Mailman 3 March 2005 - Python-checkins

python/dist/src/Misc NEWS,1.1280,1.1281
by doerwalter＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Misc In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv27988/Misc Modified Files: NEWS Log Message: Since PyPI only accepts UTF-8 encoded data now, make sure that the data is properly encoded and include the encoding in the Content-Type header. Index: NEWS =================================================================== RCS file: /cvsroot/python/python/dist/src/Misc/NEWS,v retrieving revision 1.1280 retrieving revision 1.1281 diff -u -d -r1.1280 -r1.1281 --- NEWS 31 Mar 2005 09:35:44 -0000 1.1280 +++ NEWS 31 Mar 2005 13:57:35 -0000 1.1281 @@ -102,6 +102,9 @@ - distutils.commands.upload was added to support uploading distribution files to PyPI. +- distutils.commands.register now encodes the data as UTF-8 before posting + them to PyPI. + - decimal operator and comparison methods now return NotImplemented instead of raising a TypeError when interacting with other types. This allows other classes to implement __radd__ style methods and have them

1 0

python/dist/src/Misc NEWS, 1.1193.2.45, 1.1193.2.46
by mwh＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Misc In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv8541 Modified Files: Tag: release24-maint NEWS Log Message: I meant to check this in with the last check in (CVS disturbs me sometimes) Index: NEWS =================================================================== RCS file: /cvsroot/python/python/dist/src/Misc/NEWS,v retrieving revision 1.1193.2.45 retrieving revision 1.1193.2.46 diff -u -d -r1.1193.2.45 -r1.1193.2.46 --- NEWS 31 Mar 2005 10:20:34 -0000 1.1193.2.45 +++ NEWS 31 Mar 2005 10:28:30 -0000 1.1193.2.46 @@ -7,6 +7,14 @@ What's New in Python 2.4.2a =========================== +*Release date: XX-XXX-2005* + +Core and builtins +----------------- + +- Bug #1165306: instancemethod_new allowed the creation of a method + with im_class == im_self == NULL, which caused a crash when called. + Extension Modules -----------------

1 0

python/dist/src/Lib/test test_new.py, 1.19, 1.19.2.1
by mwh＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Lib/test In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv5596/Lib/test Modified Files: Tag: release24-maint test_new.py Log Message: Backport: Fix for rather inaccurately titled bug [ 1165306 ] Property access with decorator makes interpreter crash Don't allow the creation of unbound methods with NULL im_class, because attempting to call such crashes. Backport candidate. Index: test_new.py =================================================================== RCS file: /cvsroot/python/python/dist/src/Lib/test/test_new.py,v retrieving revision 1.19 retrieving revision 1.19.2.1 diff -u -d -r1.19 -r1.19.2.1 --- test_new.py 13 Aug 2004 03:57:22 -0000 1.19 +++ test_new.py 31 Mar 2005 10:22:43 -0000 1.19.2.1 @@ -47,6 +47,16 @@ verify(c.get_yolks() == 1 and c.get_more_yolks() == 4, 'Broken call of hand-crafted instance method') +im = new.instancemethod(break_yolks, c) +im() +verify(c.get_yolks() == -1) +try: + new.instancemethod(break_yolks, None) +except TypeError: + pass +else: + raise TestFailed, "dangerous instance method creation allowed" + # It's unclear what the semantics should be for a code object compiled at # module scope, but bound and run in a function. In CPython, `c' is global # (by accident?) while in Jython, `c' is local. The intent of the test

