5 (editors: check NEWS.help for information about editing NEWS using ReST.)
7 What's New in Python 2.5 beta 1?
8 =================================
10 *Release date: XX-JUN-2006*
15 - Bug #992017: A classic class that defined a __coerce__() method that returned
16 its arguments swapped would infinitely recurse and segfault the interpreter.
18 - Fix the socket tests so they can be run concurrently.
20 - Removed 5 integers from C frame objects (PyFrameObject).
21 f_nlocals, f_ncells, f_nfreevars, f_stack_size, f_restricted.
23 - Bug #532646: object.__call__() will continue looking for the __call__
24 attribute on objects until one without one is found. This leads to recursion
25 when you take a class and set its __call__ attribute to an instance of the
26 class. Originally fixed for classic classes, but this fix is for new-style.
27 Removes the infinite_rec_3 crasher.
29 - The string and unicode methods startswith() and endswith() now accept
30 a tuple of prefixes/suffixes to look for. Implements RFE #1491485.
32 - Buffer objects, at the C level, never used the char buffer
33 implementation even when the char buffer for the wrapped object was
34 explicitly requested (originally returned the read or write buffer).
35 Now a TypeError is raised if the char buffer is not present but is
38 - Patch #1346214: Statements like "if 0: suite" are now again optimized
39 away like they were in Python 2.4.
41 - Builtin exceptions are now full-blown new-style classes instead of
42 instances pretending to be classes, which speeds up exception handling
43 by about 80% in comparison to 2.5a2.
45 - Patch #1494554: Update unicodedata.numeric and unicode.isnumeric to
48 - Patch #921466: sys.path_importer_cache is now used to cache valid and
49 invalid file paths for the built-in import machinery which leads to
50 fewer open calls on startup.
52 - Patch #1442927: ``long(str, base)`` is now up to 6x faster for non-power-
53 of-2 bases. The largest speedup is for inputs with about 1000 decimal
54 digits. Conversion from non-power-of-2 bases remains quadratic-time in
55 the number of input digits (it was and remains linear-time for bases
58 - Bug #1334662: ``int(string, base)`` could deliver a wrong answer
59 when ``base`` was not 2, 4, 8, 10, 16 or 32, and ``string`` represented
60 an integer close to ``sys.maxint``. This was repaired by patch
61 #1335972, which also gives a nice speedup.
63 - Patch #1337051: reduced size of frame objects.
65 - PyErr_NewException now accepts a tuple of base classes as its
68 - Patch #876206: function call speedup by retaining allocated frame
71 - Bug #1462152: file() now checks more thoroughly for invalid mode
72 strings and removes a possible "U" before passing the mode to the
75 - Patch #1488312, Fix memory alignment problem on SPARC in unicode
77 - Bug #1487966: Fix SystemError with conditional expression in assignment
79 - WindowsError now has two error code attributes: errno, which carries
80 the error values from errno.h, and winerror, which carries the error
81 values from winerror.h. Previous versions put the winerror.h values
82 (from GetLastError()) into the errno attribute.
84 - Patch #1475845: Raise IndentationError for unexpected indent.
86 - Patch #1479181: split open() and file() from being aliases for each other.
88 - Patch #1497053 & bug #1275608: Exceptions occurring in ``__eq__()``
89 methods were always silently ignored by dictionaries when comparing keys.
90 They are now passed through (except when using the C API function
91 ``PyDict_GetItem()``, whose semantics did not change).
93 - Bug #1456209: In some obscure cases it was possible for a class with a
94 custom ``__eq__()`` method to confuse dict internals when class instances
95 were used as a dict's keys and the ``__eq__()`` method mutated the dict.
96 No, you don't have any code that did this ;-)
101 - Change binascii.hexlify to accept a read-only buffer instead of only a char
102 buffer and actually follow its documentation.
104 - Fixed a potentially invalid memory access of CJKCodecs' shift-jis decoder.
106 - Patch #1478788 (modified version): The functional extension module has
107 been renamed to _functools and a functools Python wrapper module added.
108 This provides a home for additional function related utilities that are
109 not specifically about functional programming. See PEP 309.
111 - Patch #1493701: performance enhancements for struct module.
113 - Patch #1490224: time.altzone is now set correctly on Cygwin.
115 - Patch #1435422: zlib's compress and decompress objects now have a
118 - Patch #1454481: thread stack size is now tunable at runtime for thread
119 enabled builds on Windows and systems with Posix threads support.
121 - On Win32, os.listdir now supports arbitrarily-long Unicode path names
122 (up to the system limit of 32K characters).
124 - Use Win32 API to implement os.{access,chdir,chmod,mkdir,remove,rename,rmdir,utime}.
125 As a result, these functions now raise WindowsError instead of OSError.
127 - ``time.clock()`` on Win64 should use the high-performance Windows
128 ``QueryPerformanceCounter()`` now (as was already the case on 32-bit
131 - Calling Tk_Init twice is refused if the first call failed as that
134 - bsddb: added the DB_ARCH_REMOVE flag and fixed db.DBEnv.log_archive() to
135 accept it without potentially using an uninitialized pointer.
137 - bsddb: added support for the DBEnv.log_stat() and DBEnv.lsn_reset() methods
138 assuming BerkeleyDB >= 4.0 and 4.4 respectively. [pybsddb project SF
139 patch numbers 1494885 and 1494902]
141 - bsddb: added an interface for the BerkeleyDB >= 4.3 DBSequence class.
142 [pybsddb project SF patch number 1466734]
144 - bsddb: fix DBCursor.pget() bug with keyword argument names when no data
145 parameter is supplied. [SF pybsddb bug #1477863]
147 - bsddb: the __len__ method of a DB object has been fixed to return correct
148 results. It could previously incorrectly return 0 in some cases.
149 Fixes SF bug 1493322 (pybsddb bug 1184012).
151 - bsddb: the bsddb.dbtables Modify method now raises the proper error and
152 aborts the db transaction safely when a modifier callback fails.
153 Fixes SF python patch/bug #1408584.
155 - bsddb: multithreaded DB access using the simple bsddb module interface
156 now works reliably. It has been updated to use automatic BerkeleyDB
157 deadlock detection and the bsddb.dbutils.DeadlockWrap wrapper to retry
158 database calls that would previously deadlock. [SF python bug #775414]
160 - Patch #1446489: add support for the ZIP64 extensions to zipfile.
166 - Bug #1117556: SimpleHTTPServer now tries to find and use the system's
167 mime.types file for determining MIME types.
169 - Bug #1339007: Shelf objects now don't raise an exception in their
170 __del__ method when initialization failed.
172 - Patch #1455898: The MBCS codec now supports the incremental mode for
173 double-byte encodings.
175 - ``difflib``'s ``SequenceMatcher.get_matching_blocks()`` was changed to
176 guarantee that adjacent triples in the return list always describe
177 non-adjacent blocks. Previously, a pair of matching blocks could end
178 up being described by multiple adjacent triples that formed a partition
179 of the matching pair.
181 - Bug #1498146: fix optparse to handle Unicode strings in option help,
182 description, and epilog.
184 - Bug #1366250: minor optparse documentation error.
186 - Bug #1361643: fix textwrap.dedent() so it handles tabs appropriately;
189 - The wsgiref package has been added to the standard library.
191 - The functions update_wrapper() and wraps() have been added to the functools
192 module. These make it easier to copy relevant metadata from the original
193 function when writing wrapper functions.
195 - The optional ``isprivate`` argument to ``doctest.testmod()``, and the
196 ``doctest.is_private()`` function, both deprecated in 2.4, were removed.
