Yes, the SGI appears not to return a grouping ([3, 0] is expected) for
the en_US locale (the rest of it looks OK).
However, there is still a bug in Lib/locale.py - the code currently
tries to allow for the possibility that an empty grouping may be
returned from localeconv() (and there must be some locales where this is
correct):
def _group(s):
conv=localeconv()
grouping=conv['grouping']
if not grouping:return s
The code calling _group() assumes that the object returned will always
be a tuple, and hence the above will cause a traceback when localeconv()
returns an empty grouping. The correct code should be:
def _group(s):
conv=localeconv()
grouping=conv['grouping']
if not grouping:return (s, 0)
test_locale will still fail on the SGI, but now because of a bug in the
platform implementation of the en_US locale, rather than a bug in the
Python locale.py code. It's better than a traceback.
BTW, mail to Martin doesn't seem to be getting through.
Guido van Rossum wrote:
>
> > localeconv()['grouping'] is "[]" at this point...
>
> It is looking like either the _locale.c module is not working right or
> the platform's implementation of the en_US locals is broken. I'm
> afraid that only you will be able to make the determination. :-(
>
> --Guido van Rossum (home page: http://www.python.org/~guido/)
--
Mark Favas - m.favas@per.dem.csiro.au
CSIRO, Private Bag No 5, Wembley, Western Australia 6913, AUSTRALIA
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4