]> granicus.if.org Git - python/log
python
24 years agomake_pair(): When comparing the pointers, they must be cast to integer
Barry Warsaw [Fri, 18 Aug 2000 05:01:19 +0000 (05:01 +0000)]
make_pair(): When comparing the pointers, they must be cast to integer
types (i.e. Py_uintptr_t, our spelling of C9X's uintptr_t).  ANSI
specifies that pointer compares other than == and != to non-related
structures are undefined.  This quiets an Insure portability warning.

24 years agoPyFloat_FromString(): Move s_buffer[] up to the top-level function
Barry Warsaw [Fri, 18 Aug 2000 05:00:03 +0000 (05:00 +0000)]
PyFloat_FromString(): Move s_buffer[] up to the top-level function
scope.  Previously, s_buffer[] was defined inside the
PyUnicode_Check() scope, but referred to in the outer scope via
assignment to s.  This quiets an Insure portability warning.

24 years agoPyInstance_DoBinOp(): When comparing the pointers, they must be cast
Barry Warsaw [Fri, 18 Aug 2000 04:57:32 +0000 (04:57 +0000)]
PyInstance_DoBinOp(): When comparing the pointers, they must be cast
to integer types (i.e. Py_uintptr_t, our spelling of C9X's uintptr_t).
ANSI specifies that pointer compares other than == and != to
non-related structures are undefined.  This quiets an Insure
portability warning.

24 years agoAdded test for uintptr_t, the C9X acceptable way to spell "type to
Barry Warsaw [Fri, 18 Aug 2000 04:53:33 +0000 (04:53 +0000)]
Added test for uintptr_t, the C9X acceptable way to spell "type to
which I can cast void* to and back again without losing information".
In pyport.h, we typedef Py_uintptr_t to mean this thing, which if the
platform supports, will be uintptr_t (otherwise, other accomodations
are made).

24 years agoAdded #undef HAVE_UINTPTR_T for autoconf's delight.
Barry Warsaw [Fri, 18 Aug 2000 04:50:27 +0000 (04:50 +0000)]
Added #undef HAVE_UINTPTR_T for autoconf's delight.

24 years agoMoved LONG_LONG #define from here to pyport.h.
Barry Warsaw [Fri, 18 Aug 2000 04:48:56 +0000 (04:48 +0000)]
Moved LONG_LONG #define from here to pyport.h.

24 years agoMoved LONG_LONG #define from longobject.h to here, since it's needed
Barry Warsaw [Fri, 18 Aug 2000 04:48:18 +0000 (04:48 +0000)]
Moved LONG_LONG #define from longobject.h to here, since it's needed
by the following.

typedef in a portable way the Python name for the C9X uintptr_t type.
This latter is the most portable way to spell an integral type to
which a void* can be cast to and back again without losing
information.  Parallel checkin hacks configure to check if the
platform/compiler supports the C9X name.

24 years agoAs pointed out by Denis S. Otkidach <den@analyt.chem.msu.ru>, xrange()
Fred Drake [Fri, 18 Aug 2000 03:12:38 +0000 (03:12 +0000)]
As pointed out by Denis S. Otkidach <den@analyt.chem.msu.ru>, xrange()
returns an xrange object, not a range object, despite the name of the
source file they're implemented in.

In the list of comparison operators, list != before <>, since <> is
described as obsolescent.

24 years agoAdjust the way __getslice__() is marked as deprecated; this will also
Fred Drake [Fri, 18 Aug 2000 02:42:14 +0000 (02:42 +0000)]
Adjust the way __getslice__() is marked as deprecated; this will also
stand out more.

24 years agoFix a markup error that caused formatting to fail.
Fred Drake [Fri, 18 Aug 2000 02:15:55 +0000 (02:15 +0000)]
Fix a markup error that caused formatting to fail.
Lots of minor markup adjustments as well.

