Stefan Holek added the comment: A little more context perhaps:
The use-case is building Python distributions containing non-ASCII filenames. These seemingly "invalid" filenames can occur in real-life when the files have been created by, say, a 'git clone' operation. So yes, I have Latin-1 bytes on the filesystem, even though my locale is UTF-8. And yes, Python 3 decodes that filename using surrogates. Creating .tar.gz distributions in this situation appears to work (even re-creating the foreign bytes when the archive is later extracted), whereas .zip archives fail in the way described above. I was hoping zipfile could be made to work the same as tarfile in this regard. Concerns for standards certainly didn't keep tarfile from supporting surrogates. ;-) ---------- _______________________________________ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue16310> _______________________________________ _______________________________________________ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
