Without getting into a discussion of exactly which Perl modules to use, I would approach this problem in this way (using existing very-high-level modules throughout):

1. Parse the HTML file into an XML data-structure.
2. Use XPath-expressions to search for what you want. (It helps considerably if there is an "id" field.)
3. Iterate through the search-list, extracting the necessary parts (using XPath again if necessary) and processing them.

Strange as it may seem, “this sort of task should be as easy as falling off a log.” No amount of fancy, complicated programming should be needed on your part:   all of the heavy-lifting has been done for you.

Depending on your actual requirements, it might be possible to do the task with no programming at all. XSL stylesheets are an extremely powerful transformational tool.