[Biopython-dev] Level of Python 3 support?

[Eric Talevich]

On Sun, Apr 14, 2013 at 2:10 PM, Peter Cock <p.j.a.cock at googlemail.com>wrote:

> Hi all,
>
> We've have Python 3 coverage on the Travis continuous
> integration tests & the nightly builtbot tests for quite some
> time now.
> http://travis-ci.org/biopython/biopython
> http://testing.open-bio.org/biopython/tgrid
>
> There have been a few Python 3 specific issues reported
> (which have been fixed), but I think we should probably
> go ahead and say we support Python 3.1 or later (with
> the exception of some minor C code which has not been
> ported).
>
> Does this seem like a good move for the next release?
>
>
Sounds good to me. Looking at the changes between 3.1 and 3.2:
http://docs.python.org/3.2/whatsnew/3.2.html

I don't see any features unique to 3.2+ that would justify us skipping
version 3.1 support. The main changes that might affect compatibility
between 3.1 and 3.2 are:

- More attention was paid to strings vs. bytes, potentially fixing some
bugs and making other functions in the standard library pickier about which
types they accept.

- Similarly for context management, the __enter__ and __exit__ methods were
implemented on more resource types so the "with" statement will work.

- ElementTree (xml.etree) got a long-awaited update.

- In the gzip module, functions compress() and decompress() were added and
GzipFile grew a peek() method. Would you want to use any of those?

- There are some new unittest functions in both 3.2 and 2.7 that would
probably be handy.