
  
  #ifdef HAVE_STDARG_PROTOTYPES
--- 209,215 ----
  	n = strlen(p);
  	if (n > 0 && p[n-1] != '\n')
  		p[n-1] = '\n';
! 	return PyMem_REALLOC(p, n+1);
  }
  
  #ifdef HAVE_STDARG_PROTOTYPES
diff -cr PyCVS/Parser/tokenizer.c PyMem/Parser/tokenizer.c
*** PyCVS/Parser/tokenizer.c	Sat Apr 29 00:24:10 2000
--- PyMem/Parser/tokenizer.c	Sun Apr 30 08:14:20 2000
***************
*** 219,244 ****
  			if (new == NULL)
  				tok->done = E_INTR;
  			else if (*new == '\0') {
! 				free(new);
  				tok->done = E_EOF;
  			}
  			else if (tok->start != NULL) {
  				int start = tok->start - tok->buf;
  				int oldlen = tok->cur - tok->buf;
  				int newlen = oldlen + strlen(new);
! 				char *buf = realloc(tok->buf, newlen+1);
  				tok->lineno++;
  				if (buf == NULL) {
! 					free(tok->buf);
  					tok->buf = NULL;
! 					free(new);
  					tok->done = E_NOMEM;
  					return EOF;
  				}
  				tok->buf = buf;
  				tok->cur = tok->buf + oldlen;
  				strcpy(tok->buf + oldlen, new);
! 				free(new);
  				tok->inp = tok->buf + newlen;
  				tok->end = tok->inp + 1;
  				tok->start = tok->buf + start;
--- 219,245 ----
  			if (new == NULL)
  				tok->done = E_INTR;
  			else if (*new == '\0') {
! 				PyMem_FREE(new);
  				tok->done = E_EOF;
  			}
  			else if (tok->start != NULL) {
  				int start = tok->start - tok->buf;
  				int oldlen = tok->cur - tok->buf;
  				int newlen = oldlen + strlen(new);
! 				char *buf = tok->buf;
! 				PyMem_RESIZE(buf, char, newlen+1);
  				tok->lineno++;
  				if (buf == NULL) {
! 					PyMem_DEL(tok->buf);
  					tok->buf = NULL;
! 					PyMem_FREE(new);
  					tok->done = E_NOMEM;
  					return EOF;
  				}
  				tok->buf = buf;
  				tok->cur = tok->buf + oldlen;
  				strcpy(tok->buf + oldlen, new);
! 				PyMem_FREE(new);
  				tok->inp = tok->buf + newlen;
  				tok->end = tok->inp + 1;
  				tok->start = tok->buf + start;
***************
*** 246,252 ****
  			else {
  				tok->lineno++;
  				if (tok->buf != NULL)
! 					free(tok->buf);
  				tok->buf = new;
  				tok->cur = tok->buf;
  				tok->inp = strchr(tok->buf, '\0');
--- 247,253 ----
  			else {
  				tok->lineno++;
  				if (tok->buf != NULL)
! 					PyMem_DEL(tok->buf);
  				tok->buf = new;
  				tok->cur = tok->buf;
  				tok->inp = strchr(tok->buf, '\0');


From Vladimir.Marangozov@inrialpes.fr  Sun Apr 30 20:43:38 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sun, 30 Apr 2000 21:43:38 +0200 (CEST)
Subject: [Patches] PyMem [5/8] - Python/*
Message-ID: <200004301943.VAA15088@python.inrialpes.fr>


[ Python/* ]

Nothing new here either.
Malloc cleanup + small syntaxic code improvement in import.c

The docs should mention that PyOS_Readline, when hooked by embedding
apps, expects in return a buffer allocated with PyMem_Malloc, because
this buffer is subsequently released by Python with PyMem_FREE/DEL.

--
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--

Disclaimer:

I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------
diff -cr PyCVS/Python/bltinmodule.c PyMem/Python/bltinmodule.c
*** PyCVS/Python/bltinmodule.c	Sat Apr 29 00:24:10 2000
--- PyMem/Python/bltinmodule.c	Sat Apr 29 00:50:10 2000
***************
*** 1875,1881 ****
  		else { /* strip trailing '\n' */
  			result = PyString_FromStringAndSize(s, strlen(s)-1);
  		}
! 		free(s);
  		return result;
  	}
  	if (v != NULL) {
--- 1875,1881 ----
  		else { /* strip trailing '\n' */
  			result = PyString_FromStringAndSize(s, strlen(s)-1);
  		}
! 		PyMem_FREE(s);
  		return result;
  	}
  	if (v != NULL) {
diff -cr PyCVS/Python/ceval.c PyMem/Python/ceval.c
*** PyCVS/Python/ceval.c	Sat Apr 29 00:24:10 2000
--- PyMem/Python/ceval.c	Sat Apr 29 00:54:34 2000
***************
*** 2558,2564 ****
  		class);
  	
  	Py_DECREF(arg);
! 	PyMem_XDEL(k);
  	
  	return result;
  }
--- 2558,2565 ----
  		class);
  	
  	Py_DECREF(arg);
! 	if (k != NULL)
! 		PyMem_DEL(k);
  	
  	return result;
  }
diff -cr PyCVS/Python/compile.c PyMem/Python/compile.c
*** PyCVS/Python/compile.c	Sat Apr 29 00:24:10 2000
--- PyMem/Python/compile.c	Sat Apr 29 00:57:18 2000
***************
*** 112,118 ****
  	Py_XDECREF(co->co_filename);
  	Py_XDECREF(co->co_name);
  	Py_XDECREF(co->co_lnotab);
! 	PyMem_DEL(co);
  }
  
  static PyObject *
--- 112,118 ----
  	Py_XDECREF(co->co_filename);
  	Py_XDECREF(co->co_name);
  	Py_XDECREF(co->co_lnotab);
! 	PyObject_DEL(co);
  }
  
  static PyObject *
diff -cr PyCVS/Python/import.c PyMem/Python/import.c
*** PyCVS/Python/import.c	Sat Apr 29 00:24:10 2000
--- PyMem/Python/import.c	Sat Apr 29 01:06:21 2000
***************
*** 119,125 ****
  		++countD;
  	for (scan = _PyImport_StandardFiletab; scan->suffix != NULL; ++scan)
  		++countS;
! 	filetab = malloc((countD + countS + 1) * sizeof(struct filedescr));
  	memcpy(filetab, _PyImport_DynLoadFiletab,
  	       countD * sizeof(struct filedescr));
  	memcpy(filetab + countD, _PyImport_StandardFiletab,
--- 119,125 ----
  		++countD;
  	for (scan = _PyImport_StandardFiletab; scan->suffix != NULL; ++scan)
  		++countS;
! 	filetab = PyMem_NEW(struct filedescr, countD + countS + 1);
  	memcpy(filetab, _PyImport_DynLoadFiletab,
  	       countD * sizeof(struct filedescr));
  	memcpy(filetab + countD, _PyImport_StandardFiletab,
***************
*** 2386,2395 ****
  }
  
  
! /* API for embedding applications that want to add their own entries to the
!    table of built-in modules.  This should normally be called *before*
!    Py_Initialize().  When the malloc() or realloc() call fails, -1 is returned
!    and the existing table is unchanged.
  
     After a similar function by Just van Rossum. */
  
--- 2386,2395 ----
  }
  
  
! /* API for embedding applications that want to add their own entries
!    to the table of built-in modules.  This should normally be called
!    *before* Py_Initialize().  When the table resize fails, -1 is
!    returned and the existing table is unchanged.
  
     After a similar function by Just van Rossum. */
  
***************
*** 2410,2419 ****
  		;
  
  	/* Allocate new memory for the combined table */
! 	if (our_copy == NULL)
! 		p = malloc((i+n+1) * sizeof(struct _inittab));
! 	else
! 		p = realloc(our_copy, (i+n+1) * sizeof(struct _inittab));
  	if (p == NULL)
  		return -1;
  
--- 2410,2417 ----
  		;
  
  	/* Allocate new memory for the combined table */
! 	p = our_copy;
! 	PyMem_RESIZE(p, struct _inittab, i+n+1);
  	if (p == NULL)
  		return -1;
  
diff -cr PyCVS/Python/marshal.c PyMem/Python/marshal.c
*** PyCVS/Python/marshal.c	Sat Apr 29 00:24:10 2000
--- PyMem/Python/marshal.c	Sun Apr 30 16:49:11 2000
***************
*** 514,530 ****
  			PyErr_SetString(PyExc_ValueError, "bad marshal data");
  			return NULL;
  		}
! 		buffer = (char *)Py_Malloc(n);
  		if (buffer == NULL)
! 		    return NULL;
  		if (r_string(buffer, (int)n, p) != n) {
! 			free(buffer);
  			PyErr_SetString(PyExc_EOFError,
  				"EOF read where object expected");
  			return NULL;
  		}
  		v = PyUnicode_DecodeUTF8(buffer, n, NULL);
! 		free(buffer);
  		return v;
  	    }
  	    
--- 514,530 ----
  			PyErr_SetString(PyExc_ValueError, "bad marshal data");
  			return NULL;
  		}
! 		buffer = PyMem_NEW(char, n);
  		if (buffer == NULL)
! 			return PyErr_NoMemory();
  		if (r_string(buffer, (int)n, p) != n) {
! 			PyMem_DEL(buffer);
  			PyErr_SetString(PyExc_EOFError,
  				"EOF read where object expected");
  			return NULL;
  		}
  		v = PyUnicode_DecodeUTF8(buffer, n, NULL);
! 		PyMem_DEL(buffer);
  		return v;
  	    }
  	    
diff -cr PyCVS/Python/pythonrun.c PyMem/Python/pythonrun.c
*** PyCVS/Python/pythonrun.c	Sat Apr 29 00:24:10 2000
--- PyMem/Python/pythonrun.c	Sat Apr 29 01:07:45 2000
***************
*** 542,548 ****
  	if (n == NULL) {
  		if (err.error == E_EOF) {
  			if (err.text)
! 				free(err.text);
  			return E_EOF;
  		}
  		err_input(&err);
--- 542,548 ----
  	if (n == NULL) {
  		if (err.error == E_EOF) {
  			if (err.text)
! 				PyMem_DEL(err.text);
  			return E_EOF;
  		}
  		err_input(&err);
***************
*** 1008,1014 ****
  	v = Py_BuildValue("(ziiz)", err->filename,
  			    err->lineno, err->offset, err->text);
  	if (err->text != NULL) {
! 		free(err->text);
  		err->text = NULL;
  	}
  	switch (err->error) {
--- 1008,1014 ----
  	v = Py_BuildValue("(ziiz)", err->filename,
  			    err->lineno, err->offset, err->text);
  	if (err->text != NULL) {
! 		PyMem_DEL(err->text);
  		err->text = NULL;
  	}
  	switch (err->error) {
diff -cr PyCVS/Python/traceback.c PyMem/Python/traceback.c
*** PyCVS/Python/traceback.c	Sat Apr 29 00:24:10 2000
--- PyMem/Python/traceback.c	Sat Apr 29 01:09:25 2000
***************
*** 71,77 ****
  	Py_TRASHCAN_SAFE_BEGIN(tb)
  	Py_XDECREF(tb->tb_next);
  	Py_XDECREF(tb->tb_frame);
! 	PyMem_DEL(tb);
  	Py_TRASHCAN_SAFE_END(tb)
  }
  
--- 71,77 ----
  	Py_TRASHCAN_SAFE_BEGIN(tb)
  	Py_XDECREF(tb->tb_next);
  	Py_XDECREF(tb->tb_frame);
! 	PyObject_DEL(tb);
  	Py_TRASHCAN_SAFE_END(tb)
  }
  


From Vladimir.Marangozov@inrialpes.fr  Sun Apr 30 20:44:28 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sun, 30 Apr 2000 21:44:28 +0200 (CEST)
Subject: [Patches] PyMem [6/8] - Objects/*
Message-ID: <200004301944.VAA15111@python.inrialpes.fr>


[ Objects/* ]

Major interventions w.r.t the new interfaces.

All builtin objects use the macro versions of the Object memory API.

Object constructors that inline PyObject_New/NewVar have been marked
as such. Object initializations with _Py_NewReference and the like
have been hidden (at best) with PyObject_INIT/INIT_VAR.

The functions for all memory APIs are implemented in object.c

--
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--

Disclaimer:

I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------
diff -cr PyCVS/Objects/bufferobject.c PyMem/Objects/bufferobject.c
*** PyCVS/Objects/bufferobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/bufferobject.c	Sat Apr 29 23:28:21 2000
***************
*** 188,198 ****
  				"size must be zero or positive");
  		return NULL;
  	}
! 	b = (PyBufferObject *)malloc(sizeof(*b) + size);
  	if ( b == NULL )
  		return PyErr_NoMemory();
! 	b->ob_type = &PyBuffer_Type;
! 	_Py_NewReference((PyObject *)b);
  
  	b->b_base = NULL;
  	b->b_ptr = (void *)(b + 1);
--- 188,198 ----
  				"size must be zero or positive");
  		return NULL;
  	}
! 	/* PyObject_New is inlined */
! 	b = (PyBufferObject *) PyObject_MALLOC(sizeof(*b) + size);
  	if ( b == NULL )
  		return PyErr_NoMemory();
! 	PyObject_INIT((PyObject *)b, &PyBuffer_Type);
  
  	b->b_base = NULL;
  	b->b_ptr = (void *)(b + 1);
***************
*** 212,218 ****
  	PyBufferObject *self;
  {
  	Py_XDECREF(self->b_base);
! 	free((void *)self);
  }
  
  static int
--- 212,218 ----
  	PyBufferObject *self;
  {
  	Py_XDECREF(self->b_base);
! 	PyObject_DEL(self);
  }
  
  static int
diff -cr PyCVS/Objects/classobject.c PyMem/Objects/classobject.c
*** PyCVS/Objects/classobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/classobject.c	Sun Apr 30 07:25:15 2000
***************
*** 147,153 ****
  	Py_XDECREF(op->cl_getattr);
  	Py_XDECREF(op->cl_setattr);
  	Py_XDECREF(op->cl_delattr);
! 	free((ANY *)op);
  }
  
  static PyObject *
--- 147,153 ----
  	Py_XDECREF(op->cl_getattr);
  	Py_XDECREF(op->cl_setattr);
  	Py_XDECREF(op->cl_delattr);
! 	PyObject_DEL(op);
  }
  
  static PyObject *
***************
*** 561,567 ****
  #endif /* Py_TRACE_REFS */
  	Py_DECREF(inst->in_class);
  	Py_XDECREF(inst->in_dict);
! 	free((ANY *)inst);
  }
  
  static PyObject *
--- 561,567 ----
  #endif /* Py_TRACE_REFS */
  	Py_DECREF(inst->in_class);
  	Py_XDECREF(inst->in_dict);
! 	PyObject_DEL(inst);
  }
  
  static PyObject *
***************
*** 1498,1505 ****
  	im = free_list;
  	if (im != NULL) {
  		free_list = (PyMethodObject *)(im->im_self);
! 		im->ob_type = &PyMethod_Type;
! 		_Py_NewReference((PyObject *)im);
  	}
  	else {
  		im = PyObject_NEW(PyMethodObject, &PyMethod_Type);
--- 1498,1504 ----
  	im = free_list;
  	if (im != NULL) {
  		free_list = (PyMethodObject *)(im->im_self);
! 		PyObject_INIT(im, &PyMethod_Type);
  	}
  	else {
  		im = PyObject_NEW(PyMethodObject, &PyMethod_Type);
***************
*** 1691,1698 ****
  PyMethod_Fini()
  {
  	while (free_list) {
! 		PyMethodObject *v = free_list;
! 		free_list = (PyMethodObject *)(v->im_self);
! 		PyMem_DEL(v);
  	}
  }
--- 1690,1697 ----
  PyMethod_Fini()
  {
  	while (free_list) {
! 		PyMethodObject *im = free_list;
! 		free_list = (PyMethodObject *)(im->im_self);
! 		PyObject_DEL(im);
  	}
  }
diff -cr PyCVS/Objects/cobject.c PyMem/Objects/cobject.c
*** PyCVS/Objects/cobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/cobject.c	Sat Apr 29 23:31:09 2000
***************
*** 151,157 ****
  	    else
  	          (self->destructor)(self->cobject);
  	  }
! 	PyMem_DEL(self);
  }
  
