*ordering* between objects; there is only a default equality test
(defined by an object being equal to itself only). Read the comment
in object.c. The current implementation never uses a three-way
comparison to compute a rich comparison, but it does use a rich
comparison to compute a three-way comparison. I'm not quite done
ripping out all the calls to PyObject_Compare/Cmp, or replacing
tp_compare implementations with tp_richcompare implementations;
but much of that has happened (to make most unit tests pass).
The following tests still fail, because I need help deciding
or understanding:
test_codeop -- depends on comparing code objects
test_datetime -- need Tim Peters' opinion
test_marshal -- depends on comparing code objects
test_mutants -- need help understanding it
The problem with test_codeop and test_marshal is this: these tests
compare two different code objects and expect them to be equal.
Is that still a feature we'd like to support? I've temporarily
removed the comparison and hash code from code objects, so they
use the default (equality by pointer only) comparison.
For the other two tests, run them to see for yourself.
(There may be more failing test with "-u all".)
A general problem with getting lots of these tests to pass is
the reality that for object types that have a natural total ordering,
implementing __cmp__ is much more convenient than implementing
__eq__, __ne__, __lt__, and so on. Should we go back to allowing
__cmp__ to provide a total ordering? Should we provide some other
way to implement rich comparison with a single method override?
Alex proposed a __key__() method; I've considered a __richcmp__()
method. Or perhaps __cmp__() just shouldn't be killed off...
which unfortunately means the errors from the bytes type change somewhat:
bytes([300]) still raises a ValueError, but bytes([10**100]) now raises a
TypeError (either that, or bytes(1.0) also raises a ValueError --
PyNumber_AsSsize_t() can only raise one type of exception.)
Merged revisions 51188-51433 via svnmerge from
svn+ssh://pythondev@svn.python.org/python/trunk
........
r51189 | kurt.kaiser | 2006-08-10 19:11:09 +0200 (Thu, 10 Aug 2006) | 4 lines
Retrieval of previous shell command was not always preserving indentation
since 1.2a1) Patch 1528468 Tal Einat.
........
r51190 | guido.van.rossum | 2006-08-10 19:41:07 +0200 (Thu, 10 Aug 2006) | 3 lines
Chris McDonough's patch to defend against certain DoS attacks on FieldStorage.
SF bug #1112549.
........
r51191 | guido.van.rossum | 2006-08-10 19:42:50 +0200 (Thu, 10 Aug 2006) | 2 lines
News item for SF bug 1112549.
........
r51192 | guido.van.rossum | 2006-08-10 20:09:25 +0200 (Thu, 10 Aug 2006) | 2 lines
Fix title -- it's rc1, not beta3.
........
r51194 | martin.v.loewis | 2006-08-10 21:04:00 +0200 (Thu, 10 Aug 2006) | 3 lines
Update dangling references to the 3.2 database to
mention that this is UCD 4.1 now.
........
r51195 | tim.peters | 2006-08-11 00:45:34 +0200 (Fri, 11 Aug 2006) | 6 lines
Followup to bug #1069160.
PyThreadState_SetAsyncExc(): internal correctness changes wrt
refcount safety and deadlock avoidance. Also added a basic test
case (relying on ctypes) and repaired the docs.
........
r51196 | tim.peters | 2006-08-11 00:48:45 +0200 (Fri, 11 Aug 2006) | 2 lines
Whitespace normalization.
........
r51197 | tim.peters | 2006-08-11 01:22:13 +0200 (Fri, 11 Aug 2006) | 5 lines
Whitespace normalization broke test_cgi, because a line
of quoted test data relied on preserving a single trailing
blank. Changed the string from raw to regular, and forced
in the trailing blank via an explicit \x20 escape.
........
r51198 | tim.peters | 2006-08-11 02:49:01 +0200 (Fri, 11 Aug 2006) | 10 lines
test_PyThreadState_SetAsyncExc(): This is failing on some
64-bit boxes. I have no idea what the ctypes docs mean
by "integers", and blind-guessing here that it intended to
mean the signed C "int" type, in which case perhaps I can
repair this by feeding the thread id argument to type
ctypes.c_long().
Also made the worker thread daemonic, so it doesn't hang
Python shutdown if the test continues to fail.
........
r51199 | tim.peters | 2006-08-11 05:49:10 +0200 (Fri, 11 Aug 2006) | 6 lines
force_test_exit(): This has been completely ineffective
at stopping test_signal from hanging forever on the Tru64
buildbot. That could be because there's no such thing as
signal.SIGALARM. Changed to the idiotic (but standard)
signal.SIGALRM instead, and added some more debug output.
........
r51202 | neal.norwitz | 2006-08-11 08:09:41 +0200 (Fri, 11 Aug 2006) | 6 lines
Fix the failures on cygwin (2006-08-10 fixed the actual locking issue).
The first hunk changes the colon to an ! like other Windows variants.
We need to always wait on the child so the lock gets released and
no other tests fail. This is the try/finally in the second hunk.
........
r51205 | georg.brandl | 2006-08-11 09:15:38 +0200 (Fri, 11 Aug 2006) | 3 lines
Add Chris McDonough (latest cgi.py patch)
........
r51206 | georg.brandl | 2006-08-11 09:26:10 +0200 (Fri, 11 Aug 2006) | 3 lines
logging's atexit hook now runs even if the rest of the module has
already been cleaned up.
........
r51212 | thomas.wouters | 2006-08-11 17:02:39 +0200 (Fri, 11 Aug 2006) | 4 lines
Add ignore of *.pyc and *.pyo to Lib/xml/etree/.
........
r51215 | thomas.heller | 2006-08-11 21:55:35 +0200 (Fri, 11 Aug 2006) | 7 lines
When a ctypes C callback function is called, zero out the result
storage before converting the result to C data. See the comment in
the code for details.
Provide a better context for errors when the conversion of a callback
function's result cannot be converted.
........
r51218 | neal.norwitz | 2006-08-12 03:43:40 +0200 (Sat, 12 Aug 2006) | 6 lines
Klocwork made another run and found a bunch more problems.
This is the first batch of fixes that should be easy to verify based on context.
This fixes problem numbers: 220 (ast), 323-324 (symtable),
321-322 (structseq), 215 (array), 210 (hotshot), 182 (codecs), 209 (etree).
........
r51219 | neal.norwitz | 2006-08-12 03:45:47 +0200 (Sat, 12 Aug 2006) | 9 lines
Even though _Py_Mangle() isn't truly public anyone can call it and
there was no verification that privateobj was a PyString. If it wasn't
a string, this could have allowed a NULL pointer to creep in below and crash.
I wonder if this should be PyString_CheckExact? Must identifiers be strings
or can they be subclasses?
Klocwork #275
........
r51220 | neal.norwitz | 2006-08-12 03:46:42 +0200 (Sat, 12 Aug 2006) | 5 lines
It's highly unlikely, though possible for PyEval_Get*() to return NULLs.
So be safe and do an XINCREF.
Klocwork # 221-222.
........
r51221 | neal.norwitz | 2006-08-12 03:47:59 +0200 (Sat, 12 Aug 2006) | 7 lines
This code is actually not used unless WITHOUT_COMPLEX is defined.
However, there was no error checking that PyFloat_FromDouble returned
a valid pointer. I believe this change is correct as it seemed
to follow other code in the area.
Klocwork # 292.
........
r51222 | neal.norwitz | 2006-08-12 03:49:12 +0200 (Sat, 12 Aug 2006) | 5 lines
Handle NULL nodes while parsing. I'm not entirely sure this is correct.
There might be something else that needs to be done to setup the error.
Klocwork #295.
........
r51223 | neal.norwitz | 2006-08-12 03:50:38 +0200 (Sat, 12 Aug 2006) | 6 lines
If _stat_float_times is false, we will try to INCREF ival which could be NULL.
Return early in that case. The caller checks for PyErr_Occurred so this
should be ok.
Klocwork #297
........
r51224 | neal.norwitz | 2006-08-12 03:51:12 +0200 (Sat, 12 Aug 2006) | 3 lines
Move the assert which checks for a NULL pointer first.
Klocwork #274.
........
r51225 | neal.norwitz | 2006-08-12 03:53:28 +0200 (Sat, 12 Aug 2006) | 5 lines
Try to handle a malloc failure. I'm not entirely sure this is correct.
There might be something else we need to do to handle the exception.
Klocwork # 212-213
........
r51226 | neal.norwitz | 2006-08-12 03:57:47 +0200 (Sat, 12 Aug 2006) | 6 lines
I'm not sure why this code allocates this string for the error message.
I think it would be better to always use snprintf and have the format
limit the size of the name appropriately (like %.200s).
Klocwork #340
........
r51227 | neal.norwitz | 2006-08-12 04:06:34 +0200 (Sat, 12 Aug 2006) | 3 lines
Check returned pointer is valid.
Klocwork #233
........
r51228 | neal.norwitz | 2006-08-12 04:12:30 +0200 (Sat, 12 Aug 2006) | 1 line
Whoops, how did that get in there. :-) Revert all the parts of 51227 that were not supposed to go it. Only Modules/_ctypes/cfields.c was supposed to be changed
........
r51229 | neal.norwitz | 2006-08-12 04:33:36 +0200 (Sat, 12 Aug 2006) | 4 lines
Don't deref v if it's NULL.
Klocwork #214
........
r51230 | neal.norwitz | 2006-08-12 05:16:54 +0200 (Sat, 12 Aug 2006) | 5 lines
Check return of PyMem_MALLOC (garbage) is non-NULL.
Check seq in both portions of if/else.
Klocwork #289-290.
........
r51231 | neal.norwitz | 2006-08-12 05:17:41 +0200 (Sat, 12 Aug 2006) | 4 lines
PyModule_GetDict() can fail, produce fatal errors if this happens on startup.
Klocwork #298-299.
........
r51232 | neal.norwitz | 2006-08-12 05:18:50 +0200 (Sat, 12 Aug 2006) | 5 lines
Verify verdat which is returned from malloc is not NULL.
Ensure we don't pass NULL to free.
Klocwork #306 (at least the first part, checking malloc)
........
r51233 | tim.peters | 2006-08-12 06:42:47 +0200 (Sat, 12 Aug 2006) | 35 lines
test_signal: Signal handling on the Tru64 buildbot
appears to be utterly insane. Plug some theoretical
insecurities in the test script:
- Verify that the SIGALRM handler was actually installed.
- Don't call alarm() before the handler is installed.
- Move everything that can fail inside the try/finally,
so the test cleans up after itself more often.
- Try sending all the expected signals in
force_test_exit(), not just SIGALRM. Since that was
fixed to actually send SIGALRM (instead of invisibly
dying with an AttributeError), we've seen that sending
SIGALRM alone does not stop this from hanging.
- Move the "kill the child" business into the finally
clause, so the child doesn't survive test failure
to send SIGALRM to other tests later (there are also
baffling SIGALRM-related failures in test_socket).
- Cancel the alarm in the finally clause -- if the
test dies early, we again don't want SIGALRM showing
up to confuse a later test.
Alas, this still relies on timing luck wrt the spawned
script that sends the test signals, but it's hard to see
how waiting for seconds can so often be so unlucky.
test_threadedsignals: curiously, this test never fails
on Tru64, but doesn't normally signal SIGALRM. Anyway,
fixed an obvious (but probably inconsequential) logic
error.
........
r51234 | tim.peters | 2006-08-12 07:17:41 +0200 (Sat, 12 Aug 2006) | 8 lines
Ah, fudge. One of the prints here actually "shouldn't be"
protected by "if verbose:", which caused the test to fail on
all non-Windows boxes.
Note that I deliberately didn't convert this to unittest yet,
because I expect it would be even harder to debug this on Tru64
after conversion.
........
r51235 | georg.brandl | 2006-08-12 10:32:02 +0200 (Sat, 12 Aug 2006) | 3 lines
Repair logging test spew caused by rev. 51206.
........
r51236 | neal.norwitz | 2006-08-12 19:03:09 +0200 (Sat, 12 Aug 2006) | 8 lines
Patch #1538606, Patch to fix __index__() clipping.
I modified this patch some by fixing style, some error checking, and adding
XXX comments. This patch requires review and some changes are to be expected.
I'm checking in now to get the greatest possible review and establish a
baseline for moving forward. I don't want this to hold up release if possible.
........
r51238 | neal.norwitz | 2006-08-12 20:44:06 +0200 (Sat, 12 Aug 2006) | 10 lines
Fix a couple of bugs exposed by the new __index__ code. The 64-bit buildbots
were failing due to inappropriate clipping of numbers larger than 2**31
with new-style classes. (typeobject.c) In reviewing the code for classic
classes, there were 2 problems. Any negative value return could be returned.
Always return -1 if there was an error. Also make the checks similar
with the new-style classes. I believe this is correct for 32 and 64 bit
boxes, including Windows64.
Add a test of classic classes too.
........
r51240 | neal.norwitz | 2006-08-13 02:20:49 +0200 (Sun, 13 Aug 2006) | 1 line
SF bug #1539336, distutils example code missing
........
r51245 | neal.norwitz | 2006-08-13 20:10:10 +0200 (Sun, 13 Aug 2006) | 6 lines
Move/copy assert for tstate != NULL before first use.
