Re: How would I write this multiple substitution script?

[TomDLux]

You want to do a number of substitutions to each file in a directory, with different substitutions for each of a large number of directories.

I would suggest putting the directories and associated substitutions into a data structure. It's a little simpler to put at the top of the file, but harder to modify. Reading in a YAML file containing the rules is a little more complicated to get working, but simpler to maintain over the years, if this is something that will be used for a long time.

Since you're a novice, I'll suggest putting the data in the file:

use autodie;
use File::Temp qw(tempfile);
my @RULES = (
             { dir  => '/path/to/dir1',
               mods => [ 'yellow', 'green' ], [ 'red', 'blue' ] ],
             },
             { dir  => '/path/to/dir2',
               mods => [ [ 'puce', 'chartreuse' ], [ 'lime green', 'ma
+genta' ] ],
             },
             );
...
for my $rule ( @RULES ) {
    chdir $rule->{dir};
    for my $file ( glob( .* * ) ) {
       open my $input, '<', $file; 
       my ($output, $outfile) = tempfile(); 
       while ( my $line = <$input> ) {
           chomp $line;
           for my $sub ( @{ $rule->{mods} } ) {
               $line =~ s/$sub->[0]/$sub->[1]/
           }
           print $output $line;
       }
       close $input;
       close $output;
       replace_file( $file, $outfile );
    }
}

I leave for you the process of verifying the old file is successfully replaced by the new file. The regexes will be recompiled on each search-and-replace, but if they are really literal strings, that won't be too expensive. If there were matches, I would suggest precompiling the exporessions, but I'm not sure that works for search-and-replace.