  
--- 151,157 ----
  	    else
  	          (self->destructor)(self->cobject);
  	  }
! 	PyObject_DEL(self);
  }
  
  
diff -cr PyCVS/Objects/complexobject.c PyMem/Objects/complexobject.c
*** PyCVS/Objects/complexobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/complexobject.c	Sat Apr 29 23:31:23 2000
***************
*** 166,178 ****
  PyComplex_FromCComplex(cval)
  	Py_complex cval;
  {
! 	register PyComplexObject *op =
! 		(PyComplexObject *) malloc(sizeof(PyComplexObject));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	op->ob_type = &PyComplex_Type;
  	op->cval = cval;
- 	_Py_NewReference((PyObject *)op);
  	return (PyObject *) op;
  }
  
--- 166,179 ----
  PyComplex_FromCComplex(cval)
  	Py_complex cval;
  {
! 	register PyComplexObject *op;
! 
! 	/* PyObject_New is inlined */
! 	op = (PyComplexObject *) PyObject_MALLOC(sizeof(PyComplexObject));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	PyObject_INIT(op, &PyComplex_Type);
  	op->cval = cval;
  	return (PyObject *) op;
  }
  
***************
*** 226,232 ****
  complex_dealloc(op)
  	PyObject *op;
  {
! 	PyMem_DEL(op);
  }
  
  
--- 227,233 ----
  complex_dealloc(op)
  	PyObject *op;
  {
! 	PyObject_DEL(op);
  }
  
  
diff -cr PyCVS/Objects/dictobject.c PyMem/Objects/dictobject.c
*** PyCVS/Objects/dictobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/dictobject.c	Sun Apr 30 07:27:21 2000
***************
*** 277,283 ****
  			break;
  		}
  	}
! 	newtable = (dictentry *) malloc(sizeof(dictentry) * newsize);
  	if (newtable == NULL) {
  		PyErr_NoMemory();
  		return -1;
--- 277,283 ----
  			break;
  		}
  	}
! 	newtable = PyMem_NEW(dictentry, newsize);
  	if (newtable == NULL) {
  		PyErr_NoMemory();
  		return -1;
***************
*** 301,307 ****
  		}
  	}
  
! 	PyMem_XDEL(oldtable);
  	return 0;
  }
  
--- 301,308 ----
  		}
  	}
  
! 	if (oldtable != NULL)
! 		PyMem_DEL(oldtable);
  	return 0;
  }
  
***************
*** 488,495 ****
  			Py_DECREF(ep->me_value);
  		}
  	}
! 	PyMem_XDEL(mp->ma_table);
! 	PyMem_DEL(mp);
  	Py_TRASHCAN_SAFE_END(mp)
  }
  
--- 489,497 ----
  			Py_DECREF(ep->me_value);
  		}
  	}
! 	if (mp->ma_table != NULL)
! 		PyMem_DEL(mp->ma_table);
! 	PyObject_DEL(mp);
  	Py_TRASHCAN_SAFE_END(mp)
  }
  
diff -cr PyCVS/Objects/fileobject.c PyMem/Objects/fileobject.c
*** PyCVS/Objects/fileobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/fileobject.c	Sat Apr 29 23:41:14 2000
***************
*** 215,221 ****
  	if (f->f_mode != NULL) {
  		Py_DECREF(f->f_mode);
  	}
! 	free((char *)f);
  }
  
  static PyObject *
--- 215,221 ----
  	if (f->f_mode != NULL) {
  		Py_DECREF(f->f_mode);
  	}
! 	PyObject_DEL(f);
  }
  
  static PyObject *
diff -cr PyCVS/Objects/floatobject.c PyMem/Objects/floatobject.c
*** PyCVS/Objects/floatobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/floatobject.c	Sun Apr 30 09:22:23 2000
***************
*** 98,106 ****
  #define BHEAD_SIZE	8	/* Enough for a 64-bit pointer */
  #define N_FLOATOBJECTS	((BLOCK_SIZE - BHEAD_SIZE) / sizeof(PyFloatObject))
  
- #define PyMem_MALLOC	malloc
- #define PyMem_FREE	free
- 
  struct _floatblock {
  	struct _floatblock *next;
  	PyFloatObject objects[N_FLOATOBJECTS];
--- 98,103 ----
***************
*** 115,123 ****
  fill_free_list()
  {
  	PyFloatObject *p, *q;
! 	p = (PyFloatObject *)PyMem_MALLOC(sizeof(PyFloatBlock));
  	if (p == NULL)
! 		return (PyFloatObject *)PyErr_NoMemory();
  	((PyFloatBlock *)p)->next = block_list;
  	block_list = (PyFloatBlock *)p;
  	p = &((PyFloatBlock *)p)->objects[0];
--- 112,121 ----
  fill_free_list()
  {
  	PyFloatObject *p, *q;
! 	/* XXX Float blocks escape the object heap. Use PyObject_MALLOC ??? */
! 	p = (PyFloatObject *) PyMem_MALLOC(sizeof(PyFloatBlock));
  	if (p == NULL)
! 		return (PyFloatObject *) PyErr_NoMemory();
  	((PyFloatBlock *)p)->next = block_list;
  	block_list = (PyFloatBlock *)p;
  	p = &((PyFloatBlock *)p)->objects[0];
***************
*** 141,151 ****
  		if ((free_list = fill_free_list()) == NULL)
  			return NULL;
  	}
  	op = free_list;
  	free_list = (PyFloatObject *)op->ob_type;
! 	op->ob_type = &PyFloat_Type;
  	op->ob_fval = fval;
- 	_Py_NewReference((PyObject *)op);
  	return (PyObject *) op;
  }
  
--- 139,149 ----
  		if ((free_list = fill_free_list()) == NULL)
  			return NULL;
  	}
+ 	/* PyObject_New is inlined */
  	op = free_list;
  	free_list = (PyFloatObject *)op->ob_type;
! 	PyObject_INIT(op, &PyFloat_Type);
  	op->ob_fval = fval;
  	return (PyObject *) op;
  }
  
***************
*** 779,785 ****
  			}
  		}
  		else {
! 			PyMem_FREE(list);
  			bf++;
  		}
  		fsum += frem;
--- 777,783 ----
  			}
  		}
  		else {
! 			PyMem_FREE(list); /* XXX PyObject_FREE ??? */
  			bf++;
  		}
  		fsum += frem;
diff -cr PyCVS/Objects/frameobject.c PyMem/Objects/frameobject.c
*** PyCVS/Objects/frameobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/frameobject.c	Sun Apr 30 07:33:22 2000
***************
*** 180,207 ****
  	if (builtins != NULL && !PyDict_Check(builtins))
  		builtins = NULL;
  	if (free_list == NULL) {
  		f = (PyFrameObject *)
! 			malloc(sizeof(PyFrameObject) +
! 			       extras*sizeof(PyObject *));
  		if (f == NULL)
  			return (PyFrameObject *)PyErr_NoMemory();
! 		f->ob_type = &PyFrame_Type;
! 		_Py_NewReference((PyObject *)f);
  	}
  	else {
  		f = free_list;
  		free_list = free_list->f_back;
  		if (f->f_nlocals + f->f_stacksize < extras) {
  			f = (PyFrameObject *)
! 				realloc(f, sizeof(PyFrameObject) +
! 					extras*sizeof(PyObject *));
  			if (f == NULL)
  				return (PyFrameObject *)PyErr_NoMemory();
  		}
  		else
  			extras = f->f_nlocals + f->f_stacksize;
! 		f->ob_type = &PyFrame_Type;
! 		_Py_NewReference((PyObject *)f);
  	}
  	if (builtins == NULL) {
  		/* No builtins!  Make up a minimal one. */
--- 180,206 ----
  	if (builtins != NULL && !PyDict_Check(builtins))
  		builtins = NULL;
  	if (free_list == NULL) {
+ 		/* PyObject_New is inlined */
  		f = (PyFrameObject *)
! 			PyObject_MALLOC(sizeof(PyFrameObject) +
! 					extras*sizeof(PyObject *));
  		if (f == NULL)
  			return (PyFrameObject *)PyErr_NoMemory();
! 		PyObject_INIT(f, &PyFrame_Type);
  	}
  	else {
  		f = free_list;
  		free_list = free_list->f_back;
  		if (f->f_nlocals + f->f_stacksize < extras) {
  			f = (PyFrameObject *)
! 				PyObject_REALLOC(f, sizeof(PyFrameObject) +
! 						 extras*sizeof(PyObject *));
  			if (f == NULL)
  				return (PyFrameObject *)PyErr_NoMemory();
  		}
  		else
  			extras = f->f_nlocals + f->f_stacksize;
! 		PyObject_INIT(f, &PyFrame_Type);
  	}
  	if (builtins == NULL) {
  		/* No builtins!  Make up a minimal one. */
***************
*** 376,381 ****
  	while (free_list != NULL) {
  		PyFrameObject *f = free_list;
  		free_list = free_list->f_back;
! 		PyMem_DEL(f);
  	}
  }
--- 375,380 ----
  	while (free_list != NULL) {
  		PyFrameObject *f = free_list;
  		free_list = free_list->f_back;
! 		PyObject_DEL(f);
  	}
  }
diff -cr PyCVS/Objects/funcobject.c PyMem/Objects/funcobject.c
*** PyCVS/Objects/funcobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/funcobject.c	Sat Apr 29 23:55:01 2000
***************
*** 191,197 ****
  	Py_DECREF(op->func_name);
  	Py_XDECREF(op->func_defaults);
  	Py_XDECREF(op->func_doc);
! 	PyMem_DEL(op);
  }
  
  static PyObject*
--- 191,197 ----
  	Py_DECREF(op->func_name);
  	Py_XDECREF(op->func_defaults);
  	Py_XDECREF(op->func_doc);
! 	PyObject_DEL(op);
  }
  
  static PyObject*
diff -cr PyCVS/Objects/intobject.c PyMem/Objects/intobject.c
*** PyCVS/Objects/intobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/intobject.c	Sun Apr 30 09:23:01 2000
***************
*** 94,102 ****
  #define BHEAD_SIZE	8	/* Enough for a 64-bit pointer */
  #define N_INTOBJECTS	((BLOCK_SIZE - BHEAD_SIZE) / sizeof(PyIntObject))
  
- #define PyMem_MALLOC	malloc
- #define PyMem_FREE	free
- 
  struct _intblock {
  	struct _intblock *next;
  	PyIntObject objects[N_INTOBJECTS];
--- 94,99 ----
***************
*** 111,119 ****
  fill_free_list()
  {
  	PyIntObject *p, *q;
! 	p = (PyIntObject *)PyMem_MALLOC(sizeof(PyIntBlock));
  	if (p == NULL)
! 		return (PyIntObject *)PyErr_NoMemory();
  	((PyIntBlock *)p)->next = block_list;
  	block_list = (PyIntBlock *)p;
  	p = &((PyIntBlock *)p)->objects[0];
--- 108,117 ----
  fill_free_list()
  {
  	PyIntObject *p, *q;
! 	/* XXX Int blocks escape the object heap. Use PyObject_MALLOC ??? */
! 	p = (PyIntObject *) PyMem_MALLOC(sizeof(PyIntBlock));
  	if (p == NULL)
! 		return (PyIntObject *) PyErr_NoMemory();
  	((PyIntBlock *)p)->next = block_list;
  	block_list = (PyIntBlock *)p;
  	p = &((PyIntBlock *)p)->objects[0];
***************
*** 164,174 ****
  		if ((free_list = fill_free_list()) == NULL)
  			return NULL;
  	}
  	v = free_list;
  	free_list = (PyIntObject *)v->ob_type;
! 	v->ob_type = &PyInt_Type;
  	v->ob_ival = ival;
- 	_Py_NewReference((PyObject *)v);
  #if NSMALLNEGINTS + NSMALLPOSINTS > 0
  	if (-NSMALLNEGINTS <= ival && ival < NSMALLPOSINTS) {
  		/* save this one for a following allocation */
--- 162,172 ----
  		if ((free_list = fill_free_list()) == NULL)
  			return NULL;
  	}
+ 	/* PyObject_New is inlined */
  	v = free_list;
  	free_list = (PyIntObject *)v->ob_type;
! 	PyObject_INIT(v, &PyInt_Type);
  	v->ob_ival = ival;
  #if NSMALLNEGINTS + NSMALLPOSINTS > 0
  	if (-NSMALLNEGINTS <= ival && ival < NSMALLPOSINTS) {
  		/* save this one for a following allocation */
***************
*** 933,939 ****
  			}
  		}
  		else {
! 			PyMem_FREE(list);
  			bf++;
  		}
  		isum += irem;
--- 931,937 ----
  			}
  		}
  		else {
! 			PyMem_FREE(list); /* XXX PyObject_FREE ??? */
  			bf++;
  		}
  		isum += irem;
diff -cr PyCVS/Objects/listobject.c PyMem/Objects/listobject.c
*** PyCVS/Objects/listobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/listobject.c	Sun Apr 30 00:48:14 2000
***************
*** 70,76 ****
  	if (nbytes / sizeof(PyObject *) != (size_t)size) {
  		return PyErr_NoMemory();
  	}
! 	op = (PyListObject *) malloc(sizeof(PyListObject));
  	if (op == NULL) {
  		return PyErr_NoMemory();
  	}
--- 70,77 ----
  	if (nbytes / sizeof(PyObject *) != (size_t)size) {
  		return PyErr_NoMemory();
  	}
! 	/* PyObject_NewVar is inlined */
! 	op = (PyListObject *) PyObject_MALLOC(sizeof(PyListObject));
  	if (op == NULL) {
  		return PyErr_NoMemory();
  	}
***************
*** 78,94 ****
  		op->ob_item = NULL;
  	}
  	else {
! 		op->ob_item = (PyObject **) malloc(nbytes);
  		if (op->ob_item == NULL) {
! 			free((ANY *)op);
  			return PyErr_NoMemory();
  		}
  	}
! 	op->ob_type = &PyList_Type;
! 	op->ob_size = size;
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
- 	_Py_NewReference((PyObject *)op);
  	return (PyObject *) op;
  }
  
--- 79,93 ----
  		op->ob_item = NULL;
  	}
  	else {
! 		op->ob_item = (PyObject **) PyMem_MALLOC(nbytes);
  		if (op->ob_item == NULL) {
! 			PyObject_FREE(op);
  			return PyErr_NoMemory();
  		}
  	}
! 	PyObject_INIT_VAR(op, &PyList_Type, size);
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
  	return (PyObject *) op;
  }
  
***************
*** 225,233 ****
  		while (--i >= 0) {
  			Py_XDECREF(op->ob_item[i]);
  		}
! 		free((ANY *)op->ob_item);
  	}
! 	free((ANY *)op);
  	Py_TRASHCAN_SAFE_END(op)
  }
  
--- 224,232 ----
  		while (--i >= 0) {
  			Py_XDECREF(op->ob_item[i]);
  		}
! 		PyMem_FREE(op->ob_item);
  	}
! 	PyObject_DEL(op);
  	Py_TRASHCAN_SAFE_END(op)
  }
  