198 - Patch #1359618: Speed up charmap encoder by using a trie structure
201 - The functions in the ``pprint`` module now sort dictionaries by key
202 before computing the display. Before 2.5, ``pprint`` sorted a dictionary
203 if and only if its display required more than one line, although that
204 wasn't documented. The new behavior increases predictability; e.g.,
205 using ``pprint.pprint(a_dict)`` in a doctest is now reliable.
207 - Patch #1497027: try HTTP digest auth before basic auth in urllib2
208 (thanks for J. J. Lee).
210 - Patch #1496206: improve urllib2 handling of passwords with respect to
211 default HTTP and HTTPS ports.
213 - Patch #1080727: add "encoding" parameter to doctest.DocFileSuite.
215 - Patch #1281707: speed up gzip.readline.
217 - Patch #1180296: Two new functions were added to the locale module:
218 format_string() to get the effect of "format % items" but locale-aware,
219 and currency() to format a monetary number with currency sign.
221 - Patch #1486962: Several bugs in the turtle Tk demo module were fixed
222 and several features added, such as speed and geometry control.
224 - Patch #1488881: add support for external file objects in bz2 compressed
227 - Patch #721464: pdb.Pdb instances can now be given explicit stdin and
228 stdout arguments, making it possible to redirect input and output
229 for remote debugging.
231 - Patch #1484695: Update the tarfile module to version 0.8. This fixes
232 a couple of issues, notably handling of long file names using the
233 GNU LONGNAME extension.
235 - Patch #1478292. ``doctest.register_optionflag(name)`` shouldn't create a
236 new flag when ``name`` is already the name of an option flag.
238 - Bug #1385040: don't allow "def foo(a=1, b): pass" in the compiler
241 - Patch #1472854: make the rlcompleter.Completer class usable on non-
244 - Patch #1470846: fix urllib2 ProxyBasicAuthHandler.
246 - Bug #1472827: correctly escape newlines and tabs in attribute values in
247 the saxutils.XMLGenerator class.
253 - Bug #1502728: Correctly link against librt library on HP-UX.
255 - OpenBSD 3.9 is supported now.
257 - Patch #1492356: Port to Windows CE.
259 - Bug/Patch #1481770: Use .so extension for shared libraries on HP-UX for ia64.
261 - Patch #1471883: Add --enable-universalsdk.
277 What's New in Python 2.5 alpha 2?
278 =================================
280 *Release date: 27-APR-2006*
285 - Bug #1465834: 'bdist_wininst preinstall script support' was fixed
286 by converting these apis from macros into exported functions again:
288 PyParser_SimpleParseFile PyParser_SimpleParseString PyRun_AnyFile
289 PyRun_AnyFileEx PyRun_AnyFileFlags PyRun_File PyRun_FileEx
290 PyRun_FileFlags PyRun_InteractiveLoop PyRun_InteractiveOne
291 PyRun_SimpleFile PyRun_SimpleFileEx PyRun_SimpleString
292 PyRun_String Py_CompileString
294 - Under COUNT_ALLOCS, types are not necessarily immortal anymore.
296 - All uses of PyStructSequence_InitType have been changed to initialize
297 the type objects only once, even if the interpreter is initialized
300 - Bug #1454485, array.array('u') could crash the interpreter. This was
301 due to PyArgs_ParseTuple(args, 'u#', ...) trying to convert buffers (strings)
302 to unicode when it didn't make sense. 'u#' now requires a unicode string.
304 - Py_UNICODE is unsigned. It was always documented as unsigned, but
305 due to a bug had a signed value in previous versions.
307 - Patch #837242: ``id()`` of any Python object always gives a positive
308 number now, which might be a long integer. ``PyLong_FromVoidPtr`` and
309 ``PyLong_AsVoidPtr`` have been changed accordingly. Note that it has
310 never been correct to implement a ``__hash()__`` method that returns the
311 ``id()`` of an object:
314 return id(self) # WRONG
316 because a hash result must be a (short) Python int but it was always
317 possible for ``id()`` to return a Python long. However, because ``id()``
318 could return negative values before, on a 32-bit box an ``id()`` result
319 was always usable as a hash value before this patch. That's no longer
322 - Python on OS X 10.3 and above now uses dlopen() (via dynload_shlib.c)
323 to load extension modules and now provides the dl module. As a result,
324 sys.setdlopenflags() now works correctly on these systems. (SF patch
327 - Patch #1463867: enhanced garbage collection to allow cleanup of cycles
328 involving generators that have paused outside of any ``try`` or ``with``
329 blocks. (In 2.5a1, a paused generator that was part of a reference
330 cycle could not be garbage collected, regardless of whether it was
331 paused in a ``try`` or ``with`` block.)
336 - Patch #1191065: Fix preprocessor problems on systems where recvfrom
339 - Bug #1467952: os.listdir() now correctly raises an error if readdir()
340 fails with an error condition.
342 - Fixed bsddb.db.DBError derived exceptions so they can be unpickled.
344 - Bug #1117761: bsddb.*open() no longer raises an exception when using
345 the cachesize parameter.
347 - Bug #1149413: bsddb.*open() no longer raises an exception when using
348 a temporary db (file=None) with the 'n' flag to truncate on open.
350 - Bug #1332852: bsddb module minimum BerkeleyDB version raised to 3.3
351 as older versions cause excessive test failures.
353 - Patch #1062014: AF_UNIX sockets under Linux have a special
354 abstract namespace that is now fully supported.
359 - Patch #1475231: ``doctest`` has a new ``SKIP`` option, which causes
360 a doctest to be skipped (the code is not run, and the expected output
361 or exception is ignored).
363 - Fixed contextlib.nested to cope with exceptions being raised and
364 caught inside exit handlers.
366 - Updated optparse module to Optik 1.5.1 (allow numeric constants in
367 hex, octal, or binary; add ``append_const`` action; keep going if
368 gettext cannot be imported; added ``OptionParser.destroy()`` method;
369 added ``epilog`` for better help generation).
371 - Bug #1473760: ``tempfile.TemporaryFile()`` could hang on Windows, when
372 called from a thread spawned as a side effect of importing a module.
374 - The pydoc module now supports documenting packages contained in
377 - The pkgutil module now has several new utility functions, such
378 as ``walk_packages()`` to support working with packages that are either
379 in the filesystem or zip files.
381 - The mailbox module can now modify and delete messages from
382 mailboxes, in addition to simply reading them. Thanks to Gregory
383 K. Johnson for writing the code, and to the 2005 Google Summer of
384 Code for funding his work.
386 - The ``__del__`` method of class ``local`` in module ``_threading_local``
387 returned before accomplishing any of its intended cleanup.
389 - Patch #790710: Add breakpoint command lists in pdb.
391 - Patch #1063914: Add Tkinter.Misc.clipboard_get().
393 - Patch #1191700: Adjust column alignment in bdb breakpoint lists.
395 - SimpleXMLRPCServer relied on the fcntl module, which is unavailable on
396 Windows. Bug #1469163.
398 - The warnings, linecache, inspect, traceback, site, and doctest modules
399 were updated to work correctly with modules imported from zipfiles or
400 via other PEP 302 __loader__ objects.
402 - Patch #1467770: Reduce usage of subprocess._active to processes which
403 the application hasn't waited on.
405 - Patch #1462222: Fix Tix.Grid.
407 - Fix exception when doing glob.glob('anything*/')
409 - The pstats.Stats class accepts an optional stream keyword argument to
410 direct output to an alternate file-like object.
415 - The Makefile now has a reindent target, which runs reindent.py on
418 - Patch #1470875: Building Python with MS Free Compiler
420 - Patch #1161914: Add a python-config script.
