Made the slugger compatible with every language #389
Conversation
| @@ -27,6 +27,6 @@ class Slugger | |||
| */ | |||
| public function slugify($string) | |||
| { | |||
| return trim(preg_replace('/[^a-z0-9]+/', '-', strtolower(strip_tags($string))), '-'); | |||
| return str_replace(' ', '-', strtolower(trim(strip_tags($string)))); | |||
There was a problem hiding this comment.
you should replace any whitespace char, not only spaces. And you should not put several dashes for consecutive spaces.
There was a problem hiding this comment.
Will /\s+/ regex pattern be enough here?
There was a problem hiding this comment.
@bocharsky-bw yes, I did that at the end 👍
| @@ -27,6 +27,6 @@ class Slugger | |||
| */ | |||
| public function slugify($string) | |||
| { | |||
| return trim(preg_replace('/[^a-z0-9]+/', '-', strtolower(strip_tags($string))), '-'); | |||
| return str_replace(' ', '-', strtolower(trim(strip_tags($string)))); | |||
There was a problem hiding this comment.
btw, if you want to support any language, you should use mb_strtolower
|
There are some failing tests only for PHP 5.5. I don't understand why because we use Symfony's polyfill for mb_strtolower(): |
|
@javiereguiluz, Could we use the slugs without changing their cases? |
|
@javiereguiluz, I think it doesn't work because you use unsupported symbols like |
|
@voronkovich I took that from the Wikipedia 😱 Is it wrong?
|
| @@ -27,6 +27,6 @@ class Slugger | |||
| */ | |||
| public function slugify($string) | |||
| { | |||
| return trim(preg_replace('/[^a-z0-9]+/', '-', strtolower(strip_tags($string))), '-'); | |||
| return preg_replace('/\s+/', '-', mb_strtolower(trim(strip_tags($string)))); | |||
There was a problem hiding this comment.
you should force the encoding to UTF-8 for mb_strtolower. The internal encoding of mbstring might be configured to a different one
There was a problem hiding this comment.
note that this is very likely, as the default one was not UTF-8 on PHP 5 (PHP 7 changed the default encoding to UTF-8)
|
@javiereguiluz, it's not wrong but those symbols are used rarely. Try to use: |
| @@ -44,7 +44,7 @@ public function getSlugs() | |||
| yield ['!Lorem Ipsum!', '!lorem-ipsum!']; | |||
| yield ['lorem-ipsum', 'lorem-ipsum']; | |||
| yield ['lorem 日本語 ipsum', 'lorem-日本語-ipsum']; | |||
| yield ['lorem ру́сский язы́к ipsum', 'lorem-ру́сский-язы́к-ipsum']; | |||
| yield ['lorem русский языќк ipsum', 'lorem-русский язык-ipsum']; | |||
There was a problem hiding this comment.
@javiereguiluz, you forgot to delete odd ќ
|
And it's working 🎉 Thank you all for your help! |
In the Symfony Installer we recently made a change to forbid any non-Latin character in the project names (see details). This is needed because Composer itself requires that for package names.
In the Symfony Demo app we had a similar problem with the slugs of the blog posts. We removed any non-Latin character ... but that caused bugs like #136. I propose to change the slugger completely and accept any character.
For example, you can now create this blog post:
And the blog listing no longer displays an error:
And the URL is perfectly fine although it contains non-Latin chars: