[Biopython-dev] Status of SearchIO

[Wibowo Arindrarto]

>>> 1. Bio.seq.search (in line with the namespace change)
>>>> 2. Bio.seqsearch (top-level module, separate from Bio.seq. This used
>>>>    to be Bio.SeqSearch, now adjusted for PEP8 compliance)
>>>> 3. Bio.search (same reasoning + explanation like Bio.seqsearch).
>>>> 4. Bio.SearchIO / Bio.searchio
>>>> 5. Bio.psearch (p for pairwise)
>>>
>>>> If it's down to a vote, I would vote to merge this branch as
>>>> Bio.SearchIO, and perhaps lowercase it to Bio.searchio or
>>>> biopy.searchio in the Py3 lowercase branch.
>>>>
>>>> Rationale: We already follow BioPerl with SeqIO and AlignIO, and it
>>>> seems to help users. It's also Google-friendly.
>>>
>>> I would vote for Bio.seq.search.
>>
>> And would you support moving other existing Bio.SeqXXX
>> modules under Bio.seq.* as for example outlined here?:
>> http://lists.open-bio.org/pipermail/biopython-dev/2012-October/009999.html
>> If so then I think we should go with that plan.
>
> I have started exploring that idea on this new branch,
> https://github.com/peterjc/biopython/tree/bioseq
>
> Does anyone object to me applying the first commit to the master
> branch (defining the previously discussed new warning for 'beta' code)?
> https://github.com/peterjc/biopython/commit/97485d5dcf2620f7664ae46a7897c1203847538d

No objection from me for the commit :).

But I have some concerns for the SearchIO naming. I like Bio.seqsearch
best at the moment. Bio.seq.search is good, but I understand that
Bio.SearchIO will eventually contain app wrappers and code for remote
searches as well. Putting it three levels-deep doesn't feel nice to
me. As comparisons, submodules with similar features (Bio.Phylo, and
possibly Bio.AlignIO, if in the future it will be merged with
alignment app wrappers and the alignment object model) are available
under Bio.

> Note that introducing Bio.seq now (and any relocations under this)
> can (I believe) still be combined with the lower-case modules under
> Python 3 idea as well. This just requires the public classes and
> functions defined under Bio.Seq.* remains mirrored under Bio.Seq.*
> (this means assorted Seq objects and some functions like translate).
>
> Peter

regards,
Bow