модуль для СЕО

<?php
/** 
 * Autolink plugin
 * 
 * Author: Garry Nutting
 * Original Code and Concept: phpGiggle by Biermana
 * 
 * Version: 1.1
 * Date: 03/05/06
 * 
 * CSS Styles:
 * 
 * autolink_highlight
 * autolink_link
 * autolink_abbr
 * autolink_acro
 * autolink (for Google, CPAN etc. - can be changed in respective function)
 * 
 * updated: support for accented characters
 */

//-- set the template function target
//-- use with externalLinks js function to open template function links in new window
//-- http://pure-essence.net/archives/2006/02/02/relexternal/
define('autolink_target', 'rel="external"');

$e = & $modx->Event;
switch ($e->name) {
	case "OnWebPagePrerender" :
		//-- get current template value
		$table = $modx->getFullTableName('site_content');
		$sql = $modx->db->select('template', $table, 'id="' . $modx->documentIdentifier . '"');
		$row = $modx->db->getRow($sql);
		//-- do not run if on blank template (so it doesn't interfere with RSS feeds)
		if ($row['template'] <> 0) {
			$pg = & $modx->documentOutput;
			$pg = autolink($pg);
		} else {
			return $modx->documentOutput;
		}
		break;
	default :
		return;
		break;
}

/**
 * Main autolink function - replaces keywords with desired values
 */
function autolink($string) {
	global $modx;

	//-- replace with match (1) or value entered in table (0)
	$match_replace = 1;

	//-- blocked tags ie. those that the tokenizer will ignore
	$blocked_tags = 'code|pre';
	$pagetitle_r = array (); //-- stores the link values
	$autolinkTitle = array (); //-- stores the title attribute values
	$autolinkDefault = array (); //-- will store the abbreviation/acronym value

	//-- setup db table name
	$autolinkTable = $modx->getFullTableName('site_content');

	//-- get keywords from db
	$rs = $modx->db->select('*', $autolinkTable);

	while ($row = $modx->db->getRow($rs)) {
		$pagetitle_r[$row['id']] = $row['pagetitle'];		
	}

	//-- get the body tags
	preg_match('/(<body.*?>)/', $string, $open);
	preg_match('/(<\/body.*?>)/', $string, $close);
	$body_open = $open[1];
	$body_close = $close[1];

	unset ($open);
	unset ($close);

	//-- get only body html - will add head and footer back in later
	$temp = explode("$body_open", $string);
	$head = $temp[0];
	$string = $temp[1];
	$temp = explode("$body_close", $string);
	$string = $temp[0];
	$footer = $temp[1];

	//-- ensure that first autolink match will always be second entry in array
	$text = " $string ";

	//-- split the content around the <autolink> tags
	$textarr = preg_split("/<autolink>(.*)<\/autolink>/Us", $text, -1, PREG_SPLIT_DELIM_CAPTURE);
	//-- setup output string
	$out = '';

	//-- start main loop of content array
	for ($i = 0; $i < count($textarr); $i++) {
		$text = $textarr[$i];

		//-- skip the entries outside autolink tags
		if ($i % 2 == 0) {
			$out .= $text;
			continue;
		}

		//-- loop through each keyword in the array
		foreach ($pagetitle_r as $id => $pagetitle) {

	
		
		
			//-- setup the tokenizer variables
			$anchor_level = 0;
			$tag_level = 0;
			$must_tokenize = TRUE; //-- will perform basic tokenization
			$tokens = NULL; //-- two kinds of tokens: markup and text

			$filtered = ''; //-- will filter text token by token
			$match = "/\b" . preg_quote($pagetitle, "/") . "\b/i"; //-- check for encoded match
			$match_noencode = "/\b" . preg_quote($pagetitle, "/") . "\b/i"; //-- check for non-encoded match

			$substitute = '<a class="autolink_link" href="index.php?id=' . $id. '" title="' . $pagetitle . '">' . $pagetitle . '</a>';
		
