<?php

/**
*
* Parse for links to wiki pages.
*
* @category Text
*
* @package Text_Wiki
*
* @author Paul M. Jones <pmjones@php.net>
*
* @license LGPL
*
* @version $Id$
*
*/

/**
*
* Parse for links to wiki pages.
*
* Wiki page names are typically in StudlyCapsStyle made of
* WordsSmashedTogether.
*
* You can also create described links to pages in this style:
* [WikiPageName nice text link to use for display]
*
* The token options for this rule are:
*
* 'page' => the wiki page name.
*
* 'text' => the displayed link text.
*
* 'anchor' => a named anchor on the target wiki page.
*
* @category Text
*
* @package Text_Wiki
*
* @author Paul M. Jones <pmjones@php.net>
*
*/

class Text_Wiki_Parse_Default_Wikilink extends Text_Wiki_Parse {

    var $conf = array (
                       "ext_chars" => false,
                       "utf-8" => false
    );

    /**
    *
    * Constructor.
    *
    * We override the Text_Wiki_Parse_Default constructor so we can
    * explicitly comment each part of the $regex property.
    *
    * @access public
    *
    * @param object &$obj The calling "parent" Text_Wiki object.
    *
    */

    function __construct(&$obj)
    {
        parent::__construct($obj);

        if ($this->getConf("utf-8")) {
			$upper = 'A-Z\p{Lu}';
			$lower = 'a-z0-9\p{Ll}';
			$either = 'A-Za-z0-9\p{L}';
        } else if ($this->getConf("ext_chars")) {
        	// use an extended character set; this should
        	// allow for umlauts and so on.  taken from the
        	// Tavi project defaults.php file.
			$upper = 'A-Z\xc0-\xde';
			$lower = 'a-z0-9\xdf-\xfe';
			$either = 'A-Za-z0-9\xc0-\xfe';
		} else {
			// the default character set, should be fine
			// for most purposes.
			$upper = "A-Z";
			$lower = "a-z0-9";
			$either = "A-Za-z0-9";
		}

        // build the regular expression for finding WikiPage names.
        $this->regex =
            "(!?" .            // START WikiPage pattern (1)
            "[$upper]" .       // 1 upper
            "[$either]*" .     // 0+ alpha or digit
            "[$lower]+" .      // 1+ lower or digit
            "[$upper]" .       // 1 upper
            "[$either]*" .     // 0+ or more alpha or digit
            ")" .              // END WikiPage pattern (/1)
            "((\#" .           // START Anchor pattern (2)(3)
            "[$either]" .      // 1 alpha
            "(" .              // start sub pattern (4)
            "[-_$either:.]*" . // 0+ dash, alpha, digit, underscore, colon, dot
            "[-_$either]" .    // 1 dash, alpha, digit, or underscore
            ")?)?)";           // end subpatterns (/4)(/3)(/2)
    }

    /**
    *
    * Constructor.
    *
    * We override the Text_Wiki_Parse_Default constructor so we can
    * explicitly comment each part of the $regex property.
    *
    * @access public
    *
    * @param object &$obj The calling "parent" Text_Wiki object.
    *
    */

    function Text_Wiki_Parse_Default_Wikilink(&$obj)
    {
        self::__construct($obj);
    }


    /**
    *
    * First parses for described links, then for standalone links.
    *
    * @access public
    *
    * @return void
    *
    */

    function parse()
    {
        // described wiki links
        $tmp_regex = '/\[' . $this->regex . ' (.+?)\]/'.($this->getConf("utf-8") ? "u" : "");
        $this->wiki->source = preg_replace_callback(
            $tmp_regex,
            array(&$this, "processDescr"),
            $this->wiki->source
        );

        // standalone wiki links
        if ($this->getConf("utf-8")) {
			$either = 'A-Za-z0-9\p{L}';
        } else
			$either = ( $this->getConf( "ext_chars" ) ) ? "A-Za-z0-9\xc0-\xfe" : "A-Za-z0-9";

        $tmp_regex = "/(^|[^{$either}\-_]){$this->regex}/".($this->getConf("utf-8") ? "u" : "");
        $this->wiki->source = preg_replace_callback(
            $tmp_regex,
            array(&$this, "process"),
            $this->wiki->source
        );
    }


    /**
    *
    * Generate a replacement for described links.
    *
    * @access public
    *
    * @param array &$matches The array of matches from parse().
    *
    * @return A delimited token to be used as a placeholder in
    * the source text, plus any text priot to the match.
    *
    */

    function processDescr(&$matches)
    {
        // set the options
        $options = array(
            "page"   => $matches[1],
            "text"   => $matches[5],
            "anchor" => $matches[3]
        );

        // create and return the replacement token and preceding text
        return $this->wiki->addToken($this->rule, $options); // . $matches[7];
    }


    /**
    *
    * Generate a replacement for standalone links.
    *
    *
    * @access public
    *
    * @param array &$matches The array of matches from parse().
    *
    * @return A delimited token to be used as a placeholder in
    * the source text, plus any text prior to the match.
    *
    */

    function process(&$matches)
    {
        // when prefixed with !, it's explicitly not a wiki link.
        // return everything as it was.
        if ($matches[2][0] == "!") {
            return $matches[1] . substr($matches[2], 1) . $matches[3];
        }

        // set the options
        $options = array(
            "page" => $matches[2],
            "text" => $matches[2] . $matches[3],
            "anchor" => $matches[3]
        );

        // create and return the replacement token and preceding text
        return $matches[1] . $this->wiki->addToken($this->rule, $options);
    }
}