Merge branch 'v25-12' into development

Author: ssddanbrown

.env.example.complete

@@ -351,10 +351,25 @@ EXPORT_PDF_COMMAND_TIMEOUT=15
 # Only used if 'ALLOW_UNTRUSTED_SERVER_FETCHING=true' which disables security protections.
 WKHTMLTOPDF=false
 
-# Allow <script> tags in page content
+# Allow JavaScript, and other potentiall dangerous content in page content.
+# This also removes CSP-level JavaScript control.
 # Note, if set to 'true' the page editor may still escape scripts.
+# DEPRECATED: Use 'APP_CONTENT_FILTERING' instead as detailed below. Activiting this option
+# effectively sets APP_CONTENT_FILTERING='' (No filtering)
 ALLOW_CONTENT_SCRIPTS=false
 
+# Control the behaviour of content filtering, primarily used for page content.
+# This setting is a string of characters which represent different available filters:
+# - j - Filter out JavaScript and unknown binary data based content
+# - h - Filter out unexpected, and potentially dangerous, HTML elements
+# - f - Filter out unexpected form elements
+# - a - Run content through a more complex allowlist filter
+# This defaults to using all filters, unless ALLOW_CONTENT_SCRIPTS is set to true in which case no filters are used.
+# Note: These filters are a best-attempt and may not be 100% effective. They are typically a layer used in addition to other security measures.
+# Note: The default value will always be the most-strict, so it's advised to leave this unset in your own configuration
+# to ensure you are always using the full range of filters.
+APP_CONTENT_FILTERING="jfha"
+
 # Indicate if robots/crawlers should crawl your instance.
 # Can be 'true', 'false' or 'null'.
 # The behaviour of the default 'null' option will depend on the 'app-public' admin setting.

app/Activity/Models/Comment.php

@@ -8,6 +8,7 @@
 use BookStack\Users\Models\HasCreatorAndUpdater;
 use BookStack\Users\Models\OwnableInterface;
 use BookStack\Util\HtmlContentFilter;
+use BookStack\Util\HtmlContentFilterConfig;
 use Illuminate\Database\Eloquent\Builder;
 use Illuminate\Database\Eloquent\Factories\HasFactory;
 use Illuminate\Database\Eloquent\Relations\BelongsTo;
@@ -82,7 +83,8 @@ public function logDescriptor(): string
 
     public function safeHtml(): string
     {
-        return HtmlContentFilter::removeActiveContentFromHtmlString($this->html ?? '');
+        $filter = new HtmlContentFilter(new HtmlContentFilterConfig());
+        return $filter->filterString($this->html ?? '');
     }
 
     public function jointPermissions(): HasMany

app/Config/app.php

@@ -37,19 +37,24 @@
     // The limit for all uploaded files, including images and attachments in MB.
     'upload_limit' => env('FILE_UPLOAD_SIZE_LIMIT', 50),
 
-    // Allow <script> tags to entered within page content.
-    // <script> tags are escaped by default.
-    // Even when overridden the WYSIWYG editor may still escape script content.
-    'allow_content_scripts' => env('ALLOW_CONTENT_SCRIPTS', false),
+    // Control the behaviour of content filtering, primarily used for page content.
+    // This setting is a string of characters which represent different available filters:
+    // - j - Filter out JavaScript and unknown binary data based content
+    // - h - Filter out unexpected, and potentially dangerous, HTML elements
+    // - f - Filter out unexpected form elements
+    // - a - Run content through a more complex allowlist filter
+    // This defaults to using all filters, unless ALLOW_CONTENT_SCRIPTS is set to true in which case no filters are used.
+    // Note: These filters are a best-attempt and may not be 100% effective. They are typically a layer used in addition to other security measures.
+    'content_filtering' => env('APP_CONTENT_FILTERING', env('ALLOW_CONTENT_SCRIPTS', false) === true ? '' : 'jhfa'),
 