24 years agoMention the new 'import X as Y' syntax
Andrew M. Kuchling [Thu, 17 Aug 2000 23:37:01 +0000 (23:37 +0000)]
Mention the new 'import X as Y' syntax

24 years agoDocument the returns_unicode attribute
Andrew M. Kuchling [Thu, 17 Aug 2000 23:15:21 +0000 (23:15 +0000)]
Document the returns_unicode attribute

24 years agoUpdate to reflect the recent Grammar changes.
Fred Drake [Thu, 17 Aug 2000 23:08:05 +0000 (23:08 +0000)]
Update to reflect the recent Grammar changes.

24 years agoApply SF patch #101135, adding 'import module as m' and 'from module import
Thomas Wouters [Thu, 17 Aug 2000 22:55:00 +0000 (22:55 +0000)]
Apply SF patch #101135, adding 'import module as m' and 'from module import
name as n'. By doing some twists and turns, "as" is not a reserved word.

There is a slight change in semantics for 'from module import name' (it will
now honour the 'global' keyword) but only in cases that are explicitly
undocumented.

24 years agoApply SF patch #101029: call __getitem__ with a proper slice object if there
Thomas Wouters [Thu, 17 Aug 2000 22:37:32 +0000 (22:37 +0000)]
Apply SF patch #101029: call __getitem__ with a proper slice object if there
is no __getslice__ available. Also does the same for C extension types.
Includes rudimentary documentation (it could use a cross reference to the
section on slice objects, I couldn't figure out how to do that) and a test
suite for all Python __hooks__ I could think of, including the new
behaviour.

24 years agoRemove an item that's been done.
Fred Drake [Thu, 17 Aug 2000 22:32:01 +0000 (22:32 +0000)]
Remove an item that's been done.

24 years agoA little more text, and some really minor cleanups.
Fred Drake [Thu, 17 Aug 2000 22:31:23 +0000 (22:31 +0000)]
A little more text, and some really minor cleanups.

24 years agoSmall cleanups, and note when zip() entered the menagerie.
Fred Drake [Thu, 17 Aug 2000 22:30:30 +0000 (22:30 +0000)]
Small cleanups, and note when zip() entered the menagerie.

24 years agoRevise the comments about the exceptions module to not refer to source
Fred Drake [Thu, 17 Aug 2000 22:29:31 +0000 (22:29 +0000)]
Revise the comments about the exceptions module to not refer to source
code; it is not sufficiently readable now that it in written in C, and
is less likely to be available to end users.

24 years agoMartin v. Loewis <martin@loewis.home.cs.tu-berlin.de>:
Fred Drake [Thu, 17 Aug 2000 22:19:26 +0000 (22:19 +0000)]
Martin v. Loewis <martin@loewis.home.cs.tu-berlin.de>:
Document opcodes added to support extended call syntax.

24 years agoPackage with suites for talking to CodeWarrior.
Jack Jansen [Thu, 17 Aug 2000 22:16:11 +0000 (22:16 +0000)]
Package with suites for talking to CodeWarrior.

24 years agoPackage with suites for talking to netscape.
Jack Jansen [Thu, 17 Aug 2000 22:15:41 +0000 (22:15 +0000)]
Package with suites for talking to netscape.

24 years agoThe package with standard suites. These are used separately and as base classes for...
Jack Jansen [Thu, 17 Aug 2000 22:14:57 +0000 (22:14 +0000)]
The package with standard suites. These are used separately and as base classes for other suite packages (StdSuite is slightly magical, in that it is the gensuitemodule default base package).

24 years agoUse the new package-based scripting support.
Jack Jansen [Thu, 17 Aug 2000 22:12:12 +0000 (22:12 +0000)]
Use the new package-based scripting support.

24 years agoGenerate packages in stead of separate modules. The package main module imports every...
Jack Jansen [Thu, 17 Aug 2000 22:11:45 +0000 (22:11 +0000)]
Generate packages in stead of separate modules. The package main module imports everything, it knows about the app signature, suites can extend standard suites, and lots more. Automatically finding declarations in other suites TBD.

24 years agoRemoved temporary code to disable OT networking (this was a workaround for getpeernam...
Jack Jansen [Thu, 17 Aug 2000 20:04:26 +0000 (20:04 +0000)]
Removed temporary code to disable OT networking (this was a workaround for getpeername() not working in a previous release of GUSI, but it has been fixed).

