Thomas Heller [Wed, 10 Nov 2004 09:01:41 +0000 (09:01 +0000)]
Avoid a linker warning: MSVC 7 doesn't support /pdb:None, the debug
info will always be in a .pdb file.
Fred Drake [Wed, 10 Nov 2004 08:08:26 +0000 (08:08 +0000)]
more more styling to the CSS
Fred Drake [Wed, 10 Nov 2004 08:07:00 +0000 (08:07 +0000)]
more XHTML friendliness: <dl compact> becomes well-formed
Fred Drake [Wed, 10 Nov 2004 07:48:17 +0000 (07:48 +0000)]
override a bit of LaTeX2HTML so empty table cells don't disappear from
the output
Raymond Hettinger [Tue, 9 Nov 2004 07:27:35 +0000 (07:27 +0000)]
SF patch #
1062279: deque pickling problems
(Contributed by Dima Dorfman.)
* Support pickling of dictionaries in instances of deque subclasses.
* Support pickling of recursive deques.
Raymond Hettinger [Tue, 9 Nov 2004 07:25:31 +0000 (07:25 +0000)]
SF
1062353: set pickling problems
Support automatic pickling of dictionaries in instance of set subclasses.
Andrew M. Kuchling [Tue, 9 Nov 2004 02:58:02 +0000 (02:58 +0000)]
Bump version number
Add doctest section
Wordsmithing
Tim Peters [Mon, 8 Nov 2004 22:31:09 +0000 (22:31 +0000)]
And delete a useless comment.
Tim Peters [Mon, 8 Nov 2004 22:30:28 +0000 (22:30 +0000)]
_OutputRedirectingPdb.trace_dispatch(): Return the base class's
trace_dispatch() result in a more obvious, and more robust way.
Tim Peters [Mon, 8 Nov 2004 22:07:37 +0000 (22:07 +0000)]
test_doctest.py test_pdb_set_trace_nested(): A new test from Jim Fulton
showing that doctest's pdb.set_trace() support was dramatically broken.
doctest.py _OutputRedirectingPdb.trace_dispatch(): Return a local trace
function instead of (implicitly) None. Else interaction with pdb was
bizarre, noticing only 'call' events. Amazingly, the existing set_trace()
tests didn't care.
Michael W. Hudson [Mon, 8 Nov 2004 16:46:02 +0000 (16:46 +0000)]
test for fixedness of bug #
1057835.
(thanks to Raymond for the prod).
Michael W. Hudson [Mon, 8 Nov 2004 12:17:34 +0000 (12:17 +0000)]
Fix bug
[
1057835 ] compiler.transformer, "from module import *"
Raymond Hettinger [Mon, 8 Nov 2004 06:36:42 +0000 (06:36 +0000)]
SF #
1062190. Removed an assertion that rendered trace.py unnecessarily
inflexibile.
Tim Peters [Mon, 8 Nov 2004 04:30:21 +0000 (04:30 +0000)]
SF bug
1061968: threads: segfault or Py_FatalError at exit
PyGILState_Ensure(): The fix in 2.4a3 for bug
1010677 reintroduced thread
shutdown race bug 225673. Repaired by (once again) ensuring the GIL is
held whenever deleting a thread state.
Alas, there's no useful test case for this shy bug. Four years ago, only
Guido could provoke it, on his box, and today only Armin can provoke it
on his box. I've never been able to provoke it (but not for lack of
trying!).
This is a critical fix for 2.3.5 too, since the fix for
1010677 got
backported there already and so also reintroduced 225673. I don't intend to
backport this fix. For whoever (if anyone) does, there are other thread
fixes in 2.4 that need backporting too, and I bet they need to happen first
for this patch to apply cleanly.
Martin v. Löwis [Sun, 7 Nov 2004 20:01:56 +0000 (20:01 +0000)]
Stop printing listdir bytestring output, as the precise list of strings
returned depends on the filesystem encoding.
