Question to behaviour of Paffy upconvert

[Maed0x]

Hi @glennhickey ,

I’m using paffy upconvert to adjust alignment coordinates to extracted subsequences. The tool checks if an alignment is fully inside the subsequence boundaries (start >= subsequence_start and end <= subsequence_end). I found that if an alignment starts inside the subsequence but extends past its end, the tool exits early instead of skipping that alignment and continuing with the next ones.

I wanted to check if this early exit is intended?

E.g., My current command:
paffy upconvert 1.fa -i testPaf.paf -o upconverted.paf

My PAF file (testPaf.paf) has a target sequence length of 602,619. When upconverting using a 100k to 200k subsequence of the target sequence (1.fa), alignments fully within this range convert correctly. However, row 155 spans 184,314 to 203,093 and causes an early exit. Should this row be skipped so rows 156 and 157 can still be upconverted?

The files
data.zip
For demonstration, I changed the assert in line 46 of paf_upconvert.c to an if-else statement. Because of this, it does not exit early so an upconverted.paf file could be generated.

Best,
Niklas