Commit Graph

259 Commits

Author SHA1 Message Date
Hye-Shik Chang ff365c931b Get rid of gcc warning. 2004-03-25 16:37:03 +00:00
Phillip J. Eby 91a968af76 Ensure super() lookup of descriptor from classmethod works (SF #743627) 2004-03-25 02:19:34 +00:00
Jim Fulton 8a1a594590 Fixed a bug in object.__reduce_ex__ (reduce_2) when using protocol
2.  Failure to clear the error when attempts to get the __getstate__
  attribute fail caused intermittent errors and odd behavior.
2004-02-08 04:21:26 +00:00
Raymond Hettinger 6a8bbdbe7b Improve argument checking speed. 2003-12-13 15:21:55 +00:00
Raymond Hettinger 8f5cdaa784 * Added a new method flag, METH_COEXIST.
* Used the flag to optimize set.__contains__(), dict.__contains__(),
  dict.__getitem__(), and list.__getitem__().
2003-12-13 11:26:12 +00:00
Tim Peters 0bd743cee1 subtype_dealloc(): Simplified overly contorted retracking logic. With
this change, I think subtype_dealloc is actually a smidgen less obscure
than it was in 2.3 -- we got rid of a negation in an "if" <wink>.
2003-11-13 22:50:00 +00:00
Tim Peters f7f9e9966b subtype_dealloc(): A more complete fix for critical bug 840829 +
expanded the test case with a piece that needs the more-complete fix.

I'll backport this to 2.3 maint.
2003-11-13 21:59:32 +00:00
Tim Peters add09b4149 SF bug 840829: weakref callbacks and gc corrupt memory.
subtype_dealloc():  This left the dying object exposed to gc, so that
if cyclic gc triggered during the weakref callback, gc tried to delete
the dying object a second time.  That's a disaster.  subtype_dealloc()
had a (I hope!) unique problem here, as every normal dealloc routine
untracks the object (from gc) before fiddling with weakrefs etc.  But
subtype_dealloc has obscure technical reasons for re-registering the
dying object with gc (already explained in a large comment block at
the bottom of the function).

The fix amounts to simply refraining from reregistering the dying object
with gc until after the weakref callback (if any) has been called.

This is a critical bug (hard to predict, and causes seemingly random
memory corruption when it occurs).  I'll backport it to 2.3 later.
2003-11-12 20:43:28 +00:00
Walter Dörwald f0dfc7ac5c Fix a bunch of typos in documentation, docstrings and comments.
(From SF patch #810751)
2003-10-20 14:01:56 +00:00
Raymond Hettinger 8ae4689657 Simplify and speedup uses of Py_BuildValue():
* Py_BuildValue("(OOO)",a,b,c)  -->  PyTuple_Pack(3,a,b,c)
* Py_BuildValue("()",a)         -->  PyTuple_New(0)
* Py_BuildValue("O", a)         -->  Py_INCREF(a)
2003-10-12 19:09:37 +00:00
Raymond Hettinger 56bb16f1b3 Use the simpler and faster PyArg_UnpackTuple() instead of
PyArg_ParseTuple() where possible.
2003-10-11 19:32:18 +00:00
Raymond Hettinger f34f2646a1 SF bug #820397: __nonzero__() returns 1/0
Altered to return a PyBool instead of a PyInt.

Backport candidate.
2003-10-11 17:29:04 +00:00
Guido van Rossum 22c3dda1e6 Fix leak introduced by previous typeobject.c checkin. 2003-10-09 03:46:35 +00:00
Guido van Rossum 02c58f865c SF patch #820195 by Wojtek Walczak (gminick at users.sourceforge.net):
make obj.__contains__() returns True/False instead of 1/0.
2003-10-08 21:08:29 +00:00
Michael W. Hudson b2c7de4667 Fix for
[ 784825 ] fix obscure crash in descriptor handling

Should be applied to release23-maint and in all likelyhood
release22-maint, too.

