What I would like to be able to do, is define a section of a page, containing several different _tag's. The HTML in question is:

<table border=0 cellspacing=2 width=100%>
...
<tr>
  <td class=td1 id=centered valign=top><A HREF=http://SOME_LINK>1   </
+A></td>
  <td class=td1 valign=top>
  <script language="javascript">
  var recordLink="<A HREF=SOME_LINK>1   </A>";
  recordLink = recordLink.substring(0, recordLink.indexOf(">"));
  document.write(recordLink);
  document.write(">");
  </script>
  SOME_TEXT</a></td>
<!--
  <td class=td1 valign=top>SOME_DATA</td>
-->
  </tr>
[download]

, and I'm currently reading it with:

my @comments = $tree->look_down ('_tag', '~comment');
my @lnk_details = $tree->look_down ('_tag', 'script');
[download]

The problem with this is that I have no (easy) control over checking whether there actually exist a "comment" part together with the link (contained in the "script"-tag). So I'm reading both the arrays of data together, without knowing whether they have the same number of elements, or even if one element in "comments" is paired with the element in the same position in "lnk_details".

Is there a way to make TreeBuilder extract sections of an HTML-page defined by consecutive tags, and pair those together for further parsing?