Fixes bug of having default argument for TimeRE's __init__ that caused the
LocaleTime instance to only be created once and thus not be recreated when
the locale changed.
Remove caching of TimeRE (and thus LocaleTime) instance. Error was being
caught when executing test_strptime, test_logging, and test_time in that order
when the testing of "%c" occured. Suspect the cache was not being recreated
(the test passed when test_logging was forced to re-establish the locale).
Kurt B. Kaiser [Wed, 23 Jul 2003 15:42:14 +0000 (15:42 +0000)]
1. Python Bug 775541: Calltips error when docstring is None. Introduced
by patch 769142. Fixed by patch 776062. KBK will backport net result
to IDLE release22-maint and IDLEfork.
2. Update NEWS.txt and idlever for release.
Jack Jansen [Wed, 23 Jul 2003 10:51:55 +0000 (10:51 +0000)]
Don't force boot-disk-only install, for reasons unknown it causes more
problems than it solves. In stead, put a warning near the top of the
welcome message. Fixes (or, rather works around) bug #764975.
Jack Jansen [Wed, 23 Jul 2003 10:49:17 +0000 (10:49 +0000)]
Scripts runs with pythonw no longer had full window manager access due
to the name change of Python.app/Contents/MacOS/python to
Python.app/Contents/MacOS/Python. Fixes #776116.
Tim Peters [Wed, 23 Jul 2003 00:30:11 +0000 (00:30 +0000)]
locale-restoration code: Don't leave comparison to None implicit. For
all I know, the original locale may be '' (I don't think that's possible,
but ...), and if so we would certainly want to restore it.
Tim Peters [Wed, 23 Jul 2003 00:05:07 +0000 (00:05 +0000)]
Fred wasn't kidding -- there really are docs for the locale module <wink>.
Obtain the original locale in the documented way. This way actually
works for me.
Restore the original locale at the end, instead of forcing to "C".
Move the locale fiddling into the test driver instead of doing it as a
side effect of merely importing the module. I don't know why the test
is mucking with locale (and also added a comment saying so), but it
surely has no justification for doing that as an import side-effect.
Now whenever the locale-changing code executes, the locale-restoring code
will also get run.
Thomas Heller [Tue, 22 Jul 2003 18:10:15 +0000 (18:10 +0000)]
Change the zipimport implementation to accept files containing
arbitrary bytes before the actual zip compatible archive. Zipfiles
containing comments at the end of the file are still not supported.
Add a testcase to test_zipimport, and update NEWS.
Cygwin's pthread_sigmask() implementation appears to be buggy. This
patch works around this problem by using sigprocmask() instead.
This patch is implemented in a general way so it could be used by other
platforms too. If this approach is deemed too risky, then I can work up
a patch that just hacks Python/thread_pthread.h for Cygwin.
Note that I tested this patch against 2.3c1 under Red Hat Linux 8.0 too.
[snip]
And finally, I need someone to regenerate pyconfig.h.in and configure
with the same versions of the autotools that are normally used by
Python.
Neal kindly regenerated pyconfig.h.in and configure for me.
Jack Jansen [Tue, 22 Jul 2003 13:45:26 +0000 (13:45 +0000)]
Files used for the 2.3rc1+ binary installer. This one has the size problem
fixed. It also attempts to force boot-disk-only installs, but this seems to
lead to a problem that you have to deselect and reselect the installation disk.
If no-one comes up with a fix RSN I'll revert to the previous situation and
explain in the readme that you cannot install on a non-system-disk.
Tim Peters [Tue, 22 Jul 2003 02:50:01 +0000 (02:50 +0000)]
Windows fix: When PYTHONCASEOK is set, or for any other reason imports
are satisfied in a case-insensitive manner, the attempt to import (the
non-existent) fcntl gets satisfied by FCNTL.py instead, and the tempfile
module defines a Unix-specific _set_cloexec() function in that case. As
a result, temp files can't be created then (blows up with an AttributeError
trying to reference fcntl.fcntl). This just popped up in the spambayes
project, where there is no apparent workaround (which is why I'm pushing
this in now).
Jack Jansen [Mon, 21 Jul 2003 22:11:07 +0000 (22:11 +0000)]
We erronuously re-used the pimpinstaller object if there were multiple installs.
This lead to a duplication of error messages (and installs). Fixes #764615.
