<?php

class SimpleTags_Client_Autolinks {

	static $posts = array();
	static $link_tags = array();

	/**
	 * Constructor
	 *
	 * @return void
	 * @author Amaury Balmer
	 */
	public function __construct() {
		$auto_link_priority = SimpleTags_Plugin::get_option_value( 'auto_link_priority' );
		if ( (int) $auto_link_priority == 0 ) {
			$auto_link_priority = 12;
		}

		// Auto link tags
		add_filter( 'the_posts', array( __CLASS__, 'the_posts' ), 10, 2 );

		if ( SimpleTags_Plugin::get_option_value( 'auto_link_views' ) != 'no' ) {
			add_filter( 'the_content', array( __CLASS__, 'the_content' ), $auto_link_priority );
		}
	}

	/**
	 * Stock posts ID as soon as possible
	 * TODO: test if post_type allow post_tag before keep post ID
	 *
	 * @param array $posts
	 *
	 * @return array
	 */
	public static function the_posts( $posts, $query ) {
		if ( ! empty( $posts ) && is_array( $posts ) ) {
			foreach ( (array) $posts as $post ) {
				self::$posts[] = (int) $post->ID;
			}

			self::$posts = array_unique( self::$posts );
		}

		return $posts;
	}

	/**
	 * Get tags from current post views
	 *
	 * @return boolean
	 */
	public static function get_tags_from_current_posts() {
		if ( is_array( self::$posts ) && count( self::$posts ) > 0 ) {
			// Generate SQL from post id
			$postlist = implode( "', '", self::$posts );

			// Generate key cache
			$key = md5( maybe_serialize( $postlist ) );

			$results = array();

			// Get cache if exist
			$cache = wp_cache_get( 'generate_keywords', 'simpletags' );
			if ( $cache === false ) {
				foreach ( self::$posts as $object_id ) {
					// Get terms
					$terms = get_object_term_cache( $object_id, 'post_tag' );
					if ( false === $terms ) {
						$terms = wp_get_object_terms( $object_id, 'post_tag' );
					}

					if ( $terms != false ) {
						$results = array_merge( $results, $terms );
					}
				}

				$cache[ $key ] = $results;
				wp_cache_set( 'generate_keywords', $cache, 'simpletags' );
			} else {
				if ( isset( $cache[ $key ] ) ) {
					return $cache[ $key ];
				}
			}

			return $results;
		}

		return array();
	}

	/**
	 * Get links for each tag for auto link feature
	 *
	 */
	public static function prepare_auto_link_tags() {
		$auto_link_min = (int) SimpleTags_Plugin::get_option_value( 'auto_link_min' );
		if ( $auto_link_min == 0 ) {
			$auto_link_min = 1;
		}

		foreach ( (array) self::get_tags_from_current_posts() as $term ) {
			if ( $term->count >= $auto_link_min ) {
				self::$link_tags[ $term->name ] = esc_url( get_term_link( $term, $term->taxonomy ) );
			}
		}

		return true;
	}

	/**
	 * Replace text by link to tag
	 *
	 * @param string $content
	 *
	 * @return string
	 */
	public static function the_content( $content = '' ) {
		global $post;

		// Show only on singular view ? Check context
		if ( SimpleTags_Plugin::get_option_value( 'auto_link_views' ) == 'singular' && ! is_singular() ) {
			return $content;
		}

		// user preference for this post ?
		$meta_value = get_post_meta( $post->ID, '_exclude_autolinks', true );
		if ( ! empty( $meta_value ) ) {
			return $content;
		}

		// Get currents tags if no exists
		self::prepare_auto_link_tags();

		// Shuffle array
		SimpleTags_Client::random_array( self::$link_tags );

		// HTML Rel (tag/no-follow)
		$rel = SimpleTags_Client::get_rel_attribut();

		// only continue if the database actually returned any links
		if ( ! isset( self::$link_tags ) || ! is_array( self::$link_tags ) || empty( self::$link_tags ) ) {
			return $content;
		}

		// Case option ?
		$case       = ( (int) SimpleTags_Plugin::get_option_value( 'auto_link_case' ) == 1 ) ? 'i' : '';
		$strpos_fnc = ( $case == 'i' ) ? 'stripos' : 'strpos';

		// Prepare exclude terms array
		$excludes_terms = explode( ',', SimpleTags_Plugin::get_option_value( 'auto_link_exclude' ) );
		if ( $excludes_terms == false ) {
			$excludes_terms = array();
		} else {
			$excludes_terms = array_filter( $excludes_terms, '_delete_empty_element' );
			$excludes_terms = array_unique( $excludes_terms );
		}

		$z = 0;
		foreach ( (array) self::$link_tags as $term_name => $term_link ) {
			// Force string for tags "number"
			$term_name = (string) $term_name;
			
			// Exclude terms ? next...
			if ( in_array( $term_name, (array) $excludes_terms ) ) {
				continue;
			}

			// Make a first test with PHP function, economize CPU with regexp
			if ( $strpos_fnc( $content, $term_name ) === false ) {
				continue;
			}

			if ( (int) SimpleTags_Plugin::get_option_value( 'auto_link_dom' ) == 1 && class_exists( 'DOMDocument' ) && class_exists( 'DOMXPath' ) ) {
				self::_replace_by_links_dom( $content, $term_name, $term_link, $case, $rel );
			} else {
				self::_replace_by_links_regexp( $content, $term_name, $term_link, $case, $rel );
			}

