in reply to Re: Re: Re: Bulk Regex?
in thread Bulk Regex?

I wish it were that simple! What I have are dept code *prefixes* which are not necessarily all the same length. For example, if our department codes were like:

ABCD123456 DEFG123456 DEFG998800 DEEE11223344

The code prefixes I have are like:

ABCD12 DEF DEEE11223

In other words, they can be any length. I tried using something like this: 

foreach my $pattern (@patterns) {
   if (substr($fields[7],0,length($pattern)) eq $pattern) {
       ### Code here
       last;
   }
}

[download]
but that was slightly slower than the regex methods.

Replies are listed 'Best First'.
Re^5: Bulk Regex?
by RMGir (Prior) on Aug 29, 2002 at 16:58 UTC
    Ouch.

    How many prefixes of each length, though? If you have many of a given length, maybe you could do something like this (writing support code to generate this HOH, of course): 

    my %prefixList=(
   3 => {DEF=>1, SRC=>1},
   6 => {ABCD12=>1},
  10 => {DEEEE11223},
);

#....
# inside the loop
foreach my $length(%prefixList) {
    if($prefixList{$length}->{substr($fields[7],0,$length)}){
        ### code here
        last;
    }
}

    [download]
    If the set of sizes isn't dynamic, you could make that an AOH, with some possibly empty hashes, I guess... You could also get rid of the loop and inline the check for each length, but I don't think that would save you much.

    If you're only ever going to have 1 or 2 prefixes of a given length, then this isn't worth it. But given your initial description, I think with a LOT of prefixes and very few different lengths, this could start to pull ahead of the regex.
    --
    Mike

[reply]
[d/l]

In Section Seekers of Perl Wisdom