PEP on path module for standard library

paul at boddie.org.uk paul at boddie.org.uk
Sat Jul 23 15:23:02 CEST 2005


Michael Hoffman wrote:
> Reinhold Birkenfeld wrote:
> > Probably as Terry said: a path is both a list and a string.

[...]

> One way to divide this is solely based on path separators:
>
> ['c:', 'windows', 'system32:altstream', 'test.dir',
> 'myfile.txt.zip:altstream']

I would argue that any proposed solution has to work with VMS
pathnames. ;-)

> The current stdlib solution, os.path.splitext(os.path.splitext(filename)
> [0])[0] is extremely clunky, and I have long desired something better.
> (OK, using filename.split(os.extsep) works a little better, but you get
> the idea.)

And also with unusual (eg. RISC OS) filename extensions.

To do any justice to the existing solutions, any PEP should review at
least the following projects:

 * The path module (of course):
   http://www.jorendorff.com/articles/python/path/

 * The py.path module (or at least the ideas for it):
   http://codespeak.net/py/current/doc/future.html

 * itools.uri
   http://www.ikaaro.org/itools

 * Results from the "Object-Oriented File System Virtualisation"
   project in the "Summer of Code" programme:
   http://wiki.python.org/moin/SummerOfCode

And I hope that the latter project is reviewing some of the other work,
if only to avoid the "framework proliferation" that people keep
complaining about.

Paul




More information about the Python-list mailing list