Certainly doesn't apply to release21-maint.
2003-08-15 13:07:47 +00:00
Michael W. Hudson a6a277d831 /* XXX From here until type is allocated, "return NULL" leaks bases! */
Sure looks like it to me! <wink>

When I run the leak2.py script I posted to python-dev, I only see
three reference leaks in all of test_descr.  When I run
test_descr.test_main, I still see 46 leaks.  This clearly demands
posting a yelp to python-dev :-)

This certainly should be applied to release23-maint, and in all
likelyhood release22-maint as well.
2003-08-08 13:57:22 +00:00
Michael W. Hudson e723e453a1 Repair refcounting on error return from type_set_bases.
Include a test case that failed for one of my efforts to repair this.
2003-08-07 14:58:10 +00:00
Jeremy Hylton f75d9fce16 Remove stray comments. 2003-07-16 16:17:57 +00:00
Jeremy Hylton 1c7a0ea056 Remove unnecessary check in tests for slots allowed.
The !PyType_Check(base) check snuck in as part of rev 2.215, but was
unrelated to the SF patch that is mentioned in the checkin comment.
The test is currently unnecessary because base is set to the return
value of best_bases(), which returns a type or NULL.
2003-07-16 16:08:23 +00:00
Jeremy Hylton 3e3159ce6a Require that __nonzero__() return a bool or exactly an int. 2003-06-27 17:38:27 +00:00
Jeremy Hylton 090a3495b3 Check return type of __nonzero__() method.
The language reference says you must return an int or a bool.  This
fix limits the scope of SF bug 759227 (infinite recursion) to
subclasses of int.
2003-06-27 16:46:45 +00:00
Guido van Rossum 59195fdf40 - SF patch 751998 fixes an unwanted side effect of the previous fix
for SF bug 742860 (the next item).
2003-06-13 20:54:40 +00:00
Brett Cannon 10147f7d13 Fixed a comment. 2003-06-11 20:50:33 +00:00
Neal Norwitz e2fdc61004 Fix SF #749831, copy raises SystemError when getstate raises exception 2003-06-08 13:19:58 +00:00
Guido van Rossum 1987c6693b Fix for SF 742911. We now clear the weakrefs *before* calling __del__
or emptying __dict__, just as we do for classic classes.
2003-05-29 14:29:23 +00:00
Tim Peters 3cfe75470d PyType_Ready(): Complain if the type is a base type, and gc'able, and
tp_free is NULL or PyObject_Del at the end.  Because it's a base type
it must call tp_free in its dealloc function, and because it's gc'able
it must not call PyObject_Del.

inherit_slots():  Don't inherit tp_free unless the type and its base
agree about whether they're gc'able.  If the type is gc'able and the
base is not, and the base uses the default PyObject_Del for its
tp_free, give the type PyObject_GC_Del for its tp_free (the appropriate
default for a gc'able type).