Jack Jansen [Mon, 21 Jul 2003 22:03:14 +0000 (22:03 +0000)]
Because plistlib doesn't catch all possible errors that can
happen while parsing a plistfile feeding an ill-formatted file to
pimp may cause an exception. As a stopgap we use an unqualified except
and print an error message "Unspecified error, probably ill-formatted
database". Fixes #765621.
Jack Jansen [Mon, 21 Jul 2003 20:47:11 +0000 (20:47 +0000)]
Backport from r23c1-branch:
Pimp crashed if you tried to install a non-installable package, in stead
of printing a decent error message. Fixes #773450..
Fred Drake [Mon, 21 Jul 2003 17:05:56 +0000 (17:05 +0000)]
Fix memory leak reported & discussed on the Python XML-SIG mailing list.
This patch was provided by Jeremy Kloth, and corresponds to pyexpat.c
1.77 in the PyXML CVS.
Tim Peters [Mon, 21 Jul 2003 14:25:23 +0000 (14:25 +0000)]
Merge 23c1-branch back into the head. Barry will send email about the
New Plan (releases to be made off the head, ongoing random 2.4 stuff
to be done on a short-lived branch, provided anyone is motivated enough
to create one).
Change \code to \verb, because TeX merges the two dashes into a
hyphen, resulting in the output '-list-classifiers', not the correct
'--list-classifiers'. I've verified that the line is now correct
in the DVI, PDF, HTML, and ASCII versions.
Jeremy Hylton [Fri, 18 Jul 2003 03:19:20 +0000 (03:19 +0000)]
Restore the locale to "C" on exit.
If this doesn't happen, it leaves the locale in a state that can cause
other tests to fail. For example, running test_strptime,
test_logging, and test_time in that order.
Add introductory paragraphs
Remove comment about MacOS changes; I'm not going to have time to figure
out what they are
Move PEP 273 section into numeric order
Jeremy Hylton [Thu, 17 Jul 2003 16:58:48 +0000 (16:58 +0000)]
Get socketmodule compiling on IRIX 6.5.10.
I don't think the fix here is very good, but I'm not sure what would
be better. In particular, we should not be defining _SGIAPI, but lots
of things break if we remove it.
Fred Drake [Thu, 17 Jul 2003 15:22:47 +0000 (15:22 +0000)]
- remove mention of the isprivate flag, since that isn't directly
documented here, and according to Tim, should never have been there
- misc. cleanups for consistency
Fred Drake [Thu, 17 Jul 2003 05:26:53 +0000 (05:26 +0000)]
- improve the description of how user-defined method
objects get made
- improve the description of attribute retrieval from
classes and class instances
- add brief documentation of static method and
class method objects.
expose the C API subsection which was hidden from LaTeX in a comment. In
the info conversion the \comment LaTeX macro mapped to a Texinfo @ignore
macro. Unfortunately, py2texi.el is not smart enough to avoid generating
links to the @ignore'd section, which causes makeinfo to croak.
Exposing this text is probably not the most correct thing to do, as this
documentation really belongs in the C API manual. This does get the info
files generated, however, which is a more practical goal considering the
impending release of 2.3rc1.
Doctest now examines all docstrings by default. Previously, it would
skip over functions with private names (as indicated by the underscore
naming convention). The old default created too much of a risk that
user tests were being skipped inadvertently. Note, this change could
break code in the unlikely case that someone had intentionally put
failing tests in the docstrings of private functions. The breakage
is easily fixable by specifying the old behavior when calling testmod()
or Tester(). The more likely case is that the silent failure was
unintended and that the user needed to be informed so the test could be
fixed.
Fred Drake [Wed, 16 Jul 2003 16:19:08 +0000 (16:19 +0000)]
Remove \versionchanged; the text was too complex for the GNU info
conversion to support. Keep the content as normal content, with a
note that this applies starting in Python 2.3.
Jeremy Hylton [Wed, 16 Jul 2003 16:08:23 +0000 (16:08 +0000)]
Remove unnecessary check in tests for slots allowed.
The !PyType_Check(base) check snuck in as part of rev 2.215, but was
unrelated to the SF patch that is mentioned in the checkin comment.
The test is currently unnecessary because base is set to the return
value of best_bases(), which returns a type or NULL.
Exercise Jim Fulton's new doctest extension for running doctests in a
unittest environment. Since his extension finds docstrings in private
functions, it exposed a bug in the difflib doctests.