If you treat the arg as a regex pattern, "[P/R]" actually means "P", "/" or "R". You simply want "[PR]".

"." means any character

"[QRG]" will match one of those three characters.

And if you want to format you specified:

my $pat = '';
for ($ARGV[0]) {
   /\G ([A-EG-WYZ]+)            /xgc && do { $pat .= $1;       redo };
   /\G X                        /xgc && do { $pat .= '.';      redo };
   /\G F                        /xgc && do { $pat .= '[QRG]';  redo };
   /\G \[ ([A-Z]) \/ ([A-Z]) \] /xgc && do { $pat .= "[$1$2]"; redo };
   /\G \z                       /xgc && last;

   my $pos  = pos();
   my $next = substr($_, $pos, 1);
   die("Unrecognized character \"$next\" at pos $pos\n");
}
[download]

Update: Added missing /xgc.

I found very interesting this piece of code as an example for the \G pattern, but I noticed that x modifier was missing on each regexp AFAIK, but then tested if there was anything else I'm not awared.

Just as it is, it returns nothing on $pat. Then added the x modifier and got an endless loop!

I thought it was related to a note in perlre: "It is worth noting that \G improperly used can result in an infinite loop.", but a small trace using print "($pat)\n"; just before each redo shows that \G is always at the beginning of the argument, not after what was already matched.

I tried again removing pos() to see if that function resets the pointer... No luck!

Am I doing something wrong?

It's because 'x' wasn't the only missing modifier. 'g' (to make the engine pay attn to pos) and 'c' (to prevent the engine from resetting pos) were also missing. Fixed.

m/ASP[a-z]/i
# thats a literal 'ASP'
# followed by a character class of 1 of any letter from a-z
[download]