Martin v. Löwis [Sun, 7 Nov 2004 19:57:35 +0000 (19:57 +0000)]
Try a different filename if the Latin-1 file name cannot
be represented in the file system. Fixes #989338.
Johannes Gijsbers [Sun, 7 Nov 2004 19:55:18 +0000 (19:55 +0000)]
Patch #
1062033 / bug #
1027771: clarify what happens when calling
insertBefore() with refChild=None.
Johannes Gijsbers [Sun, 7 Nov 2004 19:36:48 +0000 (19:36 +0000)]
Patch #
1062018 / bug #
1038693: comment out dead link in tkinter docs.
Johannes Gijsbers [Sun, 7 Nov 2004 19:16:05 +0000 (19:16 +0000)]
Patch #
1061931 / bug #971872: factor out part of spillproperties, so
properties are also documented if help(Class.<property>) is called
instead of help(Class).
Peter Astrand [Sun, 7 Nov 2004 16:38:08 +0000 (16:38 +0000)]
Added more documentation about the executable argument.
Fixes #
1056441.
Jeremy Hylton [Sun, 7 Nov 2004 16:13:49 +0000 (16:13 +0000)]
Fix for SF bug 988120 via patch
1061941.
If read() returned less than the number of bytes request, the full amount was subtracted from length instead of the actually read amount.
Johannes Gijsbers [Sun, 7 Nov 2004 16:11:35 +0000 (16:11 +0000)]
Patch #
1061924: add documentation for BaseRotatingHandler and correct
reference to non-existent function 'setRollover()'.
Johannes Gijsbers [Sun, 7 Nov 2004 16:02:07 +0000 (16:02 +0000)]
Clarify that it's not necessary to subclass from TestCase to create a
test case. As Jeremy put it: "subclassing is an implementation
technique, not a type declaration".
Johannes Gijsbers [Sun, 7 Nov 2004 15:46:25 +0000 (15:46 +0000)]
Patch #
1061904 / bug #878275: give a nicer error message when someone
accidentally derives from TestSuite instead of TestCase.
Peter Astrand [Sun, 7 Nov 2004 14:30:34 +0000 (14:30 +0000)]
When using shell=True on Windows, don't display a shell window by default. Fixes #
1057061.
Jeremy Hylton [Sun, 7 Nov 2004 14:24:25 +0000 (14:24 +0000)]
Fix apparently trivial buffer overflow (SF bug
1060396).
memset() wrote one past the end of the buffer, which was likely to be unused padding or a yet-to-be-initialized local variable. This routine is already tested by test_socket.
Peter Astrand [Sun, 7 Nov 2004 14:15:28 +0000 (14:15 +0000)]
If close() fails in file_dealloc, then print an error message to
stderr. close() can fail if the user is out-of-quota, for example.
Fixes #959379.
Johannes Gijsbers [Sun, 7 Nov 2004 14:14:27 +0000 (14:14 +0000)]
Patch #
1061857: add documentation for previously undocumented
TimedRotatingFileHandler class. Thanks Jeroen Vloothuis!
Jeremy Hylton [Sun, 7 Nov 2004 14:04:00 +0000 (14:04 +0000)]
SF patch
1025636: Check for NULL returns in compile.c:com_import_stmt
There is no test for this change, because there is no way to provoke memory errors on demand. Test suite passes, though.
Johannes Gijsbers [Sun, 7 Nov 2004 11:35:30 +0000 (11:35 +0000)]
Bug #
1055168: calling pdb.set_trace() calls Bdb.set_trace, which made
the debugger enter inside pdb.set_trace.
Patch #
1061767: make pdb.set_trace enter enter at the stack frame
calling pdb.set_trace().