			
		
			if ($must_tokenize) // for efficiency only tokenize if forced to do so
				{
				//-- 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)}", $text, -1, $flags);
				$must_tokenize = FALSE;
			}

			//-- there should always be at least one token, but check just in case
			if (isset ($tokens) && is_array($tokens) && count($tokens) > 0) {
				$j = 0;
				foreach ($tokens as $token) {
					if (++ $j % 2 && $token != '') //-- this token is (non-markup) text
						{
						if ($anchor_level == 0 && $tag_level == 0) // linkify if not inside anchor tags
							{
							

// ----------------- ВОТ от сюда по регулярным выражениям не сравнивается $token
							//-- test for encoded keyword
							if (preg_match("/\b" .$pagetitle. "\b/", $token, $kmatch)) {
								//-- ensure that the substitute replaces value with the case-sensitive match
								if ($match_replace == 1) {
									$substitute = preg_replace('/\b' . $pagetitle . '\b/', $kmatch[0], $substitute);
								}
								//-- replace keyword with the substitute
								$token = preg_replace("/\b" .$pagetitle. "\b/", $substitute, $token);
								$must_tokenize = TRUE; //-- re-tokenize next time around						

								//-- test for non-encoded keyword
							}
							
						}

					} 
					$filtered .= $token; //-- this token has now been filtered
					
					
				}
			}
			$text = $filtered; //-- filtering completed for this link
		
		}
		//-- add section to the output
		$out .= $text;
	}
	//-- return the compiled output to the browser
	return $head . $body_open . trim($out) . $body_close . $footer;
	}

/******************************************************************
 * 
 * Template Functions
 * 
 */

/*
    template function for search.cpan.org
*/

function autolink_cpan($autolink_it) {
	return "<a " . autolink_target . " class=\"autolink\" title=\"Search CPAN for '" .
	$autolink_it . "' !\" href=\"http://search.cpan.org/search?query=" . urlencode($autolink_it) . "\">" .
	$autolink_it . "</a>";
}

/*
    template function for google.com
*/
function autolink_google($autolink_it) {
	return "<a " . autolink_target . " class=\"autolink\" title=\"Google.com search for '" .
	$autolink_it . "' !\" href=\"http://www.google.com/search?q=" . $autolink_it . "\">" .
	$autolink_it . "</a>";
}

/*
    template function for dictionary.com
*/
function autolink_dictionary($autolink_it) {
	return "<a " . autolink_target . " class=\"autolink\" title=\"Dictionary.com definition for '" .
	$autolink_it . "' !\" href=\"http://dictionary.reference.com/search?q=" . $autolink_it . "\">" .
	$autolink_it . "</a>";
}

/*
    template function for freshmeat.net
*/
function autolink_freshmeat($autolink_it) {
	return "<a " . autolink_target . " class=\"autolink\" title=\"Freshmeat.net search projects for '" .
	$autolink_it . "' !\" href=\"http://freshmeat.net/search/?q=" . urlencode($autolink_it) . "\">" .
	$autolink_it . "</a>";
}

/*
    template function for vivisimo.com
*/
function autolink_vivisimo($autolink_it) {
	return "<a " . autolink_target . " class=\"autolink\" title=\"Vivisimo.com document cluster for '" .
	$autolink_it . "' !\" href=\"http://www.vivisimo.com/search?query=" . urlencode($autolink_it) . "\">" .
	$autolink_it . "</a>";
}

/*
    template function for webopedia.com
    note:  no gaurantee the webopedia page exists - better to manually check it first
*/
function autolink_webopedia($autolink_it) {
	return "<a " . autolink_target . " class=\"autolink\" title=\"Webopedia.com definition for '" .
	$autolink_it . "' !\" href=\"http://www.webopedia.com/TERM/" . substr($autolink_it, 0, 1) . "/" . str_replace(" ", "_", $autolink_it) . ".html" . "\">" .
	$autolink_it . "</a>";
}

?>