[Patches] [ python-Patches-680146 ] _iconv_module type casting and spelling errors
SourceForge.net
noreply@sourceforge.net
Tue, 04 Feb 2003 09:58:38 -0800
Patches item #680146, was opened at 2003-02-04 13:09
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=305470&aid=680146&group_id=5470
Category: Modules
Group: Python 2.3
>Status: Closed
Resolution: Accepted
Priority: 5
Submitted By: Christos Georgiou (tzot)
Assigned to: Walter Dörwald (doerwalter)
Summary: _iconv_module type casting and spelling errors
Initial Comment:
Compiling under SGI Irix and MIPSPro, I corrected two
type casts which under gcc are plain warnings. They
are both related to size_t and int comparisons, since
size_t is unsigned.
I also corrected the spelling mistake of "choosen"
instead of "chosen"
----------------------------------------------------------------------
>Comment By: Walter Dörwald (doerwalter)
Date: 2003-02-04 18:58
Message:
Logged In: YES
user_id=89016
Argl, Neal was faster! ;)
----------------------------------------------------------------------
Comment By: Neal Norwitz (nnorwitz)
Date: 2003-02-04 18:58
Message:
Logged In: YES
user_id=33168
Attaching a patch with more changes.
----------------------------------------------------------------------
Comment By: Walter Dörwald (doerwalter)
Date: 2003-02-04 18:54
Message:
Logged In: YES
user_id=89016
OK, I checked in a slightly different version. Can you check
whether the warnings are still gone?
----------------------------------------------------------------------
Comment By: Christos Georgiou (tzot)
Date: 2003-02-04 18:51
Message:
Logged In: YES
user_id=539787
Yes, it makes sense (and no compilation errors or warnings).
Now, if only I could manage to ignore completely iconv,
because Irix iconv has no ASCII encoding and there are
problems with the big-endianess (see my notes to patch
670715).
I need to delve into the problem with big-endian architecture,
but first I would like to compile ignoring completely the
_iconv_codec module; I do not know how to do that, though.
(I'm messing with setup.py now).
We can close this patch here, thank you.
----------------------------------------------------------------------
Comment By: Christos Georgiou (tzot)
Date: 2003-02-04 18:32
Message:
Logged In: YES
user_id=539787
The code is:
res = iconv(hdl, &inptr, &insize, &outptr, (size_t *)&outsize);
if (res == (size_t)-1)
MIPSPro compiler 7.3 unfortunately throws an error instead of
a warning, even if res is int, because iconv returns size_t,
which is unsigned. This is erroneous on MIPSPro's behalf,
but it's harmless.
I'm gonna check your patch in a quarter.
----------------------------------------------------------------------
Comment By: Neal Norwitz (nnorwitz)
Date: 2003-02-04 18:26
Message:
Logged In: YES
user_id=33168
I'm not sure why you've case res == (size_t)-1. All other
changes make sense. I've modified the patch a bit so that
insize and outsize are size_t instead of int. Does this
make sense and work for you?
----------------------------------------------------------------------
Comment By: Christos Georgiou (tzot)
Date: 2003-02-04 13:28
Message:
Logged In: YES
user_id=539787
Third attempt: the checkbox is checked
----------------------------------------------------------------------
Comment By: Christos Georgiou (tzot)
Date: 2003-02-04 13:12
Message:
Logged In: YES
user_id=539787
For some reason the attachment was not uploaded the first
time.
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=305470&aid=680146&group_id=5470