Martin v. Löwis [Sun, 7 Nov 2004 10:08:29 +0000 (10:08 +0000)]
Bump version for 2.4c1
Raymond Hettinger [Sun, 7 Nov 2004 07:08:25 +0000 (07:08 +0000)]
SF bug #
1061457: spelling error in win installer
Raymond Hettinger [Sun, 7 Nov 2004 06:18:37 +0000 (06:18 +0000)]
SF patch #
1061780: use a new -m option in pdb and profile invocations
Doc patch submitted by Ilya Sandler.
Tim Peters [Sun, 7 Nov 2004 04:52:29 +0000 (04:52 +0000)]
Whitespace normalization.
Brett Cannon [Sun, 7 Nov 2004 01:24:12 +0000 (01:24 +0000)]
Move the AC_MSG_RESULT function call for checking for pthread options up into
the 'if' statement that performed the test. Not all platforms run the test and
on those tests configure outputted a rogue 'no' line.
Brett Cannon [Sun, 7 Nov 2004 01:19:00 +0000 (01:19 +0000)]
Add comment about removal of CoreServices/CoreFoundation compilation against
the core on OS X (also specifically mention removal of
PyMac_GetAppletScriptFile() ).
Tim Peters [Sat, 6 Nov 2004 23:45:48 +0000 (23:45 +0000)]
Essentially SF patch
1061679: add missing __all__ to pickletools.py.
Harmless.
Brett Cannon [Sat, 6 Nov 2004 19:56:45 +0000 (19:56 +0000)]
Comment cleanup.
Anthony Baxter [Sat, 6 Nov 2004 04:45:33 +0000 (04:45 +0000)]
regenerated configure from configure.in
Raymond Hettinger [Sat, 6 Nov 2004 00:31:51 +0000 (00:31 +0000)]
Don't choke on modes like rb or wb.
Barry Warsaw [Sat, 6 Nov 2004 00:04:52 +0000 (00:04 +0000)]
get_boundary(): Fix for SF bug #
1060941. RFC 2046 says boundaries may begin
-- but not end -- with whitespace.
I will backport to Python 2.3.
Barry Warsaw [Sat, 6 Nov 2004 00:04:20 +0000 (00:04 +0000)]
test_boundary_with_leading_space(): Test case for SF bug #
1060941. RFC 2046
says boundaries may begin -- but not end -- with whitespace.
I will backport to Python 2.3.
Raymond Hettinger [Fri, 5 Nov 2004 16:38:08 +0000 (16:38 +0000)]
SF bug #
1060825: Error in difflib docs
Raymond Hettinger [Fri, 5 Nov 2004 07:02:59 +0000 (07:02 +0000)]
SF patch #
1035255: Remove CoreServices / CoreFoundation dependencies in core
(Contributed by Bob Ippolito.)
This patch trims down the Python core on Darwin by making it
independent of CoreFoundation and CoreServices. It does this by:
Changed linker flags in configure/configure.in
Removed the unused PyMac_GetAppletScriptFile
Moved the implementation of PyMac_StrError to the MacOS module
Moved the implementation of PyMac_GetFullPathname to the
Carbon.File module
Fred Drake [Fri, 5 Nov 2004 06:42:22 +0000 (06:42 +0000)]
- remove some bogus <meta> tags from the document head
- clean up some of the generated markup
Fred Drake [Fri, 5 Nov 2004 05:06:08 +0000 (05:06 +0000)]
generate the "type" attribute on the favicon link
Fred Drake [Fri, 5 Nov 2004 04:51:05 +0000 (04:51 +0000)]
switch remaining icon references to the PNG icons
Fred Drake [Fri, 5 Nov 2004 04:23:25 +0000 (04:23 +0000)]
add encouragement to contribute without learning LaTeX
(closes SF bug #948517)
Fred Drake [Fri, 5 Nov 2004 04:05:06 +0000 (04:05 +0000)]
- make the default image type PNG, to match mkhowto
- add a command-line option to control the image type
Tim Peters [Thu, 4 Nov 2004 21:27:48 +0000 (21:27 +0000)]
Point out some platform vagaries in stat() and utime().