422 - Patch #1324762:Remove ccpython.cc; replace --with-cxx with
423 --with-cxx-main. Link with C++ compiler only if --with-cxx-main was
424 specified. (Can be overridden by explicitly setting LINKCC.) Decouple
425 CXX from --with-cxx-main, see description in README.
427 - Patch #1429775: Link extension modules with the shared libpython.
429 - Fixed a libffi build problem on MIPS systems.
431 - ``PyString_FromFormat``, ``PyErr_Format``, and ``PyString_FromFormatV``
432 now accept formats "%u" for unsigned ints, "%lu" for unsigned longs,
433 and "%zu" for unsigned integers of type ``size_t``.
438 - test_contextlib now checks contextlib.nested can cope with exceptions
439 being raised and caught inside exit handlers.
441 - test_cmd_line now checks operation of the -m and -c command switches
443 - The test_contextlib test in 2.5a1 wasn't actually run unless you ran
444 it separately and by hand. It also wasn't cleaning up its changes to
445 the current Decimal context.
447 - regrtest.py now has a -M option to run tests that test the new limits of
448 containers, on 64-bit architectures. Running these tests is only sensible
449 on 64-bit machines with more than two gigabytes of memory. The argument
450 passed is the maximum amount of memory for the tests to use.
455 - Added the Python benchmark suite pybench to the Tools/ directory;
456 contributed by Marc-Andre Lemburg.
461 - Patch #1473132: Improve docs for ``tp_clear`` and ``tp_traverse``.
463 - PEP 343: Added Context Types section to the library reference
464 and attempted to bring other PEP 343 related documentation into
465 line with the implementation and/or python-dev discussions.
467 - Bug #1337990: clarified that ``doctest`` does not support examples
468 requiring both expected output and an exception.
471 What's New in Python 2.5 alpha 1?
472 =================================
474 *Release date: 05-APR-2006*
479 - PEP 338: -m command line switch now delegates to runpy.run_module
480 allowing it to support modules in packages and zipfiles
482 - On Windows, .DLL is not an accepted file name extension for
483 extension modules anymore; extensions are only found if they
486 - Bug #1421664: sys.stderr.encoding is now set to the same value as
489 - __import__ accepts keyword arguments.
491 - Patch #1460496: round() now accepts keyword arguments.
493 - Fixed bug #1459029 - unicode reprs were double-escaped.
495 - Patch #1396919: The system scope threads are reenabled on FreeBSD
496 5.4 and later versions.
498 - Bug #1115379: Compiling a Unicode string with an encoding declaration
499 now gives a SyntaxError.
501 - Previously, Python code had no easy way to access the contents of a
502 cell object. Now, a ``cell_contents`` attribute has been added
503 (closes patch #1170323).
505 - Patch #1123430: Python's small-object allocator now returns an arena to
506 the system ``free()`` when all memory within an arena becomes unused
507 again. Prior to Python 2.5, arenas (256KB chunks of memory) were never
508 freed. Some applications will see a drop in virtual memory size now,
509 especially long-running applications that, from time to time, temporarily
510 use a large number of small objects. Note that when Python returns an
511 arena to the platform C's ``free()``, there's no guarantee that the
512 platform C library will in turn return that memory to the operating system.
513 The effect of the patch is to stop making that impossible, and in tests it
514 appears to be effective at least on Microsoft C and gcc-based systems.
515 Thanks to Evan Jones for hard work and patience.
517 - Patch #1434038: property() now uses the getter's docstring if there is
518 no "doc" argument given. This makes it possible to legitimately use
519 property() as a decorator to produce a read-only property.
521 - PEP 357, patch 1436368: add an __index__ method to int/long and a matching
522 nb_index slot to the PyNumberMethods struct. The slot is consulted instead
523 of requiring an int or long in slicing and a few other contexts, enabling
524 other objects (e.g. Numeric Python's integers) to be used as slice indices.
526 - Fixed various bugs reported by Coverity's Prevent tool.
528 - PEP 352, patch #1104669: Make exceptions new-style objects. Introduced the
529 new exception base class, BaseException, which has a new message attribute.
530 KeyboardInterrupt and SystemExit to directly inherit from BaseException now.
531 Raising a string exception now raises a DeprecationWarning.
533 - Patch #1438387, PEP 328: relative and absolute imports. Imports can now be
534 explicitly relative, using 'from .module import name' to mean 'from the same
535 package as this module is in. Imports without dots still default to the
536 old relative-then-absolute, unless 'from __future__ import
537 absolute_import' is used.
539 - Properly check if 'warnings' raises an exception (usually when a filter set
540 to "error" is triggered) when raising a warning for raising string
543 - CO_GENERATOR_ALLOWED is no longer defined. This behavior is the default.
544 The name was removed from Include/code.h.
546 - PEP 308: conditional expressions were added: (x if cond else y).
549 - The copy module now "copies" function objects (as atomic objects).
550 - dict.__getitem__ now looks for a __missing__ hook before raising
553 - PEP 343: with statement implemented. Needs ``from __future__ import
554 with_statement``. Use of 'with' as a variable will generate a warning.
555 Use of 'as' as a variable will also generate a warning (unless it's
556 part of an import statement).
557 The following objects have __context__ methods:
558 - The built-in file type.
559 - The thread.LockType type.
560 - The following types defined by the threading module:
561 Lock, RLock, Condition, Semaphore, BoundedSemaphore.
562 - The decimal.Context class.
564 - Fix the encodings package codec search function to only search
565 inside its own package. Fixes problem reported in patch #1433198.
567 Note: Codec packages should implement and register their own
568 codec search function. PEP 100 has the details.
570 - PEP 353: Using ``Py_ssize_t`` as the index type.
572 - ``PYMALLOC_DEBUG`` builds now add ``4*sizeof(size_t)`` bytes of debugging
573 info to each allocated block, since the ``Py_ssize_t`` changes (PEP 353)
574 now allow Python to make use of memory blocks exceeding 2**32 bytes for
575 some purposes on 64-bit boxes. A ``PYMALLOC_DEBUG`` build was limited
576 to 4-byte allocations before.
578 - Patch #1400181, fix unicode string formatting to not use the locale.
579 This is how string objects work. u'%f' could use , instead of .
580 for the decimal point. Now both strings and unicode always use periods.
582 - Bug #1244610, #1392915, fix build problem on OpenBSD 3.7 and 3.8.
583 configure would break checking curses.h.
585 - Bug #959576: The pwd module is now builtin. This allows Python to be
586 built on UNIX platforms without $HOME set.
588 - Bug #1072182, fix some potential problems if characters are signed.
590 - Bug #889500, fix line number on SyntaxWarning for global declarations.
592 - Bug #1378022, UTF-8 files with a leading BOM crashed the interpreter.
594 - Support for converting hex strings to floats no longer works.
595 This was not portable. float('0x3') now raises a ValueError.
597 - Patch #1382163: Expose Subversion revision number to Python. New C API
598 function Py_GetBuildNumber(). New attribute sys.subversion. Build number
599 is now displayed in interactive prompt banner.
601 - Implementation of PEP 341 - Unification of try/except and try/finally.
602 "except" clauses can now be written together with a "finally" clause in
603 one try statement instead of two nested ones. Patch #1355913.
605 - Bug #1379994: Builtin unicode_escape and raw_unicode_escape codec
606 now encodes backslash correctly.
608 - Patch #1350409: Work around signal handling bug in Visual Studio 2005.
610 - Bug #1281408: Py_BuildValue now works correctly even with unsigned longs
613 - SF Bug #1350188, "setdlopenflags" leads to crash upon "import"
614 It was possible for dlerror() to return a NULL pointer, so
615 it will now use a default error message in this case.
