Neil Hillman

As a web developer, I’m often trying to squeeze text into small spaces, and sometimes the only way is to create a summary, or abstract of the text.

A lot of abstract functions will truncate after n words:

<?php
preg_match('/^([^.!?\s]*[\.!?\s]+){0,n}/', strip_tags($text), $abstract);
echo $abstract[0];
?>

… or truncate after n sentences:

<?php
preg_match('/^([^.!?]*[\.!?]+){0,n}/', strip_tags($text), $abstract);
echo $abstract[0];
?>

… but this isn’t always enough. Three sentences, (or 50 words for that matter), could run anywhere between 150 and 450 characters long, which is a big variation if you only have room to display 180 characters! So here’s my reusable abstract function that will…

… truncate after the last word that appears within your specified character limit:

<?php
function abstract( $string, $len ) {
  if ( strlen($string) > $len ) {
    $pos = ( $len - stripos(strrev(substr($string, 0, $len)), ' ') );
    $sum = substr($string, 0, $pos-1);
    $chr = $sum[strlen($sum)-1];
    if ( strpos($chr, '.,!?;:') ) {
       // STRIP LAST CHAR
       $sum = substr($sum, 0, strlen($sum)-1);
    }
    // ADD ELLIPSIS
    return $sum."&#8230;";
  } else {
    return $string;
  }
}
echo abstract( $string, 180 );
?>

Extracts from “I Hate the French”, a satirical comedy song written by Richard Curtis and performed by Howard Goodall, live during Rowan Atkinson’s 1980 UK tour.

They all wear berets and they’re all called “Jacques”,
They even steal from us the words they lack:
“le Weekend”, “le Camping” and “Cul-de-sac”,
That’s why I hate the French…

They bake their bread in such a naughty shape,
They brag about their wine, and worship the grape,
They criticise our food but then they eat “Crêpe”!
That’s why I hate the French…

And now they’ve started coming here in droves,
“French cigarettes”, “French letters”, and “French clothes”.

I’m sick and tired of eating all this “Brie”,
And I’ll be buggered if I go to “Gay Paris”!

They’re pretty cocky ’bout their “games in the dark”,
They think with girls they light a “special spark”,
But look what the bastards did to Joan of Arc!
That’s why I hate the French…

If like me, you often need to use regular expressions, but still after over 10 years of using them can waste an entire morning just trying to write a simple search and replace, here are some really useful snippets I have used along the way.

To match an HTML link within a string, and to isolate the inner HTML, (whatever occurs between the opening and closing anchor tags), use:

<?php
preg_match('#<a\s+.*?href=[\'"]([^\'"]+)[\'"]\s*
    (?:title=[\'"]([^\'"]+)[\'"])?.*?>((?:(?!</a>).)*)</a>#i', $input, $match);

// $input - A STRING CONTAINING AN ANCHOR LINK WITH INNER HTML
$input = 'This is some text containing a linked image <a
    href="http://www.google.com"><img src="google.jpg" alt="Google"
    width="200" height="50" /></a>. Goodbye.';

// $match[0] RETURNS THE ENTIRE LINK INCLUDING THE INNER HTML AND ANCHOR TAGS
$match[0] == '<a href="http://www.google.com"><img src="google.jpg"
    alt="Google" width="200" height="50" /></a>';

// $match[1] RETURNS JUST THE ANCHOR LINK
$match[1] == 'http://www.google.com';

// $match[2] RETURNS THE INNER HTML, IN THIS CASE, JUST THE IMAGE TAG
$match[2] == '<img src="google.jpg" alt="Google" width="200" height="50" />';
?>

To match an email address, and to isolate either the username or domain name part of the address, use:

<?php
preg_match('#([a-z0-9&\-_.]+?)@([\w\-]+\.([\w\-\.]+\.)*[\w]+)#i',
    $input, $match);

// $input - A STRING CONTAINING AN EMAIL ADDRESS
$input = 'My email address is spam-a-lot@mydomain.com. Write me!';

// $match[0] RETURNS ONLY THE VALID EMAIL ADDRESS
$match[0] == 'spam-a-lot@mydomain.com';

// $match[1] RETURNS ONLY THE USERNAME PART
$match[1] == 'spam-a-lot';

// $match[2] RETURNS ONLY THE DOMAIN NAME PART
$match[2] == 'mydomain.com';
?>

To match an email address, and replace it with a clickable link to the email address, use:

<?php
$ret = preg_replace("#([a-z0-9&\-_.]+?)@([\w\-]+\.([\w\-\.]+\.)*[\w]+)#i",
    "\\1<a href="mailto:\\2@\\3">\\2@\\3</a>", $input);

// $input - A STRING CONTAINING AN EMAIL ADDRESS
$input = 'My email address is spam-a-lot@mydomain.com. Write me!';

// $ret - RETURNED STRING CONTAINING A CLICKABLE EMAIL LINK
$ret == 'My email address is <a
    href="mailto:spam-a-lot@mydomain.com">spam-a-lot@mydomain.com</a>.
    Write me!';
?>