RE: Use of regular expression

I completely understand.

Actually, I should be careful about the word "word" here. It will find the
"word" rick, but it will also find bricks and brick and ricky (for example).
In regular-expression speak, to find the *word* rick, you could use:

'\brick\b'

I don't know a way to do the same in SQL, especially when word boundaries
can be any of a number of characters: <space> <comma> <period> <apost>
<quote> <semicolon> et cetera.

Dennis Lovelady
http://www.linkedin.com/in/dennislovelady
--
CONGRESS.SYS corrupted. Reboot Washington DC <Y/N>?

-----Original Message-----
From: rpg400-l-bounces@xxxxxxxxxxxx [mailto:rpg400-l-
bounces@xxxxxxxxxxxx] On Behalf Of Rick.Chevalier@xxxxxxxxxxxxxxx
Sent: Monday, July 19, 2010 9:06 AM
To: rpg400-l@xxxxxxxxxxxx
Subject: RE: Use of regular expression

Dennis,

Thanks for clarifying the regular expression. As usual, I have
overcomplicated it.

To search for the word rick anywhere with a string, your exact regular
expression would be (are you ready for this complexity?) 'rick'. That
will find the string 'rick' anywhere within the search field.

This is exactly the same as your r{1}i{1}... syntax, and _so_ much more
legible. While the SQL suggestion is an option, it is likely to add to
your run time. But (like all performance topics) it depends on what
else is happening in your code.

I'm wondering if using regular expressions is a good way to handle a
word search. What I need is an exact match of a series of characters,
say 'rick', in a character field. I have only used regular
expressions a couple of times and the syntax still gets me confused.
I think I see a way to do this with with something like

r{1}i{1}c{1}k{1}.

I know I can do this with the %SCAN op code. The reason I'm thinking
about using regular expressions is the search will be over a table of
close to 1 billion records and I'm looking for efficiencies any place
I can find them. If an expression can be created I want to do some
comparisons between using a regular expression and %SCAN to see which
is really the best performer.