]> granicus.if.org Git - python/log
python
24 years agoKeepconsole is now a 4-way option: never/errorexit/unseen output/always. Default...
Jack Jansen [Fri, 22 Sep 2000 21:50:11 +0000 (21:50 +0000)]
Keepconsole is now a 4-way option: never/errorexit/unseen output/always. Default is "unseen output". Upped the Popt version number.

24 years agoPoke-and-hope attempt to fix Bugs #115006 and #114324: fix the test
Guido van Rossum [Fri, 22 Sep 2000 19:41:56 +0000 (19:41 +0000)]
Poke-and-hope attempt to fix Bugs #115006 and #114324: fix the test
for pthread_t (to calculate its size) to work even if pthread_t is a
struct.

24 years agoMaildir.__init__(): Use the correct filter for filenames, so that this
Fred Drake [Fri, 22 Sep 2000 18:41:50 +0000 (18:41 +0000)]
Maildir.__init__():  Use the correct filter for filenames, so that this
                     class conforms to the maildir specification.

24 years agoAdded refcount information for the *_InPlace*() API series.
Fred Drake [Fri, 22 Sep 2000 18:19:37 +0000 (18:19 +0000)]
Added refcount information for the *_InPlace*() API series.
This closes SourceForge bug #114287.

24 years agoPyNumber_Coerce() returns an int, not a PyObject *.
Fred Drake [Fri, 22 Sep 2000 18:17:49 +0000 (18:17 +0000)]
PyNumber_Coerce() returns an int, not a PyObject *.

24 years agoFix the way we found relevant cfuncdesc lines; PREFIX was not a regular
Fred Drake [Fri, 22 Sep 2000 17:55:32 +0000 (17:55 +0000)]
Fix the way we found relevant cfuncdesc lines; PREFIX was not a regular
expression!

24 years agoIt's better to test for __hpux rather than __hppa, and hpux or hppa is
Guido van Rossum [Fri, 22 Sep 2000 17:26:14 +0000 (17:26 +0000)]
It's better to test for __hpux rather than __hppa, and hpux or hppa is
unnecessary.  Sez edg@SF

24 years agouse_sans_serif(),
Fred Drake [Fri, 22 Sep 2000 17:05:04 +0000 (17:05 +0000)]
use_sans_serif(),
use_italics():  Remove both functions, inlining use_italics() at its
                only call site.

init_myformat():  Uncomment line so that some internal markup does not
                  get generated, since it is not properly removed later.
                  (Fix on aspect of SourceForge bug #114749.)

Modified call to process_commands_wrap_deferred(), removing \code from
the list since it had a bad interaction with other changes in some contexts.

24 years agoUpdate versioning for the next Python release.
Fred Drake [Fri, 22 Sep 2000 16:20:23 +0000 (16:20 +0000)]
Update versioning for the next Python release.

24 years agoUpdate RELEASE for the next Python release.
Fred Drake [Fri, 22 Sep 2000 16:18:19 +0000 (16:18 +0000)]
Update RELEASE for the next Python release.

24 years agoChange HP=UX compiler options from -Aa to -Ae, which implies
Guido van Rossum [Fri, 22 Sep 2000 16:15:54 +0000 (16:15 +0000)]
Change HP=UX compiler options from -Aa to -Ae, which implies
-D_HPUX_SOURCE and also turns on long long support.

Suggestion by stnor@sweden.hp.com (Stefan Norberg).

Please test this if you have access to HP-UX!!!

24 years agoHopefully fix the problem with undeclared fdatasync() on HP-UX that
Guido van Rossum [Fri, 22 Sep 2000 16:01:05 +0000 (16:01 +0000)]
Hopefully fix the problem with undeclared fdatasync() on HP-UX that
was reported twice so far.

Someone with access to HP-UX, please test this!  (Is '__hppa' or
'hppa' really the correct symbol to test for?)

24 years agoFix some minor nits about the use of \optional in parameter lists.
Fred Drake [Fri, 22 Sep 2000 15:46:35 +0000 (15:46 +0000)]
Fix some minor nits about the use of \optional in parameter lists.

