Author: Alex Gaynor <alex.gay...@gmail.com>
Branch:
Changeset: r44716:572f0d8242f6
Date: 2011-06-05 14:40 +0200
http://bitbucket.org/pypy/pypy/changeset/572f0d8242f6/
Log: merged upstream.
diff --git a/pypy/doc/cpython_differences.rst b/pypy/doc/cpython_differences.rst
--- a/pypy/doc/cpython_differences.rst
+++ b/pypy/doc/cpython_differences.rst
@@ -173,6 +173,11 @@
>>>> A.__del__ = lambda self: None
__main__:1: RuntimeWarning: a __del__ method added to an existing type
will not be called
+Even more obscure: the same is true, for old-style classes, if you attach
+the ``__del__`` to an instance (even in CPython this does not work with
+new-style classes). You get a RuntimeWarning in PyPy. To fix these cases
+just make sure there is a ``__del__`` method in the class to start with.
+
Subclasses of built-in types
----------------------------
diff --git a/pypy/doc/project-ideas.rst b/pypy/doc/project-ideas.rst
--- a/pypy/doc/project-ideas.rst
+++ b/pypy/doc/project-ideas.rst
@@ -11,8 +11,11 @@
`mailing list`_. This is simply for the reason that small possible projects
tend to change very rapidly.
-XXX: write a paragraph that this is a loose collection and where to go
-from here
+This list is mostly for having on overview on potential projects. This list is
+by definition not exhaustive and we're pleased if people come up with their
+own improvement ideas. In any case, if you feel like working on some of those
+projects, or anything else in PyPy, pop up on IRC or write to us on the
+`mailing list`_.
Numpy improvements
------------------
diff --git a/pypy/jit/metainterp/test/test_compile.py
b/pypy/jit/metainterp/test/test_compile.py
--- a/pypy/jit/metainterp/test/test_compile.py
+++ b/pypy/jit/metainterp/test/test_compile.py
@@ -63,6 +63,8 @@
call_pure_results = {}
class jitdriver_sd:
warmstate = FakeState()
+ on_compile = staticmethod(lambda *args: None)
+ on_compile_bridge = staticmethod(lambda *args: None)
def test_compile_new_loop():
cpu = FakeCPU()
diff --git a/pypy/jit/metainterp/test/test_warmstate.py
b/pypy/jit/metainterp/test/test_warmstate.py
--- a/pypy/jit/metainterp/test/test_warmstate.py
+++ b/pypy/jit/metainterp/test/test_warmstate.py
@@ -181,6 +181,7 @@
cpu = None
memory_manager = None
class FakeJitDriverSD:
+ jitdriver = None
_green_args_spec = [lltype.Signed, lltype.Float]
_get_printable_location_ptr = None
_confirm_enter_jit_ptr = None
@@ -207,6 +208,7 @@
cpu = None
memory_manager = None
class FakeJitDriverSD:
+ jitdriver = None
_green_args_spec = [lltype.Signed, lltype.Float]
_get_printable_location_ptr = llhelper(GET_LOCATION, get_location)
_confirm_enter_jit_ptr = None
@@ -230,6 +232,7 @@
cpu = None
memory_manager = None
class FakeJitDriverSD:
+ jitdriver = None
_green_args_spec = [lltype.Signed, lltype.Float]
_get_printable_location_ptr = None
_confirm_enter_jit_ptr = llhelper(ENTER_JIT, confirm_enter_jit)
@@ -253,6 +256,7 @@
cpu = None
memory_manager = None
class FakeJitDriverSD:
+ jitdriver = None
_green_args_spec = [lltype.Signed, lltype.Float]
_get_printable_location_ptr = None
_confirm_enter_jit_ptr = None
diff --git a/pypy/module/_multibytecodec/c_codecs.py
b/pypy/module/_multibytecodec/c_codecs.py
--- a/pypy/module/_multibytecodec/c_codecs.py
