Tabs in FASTA record headers cause Kaptive to fall over

[CorinYeatsCGPS]

Hi,

I found we have a few FASTAs with tab characters in them, which causes Kaptive to fall over:

Assembly        Best match locus        Best match type Match confidence        Problems        Identity        Coverage        Length discrepancy      Expected genes in locus Expected genes in locus, details  Missing expected genes  Other genes in locus    Other genes in locus, details   Expected genes outside locus    Expected genes outside locus, details   Other genes outside locus       Other genes outside locus, details        Truncated genes, details        Extra genes, details
Traceback (most recent call last):
  File "/usr/local/bin/kaptive", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/usr/local/lib/python3.11/site-packages/kaptive/__main__.py", line 234, in main
    if result := typing_pipeline(assembly, args.db, args.threads, args.score_metric, args.weight_metric,
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/kaptive/assembly.py", line 278, in typing_pipeline
    partial=a.partial, dna_seq=assembly.seq(a.ctg, a.r_st, a.r_en, a.strand))
                               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/kaptive/assembly.py", line 60, in seq
    return self.contigs[ctg].seq[start:end] if strand == "+" else self.contigs[ctg].seq[
           ~~~~~~~~~~~~^^^^^
KeyError: 'OW967300.1'

It's a bit unusual but technically it's allowed to have tabs in the headers, and there appears to be a pipeline in use that produces them.

[tomdstanton]

See klebgenomics/Kleborate#85

[tomdstanton]

Fixed in b94e596.
@CorinYeatsCGPS can you double check with the problematic assembly on your end? I made some synthetic tabbed headers and seemed to work fine!