application/Url.php

   1 <?php
   2 /**
   3  * Converts an array-represented URL to a string
   4  *
   5  * Source: http://php.net/manual/en/function.parse-url.php#106731
   6  *
   7  * @see http://php.net/manual/en/function.parse-url.php
   8  *
   9  * @param array $parsedUrl an array-represented URL
  10  *
  11  * @return string the string representation of the URL
  12  */
  13 function unparse_url($parsedUrl)
  14 {
  15     $scheme   = isset($parsedUrl['scheme']) ? $parsedUrl['scheme'].'://' : '';
  16     $host     = isset($parsedUrl['host']) ? $parsedUrl['host'] : '';
  17     $port     = isset($parsedUrl['port']) ? ':'.$parsedUrl['port'] : '';
  18     $user     = isset($parsedUrl['user']) ? $parsedUrl['user'] : '';
  19     $pass     = isset($parsedUrl['pass']) ? ':'.$parsedUrl['pass']  : '';
  20     $pass     = ($user || $pass) ? "$pass@" : '';
  21     $path     = isset($parsedUrl['path']) ? $parsedUrl['path'] : '';
  22     $query    = isset($parsedUrl['query']) ? '?'.$parsedUrl['query'] : '';
  23     $fragment = isset($parsedUrl['fragment']) ? '#'.$parsedUrl['fragment'] : '';
  24
  25     return "$scheme$user$pass$host$port$path$query$fragment";
  26 }
  27
  28 /**
  29  * Removes undesired query parameters and fragments
  30  *
  31  * @param string url Url to be cleaned
  32  *
  33  * @return string the string representation of this URL after cleanup
  34  */
  35 function cleanup_url($url)
  36 {
  37     $obj_url = new Url($url);
  38     return $obj_url->cleanup();
  39 }
  40
  41 /**
  42  * Get URL scheme.
  43  *
  44  * @param string url Url for which the scheme is requested
  45  *
  46  * @return mixed the URL scheme or false if none is provided.
  47  */
  48 function get_url_scheme($url)
  49 {
  50     $obj_url = new Url($url);
  51     return $obj_url->getScheme();
  52 }
  53
  54 /**
  55  * Adds a trailing slash at the end of URL if necessary.
  56  *
  57  * @param string $url URL to check/edit.
  58  *
  59  * @return string $url URL with a end trailing slash.
  60  */
  61 function add_trailing_slash($url)
  62 {
  63     return $url . (!endsWith($url, '/') ? '/' : '');
  64 }
  65
  66 /**
  67  * Replace not whitelisted protocols by 'http://' from given URL.
  68  *
  69  * @param string $url       URL to clean
  70  * @param array  $protocols List of allowed protocols (aside from http(s)).
  71  *
  72  * @return string URL with allowed protocol
  73  */
  74 function whitelist_protocols($url, $protocols)
  75 {
  76     if (startsWith($url, '?') || startsWith($url, '/')) {
  77         return $url;
  78     }
  79     $protocols = array_merge(['http', 'https'], $protocols);
  80     $protocol = preg_match('#^(\w+):/?/?#', $url, $match);
  81     // Protocol not allowed: we remove it and replace it with http
  82     if ($protocol === 1 && ! in_array($match[1], $protocols)) {
  83         $url = str_replace($match[0], 'http://', $url);
  84     } elseif ($protocol !== 1) {
  85         $url = 'http://' . $url;
  86     }
  87     return $url;
  88 }
  89
  90 /**
  91  * URL representation and cleanup utilities
  92  *
  93  * Form
  94  *   scheme://[username:password@]host[:port][/path][?query][#fragment]
  95  *
  96  * Examples
  97  *   http://username:password@hostname:9090/path?arg1=value1&arg2=value2#anchor
  98  *   https://host.name.tld
  99  *   https://h2.g2/faq/?vendor=hitchhiker&item=guide&dest=galaxy#answer
 100  *
 101  * @see http://www.faqs.org/rfcs/rfc3986.html
 102  */
 103 class Url
 104 {
 105     private static $annoyingQueryParams = array(
 106         // Facebook
 107         'action_object_map=',
 108         'action_ref_map=',
 109         'action_type_map=',
 110         'fb_',
 111         'fb=',
 112         'PHPSESSID=',
 113
 114         // Scoop.it
 115         '__scoop',
 116
 117         // Google Analytics & FeedProxy
 118         'utm_',
 119
 120         // ATInternet
 121         'xtor=',
 122
 123         // Other
 124         'campaign_'
 125     );
 126
 127     private static $annoyingFragments = array(
 128         // ATInternet
 129         'xtor=RSS-',
 130
 131         // Misc.