24 years agoAddress Bug #115057: add a --with-suffix option to set the EXE
Guido van Rossum [Fri, 22 Sep 2000 15:38:21 +0000 (15:38 +0000)]
Address Bug #115057: add a --with-suffix option to set the EXE
variable in the Makefiles from the configure script.  Usefil for
Cygwin and Mac OS X builds.

24 years ago- plug a memory leak due to circular lists
Neil Schemenauer [Fri, 22 Sep 2000 15:30:16 +0000 (15:30 +0000)]
- plug a memory leak due to circular lists

24 years ago- Replace debugleak flag with findleaks flag. The new SAVEALL GC option is
Neil Schemenauer [Fri, 22 Sep 2000 15:29:28 +0000 (15:29 +0000)]
- Replace debugleak flag with findleaks flag.  The new SAVEALL GC option is
  used to find cyclic garbage produced by tests.

24 years ago- Add test for new SAVEALL debugging flag
Neil Schemenauer [Fri, 22 Sep 2000 15:26:20 +0000 (15:26 +0000)]
- Add test for new SAVEALL debugging flag
- Use exceptions rather than asserts for failing tests.
- Reorganize tests and produce some output if verbose option is set.

24 years ago- Add DEBUG_SAVEALL option. When enabled all garbage objects found by the
Neil Schemenauer [Fri, 22 Sep 2000 15:22:38 +0000 (15:22 +0000)]
- Add DEBUG_SAVEALL option.  When enabled all garbage objects found by the
  collector will be saved in gc.garbage.  This is useful for debugging a
  program that creates reference cycles.

- Fix else statements in gcmodule.c to conform to Python coding standards.

24 years agoWhen PyInt_FromLong() returns NULL, you do not need to check
Fred Drake [Fri, 22 Sep 2000 15:21:31 +0000 (15:21 +0000)]
When PyInt_FromLong() returns NULL, you do not need to check
PyErr_Occurred().  Removed the extra test and setting of a
bogus exception.

24 years agoContributed modules by Riccardo Trocca. Extended pixmap wrapper, NumPy visualiser...
Jack Jansen [Fri, 22 Sep 2000 12:46:19 +0000 (12:46 +0000)]
Contributed modules by Riccardo Trocca. Extended pixmap wrapper, NumPy visualiser and QuickTime to images.

24 years agoAllow lists of files/fsspecs as the source for copy() and move(). By
Jack Jansen [Fri, 22 Sep 2000 12:17:14 +0000 (12:17 +0000)]
Allow lists of files/fsspecs as the source for copy() and move(). By
Bill Bedford, slightly edited by me.

24 years agoImplemented new os.startfile function, unique to Windows, exposing a
Tim Peters [Fri, 22 Sep 2000 10:05:54 +0000 (10:05 +0000)]
Implemented new os.startfile function, unique to Windows, exposing a
subset of Win32 ShellExecute's functionality.  Guido wants this because
IDLE's Help -> Docs function currently crashes his machine because of a
conflict between his version of Norton AntiVirus (6.10.20) and MS's
_popen.  Docs for startfile are being mailed to Fred (or just read the
docstring -- it tells the whole story).
Changed webbrowser.py to use os.startfile instead of os.popen on Windows.
Changed IDLE's EditorWindow.py to pass an absolute path for the docs
(hardcoding ShellExecute's "directory" arg to "." as used to be done let
IDLE work, but made the startfile command exceedingly obscure for other
uses -- the MS docs are terrible, of course, & still not sure I
understand it).
Note that Windows Python must link with shell32.lib now!  That's where
ShellExecute lives.

24 years agoMore whitespace cleanup, to satisfy tabnanny.py. Don't trust -tt!
Guido van Rossum [Fri, 22 Sep 2000 09:30:29 +0000 (09:30 +0000)]
More whitespace cleanup, to satisfy tabnanny.py.  Don't trust -tt!

24 years agoGet rid of the one tab in the file.
Guido van Rossum [Fri, 22 Sep 2000 09:23:08 +0000 (09:23 +0000)]
Get rid of the one tab in the file.
Closes Bug #115054.