24 years agoRemoved temporary code that always set creator to SimpleText.
Jack Jansen [Thu, 17 Aug 2000 20:01:24 +0000 (20:01 +0000)]
Removed temporary code that always set creator to SimpleText.

24 years agoReran autoconf.
Sjoerd Mullender [Thu, 17 Aug 2000 11:41:06 +0000 (11:41 +0000)]
Reran autoconf.

24 years agoFix up problems when compiling in a directory other than the source
Sjoerd Mullender [Thu, 17 Aug 2000 11:38:18 +0000 (11:38 +0000)]
Fix up problems when compiling in a directory other than the source
directory.

24 years agoConvert some old-style string exceptions to class exceptions.
Fred Drake [Thu, 17 Aug 2000 05:06:49 +0000 (05:06 +0000)]
Convert some old-style string exceptions to class exceptions.

24 years agoConvert some old-style string exceptions to class exceptions.
Fred Drake [Thu, 17 Aug 2000 04:45:13 +0000 (04:45 +0000)]
Convert some old-style string exceptions to class exceptions.

24 years ago\citetitle: Format the title, not the optional URL!
Fred Drake [Thu, 17 Aug 2000 04:25:59 +0000 (04:25 +0000)]
\citetitle:  Format the title, not the optional URL!

24 years agoAdd section on list comprehension
Andrew M. Kuchling [Thu, 17 Aug 2000 00:27:06 +0000 (00:27 +0000)]
Add section on list comprehension
Comment out the unwritten XML section
mymalloc.h -> pymem.h

24 years agoInsure properly identifies the `interned' dictionary as leaking at
Barry Warsaw [Wed, 16 Aug 2000 23:41:01 +0000 (23:41 +0000)]
Insure properly identifies the `interned' dictionary as leaking at
shutdown time, but CVS log entry for revision 2.45 explains why this
is so.  Simply include a comment so we don't have to re-figure it out
again 5 years from now.

24 years ago_PySys_Init(): Fix another Insure discovered memory leak; the PyString
Barry Warsaw [Wed, 16 Aug 2000 23:03:57 +0000 (23:03 +0000)]
_PySys_Init(): Fix another Insure discovered memory leak; the PyString
created from the "big"/"little" constant needs to be decref'd.

24 years agoThe sre test suite currently overruns the stack on Win64, Linux64, and Monterey
Trent Mick [Wed, 16 Aug 2000 22:29:55 +0000 (22:29 +0000)]
The sre test suite currently overruns the stack on Win64, Linux64, and Monterey
(64-bit AIX) This is because the RECURSION_LIMIT is too low. This patch lowers
to recusion limit to 7500 such that the recusion check fires before a segfault.

Fredrik suggested/approved the fix in private email, modulo sre's recusion
limit checking no being necessary when PyOS_CheckStack is implemented for
Windows.

24 years agoKa-Ping Yee <ping@lfw.org>:
Fred Drake [Wed, 16 Aug 2000 21:44:03 +0000 (21:44 +0000)]
Ka-Ping Yee <ping@lfw.org>:
Further examples of list comprehensions.

24 years agoupdated occurences of fqdn algorithm (closes patch #101197)
Peter Schneider-Kamp [Wed, 16 Aug 2000 20:30:21 +0000 (20:30 +0000)]
updated occurences of fqdn algorithm (closes patch #101197)

24 years agoSyntaxError__str__(): Fix two memory problems discovered by Insure.
Barry Warsaw [Wed, 16 Aug 2000 19:43:17 +0000 (19:43 +0000)]
SyntaxError__str__(): Fix two memory problems discovered by Insure.
First, the allocated buffer was never freed after using it to create
the PyString object.  Second, it was possible that have_filename would
be false (meaning that filename was not a PyString object), but that
the code would still try to PyString_GET_SIZE() it.

