<?php

/**
* 
* Parse for URLS in the source text.
* 
* @category Text
* 
* @package Text_Wiki
* 
* @author Paul M. Jones <pmjones@php.net>
* 
* @license LGPL
* 
* @version $Id: Url.php,v 1.3 2005/02/23 17:38:29 pmjones Exp $
* 
*/

/**
* 
* Parse for URLS in the source text.
* 
* Various URL markings are supported: inline (the URL by itself),
* numbered or footnote reference (where the URL is enclosed in square
* brackets), and named reference (where the URL is enclosed in square
* brackets and has a name included inside the brackets).  E.g.:
*
* inline    -- http://example.com
* numbered  -- [http://example.com]
* described -- [http://example.com Example Description]
*
* When rendering a URL token, this will convert URLs pointing to a .gif,
* .jpg, or .png image into an inline <img /> tag (for the 'xhtml'
* format).
*
* Token options are:
* 
* 'type' => ['inline'|'footnote'|'descr'] the type of URL
* 
* 'href' => the URL link href portion
* 
* 'text' => the displayed text of the URL link
* 
* @category Text
* 
* @package Text_Wiki
* 
* @author Paul M. Jones <pmjones@php.net>
* 
*/

class Text_Wiki_Parse_Url extends Text_Wiki_Parse {
    
    
    /**
    * 
    * Keeps a running count of numbered-reference URLs.
    * 
    * @access public
    * 
    * @var int
    * 
    */
    
    var $footnoteCount = 0;
    
    
    /**
    * 
    * URL schemes recognized by this rule.
    * 
    * @access public
    * 
    * @var array
    * 
    */
    
    var $conf = array(
        'schemes' => array(
            'http://',
            'https://',
            'ftp://',
            'gopher://',
            'news://',
            'mailto:'
        )
    );
    
    
    /**
    * 
    * Constructor.
    * 
    * We override the constructor so we can comment the regex nicely.
    * 
    * @access public
    * 
    */
    
    function Text_Wiki_Parse_Url(&$obj)
    {
        parent::Text_Wiki_Parse($obj);
        
        // convert the list of recognized schemes to a regex-safe string,
        // where the pattern delim is a slash
        $tmp = array();
        $list = $this->getConf('schemes', array());
        foreach ($list as $val) {
            $tmp[] = preg_quote($val, '/');
        }
        $schemes = implode('|', $tmp);
        
        // build the regex
        $this->regex =
            "($schemes)" . // allowed schemes
            "(" . // start pattern
            "[^ \\/".'"'."\'{$this->wiki->delim}]*\\/" . // no spaces, backslashes, slashes, double-quotes, single quotes, or delimiters;
            ")*" . // end pattern
            "[^ \\t\\n\\/".'"'."\'{$this->wiki->delim}]*" .
            "[A-Za-z0-9\\/?=&~_]";
    }
    
    
    /**
    * 
    * Find three different kinds of URLs in the source text.
    *
    * @access public
    * 
    */
    
    function parse()
    {
        // -------------------------------------------------------------
        // 
        // Described-reference (named) URLs.
        // 
        
        // the regular expression for this kind of URL
        $tmp_regex = '/\[(' . $this->regex . ') ([^\]]+)\]/';
        
        // use a custom callback processing method to generate
        // the replacement text for matches.
        $this->wiki->source = preg_replace_callback(
            $tmp_regex,
            array(&$this, 'processDescr'),
            $this->wiki->source
        );
        
        
        // -------------------------------------------------------------
        // 
        // Numbered-reference (footnote-style) URLs.
        // 
        
        // the regular expression for this kind of URL
        $tmp_regex = '/\[(' . $this->regex . ')\]/U';
        
        // use a custom callback processing method to generate
        // the replacement text for matches.
        $this->wiki->source = preg_replace_callback(
            $tmp_regex,
            array(&$this, 'processFootnote'),
            $this->wiki->source
        );
        
        
        // -------------------------------------------------------------
        // 
        // Normal inline URLs.
        // 
        
        // the regular expression for this kind of URL
        
        $tmp_regex = '/(^|[^A-Za-z])(' . $this->regex . ')(.*?)/';
        
        // use the standard callback for inline URLs
        $this->wiki->source = preg_replace_callback(
            $tmp_regex,
            array(&$this, 'process'),
            $this->wiki->source
        );
    }
    
    
    /**
    * 
    * Process inline URLs.
    * 
    * @param array &$matches
    * 
    * @param array $matches An array of matches from the parse() method
    * as generated by preg_replace_callback.  $matches[0] is the full
    * matched string, $matches[1] is the first matched pattern,
    * $matches[2] is the second matched pattern, and so on.
    * 
    * @return string The processed text replacement.
    * 
    */ 
    
    function process(&$matches)
    {
        // set options
        $options = array(
            'type' => 'inline',
            'href' => $matches[2],
            'text' => $matches[2]
        );
        
        // tokenize
        return $matches[1] . $this->wiki->addToken($this->rule, $options) . $matches[5];
    }
    
    
    /**
    * 
    * Process numbered (footnote) URLs.
    * 
    * Token options are:
    * @param array &$matches
    * 
    * @param array $matches An array of matches from the parse() method
    * as generated by preg_replace_callback.  $matches[0] is the full
    * matched string, $matches[1] is the first matched pattern,
    * $matches[2] is the second matched pattern, and so on.
    * 
    * @return string The processed text replacement.
    * 
    */ 
    
    function processFootnote(&$matches)
    {
        // keep a running count for footnotes 
        $this->footnoteCount++;
        
        // set options
        $options = array(
            'type' => 'footnote',
            'href' => $matches[1],
            'text' => $this->footnoteCount
        );
        
        // tokenize
        return $this->wiki->addToken($this->rule, $options);
    }
    
    
    /**
    * 
    * Process described-reference (named-reference) URLs.
    * 
    * Token options are:
    *     'type' => ['inline'|'footnote'|'descr'] the type of URL
    *     'href' => the URL link href portion
    *     'text' => the displayed text of the URL link
    * 
    * @param array &$matches
    * 
    * @param array $matches An array of matches from the parse() method
    * as generated by preg_replace_callback.  $matches[0] is the full
    * matched string, $matches[1] is the first matched pattern,
    * $matches[2] is the second matched pattern, and so on.
    * 
    * @return string The processed text replacement.
    * 
    */ 
    
    function processDescr(&$matches)
    {
        // set options
        $options = array(
            'type' => 'descr',
            'href' => $matches[1],
            'text' => $matches[4]
        );
        
        // tokenize
        return $this->wiki->addToken($this->rule, $options);
    }
}
?>