Verify that PyEval_Get{Globals,Locals} returned valid pointers.
Klocwork 231-232
........
r51246 | neal.norwitz | 2006-08-13 20:10:28 +0200 (Sun, 13 Aug 2006) | 5 lines
Handle a whole lot of failures from PyString_FromInternedString().
Should fix most of Klocwork 234-272.
........
r51247 | neal.norwitz | 2006-08-13 20:10:47 +0200 (Sun, 13 Aug 2006) | 8 lines
cpathname could be NULL if it was longer than MAXPATHLEN. Don't try
to write the .pyc to NULL.
Check results of PyList_GetItem() and PyModule_GetDict() are not NULL.
Klocwork 282, 283, 285
........
r51248 | neal.norwitz | 2006-08-13 20:11:08 +0200 (Sun, 13 Aug 2006) | 6 lines
Fix segfault when doing string formatting on subclasses of long if
__oct__, __hex__ don't return a string.
Klocwork 308
........
r51250 | neal.norwitz | 2006-08-13 20:11:27 +0200 (Sun, 13 Aug 2006) | 5 lines
Check return result of PyModule_GetDict().
Fix a bunch of refleaks in the init of the module. This would only be found
when running python -v.
........
r51251 | neal.norwitz | 2006-08-13 20:11:43 +0200 (Sun, 13 Aug 2006) | 5 lines
Handle malloc and fopen failures more gracefully.
Klocwork 180-181
........
r51252 | neal.norwitz | 2006-08-13 20:12:03 +0200 (Sun, 13 Aug 2006) | 7 lines
It's very unlikely, though possible that source is not a string. Verify
that PyString_AsString() returns a valid pointer. (The problem can
arise when zlib.decompress doesn't return a string.)
Klocwork 346
........
r51253 | neal.norwitz | 2006-08-13 20:12:26 +0200 (Sun, 13 Aug 2006) | 5 lines
Handle failures from lookup.
Klocwork 341-342
........
r51254 | neal.norwitz | 2006-08-13 20:12:45 +0200 (Sun, 13 Aug 2006) | 6 lines
Handle failure from PyModule_GetDict() (Klocwork 208).
Fix a bunch of refleaks in the init of the module. This would only be found
when running python -v.
........
r51255 | neal.norwitz | 2006-08-13 20:13:02 +0200 (Sun, 13 Aug 2006) | 4 lines
Really address the issue of where to place the assert for leftblock.
(Followup of Klocwork 274)
........
r51256 | neal.norwitz | 2006-08-13 20:13:36 +0200 (Sun, 13 Aug 2006) | 4 lines
Handle malloc failure.
Klocwork 281
........
r51258 | neal.norwitz | 2006-08-13 20:40:39 +0200 (Sun, 13 Aug 2006) | 4 lines
Handle alloca failures.
Klocwork 225-228
........
r51259 | neal.norwitz | 2006-08-13 20:41:15 +0200 (Sun, 13 Aug 2006) | 1 line
Get rid of compiler warning
........
r51261 | neal.norwitz | 2006-08-14 02:51:15 +0200 (Mon, 14 Aug 2006) | 1 line
Ignore pgen.exe and kill_python.exe for cygwin
........
r51262 | neal.norwitz | 2006-08-14 02:59:03 +0200 (Mon, 14 Aug 2006) | 4 lines
Can't return NULL from a void function. If there is a memory error,
about the best we can do is call PyErr_WriteUnraisable and go on.
We won't be able to do the call below either, so verify delstr is valid.
........
r51263 | neal.norwitz | 2006-08-14 03:49:54 +0200 (Mon, 14 Aug 2006) | 1 line
Update purify doc some.
........
r51264 | thomas.heller | 2006-08-14 09:13:05 +0200 (Mon, 14 Aug 2006) | 2 lines
Remove unused, buggy test function.
Fixes klockwork issue #207.
........
r51265 | thomas.heller | 2006-08-14 09:14:09 +0200 (Mon, 14 Aug 2006) | 2 lines
Check for NULL return value from new_CArgObject().
Fixes klockwork issues #183, #184, #185.
........
r51266 | thomas.heller | 2006-08-14 09:50:14 +0200 (Mon, 14 Aug 2006) | 2 lines
Check for NULL return value of GenericCData_new().
Fixes klockwork issues #188, #189.
........
r51274 | thomas.heller | 2006-08-14 12:02:24 +0200 (Mon, 14 Aug 2006) | 2 lines
Revert the change that tries to zero out a closure's result storage
area because the size if unknown in source/callproc.c.
........
r51276 | marc-andre.lemburg | 2006-08-14 12:55:19 +0200 (Mon, 14 Aug 2006) | 11 lines
Slightly revised version of patch #1538956:
Replace UnicodeDecodeErrors raised during == and !=
compares of Unicode and other objects with a new
UnicodeWarning.
All other comparisons continue to raise exceptions.
Exceptions other than UnicodeDecodeErrors are also left
untouched.
........
r51277 | thomas.heller | 2006-08-14 13:17:48 +0200 (Mon, 14 Aug 2006) | 13 lines
Apply the patch #1532975 plus ideas from the patch #1533481.
ctypes instances no longer have the internal and undocumented
'_as_parameter_' attribute which was used to adapt them to foreign
function calls; this mechanism is replaced by a function pointer in
the type's stgdict.
In the 'from_param' class methods, try the _as_parameter_ attribute if
other conversions are not possible.
This makes the documented _as_parameter_ mechanism work as intended.
Change the ctypes version number to 1.0.1.
........
r51278 | marc-andre.lemburg | 2006-08-14 13:44:34 +0200 (Mon, 14 Aug 2006) | 3 lines
Readd NEWS items that were accidentally removed by r51276.
........
r51279 | georg.brandl | 2006-08-14 14:36:06 +0200 (Mon, 14 Aug 2006) | 3 lines
Improve markup in PyUnicode_RichCompare.
........
r51280 | marc-andre.lemburg | 2006-08-14 14:57:27 +0200 (Mon, 14 Aug 2006) | 3 lines
Correct an accidentally removed previous patch.
........
r51281 | thomas.heller | 2006-08-14 18:17:41 +0200 (Mon, 14 Aug 2006) | 3 lines
Patch #1536908: Add support for AMD64 / OpenBSD.
Remove the -no-stack-protector compiler flag for OpenBSD
as it has been reported to be unneeded.
........
r51282 | thomas.heller | 2006-08-14 18:20:04 +0200 (Mon, 14 Aug 2006) | 1 line
News item for rev 51281.
