Parse.php

Go to the documentation of this file.

<?php

require_once "Auth/OpenID.php";

class Auth_OpenID_Parse {

    var $_re_flags = "si";

    var $_removed_re =
           "<!--.*?-->|<!\[CDATA\[.*?\]\]>|<script\b(?!:)[^>]*>.*?<\/script>";

    var $_tag_expr = "<%s\b(?!:)([^>]*?)(?:\/>|>(.*)(?:<\/?%s\s*>|\Z))";

    var $_attr_find = '\b(\w+)=("[^"]*"|\'[^\']*\'|[^\'"\s\/<>]+)';

    var $_open_tag_expr = "<%s\b";
    var $_close_tag_expr = "<((\/%s\b)|(%s[^>\/]*\/))>";

    function Auth_OpenID_Parse()
    {
        $this->_link_find = sprintf("/<link\b(?!:)([^>]*)(?!<)>/%s",
                                    $this->_re_flags);

        $this->_entity_replacements = array(
                                            'amp' => '&',
                                            'lt' => '<',
                                            'gt' => '>',
                                            'quot' => '"'
                                            );

        $this->_attr_find = sprintf("/%s/%s",
                                    $this->_attr_find,
                                    $this->_re_flags);

        $this->_removed_re = sprintf("/%s/%s",
                                     $this->_removed_re,
                                     $this->_re_flags);

        $this->_ent_replace =
            sprintf("&(%s);", implode("|",
                                      $this->_entity_replacements));
    }

    function tagMatcher($tag_name, $close_tags = null)
    {
        $expr = $this->_tag_expr;

        if ($close_tags) {
            $options = implode("|", array_merge(array($tag_name), $close_tags));
            $closer = sprintf("(?:%s)", $options);
        } else {
            $closer = $tag_name;
        }

        $expr = sprintf($expr, $tag_name, $closer);
        return sprintf("/%s/%s", $expr, $this->_re_flags);
    }

    function openTag($tag_name)
    {
        $expr = sprintf($this->_open_tag_expr, $tag_name);
        return sprintf("/%s/%s", $expr, $this->_re_flags);
    }

    function closeTag($tag_name)
    {
        $expr = sprintf($this->_close_tag_expr, $tag_name, $tag_name);
        return sprintf("/%s/%s", $expr, $this->_re_flags);
    }

    function htmlBegin($s)
    {
        $matches = array();
        $result = preg_match($this->openTag('html'), $s,
                             $matches, PREG_OFFSET_CAPTURE);
        if ($result === false || !$matches) {
            return false;
        }
        // Return the offset of the first match.
        return $matches[0][1];
    }

    function htmlEnd($s)
    {
        $matches = array();
        $result = preg_match($this->closeTag('html'), $s,
                             $matches, PREG_OFFSET_CAPTURE);
        if ($result === false || !$matches) {
            return false;
        }
        // Return the offset of the first match.
        return $matches[count($matches) - 1][1];
    }

    function headFind()
    {
        return $this->tagMatcher('head', array('body', 'html'));
    }

    function replaceEntities($str)
    {
        foreach ($this->_entity_replacements as $old => $new) {
            $str = preg_replace(sprintf("/&%s;/", $old), $new, $str);
        }
        return $str;
    }

    function removeQuotes($str)
    {
        $matches = array();
        $double = '/^"(.*)"$/';
        $single = "/^\'(.*)\'$/";

        if (preg_match($double, $str, $matches)) {
            return $matches[1];
        } else if (preg_match($single, $str, $matches)) {
            return $matches[1];
        } else {
            return $str;
        }
    }
    
    function match($regexp, $text, &$match)
    {
        if (!is_callable('mb_ereg_search_init')) {
            return preg_match($regexp, $text, $match);
        }

        $regexp = substr($regexp, 1, strlen($regexp) - 2 - strlen($this->_re_flags));
        mb_ereg_search_init($text);
        if (!mb_ereg_search($regexp)) {
            return false;
        }
        list($match) = mb_ereg_search_getregs();
        return true;
    }

    function parseLinkAttrs($html)
    {
        $stripped = preg_replace($this->_removed_re,
                                 "",
                                 $html);

        $html_begin = $this->htmlBegin($stripped);
        $html_end = $this->htmlEnd($stripped);

        if ($html_begin === false) {
            return array();
        }

        if ($html_end === false) {
            $html_end = strlen($stripped);
        }

        $stripped = substr($stripped, $html_begin,
                           $html_end - $html_begin);

        // Workaround to prevent PREG_BACKTRACK_LIMIT_ERROR:
        $old_btlimit = ini_set( 'pcre.backtrack_limit', -1 );

        // Try to find the <HEAD> tag.
        $head_re = $this->headFind();
        $head_match = '';
        if (!$this->match($head_re, $stripped, $head_match)) {
                     ini_set( 'pcre.backtrack_limit', $old_btlimit );
                     return array();
        }

        $link_data = array();
        $link_matches = array();

        if (!preg_match_all($this->_link_find, $head_match,
                            $link_matches)) {
            ini_set( 'pcre.backtrack_limit', $old_btlimit );
            return array();
        }

        foreach ($link_matches[0] as $link) {
            $attr_matches = array();
            preg_match_all($this->_attr_find, $link, $attr_matches);
            $link_attrs = array();
            foreach ($attr_matches[0] as $index => $full_match) {
                $name = $attr_matches[1][$index];
                $value = $this->replaceEntities(
                              $this->removeQuotes($attr_matches[2][$index]));

                $link_attrs[strtolower($name)] = $value;
            }
            $link_data[] = $link_attrs;
        }

        ini_set( 'pcre.backtrack_limit', $old_btlimit );
        return $link_data;
    }

    function relMatches($rel_attr, $target_rel)
    {
        // Does this target_rel appear in the rel_str?
        // XXX: TESTME
        $rels = preg_split("/\s+/", trim($rel_attr));
        foreach ($rels as $rel) {
            $rel = strtolower($rel);
            if ($rel == $target_rel) {
                return 1;
            }
        }

        return 0;
    }

    function linkHasRel($link_attrs, $target_rel)
    {
        // Does this link have target_rel as a relationship?
        // XXX: TESTME
        $rel_attr = Auth_OpeniD::arrayGet($link_attrs, 'rel', null);
        return ($rel_attr && $this->relMatches($rel_attr,
                                               $target_rel));
    }

    function findLinksRel($link_attrs_list, $target_rel)
    {
        // Filter the list of link attributes on whether it has
        // target_rel as a relationship.
        // XXX: TESTME
        $result = array();
        foreach ($link_attrs_list as $attr) {
            if ($this->linkHasRel($attr, $target_rel)) {
                $result[] = $attr;
            }
        }

        return $result;
    }

    function findFirstHref($link_attrs_list, $target_rel)
    {
        // Return the value of the href attribute for the first link
        // tag in the list that has target_rel as a relationship.
        // XXX: TESTME
        $matches = $this->findLinksRel($link_attrs_list,
                                       $target_rel);
        if (!$matches) {
            return null;
        }
        $first = $matches[0];
        return Auth_OpenID::arrayGet($first, 'href', null);
    }
}

function Auth_OpenID_legacy_discover($html_text, $server_rel,
                                     $delegate_rel)
{
    $p = new Auth_OpenID_Parse();

    $link_attrs = $p->parseLinkAttrs($html_text);

    $server_url = $p->findFirstHref($link_attrs,
                                    $server_rel);

    if ($server_url === null) {
        return false;
    } else {
        $delegate_url = $p->findFirstHref($link_attrs,
                                          $delegate_rel);
        return array($delegate_url, $server_url);
    }
}