help!

Anon

I'm trying to create a non repeating counter. I am able to log the IP address in one file and in the other file have the number of hits.

Right now I am able to stop counting the reloads. However, if someone else on a different IP address visits my site, and the person who already visited it reloads it, it will count.

I would like to have a whole list of IP address instead of deleting the file and then recreating it during every visit.

So I could have a bunch of IP address in one file eg..

255.255.255.255
123.214.152.102
96.256.32.100
98.25.147.240

and in my other file have the number of hits which would be
4 since there are 4 unique IP addresses.

Now how can I get the script to read all the IP address and create and if statement.

eg..

read the Ip file
$ip = "a list of ip address in the file";

// create some type of array or loop...

// get the current addresss
$addr = getenv("REMOTE_ADDR");

// see if any of them match the current IP address
if ($ip != $addr) {
// log the new IP

// increment the counter

}

Anon

One way would be to read all your IPs into an array, then have it increment through each one to see if that IP is already listed, in which case you would have it continue without incrementing the counter.

The problem would be that it's terribly innefficient and, if it's a big site, would be really...um...innefficient.

Another way is to read your IP list into a variable, then simply do a ereg() looking for that IP.

If it finds the IP, you may wany to increment a special counter elsewhere so you can track how often that visitor is loading up your pages.

If it doesn't find the IP, add the IP, increment the counter, then move along with your script.

Simple and fast, and should work fine.

Anon

Here's another option.

Instead of creating a file to hold your IP records, create a directory.

In that directory, create a separate file for each IP address. The filename can be the same as the IP address.

That means you don't have to open (and lock) the entire dataset in order to manipulate a single value.

Anon

This is a step in the right direction; you absolutely don't want this
in a single (non-database) file because the synchronization will
create an incredible bottleneck.

But I'd be cautious about a solution that created thousands of files
in a single directory, too: you might be better having a separate
directory for each octet. That way, any given directory search
will only be examining at most 256 files, rather than 64k or more.

But now that I mentioned databases, this would be a great place
to employ a write-efficient database such as PostgreSQL. Not
sure how MySql would do due to the table locking that might be
required.

And finally, do you really need this information in real time?
It's almost always available from your server logs, if you
can afford to just generate the stats when you want to
see them.