Mailman 3 August 2006 - Python-checkins

r51479 - peps/trunk/pep-0362.txt
by brett.cannon Aug. 22, 2006

Aug. 22, 2006

Author: brett.cannon Date: Tue Aug 22 19:30:16 2006 New Revision: 51479 Modified: peps/trunk/pep-0362.txt Log: Remove Open Issue of whether keyword-only parameters should be kept in a separate set instead of the 'parameters' attribute on Signature objects. Also strengthen the argument for keeping keyword_only as an attribute instead of setting position to None to flag this state. Modified: peps/trunk/pep-0362.txt ============================================================================== --- peps/trunk/pep-0362.txt (original) +++ peps/trunk/pep-0362.txt Tue Aug 22 19:30:16 2006 @@ -112,9 +112,13 @@ that holds a boolean representing whether the parameter is keyword-only or not. -As to whether the keyword-only parameters should be stored in the -``parameters`` attribute of a Signature object or in a set is -discussed in `Open Issues`_. +Nick Coghlan suggested to set 'position' to None to signal that the +argument is keyword-only and thus remove the need for the new +attribute. But that would cause different types to be used in the +attribute that are in no way compatible. It also removes the ability +to know the position number within the signature from the Paramter +object itself. Plus Guido preferred the original approach over this +alternative. Open Issues @@ -131,23 +135,6 @@ needed, and then return the value of ``__signature__``. -Where should keyword-only parameters be stored in the Signature object? ------------------------------------------------------------------------ - -If PEP 3102 [#pep-3102]_ ends up being accepted, there is the -possibility that storing keyword-only parameters in a set instead of -in the ``parameters`` attribute of the Signature object makes more -sense. Since keyword-only parameters do not have any semantic meaning -in terms of their position within the signature, there is no direct -semantic gain in storing it in the parameter list. - -Storing keyword-only paramaters in a set makes it much more explicit -that keyword-only parameters have no inherent order. It does have the -drawback, though, that if one wants to process all parameters at once -they would need to perform extra work to make sure to go through both -the parameter list and set. - - References ==========

1 0

r51477 - python/branches/int_unification/INTBENCH
by martin.v.loewis Aug. 22, 2006

Aug. 22, 2006

Author: martin.v.loewis Date: Tue Aug 22 17:23:54 2006 New Revision: 51477 Added: python/branches/int_unification/INTBENCH Log: Add measurement results. Added: python/branches/int_unification/INTBENCH ============================================================================== --- (empty file) +++ python/branches/int_unification/INTBENCH Tue Aug 22 17:23:54 2006 @@ -0,0 +1,9 @@ +Measurements of _testcapi.profile_int, on an 800MHz G3, OSX 10.4.7 +best of three runs + +r51476 (original p3yk) +Test 1: 1.601978s +Test 2: 1.696623s +Test 3: 1.900683s +Test 4: 5.307155s +

1 0

r51476 - python/branches/int_unification/Modules/_testcapimodule.c
by martin.v.loewis Aug. 22, 2006