........
r51283 | georg.brandl | 2006-08-14 22:25:39 +0200 (Mon, 14 Aug 2006) | 3 lines
Fix refleak introduced in rev. 51248.
........
r51284 | georg.brandl | 2006-08-14 23:34:08 +0200 (Mon, 14 Aug 2006) | 5 lines
Make tabnanny recognize IndentationErrors raised by tokenize.
Add a test to test_inspect to make sure indented source
is recognized correctly. (fixes#1224621)
........
r51285 | georg.brandl | 2006-08-14 23:42:55 +0200 (Mon, 14 Aug 2006) | 3 lines
Patch #1535500: fix segfault in BZ2File.writelines and make sure it
raises the correct exceptions.
........
r51287 | georg.brandl | 2006-08-14 23:45:32 +0200 (Mon, 14 Aug 2006) | 3 lines
Add an additional test: BZ2File write methods should raise IOError
when file is read-only.
........
r51289 | georg.brandl | 2006-08-14 23:55:28 +0200 (Mon, 14 Aug 2006) | 3 lines
Patch #1536071: trace.py should now find the full module name of a
file correctly even on Windows.
........
r51290 | georg.brandl | 2006-08-15 00:01:24 +0200 (Tue, 15 Aug 2006) | 3 lines
Cookie.py shouldn't "bogusly" use string._idmap.
........
r51291 | georg.brandl | 2006-08-15 00:10:24 +0200 (Tue, 15 Aug 2006) | 3 lines
Patch #1511317: don't crash on invalid hostname info
........
r51292 | tim.peters | 2006-08-15 02:25:04 +0200 (Tue, 15 Aug 2006) | 2 lines
Whitespace normalization.
........
r51293 | neal.norwitz | 2006-08-15 06:14:57 +0200 (Tue, 15 Aug 2006) | 3 lines
Georg fixed one of my bugs, so I'll repay him with 2 NEWS entries.
Now we're even. :-)
........
r51295 | neal.norwitz | 2006-08-15 06:58:28 +0200 (Tue, 15 Aug 2006) | 8 lines
Fix the test for SocketServer so it should pass on cygwin and not fail
sporadically on other platforms. This is really a band-aid that doesn't
fix the underlying issue in SocketServer. It's not clear if it's worth
it to fix SocketServer, however, I opened a bug to track it:
http://python.org/sf/1540386
........
r51296 | neal.norwitz | 2006-08-15 06:59:30 +0200 (Tue, 15 Aug 2006) | 3 lines
Update the docstring to use a version a little newer than 1999. This was
taken from a Debian patch. Should we update the version for each release?
........
r51298 | neal.norwitz | 2006-08-15 08:29:03 +0200 (Tue, 15 Aug 2006) | 2 lines
Subclasses of int/long are allowed to define an __index__.
........
r51300 | thomas.heller | 2006-08-15 15:07:21 +0200 (Tue, 15 Aug 2006) | 1 line
Check for NULL return value from new_CArgObject calls.
........
r51303 | kurt.kaiser | 2006-08-16 05:15:26 +0200 (Wed, 16 Aug 2006) | 2 lines
The 'with' statement is now a Code Context block opener
........
r51304 | anthony.baxter | 2006-08-16 05:42:26 +0200 (Wed, 16 Aug 2006) | 1 line
preparing for 2.5c1
........
r51305 | anthony.baxter | 2006-08-16 05:58:37 +0200 (Wed, 16 Aug 2006) | 1 line
preparing for 2.5c1 - no, really this time
........
r51306 | kurt.kaiser | 2006-08-16 07:01:42 +0200 (Wed, 16 Aug 2006) | 9 lines
Patch #1540892: site.py Quitter() class attempts to close sys.stdin
before raising SystemExit, allowing IDLE to honor quit() and exit().
M Lib/site.py
M Lib/idlelib/PyShell.py
M Lib/idlelib/CREDITS.txt
M Lib/idlelib/NEWS.txt
M Misc/NEWS
........
r51307 | ka-ping.yee | 2006-08-16 09:02:50 +0200 (Wed, 16 Aug 2006) | 6 lines
Update code and tests to support the 'bytes_le' attribute (for
little-endian byte order on Windows), and to work around clocks
with low resolution yielding duplicate UUIDs.
Anthony Baxter has approved this change.
........
r51308 | kurt.kaiser | 2006-08-16 09:04:17 +0200 (Wed, 16 Aug 2006) | 2 lines
Get quit() and exit() to work cleanly when not using subprocess.
........
r51309 | marc-andre.lemburg | 2006-08-16 10:13:26 +0200 (Wed, 16 Aug 2006) | 2 lines
Revert to having static version numbers again.
........
r51310 | martin.v.loewis | 2006-08-16 14:55:10 +0200 (Wed, 16 Aug 2006) | 2 lines
Build _hashlib on Windows. Build OpenSSL with masm assembler code.
Fixes#1535502.
........
r51311 | thomas.heller | 2006-08-16 15:03:11 +0200 (Wed, 16 Aug 2006) | 6 lines
Add commented assert statements to check that the result of
PyObject_stgdict() and PyType_stgdict() calls are non-NULL before
dereferencing the result. Hopefully this fixes what klocwork is
complaining about.
Fix a few other nits as well.
........
r51312 | anthony.baxter | 2006-08-16 15:08:25 +0200 (Wed, 16 Aug 2006) | 1 line
news entry for 51307
........
r51313 | andrew.kuchling | 2006-08-16 15:22:20 +0200 (Wed, 16 Aug 2006) | 1 line
Add UnicodeWarning
........
r51314 | andrew.kuchling | 2006-08-16 15:41:52 +0200 (Wed, 16 Aug 2006) | 1 line
Bump document version to 1.0; remove pystone paragraph
........
r51315 | andrew.kuchling | 2006-08-16 15:51:32 +0200 (Wed, 16 Aug 2006) | 1 line
Link to docs; remove an XXX comment
........
r51316 | martin.v.loewis | 2006-08-16 15:58:51 +0200 (Wed, 16 Aug 2006) | 1 line
Make cl build step compile-only (/c). Remove libs from source list.
........
r51317 | thomas.heller | 2006-08-16 16:07:44 +0200 (Wed, 16 Aug 2006) | 5 lines
The __repr__ method of a NULL py_object does no longer raise an
exception. Remove a stray '?' character from the exception text
when the value is retrieved of such an object.
Includes tests.
