Hi all,

I am delighted to report that I have figured out how to address both
Unicode and File aspects of Python 3 for pycurl and therefore, a
release of pycurl with Python 3 support is now closer than ever!

I have a patch[1] that passes tests on my machine (but not on travis,
not sure why yet). It leaks memory in string conversions and needs some
other minor cleanup, but appears to be completely functional otherwise.

This patch is only based on Python 3 patches posted to SourceForge and
my own work, therefore it should be as license-ok as any other patch
submitted to the project (unlike patches in various GNU/Linux
distributions which may be partially covered by GPL).

Please feel free to review the patch and/or try it in your code. Please
note files added under doc directory that document behavior under
Python 3.

[1] https://github.com/pycurl/pycurl/pull/69

Oleg
_______________________________________________
http://cool.haxx.se/cgi-bin/mailman/listinfo/curl-and-python
Received on 2013-12-09