***************
*** 501,507 ****
  	else { /* Insert d items; recycle ihigh-ilow items */
  		NRESIZE(item, PyObject *, a->ob_size + d);
  		if (item == NULL) {
! 			PyMem_XDEL(recycle);
  			PyErr_NoMemory();
  			return -1;
  		}
--- 500,507 ----
  	else { /* Insert d items; recycle ihigh-ilow items */
  		NRESIZE(item, PyObject *, a->ob_size + d);
  		if (item == NULL) {
! 			if (recycle != NULL)
! 				PyMem_DEL(recycle);
  			PyErr_NoMemory();
  			return -1;
  		}
diff -cr PyCVS/Objects/longobject.c PyMem/Objects/longobject.c
*** PyCVS/Objects/longobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/longobject.c	Sun Apr 30 00:13:09 2000
***************
*** 995,1001 ****
  long_dealloc(v)
  	PyObject *v;
  {
! 	PyMem_DEL(v);
  }
  
  static PyObject *
--- 995,1001 ----
  long_dealloc(v)
  	PyObject *v;
  {
! 	PyObject_DEL(v);
  }
  
  static PyObject *
diff -cr PyCVS/Objects/methodobject.c PyMem/Objects/methodobject.c
*** PyCVS/Objects/methodobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/methodobject.c	Sun Apr 30 00:15:12 2000
***************
*** 46,53 ****
  	op = free_list;
  	if (op != NULL) {
  		free_list = (PyCFunctionObject *)(op->m_self);
! 		op->ob_type = &PyCFunction_Type;
! 		_Py_NewReference((PyObject *)op);
  	}
  	else {
  		op = PyObject_NEW(PyCFunctionObject, &PyCFunction_Type);
--- 46,52 ----
  	op = free_list;
  	if (op != NULL) {
  		free_list = (PyCFunctionObject *)(op->m_self);
! 		PyObject_INIT(op, &PyCFunction_Type);
  	}
  	else {
  		op = PyObject_NEW(PyCFunctionObject, &PyCFunction_Type);
***************
*** 288,293 ****
  	while (free_list) {
  		PyCFunctionObject *v = free_list;
  		free_list = (PyCFunctionObject *)(v->m_self);
! 		PyMem_DEL(v);
  	}
  }
--- 287,292 ----
  	while (free_list) {
  		PyCFunctionObject *v = free_list;
  		free_list = (PyCFunctionObject *)(v->m_self);
! 		PyObject_DEL(v);
  	}
  }
diff -cr PyCVS/Objects/moduleobject.c PyMem/Objects/moduleobject.c
*** PyCVS/Objects/moduleobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/moduleobject.c	Sun Apr 30 00:16:34 2000
***************
*** 170,176 ****
  		_PyModule_Clear((PyObject *)m);
  		Py_DECREF(m->md_dict);
  	}
! 	free((char *)m);
  }
  
  static PyObject *
--- 170,176 ----
  		_PyModule_Clear((PyObject *)m);
  		Py_DECREF(m->md_dict);
  	}
! 	PyObject_DEL(m);
  }
  
  static PyObject *
diff -cr PyCVS/Objects/object.c PyMem/Objects/object.c
*** PyCVS/Objects/object.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/object.c	Sun Apr 30 07:34:17 2000
***************
*** 112,161 ****
  }
  #endif
  
- #ifndef MS_COREDLL
  PyObject *
! _PyObject_New(tp)
! 	PyTypeObject *tp;
! #else
! PyObject *
! _PyObject_New(tp,op)
! 	PyTypeObject *tp;
  	PyObject *op;
! #endif
  {
! #ifndef MS_COREDLL
! 	PyObject *op = (PyObject *) malloc(tp->tp_basicsize);
! #endif
! 	if (op == NULL)
! 		return PyErr_NoMemory();
! 	op->ob_type = tp;
  	_Py_NewReference(op);
  	return op;
  }
  
- #ifndef MS_COREDLL
  PyVarObject *
! _PyObject_NewVar(tp, size)
  	PyTypeObject *tp;
  	int size;
! #else
  PyVarObject *
! _PyObject_NewVar(tp, size, op)
  	PyTypeObject *tp;
  	int size;
- 	PyVarObject *op;
- #endif
  {
! #ifndef MS_COREDLL
! 	PyVarObject *op = (PyVarObject *)
! 		malloc(tp->tp_basicsize + size * tp->tp_itemsize);
! #endif
  	if (op == NULL)
  		return (PyVarObject *)PyErr_NoMemory();
! 	op->ob_type = tp;
! 	op->ob_size = size;
! 	_Py_NewReference((PyObject *)op);
! 	return op;
  }
  
  int
--- 112,179 ----
  }
  #endif
  
  PyObject *
! PyObject_Init(op, tp)
  	PyObject *op;
! 	PyTypeObject *tp;
  {
! 	if (op == NULL) {
! 		PyErr_SetString(PyExc_SystemError,
! 				"NULL object passed to PyObject_Init");
! 		return op;
!   	}
! 	/* Any changes should be reflected in PyObject_INIT (objimpl.h) */
  	_Py_NewReference(op);
+ 	op->ob_type = tp;
  	return op;
  }
  
  PyVarObject *
! PyObject_InitVar(op, tp, size)
! 	PyVarObject *op;
  	PyTypeObject *tp;
  	int size;
! {
! 	if (op == NULL) {
! 		PyErr_SetString(PyExc_SystemError,
! 				"NULL object passed to PyObject_InitVar");
! 		return op;
! 	}
! 	/* Any changes should be reflected in PyObject_INIT_VAR */
! 	_Py_NewReference(op);
! 	op->ob_type = tp;
! 	op->ob_size = size;
! 	return op;
! }
! 
! PyObject *
! _PyObject_New(tp)
! 	PyTypeObject *tp;
! {
! 	PyObject *op;
! 	op = (PyObject *) PyObject_MALLOC(_PyObject_SIZE(tp));
! 	if (op == NULL)
! 		return PyErr_NoMemory();
! 	return PyObject_INIT(op, tp);
! }
! 
  PyVarObject *
! _PyObject_NewVar(tp, size)
  	PyTypeObject *tp;
  	int size;
  {
! 	PyVarObject *op;
! 	op = (PyVarObject *) PyObject_MALLOC(_PyObject_VAR_SIZE(tp, size));
  	if (op == NULL)
  		return (PyVarObject *)PyErr_NoMemory();
! 	return PyObject_INIT_VAR(op, tp, size);
! }
! 
! void
! _PyObject_Del(op)
! 	PyObject *op;
! {
! 	PyObject_FREE(op);
  }
  
  int
***************
*** 888,917 ****
  int (*_Py_abstract_hack) Py_FPROTO((PyObject *)) = &PyObject_Length;
  
  
! /* Malloc wrappers (see mymalloc.h) */
! 
! /* The Py_{Malloc,Realloc} wrappers call PyErr_NoMemory() on failure */
  
  ANY *
! Py_Malloc(nbytes)
  	size_t nbytes;
  {
- 	ANY *p;
  #if _PyMem_EXTRA > 0
  	if (nbytes == 0)
  		nbytes = _PyMem_EXTRA;
  #endif
! 	p = malloc(nbytes);
! 	if (p != NULL)
! 		return p;
! 	else {
! 		PyErr_NoMemory();
! 		return NULL;
! 	}
  }
  
  ANY *
! Py_Realloc(p, nbytes)
  	ANY *p;
  	size_t nbytes;
  {
--- 906,926 ----
  int (*_Py_abstract_hack) Py_FPROTO((PyObject *)) = &PyObject_Length;
  
  
! /* Python's malloc wrappers (see mymalloc.h) */
  
  ANY *
! PyMem_Malloc(nbytes)
  	size_t nbytes;
  {
  #if _PyMem_EXTRA > 0
  	if (nbytes == 0)
  		nbytes = _PyMem_EXTRA;
  #endif
! 	return PyMem_MALLOC(nbytes);
  }
  
  ANY *
! PyMem_Realloc(p, nbytes)
  	ANY *p;
  	size_t nbytes;
  {
***************
*** 919,970 ****
  	if (nbytes == 0)
  		nbytes = _PyMem_EXTRA;
  #endif
! 	p = realloc(p, nbytes);
! 	if (p != NULL)
! 		return p;
! 	else {
! 		PyErr_NoMemory();
! 		return NULL;
! 	}
  }
  
  void
! Py_Free(p)
  	ANY *p;
  {
! 	free(p);
  }
  
! /* The PyMem_{Malloc,Realloc} wrappers don't call anything on failure */
  
  ANY *
! PyMem_Malloc(nbytes)
  	size_t nbytes;
  {
! #if _PyMem_EXTRA > 0
! 	if (nbytes == 0)
! 		nbytes = _PyMem_EXTRA;
! #endif
! 	return malloc(nbytes);
  }
  
  ANY *
! PyMem_Realloc(p, nbytes)
  	ANY *p;
  	size_t nbytes;
  {
! #if _PyMem_EXTRA > 0
! 	if (nbytes == 0)
! 		nbytes = _PyMem_EXTRA;
! #endif
! 	return realloc(p, nbytes);
  }
  
  void
! PyMem_Free(p)
  	ANY *p;
  {
! 	free(p);
  }
  
  
--- 928,966 ----
  	if (nbytes == 0)
  		nbytes = _PyMem_EXTRA;
  #endif
! 	return PyMem_REALLOC(p, nbytes);
  }
  
  void
! PyMem_Free(p)
  	ANY *p;
  {
! 	PyMem_FREE(p);
  }
  
! 
! /* Python's object malloc wrappers (see objimpl.h) */
  
  ANY *
! PyObject_Malloc(nbytes)
  	size_t nbytes;
  {
! 	return PyObject_MALLOC(nbytes);
  }
  
  ANY *
! PyObject_Realloc(p, nbytes)
  	ANY *p;
  	size_t nbytes;
  {
! 	return PyObject_REALLOC(p, nbytes);
  }
  
  void
! PyObject_Free(p)
  	ANY *p;
  {
! 	PyObject_FREE(p);
  }
  
  
diff -cr PyCVS/Objects/rangeobject.c PyMem/Objects/rangeobject.c
*** PyCVS/Objects/rangeobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/rangeobject.c	Sun Apr 30 00:33:25 2000
***************
*** 61,67 ****
  range_dealloc(r)
  	rangeobject *r;
  {
! 	PyMem_DEL(r);
  }
  
  static PyObject *
--- 61,67 ----
  range_dealloc(r)
  	rangeobject *r;
  {
! 	PyObject_DEL(r);
  }
  
  static PyObject *
diff -cr PyCVS/Objects/sliceobject.c PyMem/Objects/sliceobject.c
*** PyCVS/Objects/sliceobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/sliceobject.c	Sun Apr 30 00:44:54 2000
***************
*** 57,64 ****
  	PyObject *stop;
  	PyObject *step;
  {
! 	PySliceObject *obj =
! 		(PySliceObject *) PyObject_NEW(PySliceObject, &PySlice_Type);
  
  	if (step == NULL) step = Py_None;
  	Py_INCREF(step);
--- 57,63 ----
  	PyObject *stop;
  	PyObject *step;
  {
! 	PySliceObject *obj = PyObject_NEW(PySliceObject, &PySlice_Type);
  
  	if (step == NULL) step = Py_None;
  	Py_INCREF(step);
***************
*** 115,121 ****
  	Py_DECREF(r->step);
  	Py_DECREF(r->start);
  	Py_DECREF(r->stop);
! 	PyMem_DEL(r);
  }
  
  static PyObject *
--- 114,120 ----
  	Py_DECREF(r->step);
  	Py_DECREF(r->start);
  	Py_DECREF(r->stop);
! 	PyObject_DEL(r);
  }
  
  static PyObject *
diff -cr PyCVS/Objects/stringobject.c PyMem/Objects/stringobject.c
*** PyCVS/Objects/stringobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/stringobject.c	Sun Apr 30 01:02:47 2000
***************
*** 92,110 ****
  		return (PyObject *)op;
  	}
  #endif /* DONT_SHARE_SHORT_STRINGS */
  	op = (PyStringObject *)
! 		malloc(sizeof(PyStringObject) + size * sizeof(char));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	op->ob_type = &PyString_Type;
! 	op->ob_size = size;
  #ifdef CACHE_HASH
  	op->ob_shash = -1;
  #endif
  #ifdef INTERN_STRINGS
  	op->ob_sinterned = NULL;
  #endif
- 	_Py_NewReference((PyObject *)op);
  	if (str != NULL)
  		memcpy(op->ob_sval, str, size);
  	op->ob_sval[size] = '\0';
--- 92,110 ----
  		return (PyObject *)op;
  	}
  #endif /* DONT_SHARE_SHORT_STRINGS */
+ 
+ 	/* PyObject_NewVar is inlined */
  	op = (PyStringObject *)
! 		PyObject_MALLOC(sizeof(PyStringObject) + size * sizeof(char));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	PyObject_INIT_VAR(op, &PyString_Type, size);
  #ifdef CACHE_HASH
  	op->ob_shash = -1;
  #endif
  #ifdef INTERN_STRINGS
  	op->ob_sinterned = NULL;
  #endif
  	if (str != NULL)
  		memcpy(op->ob_sval, str, size);
  	op->ob_sval[size] = '\0';
***************
*** 142,160 ****
  		return (PyObject *)op;
  	}
  #endif /* DONT_SHARE_SHORT_STRINGS */
  	op = (PyStringObject *)
! 		malloc(sizeof(PyStringObject) + size * sizeof(char));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	op->ob_type = &PyString_Type;
! 	op->ob_size = size;
  #ifdef CACHE_HASH
  	op->ob_shash = -1;
  #endif
  #ifdef INTERN_STRINGS
  	op->ob_sinterned = NULL;
  #endif
- 	_Py_NewReference((PyObject *)op);
  	strcpy(op->ob_sval, str);
  #ifndef DONT_SHARE_SHORT_STRINGS
  	if (size == 0) {
--- 142,160 ----
  		return (PyObject *)op;
  	}
  #endif /* DONT_SHARE_SHORT_STRINGS */
+ 
+ 	/* PyObject_NewVar is inlined */
  	op = (PyStringObject *)
! 		PyObject_MALLOC(sizeof(PyStringObject) + size * sizeof(char));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	PyObject_INIT_VAR(op, &PyString_Type, size);
  #ifdef CACHE_HASH
  	op->ob_shash = -1;
  #endif
  #ifdef INTERN_STRINGS
  	op->ob_sinterned = NULL;
  #endif
  	strcpy(op->ob_sval, str);
  #ifndef DONT_SHARE_SHORT_STRINGS
  	if (size == 0) {
***************
*** 172,178 ****
  string_dealloc(op)
  	PyObject *op;
  {
! 	PyMem_DEL(op);
  }
  
  int
--- 172,178 ----
  string_dealloc(op)
  	PyObject *op;
  {
! 	PyObject_DEL(op);
  }
  
  int
***************
*** 307,325 ****
  		return (PyObject *)a;
  	}
  	size = a->ob_size + b->ob_size;
  	op = (PyStringObject *)
! 		malloc(sizeof(PyStringObject) + size * sizeof(char));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	op->ob_type = &PyString_Type;
! 	op->ob_size = size;
  #ifdef CACHE_HASH
  	op->ob_shash = -1;
  #endif
  #ifdef INTERN_STRINGS
  	op->ob_sinterned = NULL;
  #endif
- 	_Py_NewReference((PyObject *)op);
  	memcpy(op->ob_sval, a->ob_sval, (int) a->ob_size);
  	memcpy(op->ob_sval + a->ob_size, b->ob_sval, (int) b->ob_size);
  	op->ob_sval[size] = '\0';
--- 307,324 ----
  		return (PyObject *)a;
  	}
  	size = a->ob_size + b->ob_size;
+ 	/* PyObject_NewVar is inlined */
  	op = (PyStringObject *)
! 		PyObject_MALLOC(sizeof(PyStringObject) + size * sizeof(char));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	PyObject_INIT_VAR(op, &PyString_Type, size);
  #ifdef CACHE_HASH
  	op->ob_shash = -1;
  #endif
  #ifdef INTERN_STRINGS
  	op->ob_sinterned = NULL;
  #endif
  	memcpy(op->ob_sval, a->ob_sval, (int) a->ob_size);
  	memcpy(op->ob_sval + a->ob_size, b->ob_sval, (int) b->ob_size);
  	op->ob_sval[size] = '\0';
***************
*** 342,360 ****
  		Py_INCREF(a);
  		return (PyObject *)a;
  	}
  	op = (PyStringObject *)
! 		malloc(sizeof(PyStringObject) + size * sizeof(char));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	op->ob_type = &PyString_Type;
! 	op->ob_size = size;
  #ifdef CACHE_HASH
  	op->ob_shash = -1;
  #endif
  #ifdef INTERN_STRINGS
  	op->ob_sinterned = NULL;
  #endif
- 	_Py_NewReference((PyObject *)op);
  	for (i = 0; i < size; i += a->ob_size)
  		memcpy(op->ob_sval+i, a->ob_sval, (int) a->ob_size);
  	op->ob_sval[size] = '\0';
--- 341,358 ----
  		Py_INCREF(a);
  		return (PyObject *)a;
  	}
+ 	/* PyObject_NewVar is inlined */
  	op = (PyStringObject *)
! 		PyObject_MALLOC(sizeof(PyStringObject) + size * sizeof(char));
  	if (op == NULL)
  		return PyErr_NoMemory();
! 	PyObject_INIT_VAR(op, &PyString_Type, size);
  #ifdef CACHE_HASH
  	op->ob_shash = -1;
  #endif
  #ifdef INTERN_STRINGS
  	op->ob_sinterned = NULL;
  #endif
  	for (i = 0; i < size; i += a->ob_size)
  		memcpy(op->ob_sval+i, a->ob_sval, (int) a->ob_size);
  	op->ob_sval[size] = '\0';
***************
*** 1498,1504 ****
  		goto return_same;
  	new_len = len + nfound*(sub_len - pat_len);
  
