Author jeff.allen
Recipients dani_kase, jeff.allen
Date 2018-03-14.09:08:07
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <>
The code in the SO post:
 FS_ENCODING = 'ascii'; fn = b('[abc\xff]'); encoded = u('[abc\udcff]')
appears to be testing/expecting that the invalid ascii character '\xff' will be encoded into the invalid unicode string '\udcff'. This is used in CPython to byte-smuggle non-ascii file names through unicode strings.

Jython approaches differently. Versions after 2.7.0 may be able to treat this code more gently.

A much fuller consideration is given in #2340, #2342. Maybe allow one/both of these to supersede this ticket?
Date User Action Args
2018-03-14 09:08:07jeff.allensetmessageid: <>
2018-03-14 09:08:07jeff.allensetrecipients: + jeff.allen, dani_kase
2018-03-14 09:08:07jeff.allenlinkissue2424 messages
2018-03-14 09:08:07jeff.allencreate