From: Mark Thomas <mrt@...>
Date: 2008-02-29T03:54:55+09:00
Subject: Re: Need a regex searching html code

A regex will break too easily when parsing HTML. A real parser will do
a much better job, and often be more concise and readable, too.

This does what you want:

#-------
require 'rubygems'
require 'hpricot'
@doc = Hpricot(html) # or Hpricot(open("filename"))

def find(term)
  @doc.search("//div[@class='info']").each do |info|
    header = info.search("h5").remove
    if header.inner_text == term
      puts info.inner_html
    end
  end
end
#-------

> find("Plot Outline:")
John McClane takes on an Internet-based terrorist organization who is
systematically shutting down the United States. <a href="http://
www.imdb.com/title/tt0337978/plotsummary" class="tn15more
inline" onclick="(new Image()).src='/rg/title-tease/plotsummary/images/
b.gif?
link=/title/tt0337978/plotsummary';">more</a>

Mark