View only domain name for links

A link shortener is a web service that creates url forwards.

You just want some text manipulation.

The following will return the links you want:

function pretty_domain_link($url) {
    preg_match('@^(?:http://)?(?:[^.]+\.)?([^.]+)\.@i', $url, $matches);
    return "<a href=\"$url\">".ucfirst($matches[1])."</a>";
}

But if you want to automatically parse and modify the links in your posts, you will need…

Alright, I am going to assume that you want to just post plaintext urls into your post and have those urls replaced by pretty links.

This is a function that can replace all plaintext urls with <a href=""></a> hyperlinks.

function prettify_links($content) {
    $pattern = '@(^|\s|>)(https?://(?:[\w]+\.)?([\w]+)\.(?:[\w]+)(?:/(?:[^\s<]*)?)?)(\s|$|<)@i';
    return preg_replace_callback($pattern,
        function ($matches) {
            return $matches[1]."<a href=\"$matches[2]\">".ucfirst($matches[3])."</a>".$matches[4];
        },
        $content);
}

Next we add a content filter for posts:

add_filter('the_content', 'prettify_links');

Et voilà, all urls that are enclosed in either whitespace or tags get prettified.

Check my site to see an example. The front page is a mess, but if you scroll down you will see a Test Post. Observe the links. If you then open the Test Post itself in single view, you will see the plain text.

You can add this code to your theme’s functions.php. That is to say, add both the prettify_links function and the add_filter call to that file.

Give it a try, and tell me what you think.

It’s taken me hours to figure out the regular expression that I needed lol, and it is probably still not perfect, because I am too lazy to think of all the different cases of the characters that are valid in urls.

Anyway, it will accept the following urls:

>http://bla.com/bla<
. http://bla.com/bla . (adding the dots to show the spaces, what I mean is that it is a url enclosed in whitespace)

which normally translates to
<p>http://bla.com/bla</p>

So when you enter a url and WordPress turns it into a paragraph it should always work.

It will NOT process urls enclosed in anything else (except for newlines and beginning/end of data) so it will not touch anything like "http://bla.com/bla" which would be a case if it was ALREADY a hyperlink.

It’s been fun messing with that code. I hope you enjoy it.

Alright but that will be a rather specialized use case.

What is doing the pasting?

I need to know the exact html markup of those three lines.

You can change it in CSS perhaps but only if the three lines are already identified with some class or id.

I’d need to do some second stage to merge the lines.

So the links are actually put in special <li> tags?

No there is no way for that because then you’d end up with

Google | Yahoo | Reddit |

You need a second stage, so I need to know the format of the resulting html (before my filter gets applied).

Okay I may have a foolproof answer for you:

function prettify_links($content) {
    return preg_replace_callback(
        '@(^|\s|>)(https?://(?:[\w]+?\.)?([\w]+?)\.(?:[\w]+)(?:/(?:[^\s<\]"]+)?)?)@i',
        function ($matches) {
            return $matches[1]."<a href=\"$matches[2]\">".ucfirst($matches[3])."</a>";
        },
        $content
    );
}
function remove_p_br($content) {
    return preg_replace(
        '@<p>\s*(<a [^>]+>[^<]*</a>)\s*</p>@i',
        "$1",
        preg_replace('@<br.?.?>@', '', $content)
    );
}
function group_a($content) {
    $a = "(<a [^>]+>[^<]*</a>)";
    return preg_replace(
        '@'.$a.'\s*'.$a.'\s*'.$a.'@i',
        "<p>$1 | $2 | $3</p>",
        $content
    );
}
function prettify_and_merge($content) {
    preg_match('@^(.*?)(<li class="ddl">.*?</li>)(.*)$@is', $content, $split);
    return count($split) ? $split[1] . group_a(remove_p_br(prettify_links($split[2]))) . $split [3] : $content;
}

add_filter('the_content', 'prettify_and_merge');

Alternatively these couple of lines would do as well:

function prettify_and_merge($content) {
    preg_match('@^(.*?)(<li class="ddl">.*?</li>)(.*)$@is', $content, $split);
    $a = "(<a [^>]+>[^<]*</a>)";
    return count($split) ? $split[1] . preg_replace( '@'.$a.'\s*'.$a.'\s*'.$a.'@i', "<p>$1 | $2 | $3</p>", preg_replace('@<p>\s*(<a [^>]+>[^<]*</a>)\s*</p>@i', "$1", preg_replace('@<br.?.?>@', '', preg_replace_callback('@(^|\s|>)(https?://(?:[\w]+?\.)?([\w]+?)\.(?:[\w]+)(?:/(?:[^\s<\]"]+)?)?)@i', function ($matches) { return $matches[1]."<a href=\"$matches[2]\">".ucfirst($matches[3])."</a>"; }, $split[2])))) . $split [3] : $content;
}
add_filter('the_content', 'prettify_and_merge');

(It’s a tad long).

Give it a try.

Hey,

I didn’t really have time for it today. But I suspect you just had to remove one character right. I think I treated ] as a delimiter that could end the url rather than be part of it.