Bugfix candidate (the vagaries aren't new <wink>), but I don't intend to
backport this.
Armin Rigo [Thu, 4 Nov 2004 11:29:09 +0000 (11:29 +0000)]
Mistakes in the "sequence types" page:
* explanation for example with lists of lists made confusing use of
the word "contains" to mean "is built out of".
* wrong formula for slices with step. Is it ok to use LaTeX formulas
(which become images in the html document)? This version needs one
because it's based on a fraction. Just writing "\code{(j-i)/k}" here would
be ambiguous because it looks like a rounding-down-to-the-previous-integer
division, which is not what we need here. Of course we could write
"\code{float(j-i)/k}" but it just looks confusing.
Fred Drake [Thu, 4 Nov 2004 05:45:44 +0000 (05:45 +0000)]
markup nit
Anthony Baxter [Thu, 4 Nov 2004 05:23:17 +0000 (05:23 +0000)]
post-release
Skip Montanaro [Thu, 4 Nov 2004 04:31:30 +0000 (04:31 +0000)]
Fix bug
1052242. Also includes rewrite of test case using unittest and
avoiding use of popen.
Fred Drake [Thu, 4 Nov 2004 03:23:04 +0000 (03:23 +0000)]
fix markup in "title" attribute for \citetitle{}
(closes SF patch #
1054715; backporting to release23-maint branch)
cvs2svn [Wed, 3 Nov 2004 06:21:37 +0000 (06:21 +0000)]
This commit was manufactured by cvs2svn to create tag 'r24b2'.
Anthony Baxter [Wed, 3 Nov 2004 06:21:37 +0000 (06:21 +0000)]
release shenanigans
Martin v. Löwis [Tue, 2 Nov 2004 22:59:56 +0000 (22:59 +0000)]
Exclude badsyntax from compileall; adjust options to what Makefile.pre.in does.
Fixes #
1049003.
Martin v. Löwis [Tue, 2 Nov 2004 22:28:07 +0000 (22:28 +0000)]
Bump buildno for 2.4b2.
Fred Drake [Tue, 2 Nov 2004 19:20:43 +0000 (19:20 +0000)]
fix accidental commit
Fred Drake [Tue, 2 Nov 2004 19:18:20 +0000 (19:18 +0000)]
clarify discussion of iteration in the section on the "for" statement
(SF bug #829073)
Fred Drake [Tue, 2 Nov 2004 19:18:20 +0000 (19:18 +0000)]
clarify discussion of iteration in the section on the "for" statement
(SF bug #829073)
Fred Drake [Tue, 2 Nov 2004 18:57:33 +0000 (18:57 +0000)]
clarify trailing comma in function argument list
(SF bug #798652)
Fred Drake [Tue, 2 Nov 2004 18:24:26 +0000 (18:24 +0000)]
- show how to use file.write() with a non-string value
(closes SF bug #621057)
- add missing whitespace around assignment operator
Anthony Baxter [Tue, 2 Nov 2004 13:03:54 +0000 (13:03 +0000)]
release bit
Raymond Hettinger [Tue, 2 Nov 2004 04:20:10 +0000 (04:20 +0000)]
Maintain peepholer's cumlc invariant by updating the running total
everytime a LOAD_CONSTANT is encountered, created, or overwritten.
Added two tests to cover cases affected by the patch.
Raymond Hettinger [Tue, 2 Nov 2004 02:11:35 +0000 (02:11 +0000)]
Bump-up block size.
Raymond Hettinger [Mon, 1 Nov 2004 22:27:14 +0000 (22:27 +0000)]
* Bumped up the sleep() delay to four seconds so this test will run
reliably on WinME with FAT32.
* Native speaker rewrite of the comment block.
* Removed unnecessary backslashes from the multi-line function defintions.
Walter Dörwald [Mon, 1 Nov 2004 17:10:19 +0000 (17:10 +0000)]
Add error checks for the bz2, cStringIO and operator modules.
