Hi luxs,

What code have you tried?

Have a look at the "~" CSS selector in Mojo::DOM::CSS.

Update: The AM gave it away :-)

Hope this helps,
-- Hauke D

The only script I've manage to do is

my $tmpi = 0;
$fimg = $dom
  ->find('img, h1')
  ->map( sub
         {
          if( $_->tag eq 'h1' )
             { $tmpi++; }
          if( $tmpi > 0 && defined $_->attr->{'src'} )
             { $_->attr->{'src'} } # without semicolumn!!!!
         }
       );
for( my $ii = 0; $ii <= $#$fimg; $ii++ )
   {
    if( length( $fimg->[$ii] ) > 0 )
        {
         print $fimg->[$ii];
        }
    }
[download]

looks too complicated and not nice (but working).

Hi luxs,

Since I'm guessing the <img> tags can be nested arbitrarily deep, the CSS selector "img, h1" was going to be my next suggestion.

There are many other ways to write the logic, here's one:

use warnings;
use strict;
use Mojo::DOM;

my $data = <<'ENDDATA';
yada...yada...yada..
<img src="111">
yada...yada...yada..
<img src="222">
yada...yada...yada..
<h1>Some title</h1>
yada...yada...yada..
<a href="444444">
<img src="333">
yada...yada...yada..
<img src="444">
ENDDATA

my $dom = Mojo::DOM->new($data);
my (@imgs,$seen_h1);
for my $tag ( $dom->find('h1, img')->each ) {
    if ($seen_h1 && $tag->tag eq 'img' && length $tag->attr('src'))
        { push @imgs, $tag }
    elsif ($tag->tag eq 'h1')
        { $seen_h1 = 'true' }
}
print $_->attr('src'),"\n" for @imgs;

__END__

333
444
[download]

You don't even need the intermediate @imgs array and can print directly from the first loop if you like.

Hope this helps,
-- Hauke D

 $ perl -Mojo -le " my $dom = x(b(q{2.html})->slurp); for my $img ( $d
+om->find(q{h1 ~ img})->each ){ print $img->attr(q{src});  } "
333
444
[download]

Nice script, but it fails if between H1 and IMG occur other tags.

my $data = 'yada...yada...yada..
<img src="111">
yada...yada...yada..
<img src="222">
yada...yada...yada..
<h1>Some title</h1>
yada...yada...yada..
<a href="444444">
<img src="333">
yada...yada...yada..
<img src="444">';

my $dom = Mojo::DOM->new($data);
for my $img ( $dom->find(q{h1 ~ img})->each )
        {
         print $img->attr(q{src});
        }
[download]