I used to work in the "E Discovery" business where millions of documents where compared and screened for similarities, key words, and other patterns. We had tools such as Equivio that would do all of that and tools such a Lucene for certain full text searches and other pattern recognition. Those tools may have used Regex but it seems to may that with the speed they actually had, they probably used something a whole lot more directed and effecient and that makes a whole lot of sense because code to do something specific is almost always a whole lot faster than generic code.

I appreciate the thoughts you folks have written down but, except for possible ad hoc one off searches where performance doesn't really seem to matter to some folks, I'm not seeing much that I like in Regex. I'll keep an open mind about certain aspects of it but I'm not seeing it as a valuable large volume processing tool.

Heh... and you might be able to cut yourself or the guy standing behind you when you swing a 4 handled, 5 headed axe and it'll look really cool on the shelf, but you're not going to cut much wood with it. 😛