X-Git-Url: http://source.bookstackapp.com/bookstack/blobdiff_plain/d3a96451614201ca1b92bd3dffa9726d0b48a203..refs/pull/3918/head:/app/Entities/Tools/PageContent.php diff --git a/app/Entities/Tools/PageContent.php b/app/Entities/Tools/PageContent.php index 661c554da..9aa2e8d35 100644 --- a/app/Entities/Tools/PageContent.php +++ b/app/Entities/Tools/PageContent.php @@ -3,26 +3,23 @@ namespace BookStack\Entities\Tools; use BookStack\Entities\Models\Page; -use BookStack\Entities\Tools\Markdown\CustomListItemRenderer; -use BookStack\Entities\Tools\Markdown\CustomStrikeThroughExtension; +use BookStack\Entities\Tools\Markdown\MarkdownToHtml; 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 DOMElement; +use DOMNode; use DOMNodeList; use DOMXPath; use Illuminate\Support\Str; -use League\CommonMark\Block\Element\ListItem; -use League\CommonMark\CommonMarkConverter; -use League\CommonMark\Environment; -use League\CommonMark\Extension\Table\TableExtension; -use League\CommonMark\Extension\TaskList\TaskListExtension; class PageContent { - protected $page; + protected Page $page; /** * PageContent constructor. @@ -37,7 +34,7 @@ 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 = ''; @@ -48,33 +45,17 @@ class PageContent */ public function setNewMarkdown(string $markdown) { + $markdown = $this->extractBase64ImagesFromMarkdown($markdown); $this->page->markdown = $markdown; - $html = $this->markdownToHtml($markdown); + $html = (new MarkdownToHtml($markdown))->convert(); $this->page->html = $this->formatHtml($html); $this->page->text = $this->toPlainText(); } - /** - * Convert the given Markdown content to a HTML string. - */ - protected function markdownToHtml(string $markdown): string - { - $environment = Environment::createCommonMarkEnvironment(); - $environment->addExtension(new TableExtension()); - $environment->addExtension(new TaskListExtension()); - $environment->addExtension(new CustomStrikeThroughExtension()); - $environment = Theme::dispatch(ThemeEvents::COMMONMARK_ENVIRONMENT_CONFIGURE, $environment) ?? $environment; - $converter = new CommonMarkConverter([], $environment); - - $environment->addBlockRenderer(ListItem::class, new CustomListItemRenderer(), 10); - - return $converter->convertToHtml($markdown); - } - /** * 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 +66,13 @@ 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 +84,90 @@ class PageContent return $html; } + /** + * Convert all inline base64 content to uploaded image files. + * Regex is used to locate the start of data-uri definitions then + * manual looping over content is done to parse the whole data uri. + * Attempting to capture the whole data uri using regex can cause PHP + * PCRE limits to be hit with larger, multi-MB, files. + */ + protected function extractBase64ImagesFromMarkdown(string $markdown) + { + $matches = []; + $contentLength = strlen($markdown); + $replacements = []; + preg_match_all('/!\[.*?]\(.*?(data:image\/.{1,6};base64,)/', $markdown, $matches, PREG_OFFSET_CAPTURE); + + foreach ($matches[1] as $base64MatchPair) { + [$dataUri, $index] = $base64MatchPair; + + for ($i = strlen($dataUri) + $index; $i < $contentLength; $i++) { + $char = $markdown[$i]; + if ($char === ')' || $char === ' ' || $char === "\n" || $char === '"') { + break; + } + $dataUri .= $char; + } + + $newUrl = $this->base64ImageUriToUploadedImageUrl($dataUri); + $replacements[] = [$dataUri, $newUrl]; + } + + foreach ($replacements as [$dataUri, $newUrl]) { + $markdown = str_replace($dataUri, $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: string, 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 +192,15 @@ 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) { @@ -160,6 +216,9 @@ class PageContent $html .= $doc->saveHTML($childNode); } + // Perform required string-level tweaks + $html = str_replace(' ', ' ', $html); + return $html; } @@ -180,9 +239,9 @@ class PageContent * A map for existing ID's should be passed in to check for current existence. * Returns a pair of strings in the format [old_id, new_id]. */ - protected function setUniqueId(\DOMNode $element, array &$idMap): array + protected function setUniqueId(DOMNode $element, array &$idMap): array { - if (get_class($element) !== 'DOMElement') { + if (!$element instanceof DOMElement) { return ['', '']; } @@ -194,7 +253,7 @@ class PageContent return [$existingId, $existingId]; } - // Create an unique id for the element + // Create a unique id for the element // Uses the content as a basis to ensure output is the same every time // the same content is passed through. $contentId = 'bkmrk-' . mb_substr(strtolower(preg_replace('/\s+/', '-', trim($element->nodeValue))), 0, 20); @@ -264,7 +323,7 @@ class PageContent */ protected function headerNodesToLevelList(DOMNodeList $nodeList): array { - $tree = collect($nodeList)->map(function ($header) { + $tree = collect($nodeList)->map(function (DOMElement $header) { $text = trim(str_replace("\xc2\xa0", '', $header->nodeValue)); $text = mb_substr($text, 0, 100); @@ -315,23 +374,30 @@ class PageContent continue; } - // Find page and skip this if page not found + // Find page to use, and default replacement to empty string for non-matches. /** @var ?Page $matchedPage */ $matchedPage = Page::visible()->find($pageId); - if ($matchedPage === null) { - $html = str_replace($fullMatch, '', $html); - continue; + $replacement = ''; + + if ($matchedPage && count($splitInclude) === 1) { + // If we only have page id, just insert all page html and continue. + $replacement = $matchedPage->html; + } elseif ($matchedPage && count($splitInclude) > 1) { + // Otherwise, if our include tag defines a section, load that specific content + $innerContent = $this->fetchSectionOfPage($matchedPage, $splitInclude[1]); + $replacement = trim($innerContent); } - // If we only have page id, just insert all page html and continue. - if (count($splitInclude) === 1) { - $html = str_replace($fullMatch, $matchedPage->html, $html); - continue; - } + $themeReplacement = Theme::dispatch( + ThemeEvents::PAGE_INCLUDE_PARSE, + $includeId, + $replacement, + clone $this->page, + $matchedPage ? (clone $matchedPage) : null, + ); - // Create and load HTML into a document - $innerContent = $this->fetchSectionOfPage($matchedPage, $splitInclude[1]); - $html = str_replace($fullMatch, trim($innerContent), $html); + // Perform the content replacement + $html = str_replace($fullMatch, $themeReplacement ?? $replacement, $html); } return $html; @@ -342,7 +408,7 @@ 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.