24 years agoadd dummy 'add2lib' target to Grammar/Makefile so non-GNU makes don't bail out
Trent Mick [Wed, 16 Aug 2000 19:02:06 +0000 (19:02 +0000)]
add dummy 'add2lib' target to Grammar/Makefile so non-GNU makes don't bail out

This closes patch:
http://sourceforge.net/patch/index.php?func=detailpatch&patch_id=101176&group_id=5470

24 years agoUse socket.getfqdn() instead of defining make_fqdn().
Fred Drake [Wed, 16 Aug 2000 14:26:22 +0000 (14:26 +0000)]
Use socket.getfqdn() instead of defining make_fqdn().

24 years agoDocument socket.getfqdn().
Fred Drake [Wed, 16 Aug 2000 14:21:42 +0000 (14:21 +0000)]
Document socket.getfqdn().

24 years agoThe socket module is now _socket on all platforms.
Fred Drake [Wed, 16 Aug 2000 14:19:43 +0000 (14:19 +0000)]
The socket module is now _socket on all platforms.

24 years agoRemove a lot of the confusing conditional compilation from the beginning
Fred Drake [Wed, 16 Aug 2000 14:18:30 +0000 (14:18 +0000)]
Remove a lot of the confusing conditional compilation from the beginning
of the init_socket() function.  This module is now *always* _socket.

24 years agoRevise the wrapper structure for the socket module:
Fred Drake [Wed, 16 Aug 2000 14:14:32 +0000 (14:14 +0000)]
Revise the wrapper structure for the socket module:

socket.py is used for all platforms, and it defines the additional
classes and alternate socket() function for Windows and BeOS systems.

The plat-*/socket.py files are no longer needed, since there is a
shared socket.py.

make_fqdn() is provided, but I decided to call it getfqdn() to be
consistent with the other names in the socket module.  Since it is
really a "get" operation and does not create a new name, this is
the right name to give it.

Move the docstring here from the _socket module.

24 years agoUpdated comment
Andrew M. Kuchling [Wed, 16 Aug 2000 12:27:23 +0000 (12:27 +0000)]
Updated comment

24 years agoDeleted now-unused include files
Andrew M. Kuchling [Wed, 16 Aug 2000 12:22:25 +0000 (12:22 +0000)]
Deleted now-unused include files

24 years agoMention setdefault() method for dicts
Andrew M. Kuchling [Wed, 16 Aug 2000 02:52:37 +0000 (02:52 +0000)]
Mention setdefault() method for dicts

24 years agoFrom Rene Liebscher:
Mark Hammond [Tue, 15 Aug 2000 22:33:59 +0000 (22:33 +0000)]
From Rene Liebscher:
This patch makes it possible to use gnu-win32 and lcc-win32
(http://www.cs.virginia.edu/~lcc-win32/) compilers to build
extension modules. It adds compiler specific sections to
PC/config.h .
It also extends the Borland compiler section. This has then two parts,
one for Win32 and the other one for the rest. The Win32 part
should be almost complete.

*** This patch is not intended to make it possible to compile
     Python with these compilers, it is intended to be able to
     use these compilers to build extension modules. ****

24 years agoApply SF patch #101151, by Peter S-K, which fixes smtplib's passing of the
Thomas Wouters [Tue, 15 Aug 2000 19:30:36 +0000 (19:30 +0000)]
Apply SF patch #101151, by Peter S-K, which fixes smtplib's passing of the
'helo' and 'ehlo' message, and exports the 'make_fqdn' function. This
function should be moved to socket.py, if that module ever gets a Python
wrapper.

24 years agoChris Herborth <chrish@pobox.com>:
Fred Drake [Tue, 15 Aug 2000 18:52:33 +0000 (18:52 +0000)]
Chris Herborth <chrish@pobox.com>:
Minor updates for BeOS R5.

Use of OSError in test.test_fork1 changed to TestSkipped, with corresponding
change in BeOS/README (by Fred).

This closes SourceForge patch #100978.

24 years agoFix error made in applying Thomas's patch.
Fred Drake [Tue, 15 Aug 2000 18:44:10 +0000 (18:44 +0000)]
Fix error made in applying Thomas's patch.

