Industrial Strength WordPress Dofollow Upgrade

By Jeff Starr on Tuesday, September 11, 2007 — 15 Responses metadata

Post #402 categorized as Function, WordPress, last updated on Nov 5, 2007
Tagged with comments, dofollow, hack, links, nofollow, php, tutorials, WordPress

Encourage Comments by Completely Eliminating All Nofollow Links

Want to remove all traces of the hideous nofollow attribute without having to install yet another unnecessary plugin? By default, WordPress generates nofollow links in three different ways — this article will show you how to eliminate all of them..

Some context please..

Note: if you are already familiar with the various functions involved in the nofollow-removal process, please feel free to skip the proceeding discussion and jump directly to the tutorial.

WordPress adds nofollow attributes to all trackbacks, pingbacks, and commentator links

We have seen how simple it is to eradicate nofollow from comment-related content, which includes the three different types of $author URLs: trackbacks, pingbacks, and commentator links. In fact, WordPress generates hyperlinks for each of these comment-author URLs via the function get_comment_author_link(), which is conveniently located in the file wp-includes/comment-functions.php in WordPress 2.0 and wp-includes/comment-template.php in WordPress 2.1 and 2.2:

function get_comment_author_link() {
	global $comment;
	$url    = get_comment_author_url();
	$author = get_comment_author();

	if ( empty( $url ) || 'http://' == $url )
		$return = $author;
	else
		$return = "<a href='$url' rel='external nofollow'>$author</a>";
	return apply_filters('get_comment_author_link', $return);
}

Here, WordPress simply differentiates between linked and unlinked author data. Commentators with an associated URL receive a nasty nofollow link, regardless of link type (trackback, pingback, or commentator link). Those without an associated URL receive no link, but rather a text-only rendering of their name. Thus, our first type of link for which WordPress generates nofollow markup involves comment-related author links.

WordPress adds nofollow attributes to all unformatted, text-only comment links

The second type of link for which WordPress generates nofollow markup involves hyperlinks generated automatically from unformatted, text-only (non-hypertext) URL information. WordPress filters all comment content (post content is not filtered) for strings of text matching the generalized pattern of a URL: http://domain.tld/directory/file.html. Each detected instance of such a string is transformed into an actual hypertext link via the function make_clickable(), which is located in wp-includes/functions-formatting.php in WordPress 2.0, and wp-includes/formatting.php in WordPress 2.1 and 2.2:

function make_clickable($ret) {
	$ret = ' ' . $ret;
	// in testing, using arrays here was found to be faster
	$ret = preg_replace(
		array(
			'#([\s>])([\w]+?://[\w\#$%&~/.\-;:=,?@\[\]+]*)#is',
			'#([\s>])((www|ftp)\.[\w\#$%&~/.\-;:=,?@\[\]+]*)#is',
			'#([\s>])([a-z0-9\-_.]+)@([^,< \n\r]+)#i'),
		array(
			'$1<a href="$2" rel="nofollow">$2</a>',
			'$1<a href="http://$2" rel="nofollow">$2</a>',
			'$1<a href="mailto:$2@$3">$2@$3</a>'),$ret);
	// this one is not in an array because we need it to run last, for cleanup of accidental links within links
	$ret = preg_replace("#(<a( [^>]+?>|>))<a [^>]+?>([^>]+?)</a></a>#i", "$1$3</a>", $ret);
	$ret = trim($ret);
	return $ret;
}

Although it looks hideously complicated, the make_clickable() function simply checks comments for unformatted (non-hypertext/text-only) URLs and transforms them into genuine, clickable hyperlinks for all to enjoy. Of course, in the process of generating the links, WordPress takes the liberty of injecting those nasty nofollow tags. Yuck. It is also important to note that WordPress calls this function into action via the following filter (located in wp-includes/default-filters.php in WP 2+):

add_filter('comment_text', 'make_clickable');

