Regular expression - any text to URL friendly one

PHP regular expression script to remove anything that is not a alphabetical letter or number 0 to 9 and replace space to a hyphen - change to lowercase make sure there is only one hyphen - between words no -- or --- etc.

For example:

Example: The quick brown fox jumped Result: the-quick-brown-fox-jumped

Example: The quick brown开发者_StackOverflow社区 fox jumped! Result: the-quick-brown-fox-jumped

Example: The quick brown fox - jumped! Result: the-quick-brown-fox-jumped

Example: The quick ~`!@#$%^ &*()_+= ------- brown {}|][ :"'; <>?.,/ fox - jumped! Result: the-quick-brown-fox-jumped

Example: The quick 1234567890 ~`!@#$%^ &*()_+= ------- brown {}|][ :"'; <>?.,/ fox - jumped! Result: the-quick-1234567890-brown-fox-jumped

Anybody have idea for the regular expression?

Thanks!

Since you seem to want all sequences of non-alphanumeric characters being replaced by a single hyphen, you can use this:

$str = preg_replace('/[^a-zA-Z0-9]+/', '-', $str);

But this can result in leading or trailing hyphens that can be removed with trim:

$str = trim($str, '-');

And to convert the result into lowercase, use strtolower:

$str = strtolower($str);

So all together:

$str = strtolower($str);
$str = trim($str, '-');
$str = preg_replace('/[^a-z0-9]+/', '-', $str);

Or in a compact one-liner:

$str = strtolower(trim(preg_replace('/[^a-zA-Z0-9]+/', '-', $str), '-'));

I was just working with something similar, and I came up with this little piece of code, it also contemplates the use of latin characters.

This is the sample string:

$str = 'El veloz murciélago hindú comía fe<!>&@#$%&!"#%&?¡?*liz cardillo y kiwi. La cigüeña ¨^;.-|°¬tocaba el saxofón detrás del palenque de paja';

First I convert the string to htmlentities just to make it easier to use later.

$friendlyURL = htmlentities($str, ENT_COMPAT, "UTF-8", false);

Then I replace latin characters with their corresponding ascii characters (á becomes a, Ü becomes U, and so on):

$friendlyURL = preg_replace('/&([a-z]{1,2})(?:acute|circ|lig|grave|ring|tilde|uml|cedil|caron);/i','\1',$friendlyURL);

Then I convert the string back from html entities to symbols, again for easier use later.

$friendlyURL = html_entity_decode($friendlyURL,ENT_COMPAT, "UTF-8");

Next I replace all non alphanumeric characters into hyphens.

$friendlyURL = preg_replace('/[^a-z0-9-]+/i', '-', $friendlyURL);

I remove extra hyphens inside the string:

$friendlyURL = preg_replace('/-+/', '-', $friendlyURL);

I remove leading and trailing hyphens:

$friendlyURL = trim($friendlyURL, '-');

And finally convert all into lowercase:

$friendlyURL = strtolower($friendlyURL);

All together:

function friendlyUrl ($str = '') {

    $friendlyURL = htmlentities($str, ENT_COMPAT, "UTF-8", false); 
    $friendlyURL = preg_replace('/&([a-z]{1,2})(?:acute|circ|lig|grave|ring|tilde|uml|cedil|caron);/i','\1',$friendlyURL);
    $friendlyURL = html_entity_decode($friendlyURL,ENT_COMPAT, "UTF-8"); 
    $friendlyURL = preg_replace('/[^a-z0-9-]+/i', '-', $friendlyURL);
    $friendlyURL = preg_replace('/-+/', '-', $friendlyURL);
    $friendlyURL = trim($friendlyURL, '-');
    $friendlyURL = strtolower($friendlyURL);
    return $friendlyURL;

}

Test:

$str = 'El veloz murciélago hindú comía fe<!>&@#$%&!"#%&-?¡?*-liz cardillo y kiwi. La cigüeña ¨^`;.-|°¬tocaba el saxofón detrás del palenque de paja';

echo friendlyUrl($str);

Outcome:

el-veloz-murcielago-hindu-comia-fe-liz-cardillo-y-kiwi-la-ciguena-tocaba-el-saxofon-detras-del-palenque-de-paja

I guess Gumbo's answer fits your problem better, and it's a shorter code, but I thought it would be useful for others.

Cheers, Adriana

In a function:

function sanitize_text_for_urls ($str) 
{
    return trim( strtolower( preg_replace(
        array('/[^a-z0-9-\s]/ui', '/\s/', '/-+/'),
        array('', '-', '-'),
        iconv('UTF-8', 'ASCII//TRANSLIT', $str) )), '-');
}

What it does:

// Solve accents and diacritics
$str = iconv('UTF-8', 'ASCII//TRANSLIT', $str);

// Leave only alphanumeric (respect existing hyphens)
$str = preg_replace('/[^a-z0-9-\s]/ui', '', $str);

// Turn spaces to hyphens
$str = preg_replace('/\s+/', '-', $str);

// Remove duplicate hyphens
$str = preg_replace('/-+/', '-', $str);

// Remove trailing hyphens
$str = trim($str, '-');

// Turn to lowercase
$str = strtolower($str);

Note:
You can combine multiple preg_replace by passing an array. See the function at the top.

For example:

// Électricité, plâtrerie    -->  electricite-platrerie
// St. Lücie-Pétêrès         -->  st-lucie-peteres
// -Façade- & gros œuvre     -->  facade-gros-oeuvre

// _-Thè quîck ~`!@#&$%^ &*()_+= ---{}|][ :"; <>?.,/ fóx - jümpëd_-
// the-quick-fox-jumped

EDIT: added "/u" at the end of the regex to use UTF8
EDIT: accounted for duplicated and leading/trailing hyphens, thanks to @LuBre

If you're using this for filenames in PHP, the answer by Gumbo would be

$str = preg_replace('/[^a-zA-Z0-9.]+/', '-', $str);
$str = trim($str, '-');
$str = strtolower($str);

Added a period for file names and it's strtolower(), not strtolowercase().

$str = preg_replace('/[^a-zA-Z0-9]/', '-', $str);