X-Git-Url: http://source.bookstackapp.com/bookstack/blobdiff_plain/db7b11fe933b6f9b9962aec18f45a74b56458db5..refs/pull/5721/head:/app/Entities/Tools/PageContent.php diff --git a/app/Entities/Tools/PageContent.php b/app/Entities/Tools/PageContent.php index 3e75bd5bb..d2f5de65c 100644 --- a/app/Entities/Tools/PageContent.php +++ b/app/Entities/Tools/PageContent.php @@ -3,14 +3,18 @@ namespace BookStack\Entities\Tools; use BookStack\Entities\Models\Page; +use BookStack\Entities\Queries\PageQueries; 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\Users\Models\User; use BookStack\Util\HtmlContentFilter; use BookStack\Util\HtmlDocument; +use BookStack\Util\WebSafeMimeSniffer; +use Closure; use DOMElement; use DOMNode; use DOMNodeList; @@ -18,17 +22,20 @@ use Illuminate\Support\Str; class PageContent { + protected PageQueries $pageQueries; + public function __construct( protected Page $page ) { + $this->pageQueries = app()->make(PageQueries::class); } /** * Update the content of the page with new provided HTML. */ - public function setNewHTML(string $html): void + public function setNewHTML(string $html, User $updater): void { - $html = $this->extractBase64ImagesFromHtml($html); + $html = $this->extractBase64ImagesFromHtml($html, $updater); $this->page->html = $this->formatHtml($html); $this->page->text = $this->toPlainText(); $this->page->markdown = ''; @@ -37,9 +44,9 @@ class PageContent /** * Update the content of the page with new provided Markdown content. */ - public function setNewMarkdown(string $markdown): void + public function setNewMarkdown(string $markdown, User $updater): void { - $markdown = $this->extractBase64ImagesFromMarkdown($markdown); + $markdown = $this->extractBase64ImagesFromMarkdown($markdown, $updater); $this->page->markdown = $markdown; $html = (new MarkdownToHtml($markdown))->convert(); $this->page->html = $this->formatHtml($html); @@ -49,7 +56,7 @@ class PageContent /** * Convert all base64 image data to saved images. */ - protected function extractBase64ImagesFromHtml(string $htmlText): string + protected function extractBase64ImagesFromHtml(string $htmlText, User $updater): string { if (empty($htmlText) || !str_contains($htmlText, 'data:image')) { return $htmlText; @@ -59,9 +66,10 @@ class PageContent // Get all img elements with image data blobs $imageNodes = $doc->queryXPath('//img[contains(@src, \'data:image\')]'); + /** @var DOMElement $imageNode */ foreach ($imageNodes as $imageNode) { $imageSrc = $imageNode->getAttribute('src'); - $newUrl = $this->base64ImageUriToUploadedImageUrl($imageSrc); + $newUrl = $this->base64ImageUriToUploadedImageUrl($imageSrc, $updater); $imageNode->setAttribute('src', $newUrl); } @@ -75,7 +83,7 @@ class PageContent * 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): string + protected function extractBase64ImagesFromMarkdown(string $markdown, User $updater): string { $matches = []; $contentLength = strlen($markdown); @@ -93,7 +101,7 @@ class PageContent $dataUri .= $char; } - $newUrl = $this->base64ImageUriToUploadedImageUrl($dataUri); + $newUrl = $this->base64ImageUriToUploadedImageUrl($dataUri, $updater); $replacements[] = [$dataUri, $newUrl]; } @@ -108,16 +116,28 @@ class PageContent * 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 + protected function base64ImageUriToUploadedImageUrl(string $uri, User $updater): string { $imageRepo = app()->make(ImageRepo::class); $imageInfo = $this->parseBase64ImageUri($uri); + // Validate user has permission to create images + if (!$updater->can('image-create-all')) { + return ''; + } + // Validate extension and content if (empty($imageInfo['data']) || !ImageService::isExtensionSupported($imageInfo['extension'])) { return ''; } + // Validate content looks like an image via sniffing mime type + $mimeSniffer = new WebSafeMimeSniffer(); + $mime = $mimeSniffer->sniff($imageInfo['data']); + if (!str_starts_with($mime, 'image/')) { + return ''; + } + // Validate that the content is not over our upload limit $uploadLimitBytes = (config('app.upload_limit') * 1000000); if (strlen($imageInfo['data']) > $uploadLimitBytes) { @@ -275,21 +295,66 @@ class PageContent */ public function render(bool $blankIncludes = false): string { - $content = $this->page->html ?? ''; + $html = $this->page->html ?? ''; + + if (empty($html)) { + return $html; + } + + $doc = new HtmlDocument($html); + $contentProvider = $this->getContentProviderClosure($blankIncludes); + $parser = new PageIncludeParser($doc, $contentProvider); + + $nodesAdded = 1; + for ($includeDepth = 0; $includeDepth < 3 && $nodesAdded !