Tim Peters [Sat, 10 Feb 2001 08:00:53 +0000 (08:00 +0000)]
Moved SequenceMatcher from ndiff into new std library module difflib.py.
Guido told me to do this <wink>.
Greatly expanded docstrings, and fleshed out with examples.
New std test.
Added new get_close_matches() function for ESR.
Needs docs, but LaTeXification of the module docstring is all it needs.
\CVS: ----------------------------------------------------------------------
Jeremy Hylton [Fri, 9 Feb 2001 22:22:18 +0000 (22:22 +0000)]
Relax the rules for using 'from ... import *' and exec in the presence
of nested functions. Either is allowed in a function if it contains
no defs or lambdas or the defs and lambdas it contains have no free
variables. If a function is itself nested and has free variables,
either is illegal.
Revise the symtable to use a PySymtableEntryObject, which holds all
the revelent information for a scope, rather than using a bunch of
st_cur_XXX pointers in the symtable struct. The changes simplify the
internal management of the current symtable scope and of the stack.
Added new C source file: Python/symtable.c. (Does the Windows build
process need to be updated?)
As part of these changes, the initial _symtable module interface
introduced in 2.1a2 is replaced. A dictionary of
PySymtableEntryObjects are returned.
This modified version of a patch by Thomas Heller allows __import__
hooks to take over the Python import machinery at a very early stage
in the Python startup phase.
If there are still places in the Python interpreter which need to
bypass the __import__ hook, these places must now use
PyImport_ImportModuleEx() instead. So far no other places than in
the import mechanism itself have been identified.
Jack Jansen [Fri, 9 Feb 2001 15:58:34 +0000 (15:58 +0000)]
Replaced touched(), which no longer seems to work under os9, with something using an appleevent to the finder. More expensive, but at least it works:-)
Eric S. Raymond [Fri, 9 Feb 2001 12:10:26 +0000 (12:10 +0000)]
Eliminate use of string.whitespace and a string import with it.
Some of the characters (form feed, vertical tab) are not
legal continuation characters anyway, so this was wrong as
well as annoying.
Tim Peters [Fri, 9 Feb 2001 07:02:22 +0000 (07:02 +0000)]
SF bug #131225: sys.winver is still '2.0' in python 2.1a2.
SF patch #103683: Alternative dll version resources.
Changes similar to the patch. MarkH should review.
File version and Product version text strings now 2.1a2.
64-bit file and product version numbers are now
PY_MAJOR_VERSION, PY_MINOR_VERSION, messy, PYTHON_API_VERSION
where
messy = PY_MICRO_VERSION*1000 + PY_RELEASE_LEVEL*10 + PY_RELEASE_SERIAL
Updated company name to "Digital Creations 2".
Copyright now lists Guido; "C in a circle" symbol used instead of (C).
Comments added so this is less likely to get flubbed again, and
#if/#error guys added to trigger if the version number manipulations
above overflow.
Fred Drake [Thu, 8 Feb 2001 15:39:08 +0000 (15:39 +0000)]
set_error(): Handle construction of pyexpat.error exceptions. They
now carry a 'code' attribute that gives the Expat error
number.
Added support for additional handlers for Expat 1.95.*, including
XmlDeclHandler, EntityDeclHandler, ElementDeclHandler, and
AttlistDeclHandler. Associated constants are in the 'model'
sub-object.
Added two new attributes to the parser object: ordered_attributes and
specified_attributes. These are used to control how attributes are
reported and which attributes are reported.
Martin v. Löwis [Tue, 6 Feb 2001 01:16:06 +0000 (01:16 +0000)]
Do not allow empty qualifiedName in createDocument.
Rearrange pulldom to create documents with root element.
Provide clear methods so that the ContentHandler releases its hold on the
document.
A couple of changes to make this more conformant. MvL and Uche agree.
This will make it incompatible with the version found in Python 2.0.
Does this need to be done to PyXML too?
Changes that might break existing code are marked with (!) below.
- Formatting nit: no spaces inside parentheses: foo( a ) -> foo(a).
- Break long lines.
- (!) Fix getAttribute() and getAttributeNS() to return "" instead of
raising KeyError when the attribute is not found.
- (!) Fix getAttributeNodeNS() to return None instead of raising
KeyError. (Curiously, getAttributeNode() already did this.)
- Added hasAttributes(), which returns true iff the node has any
attributes. )This is DOM level 3.)
- (!) In createDocument(), if the qualified name is not empty,
actually create and insert the first element with that name (this
will become doc.documentElement). MvL believes that it should be an
error to specify an empty qualified name; I'm not going there today,
since it would require making a matching change to pulldom. Maybe
MvL will do this.
- In Document.writexml(), insert an xml declaration at the top. (This
doesn't include the encoding since there's no way to specify the
encoding. If that's preferred, all writexml() methods should be
fixed to support an optional encoding argument that they pass to
each other -- and they should use it to encode all text they write,
too. Later.)
Don't get fooled by an empty prefix with a valid namespaceURI -- in
this case, the code used to generate invalid tags and attribute names
with a leading colon, e.g. <:tag> or <tag :attr="foo">.