[hobbit] PCRE reference with examples

[Henrik Stoerner]

On Thu, Oct 13, 2005 at 10:50:26AM -0500, Rich Smrcina wrote:
> Ok, that's great info.  Thanks!
> 
> Based on that I'm ending up with an expression like:
> 
> HOST=$va[\d\w]*$
> 
> To match any host beginning with 'va' that has zero or more alphanumeric 
> characters after it.  Is there an easier way to write this?  If not 
> that's fine, I just want to make sure I'm using it correctly.

I'd say
  HOST=%^va[\d]*$

You need the "%" first, to tell Hobbit that what comes next is a regexp.
Then the "^" means that the expression must match at the beginning of
the string. "\w" is whitespace ? Think so - in that case it is not
needed, since Hobbit will never match with any whitepace in the
hostname.

> Another poster (Asif Iqbal) posted the MAIL directives with multiple 
> recipients.  Does that work like he posted?  Can the multiple email 
> recipients be put into a macro?  Consider the following:
> 
> $pg-tom=(tomsemail at somewhere.com|tomscell at wireless.com)
> HOST=tomshost
> 	MAIL $pg-tom DURATION=5m COLOR=red

What comes after the MAIL keyword is passed directly to your "mail"
utility. So it would probably be better to have

   $pg-tom=tomsemail at somewhere.com,tomscell at wireless.com

if your "mail" tool supports listing multiple recipients separated by
commas. Some dont - then you can use a short-hand notation like:

   $pg-tom=tomsemail at somewhere.com tomscell at wireless.com

and separate them by spaces - Hobbit will handle this as if you had
multiple "MAIL ..." lines with each of the mail recipients. So it sends
out the mail in separate e-mails, instead of one e-mail to all of the
recipients. (Same net effect).


One more thing: I'm sure "DURATION=5m" is not what you want. Make that
"DURATION>5m".



Regards,
Henrik