X-Git-Url: http://source.bookstackapp.com/bookstack/blobdiff_plain/5c5a3de7cb9ebdb1644726ec167f68384594d093..refs/pull/3918/head:/app/Entities/Tools/PageContent.php

diff --git a/app/Entities/Tools/PageContent.php b/app/Entities/Tools/PageContent.php
index 9e2b6a7b6..9aa2e8d35 100644
--- a/app/Entities/Tools/PageContent.php
+++ b/app/Entities/Tools/PageContent.php
@@ -3,8 +3,7 @@
 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;
@@ -17,15 +16,10 @@ 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.
@@ -53,34 +47,17 @@ class PageContent
     {
         $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.
      */
     protected function extractBase64ImagesFromHtml(string $htmlText): string
     {
-        if (empty($htmlText) || mb_strpos($htmlText, 'data:image') === false) {
+        if (empty($htmlText) || strpos($htmlText, 'data:image') === false) {
             return $htmlText;
         }
 
@@ -109,15 +86,35 @@ class PageContent
 
     /**
      * 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 = [];
-        preg_match_all('/!\[.*?]\(.*?(data:image\/.*?)[)"\s]/', $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 ($matches[1] as $base64Match) {
-            $newUrl = $this->base64ImageUriToUploadedImageUrl($base64Match);
-            $markdown = str_replace($base64Match, $newUrl, $markdown);
+        foreach ($replacements as [$dataUri, $newUrl]) {
+            $markdown = str_replace($dataUri, $newUrl, $markdown);
         }
 
         return $markdown;
@@ -219,6 +216,9 @@ class PageContent
             $html .= $doc->saveHTML($childNode);
         }
 
+        // Perform required string-level tweaks
+        $html = str_replace(' ', '&nbsp;', $html);
+
         return $html;
     }
 
@@ -374,23 +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;