URL Friendly Username in PHP?

Question

On my PHP site, currently users login with an email address and a password. I would like to add a username as well, this username they g\set will be unique and they cannot change it. I am wondering how I can make this name have no spaces in it and work in a URL so I can use there username to link to there profiles and other stuff. If there is a space in there username then it should add an underscore jason_davis. I am not sure the best way to do this?

Solution 1

function Slug($string)
{
    return strtolower(trim(preg_replace('~[^0-9a-z]+~i', '-', html_entity_decode(preg_replace('~&([a-z]{1,2})(?:acute|cedil|circ|grave|lig|orn|ring|slash|th|tilde|uml);~i', '$1', htmlentities($string, ENT_QUOTES, 'UTF-8')), ENT_QUOTES, 'UTF-8')), '-'));
}

$user = 'Alix Axel';
echo Slug($user); // alix-axel

$user = 'Álix Ãxel';
echo Slug($user); // alix-axel

$user = 'Álix----_Ãxel!?!?';
echo Slug($user); // alix-axel

Solution 2

In other words... you need to create a username slug. Doctrine (ORM for PHP) has a nice function to do it. Doctrine_Inflector::urlize()

EDIT: You should also keep username slug in database, as a Unique Key column. Then every search operation should be done based on that column, not original username.

Comments

[Gumbo]

There are plenty questions like this. Didn’t you get an answer with searching?

[JasonDavis]

@Gumbo I searched SO, not google. Possibly not the correct term but I did search for "URL friendly username" with not much luck. I didn't know it was called a slug before this.

[Gumbo]

Maybe not everyone is trying to convert usernames. But searching for “URL friendly string” is returning usable results.

[GG.]

Similar: stackoverflow.com/questions/5305879

[ausi]

Nowadays, you can use libraries like github.com/cocur/slugify or github.com/ausi/slug-generator to achieve that.

[John Feminella]

This is dangerous! Multiple unique user names can map to the same URL. That's not what you want, is it? Consider, e.g., AB and ab, which are unique strings but map to the same slug string. You should store the slug as the identifier.

[Pekka]

@John Feminella: He would obviously have to check for duplicates at some point before storing the slug.

[JasonDavis]

perfect, thank you. BTW what does this part look for acute|cedil|circ|grave|lig|orn|ring|slash|th|tilde|uml ?

[Pekka]

@jasondavis: För stüff lıkë thîs!

[Alix Axel]

@Pekka: Thanks! =) @jasondavid: It removes accents, instead of having a huge lookup table we convert to html entities and fetch the unaccented char.

[Frank Farmer]

As a slight improvement, using iconv() to convert to ASCII//TRANSLIT would probably catch a lot more chars.

[Alix Axel]

@Frank Farmer: iconv('UTF-8', 'ASCII//TRANSLIT', 'Álix Ãxel') returns 'Alix ~Axel. Using PHP 5.3.0 and .php file encoded as UTF-8 no BOM.

[John Conde]

Anyone know why Á and à would cause no output or error to occur when using this function?

[John Conde]

I'll answer my own question: remove the "'UTF-8'" parameter from htmlentities. That did the trick.

[Alix Axel]

@John Conde: I though you're talking about the iconv() function! You shouldn't remove the UTF-8 from htmlentities, instead you should save all your .php files encoded as UTF-8 no BOM.

[John Conde]

Alex, thanks for the info. I reutnred the "UTF-8" parameter and saved the file as UTF-8 and it worked like a charm.

[Alix Axel]

@John Conde: No problem! ;) You should always save your files UTF-8 encoded.

[s3m3n]

This function does not convert Polish chars "ąćęłńóśźż" => "acelnoszz", because they have no names in html entities (only numeric representation). You still need to replace those with table. Paste below code before return in function: $string = strtr(mb_strtolower($string), array('ą'=>'a','ć'=>'c','ę'=>'e','ł'=>'l','ń'=>'n','ó'=>'o',‌​'ś'=>'s','ź'=>'z','ż‌​'=>'z'));

[Alix Axel]

@s3men: Indeed, not only Polish characters but others as well (Chinese, Japanese, Turkish, Arabic, ...).

[s3m3n]

After few more tests iconv made the trick, as someone wrote above. setlocale(LC_ALL, "en_US.utf8"); $string = iconv("UTF-8", "ascii//TRANSLIT", '>ĄĘŁŹÓżół<'); gives me >AELZOzol<

[Jared Farrish]

codepad demo of the Slug() function, with a second identical but spaced out nSlug() function (for the eyeball impaired): codepad.org/rJNSQmGJ