[BioSequences.jl] Search

[CiaranOMara]

I think there are some things regarding BioSequence's search mehods that are worth discussing.
Since I'm short on time, what follows are notes/bullet points. But, hopefully, they are sufficient to start a consensus, and then maybe some courses of action.

I believe there is a question as to what type of search result is most useful. Is it useful to return a range, is it better to return nothing, or have both by keeping lower-level functions that return ranges available? On a related note, does chaining search queries make sense?

Subtype search queries from abstract type SearchQuery end and consolidate an interface for:

• ExactSearchQuery
• ApproximateSearchQuery
• PWM

I like findnext(predicate/query::SearchQuery, seq::BioSequence, start::Integer, stop::Integer).

New features:

• Add a new LiteralSearchQuery where N doesn't match.
• Provide a search iterator?
• Add a findall search variant, which could set and collect from an iterator.

Thoughts @TransGirlCodes @jakobnissen @kescobo?

[kescobo]

@CiaranOMara Apologies for never responding to this... better (2 years) late than never? 😬

I have a vague recollection that some work has been done on this in the meantime - is there a more updated discussion?

[CiaranOMara]

I think some relevant discussion happened at BioJulia/BioSequences.jl#190 -- there may be other places. And if I recall correctly, some of the work was done in the leadup for the v3 release.

I have a PR relating to this discussion but have let it fall by the wayside as it was out of sync with what others were doing at the time.