Add function names to various PyArg_ParseTuple calls in bz2module.c.
Tim Peters [Mon, 1 Nov 2004 16:39:57 +0000 (16:39 +0000)]
gc_list_move(): Make this truly equivalent to remove+append. While
nothing in gc currently cares, the original coding could screw up if,
e.g., you tried to move a node to the list it's already in, and the node
was already the last in its list.
Raymond Hettinger [Mon, 1 Nov 2004 15:19:11 +0000 (15:19 +0000)]
Simplify and future proof NOP counting in the peepholer.
No longer assumes that the input is NOP free.
Hye-Shik Chang [Mon, 1 Nov 2004 08:26:09 +0000 (08:26 +0000)]
Bug #
1057993: Use sleep() always instead of os.utime() possibly broken
in some platforms.
Barry Warsaw [Mon, 1 Nov 2004 03:59:24 +0000 (03:59 +0000)]
SF bug #
1030118, this should have named the email.Utils module as the one
containing these functions. (I will backport to Python 2.3)
Barry Warsaw [Mon, 1 Nov 2004 03:52:43 +0000 (03:52 +0000)]
SF patch #
1056967, changes the semantics of Template.safe_substitute() to not
raise a ValueError for dangling delimiters (the delimiter itself is returned).
Tim Peters [Mon, 1 Nov 2004 02:40:52 +0000 (02:40 +0000)]
test_on_error(): Rewrite so it works on WinXP too. Unsure about 95/98/ME.
Tim Peters [Mon, 1 Nov 2004 01:39:08 +0000 (01:39 +0000)]
gc list function cleanup.
Introduced gc_list_move(), which captures the common gc_list_remove() +
gc_list_append() sequence. In fact, no uses of gc_list_append() remained
(they were all in a gc_list_move() sequence), so commented that one out.
gc_list_merge(): assert that `from` != `to`; that was an implicit
precondition, now verified in a debug build.
Others: added comments about their purpose.
Tim Peters [Sun, 31 Oct 2004 22:12:43 +0000 (22:12 +0000)]
handle_weakrefs(): Simplification -- there's no need to make a second
pass over the unreachable weakrefs-with-callbacks to unreachable objects.
Johannes Gijsbers [Sun, 31 Oct 2004 12:19:34 +0000 (12:19 +0000)]
Bug #
1057535: add link in time module to calendar.timegm().
Johannes Gijsbers [Sun, 31 Oct 2004 12:05:31 +0000 (12:05 +0000)]
Bug #
1048941: shutil.rmtree error handling was always broken
Rewrite rmtree again, this time without os.walk(). Error handling had been
broken since Python 2.3, and the os.walk() version inherited this.
Raymond Hettinger [Sun, 31 Oct 2004 05:46:59 +0000 (05:46 +0000)]
SF patch #
1056231: typo in comment (unicodeobject.h)
Tim Peters [Sat, 30 Oct 2004 23:09:22 +0000 (23:09 +0000)]
SF
1055820: weakref callback vs gc vs threads
In cyclic gc, clear weakrefs to unreachable objects before allowing any
Python code (weakref callbacks or __del__ methods) to run.
This is a critical bugfix, affecting all versions of Python since weakrefs
were introduced. I'll backport to 2.3.
Armin Rigo [Sat, 30 Oct 2004 21:08:59 +0000 (21:08 +0000)]
Fixed a comment and added another one.
Raymond Hettinger [Sat, 30 Oct 2004 08:55:08 +0000 (08:55 +0000)]
Adopt some peepholer suggestions from Armin Rigo:
* Use simpler, faster two pass algorithm for markblocks().
* Free the blocks variable if not NULL and exiting without change.
* Verify that the rest of the compiler has not set an exception.
* Make the test for tuple of constants less restrictive.
* Embellish the comment for chained conditional jumps.