1 0

python/dist/src/Objects classobject.c, 2.176, 2.176.2.1
by mwh＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Objects In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv5596/Objects Modified Files: Tag: release24-maint classobject.c Log Message: Backport: Fix for rather inaccurately titled bug [ 1165306 ] Property access with decorator makes interpreter crash Don't allow the creation of unbound methods with NULL im_class, because attempting to call such crashes. Backport candidate. Index: classobject.c =================================================================== RCS file: /cvsroot/python/python/dist/src/Objects/classobject.c,v retrieving revision 2.176 retrieving revision 2.176.2.1 diff -u -d -r2.176 -r2.176.2.1 --- classobject.c 23 Sep 2004 02:39:36 -0000 2.176 +++ classobject.c 31 Mar 2005 10:22:42 -0000 2.176.2.1 @@ -2208,6 +2208,12 @@ } if (self == Py_None) self = NULL; + if (self == NULL && classObj == NULL) { + PyErr_SetString(PyExc_TypeError, + "unbound methods must have non-NULL im_class"); + return NULL; + } + return PyMethod_New(func, self, classObj); }

1 0

python/dist/src/Modules readline.c, 2.79.2.1, 2.79.2.2
by mwh＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Modules In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv3364/Modules Modified Files: Tag: release24-maint readline.c Log Message: Backport: Fixes for [ 1166660 ] The readline module can cause python to segfault It seems to me that the code I'm rewriting here attempted to call any user-supplied hook functions using the thread state of the thread that called the hook-setting function, as opposed to that of the thread that is currently executing. This doesn't work, in general. Fix this by using the PyGILState API (It wouldn't be that hard to define a dummy version of said API when #ifndef WITH_THREAD, would it?). Also, check the conversion to integer of the return value of a hook function for errors (this problem was mentioned in the ipython bug report linked to in the above bug). Index: readline.c =================================================================== RCS file: /cvsroot/python/python/dist/src/Modules/readline.c,v retrieving revision 2.79.2.1 retrieving revision 2.79.2.2 diff -u -d -r2.79.2.1 -r2.79.2.2 --- readline.c 27 Feb 2005 20:34:01 -0000 2.79.2.1 +++ readline.c 31 Mar 2005 10:20:33 -0000 2.79.2.2 @@ -161,8 +161,7 @@ /* Generic hook function setter */ static PyObject * -set_hook(const char *funcname, PyObject **hook_var, - PyThreadState **tstate, PyObject *args) +set_hook(const char *funcname, PyObject **hook_var, PyObject *args) { PyObject *function = Py_None; char buf[80]; @@ -172,14 +171,12 @@ if (function == Py_None) { Py_XDECREF(*hook_var); *hook_var = NULL; - *tstate = NULL; } else if (PyCallable_Check(function)) { PyObject *tmp = *hook_var; Py_INCREF(function); *hook_var = function; Py_XDECREF(tmp); - *tstate = PyThreadState_GET(); } else { PyOS_snprintf(buf, sizeof(buf), @@ -196,18 +193,15 @@ /* Exported functions to specify hook functions in Python */ static PyObject *startup_hook = NULL; -static PyThreadState *startup_hook_tstate = NULL; #ifdef HAVE_RL_PRE_INPUT_HOOK static PyObject *pre_input_hook = NULL; -static PyThreadState *pre_input_hook_tstate = NULL; #endif static PyObject * set_startup_hook(PyObject *self, PyObject *args) { - return set_hook("startup_hook", &startup_hook, - &startup_hook_tstate, args); + return set_hook("startup_hook", &startup_hook, args); } PyDoc_STRVAR(doc_set_startup_hook, @@ -224,8 +218,7 @@ static PyObject * set_pre_input_hook(PyObject *self, PyObject *args) { - return set_hook("pre_input_hook", &pre_input_hook, - &pre_input_hook_tstate, args); + return set_hook("pre_input_hook", &pre_input_hook, args); } PyDoc_STRVAR(doc_set_pre_input_hook, @@ -241,7 +234,6 @@ /* Exported function to specify a word completer in Python */ static PyObject *completer = NULL; -static PyThreadState *completer_tstate = NULL; static PyObject *begidx = NULL; static PyObject *endidx = NULL; @@ -405,7 +397,7 @@ static PyObject * set_completer(PyObject *self, PyObject *args) { - return set_hook("completer", &completer, &completer_tstate, args); + return set_hook("completer", &completer, args); } PyDoc_STRVAR(doc_set_completer, @@ -586,28 +578,33 @@ /* C function to call the Python hooks. */ static int -on_hook(PyObject *func, PyThreadState **tstate) +on_hook(PyObject *func) { int result = 0; if (func != NULL) { PyObject *r; - /* Note that readline is called with the interpreter - lock released! */ - PyEval_RestoreThread(*tstate); +#ifdef WITH_THREAD + PyGILState_STATE gilstate = PyGILState_Ensure(); +#endif r = PyObject_CallFunction(func, NULL); if (r == NULL) goto error; if (r == Py_None) result = 0; - else + else { result = PyInt_AsLong(r); + if (result == -1 && PyErr_Occurred()) + goto error; + } Py_DECREF(r); goto done; error: PyErr_Clear(); Py_XDECREF(r); done: - *tstate = PyEval_SaveThread(); +#ifdef WITH_THREAD + PyGILState_Release(gilstate); +#endif } return result; } @@ -615,14 +612,14 @@ static int on_startup_hook(void) { - return on_hook(startup_hook, &startup_hook_tstate); + return on_hook(startup_hook); } #ifdef HAVE_RL_PRE_INPUT_HOOK static int on_pre_input_hook(void) { - return on_hook(pre_input_hook, &pre_input_hook_tstate); + return on_hook(pre_input_hook); } #endif @@ -635,11 +632,9 @@ char *result = NULL; if (completer != NULL) { PyObject *r; - /* Note that readline is called with the interpreter - lock released! */ - PyEval_RestoreThread(completer_tstate); - /* Don't use the default filename completion if we - * have a custom completion function... */ +#ifdef WITH_THREAD + PyGILState_STATE gilstate = PyGILState_Ensure(); +#endif rl_attempted_completion_over = 1; r = PyObject_CallFunction(completer, "si", text, state); if (r == NULL) @@ -659,7 +654,9 @@ PyErr_Clear(); Py_XDECREF(r); done: - completer_tstate = PyEval_SaveThread(); +#ifdef WITH_THREAD + PyGILState_Release(gilstate); +#endif } return result; }