			$z ++;
			if ( $z > (int) SimpleTags_Plugin::get_option_value( 'auto_link_max_by_post' ) ) {
				break;
			}
		}


		return $content;
	}

	/**
	 * Replace text by link, except HTML tag, and already text into link, use DOMdocument.
	 * Code take from : http://stackoverflow.com/questions/4044812/regex-domdocument-match-and-replace-text-not-in-a-link
	 *
	 * @param string $content
	 * @param string $search
	 * @param string $replace
	 * @param string $case
	 * @param string $rel
	 */
	private static function _replace_by_links_dom( &$content, $search = '', $replace = '', $case = '', $rel = '' ) {
		$dom = new DOMDocument();

		// loadXml needs properly formatted documents, so it's better to use loadHtml, but it needs a hack to properly handle UTF-8 encoding
		$result = $dom->loadHtml( mb_convert_encoding( $content, 'HTML-ENTITIES', "UTF-8" ) );
		if ( $result === false ) {
			return false;
		}

		$xpath = new DOMXPath( $dom );
		foreach ( $xpath->query( '//text()[not(ancestor::a)]' ) as $node ) {
			$substitute = '<a href="' . $replace . '" class="st_tag internal_tag" ' . $rel . ' title="' . esc_attr( sprintf( SimpleTags_Plugin::get_option_value( 'auto_link_title' ), $search ) ) . "\">$search</a>";

			if ( $case == 'i' ) {
				$replaced = str_ireplace( $search, $substitute, $node->wholeText );
			} else {
				$replaced = str_replace( $search, $substitute, $node->wholeText );
			}

			$newNode = $dom->createDocumentFragment();
			$newNode->appendXML( $replaced );
			$node->parentNode->replaceChild( $newNode, $node );
		}

		// get only the body tag with its contents, then trim the body tag itself to get only the original content
		$content = mb_substr( $dom->saveHTML( $xpath->query( '//body' )->item( 0 ) ), 6, - 7, "UTF-8" );
	}

	/**
	 * Replace text by link, except HTML tag, and already text into link, use PregEXP.
	 *
	 * @param string $content
	 * @param string $search
	 * @param string $replace
	 * @param string $case
	 * @param string $rel
	 */
	private static function _replace_by_links_regexp( &$content, $search = '', $replace = '', $case = '', $rel = '' ) {
		$must_tokenize = true; // will perform basic tokenization
		$tokens        = null; // two kinds of tokens: markup and text

		$j        = 0;
		$filtered = ''; // will filter text token by token

		$match      = '/(\PL|\A)(' . preg_quote( $search, "/" ) . ')(\PL|\Z)/u' . $case;
		$substitute = '$1<a href="' . $replace . '" class="st_tag internal_tag" ' . $rel . ' title="' . esc_attr( sprintf( SimpleTags_Plugin::get_option_value( 'auto_link_title' ), $search ) ) . "\">$2</a>$3";

		//$match = "/\b" . preg_quote($search, "/") . "\b/".$case;
		//$substitute = '<a href="'.$replace.'" class="st_tag internal_tag" '.$rel.' title="'. esc_attr( sprintf( __('Posts tagged with %s', 'simpletags'), $search ) )."\">$0</a>";
		// for efficiency only tokenize if forced to do so
		if ( $must_tokenize ) {
			// this regexp is taken from PHP Markdown by Michel Fortin: http://www.michelf.com/projects/php-markdown/
			$comment                = '(?s:<!(?:--.*?--\s*)+>)|';
			$processing_instruction = '(?s:<\?.*?\?>)|';
			$tag                    = '(?:<[/!$]?[-a-zA-Z0-9:]+\b(?>[^"\'>]+|"[^"]*"|\'[^\']*\')*>)';

			$markup        = $comment . $processing_instruction . $tag;
			$flags         = PREG_SPLIT_DELIM_CAPTURE;
			$tokens        = preg_split( "{($markup)}", $content, - 1, $flags );
			$must_tokenize = false;
		}

		// there should always be at least one token, but check just in case
		$anchor_level = 0;
		if ( isset( $tokens ) && is_array( $tokens ) && count( $tokens ) > 0 ) {
			$i = 0;
			foreach ( $tokens as $token ) {
				if ( ++ $i % 2 && $token != '' ) { // this token is (non-markup) text
					if ( $anchor_level == 0 ) { // linkify if not inside anchor tags
						if ( preg_match( $match, $token ) ) { // use preg_match for compatibility with PHP 4
							$j ++;
							if ( $j <= SimpleTags_Plugin::get_option_value( 'auto_link_max_by_tag' ) || SimpleTags_Plugin::get_option_value( 'auto_link_max_by_tag' ) == 0 ) {// Limit replacement at 1 by default, or options value !
								$token = preg_replace( $match, $substitute, $token ); // only PHP 5 supports calling preg_replace with 5 arguments
							}
							$must_tokenize = true; // re-tokenize next time around
						}
					}
				} else { // this token is markup
					if ( preg_match( "#<\s*a\s+[^>]*>#i", $token ) ) { // found <a ...>
						$anchor_level ++;
					} elseif ( preg_match( "#<\s*/\s*a\s*>#i", $token ) ) { // found </a>
						$anchor_level --;
					}
				}
				$filtered .= $token; // this token has now been filtered
			}
			$content = $filtered; // filtering completed for this link
		}
	}

}