Fred Drake [Fri, 29 Oct 2004 19:47:52 +0000 (19:47 +0000)]
add xml:id attributes wherever we generate the IDs (old patch saved up)
Fred Drake [Fri, 29 Oct 2004 14:35:42 +0000 (14:35 +0000)]
style consistency:
- always include a space after the "#" that starts a comment
- easier to read imports
Vinay Sajip [Fri, 29 Oct 2004 12:30:28 +0000 (12:30 +0000)]
Added network logging example
Armin Rigo [Thu, 28 Oct 2004 16:32:00 +0000 (16:32 +0000)]
Wrote down the invariants of some common objects whose structure is
exposed in header files. Fixed a few comments in these headers.
As we might have expected, writing down invariants systematically exposed a
(minor) bug. In this case, function objects have a writeable func_code
attribute, which could be set to code objects with the wrong number of
free variables. Calling the resulting function segfaulted the interpreter.
Added a corresponding test.
Walter Dörwald [Thu, 28 Oct 2004 13:04:26 +0000 (13:04 +0000)]
Trigger a few error cases in Modules/_codecsmodule.c.
Brett Cannon [Thu, 28 Oct 2004 04:49:21 +0000 (04:49 +0000)]
Fix bug of implementation of algorithm for calculating the date from year, week
of the year, and day of the week. Was not taking into consideration properly
the issue of when %U is used for the week of the year but the year starts on
Monday.
Closes bug #
1045381 again.
Mark Hammond [Wed, 27 Oct 2004 21:54:33 +0000 (21:54 +0000)]
Fix [
1055540 ] bdist_wininst broken for pure Python distributions
Raymond Hettinger [Wed, 27 Oct 2004 06:21:46 +0000 (06:21 +0000)]
SF patch #
1053375.
(Contributed by Facundo Batista.)
Code simplification by eliminating the unnecessary and error-prone
convolutions for the previously weird sign convention in _WorkRep().
Makes the code more understandable, more reliable, and a bit faster.
Hye-Shik Chang [Wed, 27 Oct 2004 03:12:05 +0000 (03:12 +0000)]
Add a comment explains why we should modify mtime here.
Tim Peters [Wed, 27 Oct 2004 02:44:10 +0000 (02:44 +0000)]
Removed newly redundant embedded import.
Tim Peters [Wed, 27 Oct 2004 02:43:25 +0000 (02:43 +0000)]
Whitespace normalization.
Tim Peters [Wed, 27 Oct 2004 02:33:15 +0000 (02:33 +0000)]
Paper over bug
1054615 by passing sane values to os.utime().
The underlying bug still exists, but also existed in 2.3.4:
import.c's load_source_module() returns NULL if
PyOS_GetLastModificationTime() returns -1, but
PyOS_GetLastModificationTime() doesn't set any exception when it returns
-1, and neither does load_source_module() when it gets back -1. This
leads to "SystemError: NULL result without error in PyObject_Call"
on an import that fails in this way.
Greg Ward [Wed, 27 Oct 2004 02:20:04 +0000 (02:20 +0000)]
Update optparse module and test suite to Optik 1.5a2.
Facundo Batista [Tue, 26 Oct 2004 23:38:46 +0000 (23:38 +0000)]
Very few little improvements.
Tim Peters [Tue, 26 Oct 2004 14:38:48 +0000 (14:38 +0000)]
In paragraph 1, changed
Python software
to
this software ("Python")
Just van Rossum [Tue, 26 Oct 2004 11:02:08 +0000 (11:02 +0000)]
- Added tests for the string load/dump function.
- Added a chunk of plist data as generated by Cocoa's NSDictionary and
verify we output the same (including formatting)
- Changed the "literal" plist code to match the raw test data
Just van Rossum [Tue, 26 Oct 2004 10:30:55 +0000 (10:30 +0000)]
Made <data> output match Apple's exactly. To do that I had to add a custom
version of base64.encodestring() so I could control the line length of the
base64 output.