<?php
/**
 * Markup Scanner Handler: Wxml2Xml
 *
 * $Id: Xinha2Xml.php 884 2010-11-11 20:04:23Z gerd $
 *
 * @author gERD Schaufelberger <gerd@php-tools.net>
 * @license PHP License
 * @package wb
 * @subpackage Markup
 */

WBClass::load('WBMarkup_Handler_Xml2Wxml');

/**
 * Markup Scanner Handler: Wxml2Xml
 *
 * Convert Wxml html data to Wombat XML
 *
 * @version 0.2.1
 * @package wb
 * @subpackage Markup
 */
class WBMarkup_Handler_Wxml2Xml extends WBMarkup_Handler_Xml2Wxml
{
    /**
     * Cunstructor
     *
     * Flip mapping array
     */
    public function __construct()
    {
        $this->stdTagMap    =   array_flip($this->stdTagMap);
    }
    
   /**
    * Handler on start element
    *
    * @param string $ns The used namespace, if set. otherwise null
    * @param string $tag the tag itself
    * @param array  $attributes the attributes of the found element
    * @param bool $isEmpty true if it is a start- and closing-Element (e.g. <br />)
    * @return bool usually true, false to stop the scanner
    */
    public function onStartElement($ns, $tag, $attributes, $isEmpty)
    {
        if (empty($ns)) {
            if (isset($this->stdTagMap[$tag])) {
                $tag    =   $this->stdTagMap[$tag];            
            }
            if (isset($attributes['wb:dialog'])) {
                $ns     =   'wb';
                $tag    =   $attributes['wb:dialog'];
            }
        }
        
        return parent::onStartElement($ns, $tag, $attributes, $isEmpty);
    }

   /**
    * Handler for entities
    *
    * Remove non-breaking spaces. Call parent's method
    *
    * @param string $entity the entity element, e.g. nbsp for &nbsp;
    * @param boolean $isUnicode true = the element is a unicode string
    */
    public function onEntityElement($entity, $isUnicode)
    {
        if ('nbsp' == $entity) {
            $this->doc[$this->depth]['cData'][] =   ' ';
            return true;
        }
        return parent::onEntityElement($entity, $isUnicode);
    }

   /**
    * Called right after scan is complete
    *
    * Trim whitespace only cData parts. Call parent's method
    *
    * @return bool usually true, false to stop the scanner
    */
    public function onScanComplete()
    {
        foreach ($this->doc[0]['cData'] as &$cData) {
            $len    =   trim($cData);
            if (0 == strlen($len)) {
                $cData  =   '';
            }
        }

        return parent::onScanComplete();
    }

    /**
     * Convert node
     *
     * transform namespaced tags to HTML
     *
     * @todo convert empty namespace (standard HTML) to known WB namespace, if possible
     * @param string $ns
     * @param string $tag
     * @param array $node
     */
    protected function convert($ns, $tag, &$node)
    {
        $con    =   $this->getConverter($ns, $tag);
        if (!$con) {
            $node['ns']     =   null;
            $node['tag']    =   false;
            return;
        }
        return $con->toXml($node);
    }
    
    /**
     * Serialize Document Node
     *
     * Clean up rubish from WYSIWIG editor.
     * Strip nodes with whitespace content only.
     * Remove trailing BR-tags
     *
     * @param array $node
     * @return string serialized attributes
     */
    protected function node2String($node)
    {
        if ($node['isEmpty']) {
            return parent::node2String($node);            
        }
        
        if (!empty($node['ns'])) {
            return parent::node2String($node);
        }

        $this->stripBadTag($node);
        $this->stripBadAttributes($node);

        $brTagRm    =   array('p', 'td', 'th', 'div');

        // strip BR-Tags bofore closing tag
        $end    =   trim(end($node['cData']));
        while ('' === $end && !empty($node['cData'])) {
            array_pop($node['cData']);
            $end    =   end($node['cData']);
        }
        if (in_array($node['tag'], $brTagRm) && '<br />' == $end) {
            array_pop($node['cData']);
        }

        $cData  =   array_map('trim', $node['cData']);
        $cData  =   implode("\n", $cData);

        // empty cData?
        if (0 == strlen($cData)) {
            return '';
        }

        // force paragraph
        if (1 == $node['depth'] && empty($node['tag'])) {
            $node['tag']        =   'p';
            $node['isEmpty']    =   false;
        }

        return parent::node2String($node);
    }

    /**
     * Remove unwanted tags
     *
     * @param array $node
     */
    private function stripBadTag(&$node)
    {
        switch (strtolower($node['tag'])) {
            case 'font':
            case 'span':
            case 'div':
            case 'table':
            case 'tbody':
            case 'thead':
            case 'tfoot':
            case 'tr':
            case 'td':
            case 'th':
                $node['tag']        =   '';
                $node['ns']         =   '';
                $node['attributes'] =   array();
                break;

            default:
                return;
                break;
        }
    }

    /**
     * Remove unwanted attributes
     *
     * Remove all attributes besides:
     *  - class
     *  - href (a)
     *  - title (a)
     *
     * @param array $node
     */
    private function stripBadAttributes(&$node)
    {
        if (empty($node['attributes'])) {
            return;
        }

        $allowed    =   array('class');
        if ('a' == $node['tag']) {
            $allowed[]  =   'href';
            $allowed[]  =   'title';
        }

        $atts   =   array();
        foreach ($allowed as $a) {
            if (isset($node['attributes'][$a])) {
                $atts[$a]   =   $node['attributes'][$a];
            }
        }
        $node['attributes'] =   $atts;
    }

}