24 years agoThomas Wouters <thomas@xs4all.net>:
Fred Drake [Tue, 15 Aug 2000 17:54:49 +0000 (17:54 +0000)]
Thomas Wouters <thomas@xs4all.net>:
Update the grammar to reflect the most recent changes to list
comprehensions.

24 years agoFix markup bug that prevented formatting.
Fred Drake [Tue, 15 Aug 2000 17:47:09 +0000 (17:47 +0000)]
Fix markup bug that prevented formatting.

Adjusted some markup for consistency with the rest of the documentation
and creation of the proper index entries.

24 years agoFix new compiler warnings. Unused var in compile.c. Argsize mismatches
Tim Peters [Tue, 15 Aug 2000 16:41:26 +0000 (16:41 +0000)]
Fix new compiler warnings.  Unused var in compile.c.  Argsize mismatches
in binascii.c (only on platforms with signed chars -- although Py_CHARMASK
is documented as returning an int, it only does so on platforms with
signed chars).

24 years agomy_basename(): Removes the leading path components from a path name,
Fred Drake [Tue, 15 Aug 2000 16:20:36 +0000 (16:20 +0000)]
my_basename():  Removes the leading path components from a path name,
returning a pointer to the start of the file's "base" name;
similar to os.path.basename().

SyntaxError__str__():  Use my_basename() to keep the length of the
file name included in the exception message short.

24 years agoRemove the osdefs.h #include; it was not needed in the final version of
Fred Drake [Tue, 15 Aug 2000 16:13:37 +0000 (16:13 +0000)]
Remove the osdefs.h #include; it was not needed in the final version of
my last set of changes.

24 years agoUpdate test output to reflect change in SyntaxError formatting.
Fred Drake [Tue, 15 Aug 2000 15:51:18 +0000 (15:51 +0000)]
Update test output to reflect change in SyntaxError formatting.

This closes SourceForge bug #110628 (Jitterbug PR#278).

24 years agoWhen raising a SyntaxError, make a best-effort attempt to set the
Fred Drake [Tue, 15 Aug 2000 15:49:44 +0000 (15:49 +0000)]
When raising a SyntaxError, make a best-effort attempt to set the
filename and lineno attributes, but do not mask the SyntaxError if we
fail.

This is part of what is needed to close SoruceForge bug #110628
(Jitterbug PR#278).

Wrap a long line to fit in under 80 columns.

24 years agoWhen raising a SyntaxError, make a best-effort attempt to set the
Fred Drake [Tue, 15 Aug 2000 15:49:03 +0000 (15:49 +0000)]
When raising a SyntaxError, make a best-effort attempt to set the
filename and lineno attributes, but do not mask the SyntaxError if we
fail.

This is part of what is needed to close SoruceForge bug #110628
(Jitterbug PR#278).

24 years agoSyntaxError__str__(): Do more formatting of the exception here, rather
Fred Drake [Tue, 15 Aug 2000 15:46:16 +0000 (15:46 +0000)]
SyntaxError__str__():  Do more formatting of the exception here, rather
than depending on the site that raises the exception.  If the
filename and lineno attributes are set on the exception object,
use them to augment the message displayed.

This is part of what is needed to close SoruceForge bug #110628
(Jitterbug PR#278).

24 years agoBump version to 0.9.1.
Greg Ward [Tue, 15 Aug 2000 13:14:27 +0000 (13:14 +0000)]
Bump version to 0.9.1.

24 years agoFixed the move-RPM-files hack so it knows about the '--binary-only' and
Greg Ward [Tue, 15 Aug 2000 13:05:35 +0000 (13:05 +0000)]
Fixed the move-RPM-files hack so it knows about the '--binary-only' and
'--source-only' options.

24 years agoAdded support for the '--dist-dir' option, including a mildly nasty
Greg Ward [Tue, 15 Aug 2000 13:03:16 +0000 (13:03 +0000)]
Added support for the '--dist-dir' option, including a mildly nasty
hack to find the two created RPM files (source and binary) and
move them to the "dist dir" (default "dist").

