Validating URLs --thebest way is ?

Anon

Hello fellow coders.

I was just wondering what might be the best way to validate an URL. I woudl like to code a piece of equipment, that allows people to add links into a mySQL databse. I am not really great with regex, so I was wondering what woudl be actually the best way to do this ?

Either check the URL syntax using regex.

connect to the given URL and then parse the page through some function that looks if there is the occurance of 404 (maybe int he header? ).

How would you go about this problem, I am nbot exactly a Pro with php and I appreciate any help I may get.

Thank you.

Anon

Try fopen():

$addr="http://whatever.org/foo.jpg";
$fp = fopen($addr,"r");
if (!$fp) {
//your choice of things to do
}

Anon

Thank you PetrB but that does not quite solve my initial problem. I am wondering what the best way is. When I use the fopen on something like, let us say

http:///www.you.net/tron/index.html

and the server of you.net returns a 404 does that mean the fopen failed ? It read data back so I guess it would not be validated as failure, right ?

I am espcially weary of those who will enter links that have no page to link to such as the example above.

Thank you for the idea though

Anon

Maybe fsockopen() is the function you are looking for -- I got this at zend, and have not tried it yet, but maybe you can adapt it to your needs:

<?

//idban secandri July 2000 please distribute is as GPL
//
//greeting as usual to desy, nadya, my friend at satunet group!

$phpnet = fsockopen("www.php.net", 80, &$errno, &$errstr, 30);

if(!$phpnet) {

//change with your custom messages
echo "php.net down!!\n"; }

else {

echo("<a href=\"http://www.php.net\">php.net</a>");

}

Anon

There is also this from...
http://www.php.net/manual/function.header.php

Reading header of URL
php does not support reading header function. you must open socket directly. Following is the function to read URL header. (see <a href="http://www.phpwizard.net/phpTidbits/">phpTidbits</a>
<PRE>
function get_headers($host, $path = "/")
{
$fp = fsockopen ("$host", 80, &$errnr, &$errstr) or die("$errno: $errstr");
fputs($fp,"GET $path HTTP/1.0\n\n");
while (!$end)
{
$line = fgets($fp, 2048);
if (trim($line) == "")
$end = true;
else
echo $line;
}
fclose($fp);
}
</PRE>

Maybe to look for specifics in the header like:

<PRE>
header("http/1.0 404 Not Found");
</PRE>

causes a 500 error, at least in the version of Apache I'm using (1.3.3). Instead, use the Status header:

<PRE>
header("Status: 404 Not Found");
</PRE>

Anon

Very good thank you 😉)

That is a great function.

What I would like to do now is this:

1.) check the URL if the syntax is correct using regex.

2.) check wheter the domain exists by trying to get a host lookup for the domain.

3.) try to connect to the site seing if it is up and if I get a valid file or a 404 using your function

4.) checking if the file exists for example

http://www.you.net/index.html

Thank oyu very much, if anyone has other suggestions, let me know. I plan on making this publically available

Anon

Dunno why you are checking for the right syntax and doing a lookup anyway.

Just trying to open the page would do it, and makes the first two obsolate?

Anon

If I do not check for the syntax then someone could enter ht:/w.he.d/ and my script would start using the connect routine. That is a waist of system ressources, since such a syntax will never work out correctly.

The same goes for the doamin name. if someone enters a non existant doamin name and I start checking it, then the system will open a connection and then exit with no hos lookup, but I do not really know why the connect failed.

Another waist ogf system resource even though I admit, the check function could incooperate such a lookup function..good though, thank you