24 years agoFix for SF bug 115051: Dodgy use of PyTuple_SET_ITEM in pyexpat.c
Tim Peters [Fri, 22 Sep 2000 06:01:11 +0000 (06:01 +0000)]
Fix for SF bug 115051:  Dodgy use of PyTuple_SET_ITEM in pyexpat.c

24 years agoWhite space cleanup, including one item that was an error under -tt.
Fred Drake [Fri, 22 Sep 2000 05:07:56 +0000 (05:07 +0000)]
White space cleanup, including one item that was an error under -tt.

24 years agoRemove debugging print. ;(
Fred Drake [Fri, 22 Sep 2000 04:49:50 +0000 (04:49 +0000)]
Remove debugging print.  ;(

24 years agoCorrect some bitrot; some things have become inaccurate in the tutorial.
Fred Drake [Fri, 22 Sep 2000 04:12:27 +0000 (04:12 +0000)]
Correct some bitrot; some things have become inaccurate in the tutorial.

<file>.readlines() does not call <file>.readline() internally anymore,
and the sizehint parameter should be mentioned briefly.

Some displays of floating point numbers needed to be updated due to the
change in the repr() of floats (from 1.6).

Both issues were noted by Aahz <aahz@panix.com>.

24 years agoTweak what happens when run on non-Windows platforms: set install prefix
Greg Ward [Fri, 22 Sep 2000 01:32:34 +0000 (01:32 +0000)]
Tweak what happens when run on non-Windows platforms: set install prefix
as well as scheme, and don't convert all installation paths (that's now
done by the "install" command for us).

24 years agoChanged all paths in the INSTALL_SCHEMES dict to Unix syntax, and added
Greg Ward [Fri, 22 Sep 2000 01:31:08 +0000 (01:31 +0000)]
Changed all paths in the INSTALL_SCHEMES dict to Unix syntax, and added
'convert_paths()' method to convert them all to the local syntax (backslash
or colon or whatever) at the appropriate time.

Added SCHEME_KEYS to get rid of one hard-coded list of attributes (in
'select_scheme()').

Default 'install_path_file' to true, and never set it false (it's just
there in case some outsider somewhere wants to disable installation of the
.pth file for whatever reason).

Toned down the warning emitted when 'install_path_file' is false, since we
no longer know why it might be false.

Added 'warn_dir' flag to suppress warning when installing to a directory
not in sys.path (again, we never set this false -- it's there for outsiders
to use, specifically the "bdist_*" commands).

Pulled the loop of 'change_root()' calls out to new method 'change_roots()'.

Comment updates/deletions/additions.

24 years agoFix 'convert_path()' so it returns immediately under Unix -- prevents blowing
Greg Ward [Fri, 22 Sep 2000 01:05:43 +0000 (01:05 +0000)]
Fix 'convert_path()' so it returns immediately under Unix -- prevents blowing
up when the pathname starts with '/', which is needed when converting
installation directories in the "install" command.

24 years agoDenis S. Otkidach <ods@users.sourceforge.net>:
Fred Drake [Thu, 21 Sep 2000 22:27:16 +0000 (22:27 +0000)]
Denis S. Otkidach <ods@users.sourceforge.net>:
Show how code can be written to handle __getslice__ & friends in a way that
is compatible with pre-2.0 versions of Python while still working with the
"new" way of handling slicing.

Additional explanation added by Fred Drake.

This closes SourceForge patch #101388.

24 years agoUntested patch by Ty Sarna to make TELL64 work on older NetBSD systems.
Guido van Rossum [Thu, 21 Sep 2000 22:15:29 +0000 (22:15 +0000)]
Untested patch by Ty Sarna to make TELL64 work on older NetBSD systems.
According to Justin Pettit, this also works on OpenBSD, so I've added
that symbol as well.

24 years agoIndent _connection_class so that it becomes HTTPS._connection_class.
Martin v. Löwis [Thu, 21 Sep 2000 22:09:47 +0000 (22:09 +0000)]
Indent _connection_class so that it becomes HTTPS._connection_class.

24 years agoLots of minor fixes, many suggested by Detlef Lannert
Fred Drake [Thu, 21 Sep 2000 22:01:36 +0000 (22:01 +0000)]
Lots of minor fixes, many suggested by Detlef Lannert
<lannert@uni-duesseldorf.de>.

