Christian Heimes added the comment:
> * Fix the format nits. There were missing spaces around = and ==.
Oh, I missed the macro above the function. :]
> * Verify that it's faster by compiling python in a release build
It's roughly the same speed. :/ The variation between timeit runs is
quite high on my machine. The new code is neither significant faster nor
noticeable slower.
./python Lib/timeit.py "for i in xrange(100): ''.split(); ''.split(' ',
1); (1,).index(1); (1,).index(1, 0, 1)"
1000 loops, best of 3: 342 usec per loop
The timing distributes about +-15usec with 400+ every now and then.
./python Lib/timeit.py -s "def func(a, b, c=None, d=None, e=None): pass"
"for i in xrange(100): func(a=1, b=2); func(b=1, a=2, d=3, c=4, e=5)"
This benchmark shows a noticeable speedup. It's about 205usec with patch
and 225usec without patch.
Added file: http://bugs.python.org/file9557/trunk_getargs_speedup2.patch
_____________________________________
Tracker <[EMAIL PROTECTED]>
<http://bugs.python.org/issue1691070>
_____________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
