mirror of
https://github.com/ezyang/htmlpurifier.git
synced 2024-12-24 01:01:53 +00:00
92aabf2b23
Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>
65 lines
2.0 KiB
PHP
65 lines
2.0 KiB
PHP
<?php
|
|
|
|
/**
|
|
* Injector that converts http, https and ftp text URLs to actual links.
|
|
*/
|
|
class HTMLPurifier_Injector_Linkify extends HTMLPurifier_Injector
|
|
{
|
|
/**
|
|
* @type string
|
|
*/
|
|
public $name = 'Linkify';
|
|
|
|
/**
|
|
* @type array
|
|
*/
|
|
public $needed = array('a' => array('href'));
|
|
|
|
/**
|
|
* @param HTMLPurifier_Token $token
|
|
*/
|
|
public function handleText(&$token)
|
|
{
|
|
if (!$this->allowsElement('a')) {
|
|
return;
|
|
}
|
|
|
|
if (strpos($token->data, '://') === false) {
|
|
// our really quick heuristic failed, abort
|
|
// this may not work so well if we want to match things like
|
|
// "google.com", but then again, most people don't
|
|
return;
|
|
}
|
|
|
|
// there is/are URL(s). Let's split the string.
|
|
// We use this regex:
|
|
// https://gist.github.com/gruber/249502
|
|
// but with @cscott's backtracking fix and also
|
|
// the Unicode characters un-Unicodified.
|
|
$bits = preg_split(
|
|
'/\\b((?:[a-z][\\w\\-]+:(?:\\/{1,3}|[a-z0-9%])|www\\d{0,3}[.]|[a-z0-9.\\-]+[.][a-z]{2,4}\\/)(?:[^\\s()<>]|\\((?:[^\\s()<>]|(?:\\([^\\s()<>]+\\)))*\\))+(?:\\((?:[^\\s()<>]|(?:\\([^\\s()<>]+\\)))*\\)|[^\\s`!()\\[\\]{};:\'".,<>?\x{00ab}\x{00bb}\x{201c}\x{201d}\x{2018}\x{2019}]))/iu',
|
|
$token->data, -1, PREG_SPLIT_DELIM_CAPTURE);
|
|
|
|
|
|
$token = array();
|
|
|
|
// $i = index
|
|
// $c = count
|
|
// $l = is link
|
|
for ($i = 0, $c = count($bits), $l = false; $i < $c; $i++, $l = !$l) {
|
|
if (!$l) {
|
|
if ($bits[$i] === '') {
|
|
continue;
|
|
}
|
|
$token[] = new HTMLPurifier_Token_Text($bits[$i]);
|
|
} else {
|
|
$token[] = new HTMLPurifier_Token_Start('a', array('href' => $bits[$i]));
|
|
$token[] = new HTMLPurifier_Token_Text($bits[$i]);
|
|
$token[] = new HTMLPurifier_Token_End('a');
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// vim: et sw=4 sts=4
|