+++ b/pypy/module/_multibytecodec/c_codecs.py
@@ -103,8 +103,11 @@
[DECODEBUF_P], rffi.SSIZE_T)
pypy_cjk_dec_inbuf_consumed = llexternal('pypy_cjk_dec_inbuf_consumed',
[DECODEBUF_P], rffi.SSIZE_T)
+pypy_cjk_dec_inbuf_add = llexternal('pypy_cjk_dec_inbuf_add',
+ [DECODEBUF_P, rffi.SSIZE_T, rffi.INT],
+ rffi.INT)
-def decode(codec, stringdata):
+def decode(codec, stringdata, errors="strict"):
inleft = len(stringdata)
inbuf = rffi.get_nonmovingbuffer(stringdata)
try:
@@ -112,10 +115,11 @@
if not decodebuf:
raise MemoryError
try:
- r = pypy_cjk_dec_chunk(decodebuf)
- if r != 0:
- multibytecodec_decerror(decodebuf, r)
- assert False
+ while True:
+ r = pypy_cjk_dec_chunk(decodebuf)
+ if r == 0:
+ break
+ multibytecodec_decerror(decodebuf, r, errors)
src = pypy_cjk_dec_outbuf(decodebuf)
length = pypy_cjk_dec_outlen(decodebuf)
return rffi.wcharpsize2unicode(src, length)
@@ -126,7 +130,7 @@
finally:
rffi.free_nonmovingbuffer(stringdata, inbuf)
-def multibytecodec_decerror(decodebuf, e):
+def multibytecodec_decerror(decodebuf, e, errors):
if e > 0:
reason = "illegal multibyte sequence"
esize = e
@@ -138,12 +142,19 @@
else:
raise RuntimeError
#
- # if errors == ERROR_REPLACE:...
- # if errors == ERROR_IGNORE or errors == ERROR_REPLACE:...
+ if errors == "ignore":
+ pypy_cjk_dec_inbuf_add(decodebuf, esize, 0)
+ return # continue decoding
+ if errors == "replace":
+ e = pypy_cjk_dec_inbuf_add(decodebuf, esize, 1)
+ if rffi.cast(lltype.Signed, e) == MBERR_NOMEMORY:
+ raise MemoryError
+ return # continue decoding
start = pypy_cjk_dec_inbuf_consumed(decodebuf)
end = start + esize
- if 1: # errors == ERROR_STRICT:
- raise EncodeDecodeError(start, end, reason)
+ if errors != "strict":
+ reason = "not implemented: custom error handlers" # XXX implement me
+ raise EncodeDecodeError(start, end, reason)
# ____________________________________________________________
# Encoding
@@ -165,8 +176,11 @@
[ENCODEBUF_P], rffi.SSIZE_T)
pypy_cjk_enc_inbuf_consumed = llexternal('pypy_cjk_enc_inbuf_consumed',
[ENCODEBUF_P], rffi.SSIZE_T)
+pypy_cjk_enc_inbuf_add = llexternal('pypy_cjk_enc_inbuf_add',
+ [ENCODEBUF_P, rffi.SSIZE_T, rffi.INT],
+ rffi.INT)
-def encode(codec, unicodedata):
+def encode(codec, unicodedata, errors="strict"):
inleft = len(unicodedata)
inbuf = rffi.get_nonmoving_unicodebuffer(unicodedata)
try:
@@ -174,14 +188,16 @@
if not encodebuf:
raise MemoryError
try:
- r = pypy_cjk_enc_chunk(encodebuf)
- if r != 0:
- multibytecodec_encerror(encodebuf, r)
- assert False
- r = pypy_cjk_enc_reset(encodebuf)
- if r != 0:
- multibytecodec_encerror(encodebuf, r)
- assert False
+ while True:
+ r = pypy_cjk_enc_chunk(encodebuf)
+ if r == 0:
+ break
+ multibytecodec_encerror(encodebuf, r, errors)
+ while True:
+ r = pypy_cjk_enc_reset(encodebuf)
+ if r == 0:
+ break
+ multibytecodec_encerror(encodebuf, r, errors)
src = pypy_cjk_enc_outbuf(encodebuf)
length = pypy_cjk_enc_outlen(encodebuf)
return rffi.charpsize2str(src, length)
@@ -192,7 +208,7 @@
finally:
rffi.free_nonmoving_unicodebuffer(unicodedata, inbuf)
-def multibytecodec_encerror(encodebuf, e):
+def multibytecodec_encerror(encodebuf, e, errors):
if e > 0:
reason = "illegal multibyte sequence"
esize = e
@@ -204,9 +220,16 @@
else:
raise RuntimeError
#
- # if errors == ERROR_REPLACE:...
