Author: fijal
Branch: unicode-utf8
Changeset: r93327:7cd0df437105
Date: 2017-12-09 15:42 +0200
http://bitbucket.org/pypy/pypy/changeset/7cd0df437105/
Log: fix pyexpat
diff --git a/pypy/module/pyexpat/interp_pyexpat.py
b/pypy/module/pyexpat/interp_pyexpat.py
--- a/pypy/module/pyexpat/interp_pyexpat.py
+++ b/pypy/module/pyexpat/interp_pyexpat.py
@@ -587,21 +587,22 @@
def UnknownEncodingHandler(self, space, name, info):
# Yes, supports only 8bit encodings
- translationmap = space.unicode_w(
+ translationmap, lgt = space.utf8_len_w(
space.call_method(
space.newbytes(self.all_chars), "decode",
space.newtext(name), space.newtext("replace")))
- if len(translationmap) != 256:
+ if lgt != 256:
raise oefmt(space.w_ValueError,
"multi-byte encodings are not supported")
- for i in range(256):
- c = translationmap[i]
- if c == u'\ufffd':
+ i = 0
+ for c in rutf8.Utf8StringIterator(translationmap):
+ if c == 0xfffd:
info.c_map[i] = rffi.cast(rffi.INT, -1)
else:
info.c_map[i] = rffi.cast(rffi.INT, c)
+ i += 1
info.c_data = lltype.nullptr(rffi.VOIDP.TO)
info.c_convert = lltype.nullptr(rffi.VOIDP.TO)
info.c_release = lltype.nullptr(rffi.VOIDP.TO)
_______________________________________________
pypy-commit mailing list
pypy-commit@python.org
https://mail.python.org/mailman/listinfo/pypy-commit