1 0

python/dist/src/Misc NEWS, 1.1193.2.44, 1.1193.2.45
by mwh＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Misc In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv3364/Misc Modified Files: Tag: release24-maint NEWS Log Message: Backport: Fixes for [ 1166660 ] The readline module can cause python to segfault It seems to me that the code I'm rewriting here attempted to call any user-supplied hook functions using the thread state of the thread that called the hook-setting function, as opposed to that of the thread that is currently executing. This doesn't work, in general. Fix this by using the PyGILState API (It wouldn't be that hard to define a dummy version of said API when #ifndef WITH_THREAD, would it?). Also, check the conversion to integer of the return value of a hook function for errors (this problem was mentioned in the ipython bug report linked to in the above bug). Index: NEWS =================================================================== RCS file: /cvsroot/python/python/dist/src/Misc/NEWS,v retrieving revision 1.1193.2.44 retrieving revision 1.1193.2.45 diff -u -d -r1.1193.2.44 -r1.1193.2.45 --- NEWS 31 Mar 2005 04:07:48 -0000 1.1193.2.44 +++ NEWS 31 Mar 2005 10:20:34 -0000 1.1193.2.45 @@ -10,6 +10,9 @@ Extension Modules ----------------- +- Bug #1166660: The readline module could segfault if hook functions + were set in a different thread than that which called readline. + - weakref proxy has incorrect __nonzero__ behavior. SF bug #1770766.

