[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: I agree (and was going back and forth between +0 and -0). -- status: open -> closed ___ Python tracker ___ _

[issue11707] Create C version of functools.cmp_to_key()

Georg Brandl added the comment: I would have to say that this looks hardly a trivial speed patch, and chances are we cannot guarantee 100% behavior compatibility with the pure-Python version. If you disagree with these two points, then I'm okay with it going in. -- _

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: Georg, would you opine on whether this should to into 3.2.1? -- status: closed -> open ___ Python tracker ___ __

[issue11707] Create C version of functools.cmp_to_key()

Filip Gruszczyński added the comment: I see. Thanks :-) -- ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: htt

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: Georg, do you care to opine on whether this should go into 3.2.1? (I don't have any preference). There is a big thread on comp.lang.python where a number of people seem to be very concerned about the efficiency of cmp_to_key(). OTOH, we almost never bac

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: > why keyobject_type needed traverse function? I used to know this but don't any more. It might not be necessary. Most of the types I write all use GC so it may just be habit. -- ___ Python tracker

[issue11707] Create C version of functools.cmp_to_key()

Filip Gruszczyński added the comment: I have a follow-up question: why keyobject_type needed traverse function? From what I read in docs I assumed it is required for GC tracked types. Why was it required here and how it is used? -- ___ Python track

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: > functools_cmp_to_key() doesn't > check that the argument is a callable. I don't think that is necessary; it will fail with a TypeError when the attempt is made to call it. This is the approach we take elsewhere (look at the built-in filter() function f

[issue11707] Create C version of functools.cmp_to_key()

STINNER Victor added the comment: functools_cmp_to_key() doesn't check that the argument is a callable. >>> table=list(range(3)) >>> table.sort(key=functools.cmp_to_key(3)) Traceback (most recent call last): File "", line 1, in TypeError: 'int' object is not callable PyCallable_Check() shou

[issue11707] Create C version of functools.cmp_to_key()

Roundup Robot added the comment: New changeset 76ed6a061ebe by Victor Stinner in branch 'default': Issue #11707: Fix compilation errors with Visual Studio http://hg.python.org/cpython/rev/76ed6a061ebe -- ___ Python tracker

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: Thanks for the patch Filip. Closing this. If anyone wants to lobby the RM for a 3.2 backport, feel free to re-open and assign to the release manager for consideration. -- status: open -> closed ___ Python track

[issue11707] Create C version of functools.cmp_to_key()

Roundup Robot added the comment: New changeset a03fb2fc3ed8 by Raymond Hettinger in branch 'default': Issue #11707: Fast C version of functools.cmp_to_key() http://hg.python.org/cpython/rev/a03fb2fc3ed8 -- nosy: +python-dev ___ Python tracker

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: Attaching cleaned-up version of the patch, making it more like the pure python version: * Made 'obj' a member so it is an accessible field * Accept keyword arguments * Create arg tuple like was done in the Py2.7 code * Create the zero only once * Expanded t

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: Attaching timing code. Results with and without patch (as cleaned-up): 7.1 seconds without patch 3.2 seconds with patch -- Added file: http://bugs.python.org/file21537/time_cmp_to_key.py ___ Python tracker

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: I'm working on cleaning-up (and speeding-up) the patch. I'll post new timings once that's done. -- ___ Python tracker ___

[issue11707] Create C version of functools.cmp_to_key()

Filip Gruszczyński added the comment: Here are some example performance results: def cmp(x, y): return y - x sorted(range(1, 1000), key=cmp_to_key(cmp)) ''' C version: real0m19.994s user0m8.053s sys 0m1.044s Python version: real0m28.825s user0m28.046s sys

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: I was only aiming for Py3.3. If someone wanted to push for a backport to 3.2, it would be up to the release manager to decide whether a performance booster would be worth the risk of introducing a bug in a point release. ISTM that if someone really care

[issue11707] Create C version of functools.cmp_to_key()

Terry J. Reedy added the comment: [Question from python-list] Would a C version only be for 3.3 or backported to 3.2 and 2.7. The rationale for backport to 2.7 is that .cmp_to_key was added to 2.7 to aid transition to 3.x. A faster version would make use in 2.7 more inviting. I understand tha

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: FWIW, the PSF is working towards getting electronic signatures for contributor agreements. In the mean time, a scan or photo would work just fine. -- ___ Python tracker _

[issue11707] Create C version of functools.cmp_to_key()

Filip Gruszczyński added the comment: I worked a little on the tests. Only then I have noticed, that cmp_to_key tests weren't run, so I have encountered some problems, when I finally turned them on. I have added some of my tests, fixed some things in the patch and now it seems to work fine. I

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: Filip, can you please submit a contributor agreement? http://docs.python.org/devguide/coredev.html#sign-a-contributor-agreement -- ___ Python tracker _

[issue11707] Create C version of functools.cmp_to_key()

Changes by Nick Coghlan : -- nosy: +ncoghlan ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python

[issue11707] Create C version of functools.cmp_to_key()

Raymond Hettinger added the comment: Nice work. There are a few nits (dealloc and traverse need to address both python objects in the struct). I will look at the patch in detail when I get a chance (in the next week or so). It would be great if more tests were added (in general, C equivalen

[issue11707] Create C version of functools.cmp_to_key()

Filip Gruszczyński added the comment: Ok, here is patch that keep python implementation of cmp_to_key if C version can't be imported. Thanks again for you help :-) -- Added file: http://bugs.python.org/file21505/11707_2.patch ___ Python tracker

[issue11707] Create C version of functools.cmp_to_key()

Eric Smith added the comment: I haven't had time to look at this in detail, but the concept appears sound. I'll try to devote some time to it, but hopefully someone else on core-mentorship with more familiarity with this code will also be able to review it. One thing: You don't want to delet

[issue11707] Create C version of functools.cmp_to_key()

Filip Gruszczyński added the comment: Here is a patch that passes functools tests. I'll be happy to work on it further. -- keywords: +patch Added file: http://bugs.python.org/file21504/11707.patch ___ Python tracker

[issue11707] Create C version of functools.cmp_to_key()

Changes by Daniel Stutzbach : -- nosy: +stutzbach ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.p

[issue11707] Create C version of functools.cmp_to_key()

Changes by Daniel Urban : -- nosy: +durban ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.o

[issue11707] Create C version of functools.cmp_to_key()

Changes by Filip Gruszczyński : -- nosy: +gruszczy ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.

[issue11707] Create C version of functools.cmp_to_key()

New submission from Raymond Hettinger : For cases where the underlying comparison function is in C (strcoll for example), the pure Python dispatch in cmp_to_key dominates running time. This can be reduced considerably by writing cmp_to_key in C, making its overhead as low or lower than the co