........
r51318 | andrew.kuchling | 2006-08-16 16:18:23 +0200 (Wed, 16 Aug 2006) | 1 line
Update bug/patch counts
........
r51319 | andrew.kuchling | 2006-08-16 16:21:14 +0200 (Wed, 16 Aug 2006) | 1 line
Wording/typo fixes
........
r51320 | thomas.heller | 2006-08-16 17:10:12 +0200 (Wed, 16 Aug 2006) | 9 lines
Remove the special casing of Py_None when converting the return value
of the Python part of a callback function to C. If it cannot be
converted, call PyErr_WriteUnraisable with the exception we got.
Before, arbitrary data has been passed to the calling C code in this
case.
(I'm not really sure the NEWS entry is understandable, but I cannot
find better words)
........
r51321 | marc-andre.lemburg | 2006-08-16 18:11:01 +0200 (Wed, 16 Aug 2006) | 2 lines
Add NEWS item mentioning the reverted distutils version number patch.
........
r51322 | fredrik.lundh | 2006-08-16 18:47:07 +0200 (Wed, 16 Aug 2006) | 5 lines
SF#1534630
ignore data that arrives before the opening start tag
........
r51324 | andrew.kuchling | 2006-08-16 19:11:18 +0200 (Wed, 16 Aug 2006) | 1 line
Grammar fix
........
r51328 | thomas.heller | 2006-08-16 20:02:11 +0200 (Wed, 16 Aug 2006) | 12 lines
Tutorial:
Clarify somewhat how parameters are passed to functions
(especially explain what integer means).
Correct the table - Python integers and longs can both be used.
Further clarification to the table comparing ctypes types, Python
types, and C types.
Reference:
Replace integer by C ``int`` where it makes sense.
........
r51329 | kurt.kaiser | 2006-08-16 23:45:59 +0200 (Wed, 16 Aug 2006) | 8 lines
File menu hotkeys: there were three 'p' assignments. Reassign the
'Save Copy As' and 'Print' hotkeys to 'y' and 't'. Change the
Shell menu hotkey from 's' to 'l'.
M Bindings.py
M PyShell.py
M NEWS.txt
........
r51330 | neil.schemenauer | 2006-08-17 01:38:05 +0200 (Thu, 17 Aug 2006) | 3 lines
Fix a bug in the ``compiler`` package that caused invalid code to be
generated for generator expressions.
........
r51342 | martin.v.loewis | 2006-08-17 21:19:32 +0200 (Thu, 17 Aug 2006) | 3 lines
Merge 51340 and 51341 from 2.5 branch:
Leave tk build directory to restore original path.
Invoke debug mk1mf.pl after running Configure.
........
r51354 | martin.v.loewis | 2006-08-18 05:47:18 +0200 (Fri, 18 Aug 2006) | 3 lines
Bug #1541863: uuid.uuid1 failed to generate unique identifiers
on systems with low clock resolution.
........
r51355 | neal.norwitz | 2006-08-18 05:57:54 +0200 (Fri, 18 Aug 2006) | 1 line
Add template for 2.6 on HEAD
........
r51356 | neal.norwitz | 2006-08-18 06:01:38 +0200 (Fri, 18 Aug 2006) | 1 line
More post-release wibble
........
r51357 | neal.norwitz | 2006-08-18 06:58:33 +0200 (Fri, 18 Aug 2006) | 1 line
Try to get Windows bots working again
........
r51358 | neal.norwitz | 2006-08-18 07:10:00 +0200 (Fri, 18 Aug 2006) | 1 line
Try to get Windows bots working again. Take 2
........
r51359 | neal.norwitz | 2006-08-18 07:39:20 +0200 (Fri, 18 Aug 2006) | 1 line
Try to get Unix bots install working again.
........
r51360 | neal.norwitz | 2006-08-18 07:41:46 +0200 (Fri, 18 Aug 2006) | 1 line
Set version to 2.6a0, seems more consistent.
........
r51362 | neal.norwitz | 2006-08-18 08:14:52 +0200 (Fri, 18 Aug 2006) | 1 line
More version wibble
........
r51364 | georg.brandl | 2006-08-18 09:27:59 +0200 (Fri, 18 Aug 2006) | 4 lines
Bug #1541682: Fix example in the "Refcount details" API docs.
Additionally, remove a faulty example showing PySequence_SetItem applied
to a newly created list object and add notes that this isn't a good idea.
........
r51366 | anthony.baxter | 2006-08-18 09:29:02 +0200 (Fri, 18 Aug 2006) | 3 lines
Updating IDLE's version number to match Python's (as per python-dev
discussion).
........
r51367 | anthony.baxter | 2006-08-18 09:30:07 +0200 (Fri, 18 Aug 2006) | 1 line
RPM specfile updates
........
r51368 | georg.brandl | 2006-08-18 09:35:47 +0200 (Fri, 18 Aug 2006) | 2 lines
Typo in tp_clear docs.
........
r51378 | andrew.kuchling | 2006-08-18 15:57:13 +0200 (Fri, 18 Aug 2006) | 1 line
Minor edits
........
r51379 | thomas.heller | 2006-08-18 16:38:46 +0200 (Fri, 18 Aug 2006) | 6 lines
Add asserts to check for 'impossible' NULL values, with comments.
In one place where I'n not 1000% sure about the non-NULL, raise
a RuntimeError for safety.
This should fix the klocwork issues that Neal sent me. If so,
it should be applied to the release25-maint branch also.
........
r51400 | neal.norwitz | 2006-08-19 06:22:33 +0200 (Sat, 19 Aug 2006) | 5 lines
Move initialization of interned strings to before allocating the
object so we don't leak op. (Fixes an earlier patch to this code)
Klockwork #350
........
r51401 | neal.norwitz | 2006-08-19 06:23:04 +0200 (Sat, 19 Aug 2006) | 4 lines
Move assert to after NULL check, otherwise we deref NULL in the assert.
Klocwork #307
........
r51402 | neal.norwitz | 2006-08-19 06:25:29 +0200 (Sat, 19 Aug 2006) | 2 lines
SF #1542693: Remove semi-colon at end of PyImport_ImportModuleEx macro
........
r51403 | neal.norwitz | 2006-08-19 06:28:55 +0200 (Sat, 19 Aug 2006) | 6 lines
Move initialization to after the asserts for non-NULL values.
Klocwork 286-287.
(I'm not backporting this, but if someone wants to, feel free.)
........
r51404 | neal.norwitz | 2006-08-19 06:52:03 +0200 (Sat, 19 Aug 2006) | 6 lines
Handle PyString_FromInternedString() failing (unlikely, but possible).
Klocwork #325
(I'm not backporting this, but if someone wants to, feel free.)