617 - Replaced most Unicode charmap codecs with new ones using the
618 new Unicode translate string feature in the builtin charmap
619 codec; the codecs were created from the mapping tables available
620 at ftp.unicode.org and contain a few updates (e.g. the Mac OS
621 encodings now include a mapping for the Apple logo)
623 - Added a few more codecs for Mac OS encodings
625 - Sped up some Unicode operations.
627 - A new AST parser implementation was completed. The abstract
628 syntax tree is available for read-only (non-compile) access
629 to Python code; an _ast module was added.
631 - SF bug #1167751: fix incorrect code being produced for generator expressions.
632 The following code now raises a SyntaxError: foo(a = i for i in range(10))
634 - SF Bug #976608: fix SystemError when mtime of an imported file is -1.
636 - SF Bug #887946: fix segfault when redirecting stdin from a directory.
637 Provide a warning when a directory is passed on the command line.
639 - Fix segfault with invalid coding.
641 - SF bug #772896: unknown encoding results in MemoryError.
643 - All iterators now have a Boolean value of True. Formerly, some iterators
644 supported a __len__() method which evaluated to False when the iterator
647 - On 64-bit platforms, when __len__() returns a value that cannot be
648 represented as a C int, raise OverflowError.
650 - test__locale is skipped on OS X < 10.4 (only partial locale support is
653 - SF bug #893549: parsing keyword arguments was broken with a few format
656 - Changes donated by Elemental Security to make it work on AIX 5.3
657 with IBM's 64-bit compiler (SF patch #1284289). This also closes SF
658 bug #105470: test_pwd fails on 64bit system (Opteron).
660 - Changes donated by Elemental Security to make it work on HP-UX 11 on
661 Itanium2 with HP's 64-bit compiler (SF patch #1225212).
663 - Disallow keyword arguments for type constructors that don't use them
664 (fixes bug #1119418).
666 - Forward UnicodeDecodeError into SyntaxError for source encoding errors.
668 - SF bug #900092: When tracing (e.g. for hotshot), restore 'return' events for
669 exceptions that cause a function to exit.
671 - The implementation of set() and frozenset() was revised to use its
672 own internal data structure. Memory consumption is reduced by 1/3
673 and there are modest speed-ups as well. The API is unchanged.
675 - SF bug #1238681: freed pointer is used in longobject.c:long_pow().
677 - SF bug #1229429: PyObject_CallMethod failed to decrement some
678 reference counts in some error exit cases.
680 - SF bug #1185883: Python's small-object memory allocator took over
681 a block managed by the platform C library whenever a realloc specified
682 a small new size. However, there's no portable way to know then how
683 much of the address space following the pointer is valid, so there's no
684 portable way to copy data from the C-managed block into Python's
685 small-object space without risking a memory fault. Python's small-object
686 realloc now leaves such blocks under the control of the platform C
689 - SF bug #1232517: An overflow error was not detected properly when
690 attempting to convert a large float to an int in os.utime().
692 - SF bug #1224347: hex longs now print with lowercase letters just
693 like their int counterparts.
695 - SF bug #1163563: the original fix for bug #1010677 ("thread Module
696 Breaks PyGILState_Ensure()") broke badly in the case of multiple
697 interpreter states; back out that fix and do a better job (see
698 http://mail.python.org/pipermail/python-dev/2005-June/054258.html
699 for a longer write-up of the problem).
701 - SF patch #1180995: marshal now uses a binary format by default when
704 - SF patch #1181301: on platforms that appear to use IEEE 754 floats,
705 the routines that promise to produce IEEE 754 binary representations
706 of floats now simply copy bytes around.
708 - bug #967182: disallow opening files with 'wU' or 'aU' as specified by PEP
711 - patch #1109424: int, long, float, complex, and unicode now check for the
712 proper magic slot for type conversions when subclassed. Previously the
713 magic slot was ignored during conversion. Semantics now match the way
714 subclasses of str always behaved. int/long/float, conversion of an instance
715 to the base class has been moved to the proper nb_* magic slot and out of
717 Thanks Walter D�rwald.
719 - Descriptors defined in C with a PyGetSetDef structure, where the setter is
720 NULL, now raise an AttributeError when attempting to set or delete the
721 attribute. Previously a TypeError was raised, but this was inconsistent
722 with the equivalent pure-Python implementation.
724 - It is now safe to call PyGILState_Release() before
725 PyEval_InitThreads() (note that if there is reason to believe there
726 are multiple threads around you still must call PyEval_InitThreads()
727 before using the Python API; this fix is for extension modules that
728 have no way of knowing if Python is multi-threaded yet).
730 - Typing Ctrl-C whilst raw_input() was waiting in a build with threads
731 disabled caused a crash.
733 - Bug #1165306: instancemethod_new allowed the creation of a method
734 with im_class == im_self == NULL, which caused a crash when called.
736 - Move exception finalisation later in the shutdown process - this
737 fixes the crash seen in bug #1165761
739 - Added two new builtins, any() and all().
741 - Defining a class with empty parentheses is now allowed
742 (e.g., ``class C(): pass`` is no longer a syntax error).
743 Patch #1176012 added support to the 'parser' module and 'compiler' package
744 (thanks to logistix for that added support).
746 - Patch #1115086: Support PY_LONGLONG in structmember.
748 - Bug #1155938: new style classes did not check that __init__() was
751 - Patch #802188: Report characters after line continuation character
752 ('\') with a specific error message.
754 - Bug #723201: Raise a TypeError for passing bad objects to 'L' format.
756 - Bug #1124295: the __name__ attribute of file objects was
757 inadvertently made inaccessible in restricted mode.
759 - Bug #1074011: closing sys.std{out,err} now causes a flush() and
762 - min() and max() now support key= arguments with the same meaning as in
765 - The peephole optimizer now performs simple constant folding in expressions:
768 - set and frozenset objects can now be marshalled. SF #1098985.
770 - Bug #1077106: Poor argument checking could cause memory corruption
771 in calls to os.read().
773 - The parser did not complain about future statements in illegal
774 positions. It once again reports a syntax error if a future
775 statement occurs after anything other than a doc string.
777 - Change the %s format specifier for str objects so that it returns a
778 unicode instance if the argument is not an instance of basestring and
779 calling __str__ on the argument returns a unicode instance.
781 - Patch #1413181: changed ``PyThreadState_Delete()`` to forget about the
782 current thread state when the auto-GIL-state machinery knows about
783 it (since the thread state is being deleted, continuing to remember it
784 can't help, but can hurt if another thread happens to get created with
790 - Patch #1380952: fix SSL objects timing out on consecutive read()s
792 - Patch #1309579: wait3 and wait4 were added to the posix module.
794 - Patch #1231053: The audioop module now supports encoding/decoding of alaw.
795 In addition, the existing ulaw code was updated.
797 - RFE #567972: Socket objects' family, type and proto properties are
798 now exposed via new attributes.
800 - Everything under lib-old was removed. This includes the following modules:
801 Para, addpack, cmp, cmpcache, codehack, dircmp, dump, find, fmt, grep,
802 lockfile, newdir, ni, packmail, poly, rand, statcache, tb, tzparse,
803 util, whatsound, whrandom, zmod
805 - The following modules were removed: regsub, reconvert, regex, regex_syntax.
807 - re and sre were swapped, so help(re) provides full help. importing sre
808 is deprecated. The undocumented re.engine variable no longer exists.
