handling empty variables

[cavey]

I have a page that the user goes to and enters info for a new customer. That routes to a script that checks to make sure the customer doesn't already exist in the system. If they do exists, it routes them back to the page with the form on it and displays an error... I use a script that checks to see if the url contains a value for a specific variable. Ya know like: $exists = mysql_real_escape_string($_GET['exists']); and then an if statement of what to do if it is one, many or null.

For example the url may be

page.php?exists=one

or

page.php?exists=many

and I do different things depending on if there is just one match or many. When the user first goes to the form though there is no value for 'exists' in the url, so the variable is empty, and I get the error:

Notice: Undefined index: exists in /httpdocs/organic/addsubscriber.php on line 22

I realize this is with error_handing turned on to E_ALL but how do i handle this sort of thing the right way?

[bradgrafelman]

The way I define variables in my script is like so:

$exists = (isset($_GET['exists']) ? $_GET['exists'] : FALSE);

This way, I can do three things:

1. I'll thereafter use $exists, which I know is going to be defined one way or the other - no more notices.

2. I can specify the default value (FALSE, NULL, 0, empty string, whatever is most applicable for that variable).

3. I can do some transformation on the user-provided value (ex. verifying the value is of the correct type).

To illustrate points 2 and 3, when I have ID's in the URL that I might possibly use in a SQL query later, I might do this:

$id = (!empty($_GET['id']) ? (int)$_GET['id'] : 0);

In the code snippets above, I'm using the ternary syntax for brevity; the code could also be written as:

if(!empty($_GET['id']))
    $id = (int)$_GET['id'];
else
    $id = 0;

For more information on the ternary operator, see this link.

[cavey]

Perfect dude, thanks. Works well.

[bradgrafelman]

Don't forget to mark this thread resolved.