PHP - Get Website Title From User Site Input - php

Question

I'm trying to get the title of a website that is entered by the user.
Text input: website link, entered by user is sent to the server via AJAX.
The user can input anything: an actual existing link, or just single word, or something weird like 'po392#*#8'
Here is a part of my PHP script:
// Make sure the url is on another host
if(substr($url, 0, 7) !== "http://" AND substr($url, 0, 8) !== "https://") {
$url = "http://".$url;
}
// Extra confirmation for security
if (filter_var($url, FILTER_VALIDATE_URL, FILTER_FLAG_HOST_REQUIRED)) {
$urlIsValid = "1";
} else {
$urlIsValid = "0";
}
// Make sure there is a dot in the url
if (strpos($url, '.') !== false) {
$urlIsValid = "1";
} else {
$urlIsValid = "0";
}
// Retrieve title if no title is entered
if($title == "" AND $urlIsValid == "1") {
function get_http_response_code($theURL) {
$headers = get_headers($theURL);
if($headers) {
return substr($headers[0], 9, 3);
} else {
return 'error';
}
}
if(get_http_response_code($url) != "200") {
$urlIsValid = "0";
} else {
$file = file_get_contents($url);
$res = preg_match("/<title>(.*)<\/title>/siU", $file, $title_matches);
if($res === 1) {
$title = preg_replace('/\s+/', ' ', $title_matches[1]);
$title = trim($title);
$title = addslashes($title);
}
// If title is still empty, make title the url
if($title == "") {
$title = $url;
}
}
}
However, there are still errors occuring in this script.
It works perfectly if an existing url as 'https://www.youtube.com/watch?v=eB1HfI-nIRg' is entered and when a non-existing page is entered as 'https://www.youtube.com/watch?v=NON-EXISTING', but it doesn't work when the users enters something like 'twitter.com' (without http) or something like 'yikes'.
I tried literally everthing: cUrl, DomDocument...
The problem is that when an invalid link is entered, the ajax call never completes (it keeps loading), while it should $urlIsValid = "0" whenever an error occurs.
I hope someone can help you - it's appreciated.
Nathan

Answer 1

You have a relatively simple problem but your solution is too complex and also buggy.
These are the problems that I've identified with your code:
// Make sure the url is on another host
if(substr($url, 0, 7) !== "http://" AND substr($url, 0, 8) !== "https://") {
$url = "http://".$url;
}
You won't make sure that that possible url is on another host that way (it could be localhost). You should remove this code.
// Make sure there is a dot in the url
if (strpos($url, '.') !== false) {
$urlIsValid = "1";
} else {
$urlIsValid = "0";
}
This code overwrites the code above it, where you validate that the string is indeed a valid URL, so remove it.
The definition of the additional function get_http_response_code is pointless. You could use only file_get_contents to get the HTML of the remote page and check it against false to detect the error.
Also, from your code I conclude that, if the (external to context) variable $title is empty then you won't execute any external fetch so why not check it first?
To sum it up, your code should look something like this:
if('' === $title && filter_var($url, FILTER_VALIDATE_URL))
{
//# means we suppress warnings as we won't need them
//this could be done with error_reporting(0) or similar side-effect method
$html = getContentsFromUrl($url);
if(false !== $html && preg_match("/<title>(.*)<\/title>/siU", $file, $title_matches))
{
$title = preg_replace('/\s+/', ' ', $title_matches[1]);
$title = trim($title);
$title = addslashes($title);
}
// If title is still empty, make title the url
if($title == "") {
$title = $url;
}
}
function getContentsFromUrl($url)
{
//if not full/complete url
if(!preg_match('#^https?://#ims', $url))
{
$completeUrl = 'http://' . $url;
$result = #file_get_contents($completeUrl);
if(false !== $result)
{
return $result;
}
//we try with https://
$url = 'https://' . $url;
}
return #file_get_contents($url);
}