As pointed out by a friend of the show, we could optimize the conversion from plain ASCII nucleotide sequence into compressed PackedSeq stuff by using a branchless approach based on a table. Here's his demonstration:
https://gist.github.com/cgyurgyik/dc0627fb20b977e29b1d7155561bd093
Once #214/#241 are resolved, let's try implementing this and seeing how it impacts GFA->FlatGFA conversion end-to-end!
As pointed out by a friend of the show, we could optimize the conversion from plain ASCII nucleotide sequence into compressed
PackedSeqstuff by using a branchless approach based on a table. Here's his demonstration:https://gist.github.com/cgyurgyik/dc0627fb20b977e29b1d7155561bd093
Once #214/#241 are resolved, let's try implementing this and seeing how it impacts GFA->FlatGFA conversion end-to-end!