Guido van Rossum [Tue, 18 Feb 2003 22:05:12 +0000 (22:05 +0000)]
Introducing __reduce_ex__, which is called with a protocol number argument
if it exists in preference over __reduce__. Now Tim can go implement this
in cPickle.c.
Tim Peters [Tue, 18 Feb 2003 20:50:45 +0000 (20:50 +0000)]
save_global(): Trying to resolve module.name can fail for two
reasons: importing module can fail, or the attribute lookup
module.name can fail. We were giving the same error msg for
both cases, making it needlessly hard to guess what went wrong.
These cases give different error msgs now.
Guido van Rossum [Tue, 18 Feb 2003 19:22:22 +0000 (19:22 +0000)]
The recent changes to super(), in particular supercheck(), broke when
using super() for an instance in a metaclass situation. Because the
class was a metaclass, the instance was a class, and hence the
PyType_Check() branch was taken. But this branch didn't apply. Make
it so that if this branch doesn't apply, the other branch is still
tried. All tests pass.
Jeremy Hylton [Tue, 18 Feb 2003 15:06:17 +0000 (15:06 +0000)]
Copy the trace module here from Tools/scripts.
There are some problems with this module, but the tool works for
simple tasks and no one else has volunteered a better code coverage
tool. Should cleanup and document before the beta release.
Neal Norwitz [Tue, 18 Feb 2003 03:37:49 +0000 (03:37 +0000)]
Fix 64-bit problem, ParseTuple("i") needs C ints; ("l") needs C longs.
Use "l" as that *probably* makes more sense (at least to me it does :-)
And the test passes on the alpha.
Kurt B. Kaiser [Mon, 17 Feb 2003 18:57:16 +0000 (18:57 +0000)]
M NEWS.txt
M PyShell.py
M ScriptBinding.py
M rpc.py
M run.py
Clean up the way IDLEfork handles termination of the subprocess, restore
ability to interrupt user code in Windows (so long as it's doing terminal
I/O).
1. Handle subprocess interrupts in Windows with an RPC message.
2. Run/F5 will restart the subprocess even if user code is running.
3. Restart the subprocess if the link is dropped.
4. Exit IDLE cleanly even during I/O.
4. In rpc.py, remove explicit calls to statelock, let the condition
variable handle acquire() and release().
Neal Norwitz [Mon, 17 Feb 2003 18:05:20 +0000 (18:05 +0000)]
Use correct function name to PyArg_ParseTuple("is_package").
Fix off-by-1 error in normalize_line_endings():
when *p == '\0' the NUL was copied into q and q was auto-incremented,
the loop was broken out of,
then a newline was appended followed by a NUL.
So the function, in effect, was strcpy() but added two extra chars
which was caught by obmalloc in debug mode, since there was only
room for 1 additional newline.
Get test working under regrtest (added test_main).
Jack Jansen [Mon, 17 Feb 2003 16:47:12 +0000 (16:47 +0000)]
When installing resource files whose name ends in .rsrc use the
"copy anything to a data fork based resource file" trick of macresource.
Fixes #688007.
Andrew MacIntyre [Mon, 17 Feb 2003 09:17:50 +0000 (09:17 +0000)]
Tweak to Skip's checkin of patch 686397:
- 'os2' references in ntpath.py relate to the VACPP port, not the EMX port;
- the VACPP port uses the same defpath as all other ntpath.py supported
platforms except 'ce'.
Tim Peters [Sat, 15 Feb 2003 03:01:11 +0000 (03:01 +0000)]
cPickle.c, load_build(): Taught cPickle how to pick apart
the optional proto 2 slot state.
pickle.py, load_build(): CAUTION: Noted that cPickle's
load_build and pickle's load_build really don't do the same
things with the state, and didn't before this patch either.
cPickle never tries to do .update(), and has no backoff if
instance.__dict__ can't be retrieved. There are no tests
that can tell the difference, and part of what cPickle's
load_build() did looked accidental to me, so I don't know
what the true intent is here.
pickletester.py, test_pickle.py: Got rid of the hack for
exempting cPickle from running some of the proto 2 tests.
Guido van Rossum [Fri, 14 Feb 2003 19:29:22 +0000 (19:29 +0000)]
- The audio driver tests (test_ossaudiodev.py and
test_linuxaudiodev.py) are no longer run by default. This is
because they don't always work, depending on your hardware and
software. To run these tests, you must use an invocation like
./python Lib/test/regrtest.py -u audio test_ossaudiodev
Jack Jansen [Fri, 14 Feb 2003 14:11:59 +0000 (14:11 +0000)]
Factored out classes for handling source and binary distributions. Source
now means "distutils-based source", binary "bdist format archive". Also
fixed various lurking bugs.
Guido van Rossum [Thu, 13 Feb 2003 22:07:59 +0000 (22:07 +0000)]
- Finally fixed the bug in compile() and exec where a string ending
with an indented code block but no newline would raise SyntaxError.
This would have been a four-line change in parsetok.c... Except
codeop.py depends on this behavior, so a compilation flag had to be
invented that causes the tokenizer to revert to the old behavior;
this required extra changes to 2 .h files, 2 .c files, and 2 .py
files. (Fixes SF bug #501622.)
Tim Peters [Thu, 13 Feb 2003 19:30:57 +0000 (19:30 +0000)]
We didn't have any tests making pickles with one of {pickle, cPickle},
and loading them via the other, except for the special cases of this
Guido added to test_datetime.py for datetime module objects. The new
test_xpickle.py tries all of pickletester's AbstractPickleTests in
both x-module ways.
Tim Peters [Thu, 13 Feb 2003 18:24:14 +0000 (18:24 +0000)]
Taught cPickle how to read pickles containing NEWOBJ. This won't get
exercised by the test suite before cPickle knows how to create NEWOBJ
too. For now, it was just tried once by hand (via loading a NEWOBJ
pickle created by pickle.py).
Guido van Rossum [Thu, 13 Feb 2003 16:30:16 +0000 (16:30 +0000)]
SF patch #685738 by Michael Stone.
This changes the default __new__ to refuse arguments iff tp_init is the
default __init__ implementation -- thus making it a TypeError when you
try to pass arguments to a constructor if the class doesn't override at
least __init__ or __new__.
Mark Hammond [Thu, 13 Feb 2003 12:05:15 +0000 (12:05 +0000)]
Use python_d.exe to build _ssl_d.pyd - we can not express that we depend
on a release 'python.exe' for a debug build of _ssl. It may happen that
Python.exe is currently broken, and we are trying to rebuild from scratch.
Tim Peters [Thu, 13 Feb 2003 03:13:40 +0000 (03:13 +0000)]
socket_inet_aton(): ip_addr was left undefined before use in the
!HAVE_INET_ATON case. Repaired that, and tried to repair what looked
like out-of-date comments.
Guido van Rossum [Wed, 12 Feb 2003 23:08:22 +0000 (23:08 +0000)]
Addressing SF bug #643005, implement socket.inet_aton() using
inet_aton() rather than inet_addr() -- the latter is obsolete because
it has a problem: "255.255.255.255" is a valid address but
indistinguishable from an error.
(I'm not sure if inet_aton() exists everywhere -- in case it doesn't,
I've left the old code in with an #ifdef.)
Guido van Rossum [Wed, 12 Feb 2003 21:46:11 +0000 (21:46 +0000)]
Provide access to the import lock, fixing SF bug #580952. This is
mostly from SF patch #683257, but I had to change unlock_import() to
return an error value to avoid fatal error.
Should this be backported? The patch requested this, but it's a new
feature.
Guido van Rossum [Wed, 12 Feb 2003 20:48:22 +0000 (20:48 +0000)]
Issue a warning when int('0...', 0) returns an int with the sign
folded; this will change in Python 2.4. On a 32-bit machine, this
happens for 0x80000000 through 0xffffffff, and for octal constants in
the same value range. No warning is issued if an explicit base is
given, *or* if the string contains a sign (since in those cases no
sign folding ever happens).
Guido van Rossum [Wed, 12 Feb 2003 17:05:26 +0000 (17:05 +0000)]
SF #660455 : patch by NNorwitz.
"Unsigned" (i.e., positive-looking, but really negative) hex/oct
constants with a leading minus sign are once again properly negated.
The micro-optimization for negated numeric constants did the wrong
thing for such hex/oct constants. The patch avoids the optimization
for all hex/oct constants.