feasibility of porting to Python 3k

[GoogleCodeExporter]

At the moment, MDAnalysis will not run in Python 3.x.

With the move to only support Python 2.6+ (see Issue #130) it should become doable to find out how difficult it will be to make the library also run under Python 3.x.

For the foreseeable future, we will have to maintain a working library under 2.6+ but it would be useful to find out if any of the porting methods would work for us, e.g. maintaining a "conversion-ready" 2.x code base and then use the 2to3 tool to generate the 3.x code when needed.

If anyone wants to try this then let me know and I'll make you owner of this issue. We can maintain an experimental py3k branch in the repository.

-- Oliver

Original issue reported on code.google.com by orbeckst on 17 Apr 2013 at 5:35

• Blocked on: discontinue Python 2.5 support #130, SWIG -> Cython rewrite of the XDRReader interface #211

[GoogleCodeExporter]

I'll take this one

Original comment by [email protected] on 6 May 2013 at 8:02

[GoogleCodeExporter]

Original comment by orbeckst on 6 May 2013 at 4:07

[GoogleCodeExporter]

Original comment by [email protected] on 4 Feb 2015 at 12:08

• Added labels: Milestone-Release1.0

[GoogleCodeExporter]

Original comment by [email protected] on 5 Feb 2015 at 4:19

• Now blocked on: SWIG -> Cython rewrite of the XDRReader interface #211

[GoogleCodeExporter]

So apart from using print() and not using Swig, is there anything else that we 
need to keep in mind when writing new things so they're 2/3 compatible?

Original comment by richardjgowers on 6 Feb 2015 at 5:40

[GoogleCodeExporter]

AFAIK Swig is python 3 compliant, so using it should be fine. I tend to prefer 
Cython because 1. the code if closer to Python, and 2. you just need one file 
where swig generates a few which is IMHO is cumbersome.
But basically using a couple of things from __future__ (print function, unicode 
litterals...) for Python and using Cython (good!) or Swig (not my cup of tea) 
for hardcore stuff should make things smooth for py3k.

Original comment by [email protected] on 6 Feb 2015 at 8:22

[orbeckst]

See also #260.

[orbeckst]

The discussion around #260 (and in particular @kain88-de 's recent work) has established what needs to be done so I am closing this issue.