Merge pull request #15 from chrisrhymes/feature/support-relative-links

Add support for relative urls in html fields

.github/workflows/laravel.yml

@@ -5,19 +5,19 @@ on:
     branches:
       - main
       - develop
-      - "feature/**"
+      - 'feature/**'
   pull_request:
-    branches: ["main", "develop"]
+    branches: ['main', 'develop']
 
 jobs:
   laravel-tests:
-    runs-on: ubuntu-22.04
+    runs-on: ubuntu-20.04
 
     steps:
-      - uses: shivammathur/setup-php@15c43e89cdef867065b0213be354c2841860869e
+      - uses: shivammathur/setup-php@v2
         with:
-          php-version: "8.1"
-      - uses: actions/checkout@v3
+          php-version: '8.2'
+      - uses: actions/checkout@v4
       - name: Install Dependencies
         run: composer install -q --no-ansi --no-interaction --no-scripts --no-progress --prefer-dist
       - name: Create Database

.github/workflows/pint.yml

@@ -8,7 +8,7 @@ jobs:
   phplint:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
         with:
           fetch-depth: 2
       - name: 'laravel-pint'

README.md

@@ -9,6 +9,7 @@ A package that will check for broken links in the specified model's fields. It w
 
 - [Getting Started](#getting-started)
 - [Usage](#usage)
+  - [Relative links](#relative-links)
 - [Rate Limiting](#rate-limiting)
 - [User Agent](#user-agent)
 - [Verify SSL](#verify-ssl)
@@ -96,6 +97,25 @@ $post->brokenLinks[0]->broken_link; // The link that is broken
 $post->brokenLinks[0]->exception_message; // The optional exception message
 ```
 
+### Relative links
+
+If you have relative links within a html field in your model (that don't begin with 'http'), then you can pass a 3rd parameter as the base. The CheckModelForBrokenLinks job will prepend the base to the relative url before it is checked.
+
+If your relative links don't begin with `/`, then ensure your base parameter has a trailing slash, `'http://example.com/'`.
+
+```php
+use ChrisRhymes\LinkChecker\Jobs\CheckModelForBrokenLinks;
+use ChrisRhymes\LinkChecker\Facades\LinkChecker;
+
+$post = Post::first();
+
+// Dispatch the job directly
+CheckModelForBrokenLinks::dispatch($post, ['content', 'url'], 'http://example.com');
+
+// Or using the facade
+LinkChecker::checkForBrokenLinks($post, ['content', 'url'], 'http://example.com');
+```
+
 ## Rate Limiting
 
 In order to reduce the amount of requests sent to a domain at a time, this package has rate limiting enabled.

src/Jobs/CheckModelForBrokenLinks.php

@@ -24,15 +24,18 @@ class CheckModelForBrokenLinks implements ShouldQueue
 
     private array $links = [];
 
+    private ?string $base = null;
+
     /**
      * Create a new job instance.
      *
      * @return void
      */
-    public function __construct(Model $model, array $fields)
+    public function __construct(Model $model, array $fields, ?string $base = null)
     {
         $this->model = $model;
         $this->fields = $fields;
+        $this->base = $base;
     }
 
     /**
@@ -65,8 +68,16 @@ public function handle()
                     $anchorTags = $doc->getElementsByTagName('a');
 
                     foreach ($anchorTags as $anchorTag) {
+                        $href = $anchorTag->getAttribute('href');
+
+                        if (Str::startsWith($href, ['mailto', 'tel'])) {
+                            continue;
+                        }
+
                         $link = new Link;
-                        $link->url = $anchorTag->getAttribute('href');
+                        $link->url = ! Str::startsWith($href, ['http://', 'https://']) && $this->base
+                            ? $this->base.$href
+                            : $href;
                         $link->text = $anchorTag->nodeValue;
 
                         $this->links[] = $link;

src/LinkChecker.php