Aug. 22, 2006

Author: martin.v.loewis Date: Tue Aug 22 17:19:55 2006 New Revision: 51476 Modified: python/branches/int_unification/Modules/_testcapimodule.c Log: Add int benchmarks. Modified: python/branches/int_unification/Modules/_testcapimodule.c ============================================================================== --- python/branches/int_unification/Modules/_testcapimodule.c (original) +++ python/branches/int_unification/Modules/_testcapimodule.c Tue Aug 22 17:19:55 2006 @@ -713,6 +713,83 @@ Py_RETURN_NONE; } +#ifdef HAVE_GETTIMEOFDAY +/* Profiling of integer performance */ +void print_delta(int test, struct timeval *s, struct timeval *e) +{ + e->tv_sec -= s->tv_sec; + e->tv_usec -= s->tv_usec; + if (e->tv_usec < 0) { + e->tv_sec -=1; + e->tv_usec += 1000000; + } + printf("Test %d: %d.%06ds\n", test, (int)e->tv_sec, e->tv_usec); +} + +static PyObject * +profile_int(PyObject *self, PyObject* args) +{ + int i, k; + struct timeval start, stop; + PyObject *single, **multiple; + + /* Test 1: Allocate and immediately deallocate + many small integers */ + gettimeofday(&start, NULL); + for(k=0; k < 20000; k++) + for(i=0; i < 1000; i++) { + single = PyInt_FromLong(i); + Py_DECREF(single); + } + gettimeofday(&stop, NULL); + print_delta(1, &start, &stop); + + /* Test 2: Allocate and immediately deallocate + many large integers */ + gettimeofday(&start, NULL); + for(k=0; k < 20000; k++) + for(i=0; i < 1000; i++) { + single = PyInt_FromLong(i+1000000); + Py_DECREF(single); + } + gettimeofday(&stop, NULL); + print_delta(2, &start, &stop); + + /* Test 3: Allocate a few integers, then release + them all simultaneously. */ + multiple = malloc(sizeof(PyObject*) * 1000); + gettimeofday(&start, NULL); + for(k=0; k < 20000; k++) { + for(i=0; i < 1000; i++) { + multiple[i] = PyInt_FromLong(i+1000000); + } + for(i=0; i < 1000; i++) { + Py_DECREF(multiple[i]); + } + } + gettimeofday(&stop, NULL); + print_delta(3, &start, &stop); + + /* Test 4: Allocate many integers, then release + them all simultaneously. */ + multiple = malloc(sizeof(PyObject*) * 1000000); + gettimeofday(&start, NULL); + for(k=0; k < 20; k++) { + for(i=0; i < 1000000; i++) { + multiple[i] = PyInt_FromLong(i+1000000); + } + for(i=0; i < 1000000; i++) { + Py_DECREF(multiple[i]); + } + } + gettimeofday(&stop, NULL); + print_delta(4, &start, &stop); + + Py_INCREF(Py_None); + return Py_None; +} +#endif + static PyMethodDef TestMethods[] = { {"raise_exception", raise_exception, METH_VARARGS}, {"test_config", (PyCFunction)test_config, METH_NOARGS}, @@ -751,6 +828,9 @@ #ifdef WITH_THREAD {"_test_thread_state", test_thread_state, METH_VARARGS}, #endif +#ifdef HAVE_GETTIMEOFDAY + {"profile_int", profile_int, METH_NOARGS}, +#endif {NULL, NULL} /* sentinel */ };

1 0

r51475 - sandbox/branches/jim-fix-setuptools-cli
by jim.fulton Aug. 22, 2006

Aug. 22, 2006

Author: jim.fulton Date: Tue Aug 22 17:12:27 2006 New Revision: 51475 Added: sandbox/branches/jim-fix-setuptools-cli/ - copied from r51474, sandbox/trunk/ Log: Created branch to work on setuptools windows script wrapper

1 0

r51474 - python/branches/release25-maint/Modules/mmapmodule.c
by neal.norwitz Aug. 22, 2006

Aug. 22, 2006

Author: neal.norwitz Date: Tue Aug 22 15:57:07 2006 New Revision: 51474 Modified: python/branches/release25-maint/Modules/mmapmodule.c Log: Alexander Belopolsky pointed out that pos is a size_t Modified: python/branches/release25-maint/Modules/mmapmodule.c ============================================================================== --- python/branches/release25-maint/Modules/mmapmodule.c (original) +++ python/branches/release25-maint/Modules/mmapmodule.c Tue Aug 22 15:57:07 2006 @@ -470,7 +470,7 @@ mmap_tell_method(mmap_object *self, PyObject *unused) { CHECK_VALID(NULL); - return PyInt_FromSsize_t(self->pos); + return PyInt_FromSize_t(self->pos); } static PyObject *

