Closed marcelm closed 7 months ago
I haven't had the time to look at all the changes in this PR in detail. But I agree with the name changes and with what you describe, without reservation.
If this PR is holding up any other development, or if you think its mostly non-controversial changes, I suggest a merge without a careful check.
Do you want to review this PR more closely or shall I just merge it?
Won't have time anytime soon unfortunately . Yes please go ahead and merge.
This renames:
align_SE
→align_single
align_PE
→align_paired
align_SE_read
→align_or_map_single
align_PE_read
→align_or_map_paired
joint_nam_scores
→nam_pairs
(since we now use theNamPair
struct)rescue_mate
→rescue_align
(could never remember what it does differently fromrescue_read
)get_mapq
toproper_pair_mapq
(since this is only used to compute MAPQ for uniquely mapping proper pairs)Also,
InsertSizeDistribution
is moved to a separate file.The main change is to factor out an output_aligned_pairs() function. The rescue_read() and align_paired() functions previously did two things:
When working on random alignment selection for #367, I noticed that the code outputting to SAM was duplicated (and had slightly diverged already). This commit moves that part to a new function output_aligned_pairs(). Now rescue_read() and align_paired() just return a vector of ScoredAlignmentPairs.
This should fix at least two bugs: