<?php

/**
* 
* Parses for bulleted and numbered lists.
* 
* @category Text
* 
* @package Text_Wiki
* 
* @author Paul M. Jones <pmjones@php.net>
* 
* @license LGPL
* 
* @version $Id: List.php,v 1.3 2005/02/23 17:38:29 pmjones Exp $
* 
*/

/**
* 
* Parses for bulleted and numbered lists.
* 
* This class implements a Text_Wiki_Parse to find source text marked as
* a bulleted or numbered list.  In short, if a line starts with '* ' then
* it is a bullet list item; if a line starts with '# ' then it is a 
* number list item.  Spaces in front of the * or # indicate an indented
* sub-list.  The list items must be on sequential lines, and may be 
* separated by blank lines to improve readability.  Using a non-* non-#
* non-whitespace character at the beginning of a line ends the list.
*
* @category Text
* 
* @package Text_Wiki
* 
* @author Paul M. Jones <pmjones@php.net>
* 
*/

class Text_Wiki_Parse_List extends Text_Wiki_Parse {
    
    
    /**
    * 
    * The regular expression used to parse the source text and find
    * matches conforming to this rule.  Used by the parse() method.
    * 
    * @access public
    * 
    * @var string
    * 
    * @see parse()
    * 
    */
    
    var $regex = '/\n((\*|#) .*\n)(?! {0,}(\* |# |\n))/Us';
    
    
    /**
    * 
    * Generates a replacement for the matched text.  Token options are:
    * 
    * 'type' =>
    *     'bullet_start' : the start of a bullet list
    *     'bullet_end'   : the end of a bullet list
    *     'number_start' : the start of a number list
    *     'number_end'   : the end of a number list
    *     'item_start'   : the start of item text (bullet or number)
    *     'item_end'     : the end of item text (bullet or number)
    *     'unknown'      : unknown type of list or item
    *
    * 'level' => the indent level (0 for the first level, 1 for the
    * second, etc)
    *
    * 'count' => the list item number at this level. not needed for
    * xhtml, but very useful for PDF and RTF.
    * 
    * @access public
    *
    * @param array &$matches The array of matches from parse().
    *
    * @return A series of text and delimited tokens marking the different
    * list text and list elements.
    *
    */
    
    function process(&$matches)
    {
        // the replacement text we will return
        $return = '';
        
        // the list of post-processing matches
        $list = array();
        
        // a stack of list-start and list-end types; we keep this
        // so that we know what kind of list we're working with
        // (bullet or number) and what indent level we're at.
        $stack = array();
        
        // the item count is the number of list items for any
        // given list-type on the stack
        $itemcount = array();
        
        // have we processed the very first list item?
        $pastFirst = false;
        
        // populate $list with this set of matches. $matches[1] is the
        // text matched as a list set by parse().
        preg_match_all(
            '=^( {0,})(\*|#) (.*)$=Ums',
            $matches[1],
            $list,
            PREG_SET_ORDER
        );
        
        // loop through each list-item element.
        foreach ($list as $key => $val) {
            
            // $val[0] is the full matched list-item line
            // $val[1] is the number of initial spaces (indent level)
            // $val[2] is the list item type (* or #)
            // $val[3] is the list item text
            
            // how many levels are we indented? (1 means the "root"
            // list level, no indenting.)
            $level = strlen($val[1]) + 1;
            
            // get the list item type
            if ($val[2] == '*') {
                $type = 'bullet';
            } elseif ($val[2] == '#') {
                $type = 'number';
            } else {
                $type = 'unknown';
            }
            
            // get the text of the list item
            $text = $val[3];
            
            // add a level to the list?
            if ($level > count($stack)) {
                
                // the current indent level is greater than the
                // number of stack elements, so we must be starting
                // a new list.  push the new list type onto the
                // stack...
                array_push($stack, $type);
                
                // ...and add a list-start token to the return.
                $return .= $this->wiki->addToken(
                    $this->rule, 
                    array(
                        'type' => $type . '_list_start',
                        'level' => $level - 1
                    )
                );
            }
            
            // remove a level from the list?
            while (count($stack) > $level) {
                
                // so we don't keep counting the stack, we set up a temp
                // var for the count.  -1 becuase we're going to pop the
                // stack in the next command.  $tmp will then equal the
                // current level of indent.
                $tmp = count($stack) - 1;
                
                // as long as the stack count is greater than the
                // current indent level, we need to end list types. 
                // continue adding end-list tokens until the stack count
                // and the indent level are the same.
                $return .= $this->wiki->addToken(
                    $this->rule, 
                    array (
                        'type' => array_pop($stack) . '_list_end',
                        'level' => $tmp
                    )
                );
                
                // reset to the current (previous) list type so that
                // the new list item matches the proper list type.
                $type = $stack[$tmp - 1];
                
                // reset the item count for the popped indent level
                unset($itemcount[$tmp + 1]);
            }
            
            // add to the item count for this list (taking into account
            // which level we are at).
            if (! isset($itemcount[$level])) {
                // first count
                $itemcount[$level] = 0;
            } else {
                // increment count
                $itemcount[$level]++;
            }
            
            // is this the very first item in the list?
            if (! $pastFirst) {
                $first = true;
                $pastFirst = true;
            } else {
                $first = false;
            }
            
            // create a list-item starting token.
            $start = $this->wiki->addToken(
                $this->rule, 
                array(
                    'type' => $type . '_item_start',
                    'level' => $level,
                    'count' => $itemcount[$level],
                    'first' => $first
                )
            );
            
            // create a list-item ending token.
            $end = $this->wiki->addToken(
                $this->rule, 
                array(
                    'type' => $type . '_item_end',
                    'level' => $level,
                    'count' => $itemcount[$level]
                )
            );
            
            // add the starting token, list-item text, and ending token
            // to the return.
            $return .= $start . $val[3] . $end;
        }
        
        // the last list-item may have been indented.  go through the
        // list-type stack and create end-list tokens until the stack
        // is empty.
        while (count($stack) > 0) {
            $return .= $this->wiki->addToken(
                $this->rule, 
                array (
                    'type' => array_pop($stack) . '_list_end',
                    'level' => count($stack)
                )
            );
        }
        
        // we're done!  send back the replacement text.
        return "\n" . $return . "\n\n";
    }
}
?>