admin@geocodenet.com wrote :

When I use this regular expression:

^(*.)a(*.)

Isn't it ^(.*)a(.*) rather ?

Or am I getting exactly what I'm supposed to be getting here?

This is the normal behaviour. You don't get all possibilities of how the string could be matched, you only get one. In your case you have two couples of parentheses, so you capture two strings. Your "shouldn't I get 5 strings" doesn't really make sense.

1==>M 2==>nagement

This is what you get if you use (.*?) (ungreedy)

1==>Man 2==>gement

This is what you get if you use (.*) (greedy)

admin@geocodenet.com wrote:

Hi all,

I'm sure this is just a configuration issue, but I'm not sure where my config is wrong. I'm using boost_regex_search, and capturing subgroups. When I use this regular expression:

^(*.)a(*.)

On this string: "Management"

You'll end up with a syntax error, because the asterisk goes after the dot. But that's probably just a typo in the post.

I end up with the following matches:

0==> Management 1==> Man 2==> gement

Should this be returning the following matches?

0==>Management 1==>M 2==>nagement 3==>Management 4==>Man 5==>gement

Or am I getting exactly what I'm supposed to be getting here?

You're getting exactly what you're supposed to get. Neither regex_search nor any other regex function attempts to find more than one way a given string could match the regex. Thus, since * is greedy, the first .* will always match "Man", never just "M" unless the remainder of the regex cannot be satisfied otherwise.

Also, the default syntax is Perl syntax, correct?

Correct. Sebastian Redl