Author: Matti Picus <[email protected]>
Branch: unicode-utf8-py3
Changeset: r95263:78258e983d74
Date: 2018-10-31 23:14 +0200
http://bitbucket.org/pypy/pypy/changeset/78258e983d74/
Log: convert File{En,De}codingObject to unicode via traits.as_str0, fix
translation
diff --git a/pypy/interpreter/unicodehelper.py
b/pypy/interpreter/unicodehelper.py
--- a/pypy/interpreter/unicodehelper.py
+++ b/pypy/interpreter/unicodehelper.py
@@ -80,13 +80,15 @@
errorhandler=state.decode_error_handler
if _WIN32:
bytes = space.bytes_w(w_string)
- uni = str_decode_mbcs(bytes, 'strict', True, errorhandler,
- force_ignore=False)[0]
+ slen = len(bytes)
+ uni, size = runicode.str_decode_mbcs(bytes, slen, 'strict', final=True,
+ errorhandler=errorhandler, force_ignore=False)
elif _MACOSX:
bytes = space.bytes_w(w_string)
- uni = str_decode_utf8(
+ utf8 = str_decode_utf8(
bytes, 'surrogateescape', final=True,
allow_surrogates=False)[0]
+ uni = space.realunicode_w(utf8)
elif space.sys.filesystemencoding is None or state.codec_need_encodings:
# bootstrap check: if the filesystemencoding isn't initialized
# or the filesystem codec is implemented in Python we cannot
@@ -101,6 +103,7 @@
return space.call_method(w_string, 'decode',
getfilesystemencoding(space),
space.newtext('surrogateescape'))
+ assert isinstance(uni, unicode)
return space.newtext(runicode.unicode_encode_utf_8(uni,
len(uni), 'strict', allow_surrogates=True),
len(uni))
diff --git a/pypy/module/_cffi_backend/errorbox.py
b/pypy/module/_cffi_backend/errorbox.py
--- a/pypy/module/_cffi_backend/errorbox.py
+++ b/pypy/module/_cffi_backend/errorbox.py
@@ -89,7 +89,7 @@
return
w_text = self.space.call_function(w_done)
- p = rffi.unicode2wcharp(self.space.utf8_w(w_text),
+ p = rffi.unicode2wcharp(self.space.realunicode_w(w_text),
track_allocation=False)
if self.text_p:
rffi.free_wcharp(self.text_p, track_allocation=False)
diff --git a/pypy/module/posix/interp_nt.py b/pypy/module/posix/interp_nt.py
--- a/pypy/module/posix/interp_nt.py
+++ b/pypy/module/posix/interp_nt.py
@@ -96,7 +96,7 @@
raise LLNotImplemented("GetFinalPathNameByHandle not available on "
"this platform")
- hFile = win32traits.CreateFile(path, 0, 0, None,
+ hFile = win32traits.CreateFile(traits.as_str0(path), 0, 0, None,
win32traits.OPEN_EXISTING,
win32traits.FILE_FLAG_BACKUP_SEMANTICS,
rwin32.NULL_HANDLE)
diff --git a/pypy/module/posix/interp_posix.py
b/pypy/module/posix/interp_posix.py
--- a/pypy/module/posix/interp_posix.py
+++ b/pypy/module/posix/interp_posix.py
@@ -56,7 +56,10 @@
return self.space.fsencode_w(self.w_obj)
def as_unicode(self):
- return self.space.utf8_w(self.w_obj).decode('utf8')
+ ret = self.space.realunicode_w(self.w_obj)
+ if u'\x00' in ret:
+ raise oefmt(self.space.w_ValueError, "embedded null character")
+ return ret
class FileDecoder(object):
is_unicode = False
@@ -69,7 +72,10 @@
return self.space.bytesbuf0_w(self.w_obj)
def as_unicode(self):
- return self.space.fsdecode_w(self.w_obj)
+ ret = self.space.fsdecode_w(self.w_obj)
+ if u'\x00' in ret:
+ raise oefmt(self.space.w_ValueError, "embedded null character")
+ return ret
@specialize.memo()
def make_dispatch_function(func, tag, allow_fd_fn=None):
@@ -889,7 +895,7 @@
# started through main() instead of wmain()
rwin32._wgetenv(u"")
for key, value in rwin32._wenviron_items():
- space.setitem(w_env, space.newtext(key), space.newunicode(value))
+ space.setitem(w_env, space.newtext(key), space.newtext(value))
@unwrap_spec(name=unicode, value=unicode)
def putenv(space, name, value):
diff --git a/rpython/rlib/_os_support.py b/rpython/rlib/_os_support.py
--- a/rpython/rlib/_os_support.py
+++ b/rpython/rlib/_os_support.py
@@ -64,6 +64,8 @@
assert path is not None
if isinstance(path, unicode):
return path
+ elif isinstance(path, str):
+ raise RuntimeError('str given where unicode expected')
else:
return path.as_unicode()
diff --git a/rpython/rlib/rposix.py b/rpython/rlib/rposix.py
--- a/rpython/rlib/rposix.py
+++ b/rpython/rlib/rposix.py
@@ -137,10 +137,7 @@
RPY_EXTERN void exit_suppress_iph(void* handle) {};
#endif
''',]
- post_include_bits=['RPY_EXTERN int _PyVerify_fd(int);',
- 'RPY_EXTERN void* enter_suppress_iph();',
- 'RPY_EXTERN void exit_suppress_iph(void* handle);',
- ]
+ post_include_bits=['RPY_EXTERN int _PyVerify_fd(int);']
else:
separate_module_sources = []
post_include_bits = []
@@ -238,8 +235,7 @@
rthread.tlfield_rpy_errno.setraw(_get_errno())
# ^^^ keep fork() up-to-date too, below
if _WIN32:
- includes = ['io.h', 'sys/utime.h', 'sys/types.h', 'process.h', 'time.h',
- 'direct.h']
+ includes = ['io.h', 'sys/utime.h', 'sys/types.h', 'process.h', 'time.h']
libraries = []
else:
if sys.platform.startswith(('darwin', 'netbsd', 'openbsd')):
diff --git a/rpython/rlib/rposix_stat.py b/rpython/rlib/rposix_stat.py
--- a/rpython/rlib/rposix_stat.py
+++ b/rpython/rlib/rposix_stat.py
@@ -685,8 +685,7 @@
# XXX 'traverse' is ignored, and everything related to
# the "reparse points" is missing
win32traits = make_win32_traits(traits)
-
- hFile = win32traits.CreateFile(path,
+ hFile = win32traits.CreateFile(traits.as_str0(path),
win32traits.FILE_READ_ATTRIBUTES,
0,
lltype.nullptr(rwin32.LPSECURITY_ATTRIBUTES.TO),
_______________________________________________
pypy-commit mailing list
[email protected]
https://mail.python.org/mailman/listinfo/pypy-commit