24 years agoFix long-hidden inconsistency in internal interface: 'find_modules()' now
Greg Ward [Tue, 15 Aug 2000 13:01:25 +0000 (13:01 +0000)]
Fix long-hidden inconsistency in internal interface: 'find_modules()' now
represents packages as strings, not tuples.  This allowed a simplification
in 'get_package_dir()', too -- can now assume that 'package' is a string.

24 years agobinascii_unhexlify(): Better error message, courtesy effbot.
Barry Warsaw [Tue, 15 Aug 2000 06:59:58 +0000 (06:59 +0000)]
binascii_unhexlify(): Better error message, courtesy effbot.

24 years agotests for binascii.b2a_hex() and binascii.a2b_hex().
Barry Warsaw [Tue, 15 Aug 2000 06:08:31 +0000 (06:08 +0000)]
tests for binascii.b2a_hex() and binascii.a2b_hex().

24 years agoDescribe the b2a_hex() and a2b_hex() functions (a.k.a. hexlify() and
Barry Warsaw [Tue, 15 Aug 2000 06:08:00 +0000 (06:08 +0000)]
Describe the b2a_hex() and a2b_hex() functions (a.k.a. hexlify() and
unhexlify() respectively).

24 years agoAfter a brief conversation and code review with TP, adding two very
Barry Warsaw [Tue, 15 Aug 2000 06:07:13 +0000 (06:07 +0000)]
After a brief conversation and code review with TP, adding two very
commonly used functions to convert an arbitrary binary string into
a hexadecimal digit representation and back again.  These are often
(and often differently) implemented in Python.  Best to have one
common fast implementation.  Specifically,

binascii_hexlify(): a.k.a. b2a_hex() to return the hex representation
of binary data.

binascii_unhexlify(): a.k.a. a2b_hex() to do the inverse conversion
(hex digits to binary data).  The argument must have an even length,
and must contain only hex digits, otherwise a TypeError is raised.

24 years agoSHA_hexdigest(): A couple of small patches to this function, added
Barry Warsaw [Tue, 15 Aug 2000 06:03:35 +0000 (06:03 +0000)]
SHA_hexdigest(): A couple of small patches to this function, added
after a brief conversation with TP.  First, the return values of the
PyString_* function calls should be checked for errors.  Second,
bit-manipulations should be used instead of division for spliting the
byte up into its 4 bit digits.

24 years agoAdded a test of the md5.hexdigest() method. Funny enough, this test
Barry Warsaw [Tue, 15 Aug 2000 06:01:36 +0000 (06:01 +0000)]
Added a test of the md5.hexdigest() method.  Funny enough, this test
had yet-another Python implementation of a binary-data-to-hex-digit
encoder!

24 years agoAdded description of the md5.hexdigest() method.
Barry Warsaw [Tue, 15 Aug 2000 06:00:28 +0000 (06:00 +0000)]
Added description of the md5.hexdigest() method.

24 years agomd5_hexdigest(): After a brief conversation with TP, added hexdigest()
Barry Warsaw [Tue, 15 Aug 2000 05:59:44 +0000 (05:59 +0000)]
md5_hexdigest(): After a brief conversation with TP, added hexdigest()
to this module to mirror sha's hexdigest() method.

24 years agoCorrect the reference count information for the parameters of
Fred Drake [Tue, 15 Aug 2000 04:36:16 +0000 (04:36 +0000)]
Correct the reference count information for the parameters of
PyErr_Restore().

24 years agoGuido pointed out that all names in the sys module have no underscore,
Fred Drake [Tue, 15 Aug 2000 04:24:43 +0000 (04:24 +0000)]
Guido pointed out that all names in the sys module have no underscore,

24 years agoFix for http://sourceforge.net/bugs/?func=detailbug&bug_id=111866&group_id=5470.
Tim Peters [Tue, 15 Aug 2000 03:34:48 +0000 (03:34 +0000)]
Fix for http://sourceforge.net/bugs/?func=detailbug&bug_id=111866&group_id=5470.
This was a misleading bug -- the true "bug" was that hash(x) gave an error
return when x is an infinity.  Fixed that.  Added new Py_IS_INFINITY macro to
pyport.h.  Rearranged code to reduce growing duplication in hashing of float and
complex numbers, pushing Trent's earlier stab at that to a logical conclusion.
Fixed exceedingly rare bug where hashing of floats could return -1 even if there
wasn't an error (didn't waste time trying to construct a test case, it was simply
obvious from the code that it *could* happen).  Improved complex hash so that
hash(complex(x, y)) doesn't systematically equal hash(complex(y, x)) anymore.