     // Allow server-side fetches to be performed to potentially unknown
     // and user-provided locations. Primarily used in exports when loading
     // in externally referenced assets.
     'allow_untrusted_server_fetching' => env('ALLOW_UNTRUSTED_SERVER_FETCHING', false),
 
     // Override the default behaviour for allowing crawlers to crawl the instance.
-    // May be ignored if view has be overridden or modified.
-    // Defaults to null since, if not set, 'app-public' status used instead.
+    // May be ignored if the underlying view has been overridden or modified.
+    // Defaults to null in which case the 'app-public' status is used instead.
     'allow_robots' => env('ALLOW_ROBOTS', null),
 
     // Application Base URL, Used by laravel in development commands

app/Entities/Controllers/PageController.php

@@ -21,6 +21,8 @@
 use BookStack\Http\Controller;
 use BookStack\Permissions\Permission;
 use BookStack\References\ReferenceFetcher;
+use BookStack\Util\HtmlContentFilter;
+use BookStack\Util\HtmlContentFilterConfig;
 use Exception;
 use Illuminate\Database\Eloquent\Relations\BelongsTo;
 use Illuminate\Http\Request;
@@ -173,7 +175,7 @@ public function show(string $bookSlug, string $pageSlug)
     }
 
     /**
-     * Get page from an ajax request.
+     * Get a page from an ajax request.
      *
      * @throws NotFoundException
      */
@@ -183,6 +185,10 @@ public function getPageAjax(int $pageId)
         $page->setHidden(array_diff($page->getHidden(), ['html', 'markdown']));
         $page->makeHidden(['book']);
 
+        $filterConfig = HtmlContentFilterConfig::fromConfigString(config('app.content_filtering'));
+        $filter = new HtmlContentFilter($filterConfig);
+        $page->html = $filter->filterString($page->html);
+
         return response()->json($page);
     }
 

app/Entities/Tools/EntityHtmlDescription.php

@@ -6,6 +6,7 @@
 use BookStack\Entities\Models\Bookshelf;
 use BookStack\Entities\Models\Chapter;
 use BookStack\Util\HtmlContentFilter;
+use BookStack\Util\HtmlContentFilterConfig;
 
 class EntityHtmlDescription
 {
@@ -55,7 +56,8 @@ public function getHtml(bool $raw = false): string
             return '<p></p>';
         }
 
-        return HtmlContentFilter::removeActiveContentFromHtmlString($html);
+        $filter = new HtmlContentFilter(new HtmlContentFilterConfig());
+        return $filter->filterString($html);
     }
 
     public function getPlain(): string

app/Entities/Tools/PageContent.php

@@ -2,6 +2,7 @@
 
 namespace BookStack\Entities\Tools;
 
+use BookStack\App\AppVersion;
 use BookStack\Entities\Models\Page;
 use BookStack\Entities\Queries\PageQueries;
 use BookStack\Entities\Tools\Markdown\MarkdownToHtml;
@@ -13,6 +14,7 @@
 use BookStack\Uploads\ImageService;
 use BookStack\Users\Models\User;
 use BookStack\Util\HtmlContentFilter;
+use BookStack\Util\HtmlContentFilterConfig;
 use BookStack\Util\HtmlDocument;
 use BookStack\Util\WebSafeMimeSniffer;
 use Closure;
@@ -317,11 +319,30 @@ public function render(bool $blankIncludes = false): string
             $this->updateIdsRecursively($doc->getBody(), 0, $idMap, $changeMap);
         }
 
-        if (!config('app.allow_content_scripts')) {
-            HtmlContentFilter::removeActiveContentFromDocument($doc);
+        $cacheKey = $this->getContentCacheKey($doc->getBodyInnerHtml());
+        $cached = cache()->get($cacheKey, null);
+        if ($cached !== null) {
+            return $cached;
         }
 