810 - Bug #1448490: Fixed a bug that ISO-2022 codecs could not handle
811 SS2 (single-shift 2) escape sequences correctly.
813 - The unicodedata module was updated to the 4.1 version of the Unicode
814 database. The 3.2 version is still available as unicodedata.db_3_2_0
815 for applications that require this specific version (such as IDNA).
817 - The timing module is no longer built by default. It was deprecated
818 in PEP 4 in Python 2.0 or earlier.
820 - Patch 1433928: Added a new type, defaultdict, to the collections module.
821 This uses the new __missing__ hook behavior added to dict (see above).
823 - Bug #854823: socketmodule now builds on Sun platforms even when
824 INET_ADDRSTRLEN is not defined.
826 - Patch #1393157: os.startfile() now has an optional argument to specify
827 a "command verb" to invoke on the file.
829 - Bug #876637, prevent stack corruption when socket descriptor
830 is larger than FD_SETSIZE.
832 - Patch #1407135, bug #1424041: harmonize mmap behavior of anonymous memory.
833 mmap.mmap(-1, size) now returns anonymous memory in both Unix and Windows.
834 mmap.mmap(0, size) should not be used on Windows for anonymous memory.
836 - Patch #1422385: The nis module now supports access to domains other
837 than the system default domain.
839 - Use Win32 API to implement os.stat/fstat. As a result, subsecond timestamps
840 are reported, the limit on path name lengths is removed, and stat reports
841 WindowsError now (instead of OSError).
843 - Add bsddb.db.DBEnv.set_tx_timestamp allowing time based database recovery.
845 - Bug #1413192, fix seg fault in bsddb if a transaction was deleted
848 - Patch #1103116: Basic AF_NETLINK support.
850 - Bug #1402308, (possible) segfault when using mmap.mmap(-1, ...)
852 - Bug #1400822, _curses over{lay,write} doesn't work when passing 6 ints.
853 Also fix ungetmouse() which did not accept arguments properly.
854 The code now conforms to the documented signature.
856 - Bug #1400115, Fix segfault when calling curses.panel.userptr()
857 without prior setting of the userptr.
859 - Fix 64-bit problems in bsddb.
861 - Patch #1365916: fix some unsafe 64-bit mmap methods.
863 - Bug #1290333: Added a workaround for cjkcodecs' _codecs_cn build
866 - Bug #869197: os.setgroups rejects long integer arguments
868 - Bug #1346533, select.poll() doesn't raise an error if timeout > sys.maxint
870 - Bug #1344508, Fix UNIX mmap leaking file descriptors
872 - Patch #1338314, Bug #1336623: fix tarfile so it can extract
873 REGTYPE directories from tarfiles written by old programs.
875 - Patch #1407992, fixes broken bsddb module db associate when using
876 BerkeleyDB 3.3, 4.0 or 4.1.
878 - Get bsddb module to build with BerkeleyDB version 4.4
880 - Get bsddb module to build with BerkeleyDB version 3.2
882 - Patch #1309009, Fix segfault in pyexpat when the XML document is in latin_1,
883 but Python incorrectly assumes it is in UTF-8 format
885 - Fix parse errors in the readline module when compiling without threads.
887 - Patch #1288833: Removed thread lock from socket.getaddrinfo on
888 FreeBSD 5.3 and later versions which got thread-safe getaddrinfo(3).
890 - Patches #1298449 and #1298499: Add some missing checks for error
891 returns in cStringIO.c.
893 - Patch #1297028: fix segfault if call type on MultibyteCodec,
894 MultibyteStreamReader, or MultibyteStreamWriter
896 - Fix memory leak in posix.access().
898 - Patch #1213831: Fix typo in unicodedata._getcode.
900 - Bug #1007046: os.startfile() did not accept unicode strings encoded in
901 the file system encoding.
903 - Patch #756021: Special-case socket.inet_aton('255.255.255.255') for
904 platforms that don't have inet_aton().
906 - Bug #1215928: Fix bz2.BZ2File.seek() for 64-bit file offsets.
908 - Bug #1191043: Fix bz2.BZ2File.(x)readlines for files containing one
909 line without newlines.
911 - Bug #728515: mmap.resize() now resizes the file on Unix as it did
914 - Patch #1180695: Add nanosecond stat resolution, and st_gen,
915 st_birthtime for FreeBSD.
917 - Patch #1231069: The fcntl.ioctl function now uses the 'I' code for
918 the request code argument, which results in more C-like behaviour
919 for large or negative values.
921 - Bug #1234979: For the argument of thread.Lock.acquire, the Windows
922 implementation treated all integer values except 1 as false.
924 - Bug #1194181: bz2.BZ2File didn't handle mode 'U' correctly.
926 - Patch #1212117: os.stat().st_flags is now accessible as a attribute
927 if available on the platform.
929 - Patch #1103951: Expose O_SHLOCK and O_EXLOCK in the posix module if
930 available on the platform.
932 - Bug #1166660: The readline module could segfault if hook functions
933 were set in a different thread than that which called readline.
935 - collections.deque objects now support a remove() method.
937 - operator.itemgetter() and operator.attrgetter() now support retrieving
938 multiple fields. This provides direct support for sorting on multiple
939 keys (primary, secondary, etc).
941 - os.access now supports Unicode path names on non-Win32 systems.
943 - Patches #925152, #1118602: Avoid reading after the end of the buffer
944 in pyexpat.GetInputContext.
946 - Patches #749830, #1144555: allow UNIX mmap size to default to current
949 - Added functional.partial(). See PEP309.
951 - Patch #1093585: raise a ValueError for negative history items in readline.
952 {remove_history,replace_history}
954 - The spwd module has been added, allowing access to the shadow password
957 - stat_float_times is now True.
959 - array.array objects are now picklable.
961 - the cPickle module no longer accepts the deprecated None option in the
962 args tuple returned by __reduce__().
964 - itertools.islice() now accepts None for the start and step arguments.
965 This allows islice() to work more readily with slices:
966 islice(s.start, s.stop, s.step)
968 - datetime.datetime() now has a strptime class method which can be used to
969 create datetime object using a string and format.
971 - Patch #1117961: Replace the MD5 implementation from RSA Data Security Inc
972 with the implementation from http://sourceforge.net/projects/libmd5-rfc/.
977 - PEP 338: new module runpy defines a run_module function to support
978 executing modules which provide access to source code or a code object
979 via the PEP 302 import mechanisms.
981 - The email module's parsedate_tz function now sets the daylight savings
982 flag to -1 (unknown) since it can't tell from the date whether it should
985 - Patch #624325: urlparse.urlparse() and urlparse.urlsplit() results
986 now sport attributes that provide access to the parts of the result.
988 - Patch #1462498: sgmllib now handles entity and character references
991 - Added the sqlite3 package. This is based on pysqlite2.1.3, and provides
992 a DB-API interface in the standard library. You'll need sqlite 3.0.8 or
993 later to build this - if you have an earlier version, the C extension
994 module will not be built.
996 - Bug #1460340: ``random.sample(dict)`` failed in various ways. Dicts
997 aren't officially supported here, and trying to use them will probably
998 raise an exception some day. But dicts have been allowed, and "mostly
999 worked", so support for them won't go away without warning.
1001 - Bug #1445068: getpass.getpass() can now be given an explicit stream
1002 argument to specify where to write the prompt.
1004 - Patch #1462313, bug #1443328: the pickle modules now can handle classes
1005 that have __private names in their __slots__.
1007 - Bug #1250170: mimetools now handles socket.gethostname() failures gracefully.
1009 - patch #1457316: "setup.py upload" now supports --identity to select the
1010 key to be used for signing the uploaded code.
1012 - Queue.Queue objects now support .task_done() and .join() methods
1013 to make it easier to monitor when daemon threads have completed
1014 processing all enqueued tasks. Patch #1455676.