24 years agoInitial revision
David Scherer [Tue, 15 Aug 2000 01:13:23 +0000 (01:13 +0000)]
Initial revision

24 years agoFix for bug #110670 - Win32 os.listdir raises confusing errors:
Mark Hammond [Tue, 15 Aug 2000 00:46:38 +0000 (00:46 +0000)]
Fix for bug #110670 - Win32 os.listdir raises confusing errors:
The existing win32_error() function now returns the new(ish) WindowsError, ensuring we get correct error messages.

24 years agoFix the parent of WindowsError - both the comments in this source file, and the previ...
Mark Hammond [Tue, 15 Aug 2000 00:37:32 +0000 (00:37 +0000)]
Fix the parent of WindowsError - both the comments in this source file, and the previous exceptions.py have WindowsError as a sub-class of OSError.

24 years agoRemove the test for abspath with an empty path - too hard to do in a cross-platform...
Mark Hammond [Mon, 14 Aug 2000 23:06:37 +0000 (23:06 +0000)]
Remove the test for abspath with an empty path - too hard to do in a cross-platform manner.

24 years agoThe attempt to protect against MS_WIN16 compilers that do not support long
Fred Drake [Mon, 14 Aug 2000 20:59:57 +0000 (20:59 +0000)]
The attempt to protect against MS_WIN16 compilers that do not support long
string literals has not been tested on an MS_WIN16 platform; the trailing
";" was inside the #ifndef MS_WIN16, which should cause an error (missing
semi-colon) when compiled with that symbol #defined.

24 years agoFor bug reporting, point to the SourceForge bug tracker instead of the
Fred Drake [Mon, 14 Aug 2000 16:35:05 +0000 (16:35 +0000)]
For bug reporting, point to the SourceForge bug tracker instead of the
old Jitterbug interface.

24 years agoDocument the byte_order value in the sys module.
Fred Drake [Mon, 14 Aug 2000 15:47:30 +0000 (15:47 +0000)]
Document the byte_order value in the sys module.

24 years agoAdd a byte_order value to the sys module. The value is "big" for
Fred Drake [Mon, 14 Aug 2000 15:47:03 +0000 (15:47 +0000)]
Add a byte_order value to the sys module.  The value is "big" for
big-endian machines and "little" for little-endian machines.

24 years agoDocument the range type, as suggested by Denis S. Otkidach
Fred Drake [Mon, 14 Aug 2000 15:37:59 +0000 (15:37 +0000)]
Document the range type, as suggested by Denis S. Otkidach
<den@analyt.chem.msu.ru>.

24 years agoFixed a couple of instances where a 0-length string was being
Marc-André Lemburg [Mon, 14 Aug 2000 11:29:19 +0000 (11:29 +0000)]
Fixed a couple of instances where a 0-length string was being
resized after creation. 0-length strings are usually shared
and _PyString_Resize() fails on these shared strings.

Fixes [ Bug #111667 ] unicode core dump.

24 years agoTest for fix to bug #110673: os.abspatth() now always returns os.getcwd() on Windows...
Mark Hammond [Mon, 14 Aug 2000 06:21:26 +0000 (06:21 +0000)]
Test for fix to bug #110673: os.abspatth() now always returns os.getcwd() on Windows, if an empty path is specified.  It previously did not if an empty path was delegated to win32api.GetFullPathName())

24 years agoFix for Bug #110673: os.abspatth() now always returns os.getcwd() on Windows, if...
Mark Hammond [Mon, 14 Aug 2000 06:20:32 +0000 (06:20 +0000)]
Fix for Bug #110673: os.abspatth() now always returns os.getcwd() on Windows, if an empty path is specified.  It previously did not if an empty path was delegated to win32api.GetFullPathName())

24 years agoCheck in the correct output - even though the module itself may not survive!
Mark Hammond [Mon, 14 Aug 2000 06:05:40 +0000 (06:05 +0000)]
Check in the correct output - even though the module itself may not survive!