- # if errors == ERROR_IGNORE or errors == ERROR_REPLACE:...
+ if errors == 'ignore':
+ pypy_cjk_enc_inbuf_add(encodebuf, esize, 0)
+ return # continue encoding
+ if errors == "replace":
+ e = pypy_cjk_enc_inbuf_add(encodebuf, esize, 1)
+ if rffi.cast(lltype.Signed, e) == MBERR_NOMEMORY:
+ raise MemoryError
+ return # continue decoding
start = pypy_cjk_enc_inbuf_consumed(encodebuf)
end = start + esize
- if 1: # errors == ERROR_STRICT:
- raise EncodeDecodeError(start, end, reason)
+ if errors != "strict":
+ reason = "not implemented: custom error handlers" # XXX implement me
+ raise EncodeDecodeError(start, end, reason)
diff --git a/pypy/module/_multibytecodec/interp_multibytecodec.py
b/pypy/module/_multibytecodec/interp_multibytecodec.py
--- a/pypy/module/_multibytecodec/interp_multibytecodec.py
+++ b/pypy/module/_multibytecodec/interp_multibytecodec.py
@@ -13,13 +13,11 @@
@unwrap_spec(input=str, errors="str_or_None")
def decode(self, space, input, errors=None):
- if errors is not None and errors != 'strict':
- raise OperationError(space.w_NotImplementedError, # XXX
- space.wrap("errors='%s' in _multibytecodec"
- % errors))
+ if errors is None:
+ errors = 'strict'
#
try:
- output = c_codecs.decode(self.codec, input)
+ output = c_codecs.decode(self.codec, input, errors)
except c_codecs.EncodeDecodeError, e:
raise OperationError(
space.w_UnicodeDecodeError,
@@ -37,13 +35,11 @@
@unwrap_spec(input=unicode, errors="str_or_None")
def encode(self, space, input, errors=None):
- if errors is not None and errors != 'strict':
- raise OperationError(space.w_NotImplementedError, # XXX
- space.wrap("errors='%s' in _multibytecodec"
- % errors))
+ if errors is None:
+ errors = 'strict'
#
try:
- output = c_codecs.encode(self.codec, input)
+ output = c_codecs.encode(self.codec, input, errors)
except c_codecs.EncodeDecodeError, e:
raise OperationError(
space.w_UnicodeEncodeError,
diff --git a/pypy/module/_multibytecodec/test/test_app_codecs.py
b/pypy/module/_multibytecodec/test/test_app_codecs.py
--- a/pypy/module/_multibytecodec/test/test_app_codecs.py
+++ b/pypy/module/_multibytecodec/test/test_app_codecs.py
@@ -36,6 +36,22 @@
e = raises(UnicodeDecodeError, codec.decode, "~{xyz}").value
assert e.args == ('hz', '~{xyz}', 2, 4, 'illegal multibyte sequence')
+ def test_decode_hz_ignore(self):
+ import _codecs_cn
+ codec = _codecs_cn.getcodec("hz")
+ r = codec.decode("def~{}abc", errors='ignore')
+ assert r == (u'def\u5fcf', 9)
+ r = codec.decode("def~{}abc", 'ignore')
+ assert r == (u'def\u5fcf', 9)
+
+ def test_decode_hz_replace(self):
+ import _codecs_cn
+ codec = _codecs_cn.getcodec("hz")
+ r = codec.decode("def~{}abc", errors='replace')
+ assert r == (u'def\ufffd\u5fcf', 9)
+ r = codec.decode("def~{}abc", 'replace')
+ assert r == (u'def\ufffd\u5fcf', 9)
+
def test_encode_hz(self):
import _codecs_cn
codec = _codecs_cn.getcodec("hz")
@@ -54,3 +70,17 @@
assert e.start == 3
assert e.end == 4
