[Biopython-dev] flex, setup.py and Bio.PDB.mmCIF (Bug 2619)

[Peter Cock]

On Fri, Mar 15, 2013 at 1:09 PM, Michiel de Hoon <mjldehoon at yahoo.com> wrote:
> Hi everybody,
>
> I looked at the mmCIF parser again, and it turned out that the Python
> standard library contains a shlex lexical analyzer module that makes mmCIF
> parsing straightforward without relying on flex or PLY. I uploaded a
> modified version of MMCIF2Dict.py to the git repository. This parser does
> the exact same thing as the flex-based parser, but is in pure Python. If
> you're interested, have a look at MMCIF2Dict.py in the git repository;
> comments and suggestions are welcome.

That makes MMCIF2Dict look a lot shorter :)
https://github.com/biopython/biopython/commit/b2bafdfcd67c738f91722495bb732297b7936828

> If there are no objections, I think we can remove everything in
> Bio.PDB.mmCIF. Also I'm a bit unhappy with how the information in an mmCIF
> file is represented in Biopython. I think there are more Pythonic ways to
> store the contents of an mmCIF file in a Python object.
>
> Best,
> -Michiel.

Do you think we need a deprecation cycle for Bio.PDB.mmCIF? It has
been available by default on Debian etc where the dependency was
taken care of by the packagers.

I've never used this code so perhaps Eric or João's perspective would be
more helpful than mine.

Peter