24 years agoSupport for building the new w9xpopen.exe, which is used for reliable popen operation...
Mark Hammond [Mon, 14 Aug 2000 05:07:05 +0000 (05:07 +0000)]
Support for building the new w9xpopen.exe, which is used for reliable popen operation on Windows 9x.

24 years agoPatch #101032, from David Bolen:
Mark Hammond [Mon, 14 Aug 2000 05:04:28 +0000 (05:04 +0000)]
Patch #101032, from David Bolen:

Ensure the "proxied" command's return code bubbles back up.

24 years agoPatch #101032, from David Bolen:
Mark Hammond [Mon, 14 Aug 2000 04:47:33 +0000 (04:47 +0000)]
Patch #101032, from David Bolen:
This is an enhancement to a prior patch (100941) ...
[T]his patch removes the risk of deadlock waiting for the child previously present in certain cases. It adds tracking of all file handles returned from an os.popen* call and only waits for the child process, returning the exit code, on the closure of the final file handle to that child.

24 years agoIn the section on the "Very High Level Layer", address concerns brought up
Fred Drake [Mon, 14 Aug 2000 02:50:21 +0000 (02:50 +0000)]
In the section on the "Very High Level Layer", address concerns brought up
by Edward K. Ream <edream@users.sourceforge.net> about FILE* values and
incompatible C libraries in dynamically linked extensions.  It is not clear
(to me) how realistic the issue is, but it is better documented than not.

This closes SourceForge bug #111520.

24 years agoRevert this checkin:
Trent Mick [Sun, 13 Aug 2000 22:59:26 +0000 (22:59 +0000)]
Revert this checkin:
http://www.python.org/pipermail/python-checkins/2000-August/007072.html
and make PCbuild/*.dsp PCbuild/*.dsw binary again.

24 years agoCheck for overflow in list object insertion and raise OverflowError.
Trent Mick [Sun, 13 Aug 2000 22:47:45 +0000 (22:47 +0000)]
Check for overflow in list object insertion and raise OverflowError.
see: http://www.python.org/pipermail/python-dev/2000-August/014971.html

24 years agoThe list comp patch checked for the second child node of the 'listmaker'
Thomas Wouters [Sun, 13 Aug 2000 17:05:17 +0000 (17:05 +0000)]
The list comp patch checked for the second child node of the 'listmaker'
node, without checking if the node actually had more than one child. It can
have only one node, though: '[' test ']'. This fixes it.

24 years agoFix a typo in the PyMem_Resize macro, found by Andrew Kuchling
Vladimir Marangozov [Sun, 13 Aug 2000 11:59:08 +0000 (11:59 +0000)]
Fix a typo in the PyMem_Resize macro, found by Andrew Kuchling

24 years agoOverhauld 'check_config_h()': now returns a (status, details) tuple,
Greg Ward [Sun, 13 Aug 2000 01:18:55 +0000 (01:18 +0000)]
Overhauld 'check_config_h()': now returns a (status, details) tuple,
and is much better documented to boot.

24 years agoAdded a whinging comment about the ugliness of constructing the BCPP
Greg Ward [Sun, 13 Aug 2000 00:54:39 +0000 (00:54 +0000)]
Added a whinging comment about the ugliness of constructing the BCPP
argument list.

24 years agoRene Liebscher:
Greg Ward [Sun, 13 Aug 2000 00:43:56 +0000 (00:43 +0000)]
Rene Liebscher:
  * use self.debug_print() for debug messages
  * uses now copy.copy() to copy lists
  * added 'shared_lib_extension=".dll"', ... , this is necessary if you
    want use the compiler class outside of the standard distutils build
    process.
  * changed result type of check_config_h() from int to string