The string that says \] as part of

'@(^|\s|>)(https?://(?:[\w]+?\.)?([\w]+?)\.(?:[\w]+)(?:/(?:[^\s<\]"]+)?)?)@i',

would simply need to be deleted, that’s all.

I’ve been messing with a way of making this code more functional.

Currently I have a solution that can concatenate any number of links on one line. It tries to remove <p></p> from around the link, and also remove intermediate <br /> tags. Then it just greedily scans for links and puts them on the same line as per your format.

It doesn’t depend on a <li></li> block. It’s actually not all that different from what you have now, except that it doesn’t use preg_replace.

I don’t really know what the big difference is. I’m not sure if I could do it with a sophisticated preg_replace. I could try. I’m just messing about with different coding styles for fun.

I’m not sure if I’d ever use it. But I’m thinking it could be useful to have a way of including footnote links. As in:

This is some text with a note link[_1_] and as I write on I come across another thing I want to reference[_2_].

And then down below I put down links like this:

[_1_] http://wordpress.org/support/topic/view-only-domain-name-for-links?replies=10#post-5551262
[_2_] http://www.xenhideout.nl

And then I want it to turn into something like:

[1] WordPress · [2] Xenhideout

function prettify_links($content) {
    return preg_replace_callback(
        '@(^|\s|>)(https?://(?:[\w]+?\.)?([\w]+?)\.(?:[\w]+)(?:/(?:[^\s<"]+)?)?)@i',
        function ($matches) {
            return $matches[1]."<a href=\"$matches[2]\">".ucfirst($matches[3])."</a>";
        },
        $content);
}
function remove_p($content) {
    return preg_replace(
        '@<p>\s*(<a [^>]+>[^<]*</a>)\s*</p>@i',
        "$1",
        $content
    );
}
function group_a($content) {
    $a = "(<a [^>]+>[^<]*</a>)";
    $s = '\s*(?:<br.?.?>)?\s*';
    $offset = 0;
    preg_match('@'.$a.'@i', $content, $matches, PREG_OFFSET_CAPTURE, $offset);
    $begin = ""; $links = "";
    if (count($matches)) {
        $numberlinks = 0;
        $offset = $matches[0][1];
        $begin = substr($content, 0, $offset);
        $links = $matches[1][0];
        $offset += strlen($matches[0][0]);

        preg_match('@'.$s.$a.'@i', $content, $matches, PREG_OFFSET_CAPTURE, $offset);
        while (count($matches) && $offset == $matches[0][1]) {
            $numberlinks++;
            $links .= " | " . $matches[1][0];
            $offset += strlen($matches[0][0]);
            preg_match('@'.$s.$a.'@i', $content, $matches, PREG_OFFSET_CAPTURE, $offset);
        }
        $links = $numberlinks > 1 ? '<p>' . $links . '</p>' : $links;
        $remainder = substr($content, $offset);
        return $begin . $links . group_a($remainder);
    } else
        return $content;
}
function prettify_and_merge($content) {
    return group_a(remove_p(prettify_links($content)));
}

add_filter('the_content', 'prettify_and_merge');

This code is way different, it doesn’t look at <li></li> tags but it does process all the plaintext urls in the post.

Didn’t know you used more than one of those sections in each post.

That’s more of a WordPress bug. Because it writes the following line:

<p>http://wordpress.org/support/topic/view-only-domain-name-for-links?replies=1</li>

So you can see there is no closing </p> and the (x)html is malformed.

And my code doesn’t match just <p>http://link.com/ when it removes the <p> tags.

function array_add($a, $b) {
    return array( $a[0] . $b[0], $a[1] + $b[1], $a[2] + $b[2] );
}
function group_a($content) {
    $concatlinks = function ($offset) use (&$content, &$concatlinks, &$sa) {
        preg_match($sa, $content, $matches, PREG_OFFSET_CAPTURE, $offset);
        return count($matches) && $offset == $matches[0][1] ?
            array_add(
                array(" | " . $matches[1][0], 0, 1),
                $concatlinks($offset + strlen($matches[0][0]))
            ) :
            array("", $offset, 0);
    };

    $a = "(<a [^>]+>[^<]*</a>)";
    $s = '\s*(?:<br.?.?>)?\s*';
    $sa = '@'.$s.$a.'@i';
    preg_match('@'.$a.'@i', $content, $matches, PREG_OFFSET_CAPTURE, 0);
    if (count($matches)) {
        $offset = $matches[0][1];
        $begin = substr($content, 0, $offset);
        $links = $matches[1][0];
        $offset += strlen($matches[0][0]);
        list($str, $offset, $numberlinks) = $concatlinks($offset);
        $links .= $str;
        $links = $numberlinks > 1 ? '<p>' . $links . '</p>' : $links;
        $remainder = substr($content, $offset);
        return $begin . $links . group_a($remainder);
    } else
        return $content;
}

This is the exact same code, only done with recursive functional programming 😛 😛 :P. The recursive function returns a 3-tuple.