Added --trackcalls command line arg to display crude caller/callee
relationships at program exit. Output is a bit prettier than that for
--listfuncs but won't parse as easily using downstream postprocessing tools.
* Improve readability and remove data dependencies by converting
pre-increment forms to post-increment forms. Post-incrementing
also eliminates the need for negative array indices for oparg fetches.
* In exception handling code, check for class based exceptions before
the older string based exceptions.
Tim Peters [Mon, 5 Apr 2004 19:36:21 +0000 (19:36 +0000)]
Since the fast_yield branch target was introduced, it appears that most
tests of "why" against WHY_YIELD became useless. This patch removes them,
but assert()s that why != WHY_YIELD everywhere such a test was removed.
The test suite ran fine under a debug build (i.e., the asserts never
triggered).
OS/2's TCP/IP stack supports AF_UNIX sockets, with the limitation that
the socket name must start with /socket32/. Unlike Unix systems, this
file never exists in the file system.
If a file is opened with an explicit buffer size >= 1, repeated
close() calls would attempt to free() the buffer already free()ed on
the first close(). [bug introduced with patch #788249]
Making sure that the buffer is free()ed in file object deallocation is
a belt-n-braces bit of insurance against a memory leak.
Fred Drake [Thu, 1 Apr 2004 07:40:35 +0000 (07:40 +0000)]
Fix support for the "prog" keyword to the OptionParser constructor, as well
as directly setting the .prog attribute (which should be supported based on
the class docstring).
Closes SF bug #850964.
Jeremy Hylton [Thu, 1 Apr 2004 02:45:22 +0000 (02:45 +0000)]
Bump the magic number to avoid sharing bytecode between 2.3 and 2.4.
Revise the long comment that explained details of the magic number
in gory detail.
[Bugfix candidate] Escape traceback type and value. There are probably additional cases where cgitb.py doesn't escape as paranoidly as it should (e.g. attribute names)
Guido van Rossum [Wed, 31 Mar 2004 18:53:29 +0000 (18:53 +0000)]
When /tmp has certain sticky bits set, newly created subdirectories
inherit those bits, causing the test_mkdtemp.test_mode() test to fail.
Remove those before comparing the actual mode to the expected mode.
Fred Drake [Wed, 31 Mar 2004 08:08:34 +0000 (08:08 +0000)]
add a heavy box around warning notices to make them really stand out in the
PDF and PostScript versions of the docs (the CSS already does this for HTML)
Skip Montanaro [Sat, 27 Mar 2004 18:23:11 +0000 (18:23 +0000)]
- add entry for complex number
- fix a couple typos
- refine definitions for "interpreted" and "coercion" based upon updates on
the python glossary wiki
David Ascher [Fri, 26 Mar 2004 15:10:25 +0000 (15:10 +0000)]
Fix test failure for test_tcl on OS/X and Windows if a
version of Tcl other than ActiveTcl is installed (ActiveTcl
included TclX, other Tcl distros didn't).
I'm removing the package loading test because it's hard to
come up with a package that is guaranteed to be in any Tcl installation.
Special-casing darwin and windows is ok since that leaves the
only Tk platform (X) which the test was trying to address.
Simple optimizations:
* pre-build a single identity function for the fixup function
* pre-build membership tests in dictionaries instead of in-line tuples
* assign len() to a local variable
* assign append() methods to a local variable
* use xrange() instead of range()
* replace "x<<1" with "x+x"
Fred Drake [Thu, 25 Mar 2004 16:51:12 +0000 (16:51 +0000)]
Note that reading from a socket may not always return all of the
remaining content of a stream when expected to do so.
Closes SF bug #725265. Should be backported to Python 2.3.x.
Fred Drake [Thu, 25 Mar 2004 16:39:46 +0000 (16:39 +0000)]
- make sure the methods minidom adds to the basic DOM are attributed
to Node objects in the index (closes SF bug #832251)
- fix a variety of markup nits
Someone should backport this patch to Python 2.3.x.
Fred Drake [Thu, 25 Mar 2004 14:25:28 +0000 (14:25 +0000)]
Remove note that PyErr_SetInterrupt() is obsolete; add comment about the
fact that it was marked obsolete but is still needed.
Closes SF bug #919299. Someone else should backport this to Python 2.3.
Brett Cannon [Tue, 23 Mar 2004 21:26:39 +0000 (21:26 +0000)]
Replace code in urllib for basejoin (undocumented) with urlparse.urljoin .
Test suites for urllib and urlparse run with each other's function to verify
correctness of replacement and both test suites pass.
Bumped urllib's __version__ attribute up a minor number.
Fred Drake [Tue, 23 Mar 2004 18:54:12 +0000 (18:54 +0000)]
- use recommended Python style in examples (no spaces around "=" for
keyword args)
- format multi-line calls to distutils.core.setup() consistently, and
in line with general practice (one keyword arg per line,
comma/newline after the last
- fix a few typos
Anthony Baxter [Mon, 22 Mar 2004 22:22:05 +0000 (22:22 +0000)]
Basic dependency checking. setup() has two new optional arguments
requires and provides. requires is a sequence of strings, of the
form 'packagename-version'. The dependency checking so far merely
does an '__import__(packagename)' and checks for packagename.__version__
You can also leave off the version, and any version of the package
will be installed.
There's a special case for the package 'python' - sys.version_info
is used, so
requires= ( 'python-2.3', )
just works.
Provides is of the same format as requires - but if it's not supplied,
a provides is generated by adding the version to each entry in packages,
or modules if packages isn't there.
Provides is currently only used in the PKG-INFO file. Shortly, PyPI
will grow the ability to accept these lines, and register will be
updated to send them.
There's a new command 'checkdep' command that runs these checks.
For this version, only greater-than-or-equal checking is done. We'll
add the ability to specify an optional operator later.