== 0; $includeDepth++) { + $nodesAdded = $parser->parse(); + } + + if ($includeDepth > 1) { + $idMap = []; + $changeMap = []; + $this->updateIdsRecursively($doc->getBody(), 0, $idMap, $changeMap); + } if (!config('app.allow_content_scripts')) { - $content = HtmlContentFilter::removeScripts($content); + HtmlContentFilter::removeScriptsFromDocument($doc); } - if ($blankIncludes) { - $content = $this->blankPageIncludes($content); - } else { - for ($includeDepth = 0; $includeDepth < 3; $includeDepth++) { - $content = $this->parsePageIncludes($content); + return $doc->getBodyInnerHtml(); + } + + /** + * Get the closure used to fetch content for page includes. + */ + protected function getContentProviderClosure(bool $blankIncludes): Closure + { + $contextPage = $this->page; + $queries = $this->pageQueries; + + return function (PageIncludeTag $tag) use ($blankIncludes, $contextPage, $queries): PageIncludeContent { + if ($blankIncludes) { + return PageIncludeContent::fromHtmlAndTag('', $tag); } - } - return $content; + $matchedPage = $queries->findVisibleById($tag->getPageId()); + $content = PageIncludeContent::fromHtmlAndTag($matchedPage->html ?? '', $tag); + + if (Theme::hasListeners(ThemeEvents::PAGE_INCLUDE_PARSE)) { + $themeReplacement = Theme::dispatch( + ThemeEvents::PAGE_INCLUDE_PARSE, + $tag->tagContent, + $content->toHtml(), + clone $contextPage, + $matchedPage ? (clone $matchedPage) : null, + ); + + if ($themeReplacement !== null) { + $content = PageIncludeContent::fromInlineHtml(strval($themeReplacement)); + } + } + + return $content; + }; } /** @@ -314,7 +379,7 @@ class PageContent protected function headerNodesToLevelList(DOMNodeList $nodeList): array { $tree = collect($nodeList)->map(function (DOMElement $header) { - $text = trim(str_replace("\xc2\xa0", '', $header->nodeValue)); + $text = trim(str_replace("\xc2\xa0", ' ', $header->nodeValue)); $text = mb_substr($text, 0, 100); return [ @@ -337,83 +402,4 @@ class PageContent return $tree->toArray(); } - - /** - * Remove any page include tags within the given HTML. - */ - protected function blankPageIncludes(string $html): string - { - return preg_replace("/{{@\s?([0-9].*?)}}/", '', $html); - } - - /** - * Parse any include tags "{{@#section}}" to be part of the page. - */ - protected function parsePageIncludes(string $html): string - { - $matches = []; - preg_match_all("/{{@\s?([0-9].*?)}}/", $html, $matches); - - foreach ($matches[1] as $index => $includeId) { - $fullMatch = $matches[0][$index]; - $splitInclude = explode('#', $includeId, 2); - - // Get page id from reference - $pageId = intval($splitInclude[0]); - if (is_nan($pageId)) { - continue; - } - - // Find page to use, and default replacement to empty string for non-matches. - /** @var ?Page $matchedPage */ - $matchedPage = Page::visible()->find($pageId); - $replacement = ''; - - if ($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); - } - - $themeReplacement = Theme::dispatch( - ThemeEvents::PAGE_INCLUDE_PARSE, - $includeId, - $replacement, - clone $this->page, - $matchedPage ? (clone $matchedPage) : null, - ); - - // Perform the content replacement - $html = str_replace($fullMatch, $themeReplacement ?? $replacement, $html); - } - - return $html; - } - - /** - * Fetch the content from a specific section of the given page. - */ - protected function fetchSectionOfPage(Page $page, string $sectionId): string - { - $topLevelTags = ['table', 'ul', 'ol', 'pre']; - $doc = new HtmlDocument($page->html); - - // Search included content for the id given and blank out if not exists. - $matchingElem = $doc->getElementById($sectionId); - if ($matchingElem === null) { - return ''; - } - - // Otherwise replace the content with the found content - // Checks if the top-level wrapper should be included by matching on tag types - $isTopLevel = in_array(strtolower($matchingElem->nodeName), $topLevelTags); - if ($isTopLevel) { - return $doc->getNodeOuterHtml($matchingElem); - } - - return $doc->getNodeInnerHtml($matchingElem); - } }