Dan wrote:
> Hey, I posted a while ago on a thread marked "More regex reguired!",
> the solution gave there did part of the job I asked. I need something
> more out of that regex though..
>
> 1) this.is.a.string.to.match.with
> 2) this.is.another.string.to.match.with
> 3) this.is.a.totally.with.different.string
>
> What I want to be able to do, along with what's already been said, is
> to do a search for *.with and it return only strings 1 and 2 (since
> they end in .with). String 3 contains with, but at the moment, this
> code returns that one as well.
Hi Dan.
/\/\/\/\/\/\ Misconception Alert /\/\/\/\/\/\
Regular expressions don't use the star character the same way glob
does. A star is a shorthand for 'none or more of the previous
character'.
To match a star in your string you need to escape it in the regex
with a backslash like this:
/\*/
which will match a string with a star __anywhere in it__. Regexes work
the other way around from globs in that they will match anywhere
unless you specifically say otherwise, so (more special characters)
/^\*/
matches a star at the start of the string, and
/\*$/
matches a star at the end.
> I have this:
>
> $mask =~ s/\*//g;
Totally off-beam here. You're matching lines containing a
literal star character!
> foreach $key (keys %online) {
> if ($spewcount == 50) { # is there more than 50 results?
> # yes!
> $ended = 1;
> last;
> } elsif ($online{$key}->{host} =~ /$mask/) {
> # no!
> print "User with matching host: $key
> ($key!$online{$key}->{ident}\@$online{$key}->{host}\n";
> $spewcount++;
> }
> }
>
> What's the alternative solution so a search of *.with returns only
> strings 1 & 2?
Because a period is another special character in a regex (matches
any single character) you have to escape that too, so the equivalent
of the glob *.with is
/\.with$/
> Secondly, threads now, I have a thread running beside my program
> which runs a timeout sub every second, and checks to see if a
> variable has a timestamp in the past. If it does, it removes the
> information from the hash. But, at the moment, if I set another key
> on the hash, that information doesn't get passed to the sub, and it
> never checks them. How do I get the new information in the hash over
> to the sub so it can check for timeouts on the new information?
Apart from asking you what thread model you're using I can't help
here. It sounds very much as if your child process is working on its
own copy of the hash, so the deletes it is doing aren't echoed in
the parent. But I shall stop speculating and wait for wiser words on
the subjext.
HTH,
Rob
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