24 years agoVarious tweaks and bugfixes to GetArgv. It now appears good enough for Distutils.
Jack Jansen [Thu, 21 Sep 2000 22:01:08 +0000 (22:01 +0000)]
Various tweaks and bugfixes to GetArgv. It now appears good enough for Distutils.

24 years agoAdded dependencies on the bug-reporting text.
Fred Drake [Thu, 21 Sep 2000 21:37:09 +0000 (21:37 +0000)]
Added dependencies on the bug-reporting text.

24 years agoInclude the new text on reporting bugs in a few useful places.
Fred Drake [Thu, 21 Sep 2000 21:35:22 +0000 (21:35 +0000)]
Include the new text on reporting bugs in a few useful places.

This closes SourceForge bug #114792.

24 years agoNew text about how to report bugs in Python and the documentation.
Fred Drake [Thu, 21 Sep 2000 21:32:14 +0000 (21:32 +0000)]
New text about how to report bugs in Python and the documentation.

24 years agoUpdated according to the changes made to the "s#" parser marker
Marc-André Lemburg [Thu, 21 Sep 2000 21:21:59 +0000 (21:21 +0000)]
Updated according to the changes made to the "s#" parser marker
and bumped the version number to 1.7.

24 years agoAdded a true unicode_internal_encode function and fixed the
Marc-André Lemburg [Thu, 21 Sep 2000 21:09:45 +0000 (21:09 +0000)]
Added a true unicode_internal_encode function and fixed the
unicode_internal_decode function to support Unicode objects
directly rather than by generating a copy of the object.

24 years agoSpecial case the "s#" PyArg_Parse() token for Unicode objects:
Marc-André Lemburg [Thu, 21 Sep 2000 21:08:30 +0000 (21:08 +0000)]
Special case the "s#" PyArg_Parse() token for Unicode objects:
"s#" will now return a pointer to the default encoded string data
of the Unicode object instead of a pointer to the raw UTF-16
data.

The latter is still available via PyObject_AsReadBuffer().

The patch also adds an optimization for string objects which is
based on the fact that string objects return the raw character data
for getreadbuffer access and are always single-segment.

24 years agoSpecial case the "s#" PyArg_Parse() token for Unicode objects:
Marc-André Lemburg [Thu, 21 Sep 2000 21:08:08 +0000 (21:08 +0000)]
Special case the "s#" PyArg_Parse() token for Unicode objects:
"s#" will now return a pointer to the default encoded string data
of the Unicode object instead of a pointer to the raw UTF-16
data.

The latter is still available via PyObject_AsReadBuffer().

24 years agoRevise the test case for pyexpat to avoid using asserts. Conform better
Fred Drake [Thu, 21 Sep 2000 20:32:13 +0000 (20:32 +0000)]
Revise the test case for pyexpat to avoid using asserts.  Conform better
to the Python style guide, and remove unneeded imports.

24 years agoThe minidom.Node class has a debug attribute which, when its _debug
Guido van Rossum [Thu, 21 Sep 2000 20:10:39 +0000 (20:10 +0000)]
The minidom.Node class has a debug attribute which, when its _debug
flag is true, is set to a StringIO object that silently collects all
debug messages.  This is triggered by the Node._debug=1 statement at
the top of test_minidom.py.  After the tests, we better delete that
StringIO object to avoid wasting memory.  We also reset the _debug
flag.  (Note that this is an undetectable memory leak, and the memory
doesn't get collected by the cycle-gc either, because it's all
reachable -- it's just useless.)

24 years agoRemove memory leaks of strings/Unicode objects passed into the character
Fred Drake [Thu, 21 Sep 2000 20:10:23 +0000 (20:10 +0000)]
Remove memory leaks of strings/Unicode objects passed into the character
data and default handlers -- a new reference was being passed to
Py_BuildValue() for the "O" format character; using "N" plugs the leak.

Fixed two other (minor) leaks that occurred on various error conditions.

Removed uses of the UNLESS macro, which makes code hard to read, and is
Evil.

24 years agoPaul Prescod <paul@prescod.net>:
Fred Drake [Thu, 21 Sep 2000 17:43:48 +0000 (17:43 +0000)]
Paul Prescod <paul@prescod.net>:
Add support for parsing already-opened files.  Make sure the parse()
method closes exactly those files that it opens.

