feat(PageContent): set unique ids on nested headers

[bookstack] / app / Entities / Tools / PageContent.php

diff --git a/app/Entities/Tools/PageContent.php b/app/Entities/Tools/PageContent.php
index 661c554da4809c799d38f2aba8321756ca992019..c8204a18186893c874f03dcdb65e820e03c3d638 100644 (file)
--- a/app/Entities/Tools/PageContent.php
+++ b/app/Entities/Tools/PageContent.php
@@ -9,6 +9,7 @@ use BookStack\Exceptions\ImageUploadException;
 use BookStack\Facades\Theme;
 use BookStack\Theming\ThemeEvents;
 use BookStack\Uploads\ImageRepo;
+use BookStack\Uploads\ImageService;
 use BookStack\Util\HtmlContentFilter;
 use DOMDocument;
 use DOMNodeList;
@@ -37,7 +38,7 @@ class PageContent
      */
     public function setNewHTML(string $html)
     {
-        $html = $this->extractBase64Images($this->page, $html);
+        $html = $this->extractBase64ImagesFromHtml($html);
         $this->page->html = $this->formatHtml($html);
         $this->page->text = $this->toPlainText();
         $this->page->markdown = '';
@@ -48,6 +49,7 @@ class PageContent
      */
     public function setNewMarkdown(string $markdown)
     {
+        $markdown = $this->extractBase64ImagesFromMarkdown($markdown);
         $this->page->markdown = $markdown;
         $html = $this->markdownToHtml($markdown);
         $this->page->html = $this->formatHtml($html);
@@ -74,7 +76,7 @@ class PageContent
     /**
      * Convert all base64 image data to saved images.
      */
-    public function extractBase64Images(Page $page, string $htmlText): string
+    protected function extractBase64ImagesFromHtml(string $htmlText): string
     {
         if (empty($htmlText) || strpos($htmlText, 'data:image') === false) {
             return $htmlText;
@@ -85,31 +87,13 @@ class PageContent
         $body = $container->childNodes->item(0);
         $childNodes = $body->childNodes;
         $xPath = new DOMXPath($doc);
-        $imageRepo = app()->make(ImageRepo::class);
-        $allowedExtensions = ['jpg', 'jpeg', 'png', 'gif', 'webp'];
 
         // Get all img elements with image data blobs
         $imageNodes = $xPath->query('//img[contains(@src, \'data:image\')]');
         foreach ($imageNodes as $imageNode) {
             $imageSrc = $imageNode->getAttribute('src');
-            [$dataDefinition, $base64ImageData] = explode(',', $imageSrc, 2);
-            $extension = strtolower(preg_split('/[\/;]/', $dataDefinition)[1] ?? 'png');
-
-            // Validate extension
-            if (!in_array($extension, $allowedExtensions)) {
-                $imageNode->setAttribute('src', '');
-                continue;
-            }
-
-            // Save image from data with a random name
-            $imageName = 'embedded-image-' . Str::random(8) . '.' . $extension;
-
-            try {
-                $image = $imageRepo->saveNewFromData($imageName, base64_decode($base64ImageData), 'gallery', $page->id);
-                $imageNode->setAttribute('src', $image->url);
-            } catch (ImageUploadException $exception) {
-                $imageNode->setAttribute('src', '');
-            }
+            $newUrl = $this->base64ImageUriToUploadedImageUrl($imageSrc);
+            $imageNode->setAttribute('src', $newUrl);
         }
 
         // Generate inner html as a string
@@ -121,6 +105,70 @@ class PageContent
         return $html;
     }
 
+    /**
+     * Convert all inline base64 content to uploaded image files.
+     */
+    protected function extractBase64ImagesFromMarkdown(string $markdown)
+    {
+        $matches = [];
+        preg_match_all('/!\[.*?]\(.*?(data:image\/.*?)[)"\s]/', $markdown, $matches);
+
+        foreach ($matches[1] as $base64Match) {
+            $newUrl = $this->base64ImageUriToUploadedImageUrl($base64Match);
+            $markdown = str_replace($base64Match, $newUrl, $markdown);
+        }
+
+        return $markdown;
+    }
+
+    /**
+     * Parse the given base64 image URI and return the URL to the created image instance.
+     * Returns an empty string if the parsed URI is invalid or causes an error upon upload.
+     */
+    protected function base64ImageUriToUploadedImageUrl(string $uri): string
+    {
+        $imageRepo = app()->make(ImageRepo::class);
+        $imageInfo = $this->parseBase64ImageUri($uri);
+
+        // Validate extension and content
+        if (empty($imageInfo['data']) || !ImageService::isExtensionSupported($imageInfo['extension'])) {
+            return '';
+        }
+
+        // Validate that the content is not over our upload limit
+        $uploadLimitBytes = (config('app.upload_limit') * 1000000);
+        if (strlen($imageInfo['data']) > $uploadLimitBytes) {
+            return '';
+        }
+
+        // Save image from data with a random name
+        $imageName = 'embedded-image-' . Str::random(8) . '.' . $imageInfo['extension'];
+
+        try {
+            $image = $imageRepo->saveNewFromData($imageName, $imageInfo['data'], 'gallery', $this->page->id);
+        } catch (ImageUploadException $exception) {
+            return '';
+        }
+
+        return $image->url;
+    }
+
+    /**
+     * Parse a base64 image URI into the data and extension.
+     *
+     * @return array{extension: array, data: string}
+     */
+    protected function parseBase64ImageUri(string $uri): array
+    {
+        [$dataDefinition, $base64ImageData] = explode(',', $uri, 2);
+        $extension = strtolower(preg_split('/[\/;]/', $dataDefinition)[1] ?? '');
+
+        return [
+            'extension' => $extension,
+            'data'      => base64_decode($base64ImageData) ?: '',
+        ];
+    }
+
     /**
      * Formats a page's html to be tagged correctly within the system.
      */
@@ -145,6 +193,15 @@ class PageContent
             }
         }
 
+        // Set ids on nested header nodes
+        $nestedHeaders = $xPath->query('//body//*//h1|//body//*//h2|//body//*//h3|//body//*//h4|//body//*//h5|//body//*//h6');
+        foreach ($nestedHeaders as $nestedHeader) {
+            [$oldId, $newId] = $this->setUniqueId($nestedHeader, $idMap);
+            if ($newId && $newId !== $oldId) {
+                $this->updateLinks($xPath, '#' . $oldId, '#' . $newId);
+            }
+        }
+
         // Ensure no duplicate ids within child items
         $idElems = $xPath->query('//body//*//*[@id]');
         foreach ($idElems as $domElem) {
@@ -342,7 +399,7 @@ class PageContent
      */
     protected function fetchSectionOfPage(Page $page, string $sectionId): string
     {
-        $topLevelTags = ['table', 'ul', 'ol'];
+        $topLevelTags = ['table', 'ul', 'ol', 'pre'];
         $doc = $this->loadDocumentFromHtml($page->html);
 
         // Search included content for the id given and blank out if not exists.