xa2multi.pl issue creating secondary alignments from hard-clipped primary

[sjaenick]

Hi @lh3,

the attached SAM entry (read aligned to hg38)

bug_sam.txt

is hard-clipped (79H70M), i.e. only the matching 70 nucleotides are found in the SEQ field. Passing this to xa2multi.pl,
the next alignment (79S70M) is generated from the XA tag. Here, SEQ should include the soft-clipped bases, but they
are missing (and cannot be obtained from the initial entry).

Only reasonable fix would be to make xa2multi.pl obtain SEQ from the original FASTQ file, I guess.