The awk script takes the output of Infernal's cm{scan,search} fmt=2 mode and makes it suitable for use by esl-sfetch, a sequence selector

Reading the user's guide for Infernal, Version 1.1.2; July 2016 http://eddylab.org/infernal/Userguide.pdf#page=60 we see that the relevant fields in the cmscan output are: (fmt2 column number: explanation) 3: The accession of the target sequence or profile, or ’-’ if none 4: The name of the query sequence or profile 10: The start of the alignment of this hit with respect to the sequence, numbered 1..L for a sequence of L residues. 11: The end of the alignment of this hit with respect to the sequence, numbered 1..L for a sequence of L residues

Likewise the format esl-sfetch wants is: <newname> <from> <to> <source seqname>

Putting it all together we see that the newname (which esl-sfetch with output using) is a concatenation of the original name, the sequence number, and the coordinates.