... while( $stuff =~ /^ (\d+): \s+ # the number xx: => $1 (\w+) \s+ # the locus => $2 ( (?:.(?!^\d+:))+ ) # the remaining record => $3 /msgx ) { my ($locus, $name, $record, $kegg, $func, $proc, $comp) = ($2, '', $3, ('unknown')x4); $name = $1 if $record =~ /([^\n\[]+)\s*/; $kegg = $1 if $record =~ /KEGG \s+ pathway: \s+ (.+?)\s+Function \s+ Evidence/sx; $func = $1 if $record =~ /Function \s+ Evidence \s+ (.+?) (?:\n\n|\z)/sx; $proc = $1 if $record =~ /Process \s+ Evidence \s+ (.+?) (?:\n\n|\z)/sx; $comp = $1 if $record =~ /Component \s+ Evidence \s+ (.+?) (?:\n\n|\z)/sx; print join "\n\n", $locus, $name, $kegg, $func, $proc, $comp; } ...