As I understand it, this will extract all <tr> blocks from my input.
Since you have not provided that input, we can only guess. Looking at How to ask better questions using Test::More and sample data it is simple enough to come up with a demonstration which shows that regex to work for a given set of input. Note that like brother LanX I'm going to get rid of all those backslashes.
use strict;
use warnings;
use Test::More tests => 4;
my $input = '<html><tr>foo</tr><tr>bar</tr><tr>baz</tr></html>';
my @table = $input =~ m#(<tr.*?</tr>)#g;
is $#table, 2;
is $table[0], '<tr>foo</tr>';
is $table[1], '<tr>bar</tr>';
is $table[2], '<tr>baz</tr>';