! 	new_s = (char *)malloc(new_len);
  	if (new_s == NULL) return NULL;
  
  	*out_len = new_len;
--- 1496,1502 ----
  		goto return_same;
  	new_len = len + nfound*(sub_len - pat_len);
  
! 	new_s = (char *)PyMem_MALLOC(new_len);
  	if (new_s == NULL) return NULL;
  
  	*out_len = new_len;
***************
*** 1593,1599 ****
  	}
  	else {
  		new = PyString_FromStringAndSize(new_s, out_len);
! 		free(new_s);
  	}
  	return new;
  }
--- 1591,1597 ----
  	}
  	else {
  		new = PyString_FromStringAndSize(new_s, out_len);
! 		PyMem_FREE(new_s);
  	}
  	return new;
  }
***************
*** 2273,2282 ****
  #endif
  	_Py_ForgetReference(v);
  	*pv = (PyObject *)
! 		realloc((char *)v,
  			sizeof(PyStringObject) + newsize * sizeof(char));
  	if (*pv == NULL) {
! 		PyMem_DEL(v);
  		PyErr_NoMemory();
  		return -1;
  	}
--- 2271,2280 ----
  #endif
  	_Py_ForgetReference(v);
  	*pv = (PyObject *)
! 		PyObject_REALLOC((char *)v,
  			sizeof(PyStringObject) + newsize * sizeof(char));
  	if (*pv == NULL) {
! 		PyObject_DEL(v);
  		PyErr_NoMemory();
  		return -1;
  	}
diff -cr PyCVS/Objects/tupleobject.c PyMem/Objects/tupleobject.c
*** PyCVS/Objects/tupleobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/tupleobject.c	Sun Apr 30 01:20:11 2000
***************
*** 77,89 ****
  	{
  		free_tuples[size] = (PyTupleObject *) op->ob_item[0];
  		num_free_tuples[size]--;
  #ifdef COUNT_ALLOCS
  		fast_tuple_allocs++;
  #endif
- #ifdef Py_TRACE_REFS
- 		op->ob_type = &PyTuple_Type;
- 		op->ob_size = size;
- #endif
  	}
  	else
  #endif
--- 77,89 ----
  	{
  		free_tuples[size] = (PyTupleObject *) op->ob_item[0];
  		num_free_tuples[size]--;
+ 		_Py_NewReference((PyObject *)op);
+ #ifdef Py_TRACE_REFS
+ 		PyObject_INIT_VAR(op, &PyTuple_Type, size);
+ #endif
  #ifdef COUNT_ALLOCS
  		fast_tuple_allocs++;
  #endif
  	}
  	else
  #endif
***************
*** 96,112 ****
  		{
  			return PyErr_NoMemory();
  		}
! 		;
! 		op = (PyTupleObject *) malloc(nbytes);
  		if (op == NULL)
  			return PyErr_NoMemory();
  
! 		op->ob_type = &PyTuple_Type;
! 		op->ob_size = size;
  	}
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
- 	_Py_NewReference((PyObject *)op);
  #if MAXSAVESIZE > 0
  	if (size == 0) {
  		free_tuples[0] = op;
--- 96,110 ----
  		{
  			return PyErr_NoMemory();
  		}
! 		/* PyObject_NewVar is inlined */
! 		op = (PyTupleObject *) PyObject_MALLOC(nbytes);
  		if (op == NULL)
  			return PyErr_NoMemory();
  
! 		PyObject_INIT_VAR(op, &PyTuple_Type, size);
  	}
  	for (i = 0; i < size; i++)
  		op->ob_item[i] = NULL;
  #if MAXSAVESIZE > 0
  	if (size == 0) {
  		free_tuples[0] = op;
***************
*** 193,199 ****
  		}
  #endif
  	}
! 	free((ANY *)op);
  done:
  	Py_TRASHCAN_SAFE_END(op)
  }
--- 191,197 ----
  		}
  #endif
  	}
! 	PyObject_DEL(op);
  done:
  	Py_TRASHCAN_SAFE_END(op)
  }
***************
*** 530,540 ****
  #endif		
  	{
  		sv = (PyTupleObject *)
! 			realloc((char *)v,
  				sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
  		*pv = (PyObject *) sv;
  		if (sv == NULL) {
! 			PyMem_DEL(v);
  			PyErr_NoMemory();
  			return -1;
  		}
--- 528,538 ----
  #endif		
  	{
  		sv = (PyTupleObject *)
! 			PyObject_REALLOC((char *)v,
  				sizeof(PyTupleObject) + newsize * sizeof(PyObject *));
  		*pv = (PyObject *) sv;
  		if (sv == NULL) {
! 			PyObject_DEL(v);
  			PyErr_NoMemory();
  			return -1;
  		}
***************
*** 569,575 ****
  		while (p) {
  			q = p;
  			p = (PyTupleObject *)(p->ob_item[0]);
! 			PyMem_DEL(q);
  		}
  	}
  #endif
--- 567,573 ----
  		while (p) {
  			q = p;
  			p = (PyTupleObject *)(p->ob_item[0]);
! 			PyObject_DEL(q);
  		}
  	}
  #endif
diff -cr PyCVS/Objects/unicodeobject.c PyMem/Objects/unicodeobject.c
*** PyCVS/Objects/unicodeobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/unicodeobject.c	Sun Apr 30 08:01:31 2000
***************
*** 200,213 ****
          unicode = unicode_freelist;
          unicode_freelist = *(PyUnicodeObject **)unicode_freelist;
          unicode_freelist_size--;
!         unicode->ob_type = &PyUnicode_Type;
!         _Py_NewReference((PyObject *)unicode);
  	if (unicode->str) {
  	    /* Keep-Alive optimization: we only upsize the buffer,
  	       never downsize it. */
  	    if ((unicode->length < length) &&
  		_PyUnicode_Resize(unicode, length)) {
! 		free(unicode->str);
  		goto onError;
  	    }
  	}
--- 200,212 ----
          unicode = unicode_freelist;
          unicode_freelist = *(PyUnicodeObject **)unicode_freelist;
          unicode_freelist_size--;
! 	PyObject_INIT(unicode, &PyUnicode_Type);
  	if (unicode->str) {
  	    /* Keep-Alive optimization: we only upsize the buffer,
  	       never downsize it. */
  	    if ((unicode->length < length) &&
  		_PyUnicode_Resize(unicode, length)) {
! 		PyMem_DEL(unicode->str);
  		goto onError;
  	    }
  	}
***************
*** 233,239 ****
  
   onError:
      _Py_ForgetReference((PyObject *)unicode);
!     PyMem_DEL(unicode);
      return NULL;
  }
  
--- 232,238 ----
  
   onError:
      _Py_ForgetReference((PyObject *)unicode);
!     PyObject_DEL(unicode);
      return NULL;
  }
  
***************
*** 243,249 ****
      if (unicode_freelist_size < MAX_UNICODE_FREELIST_SIZE) {
          /* Keep-Alive optimization */
  	if (unicode->length >= KEEPALIVE_SIZE_LIMIT) {
! 	    free(unicode->str);
  	    unicode->str = NULL;
  	    unicode->length = 0;
  	}
--- 242,248 ----
      if (unicode_freelist_size < MAX_UNICODE_FREELIST_SIZE) {
          /* Keep-Alive optimization */
  	if (unicode->length >= KEEPALIVE_SIZE_LIMIT) {
! 	    PyMem_DEL(unicode->str);
  	    unicode->str = NULL;
  	    unicode->length = 0;
  	}
***************
*** 257,265 ****
          unicode_freelist_size++;
      }
      else {
! 	free(unicode->str);
  	Py_XDECREF(unicode->utf8str);
!         PyMem_DEL(unicode);
      }
  }
  
--- 256,264 ----
          unicode_freelist_size++;
      }
      else {
! 	PyMem_DEL(unicode->str);
  	Py_XDECREF(unicode->utf8str);
! 	PyObject_DEL(unicode);
      }
  }
  
***************
*** 4662,4670 ****
  	PyUnicodeObject *v = u;
  	u = *(PyUnicodeObject **)u;
  	if (v->str)
! 	    free(v->str);
  	Py_XDECREF(v->utf8str);
! 	free(v);
      }
      Py_XDECREF(unicode_empty);
  }
--- 4661,4669 ----
  	PyUnicodeObject *v = u;
  	u = *(PyUnicodeObject **)u;
  	if (v->str)
! 	    PyMem_DEL(v->str);
  	Py_XDECREF(v->utf8str);
! 	PyObject_DEL(v);
      }
      Py_XDECREF(unicode_empty);
  }
diff -cr PyCVS/Objects/xxobject.c PyMem/Objects/xxobject.c
*** PyCVS/Objects/xxobject.c	Sat Apr 29 00:24:07 2000
--- PyMem/Objects/xxobject.c	Sun Apr 30 08:02:58 2000
***************
*** 71,77 ****
  	xxobject *xp;
  {
  	Py_XDECREF(xp->x_attr);
! 	PyMem_DEL(xp);
  }
  
  static PyObject *
--- 71,77 ----
  	xxobject *xp;
  {
  	Py_XDECREF(xp->x_attr);
! 	PyObject_DEL(xp);
  }
  
  static PyObject *


From Vladimir.Marangozov@inrialpes.fr  Sun Apr 30 20:45:26 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sun, 30 Apr 2000 21:45:26 +0200 (CEST)
Subject: [Patches] PyMem [7/8] - Modules/*
Message-ID: <200004301945.VAA15135@python.inrialpes.fr>


[ Modules/* ]

Important: all object constructors and destructors in extension modules
use the functions PyObject_New/Del. Some extensions are enabled by default,
others aren't, so better be safe than sorry -- all PyObject_NEW/DEL have
been lowercase'd because each one of these modules may not belong to the
core. In terms of speed, this is fine, because there are no performance
critical objects in there (i.e. with intense alloc/dealloc rates).

Spread the word -- PyObject_NEW/DEL in extension modules is discouraged!

However, all existing PyMem_NEW/DEL have been left unchanged. Fine by me.
Someone may want to lowercase them too and see whether there's a
performance drop -- I have no time to dig further in this direction...

- Malloc cleanup + some "invisible bugs" corrected, like in nismodule.c

It turns out that people have used PyMem_NEW/DEL by assuming that they
correspond to standard malloc/free. Pfieu!

Calls to "malloc" must be spelled in the code as "malloc", so if
you're reading this and you're an author of an extension module in the
standard distrbution, and you feel like you could be the author of an
invisible bug, please double check your module. And if you're an author
of a module that is not in the standard distribution, just do the same!
I cannot track down invisible bugs. :-)

--
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252

--

Disclaimer:

I confirm that, to the best of my knowledge and belief, this contribution is
free of any claims of third parties under copyright, patent or other rights
or interests ("claims").  To the extent that I have any such claims, I
hereby grant to CNRI a nonexclusive, irrevocable, royalty-free, worldwide
license to reproduce, distribute, perform and/or display publicly, prepare
derivative versions, and otherwise use this contribution as part of the
Python software and its related documentation, or any derivative versions
thereof, at no cost to CNRI or its licensed users, and to authorize others
to do so.

I acknowledge that CNRI may, at its sole discretion, decide whether or not
to incorporate this contribution in the Python software and its related
documentation.  I further grant CNRI permission to use my name and other
identifying information provided to CNRI by me for use in connection with
the Python software and its related documentation.

-------------------------------[ cut here ]---------------------------
diff -cr PyCVS/Modules/_localemodule.c PyMem/Modules/_localemodule.c
*** PyCVS/Modules/_localemodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/_localemodule.c	Sun Apr 30 03:37:37 2000
***************
*** 287,303 ****
       return NULL;
    /* assume no change in size, first */
    n1=strlen(s)+1;
!   buf=Py_Malloc(n1);
!   if(!buf)return NULL;
    n2=strxfrm(buf,s,n1);
    if(n2>n1){
      /* more space needed */
!     buf=Py_Realloc(buf,n2);
!     if(!buf)return NULL;
      strxfrm(buf,s,n2);
    }
    result=PyString_FromString(buf);
!   Py_Free(buf);
    return result;
  }
  
--- 287,303 ----
       return NULL;
    /* assume no change in size, first */
    n1=strlen(s)+1;
!   buf=PyMem_Malloc(n1);
!   if(!buf)return PyErr_NoMemory();
    n2=strxfrm(buf,s,n1);
    if(n2>n1){
      /* more space needed */
!     buf=PyMem_Realloc(buf,n2);
!     if(!buf)return PyErr_NoMemory();
      strxfrm(buf,s,n2);
    }
    result=PyString_FromString(buf);
!   PyMem_Free(buf);
    return result;
  }
  
diff -cr PyCVS/Modules/_sre.c PyMem/Modules/_sre.c
*** PyCVS/Modules/_sre.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/_sre.c	Sun Apr 30 03:43:27 2000
***************
*** 848,854 ****
                            &PyString_Type, &code, &groups, &groupindex))
  		return NULL;
  
! 	self = PyObject_NEW(PatternObject, &Pattern_Type);
  	if (self == NULL)
  		return NULL;
  
--- 848,854 ----
                            &PyString_Type, &code, &groups, &groupindex))
  		return NULL;
  
! 	self = PyObject_New(PatternObject, &Pattern_Type);
  	if (self == NULL)
  		return NULL;
  
***************
*** 886,892 ****
  	if (status > 0) {
  
  		/* create match object (with room for extra group marks) */
! 		match = PyObject_NEW_VAR(MatchObject, &Match_Type, 2*pattern->groups);
  		if (match == NULL)
  			return NULL;
  
--- 886,892 ----
  	if (status > 0) {
  
  		/* create match object (with room for extra group marks) */
! 		match = PyObject_NewVar(MatchObject, &Match_Type, 2*pattern->groups);
  		if (match == NULL)
  			return NULL;
  
***************
*** 1002,1008 ****
  	Py_XDECREF(self->code);
  	Py_XDECREF(self->pattern);
  	Py_XDECREF(self->groupindex);
! 	PyMem_DEL(self);
  }
  
  static PyObject*
--- 1002,1008 ----
  	Py_XDECREF(self->code);
  	Py_XDECREF(self->pattern);
  	Py_XDECREF(self->groupindex);
! 	PyObject_Del(self);
  }
  
  static PyObject*
***************
*** 1163,1169 ****
  {
  	Py_XDECREF(self->string);
  	Py_DECREF(self->pattern);
! 	PyMem_DEL(self);
  }
  
  static PyObject*
--- 1163,1169 ----
  {
  	Py_XDECREF(self->string);
  	Py_DECREF(self->pattern);
! 	PyObject_Del(self);
  }
  
  static PyObject*
diff -cr PyCVS/Modules/_tkinter.c PyMem/Modules/_tkinter.c
*** PyCVS/Modules/_tkinter.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/_tkinter.c	Sun Apr 30 03:49:00 2000
***************
*** 469,475 ****
  	TkappObject *v;
  	char *argv0;
    
! 	v = PyObject_NEW(TkappObject, &Tkapp_Type);
  	if (v == NULL)
  		return NULL;
  
--- 469,475 ----
  	TkappObject *v;
  	char *argv0;
    
! 	v = PyObject_New(TkappObject, &Tkapp_Type);
  	if (v == NULL)
  		return NULL;
  
***************
*** 1640,1646 ****
  {
  	TkttObject *v;
    
! 	v = PyObject_NEW(TkttObject, &Tktt_Type);
  	if (v == NULL)
  		return NULL;
  
--- 1640,1646 ----
  {
  	TkttObject *v;
    
! 	v = PyObject_New(TkttObject, &Tktt_Type);
  	if (v == NULL)
  		return NULL;
  
