In PHP, how can I replicate the expand/contract feature for Tinyurl's as on search.twitter.com?
If you want to find out where a tinyurl is going, use fsockopen to get a connection to tinyurl.com on port 80, and send it an HTTP request like this
GET /dmsfm HTTP/1.0
Host: tinyurl.com
The response you get back will look like
HTTP/1.0 301 Moved Permanently
Connection: close
X-Powered-By: PHP/5.2.6
Location: http://en.wikipedia.org/wiki/TinyURL
Content-type: text/html
Content-Length: 0
Date: Mon, 15 Sep 2008 12:29:04 GMT
Server: TinyURL/1.6
example code...
<?php
$tinyurl="dmsfm";
$fp = fsockopen("tinyurl.com", 80, $errno, $errstr, 30);
if (!$fp) {
echo "$errstr ($errno)<br />\n";
} else {
$out = "GET /$tinyurl HTTP/1.0\r\n";
$out .= "Host: tinyurl.com\r\n";
$out .= "Connection: Close\r\n\r\n";
$response="";
fwrite($fp, $out);
while (!feof($fp)) {
$response.=fgets($fp, 128);
}
fclose($fp);
//now parse the Location: header out of the response
}
?>
And here is how to contract an arbitrary URL using the TinyURL API. The general call pattern goes like this, it's a simple HTTP request with parameters:
http://tinyurl.com/api-create.php?url=http://insertyourstuffhere.com
This will return the corresponding TinyURL for http://insertyourstuffhere.com. In PHP, you can wrap this in an fsockopen() call or, for convenience, just use the file() function to retrieve it:
function make_tinyurl($longurl)
{
return(implode('', file(
'http://tinyurl.com/api-create.php?url='.urlencode($longurl))));
}
// make an example call
print(make_tinyurl('http://www.joelonsoftware.com/items/2008/09/15.html'));
Another simple and easy way:
<?php
function getTinyUrl($url) {
return file_get_contents('http://tinyurl.com/api-create.php?url='.$url);
}
?>
As people have answered programatically how to create and resolve tinyurl.com redirects, I'd like to (strongly) suggest something: caching.
In the twitter example, if every time you clicked the "expand" button, it did an XmlHTTPRequest to, say, /api/resolve_tinyurl/http://tinyurl.com/abcd
, then the server created a HTTP connection to tinyurl.com, and inspected the header - it would destroy both twitter and tinyurl's servers..
An infinitely more sensible method would be to do something like this Python'y pseudo-code..
def resolve_tinyurl(url):
key = md5( url.lower_case() )
if cache.has_key(key)
return cache[md5]
else:
resolved = query_tinyurl(url)
cache[key] = resolved
return resolved
Where cache
's items magically get backed up into memory, and/or a file, and query_tinyurl() works as Paul Dixon's answer does.
If you just want the location, then do a HEAD request instead of GET.
$tinyurl = 'http://tinyurl.com/3fvbx8';
$context = stream_context_create(array('http' => array('method' => 'HEAD')));
$response = file_get_contents($tinyurl, null, $context);
$location = '';
foreach ($http_response_header as $header) {
if (strpos($header, 'Location:') === 0) {
$location = trim(strrchr($header, ' '));
break;
}
}
echo $location;
// http://www.pingdom.com/reports/vb1395a6sww3/check_overview/?name=twitter.com%2Fhome
Here is another way to decode short urls via CURL library:
function doShortURLDecode($url) {
$ch = @curl_init($url);
@curl_setopt($ch, CURLOPT_HEADER, TRUE);
@curl_setopt($ch, CURLOPT_NOBODY, TRUE);
@curl_setopt($ch, CURLOPT_FOLLOWLOCATION, FALSE);
@curl_setopt($ch, CURLOPT_RETURNTRANSFER, TRUE);
$response = @curl_exec($ch);
preg_match('/Location: (.*)\n/', $response, $a);
if (!isset($a[1])) return $url;
return $a[1];
}
It's described here.
In PHP there is also a get_headers function that can be used to decode tiny urls.