[Biopython-dev] GSoC draft proposal - lazy loading SeqIO parsers

[Evan Parker]

Thank you both for your fast and thorough evaluation of my proposal.

*Regarding time requirements:*
My adviser is aware the possibility that I may participate in this program.
During the summer I would file a "planned educational leave" instead of
enrollment to accommodate my full-time participation in GSoC.

As for the time requirements; I cannot avoid my obligations prior to ASMS
although I can promise to spend every extra minute I have to honor my
obligations to Biopython.  If my lack of full time availability prior to
June precludes me from participation I will understand.

*Regarding specific suggestions:*
I will come up with a deeper description of the relationship between
SeqRecProxy and SeqRecord before Friday. I like the idea of a SeqRecProxy
returning itself when sliced, I had not thought of it but it would be an
elegant solution to the problem of unparsed-vs-parsed annotations, this
feature would also allow more transparent use of proxy objects and would
pave the way for compatibility with SeqIO.write().

I considered using multiple proxy classes, but I prefer making a
standardized binding for a lazy parsing function function that can be
accepted by a single SeqRecProxy at run-time. I'll make this more explicit
in my proposal.

There are many other questions and points of clarification that I still
need to evaluate. I'll incorporate as much as I can in my proposal without
overloading it and without making statements that I cannot back up with my
own understanding.

Thanks again,

-Evan