1016 - popen2.Popen objects now preserve the command in a .cmd attribute.
1018 - Added the ctypes ffi package.
1020 - email 4.0 package now integrated. This is largely the same as the email 3.0
1021 package that was included in Python 2.3, except that PEP 8 module names are
1022 now used (e.g. mail.message instead of email.Message). The MIME classes
1023 have been moved to a subpackage (e.g. email.mime.text instead of
1024 email.MIMEText). The old names are still supported for now. Several
1025 deprecated Message methods have been removed and lots of bugs have been
1026 fixed. More details can be found in the email package documentation.
1028 - Patches #1436130/#1443155: codecs.lookup() now returns a CodecInfo object
1029 (a subclass of tuple) that provides incremental decoders and encoders
1030 (a way to use stateful codecs without the stream API). Python functions
1031 codecs.getincrementaldecoder() and codecs.getincrementalencoder() as well
1032 as C functions PyCodec_IncrementalEncoder() and PyCodec_IncrementalDecoder()
1035 - Patch #1359365: Calling next() on a closed StringIO.String object raises
1036 a ValueError instead of a StopIteration now (like file and cString.String do).
1037 cStringIO.StringIO.isatty() will raise a ValueError now if close() has been
1038 called before (like file and StringIO.StringIO do).
1040 - A regrtest option -w was added to re-run failed tests in verbose mode.
1042 - Patch #1446372: quit and exit can now be called from the interactive
1043 interpreter to exit.
1045 - The function get_count() has been added to the gc module, and gc.collect()
1046 grew an optional 'generation' argument.
1048 - A library msilib to generate Windows Installer files, and a distutils
1049 command bdist_msi have been added.
1051 - PEP 343: new module contextlib.py defines decorator @contextmanager
1052 and helpful context managers nested() and closing().
1054 - The compiler package now supports future imports after the module docstring.
1056 - Bug #1413790: zipfile now sanitizes absolute archive names that are
1057 not allowed by the specs.
1059 - Patch #1215184: FileInput now can be given an opening hook which can
1060 be used to control how files are opened.
1062 - Patch #1212287: fileinput.input() now has a mode parameter for
1063 specifying the file mode input files should be opened with.
1065 - Patch #1215184: fileinput now has a fileno() function for getting the
1066 current file number.
1068 - Patch #1349274: gettext.install() now optionally installs additional
1069 translation functions other than _() in the builtin namespace.
1071 - Patch #1337756: fileinput now accepts Unicode filenames.
1073 - Patch #1373643: The chunk module can now read chunks larger than
1076 - Patch #1417555: SimpleHTTPServer now returns Last-Modified headers.
1078 - Bug #1430298: It is now possible to send a mail with an empty
1079 return address using smtplib.
1081 - Bug #1432260: The names of lambda functions are now properly displayed
1084 - Patch #1412872: zipfile now sets the creator system to 3 (Unix)
1085 unless the system is Win32.
1087 - Patch #1349118: urllib now supports user:pass@ style proxy
1088 specifications, raises IOErrors when proxies for unsupported protocols
1089 are defined, and uses the https proxy on https redirections.
1091 - Bug #902075: urllib2 now supports 'host:port' style proxy specifications.
1093 - Bug #1407902: Add support for sftp:// URIs to urlparse.
1095 - Bug #1371247: Update Windows locale identifiers in locale.py.
1097 - Bug #1394565: SimpleHTTPServer now doesn't choke on query parameters
1100 - Bug #1403410: The warnings module now doesn't get confused
1101 when it can't find out the module name it generates a warning for.
1103 - Patch #1177307: Added a new codec utf_8_sig for UTF-8 with a BOM signature.
1105 - Patch #1157027: cookielib mishandles RFC 2109 cookies in Netscape mode
1107 - Patch #1117398: cookielib.LWPCookieJar and .MozillaCookieJar now raise
1108 LoadError as documented, instead of IOError. For compatibility,
1109 LoadError subclasses IOError.
1111 - Added the hashlib module. It provides secure hash functions for MD5 and
1112 SHA1, 224, 256, 384, and 512. Note that recent developments make the
1113 historic MD5 and SHA1 unsuitable for cryptographic-strength applications.
1114 In <http://mail.python.org/pipermail/python-dev/2005-December/058850.html>
1115 Ronald L. Rivest offered this advice for Python:
1117 "The consensus of researchers in this area (at least as
1118 expressed at the NIST Hash Function Workshop 10/31/05),
1119 is that SHA-256 is a good choice for the time being, but
1120 that research should continue, and other alternatives may
1121 arise from this research. The larger SHA's also seem OK."
1123 - Added a subset of Fredrik Lundh's ElementTree package. Available
1124 modules are xml.etree.ElementTree, xml.etree.ElementPath, and
1125 xml.etree.ElementInclude, from ElementTree 1.2.6.
1127 - Patch #1162825: Support non-ASCII characters in IDLE window titles.
1129 - Bug #1365984: urllib now opens "data:" URLs again.
1131 - Patch #1314396: prevent deadlock for threading.Thread.join() when an exception
1132 is raised within the method itself on a previous call (e.g., passing in an
1135 - Bug #1340337: change time.strptime() to always return ValueError when there
1136 is an error in the format string.
1138 - Patch #754022: Greatly enhanced webbrowser.py (by Oleg Broytmann).
1140 - Bug #729103: pydoc.py: Fix docother() method to accept additional
1143 - Patch #1300515: xdrlib.py: Fix pack_fstring() to really use null bytes
1146 - Bug #1296004: httplib.py: Limit maximal amount of data read from the
1147 socket to avoid a MemoryError on Windows.
1149 - Patch #1166948: locale.py: Prefer LC_ALL, LC_CTYPE and LANG over LANGUAGE
1150 to get the correct encoding.
1152 - Patch #1166938: locale.py: Parse LANGUAGE as a colon separated list of
1155 - Patch #1268314: Cache lines in StreamReader.readlines for performance.
1157 - Bug #1290505: Fix clearing the regex cache for time.strptime().
1159 - Bug #1167128: Fix size of a symlink in a tarfile to be 0.
1161 - Patch #810023: Fix off-by-one bug in urllib.urlretrieve reporthook
1164 - Bug #1163178: Make IDNA return an empty string when the input is empty.
1166 - Patch #848017: Make Cookie more RFC-compliant. Use CRLF as default output
1167 separator and do not output trailing semicolon.
1169 - Patch #1062060: urllib.urlretrieve() now raises a new exception, named
1170 ContentTooShortException, when the actually downloaded size does not
1171 match the Content-Length header.
1173 - Bug #1121494: distutils.dir_utils.mkpath now accepts Unicode strings.
1175 - Bug #1178484: Return complete lines from codec stream readers
1176 even if there is an exception in later lines, resulting in
1177 correct line numbers for decoding errors in source code.
1179 - Bug #1192315: Disallow negative arguments to clear() in pdb.
1181 - Patch #827386: Support absolute source paths in msvccompiler.py.
1183 - Patch #1105730: Apply the new implementation of commonprefix in posixpath
1184 to ntpath, macpath, os2emxpath and riscospath.
1186 - Fix a problem in Tkinter introduced by SF patch #869468: delete bogus
1187 __hasattr__ and __delattr__ methods on class Tk that were breaking
1190 - Bug #1015140: disambiguated the term "article id" in nntplib docs and
1191 docstrings to either "article number" or "message id".
1193 - Bug #1238170: threading.Thread.__init__ no longer has "kwargs={}" as a
1194 parameter, but uses the usual "kwargs=None".