WordPress adds nofollow attributes to all other comment-related links

The third and final type of link for which WordPress generates nofollow markup involves all hyperlinks within the comments themselves. Of course, hyperlinks are URLs that are originally expressed via anchor tags: <a href=""></a>. Every such link left within a WordPress-powered comment is generated by the function wp_rel_nofollow(), which is located in the file wp-includes/functions-formatting.php in WordPress 2.0:

function wp_rel_nofollow( $text ) {
	$text = preg_replace('|<a (.+?)>|ie', "'<a ' . str_replace(' rel=\"nofollow\"','',stripslashes('$1')) . ' rel=\"nofollow\">'", $text);
	return $text;
}

..and in wp-includes/formatting.php in WordPress 2.1 and 2.2:

function wp_rel_nofollow( $text ) {
	global $wpdb;
	// This is a pre save filter, so text is already escaped.
	$text = stripslashes($text);
	$text = preg_replace('|<a (.+?)>|ie', "'<a ' . str_replace(' rel=\"nofollow\"','',stripslashes('$1')) . ' rel=\"nofollow\">'", $text);
	$text = $wpdb->escape($text);
	return $text;
}

As we may see, wp_rel_nofollow() has been upgraded slightly in WP 2.1/2.2, but its overall functionality remains the same: add nofollow attributes to all hyperlinks located within comments. Further, WordPress calls upon this function via a filter located in the wp-includes/default-filters.php file (WordPress 2+):

add_filter('pre_comment_content', 'wp_rel_nofollow', 15);

Without that filter, WordPress cannot apply nofollow to links within comment content (hint, hint).

Getting on with it..

Okay, before we jump into the industrial-strength dofollow tutorial, let’s summarize the three different types of nofollow links and their associated functions:

1. get_comment_author_link() → all trackbacks, pingbacks, and commentator links
2. make_clickable() (applied via filter) → all unformatted (text-only) URL addresses
3. wp_rel_nofollow() (applied via filter) → all other comment-related hyperlinks

And with that, we have a clear understanding of the functions involved in the nofollow-attribution process, as well as a nice plan of attack for our triune dofollow hack. Ready, let’s get on with it..

Tutorial: Complete nofollow removal [ ^ ]

Before we begin, remember to backup your data!

Step 1) Remove nofollow from all trackbacks, pingbacks, and commentator links

As per the first part of the preceding discussion, locate the function get_comment_author_link() within wp-includes/comment-functions.php (WP 2.0) or wp-includes/comment-template.php (WP 2.1/2.2). In that function, replace this line:

$return = "<a href='$url' rel='external nofollow'>$author</a>";

..with this one:

$return = "<a href='$url' rel='external'>$author</a>";

Step 2) Remove nofollow from all unformatted URL addresses

As per the second part of the preceding discussion, locate the function make_clickable() within wp-includes/functions-formatting.php (WP 2.0) or wp-includes/formatting.php (WP 2.1/2.2). In that function, replace these two lines:

'$1<a href="$2" rel="nofollow">$2</a>',
'$1<a href="http://$2" rel="nofollow">$2</a>',

..with these two lines:

'$1<a href="$2">$2</a>',
'$1<a href="http://$2">$2</a>',

Step 3) Remove nofollow from all other comment-related hyperlinks

As per the third part of the preceding discussion, locate the following filter within wp-includes/defualt-filters.php:

add_filter('pre_comment_content', 'wp_rel_nofollow', 15);

..and comment it out like so:

// add_filter('pre_comment_content', 'wp_rel_nofollow', 15);

That’s it, you’re done! Upload your three files and check a few comment links to make sure everything went according to plan. If all of this seems like waay too much hassle to go through just to upgrade WordPress to dofollow status, you may prefer to simply take advantage of one of the many dofollow plugins that are currently available for WordPress users. Either way, removing nofollow tags is a great way to improve your blog. Cheers!

Subscribe to Perishable Press

Toggle Post Options 28804 visits