The easiest way to allow any wildcard host matching is a hash table.

Even without updating, you still have $regsub.

Lucius did mention that, but at the same time ini files are slower. Generally in this situation a hash table is suggested, but for this type of script performance isn't normally an issue anyway.

Some of the text in this script suggests that it's created and/or used by someone who is a little abusive while thinking they're only being funny, to be honest.
Also, most networks running Anope should at least have alias commands like /NickServ
You should use those if possible since they only work when the services are online, which is much more secure for things like identifying with a password. :)

Yes, doing it once instead of twice would make sense.

Sorry, Jethro_, the reason I said you might as well check $nick == NickHere is because the code you suggested was no more secure than that. If you are comparing the current address of the person triggering the event (nick!ident@host) against, for example, $address(NickHere,2), the script will be comparing the triggered address with the address of whoever is currently using NickHere.
With that being the case, anyone using NickHere would validate correctly as the desired person.

No, since a timer is not called from within the script you have no way of knowing whether the dialog is still open.

Jethro_, if you are going to use that first method you might as well check for $nick == NickHere.
The second method is the same deal but it looks at the current address of the nickname compared to the address of the person in the event.
Also, you don't need the iswm operator if you have no wildcard comparison, which is the case when you compare one full address to another full address.

AndroidKnight, use this.

if (*!camron@jackdaniels.aint.got.shoot.on.me iswm $fulladdress) {

Or you can use

if (Nick!*ident@host iswm $fulladdress) {

..Which is what they want.

IllogicTC, should the dialog somehow manage to close without unsetting the timer due to some error, you would be flooded by errors at what should be once every millisecond, though it probably would be much less.
Even so, I'm fine with the check to make sure that the dialog is open within a command called by a timer.

What is your basis for calling it invalid other than the fact that it's never used?
I looked at the IP standard as well as IP address allocation and nothing mentioned 0.0.0.0.
The only mention(s) of it were to say that an IP address is x.x.x.x where x is between 0 and 255.

With 0.0.0.0 alone, you must decide whether your IP address check is based on whether the values are valid (all octets between 0 and 255) or whether the IP address is considered valid (since 0.0.0.0 is never used for anything).
I would personally only change that if 0.0.0.0 is considered invalid by the IP standard.

Your regex is completely invalid.
It matches zero or more number characters (3 times), then zero or more number characters again.
It will be true if you supply a blank string or any reasonable amount of random numbers.

The regular expression for accurately matching an IP address is far more complex.

No.. Why would it?

Thanks napa. :)

I decided to share three small functions I made for IP addresses the other day in a few minutes. Their use is as follows:

It seems my test for the \Q..\E was not correct to begin with. Thanks for making me look at it again, jaytea. :P

Yes, I understand the purpose of escaping the \E to prevent your \Q..\E from breaking. It can also still be broken by a backward slash preceding the \E.
Feel free to test a string "\E " against regex "\Q\E\E" and "\Q\\E\E" which are your options here.

I would consider that sort of thing to be a tacky addon if it were to be included within this script itself, considering the complexity difference.

Jethro_ appears to have a misleading regular expression here, but upon closer inspection it seems to be (for the most part) a readability issue.
There's no need to end an escaping \Q...\E, however, to escape something. \E would suffice instead of \E\E\Q.

A similar loop is this:

while ($0) { /cs op # $1 | tokenize 32 $2- }

GET and POST are only methods of requesting/sending data. Your regex is used when receiving it.

Jethro_, probably referring to the fact that you have to join each line together to form a pseudo-switch which is actually just a single command.

Won't that fail if you use any non-word character as your case criteria? (\w+)

That explains the old programming style.

Converting the text in the reverse order, you mean. Probably easier to understand that way.

/echo -q is an echo which doesn't show up if you put the . in front of your alias call.
/.msg hides the echo of your message.
If you created your own /noop alias which pointed to a silenced (e.g. /.noop2) alias then you could use echo -q but that'd be a pain anyway.
The other alternative, of course, is to just use a dud variable in its place.

You don't even need to "negotiate" to $10 - there is a valid discount for $10 still, which has been around for years. For the amount of time I've used it, it's worth the $10.
http://www.mirc.co.uk/register.php?coupon=MIRC-SWV0-MNKL

i think someone should make a MSL program that makes the sockets for you, you type in the triggers and websites - Idea!
That's actually really easy up until the point where you decide how each website must be matched. All websites are arranged differently and sometimes the data comes through on more than one message, making that more difficult.
One solution which can't really fail is to always use a file and binvars, but that will make unnecessarily complicated scripts for most applications.

Jethro_ you /set %nicks once as well as /var %nicks twice. You should probably fix that up.