1196 - textwrap now processes text chunks at O(n) speed instead of O(n**2).
1197 Patch #1209527 (Contributed by Connelly).
1199 - urllib2 has now an attribute 'httpresponses' mapping from HTTP status code
1200 to W3C name (404 -> 'Not Found'). RFE #1216944.
1202 - Bug #1177468: Don't cache the /dev/urandom file descriptor for os.urandom,
1203 as this can cause problems with apps closing all file descriptors.
1205 - Bug #839151: Fix an attempt to access sys.argv in the warnings module;
1206 it can be missing in embedded interpreters
1208 - Bug #1155638: Fix a bug which affected HTTP 0.9 responses in httplib.
1210 - Bug #1100201: Cross-site scripting was possible on BaseHTTPServer via
1213 - Bug #1108948: Cookie.py produced invalid JavaScript code.
1215 - The tokenize module now detects and reports indentation errors.
1218 - The tokenize module has a new untokenize() function to support a full
1219 roundtrip from lexed tokens back to Python source code. In addition,
1220 the generate_tokens() function now accepts a callable argument that
1221 terminates by raising StopIteration.
1223 - Bug #1196315: fix weakref.WeakValueDictionary constructor.
1225 - Bug #1213894: os.path.realpath didn't resolve symlinks that were the first
1226 component of the path.
1228 - Patch #1120353: The xmlrpclib module provides better, more transparent,
1229 support for datetime.{datetime,date,time} objects. With use_datetime set
1230 to True, applications shouldn't have to fiddle with the DateTime wrapper
1233 - distutils.commands.upload was added to support uploading distribution
1236 - distutils.commands.register now encodes the data as UTF-8 before posting
1239 - decimal operator and comparison methods now return NotImplemented
1240 instead of raising a TypeError when interacting with other types. This
1241 allows other classes to implement __radd__ style methods and have them
1244 - Bug #1163325: Decimal infinities failed to hash. Attempting to
1245 hash a NaN raised an InvalidOperation instead of a TypeError.
1247 - Patch #918101: Add tarfile open mode r|* for auto-detection of the
1248 stream compression; add, for symmetry reasons, r:* as a synonym of r.
1250 - Patch #1043890: Add extractall method to tarfile.
1252 - Patch #1075887: Don't require MSVC in distutils if there is nothing
1255 - Patch #1103407: Properly deal with tarfile iterators when untarring
1256 symbolic links on Windows.
1258 - Patch #645894: Use getrusage for computing the time consumption in
1259 profile.py if available.
1261 - Patch #1046831: Use get_python_version where appropriate in sysconfig.py.
1263 - Patch #1117454: Remove code to special-case cookies without values
1266 - Patch #1117339: Add cookielib special name tests.
1268 - Patch #1112812: Make bsddb/__init__.py more friendly for modulefinder.
1270 - Patch #1110248: SYNC_FLUSH the zlib buffer for GZipFile.flush.
1272 - Patch #1107973: Allow to iterate over the lines of a tarfile.ExFileObject.
1274 - Patch #1104111: Alter setup.py --help and --help-commands.
1276 - Patch #1121234: Properly cleanup _exit and tkerror commands.
1278 - Patch #1049151: xdrlib now unpacks booleans as True or False.
1280 - Fixed bug in a NameError bug in cookielib. Patch #1116583.
1282 - Applied a security fix to SimpleXMLRPCserver (PSF-2005-001). This
1283 disables recursive traversal through instance attributes, which can
1284 be exploited in various ways.
1286 - Bug #1222790: in SimpleXMLRPCServer, set the reuse-address and close-on-exec
1287 flags on the HTTP listening socket.
1289 - Bug #792570: SimpleXMLRPCServer had problems if the request grew too large.
1290 Fixed by reading the HTTP body in chunks instead of one big socket.read().
1292 - Patches #893642, #1039083: add allow_none, encoding arguments to constructors of
1293 SimpleXMLRPCServer and CGIXMLRPCRequestHandler.
1295 - Bug #1110478: Revert os.environ.update to do putenv again.
1297 - Bug #1103844: fix distutils.install.dump_dirs() with negated options.
1299 - os.{SEEK_SET, SEEK_CUR, SEEK_END} have been added for convenience.
1301 - Enhancements to the csv module:
1303 + Dialects are now validated by the underlying C code, better
1304 reflecting its capabilities, and improving its compliance with
1306 + Dialect parameter parsing has been re-implemented to improve error
1308 + quotechar=None and quoting=QUOTE_NONE now work the way PEP 305
1310 + the parser now removes the escapechar prefix from escaped characters.
1311 + when quoting=QUOTE_NONNUMERIC, the writer now tests for numeric
1312 types, rather than any object that can be represented as a numeric.
1313 + when quoting=QUOTE_NONNUMERIC, the reader now casts unquoted fields
1315 + reader now allows \r characters to be quoted (previously it only allowed
1317 + writer doublequote handling improved.
1318 + Dialect classes passed to the module are no longer instantiated by
1319 the module before being parsed (the former validation scheme required
1320 this, but the mechanism was unreliable).
1321 + The dialect registry now contains instances of the internal
1322 C-coded dialect type, rather than references to python objects.
1323 + the internal c-coded dialect type is now immutable.
1324 + register_dialect now accepts the same keyword dialect specifications
1325 as the reader and writer, allowing the user to register dialects
1326 without first creating a dialect class.
1327 + a configurable limit to the size of parsed fields has been added -
1328 previously, an unmatched quote character could result in the entire
1329 file being read into the field buffer before an error was reported.
1330 + A new module method csv.field_size_limit() has been added that sets
1331 the parser field size limit (returning the former limit). The initial
1333 + A line_num attribute has been added to the reader object, which tracks
1334 the number of lines read from the source iterator. This is not
1335 the same as the number of records returned, as records can span
1337 + reader and writer objects were not being registered with the cyclic-GC.
1338 This has been fixed.
1340 - _DummyThread objects in the threading module now delete self.__block that is
1341 inherited from _Thread since it uses up a lock allocated by 'thread'. The
1342 lock primitives tend to be limited in number and thus should not be wasted on
1343 a _DummyThread object. Fixes bug #1089632.
1345 - The imghdr module now detects Exif files.
1347 - StringIO.truncate() now correctly adjusts the size attribute.
1350 - locale.py now uses an updated locale alias table (built using
1351 Tools/i18n/makelocalealias.py, a tool to parse the X11 locale
1352 alias file); the encoding lookup was enhanced to use Python's
1353 encoding alias table.
1355 - moved deprecated modules to Lib/lib-old: whrandom, tzparse, statcache.
1357 - the pickle module no longer accepts the deprecated None option in the
1358 args tuple returned by __reduce__().
1360 - optparse now optionally imports gettext. This allows its use in setup.py.
1362 - the pickle module no longer uses the deprecated bin parameter.
1364 - the shelve module no longer uses the deprecated binary parameter.
1366 - the pstats module no longer uses the deprecated ignore() method.
1368 - the filecmp module no longer uses the deprecated use_statcache argument.
1370 - unittest.TestCase.run() and unittest.TestSuite.run() can now be successfully
1371 extended or overridden by subclasses. Formerly, the subclassed method would
1372 be ignored by the rest of the module. (Bug #1078905).
1374 - heapq.nsmallest() and heapq.nlargest() now support key= arguments with
1375 the same meaning as in list.sort().
1377 - Bug #1076985: ``codecs.StreamReader.readline()`` now calls ``read()`` only
1378 once when a size argument is given. This prevents a buffer overflow in the
1379 tokenizer with very long source lines.
