[issue25596] regular files handled as directories in the glob module

[Xavier de Gaye]

Xavier de Gaye added the comment:

FWIW I have followed the idea of having _iterdir() yielding the DirEntry entry instead of the name in Serhiy's patch.  There is a slight performance gain. Now _glob0() and _glob1() do return a list of directories when dironly is true but there is now another place where OSError must be tracked, so it is not clear if this is worth it.

glob_scandir_2_diff.patch is the differential patch between glob_scandir_2.patch and glob_scandir_3.patch.

Here are the performance tests run with both patches.

$ ./python -m timeit -s "from glob import glob" -- "glob('**/*', recursive=True)"
glob_scandir_2.patch: 33.1 msec per loop
glob_scandir_3.patch: 33.8 msec per loop

$ ./python -m timeit -s "from glob import glob" -- "glob('/usr/lib*/**/*', recursive=True)"
glob_scandir_2.patch: 927 msec per loop
glob_scandir_3.patch: 850 msec per loop

$ ./python -m timeit -s "from glob import glob" -- "glob('/usr/lib*/**/', recursive=True)"
glob_scandir_2.patch: 423 msec per loop
glob_scandir_3.patch: 337 msec per loop

----------
Added file: http://bugs.python.org/file41570/glob_scandir_2_diff.patch

_______________________________________
Python tracker <report at bugs.python.org>
<http://bugs.python.org/issue25596>
_______________________________________