Peepholer could be fooled into misidentifying a tuple_of_constants.
Added code to count consecutive occurrences of LOAD_CONST.
Use the count to weed out the misidentified cases.
Added a unittest.
Just van Rossum [Tue, 26 Oct 2004 07:20:26 +0000 (07:20 +0000)]
- Removed Date class. We don't really need it for b/w compatibility since
a) the functionality depended on PyXML before and b) hardly worked to
begin with.
- Instead, output and require upon input datetime.datetime objects.
Just van Rossum [Tue, 26 Oct 2004 06:50:50 +0000 (06:50 +0000)]
- added two more convenience functions: readPlistFromString() and
writePlistToString()
- use these two in the resource functions.
- Tweaked module doc string.
Tim Peters [Sun, 24 Oct 2004 23:45:42 +0000 (23:45 +0000)]
format_paragraph_event(): Patch 961387 introduced a bug here, causing
the indentation of a comment block to be ignored when reformatting the
block, leading to overly long reformatted lines (too wide by an amount
equal to the indentation width). Looks like a typo in the original
patch, a 1-character repair.
Peter Astrand [Thu, 21 Oct 2004 19:28:34 +0000 (19:28 +0000)]
Removed test_close_fds, because it's too unreliable. We simply cannot
know that the newly-started Python process only has 3 filedescriptors
open. Fixes bug 1048808.
Tim Peters [Thu, 14 Oct 2004 04:16:54 +0000 (04:16 +0000)]
remove_stderr_debug_decorations(): Always try the substitution. Else
this test failed under the combination of passing -O to a debug-build
Python. Now all 4 of those pass ({debug, release} x {-O, no -O}).
Anthony Baxter [Wed, 13 Oct 2004 15:54:17 +0000 (15:54 +0000)]
Patch 983206: distutils obeys LDSHARED env var. Removed the code in
Python's own setup.py that did the same thing (and tested on Solaris,
where LDSHARED is needed...)
Anthony Baxter [Wed, 13 Oct 2004 12:35:28 +0000 (12:35 +0000)]
Backing out the basic dependency checking (from pycon sprint).
This support was only a first cut, and doesn't deserve to be in
a released version (where we have to support it in an ongoing
manner)
Tim Peters [Wed, 13 Oct 2004 04:07:12 +0000 (04:07 +0000)]
New helper remove_stderr_debug_decorations(). This test passes in a
debug build on Windows now. More applications of the helper may be needed
on non-Windows platforms.
Tim Peters [Wed, 13 Oct 2004 03:43:40 +0000 (03:43 +0000)]
Kill several problems at once: test_poll() failed sometimes for me.
Turns out the mysterious "expected output" file contained exactly N dots,
because test_poll() has a loop that *usually* went around N times,
printing one dot on each loop trip. But there's no guarantee of that,
because the exact value of N depended on the vagaries of scheduling
time.sleep()s across two different processes. So stopped printing dots,
and got rid of the expected output file. Add a loop counter instead,
and verify that the loop goes around at least a couple of times. Also
cut the minimum time needed for this test from 4 seconds to 1.
Tim Peters [Wed, 13 Oct 2004 03:29:54 +0000 (03:29 +0000)]
test_stdout_none(): Don't print "banana" to the screen in the middle
of the test. It's testing stdout in a different process, so it has to
print something, but I didn't find "banana" to be self-explanatory.
Tim Peters [Wed, 13 Oct 2004 03:21:35 +0000 (03:21 +0000)]
Windows test_creationflags() test: print msg to stderr informing the
tester that a DOS box is expected to flash. Slash the sleep from 2
seconds to a quarter second (why would we want to wait 2 seconds just
to stare at a DOS box?).
Tim Peters [Wed, 13 Oct 2004 03:14:40 +0000 (03:14 +0000)]
XXX about extreme expense of test_no_leaking() on Windows. I'm not sure
what this is trying to do. If it's necessary for it to create > 1000
processes, it should be controlled by a new resource and not run by
default on Windows.
Tim Peters [Tue, 12 Oct 2004 22:29:54 +0000 (22:29 +0000)]
Experience with Zope2's tests showed it's a Bad Idea to make unittest
display a test's docstring as "the name" of the test. So changed most
test docstrings to comments, and removed the clearly useless ones. Now
unittest reports the actual names of the test methods.
* Use fresh copy of globals/locals so the script being debugged can't access
the pdb namespace (e.g.: p line_prefix will no longer work).
* Remove pdb.py's path from sys.path. Having it in there is normally not a
problem, but it could prove irritating when messing with PYTHONPATH or
invoking pdb via /usr/bin/pdf.
* You can now set a breakpoint on the script being debugged, even if the script
doesn't end with a '.py' extension. Also, setting breakpoints with absolute
paths now works reliably.
Enhancements:
* Go directly to the first line of the script.
* Enter post-mortem debugging if the script being debugged doesn't catch an
exception.
* Restart the script being debugged and preserve debugger state when the script
being debugged exits.
Cleanup:
* Moved the __main__ method into a main() function.
* Kill the (undocumented, not in __all__) mainmodule/mainpyfile globals, add a
mainpyfile attribute to pdb.
projects / python / log