Hi,

I need to do massive regexp search and replace. For a given text file, I need to go through each line of this file and apply each of the (different) replacement expressions listed in a Perl script (or a separate data file). The whole thing is not too complex to implement. The problem is my replacement expressions may amount to thousands, in which case all several implementations I have tried are extremely slow.

Here is a sample implementation of mine:

open(IN, "<samplein.txt");
my @INPUT = <IN>;
close (IN);

open (OUT,">sampleout.txt");
foreach $INline (@INPUT) {
    foreach my $Rpatt (@Patts) {
        (my $Source, my $Target) = split(/\t/, $Rpatt);
        if ($Source && $Target) {
            $Source  = $Source;
            $Target = "\"$Target\"";
            $INline =~ s/$Source/$Target/gee);
        }
    }
    chomp $INline;
    print OUT $INline;
}
close (OUT);
[download]

This implementation does the job, but when the list of replacement patterns (@Patts) is big (several thousands), the replacement takes ages.

Can anyone help me find an efficient implementation.