From: William Morgan <wmorgan-ruby-talk@...>
Date: 2005-07-18T09:18:29+09:00
Subject: Re: Parsing DNS zonefiles

Excerpts from Dennis Roberts's mail of 14 Jul 2005 (EDT):
> Dan could you explain the difference between using regular expressions
> and a bonafide parser?

In addition to what others have said, if you have to handle
ungrammatical files on a regular basis (I do---xml files), I've found it
typically easier to write regexp parsers that are robust to
ungrammatical input than to try and force a parser to be robust.

Also, if you just want one particular bit of information, regexps can be
a one-line alternative to a many-line grammar.

(There are also theoretical distinctions between the two (parser
 generators win) but I don't find them all that commonplace in practice.)

-- 
William <wmorgan-ruby-talk@masanjin.net>