Hey. I'm trying to match this line with regex in a perl script:

Example line:
[STATSME] "mona<810><1657282><CT>" "weapon mac10" "shots 68" "hits 6" "headshots 0"

Every line will look like:

[STATSME] "NAMESTUFF" "weapon SOMEWEAPON" "shots NUM1" "hits NUM2" "headshots NUM3"

The bold stuff is what I need to get out of it.

The string I tried (that didn't work) is:

($s_output =~ /^\[STATSME\] "([^"]+)" "weapon ([^"]+)" "shots ([^"]+)" "hits ([^"]+)" "headshots ([^"]+)"$/)

I also tried replacing each ([^"]+) with (.*) and (.+) but no luck. And I tried replacing each space with a \s.
Any ideas as to what I'm doin wrong??

Thanks!!!
Justin

I think you have a whitespace problem, this program works as expected and isn't that different than your regexp:
#!/usr/bin/perl
while (<DATA>)
{
print "name $1 weapon $2 shots $3 hits $4 headshots $5\n"
if /^\[STATSME\]\s*\"([^\"]+)\".+weapon([^\"]+)\".+shots\s(\d+).+hits\s(\d+).+headshots\s+(\d+)/;
}

__DATA__
[STATSME] "mona<810><1657282><CT>" "weapon mac10" "shots 68" "hits 6" "headshots 0"
[STATSME] "lisa<810><1657283><CT>" "weapon flower" "shots 1" "hits 1" "headshots 1"

One thing I did was not trust what was between the quotes so I used the inneffecient .+ as in .+weapon. What I would do if I where you would be to simplify the code and then (in my example) take the if statement down to:
if /^\[STATSME\]\s*\"([^\"]+)/;
then if that works go up to:
if /^\[STATSME\]\s*\"([^\"]+)\".+weapon([^\"]+)/;

So you can see where your regexp starts going of in the weeds.

Thanks for the reply! I'll try that... I'm afraid that might not be the problem though.

After i posted, i tried matching /STATSME/ and even THAT didn't work. The really odd thing is that there were other lines in the logs that had STATSME typed in them, and it DID match those.

I'm kinda baffled

Justin

Did you pull the datafile up in something like emacs so you can see if there are some bizzarre characters in the file?
Has the file ever be DOSified? Perhaps you have some sort of line termination problem and the whole file is slurped in with one <> operator.

Perhaps you could do something like this:
less datafile > datafile2 to scrape out the crap.
You are better off not doing this long term.

Well, to make things more confusing... I have the same exact regex line in a different perl script and it works over there with the same logs :-o.

The program doesn't get it's stats from text files... it listens on a port and waits for another program (the halflife engine) to send it data. I *can* have it important logs manually, but I still get the same error.

The OTHER program (that the regex line works in) is similiar, but reads in the text files on it's own. I don't think that has anything to do with it though....

Maybe I'll try:

/S(.*)T(.*)A(.*)T(.*)S(.*)M(.*)E/

Haha. ::sigh:: signs of desperation....

If you're good with perl (and don't mind) I can send you the script and some sample logs. It's simple, it just doesn't make sense why it doesn't work.

Justin

Sure you can send me the program and the log file or just the port and IP of the server if it is open. My email address is lavelle at myrdin dot com.

What game is this for?

It's for counter-strike.

Lavelle, the person who wrote the program that creates the lines that I had to parse altered the logging format into something that was easier for me to parse.

I'm still dying to know why I couldn't get the regex to work, but it's not worth the trouble

Thanks for the offer though!

Justin