1 0

python/dist/src/Misc NEWS,1.1279,1.1280
by mwh＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Misc In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv12420 Modified Files: NEWS Log Message: NEWS entries for the bugs I fixed yesterday. Index: NEWS =================================================================== RCS file: /cvsroot/python/python/dist/src/Misc/NEWS,v retrieving revision 1.1279 retrieving revision 1.1280 diff -u -d -r1.1279 -r1.1280 --- NEWS 29 Mar 2005 13:47:59 -0000 1.1279 +++ NEWS 31 Mar 2005 09:35:44 -0000 1.1280 @@ -12,6 +12,9 @@ Core and builtins ----------------- +- Bug #1165306: instancemethod_new allowed the creation of a method + with im_class == im_self == NULL, which caused a crash when called. + - Move exception finalisation later in the shutdown process - this fixes the crash seen in bug #1165761 @@ -54,6 +57,9 @@ Extension Modules ----------------- +- Bug #1166660: The readline module could segfault if hook functions + were set in a different thread than that which called readline. + - collections.deque objects now support a remove() method. - operator.itemgetter() and operator.attrgetter() now support retrieving

1 0

python/dist/src/Lib/test test_weakref.py, 1.42, 1.42.4.1
by rhettinger＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Lib/test In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv4856/Lib/test Modified Files: Tag: release24-maint test_weakref.py Log Message: SF bug #1770766: weakref proxy has incorrect __nonzero__ behavior. Index: test_weakref.py =================================================================== RCS file: /cvsroot/python/python/dist/src/Lib/test/test_weakref.py,v retrieving revision 1.42 retrieving revision 1.42.4.1 diff -u -d -r1.42 -r1.42.4.1 --- test_weakref.py 8 Jul 2004 01:22:31 -0000 1.42 +++ test_weakref.py 31 Mar 2005 04:07:47 -0000 1.42.4.1 @@ -271,6 +271,12 @@ del f[0] self.assertEqual(f.result, 0) + def test_proxy_bool(self): + # Test clearing of SF bug #1170766 + class List(list): pass + lyst = List() + self.assertEqual(bool(weakref.proxy(lyst)), bool(lyst)) + def test_getweakrefcount(self): o = C() ref1 = weakref.ref(o)

1 0

python/dist/src/Objects weakrefobject.c, 1.20, 1.20.2.1
by rhettinger＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Objects In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv4856/Objects Modified Files: Tag: release24-maint weakrefobject.c Log Message: SF bug #1770766: weakref proxy has incorrect __nonzero__ behavior. Index: weakrefobject.c =================================================================== RCS file: /cvsroot/python/python/dist/src/Objects/weakrefobject.c,v retrieving revision 1.20 retrieving revision 1.20.2.1 diff -u -d -r1.20 -r1.20.2.1 --- weakrefobject.c 30 Oct 2004 23:09:22 -0000 1.20 +++ weakrefobject.c 31 Mar 2005 04:07:55 -0000 1.20.2.1 @@ -505,11 +505,7 @@ PyObject *o = PyWeakref_GET_OBJECT(proxy); if (!proxy_checkref(proxy)) return -1; - if (o->ob_type->tp_as_number && - o->ob_type->tp_as_number->nb_nonzero) - return (*o->ob_type->tp_as_number->nb_nonzero)(o); - else - return 1; + return PyObject_IsTrue(o); } static void

1 0

python/dist/src/Misc NEWS, 1.1193.2.43, 1.1193.2.44
by rhettinger＠users.sourceforge.net 31 Mar '05

31 Mar '05

Update of /cvsroot/python/python/dist/src/Misc In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv4856/Misc Modified Files: Tag: release24-maint NEWS Log Message: SF bug #1770766: weakref proxy has incorrect __nonzero__ behavior. Index: NEWS =================================================================== RCS file: /cvsroot/python/python/dist/src/Misc/NEWS,v retrieving revision 1.1193.2.43 retrieving revision 1.1193.2.44 diff -u -d -r1.1193.2.43 -r1.1193.2.44 --- NEWS 29 Mar 2005 12:38:37 -0000 1.1193.2.43 +++ NEWS 31 Mar 2005 04:07:48 -0000 1.1193.2.44 @@ -4,6 +4,15 @@ (editors: check NEWS.help for information about editing NEWS using ReST.) +What's New in Python 2.4.2a +=========================== + +Extension Modules +----------------- + +- weakref proxy has incorrect __nonzero__ behavior. SF bug #1770766. + + What's New in Python 2.4.1 final? =================================

1 0