Author: Tyler Wade <[email protected]>
Branch: utf8-unicode2
Changeset: r72413:02e451d4a78b
Date: 2014-07-09 00:42 -0500
http://bitbucket.org/pypy/pypy/changeset/02e451d4a78b/
Log: Handle UnicodeEncodeError better; make default error handlers
RPython
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
@@ -1,7 +1,7 @@
import sys
from rpython.rlib.rstring import StringBuilder
-from rpython.rlib.objectmodel import specialize
+from rpython.rlib.objectmodel import we_are_translated, specialize
from rpython.rlib.rarithmetic import r_uint, intmask
from rpython.rlib.unicodedata import unicodedb
from rpython.rlib.runicode import utf8_code_length
@@ -1564,7 +1564,6 @@
def default_unicode_error_decode(errors, encoding, msg, s,
startingpos, endingpos):
- """NOT_RPYTHON"""
if errors == 'replace':
return _unicode_error_replacement, endingpos
if errors == 'ignore':
@@ -1574,10 +1573,17 @@
def default_unicode_error_encode(errors, encoding, msg, u,
startingpos, endingpos):
- """NOT_RPYTHON"""
if errors == 'replace':
return '?', None, endingpos
if errors == 'ignore':
return '', None, endingpos
+
+ if we_are_translated():
+ # The constructor for UnicodeEncodeError requires an actual unicode
+ # object; a Utf8Str isn't good enough. Converting a Utf8Str to a
+ # unicode is (somewhat arbitrarily) not RPython. Since, translated
+ # built-in exceptions don't care about their arguments, only do the
+ # conversion when not translated.
+ raise UnicodeEncodeError()
raise UnicodeEncodeError(encoding, unicode(u), startingpos, endingpos, msg)
_______________________________________________
pypy-commit mailing list
[email protected]
https://mail.python.org/mailman/listinfo/pypy-commit
