[thelist] Can I capture part of a link?

[Jason Lotito]

Hugh Blair wrote:
 > I can't find the answer to this question.
 >
 > Scene:
 > Person is sent an email with a link to click on that looks
 > something like this: http://www.domain.com/AAA###### or
 > http://www.domain.com?AAA###### where "A" is an alpha
 > character and # is a number, length from 1 to 6. User
 > clicks on that link which takes them to a web page that's
 > a survey. They fill out the survey and an email is sent
 > and/or their entries are added to a database.
 >
 > What I need is to track that AAA###### with their responses.
 > The client doesn't want the user to be required to enter any
 > identifying information. This code will just be used to track
 > the distribution of responses.
 >
 > I can't figure out how to do that with just an HTML page.
 > Just looking in the server log won't be acceptable, the
 > code has to travel with the responses.
 >
 > Can that be done? How? Or will I need to use some Perl
 > routine or a PHP page? The server is 'nix so Perl/PHP are
 > available.
 >
 > I've just not run into this issue before so I'm lost.
 >
 > Thanks in advance,
 >
 > --
 >    Hugh
 >


Rather simple.  Just run a session as the user is on the site via PHP,
and you can track the individual.

domain.com?id=AAA######
<?php

start_session();
$_SESSION['id'] = $_GET['id']

db_query("INSERT INTO tracker ( id, page, datetime ) VALUES (
'{$_SESSION['id']}', '{$_SERVER['PHP_SELF']}', NOW())");

?>

That is rather basic code, needs error checking, hasn't been tested,
etc, but gives you a good idea of a simply way to do this.  Tracking the
users and storing the details in a database.

Jason