Unless you show how your BLAST report looks like (Tabular, line-wise...etc) by enclosing a subset of the report in here we are not going to be able to 'guide' you best. While BioPerl can be a way to go, Boulder::Blast can be another option.
in reply to Subroutine to parse BLAST
Whether or not to use regular expressions can not be absolutely ruled in/out because sometimes you require to combine the parsing abilities of the module you use with the prowess of regular expressions when parsing such sequence or blast objects. So without data we're just punching in the dark.
NOTE: You seem to be using a library 'BeginPerlBioinfo', where is that coming from, did you read its documentation already?
also you need to read Markup in the Monastery and Perl Monks Approved HTML tags.
Excellence is an Endeavor of Persistence.
A Year-Old Monk :D .