........
r51416 | georg.brandl | 2006-08-20 15:15:39 +0200 (Sun, 20 Aug 2006) | 2 lines
Patch #1542948: fix urllib2 header casing issue. With new test.
........
r51428 | jeremy.hylton | 2006-08-21 18:19:37 +0200 (Mon, 21 Aug 2006) | 3 lines
Move peephole optimizer to separate file.
........
r51429 | jeremy.hylton | 2006-08-21 18:20:29 +0200 (Mon, 21 Aug 2006) | 2 lines
Move peephole optimizer to separate file. (Forgot .h in previous checkin.)
........
r51432 | neal.norwitz | 2006-08-21 19:59:46 +0200 (Mon, 21 Aug 2006) | 5 lines
Fix bug #1543303, tarfile adds padding that breaks gunzip.
Patch # 1543897.
Will backport to 2.5
........
r51433 | neal.norwitz | 2006-08-21 20:01:30 +0200 (Mon, 21 Aug 2006) | 2 lines
Add assert to make Klocwork happy (#276)
........
(classobject.[ch] aren't empty yet because they also define PyMethod.)
This breaks lots of stuff, notably cPickle. But it's a step in the right
direction. I'll clean it up later.
(Also a few unrelated changes, e.g. T_NONE to define a "struct member"
that is always None, and simplification of __hash__ -- these are unfinished.)
variations of the type struct and its attachments. In Py3k, all type
structs have to have all fields -- no binary backwards compatibility.
Had to change the complex object to a new-style number!
number of tests, all because of the codecs/_multibytecodecs issue described
here (it's not a Py3K issue, just something Py3K discovers):
http://mail.python.org/pipermail/python-dev/2006-April/064051.html
Hye-Shik Chang promised to look for a fix, so no need to fix it here. The
tests that are expected to break are:
test_codecencodings_cn
test_codecencodings_hk
test_codecencodings_jp
test_codecencodings_kr
test_codecencodings_tw
test_codecs
test_multibytecodec
This merge fixes an actual test failure (test_weakref) in this branch,
though, so I believe merging is the right thing to do anyway.
In C++, it's an error to pass a string literal to a char* function
without a const_cast(). Rather than require every C++ extension
module to put a cast around string literals, fix the API to state the
const-ness.
I focused on parts of the API where people usually pass literals:
PyArg_ParseTuple() and friends, Py_BuildValue(), PyMethodDef, the type
slots, etc. Predictably, there were a large set of functions that
needed to be fixed as a result of these changes. The most pervasive
change was to make the keyword args list passed to
PyArg_ParseTupleAndKewords() to be a const char *kwlist[].
One cast was required as a result of the changes: A type object
mallocs the memory for its tp_doc slot and later frees it.
PyTypeObject says that tp_doc is const char *; but if the type was
created by type_new(), we know it is safe to cast to char *.
Rewrote Py_RETURN_{NONE, TRUE, FALSE} to expand to comma expressions
rather than "do {} while(0)" thingies. The OP complained because he
likes using MS /W4 sometimes, and then all his uses of these things
generate nuisance warnings about testing a constant expression (in
the "while(0)" part). Comma expressions don't have this problem
(although it's a lucky accident that comma expressions suffice for these
macros!).
the purpose. Increased my claim to two bits, hoping that nobody
will complain about it. I'm taking the highest two bits, whatever
the integer word size may be.
new line.
New pvt API function _Py_PrintReferenceAddresses(): Prints only the
addresses and refcnts of the live objects. This is always safe to call,
because it has no dependence on Python's C API.
Py_Finalize(): If envar PYTHONDUMPREFS is set, call (the new)
_Py_PrintReferenceAddresses() right before dumping final pymalloc stats.
We can't print the reprs of the objects here because too much of the
interpreter has been shut down. You need to correlate the addresses
displayed here with the object reprs printed by the earlier
PYTHONDUMPREFS call to _Py_PrintReferences().
Arranged that all the objects exposed by __builtin__ appear in the list
of all objects. I basically peed away two days tracking down a mystery
leak in sys.gettotalrefcount() in a ZODB app (== tons of code), because
the object leaking the references didn't appear in the sys.getobjects(0)
list. The object happened to be False. Now False is in the list, along
with other popular & previously missing leak candidates (like None).
Alas, we still don't have a choke point covering *all* Python objects,
so the list of all objects may still be incomplete.
_Py_AddToAllObjects() that simply inserts an object at the front of
the doubly-linked list of all objects. Changed PyType_Ready() (the
closest thing we've got to a choke point for type objects) to call
that.
Py_Init crash". refchain cannot be cleared because objects can live across
Py_Finalize() and Py_Initialize() if they are kept alive by circular
references.
a lot of work: it had to save and restore the current exception around
a call to lookup_maybe(), because that could fail in rare cases, and
most objects don't have a __del__ method, so the whole exercise was
usually a waste of time. Changed this to cache the __del__ method in
the type object just like all other special methods, in a new slot
tp_del. So now subtype_dealloc() can test whether tp_del is NULL and
skip the whole exercise if it is. The new slot doesn't need a new
flag bit: subtype_dealloc() is only called if the type was dynamically
allocated by type_new(), so it's guaranteed to have all current slots.
Types defined in C cannot fill in tp_del with a function of their own,
so there's no corresponding "wrapper". (That functionality is already
available through tp_dealloc.)
The staticforward define was needed to support certain broken C
compilers (notably SCO ODT 3.0, perhaps early AIX as well) botched the
static keyword when it was used with a forward declaration of a static
initialized structure. Standard C allows the forward declaration with
static, and we've decided to stop catering to broken C compilers. (In
fact, we expect that the compilers are all fixed eight years later.)
I'm leaving staticforward and statichere defined in object.h as
static. This is only for backwards compatibility with C extensions
that might still use it.
XXX I haven't updated the documentation.
helper macros to something saner, and used them appropriately in other
files too, to reduce #ifdef blocks.
classobject.c, instance_dealloc(): One of my worst Python Memories is
trying to fix this routine a few years ago when COUNT_ALLOCS was defined
but Py_TRACE_REFS wasn't. The special-build code here is way too
complicated. Now it's much simpler. Difference: in a Py_TRACE_REFS
build, the instance is no longer in the doubly-linked list of live
objects while its __del__ method is executing, and that may be visible
via sys.getobjects() called from a __del__ method. Tough -- the object
is presumed dead while its __del__ is executing anyway, and not calling
_Py_NewReference() at the start allows enormous code simplification.
typeobject.c, call_finalizer(): The special-build instance_dealloc()
pain apparently spread to here too via cut-'n-paste, and this is much
simpler now too. In addition, I didn't understand why this routine
was calling _PyObject_GC_TRACK() after a resurrection, since there's no
plausible way _PyObject_GC_UNTRACK() could have been called on the
object by this point. I suspect it was left over from pasting the
instance_delloc() code. Instead asserted that the object is still
tracked. Caution: I suspect we don't have a test that actually
exercises the subtype_dealloc() __del__-resurrected-me code.
that their uses can be prettier. I've come to despise the names I picked
for these things, though, and expect to change all of them -- I changed
a bunch of other files to use them (replacing #ifdef blocks), but the
names were so obscure out of context that I backed that all out again.
more trivial lexical helper macros so that uses of these guys expand
to nothing at all when they're not enabled. This should help sub-
standard compilers that can't do a good job of optimizing away the
previous "(void)0" expressions.
