Regexps are bad at very complicated cases, especially when things are nested. The typical example is HTML: it's easy enough to make a simple regexp to grab tags, but to get it really right and account for all cases (comments, greater than signs within quotes within HTML tags, multiline tags, etc.) you outgrow the usefullness of a regular expression and really need something that will actually parse the string (or file, whatever) that contains the HTML. Luckily, there are modules that will do just that for you.