***************
*** 1662,1668 ****
  
  	Py_XDECREF(func);
  
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 1662,1668 ----
  
  	Py_XDECREF(func);
  
! 	PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 1910,1916 ****
  	ENTER_TCL
  	Tcl_DeleteInterp(Tkapp_Interp(self));
  	LEAVE_TCL
! 	PyMem_DEL(self);
  	DisableEventHook();
  }
  
--- 1910,1916 ----
  	ENTER_TCL
  	Tcl_DeleteInterp(Tkapp_Interp(self));
  	LEAVE_TCL
! 	PyObject_Del(self);
  	DisableEventHook();
  }
  
diff -cr PyCVS/Modules/almodule.c PyMem/Modules/almodule.c
*** PyCVS/Modules/almodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/almodule.c	Sun Apr 30 06:45:25 2000
***************
*** 648,654 ****
  {
  	alcobject *self;
  	
! 	self = PyObject_NEW(alcobject, &Alctype);
  	if (self == NULL)
  		return NULL;
  	/* XXXX Add your own initializers here */
--- 648,654 ----
  {
  	alcobject *self;
  	
! 	self = PyObject_New(alcobject, &Alctype);
  	if (self == NULL)
  		return NULL;
  	/* XXXX Add your own initializers here */
***************
*** 667,673 ****
  #else
  	(void) ALfreeconfig(self->config);	/* ignore errors */
  #endif
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 667,673 ----
  #else
  	(void) ALfreeconfig(self->config);	/* ignore errors */
  #endif
! 	PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 1421,1427 ****
  {
  	alpobject *self;
  	
! 	self = PyObject_NEW(alpobject, &Alptype);
  	if (self == NULL)
  		return NULL;
  	/* XXXX Add your own initializers here */
--- 1421,1427 ----
  {
  	alpobject *self;
  	
! 	self = PyObject_New(alpobject, &Alptype);
  	if (self == NULL)
  		return NULL;
  	/* XXXX Add your own initializers here */
***************
*** 1442,1448 ****
  		ALcloseport(self->port);
  #endif
  	}
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 1442,1448 ----
  		ALcloseport(self->port);
  #endif
  	}
! 	PyObject_Del(self);
  }
  
  static PyObject *
diff -cr PyCVS/Modules/arraymodule.c PyMem/Modules/arraymodule.c
*** PyCVS/Modules/arraymodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/arraymodule.c	Sun Apr 30 04:00:25 2000
***************
*** 346,352 ****
  	if (nbytes / descr->itemsize != (size_t)size) {
  		return PyErr_NoMemory();
  	}
! 	op = PyMem_NEW(arrayobject, 1);
  	if (op == NULL) {
  		return PyErr_NoMemory();
  	}
--- 346,352 ----
  	if (nbytes / descr->itemsize != (size_t)size) {
  		return PyErr_NoMemory();
  	}
! 	op = PyObject_NewVar(arrayobject, &Arraytype, size);
  	if (op == NULL) {
  		return PyErr_NoMemory();
  	}
***************
*** 356,369 ****
  	else {
  		op->ob_item = PyMem_NEW(char, nbytes);
  		if (op->ob_item == NULL) {
! 			PyMem_DEL(op);
  			return PyErr_NoMemory();
  		}
  	}
- 	op->ob_type = &Arraytype;
- 	op->ob_size = size;
  	op->ob_descr = descr;
- 	_Py_NewReference((PyObject *)op);
  	return (PyObject *) op;
  }
  
--- 356,366 ----
  	else {
  		op->ob_item = PyMem_NEW(char, nbytes);
  		if (op->ob_item == NULL) {
! 			PyObject_Del(op);
  			return PyErr_NoMemory();
  		}
  	}
  	op->ob_descr = descr;
  	return (PyObject *) op;
  }
  
***************
*** 466,472 ****
  {
  	if (op->ob_item != NULL)
  		PyMem_DEL(op->ob_item);
! 	PyMem_DEL(op);
  }
  
  static int
--- 463,469 ----
  {
  	if (op->ob_item != NULL)
  		PyMem_DEL(op->ob_item);
! 	PyObject_Del(op);
  }
  
  static int
diff -cr PyCVS/Modules/bsddbmodule.c PyMem/Modules/bsddbmodule.c
*** PyCVS/Modules/bsddbmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/bsddbmodule.c	Sun Apr 30 04:06:15 2000
***************
*** 89,95 ****
  	bsddbobject *dp;
  	HASHINFO info;
  
! 	if ((dp = PyObject_NEW(bsddbobject, &Bsddbtype)) == NULL)
  		return NULL;
  
  	info.bsize = bsize;
--- 89,95 ----
  	bsddbobject *dp;
  	HASHINFO info;
  
! 	if ((dp = PyObject_New(bsddbobject, &Bsddbtype)) == NULL)
  		return NULL;
  
  	info.bsize = bsize;
***************
*** 143,149 ****
  	bsddbobject *dp;
  	BTREEINFO info;
  
! 	if ((dp = PyObject_NEW(bsddbobject, &Bsddbtype)) == NULL)
  		return NULL;
  
  	info.flags = btflags;
--- 143,149 ----
  	bsddbobject *dp;
  	BTREEINFO info;
  
! 	if ((dp = PyObject_New(bsddbobject, &Bsddbtype)) == NULL)
  		return NULL;
  
  	info.flags = btflags;
***************
*** 200,206 ****
  	bsddbobject *dp;
  	RECNOINFO info;
  
! 	if ((dp = PyObject_NEW(bsddbobject, &Bsddbtype)) == NULL)
  		return NULL;
  
  	info.flags = rnflags;
--- 200,206 ----
  	bsddbobject *dp;
  	RECNOINFO info;
  
! 	if ((dp = PyObject_New(bsddbobject, &Bsddbtype)) == NULL)
  		return NULL;
  
  	info.flags = rnflags;
***************
*** 261,267 ****
  				"Python bsddb: close errno %d in dealloc\n",
  				errno);
  	}
! 	PyMem_DEL(dp);
  }
  
  #ifdef WITH_THREAD
--- 261,267 ----
  				"Python bsddb: close errno %d in dealloc\n",
  				errno);
  	}
! 	PyObject_Del(dp);
  }
  
  #ifdef WITH_THREAD
diff -cr PyCVS/Modules/cPickle.c PyMem/Modules/cPickle.c
*** PyCVS/Modules/cPickle.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/cPickle.c	Sun Apr 30 04:10:32 2000
***************
*** 171,177 ****
  
      if (self->data) free(self->data);
  
!     PyMem_DEL(self);
  }
  
  static PyTypeObject PdataType = {
--- 171,177 ----
  
      if (self->data) free(self->data);
  
!     PyObject_Del(self);
  }
  
  static PyTypeObject PdataType = {
***************
*** 186,192 ****
  Pdata_New() {
      Pdata *self;
  
!     UNLESS (self = PyObject_NEW(Pdata, &PdataType)) return NULL;
      self->size=8;
      self->length=0;
      self->data=malloc(self->size * sizeof(PyObject*));
--- 186,192 ----
  Pdata_New() {
      Pdata *self;
  
!     UNLESS (self = PyObject_New(Pdata, &PdataType)) return NULL;
      self->size=8;
      self->length=0;
      self->data=malloc(self->size * sizeof(PyObject*));
***************
*** 2132,2138 ****
  newPicklerobject(PyObject *file, int bin) {
      Picklerobject *self;
  
!     UNLESS (self = PyObject_NEW(Picklerobject, &Picklertype))
          return NULL;
  
      self->fp = NULL;
--- 2132,2138 ----
  newPicklerobject(PyObject *file, int bin) {
      Picklerobject *self;
  
!     UNLESS (self = PyObject_New(Picklerobject, &Picklertype))
          return NULL;
  
      self->fp = NULL;
***************
*** 2243,2249 ****
          free(self->write_buf);
      }
  
!     PyMem_DEL(self);
  }
  
  
--- 2243,2249 ----
          free(self->write_buf);
      }
  
!     PyObject_Del(self);
  }
  
  
***************
*** 2885,2891 ****
                PyInstanceObject *inst;
  
                PyErr_Clear();
!               UNLESS (inst=PyObject_NEW(PyInstanceObject, &PyInstance_Type))
                  goto err;
                inst->in_class=(PyClassObject*)cls;
                Py_INCREF(cls);
--- 2885,2891 ----
                PyInstanceObject *inst;
  
                PyErr_Clear();
!               UNLESS (inst=PyObject_New(PyInstanceObject, &PyInstance_Type))
                  goto err;
                inst->in_class=(PyClassObject*)cls;
                Py_INCREF(cls);
***************
*** 4036,4042 ****
  newUnpicklerobject(PyObject *f) {
      Unpicklerobject *self;
  
!     UNLESS (self = PyObject_NEW(Unpicklerobject, &Unpicklertype))
          return NULL;
  
      self->file = NULL;
--- 4036,4042 ----
  newUnpicklerobject(PyObject *f) {
      Unpicklerobject *self;
  
!     UNLESS (self = PyObject_New(Unpicklerobject, &Unpicklertype))
          return NULL;
  
      self->file = NULL;
***************
*** 4141,4147 ****
          free(self->buf);
      }
      
!     PyMem_DEL(self);
  }
  
  
--- 4141,4147 ----
          free(self->buf);
      }
      
!     PyObject_Del(self);
  }
  
  
diff -cr PyCVS/Modules/cStringIO.c PyMem/Modules/cStringIO.c
*** PyCVS/Modules/cStringIO.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/cStringIO.c	Sun Apr 30 04:13:56 2000
***************
*** 56,62 ****
  "\n"
  "This module provides a simple useful replacement for\n"
  "the StringIO module that is written in C.  It does not provide the\n"
! "full generality if StringIO, but it provides anough for most\n"
  "applications and is especially useful in conjuction with the\n"
  "pickle module.\n"
  "\n"
--- 56,62 ----
  "\n"
  "This module provides a simple useful replacement for\n"
  "the StringIO module that is written in C.  It does not provide the\n"
! "full generality if StringIO, but it provides enough for most\n"
  "applications and is especially useful in conjuction with the\n"
  "pickle module.\n"
  "\n"
***************
*** 407,413 ****
  O_dealloc(Oobject *self) {
    if (self->buf != NULL)
      free(self->buf);
!   PyMem_DEL(self);
  }
  
  static PyObject *
--- 407,413 ----
  O_dealloc(Oobject *self) {
    if (self->buf != NULL)
      free(self->buf);
!   PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 465,471 ****
  newOobject(int  size) {
    Oobject *self;
  	
!   self = PyObject_NEW(Oobject, &Otype);
    if (self == NULL)
      return NULL;
    self->pos=0;
--- 465,471 ----
  newOobject(int  size) {
    Oobject *self;
  	
!   self = PyObject_New(Oobject, &Otype);
    if (self == NULL)
      return NULL;
    self->pos=0;
***************
*** 536,542 ****
  static void
  I_dealloc(Iobject *self) {
    Py_XDECREF(self->pbuf);
!   PyMem_DEL(self);
  }
  
  static PyObject *
--- 536,542 ----
  static void
  I_dealloc(Iobject *self) {
    Py_XDECREF(self->pbuf);
!   PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 586,592 ****
    }
    buf = PyString_AS_STRING(s);
    size = PyString_GET_SIZE(s);
!   UNLESS(self = PyObject_NEW(Iobject, &Itype)) return NULL;
    Py_INCREF(s);
    self->buf=buf;
    self->string_size=size;
--- 586,592 ----
    }
    buf = PyString_AS_STRING(s);
    size = PyString_GET_SIZE(s);
!   UNLESS(self = PyObject_New(Iobject, &Itype)) return NULL;
    Py_INCREF(s);
    self->buf=buf;
    self->string_size=size;
diff -cr PyCVS/Modules/cdmodule.c PyMem/Modules/cdmodule.c
*** PyCVS/Modules/cdmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/cdmodule.c	Sun Apr 30 04:16:08 2000
***************
*** 447,453 ****
  {
  	if (self->ob_cdplayer != NULL)
  		CDclose(self->ob_cdplayer);
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 447,453 ----
  {
  	if (self->ob_cdplayer != NULL)
  		CDclose(self->ob_cdplayer);
! 	PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 483,489 ****
  {
  	cdplayerobject *p;
  
! 	p = PyObject_NEW(cdplayerobject, &CdPlayertype);
  	if (p == NULL)
  		return NULL;
  	p->ob_cdplayer = cdp;
--- 483,489 ----
  {
  	cdplayerobject *p;
  
! 	p = PyObject_New(cdplayerobject, &CdPlayertype);
  	if (p == NULL)
  		return NULL;
  	p->ob_cdplayer = cdp;
***************
*** 761,767 ****
  		self->ob_cdcallbacks[i].ob_cdcallbackarg = NULL;
  	}
  	CDdeleteparser(self->ob_cdparser);
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 761,767 ----
  		self->ob_cdcallbacks[i].ob_cdcallbackarg = NULL;
  	}
  	CDdeleteparser(self->ob_cdparser);
! 	PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 799,805 ****
  	cdparserobject *p;
  	int i;
  
! 	p = PyObject_NEW(cdparserobject, &CdParsertype);
  	if (p == NULL)
  		return NULL;
  	p->ob_cdparser = cdp;
--- 799,805 ----
  	cdparserobject *p;
  	int i;
  
! 	p = PyObject_New(cdparserobject, &CdParsertype);
  	if (p == NULL)
  		return NULL;
  	p->ob_cdparser = cdp;
diff -cr PyCVS/Modules/clmodule.c PyMem/Modules/clmodule.c
*** PyCVS/Modules/clmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/clmodule.c	Sun Apr 30 04:18:47 2000
***************
*** 673,679 ****
  		else
  			clCloseDecompressor(SELF->ob_compressorHdl);
  	}
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 673,679 ----
  		else
  			clCloseDecompressor(SELF->ob_compressorHdl);
  	}
! 	PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 713,719 ****
  	if (!PyArg_Parse(args, "i", &scheme))
  		return NULL;
  
! 	new = PyObject_NEW(clobject, &Cltype);
  	if (new == NULL)
  		return NULL;
  
--- 713,719 ----
  	if (!PyArg_Parse(args, "i", &scheme))
  		return NULL;
  
! 	new = PyObject_New(clobject, &Cltype);
  	if (new == NULL)
  		return NULL;
  
diff -cr PyCVS/Modules/cursesmodule.c PyMem/Modules/cursesmodule.c
*** PyCVS/Modules/cursesmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/cursesmodule.c	Sun Apr 30 04:25:13 2000
***************
*** 273,279 ****
  	PyFileObject *in_fo;
  	PyFileObject *out_fo;
  	PyCursesScreenObject *xp;
! 	xp = PyObject_NEW(PyCursesScreenObject, &PyCursesScreen_Type);
  	if (xp == NULL)
  		return NULL;
  	return (PyObject *)xp;
--- 273,279 ----
  	PyFileObject *in_fo;
  	PyFileObject *out_fo;
  	PyCursesScreenObject *xp;
! 	xp = PyObject_New(PyCursesScreenObject, &PyCursesScreen_Type);
  	if (xp == NULL)
  		return NULL;
  	return (PyObject *)xp;
***************
*** 289,295 ****
  {
  	PyCursesWindowObject *wo;
  
! 	wo = PyObject_NEW(PyCursesWindowObject, &PyCursesWindow_Type);
  	if (wo == NULL)
  		return NULL;
  	wo->win = win;
--- 289,295 ----
  {
  	PyCursesWindowObject *wo;
  
! 	wo = PyObject_New(PyCursesWindowObject, &PyCursesWindow_Type);
  	if (wo == NULL)
  		return NULL;
  	wo->win = win;
***************
*** 303,309 ****
  {
    if (wo->win != stdscr)
      delwin(wo->win);
!   PyMem_DEL(wo);
  }
  
  static PyObject *
--- 303,309 ----
  {
    if (wo->win != stdscr)
      delwin(wo->win);
!   PyObject_Del(wo);
  }
  
  static PyObject *
***************
*** 1125,1131 ****
  	WINDOW *pad;
  {
  	PyCursesPadObject *po;
! 	po = PyObject_NEW(PyCursesPadObject, &PyCursesPad_Type);
  	if (po == NULL)
  		return NULL;
  	po->pad = pad;
--- 1125,1131 ----
  	WINDOW *pad;
  {
  	PyCursesPadObject *po;
! 	po = PyObject_New(PyCursesPadObject, &PyCursesPad_Type);
  	if (po == NULL)
  		return NULL;
  	po->pad = pad;
diff -cr PyCVS/Modules/dbmmodule.c PyMem/Modules/dbmmodule.c
*** PyCVS/Modules/dbmmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/dbmmodule.c	Sun Apr 30 04:25:55 2000
***************
*** 62,68 ****
  {
          dbmobject *dp;
  