Modified by FLD for better conformance to the Python style guide.

This closes SourceForge patch #101512.

24 years ago- fixed yet another gcc -pedantic warning
Fredrik Lundh [Thu, 21 Sep 2000 17:03:25 +0000 (17:03 +0000)]
- fixed yet another gcc -pedantic warning
- added experimental "expand" method to match objects
- don't use the buffer interface on unicode strings

24 years agoSAXException.__getitem__(): Raise AttributeError instead of NameError.
Fred Drake [Thu, 21 Sep 2000 16:32:28 +0000 (16:32 +0000)]
SAXException.__getitem__():  Raise AttributeError instead of NameError.

24 years agoAs suggested by Toby Dickenson, setting ob_type to NULL in
Guido van Rossum [Thu, 21 Sep 2000 16:25:33 +0000 (16:25 +0000)]
As suggested by Toby Dickenson, setting ob_type to NULL in
_Py_Dealloc(), is a bad idea (and always was!).  So let's drop it.

24 years agoConvert the longest two tables from tableii to longtableii so they do not
Fred Drake [Thu, 21 Sep 2000 16:04:08 +0000 (16:04 +0000)]
Convert the longest two tables from tableii to longtableii so they do not
make too big a mess.  One actually did not fit on a single page at all!

24 years agoAdd new environments: longtableii, longtableiii, and longtableiv, to
Fred Drake [Thu, 21 Sep 2000 15:58:02 +0000 (15:58 +0000)]
Add new environments: longtableii, longtableiii, and longtableiv, to
support long tables which might break across page boundaries.  Otherwise
identical to tableii, tableiii, and tableiv.

24 years agoAllow the LaTeX markup descriptions to use the \moreargs and \unspecified
Fred Drake [Thu, 21 Sep 2000 15:53:54 +0000 (15:53 +0000)]
Allow the LaTeX markup descriptions to use the \moreargs and \unspecified
macros in the argument list position.

24 years agofix bogus references to imp; makes test_minidom succeed
Jeremy Hylton [Thu, 21 Sep 2000 14:46:56 +0000 (14:46 +0000)]
fix bogus references to imp; makes test_minidom succeed

24 years agoUse PyOS_setsig() instead of signal(). Also remove redundant spaces
Guido van Rossum [Thu, 21 Sep 2000 14:32:04 +0000 (14:32 +0000)]
Use PyOS_setsig() instead of signal().  Also remove redundant spaces
from the FreeBSD code.

24 years agoNow uses make_parser to create its parser (patch 101573).
Lars Gustäbel [Thu, 21 Sep 2000 08:38:46 +0000 (08:38 +0000)]
Now uses make_parser to create its parser (patch 101573).

24 years agoAdded the make_parser function (patch 101571).
Lars Gustäbel [Thu, 21 Sep 2000 08:34:04 +0000 (08:34 +0000)]
Added the make_parser function (patch 101571).

24 years agoUpdated XMLGenerator to new DocumentHandler interface (patch 101572).
Lars Gustäbel [Thu, 21 Sep 2000 08:25:28 +0000 (08:25 +0000)]
Updated XMLGenerator to new DocumentHandler interface (patch 101572).

24 years agoUpdated to correct DocumentHandler signatures. (patch 101570)
Lars Gustäbel [Thu, 21 Sep 2000 08:18:55 +0000 (08:18 +0000)]
Updated to correct DocumentHandler signatures. (patch 101570)

24 years agoWindows installer:
Tim Peters [Thu, 21 Sep 2000 07:50:36 +0000 (07:50 +0000)]
Windows installer:
    Don't ship debug .dll, .pyd or .lib files.  Saves space.
    Bumped the title to beta 2.

