library/HTMLPurifier/URIParser.php

   1 <?php
   2
   3 /**
   4  * Parses a URI into the components and fragment identifier as specified
   5  * by RFC 3986.
   6  */
   7 class HTMLPurifier_URIParser
   8 {
   9
  10     /**
  11      * Instance of HTMLPurifier_PercentEncoder to do normalization with.
  12      */
  13     protected $percentEncoder;
  14
  15     public function __construct()
  16     {
  17         $this->percentEncoder = new HTMLPurifier_PercentEncoder();
  18     }
  19
  20     /**
  21      * Parses a URI.
  22      * @param $uri string URI to parse
  23      * @return HTMLPurifier_URI representation of URI. This representation has
  24      *         not been validated yet and may not conform to RFC.
  25      */
  26     public function parse($uri)
  27     {
  28         $uri = $this->percentEncoder->normalize($uri);
  29
  30         // Regexp is as per Appendix B.
  31         // Note that ["<>] are an addition to the RFC's recommended
  32         // characters, because they represent external delimeters.
  33         $r_URI = '!'.
  34             '(([a-zA-Z0-9\.\+\-]+):)?'. // 2. Scheme
  35             '(//([^/?#"<>]*))?'. // 4. Authority
  36             '([^?#"<>]*)'.       // 5. Path
  37             '(\?([^#"<>]*))?'.   // 7. Query
  38             '(#([^"<>]*))?'.     // 8. Fragment
  39             '!';
  40
  41         $matches = array();
  42         $result = preg_match($r_URI, $uri, $matches);
  43
  44         if (!$result) return false; // *really* invalid URI
  45
  46         // seperate out parts
  47         $scheme     = !empty($matches[1]) ? $matches[2] : null;
  48         $authority  = !empty($matches[3]) ? $matches[4] : null;
  49         $path       = $matches[5]; // always present, can be empty
  50         $query      = !empty($matches[6]) ? $matches[7] : null;
  51         $fragment   = !empty($matches[8]) ? $matches[9] : null;
  52
  53         // further parse authority
  54         if ($authority !== null) {
  55             $r_authority = "/^((.+?)@)?(\[[^\]]+\]|[^:]*)(:(\d*))?/";
  56             $matches = array();
  57             preg_match($r_authority, $authority, $matches);
  58             $userinfo   = !empty($matches[1]) ? $matches[2] : null;
  59             $host       = !empty($matches[3]) ? $matches[3] : '';
  60             $port       = !empty($matches[4]) ? (int) $matches[5] : null;
  61         } else {
  62             $port = $host = $userinfo = null;
  63         }
  64
  65         return new HTMLPurifier_URI(
  66             $scheme, $userinfo, $host, $port, $path, $query, $fragment);
  67     }
  68
  69 }
  70
  71 // vim: et sw=4 sts=4