! 	dp = PyObject_NEW(dbmobject, &Dbmtype);
  	if (dp == NULL)
  		return NULL;
  	dp->di_size = -1;
--- 62,68 ----
  {
          dbmobject *dp;
  
! 	dp = PyObject_New(dbmobject, &Dbmtype);
  	if (dp == NULL)
  		return NULL;
  	dp->di_size = -1;
***************
*** 82,88 ****
  {
          if ( dp->di_dbm )
  		dbm_close(dp->di_dbm);
! 	PyMem_DEL(dp);
  }
  
  static int
--- 82,88 ----
  {
          if ( dp->di_dbm )
  		dbm_close(dp->di_dbm);
! 	PyObject_Del(dp);
  }
  
  static int
diff -cr PyCVS/Modules/dlmodule.c PyMem/Modules/dlmodule.c
*** PyCVS/Modules/dlmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/dlmodule.c	Sun Apr 30 04:27:30 2000
***************
*** 54,60 ****
  	PyUnivPtr *handle;
  {
  	dlobject *xp;
! 	xp = PyObject_NEW(dlobject, &Dltype);
  	if (xp == NULL)
  		return NULL;
  	xp->dl_handle = handle;
--- 54,60 ----
  	PyUnivPtr *handle;
  {
  	dlobject *xp;
! 	xp = PyObject_New(dlobject, &Dltype);
  	if (xp == NULL)
  		return NULL;
  	xp->dl_handle = handle;
***************
*** 67,73 ****
  {
  	if (xp->dl_handle != NULL)
  		dlclose(xp->dl_handle);
! 	PyMem_DEL(xp);
  }
  
  static PyObject *
--- 67,73 ----
  {
  	if (xp->dl_handle != NULL)
  		dlclose(xp->dl_handle);
! 	PyObject_Del(xp);
  }
  
  static PyObject *
diff -cr PyCVS/Modules/flmodule.c PyMem/Modules/flmodule.c
*** PyCVS/Modules/flmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/flmodule.c	Sun Apr 30 04:30:00 2000
***************
*** 332,338 ****
  	fl_free_object(g->ob_generic);
  	Py_XDECREF(g->ob_callback);
  	Py_XDECREF(g->ob_callback_arg);
! 	PyMem_DEL(g);
  }
  
  #define OFF(x) offsetof(FL_OBJECT, x)
--- 332,338 ----
  	fl_free_object(g->ob_generic);
  	Py_XDECREF(g->ob_callback);
  	Py_XDECREF(g->ob_callback_arg);
! 	PyObject_Del(g);
  }
  
  #define OFF(x) offsetof(FL_OBJECT, x)
***************
*** 461,467 ****
  	PyMethodDef *methods;
  {
  	genericobject *g;
! 	g = PyObject_NEW(genericobject, &GenericObjecttype);
  	if (g == NULL)
  		return NULL;
  	g-> ob_generic = generic;
--- 461,467 ----
  	PyMethodDef *methods;
  {
  	genericobject *g;
! 	g = PyObject_New(genericobject, &GenericObjecttype);
  	if (g == NULL)
  		return NULL;
  	g-> ob_generic = generic;
***************
*** 1852,1858 ****
  	if (f->ob_form->visible)
  		fl_hide_form(f->ob_form);
  	fl_free_form(f->ob_form);
! 	PyMem_DEL(f);
  }
  
  #define OFF(x) offsetof(FL_FORM, x)
--- 1852,1858 ----
  	if (f->ob_form->visible)
  		fl_hide_form(f->ob_form);
  	fl_free_form(f->ob_form);
! 	PyObject_Del(f);
  }
  
  #define OFF(x) offsetof(FL_FORM, x)
***************
*** 1931,1937 ****
  	FL_FORM *form;
  {
  	formobject *f;
! 	f = PyObject_NEW(formobject, &Formtype);
  	if (f == NULL)
  		return NULL;
  	f->ob_form = form;
--- 1931,1937 ----
  	FL_FORM *form;
  {
  	formobject *f;
! 	f = PyObject_New(formobject, &Formtype);
  	if (f == NULL)
  		return NULL;
  	f->ob_form = form;
diff -cr PyCVS/Modules/fmmodule.c PyMem/Modules/fmmodule.c
*** PyCVS/Modules/fmmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/fmmodule.c	Sun Apr 30 04:30:39 2000
***************
*** 59,65 ****
  				"error creating new font handle");
  		return NULL;
  	}
! 	fhp = PyObject_NEW(fhobject, &Fhtype);
  	if (fhp == NULL)
  		return NULL;
  	fhp->fh_fh = fh;
--- 59,65 ----
  				"error creating new font handle");
  		return NULL;
  	}
! 	fhp = PyObject_New(fhobject, &Fhtype);
  	if (fhp == NULL)
  		return NULL;
  	fhp->fh_fh = fh;
***************
*** 196,202 ****
  	fhobject *fhp;
  {
  	fmfreefont(fhp->fh_fh);
! 	PyMem_DEL(fhp);
  }
  
  static PyTypeObject Fhtype = {
--- 196,202 ----
  	fhobject *fhp;
  {
  	fmfreefont(fhp->fh_fh);
! 	PyObject_Del(fhp);
  }
  
  static PyTypeObject Fhtype = {
diff -cr PyCVS/Modules/gdbmmodule.c PyMem/Modules/gdbmmodule.c
*** PyCVS/Modules/gdbmmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/gdbmmodule.c	Sun Apr 30 04:32:18 2000
***************
*** 93,99 ****
  {
          dbmobject *dp;
  
! 	dp = PyObject_NEW(dbmobject, &Dbmtype);
  	if (dp == NULL)
  		return NULL;
  	dp->di_size = -1;
--- 93,99 ----
  {
          dbmobject *dp;
  
! 	dp = PyObject_New(dbmobject, &Dbmtype);
  	if (dp == NULL)
  		return NULL;
  	dp->di_size = -1;
***************
*** 117,123 ****
  {
          if ( dp->di_dbm )
  		gdbm_close(dp->di_dbm);
! 	PyMem_DEL(dp);
  }
  
  static int
--- 117,123 ----
  {
          if ( dp->di_dbm )
  		gdbm_close(dp->di_dbm);
! 	PyObject_Del(dp);
  }
  
  static int
diff -cr PyCVS/Modules/getpath.c PyMem/Modules/getpath.c
*** PyCVS/Modules/getpath.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/getpath.c	Sun Apr 30 04:35:02 2000
***************
*** 529,535 ****
  	bufsz += strlen(exec_prefix) + 1;
  
  	/* This is the only malloc call in this file */
! 	buf = malloc(bufsz);
  
  	if (buf == NULL) {
  		/* We can't exit, so print a warning and limp along */
--- 529,535 ----
  	bufsz += strlen(exec_prefix) + 1;
  
  	/* This is the only malloc call in this file */
! 	buf = PyMem_Malloc(bufsz);
  
  	if (buf == NULL) {
  		/* We can't exit, so print a warning and limp along */
diff -cr PyCVS/Modules/linuxaudiodev.c PyMem/Modules/linuxaudiodev.c
*** PyCVS/Modules/linuxaudiodev.c	Sat Apr 29 00:24:07 2000
--- PyMem/Modules/linuxaudiodev.c	Sun Apr 30 04:38:45 2000
***************
*** 104,110 ****
    }
  
    /* Create and initialize the object */
!   if ((xp = PyObject_NEW(lad_t, &Ladtype)) == NULL) {
      close(fd);
      return NULL;
    }
--- 104,110 ----
    }
  
    /* Create and initialize the object */
!   if ((xp = PyObject_New(lad_t, &Ladtype)) == NULL) {
      close(fd);
      return NULL;
    }
***************
*** 118,124 ****
  lad_dealloc(lad_t *xp)
  {
    close(xp->x_fd);
!   PyMem_DEL(xp);
  }
  
  static PyObject *
--- 118,124 ----
  lad_dealloc(lad_t *xp)
  {
    close(xp->x_fd);
!   PyObject_Del(xp);
  }
  
  static PyObject *
diff -cr PyCVS/Modules/main.c PyMem/Modules/main.c
*** PyCVS/Modules/main.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/main.c	Sun Apr 30 04:40:41 2000
***************
*** 124,130 ****
  			/* -c is the last option; following arguments
  			   that look like options are left for the
  			   the command to interpret. */
! 			command = malloc(strlen(optarg) + 2);
  			if (command == NULL)
  				Py_FatalError(
  				   "not enough memory to copy -c argument");
--- 124,130 ----
  			/* -c is the last option; following arguments
  			   that look like options are left for the
  			   the command to interpret. */
! 			command = PyMem_Malloc(strlen(optarg) + 2);
  			if (command == NULL)
  				Py_FatalError(
  				   "not enough memory to copy -c argument");
***************
*** 277,283 ****
  
  	if (command) {
  		sts = PyRun_SimpleString(command) != 0;
! 		free(command);
  	}
  	else {
  		if (filename == NULL && stdin_is_interactive) {
--- 277,283 ----
  
  	if (command) {
  		sts = PyRun_SimpleString(command) != 0;
! 		PyMem_Free(command);
  	}
  	else {
  		if (filename == NULL && stdin_is_interactive) {
diff -cr PyCVS/Modules/md5module.c PyMem/Modules/md5module.c
*** PyCVS/Modules/md5module.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/md5module.c	Sun Apr 30 04:42:06 2000
***************
*** 56,62 ****
  {
  	md5object *md5p;
  
! 	md5p = PyObject_NEW(md5object, &MD5type);
  	if (md5p == NULL)
  		return NULL;
  
--- 56,62 ----
  {
  	md5object *md5p;
  
! 	md5p = PyObject_New(md5object, &MD5type);
  	if (md5p == NULL)
  		return NULL;
  
***************
*** 71,77 ****
  md5_dealloc(md5p)
  	md5object *md5p;
  {
! 	PyMem_DEL(md5p);
  }
  
  
--- 71,77 ----
  md5_dealloc(md5p)
  	md5object *md5p;
  {
! 	PyObject_Del(md5p);
  }
  
  
diff -cr PyCVS/Modules/mmapmodule.c PyMem/Modules/mmapmodule.c
*** PyCVS/Modules/mmapmodule.c	Sat Apr 29 00:24:07 2000
--- PyMem/Modules/mmapmodule.c	Sun Apr 30 04:48:32 2000
***************
*** 69,75 ****
  	}
  #endif /* UNIX */
  
! 	PyMem_DEL(m_obj);
  }
  
  static PyObject *
--- 69,75 ----
  	}
  #endif /* UNIX */
  
! 	PyObject_Del(m_obj);
  }
  
  static PyObject *
***************
*** 706,712 ****
  		)
  		return NULL;
    
! 	m_obj = PyObject_NEW (mmap_object, &mmap_object_type);
  	if (m_obj == NULL) {return NULL;}
  	m_obj->size = (size_t) map_size;
  	m_obj->pos = (size_t) 0;
--- 706,712 ----
  		)
  		return NULL;
    
! 	m_obj = PyObject_New (mmap_object, &mmap_object_type);
  	if (m_obj == NULL) {return NULL;}
  	m_obj->size = (size_t) map_size;
  	m_obj->pos = (size_t) 0;
***************
*** 757,763 ****
  		fseek(&_iob[fileno], 0, SEEK_SET);
  	}
  
! 	m_obj = PyObject_NEW (mmap_object, &mmap_object_type);
      
  	if (fh) {
  		m_obj->file_handle = fh;
--- 757,763 ----
  		fseek(&_iob[fileno], 0, SEEK_SET);
  	}
  
! 	m_obj = PyObject_New (mmap_object, &mmap_object_type);
      
  	if (fh) {
  		m_obj->file_handle = fh;
diff -cr PyCVS/Modules/mpzmodule.c PyMem/Modules/mpzmodule.c
*** PyCVS/Modules/mpzmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/mpzmodule.c	Sun Apr 30 04:49:32 2000
***************
*** 123,129 ****
  #ifdef MPZ_DEBUG
  	fputs( "mpz_object() called...\n", stderr );
  #endif /* def MPZ_DEBUG */
! 	mpzp = PyObject_NEW(mpzobject, &MPZtype);
  	if (mpzp == NULL)
  		return NULL;
  
--- 123,129 ----
  #ifdef MPZ_DEBUG
  	fputs( "mpz_object() called...\n", stderr );
  #endif /* def MPZ_DEBUG */
! 	mpzp = PyObject_New(mpzobject, &MPZtype);
  	if (mpzp == NULL)
  		return NULL;
  
***************
*** 285,291 ****
  	fputs( "mpz_dealloc() called...\n", stderr );
  #endif /* def MPZ_DEBUG */
  	mpz_clear(&mpzp->mpz);
! 	PyMem_DEL(mpzp);
  } /* mpz_dealloc() */
  
  
--- 285,291 ----
  	fputs( "mpz_dealloc() called...\n", stderr );
  #endif /* def MPZ_DEBUG */
  	mpz_clear(&mpzp->mpz);
! 	PyObject_Del(mpzp);
  } /* mpz_dealloc() */
  
  
diff -cr PyCVS/Modules/newmodule.c PyMem/Modules/newmodule.c
*** PyCVS/Modules/newmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/newmodule.c	Sun Apr 30 04:51:01 2000
***************
*** 49,55 ****
  			      &PyClass_Type, &klass,
  			      &PyDict_Type, &dict))
  		return NULL;
! 	inst = PyObject_NEW(PyInstanceObject, &PyInstance_Type);
  	if (inst == NULL)
  		return NULL;
  	Py_INCREF(klass);
--- 49,55 ----
  			      &PyClass_Type, &klass,
  			      &PyDict_Type, &dict))
  		return NULL;
! 	inst = PyObject_New(PyInstanceObject, &PyInstance_Type);
  	if (inst == NULL)
  		return NULL;
  	Py_INCREF(klass);
diff -cr PyCVS/Modules/nismodule.c PyMem/Modules/nismodule.c
*** PyCVS/Modules/nismodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/nismodule.c	Sun Apr 30 05:01:16 2000
***************
*** 353,363 ****
  	if (list->stat != NIS_TRUE)
  		goto finally;
  
! 	PyMem_DEL(server);
  	return list->maps;
  
    finally:
! 	PyMem_DEL(server);
  	return NULL;
  }
  
--- 353,363 ----
  	if (list->stat != NIS_TRUE)
  		goto finally;
  
! 	free(server);
  	return list->maps;
  
    finally:
! 	free(server);
  	return NULL;
  }
  
diff -cr PyCVS/Modules/parsermodule.c PyMem/Modules/parsermodule.c
*** PyCVS/Modules/parsermodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/parsermodule.c	Sun Apr 30 05:38:03 2000
***************
*** 295,301 ****
  static PyObject*
  parser_newastobject(node *ast, int type)
  {
!     PyAST_Object* o = PyObject_NEW(PyAST_Object, &PyAST_Type);
  
      if (o != 0) {
          o->ast_node = ast;
--- 295,301 ----
  static PyObject*
  parser_newastobject(node *ast, int type)
  {
!     PyAST_Object* o = PyObject_New(PyAST_Object, &PyAST_Type);
  
      if (o != 0) {
          o->ast_node = ast;
***************
*** 317,323 ****
  parser_free(PyAST_Object *ast)
  {
      PyNode_Free(ast->ast_node);
!     PyMem_DEL(ast);
  }
  
  
--- 317,323 ----
  parser_free(PyAST_Object *ast)
  {
      PyNode_Free(ast->ast_node);
!     PyObject_Del(ast);
  }
  