cPickle.c:  The Pickler and Unpickler types claim to be base classes
and gc'able, but their dealloc functions didn't call tp_free.
Repaired that.  Also call PyType_Ready() on these typeobjects, so
that the correct (PyObject_GC_Del) default memory-freeing function
gets plugged into these types' tp_free slots.
2003-05-21 21:29:48 +00:00
Brett Cannon be67d87e4d Fixing the previous patch to have the changes be to the proper docstrings. 2003-05-20 02:40:12 +00:00
Brett Cannon 154da9b7e2 Fix docstrings for __(get|set|del)slice__ to mention that negative indices are not supported. 2003-05-20 02:30:04 +00:00
Martin v. Löwis a94568a753 Patch #734231: Update RiscOS support. In particular, correct
riscospath.extsep, and use os.extsep throughout.
2003-05-10 07:36:56 +00:00
Guido van Rossum 636688d470 Improve the message about metatype/metaclass conflicts. 2003-04-23 12:07:22 +00:00
Guido van Rossum 6cc5bb685d Sigh. The crucial change was still missing from the previous
checkin. :-(
2003-04-16 20:01:36 +00:00
Guido van Rossum 76ba09fd81 - super() no longer ignores data descriptors, except __class__. See
the thread started at
  http://mail.python.org/pipermail/python-dev/2003-April/034338.html
2003-04-16 19:40:58 +00:00
Guido van Rossum 19a02ba69d Fix three (!) object leaks in the code for assignment to __bases__. 2003-04-15 22:09:45 +00:00
Guido van Rossum 52b2705e9c Ouch, it's Carlo Verre, not Verre Carlo. 2003-04-15 20:05:10 +00:00
Guido van Rossum 4dcdb78c6f Close off the "Verre Carlo hack" as discussed on python-dev. 2003-04-14 21:46:03 +00:00
Guido van Rossum 2fd02eb80f super_getattro(): kill some dead code; explain a mystery. 2003-04-14 21:20:26 +00:00
Jeremy Hylton d06483c2f6 Missing DECREF. 2003-04-09 21:01:42 +00:00
Raymond Hettinger f394df47fd SF bug #699934: Obscure error message
mwh pointed out that the error message did not
make sense if obtained by rearranging the bases.
2003-04-06 19:13:41 +00:00
Guido van Rossum 8d24ee97df Refactoring: rename update_these_slots() into update_subclasses() and
generalize to take a callback function and a void * data argument.
This might come in handy later... :-)
2003-03-24 23:49:49 +00:00
Tim Peters 7571a0fbcf Improved new Py_TRACE_REFS gimmicks.
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.
2003-03-23 17:52:28 +00:00
Tim Peters bf9b24464e slot_sq_contains(): This leaked a reference to the result of calling
__contains__().

Bugfix candidate.
2003-03-23 05:35:36 +00:00
Tim Peters 36eb4dfb81 Refactored some of the Py_TRACE_REFS code. New private API function
_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.
2003-03-23 03:33:13 +00:00
Raymond Hettinger 83245b5828 SF bug #699934: Obscure error message
Clarify error message for mro conflicts.
2003-03-12 04:25:42 +00:00
Guido van Rossum e5c691abe3 - The extended type structure used for heap types (new-style
classes defined by Python code using a class statement) is now
  exported from object.h as PyHeapTypeObject.  (SF patch #696193.)
2003-03-07 15:13:17 +00:00
Guido van Rossum 036f999669 Implementing the salient parts of __reduce_ex__ in C.
This still falls back to helpers in copy_reg for:
   - pickle protocols < 2
   - calculating the list of slot names (done only once per class)
   - the __newobj__ function (which is used as a token but never called)
2003-02-21 22:02:54 +00:00
Guido van Rossum c53f009f94 Introducing __reduce_ex__, which is called with a protocol number argument
if it exists in preference over __reduce__.  Now Tim can go implement this
in cPickle.c.
2003-02-18 22:05:12 +00:00
Tim Peters 97e5ff555e Removed unreferenced label. 2003-02-18 19:32:50 +00:00
Guido van Rossum 8e80a72be4 The recent changes to super(), in particular supercheck(), broke when
using super() for an instance in a metaclass situation.  Because the
class was a metaclass, the instance was a class, and hence the
PyType_Check() branch was taken.  But this branch didn't apply.  Make
it so that if this branch doesn't apply, the other branch is still
tried.  All tests pass.
2003-02-18 19:22:22 +00:00
Guido van Rossum 298e421453 SF patch #685738 by Michael Stone.
This changes the default __new__ to refuse arguments iff tp_init is the
default __init__ implementation -- thus making it a TypeError when you
try to pass arguments to a constructor if the class doesn't override at
least __init__ or __new__.
2003-02-13 16:30:16 +00:00
Guido van Rossum a89d10edc9 Implement another useful feature for proxies: in super(X, x), x may
now be a proxy for an X instance, as long as issubclass(x.__class__, X).
2003-02-12 03:58:38 +00:00