You need to turn file wildcards * (and probably ?) into the equivalent regular expression matches. The following code does that and combines the resulting match strings into a single regular expression. Note that two versions of the pattern are returned to facilitate sorting the match stings in order of most explicit to least explicit. That may not be important, but you should think about the implications.

use strict;
use warnings;

my $patFile = <<PATS;
*.pl.*
*\\PL\\*
*\\pl-*.fbrb
*_pl-*.fbrb
*\\*_pl-00.fbrb
*\\pl-00.fbrb
*\\polish\\*
*\\psarc\\polish*.psarc
*_POL.*
*_POLISH.SUB
*_POL_*
*_po.xvag
*_polish.*
*_pl.psarc
*_pl2.psarc
*_pol.*
*_por.*
PATS

open my $patIn, '<', \$patFile;
my @patterns = map {parsePattern($_)} <$patIn>;
my $matchStr = join '|',
    map  {$_->[1]}
    sort {length($b->[0]) <=> length($a->[0])} @patterns;
my $regex = qr/($matchStr)/;

print "Match regex is '$matchStr'\n";

while (<DATA>) {
    chomp;
    print "Matched '$_' on $1\n" if $_ =~ $regex;
}


sub parsePattern {
    my ($path) = @_;
    
    chomp $path;

    (my $explicit = $path) =~ tr/*//d;
    
    $path =~ s![\\/]![\\\\/]!g;
    $path =~ s/\./\\./g;
    $path =~ s/^\*//;
    $path .= '$' if $path !~ s/\*$//;
    $path =~ s/\*/.*/g;
    $path =~ s/\?/./g;
    return [$explicit, $path];
}

__DATA__
c:\Build\PL\Data\test1.dat
c:\Build\Data\test1.dat.wibble_POLISH.SUB
c:\Build\Data\test1.dat.wibble_POLISH_SUB
[download]

Prints:

Match regex is '[\\/]psarc[\\/]polish.*\.psarc$|[\\/].*_pl-00\.fbrb$|[
+\\/]pl-00\.fbrb$|_POLISH\.SUB$|_pl2\.psarc$|[\\/]pl-.*\.fbrb$|_pl-.*\
+.fbrb$|_pl\.psarc$|[\\/]polish[\\/]|_po\.xvag$|_polish\.|_POL\.|_POL_
+|_pol\.|_por\.|\.pl\.|[\\/]PL[\\/]'
Matched 'c:\Build\PL\Data\test1.dat' on \PL\
Matched 'c:\Build\Data\test1.dat.wibble_POLISH.SUB' on _POLISH.SUB
[download]

As a further generalization both / and \ path separators are accepted.