This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author vstinner
Recipients loewis, vstinner
Date 2010-09-13.12:09:42
SpamBayes Score 5.3374566e-05
Marked as misclassified No
Message-id <1284379784.92.0.239937735597.issue9820@psf.upfronthosting.co.za>
In-reply-to
Content
> I think trying to emulate, in Python, what the *A functions 
> do is futile.

My problem is that some functions will use mbcs in strict mode (functions using PyUnicode_EncodeFSDefault): raise UnicodeEncodeError, and other will use mbcs in replace mode (functions using Windows functions in ANSI mode): raise IOError (or other error depending on the function). It's inconsistent. We should try to keep the same behaviour for all functions.

Examples of functions using (indirectly) PyUnicode_EncodeFSDefault to encode unicode filenames: bz2.BZ2File() and _ssl.SSLContext.load_cert_chain().
History
Date User Action Args
2010-09-13 12:09:45vstinnersetrecipients: + vstinner, loewis
2010-09-13 12:09:44vstinnersetmessageid: <1284379784.92.0.239937735597.issue9820@psf.upfronthosting.co.za>
2010-09-13 12:09:43vstinnerlinkissue9820 messages
2010-09-13 12:09:42vstinnercreate