[Python-3000] cleaning up *path.py code duplication (original) (raw)

Jack Diederich jackdied at jackdied.com
Sat Aug 26 03:52:43 CEST 2006


Ooph, there is some dissonance in the comments and the code. Cut-n-paste errors I suppose.

-- ntpath.py -- def islink(path): """Test for symbolic link. On WindowsNT/95 always returns false""" return False

<snip 10 lines>

This follows symbolic links, so both islink() and isdir() can be true

for the same path.

def isfile(path): """Test whether a path is a regular file""" -- end exeprt --

I'll try and keep a list so those in the know can do a post mortem on the comments. I'm only useful for vetting the *nix versions.

-Jack

On Fri, Aug 25, 2006 at 05:35:33PM -0700, Guido van Rossum wrote:

Sounds like a great 2.6 project. Beware of things that are intentionally different between platforms of course!

--Guido On 8/25/06, Jack Diederich <jackdied at jackdied.com> wrote: > While checking find() uses in the stdlib I noticed that the various > path modules have duplicate code and docstrings for some generic path > manipulations. Delightfully they even have different implementations > and docstrings for identical functions. splitext() is a great bad > example - os2emxpath.splitext() builds up strings by doing char-by-char > concatenations where everyone else uses find() + slice. > > If there are no objections I'll move these into a module named > genericpath.py and change the others to do > > from genericpath import func1, func2, funcN > > where applicable. > > So, any objections? Should it be a 2.6 backport too? > > -Jack _> ________________________ > Python-3000 mailing list > Python-3000 at python.org > http://mail.python.org/mailman/listinfo/python-3000 > Unsubscribe: http://mail.python.org/mailman/options/python-3000/guido%40python.org >

-- --Guido van Rossum (home page: http://www.python.org/~guido/)


Python-3000 mailing list Python-3000 at python.org http://mail.python.org/mailman/listinfo/python-3000 Unsubscribe: http://mail.python.org/mailman/options/python-3000/jack%40performancedrivers.com



More information about the Python-3000 mailing list