Py_DECREF: There's only one definition of this now. Yay! That
was that last one in the family defined multiple times in an #ifdef
maze.
Py_FatalError(): Changed the char* signature to const char*.
_Py_NegativeRefcount(): New helper function for the Py_REF_DEBUG
expansion of Py_DECREF. Calling an external function cuts down on
the volume of generated code. The previous inline expansion of abort()
didn't work as intended on Windows (the program often kept going, and
the error msg scrolled off the screen unseen). _Py_NegativeRefcount
calls Py_FatalError instead, which captures our best knowledge of
how to abort effectively across platforms.
that have taken me "too long" to reverse-engineer over the years.
Vastly reduced the nesting level and redundancy of #ifdef-ery.
Took a light stab at repairing comments that are no longer true.
sys_gettotalrefcount(): Changed to enable under Py_REF_DEBUG.
It was enabled under Py_TRACE_REFS, which was much heavier than
necessary. sys.gettotalrefcount() is now available in a
Py_REF_DEBUG-only build.
mechanism is no longer evil: it no longer plays dangerous games with
the type pointer or refcounts, and objects in extension modules can play
along too without needing to edit the core first.
Rewrote all the comments to explain this, and (I hope) give clear
guidance to extension authors who do want to play along. Documented
all the functions. Added more asserts (it may no longer be evil, but
it's still dangerous <0.9 wink>). Rearranged the generated code to
make it clearer, and to tolerate either the presence or absence of a
semicolon after the macros. Rewrote _PyTrash_destroy_chain() to call
tp_dealloc directly; it was doing a Py_DECREF again, and that has all
sorts of obscure distorting effects in non-release builds (Py_DECREF
was already called on the object!). Removed Christian's little "embedded
change log" comments -- that's what checkin messages are for, and since
it was impossible to correlate the comments with the code that changed,
I found them merely distracting.
PEP 285. Everything described in the PEP is here, and there is even
some documentation. I had to fix 12 unit tests; all but one of these
were printing Boolean outcomes that changed from 0/1 to False/True.
(The exception is test_unicode.py, which did a type(x) == type(y)
style comparison. I could've fixed that with a single line using
issubtype(x, type(y)), but instead chose to be explicit about those
places where a bool is expected.
Still to do: perhaps more documentation; change standard library
modules to return False/True from predicates.
type.__module__ behavior.
This adds the module name and a dot in front of the type name in every
type object initializer, except for built-in types (and those that
already had this). Note that it touches lots of Mac modules -- I have
no way to test these but the changes look right. Apologies if they're
not. This also touches the weakref docs, which contains a sample type
object initializer. It also touches the mmap test output, because the
mmap type's repr is included in that output. It touches object.h to
put the correct description in a comment.
object.h: Added PyType_CheckExact macro.
typeobject.c, type_new():
+ Use the new macro.
+ Assert that the arguments have the right types rather than do incomplete
runtime checks "sometimes".
+ If this isn't the 1-argument flavor() of type, and there aren't 3 args
total, produce a "types() takes 1 or 3 args" msg before
PyArg_ParseTupleAndKeywords produces a "takes exactly 3" msg.
is a list of weak references to types (new-style classes). Make this
accessible to Python as the function __subclasses__ which returns a
list of types -- we don't want Python programmers to be able to
manipulate the raw list.
In order to make this possible, I also had to add weak reference
support to type objects.
This will eventually be used together with a trap on attribute
assignment for dynamic classes for a major speed-up without losing the
dynamic properties of types: when a __foo__ method is added to a
class, the class and all its subclasses will get an appropriate tp_foo
slot function.
instances).
Also added GC support to various auxiliary types: super, property,
descriptors, wrappers, dictproxy. (Only type objects have a tp_clear
field; the other types are.)
One change was necessary to the GC infrastructure. We have statically
allocated type objects that don't have a GC header (and can't easily
be given one) and heap-allocated type objects that do have a GC
header. Giving these different metatypes would be really ugly: I
tried, and I had to modify pickle.py, cPickle.c, copy.py, add a new
invent a new name for the new metatype and make it a built-in, change
affected tests... In short, a mess. So instead, we add a new type
slot tp_is_gc, which is a simple Boolean function that determines
whether a particular instance has GC headers or not. This slot is
only relevant for types that have the (new) GC flag bit set. If the
tp_is_gc slot is NULL (by far the most common case), all instances of
the type are deemed to have GC headers. This slot is called by the
PyObject_IS_GC() macro (which is only used twice, both times in
gcmodule.c).
I also changed the extern declarations for a bunch of GC-related
functions (_PyObject_GC_Del etc.): these always exist but objimpl.h
only declared them when WITH_CYCLE_GC was defined, but I needed to be
able to reference them without #ifdefs. (When WITH_CYCLE_GC is not
defined, they do the same as their non-GC counterparts anyway.)
no backwards compatibility to worry about, so I just pushed the
'closure' struct member to the back -- it's never used in the current
code base (I may eliminate it, but that's more work because the getter
and setter signatures would have to change.)
As examples, I added actual docstrings to the getset attributes of a
few types: file.closed, xxsubtype.spamdict.state.
compatibility, this required all places where an array of "struct
memberlist" structures was declared that is referenced from a type's
tp_members slot to change the type of the structure to PyMemberDef;
"struct memberlist" is now only used by old code that still calls
PyMember_Get/Set. The code in PyObject_GenericGetAttr/SetAttr now
calls the new APIs PyMember_GetOne/SetOne, which take a PyMemberDef
argument.
As examples, I added actual docstrings to the attributes of a few
types: file, complex, instance method, super, and xxsubtype.spamlist.
Also converted the symtable to new style getattr.
hack, and it's even more disgusting than a PyInstance_Check() call.
If the tp_compare slot is the slot used for overrides in Python,
it's always called.
Add some tests that show what should work too.