-        return $doc->getBodyInnerHtml();
+        $filterConfig = HtmlContentFilterConfig::fromConfigString(config('app.content_filtering'));
+        $filter = new HtmlContentFilter($filterConfig);
+        $filtered = $filter->filterDocument($doc);
+
+        $cacheTime = 86400 * 7; // 1 week
+        cache()->put($cacheKey, $filtered, $cacheTime);
+
+        return $filtered;
+    }
+
+    protected function getContentCacheKey(string $html): string
+    {
+        $contentHash = md5($html);
+        $contentId = $this->page->id;
+        $contentTime = $this->page->updated_at?->timestamp ?? time();
+        $appVersion = AppVersion::get();
+        $filterConfig = config('app.content_filtering') ?? '';
+        return "page-content-cache::{$filterConfig}::{$appVersion}::{$contentId}::{$contentTime}::{$contentHash}";
     }
 
     /**

app/Entities/Tools/PageEditorData.php

@@ -8,6 +8,8 @@
 use BookStack\Entities\Tools\Markdown\HtmlToMarkdown;
 use BookStack\Entities\Tools\Markdown\MarkdownToHtml;
 use BookStack\Permissions\Permission;
+use BookStack\Util\HtmlContentFilter;
+use BookStack\Util\HtmlContentFilterConfig;
 
 class PageEditorData
 {
@@ -47,6 +49,7 @@ protected function build(): array
         $isDraftRevision = false;
         $this->warnings = [];
         $editActivity = new PageEditActivity($page);
+        $lastEditorId = $page->updated_by ?? user()->id;
 
         if ($editActivity->hasActiveEditing()) {
             $this->warnings[] = $editActivity->activeEditingMessage();
@@ -58,11 +61,20 @@ protected function build(): array
             $page->forceFill($userDraft->only(['name', 'html', 'markdown']));
             $isDraftRevision = true;
             $this->warnings[] = $editActivity->getEditingActiveDraftMessage($userDraft);
+            $lastEditorId = $userDraft->created_by;
         }
 
+        // Get editor type and handle changes
         $editorType = $this->getEditorType($page);
         $this->updateContentForEditor($page, $editorType);
 
+        // Filter HTML content if required
+        if ($editorType->isHtmlBased() && !old('html') && $lastEditorId !== user()->id) {
+            $filterConfig = HtmlContentFilterConfig::fromConfigString(config('app.content_filtering'));
+            $filter = new HtmlContentFilter($filterConfig);
+            $page->html = $filter->filterString($page->html);
+        }
+
         return [
             'page'            => $page,
             'book'            => $page->book,

app/Theming/CustomHtmlHeadContentProvider.php

@@ -4,25 +4,16 @@
 
 use BookStack\Util\CspService;
 use BookStack\Util\HtmlContentFilter;
+use BookStack\Util\HtmlContentFilterConfig;
 use BookStack\Util\HtmlNonceApplicator;
 use Illuminate\Contracts\Cache\Repository as Cache;
 
 class CustomHtmlHeadContentProvider
 {
-    /**
-     * @var CspService
-     */
-    protected $cspService;
-
-    /**
-     * @var Cache
-     */
-    protected $cache;
-
-    public function __construct(CspService $cspService, Cache $cache)
-    {
-        $this->cspService = $cspService;
-        $this->cache = $cache;
+    public function __construct(
+        protected CspService $cspService,
+        protected Cache $cache
+    ) {
     }
 
     /**
@@ -50,7 +41,8 @@ public function forExport(): string
         $hash = md5($content);
 
         return $this->cache->remember('custom-head-export:' . $hash, 86400, function () use ($content) {
-            return HtmlContentFilter::removeActiveContentFromHtmlString($content);
+            $config = new HtmlContentFilterConfig(filterOutNonContentElements: false, useAllowListFilter: false);
+            return (new HtmlContentFilter($config))->filterString($content);
         });
     }