[pypy-commit] pypy utf8-unicode2: Don't pass r_uint's to Utf8Builder.append

[waedt]

Author: Tyler Wade <wayedt at gmail.com>
Branch: utf8-unicode2
Changeset: r72698:965da5f2a8a3
Date: 2014-08-05 03:27 -0500
http://bitbucket.org/pypy/pypy/changeset/965da5f2a8a3/

Log:	Don't pass r_uint's to Utf8Builder.append

diff --git a/pypy/interpreter/utf8.py b/pypy/interpreter/utf8.py
--- a/pypy/interpreter/utf8.py
+++ b/pypy/interpreter/utf8.py
@@ -642,9 +642,7 @@
                 self._is_ascii = False
             self._length += len(c)
 
-        elif isinstance(c, int) or isinstance(c, base_int):
-            c = intmask(c)
-
+        elif isinstance(c, int):
             if c < 0x80:
                 self._builder.append(chr(c))
             elif c < 0x800:
@@ -665,14 +663,13 @@
             else:
                 raise ValueError("Invalid unicode codepoint > 0x10FFFF.")
             self._length += 1
-        elif isinstance(c, str):
+        else:
+            assert isinstance(c, str)
             self._builder.append(c)
 
             # XXX The assumption here is that the bytes being appended are
             #     ASCII, ie 1:1 byte:char
             self._length += len(c)
-        else:
-            raise TypeError()
 
     @specialize.argtype(1)
     def append_slice(self, s, start, end):
diff --git a/pypy/interpreter/utf8_codecs.py b/pypy/interpreter/utf8_codecs.py
--- a/pypy/interpreter/utf8_codecs.py
+++ b/pypy/interpreter/utf8_codecs.py
@@ -163,7 +163,7 @@
         else:
             # when we get here, chr is a 32-bit unicode character
             if chr <= MAXUNICODE:
-                builder.append(chr)
+                builder.append(intmask(chr))
                 pos += digits
 
             else: