[Python-Dev] Adding the 'path' module (was Re: Some RFE for review)

[Thomas Heller]

Neil Hodgson <nyamatongwe at gmail.com> writes:

> Guido van Rossum:
>
>> Ah, sigh. I didn't know that os.listdir() behaves differently when the
>> argument is Unicode. Does os.listdir(".") really behave differently
>> than os.listdir(u".")? 
>
>    Yes:
>>>> os.listdir(".")
> ['abc', '????????????']
>>>> os.listdir(u".")
> [u'abc', u'\u0417\u0434\u0440\u0430\u0432\u0441\u0442\u0432\u0443\u0439\u0442\u0435']
>
>> Bah! I don't think that's a very good design
>> (although I see where it comes from). 
>
>    Partly my fault. At the time I was more concerned with making
> functionality possible rather than convenient.
>
>> Promoting only those entries
>> that need it seems the right solution -- user code that can't deal
>> with the Unicode entries shouldn't be used around directories
>> containing unicode -- if it needs to work around unicode it should be
>> fixed to support that!

I'm sorry but that's not my opinion.

Code that can't deal with unicode entries is broken, imo.  The
programmer does not know where the user runs this code at what he throws
at it.  I think that this will hide bugs.

When I installed the first game written in Python with pygame on my
daughter's PC it didn't run, simply because there was a font listed in
the registry which contained umlauts somewhere.

OTOH, I once had a bug report from a py2exe user who complained that the
program didn't start when installed in a path with japanese characters
on it.  I tried this out, the bug existed (and still exists), but I was
astonished how many programs behaved the same: On a PC with english
language settings, you cannot start WinZip or Acrobat Reader (to give
just some examples) on a .zip or .pdf file contained in such a
directory.

>    OK, I'll work on a patch for that but I'd like to see the opinions
> of the usual unicode guys as this will produce more opportunities for
> UnicodeDecodeError. The modification will probably work in the
> opposite way, asking for all the names in unicode and then attempting
> to convert to the default code page with failures retaining the
> unicode name.

Thomas