1381 - Bug #1083110: ``zlib.decompress.flush()`` would segfault if called
1382 immediately after creating the object, without any intervening
1383 ``.decompress()`` calls.
1385 - The reconvert.quote function can now emit triple-quoted strings. The
1386 reconvert module now has some simple documentation.
1388 - ``UserString.MutableString`` now supports negative indices in
1389 ``__setitem__`` and ``__delitem__``
1391 - Bug #1149508: ``textwrap`` now handles hyphenated numbers (eg. "2004-03-05")
1394 - Partial fixes for SF bugs #1163244 and #1175396: If a chunk read by
1395 ``codecs.StreamReader.readline()`` has a trailing "\r", read one more
1396 character even if the user has passed a size parameter to get a proper
1397 line ending. Remove the special handling of a "\r\n" that has been split
1400 - Bug #1251300: On UCS-4 builds the "unicode-internal" codec will now complain
1401 about illegal code points. The codec now supports PEP 293 style error
1404 - Bug #1235646: ``codecs.StreamRecoder.next()`` now reencodes the data it reads
1405 from the input stream, so that the output is a byte string in the correct
1406 encoding instead of a unicode string.
1408 - Bug #1202493: Fixing SRE parser to handle '{}' as perl does, rather than
1409 considering it exactly like a '*'.
1411 - Bug #1245379: Add "unicode-1-1-utf-7" as an alias for "utf-7" to
1412 ``encodings.aliases``.
1414 - ` uu.encode()`` and ``uu.decode()`` now support unicode filenames.
1416 - Patch #1413711: Certain patterns of differences were making difflib
1417 touch the recursion limit.
1419 - Bug #947906: An object oriented interface has been added to the calendar
1420 module. It's possible to generate HTML calendar now and the module can be
1421 called as a script (e.g. via ``python -mcalendar``). Localized month and
1422 weekday names can be ouput (even if an exotic encoding is used) using
1423 special classes that use unicode.
1428 - Fix test_float, test_long, and test_struct failures on Tru64 with gcc
1429 by using -mieee gcc option.
1431 - Patch #1432345: Make python compile on DragonFly.
1433 - Build support for Win64-AMD64 was added.
1435 - Patch #1428494: Prefer linking against ncursesw over ncurses library.
1437 - Patch #881820: look for openpty and forkpty also in libbsd.
1439 - The sources of zlib are now part of the Python distribution (zlib 1.2.3).
1440 The zlib module is now builtin on Windows.
1442 - Use -xcode=pic32 for CCSHARED on Solaris with SunPro.
1444 - Bug #1189330: configure did not correctly determine the necessary
1445 value of LINKCC if python was built with GCC 4.0.
1447 - Upgrade Windows build to zlib 1.2.3 which eliminates a potential security
1448 vulnerability in zlib 1.2.1 and 1.2.2.
1450 - EXTRA_CFLAGS has been introduced as an environment variable to hold compiler
1451 flags that change binary compatibility. Changes were also made to
1452 distutils.sysconfig to also use the environment variable when used during
1453 compilation of the interpreter and of C extensions through distutils.
1455 - SF patch 1171735: Darwin 8's headers are anal about POSIX compliance,
1456 and linking has changed (prebinding is now deprecated, and libcc_dynamic
1457 no longer exists). This configure patch makes things right.
1459 - Bug #1158607: Build with --disable-unicode again.
1461 - spwdmodule.c is built only if either HAVE_GETSPNAM or HAVE_HAVE_GETSPENT is
1462 defined. Discovered as a result of not being able to build on OS X.
1464 - setup.py now uses the directories specified in LDFLAGS using the -L option
1465 and in CPPFLAGS using the -I option for adding library and include
1466 directories, respectively, for compiling extension modules against. This has
1467 led to the core being compiled using the values in CPPFLAGS. It also removes
1468 the need for the special-casing of both DarwinPorts and Fink for darwin since
1469 the proper directories can be specified in LDFLAGS (``-L/sw/lib`` for Fink,
1470 ``-L/opt/local/lib`` for DarwinPorts) and CPPFLAGS (``-I/sw/include`` for
1471 Fink, ``-I/opt/local/include`` for DarwinPorts).
1473 - Test in configure.in that checks for tzset no longer dependent on tm->tm_zone
1474 to exist in the struct (not required by either ISO C nor the UNIX 2 spec).
1475 Tests for sanity in tzname when HAVE_TZNAME defined were also defined.
1476 Closes bug #1096244. Thanks Gregory Bond.
1481 - ``PyMem_{Del, DEL}`` and ``PyMem_{Free, FREE}`` no longer map to
1482 ``PyObject_{Free, FREE}``. They map to the system ``free()`` now. If memory
1483 is obtained via the ``PyObject_`` family, it must be released via the
1484 ``PyObject_`` family, and likewise for the ``PyMem_`` family. This has
1485 always been officially true, but when Python's small-object allocator was
1486 introduced, an attempt was made to cater to a few extension modules
1487 discovered at the time that obtained memory via ``PyObject_New`` but
1488 released it via ``PyMem_DEL``. It's years later, and if such code still
1489 exists it will fail now (probably with segfaults, but calling wrong
1490 low-level memory management functions can yield many symptoms).
1492 - Added a C API for set and frozenset objects.
1494 - Removed PyRange_New().
1496 - Patch #1313939: PyUnicode_DecodeCharmap() accepts a unicode string as the
1497 mapping argument now. This string is used as a mapping table. Byte values
1498 greater than the length of the string and 0xFFFE are treated as undefined
1505 - In test_os, st_?time is now truncated before comparing it with ST_?TIME.
1507 - Patch #1276356: New resource "urlfetch" is implemented. This enables
1508 even impatient people to run tests that require remote files.
1514 - Bug #1402224: Add warning to dl docs about crashes.
1516 - Bug #1396471: Document that Windows' ftell() can return invalid
1517 values for text files with UNIX-style line endings.
1519 - Bug #1274828: Document os.path.splitunc().
1521 - Bug #1190204: Clarify which directories are searched by site.py.
1523 - Bug #1193849: Clarify os.path.expanduser() documentation.
1525 - Bug #1243192: re.UNICODE and re.LOCALE affect \d, \D, \s and \S.
1527 - Bug #755617: Document the effects of os.chown() on Windows.
1529 - Patch #1180012: The documentation for modulefinder is now in the library reference.
1531 - Patch #1213031: Document that os.chown() accepts argument values of -1.
1533 - Bug #1190563: Document os.waitpid() return value with WNOHANG flag.
1535 - Bug #1175022: Correct the example code for property().
1537 - Document the IterableUserDict class in the UserDict module.
1538 Closes bug #1166582.
1540 - Remove all latent references for "Macintosh" that referred to semantics for
1541 Mac OS 9 and change to reflect the state for OS X.
1542 Closes patch #1095802. Thanks Jack Jansen.
1551 - FreeBSD 7 support is added.
1557 - Created Misc/Vim/vim_syntax.py to auto-generate a python.vim file in that
1558 directory for syntax highlighting in Vim. Vim directory was added and placed
1559 vimrc to it (was previous up a level).
1561 - Added two new files to Tools/scripts: pysource.py, which recursively
1562 finds Python source files, and findnocoding.py, which finds Python
1563 source files that need an encoding declaration.
1564 Patch #784089, credits to Oleg Broytmann.
1566 - Bug #1072853: pindent.py used an uninitialized variable.
1568 - Patch #1177597: Correct Complex.__init__.
1570 - Fixed a display glitch in Pynche, which could cause the right arrow to
1571 wiggle over by a pixel.
1575 **(For information about older versions, consult the HISTORY file.)**