feat: discussion UTF-8 slug driver (#3606)

* feat: add utf-8 slug driver
* test: add tests for slugging expectations
* fix: non-word characters aren't removed

Signed-off-by: Sami Mazouz <ilyasmazouz@gmail.com>
Co-authored-by: Alexander Skvortsov <sasha.skvortsov109@gmail.com>

framework/core/src/Discussion/Discussion.php

@@ -441,6 +441,7 @@ class Discussion extends AbstractModel
      *
      * This automatically creates a matching slug for the discussion.
      *
+     * @todo slug should be set by the slugger, drop slug column entirely?
      * @param string $title
      */
     protected function setTitleAttribute($title)

framework/core/src/Discussion/Utf8SlugDriver.php

@@ -0,0 +1,46 @@
+<?php
+
+/*
+ * This file is part of Flarum.
+ *
+ * For detailed copyright and license information, please view the
+ * LICENSE file that was distributed with this source code.
+ */
+
+namespace Flarum\Discussion;
+
+use Flarum\Database\AbstractModel;
+use Flarum\Http\SlugDriverInterface;
+use Flarum\User\User;
+
+class Utf8SlugDriver implements SlugDriverInterface
+{
+    /**
+     * @var DiscussionRepository
+     */
+    protected $discussions;
+
+    public function __construct(DiscussionRepository $discussions)
+    {
+        $this->discussions = $discussions;
+    }
+
+    public function toSlug(AbstractModel $instance): string
+    {
+        $slug = preg_replace('/[-\s]+/u', '-', $instance->title);
+        $slug = preg_replace('/[^\p{L}\p{N}\p{M}_-]+/u', '', $slug);
+        $slug = strtolower($slug);
+
+        return $instance->id.(trim($slug) ? '-'.$slug : '');
+    }
+
+    public function fromSlug(string $slug, User $actor): AbstractModel
+    {
+        if (strpos($slug, '-')) {
+            $slug_array = explode('-', $slug);
+            $slug = $slug_array[0];
+        }
+
+        return $this->discussions->findOrFail($slug, $actor);
+    }
+}

framework/core/src/Http/HttpServiceProvider.php

@@ -11,6 +11,7 @@ namespace Flarum\Http;
 
 use Flarum\Discussion\Discussion;
 use Flarum\Discussion\IdWithTransliteratedSlugDriver;
+use Flarum\Discussion\Utf8SlugDriver;
 use Flarum\Foundation\AbstractServiceProvider;
 use Flarum\Settings\SettingsRepositoryInterface;
 use Flarum\User\IdSlugDriver;
@@ -37,7 +38,8 @@ class HttpServiceProvider extends AbstractServiceProvider
         $this->container->singleton('flarum.http.slugDrivers', function () {
             return [
                 Discussion::class => [
-                    'default' => IdWithTransliteratedSlugDriver::class
+                    'default' => IdWithTransliteratedSlugDriver::class,
+                    'utf8' => Utf8SlugDriver::class,
                 ],
                 User::class => [
                     'default' => UsernameSlugDriver::class,

framework/core/tests/integration/slugger/SlugDriverTest.php

@@ -0,0 +1,93 @@
+<?php
+
+/*
+ * This file is part of Flarum.
+ *
+ * For detailed copyright and license information, please view the
+ * LICENSE file that was distributed with this source code.
+ */
+
+namespace integration\slugger;
+
+use Carbon\Carbon;
+use Flarum\Discussion\Discussion;
+use Flarum\Http\SlugManager;
+use Flarum\Testing\integration\RetrievesAuthorizedUsers;
+use Flarum\Testing\integration\TestCase;
+use Flarum\User\User;
+
+class SlugDriverTest extends TestCase
+{
+    use RetrievesAuthorizedUsers;
+
+    protected function setUp(): void
+    {
+        parent::setUp();
+
+        $this->prepareDatabase([
+            'users' => [
+                $this->normalUser(),
+            ],
+            'discussions' => [
+                ['id' => 20, 'title' => 'Empty discussion', 'created_at' => Carbon::now()->toDateTimeString(), 'user_id' => 2, 'first_post_id' => null, 'comment_count' => 0, 'is_private' => 0],
+                ['id' => 21, 'title' => 'తెలుగు', 'created_at' => Carbon::now()->toDateTimeString(), 'user_id' => 2, 'first_post_id' => null, 'comment_count' => 0, 'is_private' => 0],
+                ['id' => 22, 'title' => '支持中文吗', 'created_at' => Carbon::now()->toDateTimeString(), 'user_id' => 2, 'first_post_id' => null, 'comment_count' => 0, 'is_private' => 0],
+                ['id' => 23, 'title' => 'తెలుగు%$', 'created_at' => Carbon::now()->toDateTimeString(), 'user_id' => 2, 'first_post_id' => null, 'comment_count' => 0, 'is_private' => 0],
+                ['id' => 24, 'title' => '支持中文吗%*', 'created_at' => Carbon::now()->toDateTimeString(), 'user_id' => 2, 'first_post_id' => null, 'comment_count' => 0, 'is_private' => 0],
+            ],
+        ]);
+    }
+
+    /**
+     * @dataProvider slugInstancePairDataProvider
+     * @test
+     */
+    public function slugger_formats_the_correct_slug_from_instance(string $driver, string $modelClassName, int $id, string $slug)
+    {
+        $this->setting("slug_driver_$modelClassName", $driver);
+
+        /** @var SlugManager $slugger */
+        $slugger = $this->app()->getContainer()->make(SlugManager::class)->forResource($modelClassName);
+
+        $instance = $modelClassName::query()->find($id);
+
+        /** @see Discussion::setTitleAttribute() */
+        if ($modelClassName === Discussion::class) {
+            $instance->title = $instance->title;
+        }
+
+        $this->assertEquals($slug, $slugger->toSlug($instance));
+    }
+
+    /**
+     * @dataProvider slugInstancePairDataProvider
+     * @test
+     */
+    public function slugger_returns_the_correct_instance_from_slug(string $driver, string $modelClassName, int $id, string $slug)
+    {
+        $this->setting("slug_driver_$modelClassName", $driver);
+
+        /** @var SlugManager $slugger */
+        $slugger = $this->app()->getContainer()->make(SlugManager::class)->forResource($modelClassName);
+
+        $this->assertEquals($modelClassName::query()->find($id), $slugger->fromSlug($slug, User::query()->find(1)));
+    }
+
+    public function slugInstancePairDataProvider(): array
+    {
+        return [
+            ['default', Discussion::class, 20, '20-empty-discussion'],
+            ['default', Discussion::class, 21, '21'],
+            ['default', Discussion::class, 22, '22'],
+
+            ['utf8', Discussion::class, 20, '20-empty-discussion'],
+            ['utf8', Discussion::class, 21, '21-తెలుగు'],
+            ['utf8', Discussion::class, 22, '22-支持中文吗'],
+            ['utf8', Discussion::class, 23, '23-తెలుగు'],
+            ['utf8', Discussion::class, 24, '24-支持中文吗'],
+
+            ['default', User::class, 2, 'normal'],
+            ['id', User::class, 2, '2'],
+        ];
+    }
+}