Don't use regular expression to parse HTML. There are simply too many ways to write HTML. What happens if the closing tag is </td > (note the white space), for example?

The following code uses HTML::TokeParser and stuffs the data you want into the @data array.

#!/usr/bin/perl -w
require 5.004;
use strict;
use HTML::TokeParser;
use Data::Dumper;
    
my $doc      = shift or &usage;
my $p        = HTML::TokeParser->new($doc) || die "Can't open: $!";
my @data;

# walk through document and get each tag
while (my $token = $p->get_tag) {
    my $tag = $token->[0];
    if ( $tag eq 'td' ) {
        # get_text until a closing 'td' is found
        push @data, $p->get_trimmed_text( "/td" );
    }
}

print Dumper \@data;

sub usage {
    print "\ttest.pl some.html";
    exit;
}
[download]

Cheers,
Ovid

Vote for paco!

Join the Perlmonks Setiathome Group or just click on the the link and check out our stats.

To do this reliably (assuming that's a real-life example up there), you are going to need to use some derivative of HTML::Parser--probably HTML::TokeParser, (which is recommended for small jobs),or HTML::TableExtract.

If you really are getting things out of table rows, then of course HTML::TableExtract is an easy call, and it's an easy module to use, and you're set. If a more general HTML solution is needed, then I would recommend something along the lines of the following pseudocode (using TokeParser):

my ($temp,$flag);
while ($token = $parser->get_token) {
    if ($token eq $start_token) {
         $flag = 1
    } elsif ($token eq $end_token) {
         $flag = 0;
         return $temp if $temp =~/KEY/;
         $temp = '';
    }
    $temp .= $token if $flag;
}
[download]

foreach ($text =~ m#<tr>(.*?)</tr>#gmsi)
{
     # Note that foreach assigns the memorized text to $_
     if (/\bISTF\b/)
     {
          print "$_\n";
     }
}
[download]

This works assuming you want quick and dirty rather than using the elegant reliable Parser solution.

$string =<<'TEXT';
<tr>1 The last company to fail was ISTF with losses of  $60M.</tr>
<tr>2 The last company to fail was ITF  with losses of  $60M.</tr>
<tr>3 The last company to fail wasISTF with losses of  $60M.</tr>
< tr>4 The last company to fail was ISTF 
with losses of  $60M.</tr >
TEXT

print "Found $1\n" while $string =~ m|<\s*tr\s*>([^<]*\bISTF\b[^<]*)<\
+s*/tr\s*>|gi;
[download]

cheers

tachyon

s&&rsenoyhcatreve&&&s&n.+t&"$'$`$\"$\&"&ee&&y&srve&&d&&print

I'd say that's a little bit too quick and dirty, it breaks easily on sth like: <tr>The <em>last</em> company to fail was ISTF.</tr> Parsing HTML is tricky, that's what the parser modules (mentioned multiple times in this thread) are for.

-- Hofmator

I'm no regex master, but if it were me I'd do it in two stages:

for ($text =~ /<tr>(.*?)<\/tr>/gmsi){
  $selected=$1;
  if ($selected =~ /ISTF/is) {
    $selected_text = $1 ;
    ## Your code here
  }
}
[download]

andre germain
"Wherever you go, there you are."