preg_replace: How to remove it

Please help me how to use it
http://www.phpliveregex.com/p/eyZ

 <form method="GET">
        <div class="row">
            <div class="input-group">
            <input name="url" type="text" class="form-control" placeholder="Example http://www.site.com/id124536/05/how_to_make_a_php_site" required>
            <span class="input-group-btn">
                <button class="btn btn-info" type="submit">DOWNLOAD</button>
                    </span>
        </div><!-- /input-group -->
        </div><!-- /.row -->
    </form>
<br>

<?php
function Videos($url)
{
    $url = str_replace(' ', '', $url);
    if(!empty($url)):
        curl_setopt($curl = curl_init($url), CURLOPT_RETURNTRANSFER, 1);
        preg_match("/flv_url=(.+)url_bigthumb=([^&]+)/i", curl_exec($curl), $infos);
        curl_close($curl);
        return "<a href='".urldecode($infos[1])."'><img class='img-thumbnail' src='".urldecode($infos[2])."' width='300' height='200' border=0 alt='Download'></a>";
    endif;
}

echo Videos($_GET['url']);
?>
      </div>

Code (markup):

http://www. site. com/id124536/05/how_to_make_a_php_site
how to remove when data input

 

Some context would be good.
Why do you need to remove that bit? Will that bit always be in the same place? Do you have any control of the rest of the url (how it's formatted, which domain it is, what ID, and so on?)

 

Yeah, I'm really not getting the whole picture here. The form you included seems to have Jack-all to do with the PHP following it and is riddled with mistakes (like where's your fieldset and label, why the button tag to do a input's job, what's with the extra DIV for nothing...) and it's extremely unclear what your CURL is returning or how it relates to that URI you posted after the code.

 

Please sir see PM

 

Ok, seeing the subject matter thank you for not publicly putting that in the open area.

As to your problem, you're looking at this the wrong way.The URI you are having problems with are in fact a 301 redirect. So instead of returning the content you want, it's returning a new "Location" header. By default cURL will not try to follow redirects. You can TRY adding:
curl_setopt($curl, CURLOPT_FOLLOWLOCATION, 1);

Between your first curl_setopt and when you run the curl_exec, but that will not work if PHP's incorrectly named "safe_mode" is active (which there's no reason to have that on any time after 2003) OR if an open_basedir is set (which there are legitimate reasons to use).

Generally speaking it's sloppy/bad practice to blindly trust that a cURL operation is going to return a "200 ok" response, which is why blindly dumping the result of that exec into a function is just asking for it to break. The method I like to use is to add a loop trapping 300 series for redirects, and bombing if any other non-200 response is passed. Error handling is VERY important, more so if you're going to be extracting direct DL links from a website's flashvars.

That goes something like this:

<snip>use code in next post</snip>

That's untested but should give you an idea of the depth of testing needed to handle all those possible error conditions properly. If I have time later when I'm back at the workdesk I'll take a stab at seeing about dotting the t's and crossing the i's on that... or, uhm, whatever.

Really trying to parse out the value instead of following the redirect? That's just asking for it to break again moving forward.

 

Ok yeah, a few typo's... one uneccessary (), two $ before function names (doh), and one double quote that should have been a single. My bad, what I get for writing code in the post box here on the forums on the laptop. Get to the workstation and it's like "well DUH!"

This should work:

/* 
	NOTE redirect_url is only valid in PHP 5.3.7/newer, for older PHP
	support you'd need to parse the content of the response from cURL
	
	Though really at that point, just upgrade to a PHP from THIS century!
*/

function getUri($uri) {

	do {
	
		$curl = curl_init($uri);
		curl_setopt($curl, CURLOPT_RETURNTRANSFER, 1);
		// set it to TRY and handle it automagically
		curl_setopt($curl, CURLOPT_FOLLOWLOCATION, 1);
		
		// break the erorr condition handling into steps
		
		if (curl_errno($curl)) {
			$result = [ 'curlError' => $curl_error($curl) ];
			curl_close($curl);
			return $result;
		}
		
		$result = [ 'data' => curl_exec($curl) ];
		if (curl_errno($curl)) {
			$result['curlError'] = $curl_error($curl);
			curl_close($curl);
			return $result;
		}
		
		$result['curlError'] = false;
		
		$result['info'] = curl_getinfo($curl);
		// use switch so if we want to add more handlers it's simpler
		switch ($result['info']['http_code']) {
			case 301:
			case 302:
				$uri = $result['info']['redirect_url'];
				break;
			default:
				curl_close($curl);
				return $result;
		}
		
	} while (true);
	
} // function getUri


function videoFromUrl($url) {

	if (!isset($url) || empty($url = str_replace(' ', '', $url)))
		return '
			<div class="error">
				No File Specified
			</div>';
			
	$result = getUri($url);
	
	if ($result['curlError']) return '
			<div class="error">
				Error "' . $result['curlError'] . '"
			</div>';
	
	if ($result['info']['http_code'] == 200) {
		preg_match(
			"/flv_url=(.+)url_bigthumb=([^&]+)/i",
			$result['data'],
			$infos
		);
		return '
			<a href="' . htmlspecialchars(urldecode($infos[1])) . '">
				<img
					src="' . htmlspecialchars(urldecode($infos[2])) . '"
					class="img-thumbnail"
					width="300" height="200" 
					alt="Download"
				>
			</a>';
	}
	
	return '
			<div class="error">
				Unable to recover file. HTTP error code ' .
				$result['info']['http_code'] . '
			</div>';
			
} // function videoFromUrl

Code (markup):

Just call "videoFromUrl" where you were using your "Videos" function. Will give you much more robust error handling and should automatically deal with that redirect... whcih is good. Trying to parse someone else's URL that's redirecting yourself could bite you later if they decide to change the redirect AGAIN. You're better off just following how HTTP header responses are supposed to work in the first place.

Worked fine here calling the URL you PM'd with the extra directory in it.