Martin v. Löwis [Tue, 13 Mar 2001 10:50:13 +0000 (10:50 +0000)]
Patch #407965: Improve Level 2 conformance of minidom
- addition of a DocumentFragment implementation and createDocumentFragment method
- proper setting of ownerDocument for all nodes
- setting of namespaceURI to None in Element as a class attribute
- addition of setAttributeNodeNS and removeAttributeNodeNS as aliases
for setAttributeNode and removeAttributeNode
- support for inheriting from DOMImplementation to extend it with
additional features (to override the Document class)
in pulldom:
- support for nodes (comment and PI) that occur before he document element;
that became necessary as pulldom now delays creation of the document
until it has the document element.
Jeremy Hylton [Tue, 13 Mar 2001 01:58:22 +0000 (01:58 +0000)]
Variety of small INC/DECREF patches that fix reported memory leaks
with free variables. Thanks to Martin v. Loewis for finding two of
the problems. This fixes SF buf 405583.
There is also a C API change: PyFrame_New() is reverting to its
pre-2.1 signature. The change introduced by nested scopes was a
mistake. XXX Is this okay between beta releases?
cell_clear(), the GC helper, must decref its reference to break
cycles.
frame_dealloc() must dealloc all cell vars and free vars in addition
to locals.
eval_code2() setup code must INCREF cells it copies out of the
closure.
The STORE_DEREF opcode implementation must DECREF the object it passes
to PyCell_Set().
Tim Peters [Sun, 11 Mar 2001 18:36:13 +0000 (18:36 +0000)]
Identifiers matching _[A-Z_]\w* are reserved for C implementations.
May or may not be related to bug 407680 (obmalloc.c - looks like it's
corrupted). This repairs the illegal vrbl names, but leaves a pile of
illegal macro names (_THIS_xxx, _SYSTEM_xxx, _SET_HOOKS, _FETCH_HOOKS).
Tim Peters [Sun, 11 Mar 2001 07:34:18 +0000 (07:34 +0000)]
SF bug Win32: pydoc command isn't executable.
As the bug report notes, the Windows installer creates a useless pydoc
file in the base directory. Changed the installer to rename it pydoc.pyw
instead.
Tim Peters [Sun, 11 Mar 2001 04:30:35 +0000 (04:30 +0000)]
Magic strings in both Python and the Wise installer define the registry key
to be used in Windows. They got out of synch. Repaired that, and added
comments to each one pointing at the other.
Fred Drake [Sat, 10 Mar 2001 02:18:47 +0000 (02:18 +0000)]
Import the nested_scopes feature twice, to exercise the patch introduced
to avoid segfaults when more than one feature is named in the future
statement.
Fred Drake [Thu, 8 Mar 2001 22:17:54 +0000 (22:17 +0000)]
Add more font flavors for <var> elements; the previous incarnation caused
some NT/IE5.5 users to see only boxes for these characters. This might be
specific to NT Service Pack 6.
Jack Jansen [Tue, 6 Mar 2001 22:51:00 +0000 (22:51 +0000)]
PythonInterpreterClassic and PythonInterpreterCarbon now have filetype "Atmp". The ConfigurePythonXXX scripts will set this to APPL when copying either of those to PythonInterpreter.
Jack Jansen [Tue, 6 Mar 2001 22:46:25 +0000 (22:46 +0000)]
macostools.copy() now has an optional forcetype flag, which sets the filetype. This is used by various tools to set the type to APPL when copying interpreter template files (Atmp filetype).
Fred Drake [Tue, 6 Mar 2001 07:28:20 +0000 (07:28 +0000)]
Turn "Python Documentation Central" into just "Documentation Central"; some
Web browsers tried to wrap lines in bad ways, so we avoid the problem by
putting fewer words on the page.
Fred Drake [Tue, 6 Mar 2001 07:19:34 +0000 (07:19 +0000)]
Correct typos in Ping's email address.
Remove premature use of negative indexes in string operation examples;
negative indexes have not been explained at that point, and the use of
negative indexes are not necessary for the examples.
Fred Drake [Tue, 6 Mar 2001 05:52:16 +0000 (05:52 +0000)]
Move all knowledge that $(MAINOBJ) is built in the Modules/ directory
into Makefile.pre.in; the configure script will only determine the basename
of the file.
This fixes installation of a Python built using C++, reported by Greg
Wilson.
Fred Drake [Sat, 3 Mar 2001 19:41:55 +0000 (19:41 +0000)]
Be a bit more strict in setting up the export of the C API for this
module; do not attempt to insert the API object into the module dict
if there was an error creating it.
Fred Drake [Sat, 3 Mar 2001 18:08:52 +0000 (18:08 +0000)]
Wrap several more of the constants in #ifdef/#endif for FreeBSD; at least
some fairly recent versions have an anaemic selection of terminal-control
symbols.
Barry Warsaw [Sat, 3 Mar 2001 04:14:21 +0000 (04:14 +0000)]
Added `memtest' target which excludes the quicktest modules plus
test_dl, test___all__, test_fork1, and test_longexp. All these either
take way too long with Insure or crash it.
Fred Drake [Fri, 2 Mar 2001 21:05:58 +0000 (21:05 +0000)]
Revise the scripts I use to update the documentation on the SourceForge
site. These now seem (slightly) more reliable, and easier to work with
since update-docs.sh no longer needs to be installed ahead of time on
my account at SF.
Fred Drake [Fri, 2 Mar 2001 18:15:11 +0000 (18:15 +0000)]
There is no longer a -X option to the interpreter, so remove the comments
on how PyErr_NewException() behaves in that case.
Clarify why an owned reference is kept in an extension module's
variable that refers to the result of PyErr_NewException(); one reader
thought that was a leak.
Jack Jansen [Fri, 2 Mar 2001 16:32:03 +0000 (16:32 +0000)]
Enable FSpResourceFileAlreadyOpen only for Carbon. This is not strictly correct: OS9 InterfaceLib has it, but enabling it will make Python refuse to run on 8.6 and earlier.
Fred Drake [Fri, 2 Mar 2001 16:26:45 +0000 (16:26 +0000)]
Job.build_html(): When formatting HTML into more than one HTML page, and
not doing the \label{foo} --> foo.html transformation (--numeric was
specified on the command line), still look to see if there is an
"About this document..." node and copy that to "about.html", since the
page footers use that as the target.
Fred Drake [Fri, 2 Mar 2001 07:28:03 +0000 (07:28 +0000)]
When not copying a file because the output is up to date, make the message
slightly more brief, and more like the message that an extension will not
be built because the built copy is up to date.
UNIX style fork/execve/wait are not fully compatible with thread
support on BeOS. For Python, that means neither fork() from import
nor import from a fork work reliably. os._execvpe() does the latter,
importing tempfile to set up a tantalizing target for hackers. This
patch replaces both the tempfile name generation and the exec that
uses it, in case we're on BeOS. Need this for
setup:distutils:execvp(); symptoms are random crashes and internal
BeOS error messages about th name, in case we're on BeOS. It's an
issue because setup.py + distutils calls os.execvp(); symptoms are
random crashes during setup.py, and internal BeOS error messages
about thread IDs.