  
***************
*** 790,799 ****
                  PyObject *temp = PySequence_GetItem(elem, 1);
  
                  /* check_terminal_tuple() already verified it's a string */
!                 strn = (char *)malloc(PyString_GET_SIZE(temp) + 1);
                  if (strn != NULL)
                      (void) strcpy(strn, PyString_AS_STRING(temp));
!                 Py_XDECREF(temp);
  
                  if (PyObject_Length(elem) == 3) {
                      PyObject* temp = PySequence_GetItem(elem, 2);
--- 790,799 ----
                  PyObject *temp = PySequence_GetItem(elem, 1);
  
                  /* check_terminal_tuple() already verified it's a string */
!                 strn = (char *)PyMem_MALLOC(PyString_GET_SIZE(temp) + 1);
                  if (strn != NULL)
                      (void) strcpy(strn, PyString_AS_STRING(temp));
!                 Py_DECREF(temp);
  
                  if (PyObject_Length(elem) == 3) {
                      PyObject* temp = PySequence_GetItem(elem, 2);
diff -cr PyCVS/Modules/pcremodule.c PyMem/Modules/pcremodule.c
*** PyCVS/Modules/pcremodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/pcremodule.c	Sun Apr 30 05:39:56 2000
***************
*** 79,85 ****
  	PyObject *arg;
  {
  	PcreObject *self;
! 	self = PyObject_NEW(PcreObject, &Pcre_Type);
  	if (self == NULL)
  		return NULL;
  	self->regex = NULL;
--- 79,85 ----
  	PyObject *arg;
  {
  	PcreObject *self;
! 	self = PyObject_New(PcreObject, &Pcre_Type);
  	if (self == NULL)
  		return NULL;
  	self->regex = NULL;
***************
*** 95,101 ****
  {
  	if (self->regex) (pcre_free)(self->regex);
  	if (self->regex_extra) (pcre_free)(self->regex_extra);
! 	PyMem_DEL(self);
  }
  
  
--- 95,101 ----
  {
  	if (self->regex) (pcre_free)(self->regex);
  	if (self->regex_extra) (pcre_free)(self->regex_extra);
! 	PyObject_Del(self);
  }
  
  
diff -cr PyCVS/Modules/pyexpat.c PyMem/Modules/pyexpat.c
*** PyCVS/Modules/pyexpat.c	Sat Apr 29 00:24:07 2000
--- PyMem/Modules/pyexpat.c	Sun Apr 30 05:44:08 2000
***************
*** 471,477 ****
          int i;
          xmlparseobject *self;
          
!         self = PyObject_NEW(xmlparseobject, &Xmlparsetype);
          if (self == NULL)
                  return NULL;
  
--- 471,477 ----
          int i;
          xmlparseobject *self;
          
!         self = PyObject_New(xmlparseobject, &Xmlparsetype);
          if (self == NULL)
                  return NULL;
  
***************
*** 512,518 ****
          for( i=0; handler_info[i].name!=NULL; i++ ){
                  Py_XDECREF( self->handlers[i] );
          }
!         PyMem_DEL(self);
  }
  
  static int handlername2int( const char *name ){
--- 512,518 ----
          for( i=0; handler_info[i].name!=NULL; i++ ){
                  Py_XDECREF( self->handlers[i] );
          }
!         PyObject_Del(self);
  }
  
  static int handlername2int( const char *name ){
diff -cr PyCVS/Modules/readline.c PyMem/Modules/readline.c
*** PyCVS/Modules/readline.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/readline.c	Sun Apr 30 05:53:32 2000
***************
*** 377,383 ****
  	char *prompt;
  {
  	int n;
! 	char *p;
  	RETSIGTYPE (*old_inthandler)();
  	old_inthandler = signal(SIGINT, onintr);
  	if (setjmp(jbuf)) {
--- 377,383 ----
  	char *prompt;
  {
  	int n;
! 	char *p, *q;
  	RETSIGTYPE (*old_inthandler)();
  	old_inthandler = signal(SIGINT, onintr);
  	if (setjmp(jbuf)) {
***************
*** 391,398 ****
  	rl_event_hook = PyOS_InputHook;
  	p = readline(prompt);
  	signal(SIGINT, old_inthandler);
  	if (p == NULL) {
! 		p = malloc(1);
  		if (p != NULL)
  			*p = '\0';
  		return p;
--- 391,400 ----
  	rl_event_hook = PyOS_InputHook;
  	p = readline(prompt);
  	signal(SIGINT, old_inthandler);
+ 
+ 	/* We must return a buffer allocated with PyMem_Malloc. */
  	if (p == NULL) {
! 		p = PyMem_Malloc(1);
  		if (p != NULL)
  			*p = '\0';
  		return p;
***************
*** 400,409 ****
  	n = strlen(p);
  	if (n > 0)
  		add_history(p);
! 	if ((p = realloc(p, n+2)) != NULL) {
  		p[n] = '\n';
  		p[n+1] = '\0';
  	}
  	return p;
  }
  
--- 402,417 ----
  	n = strlen(p);
  	if (n > 0)
  		add_history(p);
! 	/* Copy the malloc'ed buffer into a PyMem_Malloc'ed one and
! 	   release the original. */
! 	q = p;
! 	p = PyMem_Malloc(n+2);
! 	if (p != NULL) {
! 		strncpy(p, q, n);
  		p[n] = '\n';
  		p[n+1] = '\0';
  	}
+ 	free(q);
  	return p;
  }
  
diff -cr PyCVS/Modules/regexmodule.c PyMem/Modules/regexmodule.c
*** PyCVS/Modules/regexmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/regexmodule.c	Sun Apr 30 06:35:56 2000
***************
*** 64,76 ****
  reg_dealloc(re)
  	regexobject *re;
  {
! 	PyMem_XDEL(re->re_patbuf.buffer);
  	Py_XDECREF(re->re_translate);
  	Py_XDECREF(re->re_lastok);
  	Py_XDECREF(re->re_groupindex);
  	Py_XDECREF(re->re_givenpat);
  	Py_XDECREF(re->re_realpat);
! 	PyMem_DEL(re);
  }
  
  static PyObject *
--- 64,77 ----
  reg_dealloc(re)
  	regexobject *re;
  {
! 	if (re->re_patbuf.buffer)
! 		PyMem_DEL(re->re_patbuf.buffer);
  	Py_XDECREF(re->re_translate);
  	Py_XDECREF(re->re_lastok);
  	Py_XDECREF(re->re_groupindex);
  	Py_XDECREF(re->re_givenpat);
  	Py_XDECREF(re->re_realpat);
! 	PyObject_Del(re);
  }
  
  static PyObject *
***************
*** 418,424 ****
  				"translation table must be 256 bytes");
  		return NULL;
  	}
! 	re = PyObject_NEW(regexobject, &Regextype);
  	if (re != NULL) {
  		char *error;
  		re->re_patbuf.buffer = NULL;
--- 419,425 ----
  				"translation table must be 256 bytes");
  		return NULL;
  	}
! 	re = PyObject_New(regexobject, &Regextype);
  	if (re != NULL) {
  		char *error;
  		re->re_patbuf.buffer = NULL;
diff -cr PyCVS/Modules/rotormodule.c PyMem/Modules/rotormodule.c
*** PyCVS/Modules/rotormodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/rotormodule.c	Sun Apr 30 06:44:01 2000
***************
*** 178,184 ****
  {
  	Rotorobj *xp;
  
! 	xp = PyObject_NEW(Rotorobj, &Rotor_Type);
  	if (xp == NULL)
  		return NULL;
  	set_key(xp, key);
--- 178,184 ----
  {
  	Rotorobj *xp;
  
! 	xp = PyObject_New(Rotorobj, &Rotor_Type);
  	if (xp == NULL)
  		return NULL;
  	set_key(xp, key);
***************
*** 204,213 ****
  	return xp;
  
    finally:
! 	PyMem_XDEL(xp->e_rotor);
! 	PyMem_XDEL(xp->d_rotor);
! 	PyMem_XDEL(xp->positions);
! 	PyMem_XDEL(xp->advances);
  	Py_DECREF(xp);
  	return (Rotorobj*)PyErr_NoMemory();
  }
--- 204,217 ----
  	return xp;
  
    finally:
! 	if (xp->e_rotor)
! 		PyMem_DEL(xp->e_rotor);
! 	if (xp->d_rotor)
! 		PyMem_DEL(xp->d_rotor);
! 	if (xp->positions)
! 		PyMem_DEL(xp->positions);
! 	if (xp->advances)
! 		PyMem_DEL(xp->advances);
  	Py_DECREF(xp);
  	return (Rotorobj*)PyErr_NoMemory();
  }
***************
*** 473,483 ****
  rotor_dealloc(xp)
  	Rotorobj *xp;
  {
! 	PyMem_XDEL(xp->e_rotor);
! 	PyMem_XDEL(xp->d_rotor);
! 	PyMem_XDEL(xp->positions);
! 	PyMem_XDEL(xp->advances);
! 	PyMem_DEL(xp);
  }
  
  static PyObject * 
--- 477,491 ----
  rotor_dealloc(xp)
  	Rotorobj *xp;
  {
! 	if (xp->e_rotor)
! 		PyMem_DEL(xp->e_rotor);
! 	if (xp->d_rotor)
! 		PyMem_DEL(xp->d_rotor);
! 	if (xp->positions)
! 		PyMem_DEL(xp->positions);
! 	if (xp->advances)
! 		PyMem_DEL(xp->advances);
! 	PyObject_Del(xp);
  }
  
  static PyObject * 
diff -cr PyCVS/Modules/selectmodule.c PyMem/Modules/selectmodule.c
*** PyCVS/Modules/selectmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/selectmodule.c	Sun Apr 30 06:42:55 2000
***************
*** 278,286 ****
  	wfd2obj = PyMem_NEW(pylist, FD_SETSIZE + 3);
  	efd2obj = PyMem_NEW(pylist, FD_SETSIZE + 3);
  	if (rfd2obj == NULL || wfd2obj == NULL || efd2obj == NULL) {
! 		PyMem_XDEL(rfd2obj);
! 		PyMem_XDEL(wfd2obj);
! 		PyMem_XDEL(efd2obj);
  		return NULL;
  	}
  #endif
--- 278,286 ----
  	wfd2obj = PyMem_NEW(pylist, FD_SETSIZE + 3);
  	efd2obj = PyMem_NEW(pylist, FD_SETSIZE + 3);
  	if (rfd2obj == NULL || wfd2obj == NULL || efd2obj == NULL) {
! 		if (rfd2obj) PyMem_DEL(rfd2obj);
! 		if (wfd2obj) PyMem_DEL(wfd2obj);
! 		if (efd2obj) PyMem_DEL(efd2obj);
  		return NULL;
  	}
  #endif
diff -cr PyCVS/Modules/shamodule.c PyMem/Modules/shamodule.c
*** PyCVS/Modules/shamodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/shamodule.c	Sun Apr 30 06:06:17 2000
***************
*** 380,386 ****
  static SHAobject *
  newSHAobject()
  {
! 	return (SHAobject *)PyObject_NEW(SHAobject, &SHAtype);
  }
  
  /* Internal methods for a hashing object */
--- 380,386 ----
  static SHAobject *
  newSHAobject()
  {
! 	return (SHAobject *)PyObject_New(SHAobject, &SHAtype);
  }
  
  /* Internal methods for a hashing object */
***************
*** 389,395 ****
  SHA_dealloc(ptr)
  	PyObject *ptr;
  {
! 	PyMem_DEL(ptr);
  }
  
  
--- 389,395 ----
  SHA_dealloc(ptr)
  	PyObject *ptr;
  {
! 	PyObject_Del(ptr);
  }
  
  
diff -cr PyCVS/Modules/socketmodule.c PyMem/Modules/socketmodule.c
*** PyCVS/Modules/socketmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/socketmodule.c	Sun Apr 30 06:13:35 2000
***************
*** 389,395 ****
  {
  	PySocketSockObject *s;
  	PySocketSock_Type.ob_type = &PyType_Type;
! 	s = PyObject_NEW(PySocketSockObject, &PySocketSock_Type);
  	if (s != NULL) {
  		s->sock_fd = fd;
  		s->sock_family = family;
--- 389,395 ----
  {
  	PySocketSockObject *s;
  	PySocketSock_Type.ob_type = &PyType_Type;
! 	s = PyObject_New(PySocketSockObject, &PySocketSock_Type);
  	if (s != NULL) {
  		s->sock_fd = fd;
  		s->sock_family = family;
***************
*** 1368,1374 ****
  {
  	if (s->sock_fd != -1)
  		(void) SOCKETCLOSE(s->sock_fd);
! 	PyMem_DEL(s);
  }
  
  
--- 1368,1374 ----
  {
  	if (s->sock_fd != -1)
  		(void) SOCKETCLOSE(s->sock_fd);
! 	PyObject_Del(s);
  }
  
  
***************
*** 1948,1954 ****
  	meth=SSLv2_client_method();
  #endif
  
! 	self = PyObject_NEW(SSLObject, &SSL_Type); /* Create new object */
  	if (self == NULL){
  		PyErr_SetObject(SSLErrorObject,
  				PyString_FromString("newSSLObject error"));
--- 1948,1954 ----
  	meth=SSLv2_client_method();
  #endif
  
! 	self = PyObject_New(SSLObject, &SSL_Type); /* Create new object */
  	if (self == NULL){
  		PyErr_SetObject(SSLErrorObject,
  				PyString_FromString("newSSLObject error"));
***************
*** 1962,1968 ****
  	if (self->ctx == NULL) {
  		PyErr_SetObject(SSLErrorObject,
  				PyString_FromString("SSL_CTX_new error"));
! 		PyMem_DEL(self);
  		return NULL;
  	}
  
--- 1962,1968 ----
  	if (self->ctx == NULL) {
  		PyErr_SetObject(SSLErrorObject,
  				PyString_FromString("SSL_CTX_new error"));
! 		PyObject_Del(self);
  		return NULL;
  	}
  
***************
*** 1971,1977 ****
  		PyErr_SetObject(SSLErrorObject,
  		      PyString_FromString(
  			"Both the key & certificate files must be specified"));
! 		PyMem_DEL(self);
  		return NULL;
  	}
  
--- 1971,1977 ----
  		PyErr_SetObject(SSLErrorObject,
  		      PyString_FromString(
  			"Both the key & certificate files must be specified"));
! 		PyObject_Del(self);
  		return NULL;
  	}
  
***************
*** 1983,1989 ****
  			PyErr_SetObject(SSLErrorObject,
  				PyString_FromString(
  				  "SSL_CTX_use_PrivateKey_file error"));
! 			PyMem_DEL(self);
  			return NULL;
  		}
  
--- 1983,1989 ----
  			PyErr_SetObject(SSLErrorObject,
  				PyString_FromString(
  				  "SSL_CTX_use_PrivateKey_file error"));
! 			PyObject_Del(self);
  			return NULL;
  		}
  