24 years agoDerived from Martin's SF patch 110609: support unbounded ints in %d,i,u,x,X,o formats.
Tim Peters [Thu, 21 Sep 2000 05:43:11 +0000 (05:43 +0000)]
Derived from Martin's SF patch 110609:  support unbounded ints in %d,i,u,x,X,o formats.
Note a curious extension to the std C rules:  x, X and o formatting can never produce
a sign character in C, so the '+' and ' ' flags are meaningless for them.  But
unbounded ints *can* produce a sign character under these conversions (no fixed-
width bitstring is wide enough to hold all negative values in 2's-comp form).  So
these flags become meaningful in Python when formatting a Python long which is too
big to fit in a C long.  This required shuffling around existing code, which hacked
x and X conversions to death when both the '#' and '0' flags were specified:  the
hacks weren't strong enough to deal with the simultaneous possibility of the ' ' or
'+' flags too, since signs were always meaningless before for x and X conversions.
Isomorphic shuffling was required in unicodeobject.c.
Also added dozens of non-trivial new unbounded-int test cases to test_format.py.

24 years agoNote that __getitem__() may receive a slice object as the index;
Fred Drake [Thu, 21 Sep 2000 05:28:26 +0000 (05:28 +0000)]
Note that __getitem__() may receive a slice object as the index;
reported by Detlef Lannert <lannert@uni-duesseldorf.de>.

24 years agoDocument directories are not always named with three characters any more;
Fred Drake [Thu, 21 Sep 2000 05:26:43 +0000 (05:26 +0000)]
Document directories are not always named with three characters any more;
reported by Detlef Lannert <lannert@uni-duesseldorf.de>.

24 years agoFixed a number of small problems reported by Detlef Lannert
Fred Drake [Thu, 21 Sep 2000 05:25:30 +0000 (05:25 +0000)]
Fixed a number of small problems reported by Detlef Lannert
<lannert@uni-duesseldorf.de>.

24 years agoCorran Webster: fix 'change_root()' to handle Mac OS paths.
Greg Ward [Thu, 21 Sep 2000 01:23:35 +0000 (01:23 +0000)]
Corran Webster: fix 'change_root()' to handle Mac OS paths.

24 years agoOn Unix, use O_EXCL when creating the .pyc/.pyo files, to avoid a race condition
Guido van Rossum [Wed, 20 Sep 2000 20:31:38 +0000 (20:31 +0000)]
On Unix, use O_EXCL when creating the .pyc/.pyo files, to avoid a race condition

24 years agoAdd a definition of rl_library_version. Reported by jpettit@sourceforge.
Guido van Rossum [Wed, 20 Sep 2000 20:24:21 +0000 (20:24 +0000)]
Add a definition of rl_library_version. Reported by jpettit@sourceforge.

24 years ago.use_latex(): Make this a little smarter so that it only runs pdflatex
Fred Drake [Wed, 20 Sep 2000 05:49:09 +0000 (05:49 +0000)]
.use_latex():  Make this a little smarter so that it only runs pdflatex
               once if all the temporary files are available from building
               a DVI file.  This can avoid two runs of pdflatex.

24 years agoClarify that the softspace attribute is used by print for state management,
Fred Drake [Wed, 20 Sep 2000 04:48:20 +0000 (04:48 +0000)]
Clarify that the softspace attribute is used by print for state management,
not as a user-controlled parameter.

24 years agoAdd entries for the xml.sax documentation.
Fred Drake [Wed, 20 Sep 2000 02:54:58 +0000 (02:54 +0000)]
Add entries for the xml.sax documentation.

24 years agoFirst cut at documentation for the xml.sax package (not including any
Fred Drake [Wed, 20 Sep 2000 02:52:20 +0000 (02:52 +0000)]
First cut at documentation for the xml.sax package (not including any
sub-modules).

24 years agochange 2-space indent to 4-space indent
Jeremy Hylton [Wed, 20 Sep 2000 02:47:28 +0000 (02:47 +0000)]
change 2-space indent to 4-space indent

24 years agodistutils setup script
Jeremy Hylton [Wed, 20 Sep 2000 02:35:19 +0000 (02:35 +0000)]
distutils setup script

24 years agoFix typo (newtabwith).
Guido van Rossum [Wed, 20 Sep 2000 00:17:39 +0000 (00:17 +0000)]
Fix typo (newtabwith).

24 years ago*Very* belated application of Thomas Heller's patch to handle
Greg Ward [Tue, 19 Sep 2000 23:56:43 +0000 (23:56 +0000)]
*Very* belated application of Thomas Heller's patch to handle
resource files.  The gist of the patch is to treat ".rc" and ".mc"
files as source files; ".mc" files are compiled to ".rc" and then
".res", and ".rc" files are compiled to ".res".  Wish I knew what
all these things stood for...

24 years agoTreat trailing colon in os.path.join("a:", "b") same way for DOS as in
Tim Peters [Tue, 19 Sep 2000 23:46:56 +0000 (23:46 +0000)]
Treat trailing colon in os.path.join("a:", "b") same way for DOS as in
recent changes to ntpath.py and posixmodule.c.  Thanks to Guido for
pointing out the inconsistency!

24 years agoAdded GetArgs dialog to EasyDialogs, a very nifty (if I may say so:-) way to create...
Jack Jansen [Tue, 19 Sep 2000 22:42:38 +0000 (22:42 +0000)]
Added GetArgs dialog to EasyDialogs, a very nifty (if I may say so:-) way to create a unix-style sys.argv.

24 years agoThis patch adds a new Python C API called PyString_AsStringAndSize()
Marc-André Lemburg [Tue, 19 Sep 2000 21:04:18 +0000 (21:04 +0000)]
This patch adds a new Python C API called PyString_AsStringAndSize()
which implements the automatic conversion from Unicode to a string
object using the default encoding.

The new API is then put to use to have eval() and exec accept
Unicode objects as code parameter. This closes bugs #110924
and #113890.

As side-effect, the traditional C APIs PyString_Size() and
PyString_AsString() will also accept Unicode objects as
parameters.

24 years agoTemporary fix for Bug #114821.
Guido van Rossum [Tue, 19 Sep 2000 20:51:17 +0000 (20:51 +0000)]
Temporary fix for Bug #114821.

The cause was that the replace code necessarily used a PCRE internal
function to to template expansion.

The fix changes the code to use an SRE internal if SRE is used, and a
PCRE internal if SRE is used; in a way that should work with 1.5.2.

The solution can be sped up tremendously under the assumption that the
choice between sre and pre is not changed during the execution of the
program; especially replace-all will be slow.

But I'll leave that to someone else.

24 years agoPart of SF patch 101481: on Windows, os.path.join("a:", "b") should yield
Tim Peters [Tue, 19 Sep 2000 20:39:32 +0000 (20:39 +0000)]
Part of SF patch 101481:  on Windows, os.path.join("a:", "b") should yield
"a:b", not "a:/b".  Similar change was made to posixmodule.c earlier.

24 years agotest_userlist.py:
Tim Peters [Tue, 19 Sep 2000 20:31:24 +0000 (20:31 +0000)]
test_userlist.py:
    Added new test for new __contains__ method.
    Extensive editing to get rid of asserts.

24 years agoSF patch 101391: implemented UserList.__contains__.
Tim Peters [Tue, 19 Sep 2000 20:29:03 +0000 (20:29 +0000)]
SF patch 101391:  implemented UserList.__contains__.

24 years agoFix test errors.
Paul Prescod [Tue, 19 Sep 2000 19:33:02 +0000 (19:33 +0000)]
Fix test errors.

24 years agoMoshe Zadka <mzadka@geocities.com>:
Fred Drake [Tue, 19 Sep 2000 18:21:25 +0000 (18:21 +0000)]
Moshe Zadka <mzadka@geocities.com>:
Document the __contains__() method.
(Patch slightly modified by FLD.)

This closes SourceForge patch #101387.

24 years agoMany small clarifications, including many suggested by email.
Fred Drake [Tue, 19 Sep 2000 18:10:05 +0000 (18:10 +0000)]
Many small clarifications, including many suggested by email.

24 years agoNew test cases for the StringIO module
Martin v. Löwis [Tue, 19 Sep 2000 16:35:39 +0000 (16:35 +0000)]
New test cases for the StringIO module

24 years agoDon't print specific Node instances unless running verbosely.
Martin v. Löwis [Tue, 19 Sep 2000 16:22:10 +0000 (16:22 +0000)]
Don't print specific Node instances unless running verbosely.
Closes Bug #114775.