@@ -7,8 +7,8 @@
 
 class LinkChecker
 {
-    public function checkForBrokenLinks(Model $model, array $fields)
+    public function checkForBrokenLinks(Model $model, array $fields, ?string $base = null)
     {
-        CheckModelForBrokenLinks::dispatch($model, $fields);
+        CheckModelForBrokenLinks::dispatch($model, $fields, $base);
     }
 }

tests/Feature/AvoidTelMailtoLinksTest.php

@@ -21,4 +21,14 @@
     expect($this->post->fresh()->brokenLinks)->toHaveCount(0);
 
     Http::assertNothingSent();
+
+    $this->assertDatabaseCount('broken_links', 0);
+});
+
+it('does not check tel and mailto links when base is added', function () {
+    CheckModelForBrokenLinks::dispatch($this->post, ['content'], 'https://this-is-a-relative-link.com/');
+
+    Http::assertNothingSent();
+
+    $this->assertDatabaseCount('broken_links', 0);
 });

tests/Feature/RelativeLinkTest.php

@@ -0,0 +1,66 @@
+<?php
+
+use ChrisRhymes\LinkChecker\Facades\LinkChecker;
+use ChrisRhymes\LinkChecker\Jobs\CheckModelForBrokenLinks;
+use ChrisRhymes\LinkChecker\Test\Models\Post;
+use Illuminate\Support\Facades\Http;
+
+beforeEach(function () {
+    Http::fake([
+        'https://this-is-a-relative-link.com/relative' => Http::response(null, 302),
+    ]);
+
+    $this->post = Post::factory()
+        ->create([
+            'content' => '
+                <a href="relative">Temporary redirect link</a>',
+        ]);
+});
+
+it('records relative urls using base url', function () {
+    CheckModelForBrokenLinks::dispatch($this->post, ['content'], 'https://this-is-a-relative-link.com/');
+
+    $this->assertDatabaseHas('broken_links', [
+        'linkable_id' => $this->post->id,
+        'linkable_type' => 'ChrisRhymes\LinkChecker\Test\Models\Post',
+        'broken_link' => 'https://this-is-a-relative-link.com/relative',
+        'link_text' => 'Temporary redirect link',
+        'exception_message' => '302 Redirect',
+    ]);
+});
+
+it('records relative urls using base url using the facade', function () {
+    LinkChecker::checkForBrokenLinks($this->post, ['content'], 'https://this-is-a-relative-link.com/');
+
+    $this->assertDatabaseHas('broken_links', [
+        'linkable_id' => $this->post->id,
+        'linkable_type' => 'ChrisRhymes\LinkChecker\Test\Models\Post',
+        'broken_link' => 'https://this-is-a-relative-link.com/relative',
+        'link_text' => 'Temporary redirect link',
+        'exception_message' => '302 Redirect',
+    ]);
+});
+
+it('records curl error for relative urls without setting base', function () {
+    CheckModelForBrokenLinks::dispatch($this->post, ['content']);
+
+    $this->assertDatabaseHas('broken_links', [
+        'linkable_id' => $this->post->id,
+        'linkable_type' => 'ChrisRhymes\LinkChecker\Test\Models\Post',
+        'broken_link' => 'relative',
+        'link_text' => 'Temporary redirect link',
+        'exception_message' => 'cURL error 6: Could not resolve host: relative (see https://curl.haxx.se/libcurl/c/libcurl-errors.html)',
+    ]);
+});
+
+it('records curl error for relative urls without setting base using the facade', function () {
+    LinkChecker::checkForBrokenLinks($this->post, ['content']);
+
+    $this->assertDatabaseHas('broken_links', [
+        'linkable_id' => $this->post->id,
+        'linkable_type' => 'ChrisRhymes\LinkChecker\Test\Models\Post',
+        'broken_link' => 'relative',
+        'link_text' => 'Temporary redirect link',
+        'exception_message' => 'cURL error 6: Could not resolve host: relative (see https://curl.haxx.se/libcurl/c/libcurl-errors.html)',
+    ]);
+});

tests/Models/Post.php

@@ -16,7 +16,7 @@
  */
 class Post extends Model
 {
-    use HasFactory, HasBrokenLinks;
+    use HasBrokenLinks, HasFactory;
 
     protected $fillable = [
         'title',