assert e.reason == 'illegal multibyte sequence'
+
+ def test_encode_hz_ignore(self):
+ import _codecs_cn
+ codec = _codecs_cn.getcodec("hz")
+ r = codec.encode(u'abc\u1234def', 'ignore')
+ assert r == ('abcdef', 7)
+ assert type(r[0]) is str
+
+ def test_encode_hz_replace(self):
+ import _codecs_cn
+ codec = _codecs_cn.getcodec("hz")
+ r = codec.encode(u'abc\u1234def', 'replace')
+ assert r == ('abc?def', 7)
+ assert type(r[0]) is str
diff --git a/pypy/module/_multibytecodec/test/test_c_codecs.py
b/pypy/module/_multibytecodec/test/test_c_codecs.py
--- a/pypy/module/_multibytecodec/test/test_c_codecs.py
+++ b/pypy/module/_multibytecodec/test/test_c_codecs.py
@@ -36,6 +36,24 @@
assert e.end == 4
assert e.reason == "illegal multibyte sequence"
+def test_decode_hz_ignore():
+ c = getcodec("hz")
+ u = decode(c, 'def~{}abc', 'ignore')
+ assert u == u'def\u5fcf'
+
+def test_decode_hz_replace():
+ c = getcodec("hz")
+ u = decode(c, 'def~{}abc', 'replace')
+ assert u == u'def\ufffd\u5fcf'
+
+def test_decode_hz_foobar():
+ # not implemented yet: custom error handlers
+ c = getcodec("hz")
+ e = py.test.raises(EncodeDecodeError, decode, c, "~{xyz}", "foobar").value
+ assert e.start == 2
+ assert e.end == 4
+ assert e.reason == "not implemented: custom error handlers"
+
def test_encode_hz():
c = getcodec("hz")
s = encode(c, u'foobar')
@@ -51,6 +69,25 @@
assert e.end == 4
assert e.reason == "illegal multibyte sequence"
+def test_encode_hz_ignore():
+ c = getcodec("hz")
+ s = encode(c, u'abc\u1234def', 'ignore')
+ assert s == 'abcdef'
+
+def test_encode_hz_replace():
+ c = getcodec("hz")
+ s = encode(c, u'abc\u1234def', 'replace')
+ assert s == 'abc?def'
+
+def test_encode_hz_foobar():
+ # not implemented yet: custom error handlers
+ c = getcodec("hz")
+ e = py.test.raises(EncodeDecodeError, encode,
+ c, u'abc\u1234def', 'foobar').value
+ assert e.start == 3
+ assert e.end == 4
+ assert e.reason == "not implemented: custom error handlers"
+
def test_encode_jisx0208():
c = getcodec('iso2022_jp')
s = encode(c, u'\u83ca\u5730\u6642\u592b')
diff --git a/pypy/module/cpyext/api.py b/pypy/module/cpyext/api.py
--- a/pypy/module/cpyext/api.py
+++ b/pypy/module/cpyext/api.py
@@ -348,6 +348,7 @@
'_Py_TrueStruct#': ('PyObject*', 'space.w_True'),
'_Py_ZeroStruct#': ('PyObject*', 'space.w_False'),
'_Py_NotImplementedStruct#': ('PyObject*', 'space.w_NotImplemented'),
+ '_Py_EllipsisObject#': ('PyObject*', 'space.w_Ellipsis'),
'PyDateTimeAPI': ('PyDateTime_CAPI*', 'None'),
}
FORWARD_DECLS = []
diff --git a/pypy/module/cpyext/test/test_sliceobject.py
b/pypy/module/cpyext/test/test_sliceobject.py
--- a/pypy/module/cpyext/test/test_sliceobject.py
+++ b/pypy/module/cpyext/test/test_sliceobject.py
@@ -67,3 +67,14 @@
"""),
])
assert module.nullslice() == slice(None, None, None)
+
+ def test_ellipsis(self):
+ module = self.import_extension('foo', [
+ ("get_ellipsis", "METH_NOARGS",
+ """
+ PyObject *ret = Py_Ellipsis;
+ Py_INCREF(ret);
+ return ret;
+ """),
+ ])
+ assert module.get_ellipsis() is Ellipsis
diff --git a/pypy/translator/c/src/cjkcodecs/multibytecodec.c
b/pypy/translator/c/src/cjkcodecs/multibytecodec.c
--- a/pypy/translator/c/src/cjkcodecs/multibytecodec.c
+++ b/pypy/translator/c/src/cjkcodecs/multibytecodec.c
@@ -1,6 +1,8 @@
#include <stdlib.h>
#include "src/cjkcodecs/multibytecodec.h"
+#define Py_UNICODE_REPLACEMENT_CHARACTER ((Py_UNICODE) 0xFFFD)
+
struct pypy_cjk_dec_s *pypy_cjk_dec_init(const MultibyteCodec *codec,
char *inbuf, Py_ssize_t inlen)
@@ -93,6 +95,20 @@
return d->inbuf - d->inbuf_start;
}
+int pypy_cjk_dec_inbuf_add(struct pypy_cjk_dec_s* d, Py_ssize_t skip,
+ int add_replacement_character)
+{
+ if (add_replacement_character)
+ {
+ if (d->outbuf >= d->outbuf_end)
+ if (expand_decodebuffer(d, 1) == -1)
+ return MBERR_NOMEMORY;
+ *d->outbuf++ = Py_UNICODE_REPLACEMENT_CHARACTER;
+ }
+ d->inbuf += skip;
+ return 0;
+}
+
/************************************************************/
struct pypy_cjk_enc_s *pypy_cjk_enc_init(const MultibyteCodec *codec,
@@ -209,3 +225,34 @@
{
return d->inbuf - d->inbuf_start;
}
+
+int pypy_cjk_enc_inbuf_add(struct pypy_cjk_enc_s* d, Py_ssize_t skip,
+ int add_replacement_character)
+{
+ if (add_replacement_character)
+ {
+ const Py_UNICODE replchar = '?', *inbuf = &replchar;
+ Py_ssize_t r;
+
+ while (1)
+ {
+ Py_ssize_t outleft = (Py_ssize_t)(d->outbuf_end - d->outbuf);
+ r = d->codec->encode(&d->state, d->codec->config,
+ &inbuf, 1, &d->outbuf, outleft, 0);
+ if (r != MBERR_TOOSMALL)
+ break;
+ /* output buffer too small; grow it and continue. */
+ if (expand_encodebuffer(d, -1) == -1)
+ return MBERR_NOMEMORY;
+ }
+ if (r != 0)
+ {
+ if (d->outbuf >= d->outbuf_end)
+ if (expand_encodebuffer(d, 1) == -1)
+ return MBERR_NOMEMORY;
+ *d->outbuf++ = '?';
+ }
+ }
+ d->inbuf += skip;
+ return 0;
+}
diff --git a/pypy/translator/c/src/cjkcodecs/multibytecodec.h
b/pypy/translator/c/src/cjkcodecs/multibytecodec.h
--- a/pypy/translator/c/src/cjkcodecs/multibytecodec.h
+++ b/pypy/translator/c/src/cjkcodecs/multibytecodec.h
@@ -102,6 +102,7 @@
Py_ssize_t pypy_cjk_dec_outlen(struct pypy_cjk_dec_s *);
Py_ssize_t pypy_cjk_dec_inbuf_remaining(struct pypy_cjk_dec_s *d);
Py_ssize_t pypy_cjk_dec_inbuf_consumed(struct pypy_cjk_dec_s* d);
+int pypy_cjk_dec_inbuf_add(struct pypy_cjk_dec_s*, Py_ssize_t, int);
struct pypy_cjk_enc_s {
const MultibyteCodec *codec;
@@ -119,6 +120,7 @@
Py_ssize_t pypy_cjk_enc_outlen(struct pypy_cjk_enc_s *);
Py_ssize_t pypy_cjk_enc_inbuf_remaining(struct pypy_cjk_enc_s *d);
Py_ssize_t pypy_cjk_enc_inbuf_consumed(struct pypy_cjk_enc_s* d);
+int pypy_cjk_enc_inbuf_add(struct pypy_cjk_enc_s*, Py_ssize_t, int);
/* list of codecs defined in the .c files */
_______________________________________________
pypy-commit mailing list
pypy-commit@python.org
http://mail.python.org/mailman/listinfo/pypy-commit