1 0

r51473 - python/trunk/Modules/mmapmodule.c
by neal.norwitz Aug. 22, 2006

Aug. 22, 2006

Author: neal.norwitz Date: Tue Aug 22 15:56:56 2006 New Revision: 51473 Modified: python/trunk/Modules/mmapmodule.c Log: Alexander Belopolsky pointed out that pos is a size_t Modified: python/trunk/Modules/mmapmodule.c ============================================================================== --- python/trunk/Modules/mmapmodule.c (original) +++ python/trunk/Modules/mmapmodule.c Tue Aug 22 15:56:56 2006 @@ -470,7 +470,7 @@ mmap_tell_method(mmap_object *self, PyObject *unused) { CHECK_VALID(NULL); - return PyInt_FromSsize_t(self->pos); + return PyInt_FromSize_t(self->pos); } static PyObject *

1 0

Re: [Python-checkins] r51438 - in python/branches/p3yk: Doc/api/abstract.tex Doc/api/concrete.tex Doc/api/exceptions.tex Doc/api/init.tex Doc/api/intro.tex Doc/api/newtypes.tex Doc/commontex/boilerplate.tex Doc/dist/dist.tex Doc/lib/libctypes.tex Doc
by Thomas Wouters Aug. 22, 2006

Aug. 22, 2006

On 8/22/06, Nick Coghlan <ncoghlan(a)gmail.com> wrote: > > The error argument for PyNumber_AsSsize_t is only used for > overflow handling (that's why it's always either IndexError or > OverflowError > in the core). The TypeError for non-integers (like 1.0) is raised > separately > (even if you set the error argument to NULL to request clipping). Thanks, fixed. -- Thomas Wouters <thomas(a)python.org> Hi! I'm a .signature virus! copy me into your .signature file to help me spread!

1 0

r51471 - sandbox/trunk/trackers/trackers.txt
by andrew.kuchling Aug. 22, 2006

Aug. 22, 2006

Author: andrew.kuchling Date: Tue Aug 22 14:37:46 2006 New Revision: 51471 Added: sandbox/trunk/trackers/trackers.txt Log: Record default login info for future reference Added: sandbox/trunk/trackers/trackers.txt ============================================================================== --- (empty file) +++ sandbox/trunk/trackers/trackers.txt Tue Aug 22 14:37:46 2006 @@ -0,0 +1,45 @@ + +Jira +========================= + +You can find the JIRA test instance at: + + http://jira.python.atlassian.com/secure/Dashboard.jspa + +All the infrastructure committee members have admin-level accounts in +the form "first initial + lastname". + + akuchling + bcannon + bwarsaw + mloewis + rjones + twouters + +You can have your password emailed to you by visiting the forgot +password page and entering your username: + + http://jira.python.atlassian.com/secure/ForgotPassword!default.jspa + + +Launchpad +========================= + +Launchpad uses email addresses to identify the user when logging in (a +user can use any of their validated addresses to log in with). For +the import, we used the $USER(a)users.sourceforge.net email addresses. +I outlined the use of the password recovery on the following wiki +page: http://wiki.python.org/moin/LaunchpadTracker + +Note that we've got an outgoing email filter in place for the demo +instance, so people not listed on that page (currently just the +infrastructure committee) won't receive any email. + +To make things a little easier for you though, I went through the list +of Python developers and did a quick search to see if they had +Launchpad accounts already. For those I noticed, I linked the SF +email address to the existing account. So you should be able to log +into the demo instance using your existing "amk" account with your +existing password. + +

1 0

r51470 - sandbox/trunk/trackers
by andrew.kuchling Aug. 22, 2006

Aug. 22, 2006

Author: andrew.kuchling Date: Tue Aug 22 14:37:29 2006 New Revision: 51470 Added: sandbox/trunk/trackers/ Log: Add tracker info

1 0