I encountered a performance problem using the ipaddr library to merge over 10000 network addresses. I sent a patch upstream to fix it, and that patch has been merged:
https://github.com/google/ipaddr-py/commit/6504b47a02739e853043f0a184f3c39462293e5c

Since ipaddr is also included in the standard lib of Python 3 (I think since version 3.3?) and above, I'd like to ask that this patch be considered for committing here too.

Background on including ipaddr into the std lib: https://bugs.python.org/issue14814

Could you provide a patch?

Perhaps I'm wrong, but a superficial inspection of the ipaddress.py seem to indicate that it's not affected by the same issue. _find_address_range is implemented as a generator, it doesn't restart comparison on every iteration. Alexander, did you experience any performance issues when using ipaddress module from standard library?

Thanks for the responses.

I'm afraid I was looking at the Python 3.3 source code on my desktop, it looks like the problem was fixed in the Mercurial repo about 10 months ago by https://hg.python.org/cpython/rev/f7508a176a09

Marked this issue as a duplicate? Feel free to change the resolution if that isn't right.

Cheers,

Alex