- Do not compile unicodeobject, unicodectype, and unicodedata if Unicode is disabled
- check for Py_USING_UNICODE in all places that use Unicode functions
- disables unicode literals, and the builtin functions
- add the types.StringTypes list
- remove Unicode literals from most tests.
This introduces:
- A new operator // that means floor division (the kind of division
where 1/2 is 0).
- The "future division" statement ("from __future__ import division)
which changes the meaning of the / operator to implement "true
division" (where 1/2 is 0.5).
- New overloadable operators __truediv__ and __floordiv__.
- New slots in the PyNumberMethods struct for true and floor division,
new abstract APIs for them, new opcodes, and so on.
I emphasize that without the future division statement, the semantics
of / will remain unchanged until Python 3.0.
Not yet implemented are warnings (default off) when / is used with int
or long arguments.
This has been on display since 7/31 as SF patch #443474.
Flames to /dev/null.
- Add an explicit call to PyType_Ready(&PyList_Type) to pythonrun.c
(just for the heck of it, really -- we should either explicitly
ready all types, or none).
sees it (test_iter.py is unchanged).
- Added a tp_iternext slot, which calls the iterator's next() method;
this is much faster for built-in iterators over built-in types
such as lists and dicts, speeding up pybench's ForLoop with about
25% compared to Python 2.1. (Now there's a good argument for
iterators. ;-)
- Renamed the built-in sequence iterator SeqIter, affecting the C API
functions for it. (This frees up the PyIter prefix for generic
iterator operations.)
- Added PyIter_Check(obj), which checks that obj's type has a
tp_iternext slot and that the proper feature flag is set.
- Added PyIter_Next(obj) which calls the tp_iternext slot. It has a
somewhat complex return condition due to the need for speed: when it
returns NULL, it may not have set an exception condition, meaning
the iterator is exhausted; when the exception StopIteration is set
(or a derived exception class), it means the same thing; any other
exception means some other error occurred.
new slot tp_iter in type object, plus new flag Py_TPFLAGS_HAVE_ITER
new C API PyObject_GetIter(), calls tp_iter
new builtin iter(), with two forms: iter(obj), and iter(function, sentinel)
new internal object types iterobject and calliterobject
new exception StopIteration
new opcodes for "for" loops, GET_ITER and FOR_ITER (also supported by dis.py)
new magic number for .pyc files
new special method for instances: __iter__() returns an iterator
iteration over dictionaries: "for x in dict" iterates over the keys
iteration over files: "for x in file" iterates over lines
TODO:
documentation
test suite
decide whether to use a different way to spell iter(function, sentinal)
decide whether "for key in dict" is a good idea
use iterators in map/filter/reduce, min/max, and elsewhere (in/not in?)
speed tuning (make next() a slot tp_next???)
discussion on python-dev. 'from mod import *' is still banned except
at the module level.
Fix value for special NOOPT entry in symtable. Initialze to 0 instead
of None, so that later uses of PyInt_AS_LONG() are valid. (Bug
reported by Donn Cave.)
replace local REPR macros with PyObject_REPR in object.h
except that it always returns Unicode objects.
A new C API PyObject_Unicode() is also provided.
This closes patch #101664.
Written by Marc-Andre Lemburg. Copyright assigned to Guido van Rossum.
It's hard to sort out what the bug was, exactly. So, Big Hammer:
1. Python shouldn't be in the business of #define'ing NULL, period.
2. Users of the Python C API shouldn't be in the business of not including
Python.h, period.
Hence:
1. Removed all #define's of NULL in Python source code (pyport.h and
object.h).
2. Since we're *relying* on stdio.h defining NULL, put an #error in
Python.h after its #include of stdio.h if NULL isn't defined then.
comments, docstrings or error messages. I fixed two minor things in
test_winreg.py ("didn't" -> "Didn't" and "Didnt" -> "Didn't").
There is a minor style issue involved: Guido seems to have preferred English
grammar (behaviour, honour) in a couple places. This patch changes that to
American, which is the more prominent style in the source. I prefer English
myself, so if English is preferred, I'd be happy to supply a patch myself ;)
Stein -- thanks!). Incidentally removed all the Py_PROTO macros
from object.h, as they prevented my editor from magically finding
the definitions of the "coercion", "cmpfunc" and "reprfunc"
typedefs that were being redundantly applied in longobject.c.
errors in some of the hash algorithms. For exmaple, in float_hash and
complex_hash a certain part of the value is not included in the hash
calculation. See Tim's, Guido's, and my discussion of this on
python-dev in May under the title "fix float_hash and complex_hash for
64-bit *nix"
(2) The hash algorithms that use pointers (e.g. func_hash, code_hash)
are universally not correct on Win64 (they assume that sizeof(long) ==
sizeof(void*))
As well, this patch significantly cleans up the hash code. It adds the
two function _Py_HashDouble and _PyHash_VoidPtr that the various
hashing routine are changed to use.
These help maintain the hash function invariant: (a==b) =>
(hash(a)==hash(b))) I have added Lib/test/test_hash.py and
Lib/test/output/test_hash to test this for some cases.
This patch modifies the type structures of objects that
participate in GC. The object's tp_basicsize is increased when
GC is enabled. GC information is prefixed to the object to
maintain binary compatibility. GC objects also define the
tp_flag Py_TPFLAGS_GC.
Improvements:
- does no longer need any extra memory
- has no relationship to tstate
- works in debug mode
- can easily be modified for free threading (hi Greg:)
Side effects:
Trashcan does change the order of object destruction.
Prevending that would be quite an immense effort, as
my attempts have shown. This version works always
the same, with debug mode or not. The slightly
changed destruction order should therefore be no problem.
Algorithm:
While the old idea of delaying the destruction of some
obejcts at a certain recursion level was kept, we now
no longer aloocate an object to hold these objects.
The delayed objects are instead chained together
via their ob_type field. The type is encoded via
ob_refcnt. When it comes to the destruction of the
chain of waiting objects, the topmost object is popped
off the chain and revived with type and refcount 1,
then it gets a normal Py_DECREF.
I am confident that this solution is near optimum
for minimizing side effects and code bloat.
/* More standard operations (at end for binary compatibility) */
should now be:
/* More standard operations (here for binary compatibility) */
since they're no longer at the end!
Jack Diederich
Neal Norwitz
Guido van Rossum
Thomas Wouters
Martin v. Löwis
Georg Brandl
Jeremy Hylton
Neil Schemenauer
Tim Peters
Raymond Hettinger
Jim Fulton
Thomas Heller
Alex Martelli
Brett Cannon
Christian Tismer
Mark Hammond
Fred Drake
Barry Warsaw
Marc-André Lemburg
Greg Stein