***************
*** 1993,1999 ****
  			PyErr_SetObject(SSLErrorObject,
  				PyString_FromString(
  				  "SSL_CTX_use_certificate_chain_file error"));
! 			PyMem_DEL(self);
  			return NULL;
  		}
  	}
--- 1993,1999 ----
  			PyErr_SetObject(SSLErrorObject,
  				PyString_FromString(
  				  "SSL_CTX_use_certificate_chain_file error"));
! 			PyObject_Del(self);
  			return NULL;
  		}
  	}
***************
*** 2008,2014 ****
  		/* Actually negotiate SSL connection */
  		PyErr_SetObject(SSLErrorObject,
  				PyString_FromString("SSL_connect error"));
! 		PyMem_DEL(self);
  		return NULL;
  	}
  	self->ssl->debug = 1;
--- 2008,2014 ----
  		/* Actually negotiate SSL connection */
  		PyErr_SetObject(SSLErrorObject,
  				PyString_FromString("SSL_connect error"));
! 		PyObject_Del(self);
  		return NULL;
  	}
  	self->ssl->debug = 1;
***************
*** 2079,2085 ****
  	SSL_free(self->ssl);
  	Py_XDECREF(self->x_attr);
  	Py_XDECREF(self->Socket);
! 	PyMem_DEL(self);
  }
  
  static PyObject *SSL_getattr(SSLObject *self, char *name)
--- 2079,2085 ----
  	SSL_free(self->ssl);
  	Py_XDECREF(self->x_attr);
  	Py_XDECREF(self->Socket);
! 	PyObject_Del(self);
  }
  
  static PyObject *SSL_getattr(SSLObject *self, char *name)
diff -cr PyCVS/Modules/stdwinmodule.c PyMem/Modules/stdwinmodule.c
*** PyCVS/Modules/stdwinmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/stdwinmodule.c	Sun Apr 30 06:17:59 2000
***************
*** 271,277 ****
  		Py_DECREF(dp->d_ref);
  		dp->d_ref = NULL;
  	}
! 	free((char *)dp);
  }
  
  static PyObject *
--- 271,277 ----
  		Py_DECREF(dp->d_ref);
  		dp->d_ref = NULL;
  	}
! 	PyObject_Del(dp);
  }
  
  static PyObject *
***************
*** 829,835 ****
  	int left, top, right, bottom;
  {
  	textobject *tp;
! 	tp = PyObject_NEW(textobject, &Texttype);
  	if (tp == NULL)
  		return NULL;
  	tp->t_attr = NULL;
--- 829,835 ----
  	int left, top, right, bottom;
  {
  	textobject *tp;
! 	tp = PyObject_New(textobject, &Texttype);
  	if (tp == NULL)
  		return NULL;
  	tp->t_attr = NULL;
***************
*** 853,859 ****
  		tefree(tp->t_text);
  	Py_XDECREF(tp->t_attr);
  	Py_XDECREF(tp->t_ref);
! 	PyMem_DEL(tp);
  }
  
  static PyObject *
--- 853,859 ----
  		tefree(tp->t_text);
  	Py_XDECREF(tp->t_attr);
  	Py_XDECREF(tp->t_ref);
! 	PyObject_Del(tp);
  }
  
  static PyObject *
***************
*** 1190,1196 ****
  	menu = wmenucreate(id + IDOFFSET, title);
  	if (menu == NULL)
  		return (menuobject *) PyErr_NoMemory();
! 	mp = PyObject_NEW(menuobject, &Menutype);
  	if (mp != NULL) {
  		mp->m_menu = menu;
  		mp->m_id = id + IDOFFSET;
--- 1190,1196 ----
  	menu = wmenucreate(id + IDOFFSET, title);
  	if (menu == NULL)
  		return (menuobject *) PyErr_NoMemory();
! 	mp = PyObject_New(menuobject, &Menutype);
  	if (mp != NULL) {
  		mp->m_menu = menu;
  		mp->m_id = id + IDOFFSET;
***************
*** 1216,1222 ****
  	if (mp->m_menu != NULL)
  		wmenudelete(mp->m_menu);
  	Py_XDECREF(mp->m_attr);
! 	PyMem_DEL(mp);
  }
  
  static PyObject *
--- 1216,1222 ----
  	if (mp->m_menu != NULL)
  		wmenudelete(mp->m_menu);
  	Py_XDECREF(mp->m_attr);
! 	PyObject_Del(mp);
  }
  
  static PyObject *
***************
*** 1386,1392 ****
  	bitmap = wnewbitmap(width, height);
  	if (bitmap == NULL)
  		return (bitmapobject *) PyErr_NoMemory();
! 	bp = PyObject_NEW(bitmapobject, &Bitmaptype);
  	if (bp != NULL) {
  		bp->b_bitmap = bitmap;
  		bp->b_attr = NULL;
--- 1386,1392 ----
  	bitmap = wnewbitmap(width, height);
  	if (bitmap == NULL)
  		return (bitmapobject *) PyErr_NoMemory();
! 	bp = PyObject_New(bitmapobject, &Bitmaptype);
  	if (bp != NULL) {
  		bp->b_bitmap = bitmap;
  		bp->b_attr = NULL;
***************
*** 1405,1411 ****
  	if (bp->b_bitmap != NULL)
  		wfreebitmap(bp->b_bitmap);
  	Py_XDECREF(bp->b_attr);
! 	PyMem_DEL(bp);
  }
  
  static PyObject *
--- 1405,1411 ----
  	if (bp->b_bitmap != NULL)
  		wfreebitmap(bp->b_bitmap);
  	Py_XDECREF(bp->b_attr);
! 	PyObject_Del(bp);
  }
  
  static PyObject *
***************
*** 1554,1560 ****
  	Py_DECREF(wp->w_title);
  	if (wp->w_attr != NULL)
  		Py_DECREF(wp->w_attr);
! 	free((char *)wp);
  }
  
  static PyObject *
--- 1554,1560 ----
  	Py_DECREF(wp->w_title);
  	if (wp->w_attr != NULL)
  		Py_DECREF(wp->w_attr);
! 	PyObject_Del(wp);
  }
  
  static PyObject *
***************
*** 1585,1591 ****
  		PyErr_SetString(StdwinError, "already drawing");
  		return NULL;
  	}
! 	dp = PyObject_NEW(drawingobject, &Drawingtype);
  	if (dp == NULL)
  		return NULL;
  	Drawing = dp;
--- 1585,1591 ----
  		PyErr_SetString(StdwinError, "already drawing");
  		return NULL;
  	}
! 	dp = PyObject_New(drawingobject, &Drawingtype);
  	if (dp == NULL)
  		return NULL;
  	Drawing = dp;
***************
*** 1986,1992 ****
  		PyErr_SetString(StdwinError, "creating too many windows");
  		return NULL;
  	}
! 	wp = PyObject_NEW(windowobject, &Windowtype);
  	if (wp == NULL)
  		return NULL;
  	Py_INCREF(title);
--- 1986,1992 ----
  		PyErr_SetString(StdwinError, "creating too many windows");
  		return NULL;
  	}
! 	wp = PyObject_New(windowobject, &Windowtype);
  	if (wp == NULL)
  		return NULL;
  	Py_INCREF(title);
diff -cr PyCVS/Modules/stropmodule.c PyMem/Modules/stropmodule.c
*** PyCVS/Modules/stropmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/stropmodule.c	Sun Apr 30 06:20:53 2000
***************
*** 1152,1158 ****
  		goto return_same;
  	new_len = len + nfound*(sub_len - pat_len);
  
! 	new_s = (char *)malloc(new_len);
  	if (new_s == NULL) return NULL;
  
  	*out_len = new_len;
--- 1152,1158 ----
  		goto return_same;
  	new_len = len + nfound*(sub_len - pat_len);
  
! 	new_s = (char *)PyMem_MALLOC(new_len);
  	if (new_s == NULL) return NULL;
  
  	*out_len = new_len;
***************
*** 1225,1231 ****
  	}
  	else {
  		new = PyString_FromStringAndSize(new_s, out_len);
! 		free(new_s);
  	}
  	return new;
  }
--- 1225,1231 ----
  	}
  	else {
  		new = PyString_FromStringAndSize(new_s, out_len);
! 		PyMem_FREE(new_s);
  	}
  	return new;
  }
diff -cr PyCVS/Modules/sunaudiodev.c PyMem/Modules/sunaudiodev.c
*** PyCVS/Modules/sunaudiodev.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/sunaudiodev.c	Sun Apr 30 06:23:26 2000
***************
*** 139,145 ****
  	PyMem_DEL(ctldev);
  
  	/* Create and initialize the object */
! 	xp = PyObject_NEW(sadobject, &Sadtype);
  	if (xp == NULL) {
  		close(fd);
  		return NULL;
--- 139,145 ----
  	PyMem_DEL(ctldev);
  
  	/* Create and initialize the object */
! 	xp = PyObject_New(sadobject, &Sadtype);
  	if (xp == NULL) {
  		close(fd);
  		return NULL;
***************
*** 158,164 ****
  	sadobject *xp;
  {
          close(xp->x_fd);
! 	PyMem_DEL(xp);
  }
  
  static PyObject *
--- 158,164 ----
  	sadobject *xp;
  {
          close(xp->x_fd);
! 	PyObject_Del(xp);
  }
  
  static PyObject *
***************
*** 412,418 ****
  
  static sadstatusobject *
  sads_alloc() {
! 	return PyObject_NEW(sadstatusobject, &Sadstatustype);
  }
  
  static void
--- 412,418 ----
  
  static sadstatusobject *
  sads_alloc() {
! 	return PyObject_New(sadstatusobject, &Sadstatustype);
  }
  
  static void
diff -cr PyCVS/Modules/svmodule.c PyMem/Modules/svmodule.c
*** PyCVS/Modules/svmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/svmodule.c	Sun Apr 30 06:25:02 2000
***************
*** 340,346 ****
  		Py_DECREF(self->ob_svideo);
  		self->ob_svideo = NULL;
  	}
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 340,346 ----
  		Py_DECREF(self->ob_svideo);
  		self->ob_svideo = NULL;
  	}
! 	PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 374,380 ****
  {
  	captureobject *p;
  
! 	p = PyObject_NEW(captureobject, &Capturetype);
  	if (p == NULL)
  		return NULL;
  	p->ob_svideo = self;
--- 374,380 ----
  {
  	captureobject *p;
  
! 	p = PyObject_New(captureobject, &Capturetype);
  	if (p == NULL)
  		return NULL;
  	p->ob_svideo = self;
***************
*** 994,1000 ****
  {
  	if (self->ob_svideo != NULL)
  		(void) svCloseVideo(self->ob_svideo);
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 994,1000 ----
  {
  	if (self->ob_svideo != NULL)
  		(void) svCloseVideo(self->ob_svideo);
! 	PyObject_Del(self);
  }
  
  static PyObject *
***************
*** 1026,1032 ****
  {
  	svobject *p;
  
! 	p = PyObject_NEW(svobject, &Svtype);
  	if (p == NULL)
  		return NULL;
  	p->ob_svideo = svp;
--- 1026,1032 ----
  {
  	svobject *p;
  
! 	p = PyObject_New(svobject, &Svtype);
  	if (p == NULL)
  		return NULL;
  	p->ob_svideo = svp;
diff -cr PyCVS/Modules/threadmodule.c PyMem/Modules/threadmodule.c
*** PyCVS/Modules/threadmodule.c	Sat Apr 29 00:24:06 2000
--- PyMem/Modules/threadmodule.c	Sun Apr 30 06:27:33 2000
***************
*** 58,69 ****
  newlockobject()
  {
  	lockobject *self;
! 	self = PyObject_NEW(lockobject, &Locktype);
  	if (self == NULL)
  		return NULL;
  	self->lock_lock = PyThread_allocate_lock();
  	if (self->lock_lock == NULL) {
! 		PyMem_DEL(self);
  		self = NULL;
  		PyErr_SetString(ThreadError, "can't allocate lock");
  	}
--- 58,69 ----
  newlockobject()
  {
  	lockobject *self;
! 	self = PyObject_New(lockobject, &Locktype);
  	if (self == NULL)
  		return NULL;
  	self->lock_lock = PyThread_allocate_lock();
  	if (self->lock_lock == NULL) {
! 		PyObject_Del(self);
  		self = NULL;
  		PyErr_SetString(ThreadError, "can't allocate lock");
  	}
***************
*** 79,85 ****
  	PyThread_release_lock(self->lock_lock);
  	
  	PyThread_free_lock(self->lock_lock);
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 79,85 ----
  	PyThread_release_lock(self->lock_lock);
  	
  	PyThread_free_lock(self->lock_lock);
! 	PyObject_Del(self);
  }
  
  static PyObject *
diff -cr PyCVS/Modules/xxmodule.c PyMem/Modules/xxmodule.c
*** PyCVS/Modules/xxmodule.c	Sat Apr 29 00:24:07 2000
--- PyMem/Modules/xxmodule.c	Sun Apr 30 06:32:10 2000
***************
*** 62,68 ****
  	PyObject *arg;
  {
  	XxoObject *self;
! 	self = PyObject_NEW(XxoObject, &Xxo_Type);
  	if (self == NULL)
  		return NULL;
  	self->x_attr = NULL;
--- 62,68 ----
  	PyObject *arg;
  {
  	XxoObject *self;
! 	self = PyObject_New(XxoObject, &Xxo_Type);
  	if (self == NULL)
  		return NULL;
  	self->x_attr = NULL;
***************
*** 76,82 ****
  	XxoObject *self;
  {
  	Py_XDECREF(self->x_attr);
! 	PyMem_DEL(self);
  }
  
  static PyObject *
--- 76,82 ----
  	XxoObject *self;
  {
  	Py_XDECREF(self->x_attr);
! 	PyObject_Del(self);
  }
  
  static PyObject *
diff -cr PyCVS/Modules/zlibmodule.c PyMem/Modules/zlibmodule.c
*** PyCVS/Modules/zlibmodule.c	Sat Apr 29 00:24:07 2000
--- PyMem/Modules/zlibmodule.c	Sun Apr 30 06:34:28 2000
***************
*** 46,52 ****
       PyTypeObject *type;
  {
          compobject *self;
!         self = PyObject_NEW(compobject, type);
          if (self == NULL)
                  return NULL;
  	self->is_initialised = 0;
--- 46,52 ----
       PyTypeObject *type;
  {
          compobject *self;
!         self = PyObject_New(compobject, type);
          if (self == NULL)
                  return NULL;
  	self->is_initialised = 0;
***************
*** 369,375 ****
      if (self->is_initialised)
        deflateEnd(&self->zst);
      Py_XDECREF(self->unused_data);
!     PyMem_DEL(self);
  }
  
  static void
--- 369,375 ----
      if (self->is_initialised)
        deflateEnd(&self->zst);
      Py_XDECREF(self->unused_data);
!     PyObject_Del(self);
  }
  
  static void
***************
*** 378,384 ****
  {
      inflateEnd(&self->zst);
      Py_XDECREF(self->unused_data);
!     PyMem_DEL(self);
  }
  
  static char comp_compress__doc__[] =
--- 378,384 ----
  {
      inflateEnd(&self->zst);
      Py_XDECREF(self->unused_data);
!     PyObject_Del(self);
  }
  
  static char comp_compress__doc__[] =


From Vladimir.Marangozov@inrialpes.fr  Sun Apr 30 20:46:52 2000
From: Vladimir.Marangozov@inrialpes.fr (Vladimir Marangozov)
Date: Sun, 30 Apr 2000 21:46:52 +0200 (CEST)
Subject: [Patches] PyMem [8/8] - Final words + TODO
Message-ID: <200004301946.VAA15158@python.inrialpes.fr>


Final words
-----------

Py_Malloc/Realloc/Free + PyMem_XDEL are gone.

Support for Py_Malloc/Realloc/Free is dropped. They were hardly ever used.
For source code compatibility, they are defined as PyMem_Malloc/Realloc/Free
PyMem_XDEL is defined as PyMem_DEL. ANSI C free(NULL) has no effect,
so this should be safe.

The new interfaces are in place. The distribution compiles flawlessly
on my Linux RH 6.2 and AIX 4.1, and passes OK the test suite. I've been
able to change easily the Python mallocs and have tested all existing
malloc permutations:
                         +-----------+-----------+
                         |  malloc1  |  malloc2  |   malloc1/malloc2 are
+------------------------+-----------+-----------+   two different malloc
|  PyCore_MALLOC         |     X     |           |   impementations that
|  PyCore_OBJECT_MALLOC  |     X     |           |   may operate on the same
+------------------------+-----------+-----------|   or on different heaps.
|  PyCore_MALLOC         |           |     X     |
|  PyCore_OBJECT_MALLOC  |           |     X     |   Each one of them can
+------------------------+-----------+-----------|   be std C malloc. Both
|  PyCore_MALLOC         |     X     |           |   are, by default.
|  PyCore_OBJECT_MALLOC  |           |     X     |
+------------------------+-----------+-----------|
|  PyCore_MALLOC         |           |     X     |
|  PyCore_OBJECT_MALLOC  |     X     |           |
+------------------------+-----------+-----------+

TODO topics
-----------

1) Integration of custom malloc libs in the build process.

   For instance, when the core malloc is changed, the link of pgen
   (in Parser) fails because of dependencies. The link of pgen would
   probably need to be deferred for the "link phase" of the distrib,
   after the compile in all subdirectories. Also, there has to be an
   automatic way of adding custom malloc libraries to the link phase
   (for now, this has to be done manually). I can't be bothered with
   this right now, though...

2) Invisible bugs

3) Malloc cleanup omissions or bugs in the distrib (nah, shouldn't be)

4) Write what's missing in the Extending/Embedding docs.

--
       Vladimir MARANGOZOV          | Vladimir.Marangozov@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252