24 years agomake_nav_sectref(): Create more minimal HTML, hanging the class attribute
Fred Drake [Tue, 19 Sep 2000 15:36:19 +0000 (15:36 +0000)]
make_nav_sectref():  Create more minimal HTML, hanging the class attribute
                     off an existing anchor tag if available (I think it
                     always is, but am not completely sure).

24 years agoClarify a number of issues about the file-like object API based on
Fred Drake [Tue, 19 Sep 2000 15:18:51 +0000 (15:18 +0000)]
Clarify a number of issues about the file-like object API based on
discussion on python-dev.

24 years agoFixed the error reporting (raise of TestFailed) for the zip() and
Barry Warsaw [Tue, 19 Sep 2000 14:42:09 +0000 (14:42 +0000)]
Fixed the error reporting (raise of TestFailed) for the zip() and
zip(None) tests.  Found by Finn Bock a while ago.

24 years agoRandall Hopper: Add fpectl functionality patch for FreeBSD.
Guido van Rossum [Tue, 19 Sep 2000 13:35:40 +0000 (13:35 +0000)]
Randall Hopper: Add fpectl functionality patch for FreeBSD.

24 years agoSupport sizehint in _fileobject.readlines, as documented.
Martin v. Löwis [Tue, 19 Sep 2000 11:25:58 +0000 (11:25 +0000)]
Support sizehint in _fileobject.readlines, as documented.

24 years agoSet the 'nt' installation scheme for the install command even if run
Thomas Heller [Tue, 19 Sep 2000 11:10:23 +0000 (11:10 +0000)]
Set the 'nt' installation scheme for the install command even if run
on other systems, so that data, headers, scripts are included in
the installer.

24 years agoSupport sizehint in StringIO.readlines, as documented.
Martin v. Löwis [Tue, 19 Sep 2000 11:07:44 +0000 (11:07 +0000)]
Support sizehint in StringIO.readlines, as documented.

24 years agoImplement readlines function. Closes Bug #110686.
Martin v. Löwis [Tue, 19 Sep 2000 11:06:46 +0000 (11:06 +0000)]
Implement readlines function. Closes Bug #110686.

24 years agoObscure marshal fixes:
Tim Peters [Tue, 19 Sep 2000 08:54:13 +0000 (08:54 +0000)]
Obscure marshal fixes:
    When reading a short, sign-extend on platforms where shorts are
    bigger than 16 bits.
    When reading a long, repair the unportable sign extension that was
    being done for 64-bit machines (it assumed that signed right shift
    sign-extends).

24 years agoPatch #101121, by Ka-Ping Yee: cosmetic cleanup of cgi.py, using my
Guido van Rossum [Tue, 19 Sep 2000 04:11:46 +0000 (04:11 +0000)]
Patch #101121, by Ka-Ping Yee: cosmetic cleanup of cgi.py, using my
style conventions.  (Ping has checkin privileges but apparently
ignores them at the moment.)

Ping improves a few doc strings and fixes style violations like foo ( bar ).

An addition of my own: rearrange the printing of various items in
test() so that the (long) environment comes at the end.  This avoids
having to scroll if you want to see the current directory or command
line arguments.

24 years agoAn honest attempt to make this work on Unix, Windows, and even
Guido van Rossum [Tue, 19 Sep 2000 04:01:01 +0000 (04:01 +0000)]
An honest attempt to make this work on Unix, Windows, and even
Macintosh (the latter untested).

This closes Bug #110839.

24 years agoOnly supply popen2, popen3 when fork exists.
Guido van Rossum [Tue, 19 Sep 2000 03:04:52 +0000 (03:04 +0000)]
Only supply popen2, popen3 when fork exists.
(This avoids defining non-working versions of these on the Mac.)

24 years agoMake better use of GNU Pth -- patch by Andy Dustman.
Guido van Rossum [Tue, 19 Sep 2000 00:46:46 +0000 (00:46 +0000)]
Make better use of GNU Pth -- patch by Andy Dustman.
I can't test this, so I'm just checking it in with blind faith in Andy.
I've tested that it doesn't broeak a non-Pth build on Linux.

Changes include:

- There's a --with-pth configure option.

- Instead of _GNU_PTH, we test for HAVE_PTH.

- Better signal handling.

- (The config.h.in file is regenerated in a slightly different order.)