 132         'tk.rss_all'
 133     );
 134
 135     /*
 136      * URL parts represented as an array
 137      *
 138      * @see http://php.net/parse_url
 139      */
 140     protected $parts;
 141
 142     /**
 143      * Parses a string containing a URL
 144      *
 145      * @param string $url a string containing a URL
 146      */
 147     public function __construct($url)
 148     {
 149         $url = self::cleanupUnparsedUrl(trim($url));
 150         $this->parts = parse_url($url);
 151
 152         if (!empty($url) && empty($this->parts['scheme'])) {
 153             $this->parts['scheme'] = 'http';
 154         }
 155     }
 156
 157     /**
 158      * Clean up URL before it's parsed.
 159      * ie. handle urlencode, url prefixes, etc.
 160      *
 161      * @param string $url URL to clean.
 162      *
 163      * @return string cleaned URL.
 164      */
 165     protected static function cleanupUnparsedUrl($url)
 166     {
 167         return self::removeFirefoxAboutReader($url);
 168     }
 169
 170     /**
 171      * Remove Firefox Reader prefix if it's present.
 172      *
 173      * @param string $input url
 174      *
 175      * @return string cleaned url
 176      */
 177     protected static function removeFirefoxAboutReader($input)
 178     {
 179         $firefoxPrefix = 'about://reader?url=';
 180         if (startsWith($input, $firefoxPrefix)) {
 181             return urldecode(ltrim($input, $firefoxPrefix));
 182         }
 183         return $input;
 184     }
 185
 186     /**
 187      * Returns a string representation of this URL
 188      */
 189     public function toString()
 190     {
 191         return unparse_url($this->parts);
 192     }
 193
 194     /**
 195      * Removes undesired query parameters
 196      */
 197     protected function cleanupQuery()
 198     {
 199         if (! isset($this->parts['query'])) {
 200             return;
 201         }
 202
 203         $queryParams = explode('&', $this->parts['query']);
 204
 205         foreach (self::$annoyingQueryParams as $annoying) {
 206             foreach ($queryParams as $param) {
 207                 if (startsWith($param, $annoying)) {
 208                     $queryParams = array_diff($queryParams, array($param));
 209                     continue;
 210                 }
 211             }
 212         }
 213
 214         if (count($queryParams) == 0) {
 215             unset($this->parts['query']);
 216             return;
 217         }
 218
 219         $this->parts['query'] = implode('&', $queryParams);
 220     }
 221
 222     /**
 223      * Removes undesired fragments
 224      */
 225     protected function cleanupFragment()
 226     {
 227         if (! isset($this->parts['fragment'])) {
 228             return;
 229         }
 230
 231         foreach (self::$annoyingFragments as $annoying) {
 232             if (startsWith($this->parts['fragment'], $annoying)) {
 233                 unset($this->parts['fragment']);
 234                 break;
 235             }
 236         }
 237     }
 238
 239     /**
 240      * Removes undesired query parameters and fragments
 241      *
 242      * @return string the string representation of this URL after cleanup
 243      */
 244     public function cleanup()
 245     {
 246         $this->cleanupQuery();
 247         $this->cleanupFragment();
 248         return $this->toString();
 249     }
 250
 251     /**
 252      * Converts an URL with an International Domain Name host to a ASCII one.
 253      * This requires PHP-intl. If it's not available, just returns this->cleanup().
 254      *
 255      * @return string converted cleaned up URL.
 256      */
 257     public function idnToAscii()
 258     {
 259         $out = $this->cleanup();
 260         if (! function_exists('idn_to_ascii') || ! isset($this->parts['host'])) {
 261             return $out;
 262         }
 263         $asciiHost = idn_to_ascii($this->parts['host'], 0, INTL_IDNA_VARIANT_UTS46);
 264         return str_replace($this->parts['host'], $asciiHost, $out);
 265     }
 266
 267     /**
 268      * Get URL scheme.
 269      *
 270      * @return string the URL scheme or false if none is provided.
 271      */
 272     public function getScheme()
 273     {
 274         if (!isset($this->parts['scheme'])) {
 275             return false;
 276         }
 277         return $this->parts['scheme'];
 278     }
 279
 280     /**
 281      * Get URL host.
 282      *
 283      * @return string the URL host or false if none is provided.
 284      */
 285     public function getHost()
 286     {
 287         if (empty($this->parts['host'])) {
 288             return false;
 289         }
 290         return $this->parts['host'];
 291     }
 292
 293     /**
 294      * Test if the Url is an HTTP one.
 295      *
 296      * @return true is HTTP, false otherwise.
 297      */
 298     public function isHttp()
 299     {
 300         return strpos(strtolower($this->parts['scheme']), 'http') !== false;
 301     }
 302 }