Merge pull request #4688 from BookStackApp/include-parser

New include tag parser

app/Entities/Tools/PageContent.php

@@ -11,6 +11,7 @@
 use BookStack\Uploads\ImageService;
 use BookStack\Util\HtmlContentFilter;
 use BookStack\Util\HtmlDocument;
+use Closure;
 use DOMElement;
 use DOMNode;
 use DOMNodeList;
@@ -275,21 +276,65 @@ protected function toPlainText(): string
      */
     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;
+
+        return function (PageIncludeTag $tag) use ($blankIncludes, $contextPage): PageIncludeContent {
+            if ($blankIncludes) {
+                return PageIncludeContent::fromHtmlAndTag('', $tag);
+            }
+
+            $matchedPage = Page::visible()->find($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;
+            return $content;
+        };
     }
 
     /**
@@ -337,83 +382,4 @@ protected function headerNodesToLevelList(DOMNodeList $nodeList): array
 
         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 "{{@<page_id>#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);
-    }
 }

app/Entities/Tools/PageIncludeContent.php

@@ -0,0 +1,85 @@
+<?php
+
+namespace BookStack\Entities\Tools;
+
+use BookStack\Util\HtmlDocument;
+use DOMNode;
+
+class PageIncludeContent
+{
+    protected static array $topLevelTags = ['table', 'ul', 'ol', 'pre'];
+
+    /**
+     * @param DOMNode[] $contents
+     * @param bool $isInline
+     */
+    public function __construct(
+        protected array $contents,
+        protected bool $isInline,
+    ) {
+    }
+
+    public static function fromHtmlAndTag(string $html, PageIncludeTag $tag): self
+    {
+        if (empty($html)) {
+            return new self([], true);
+        }
+
+        $doc = new HtmlDocument($html);
+
+        $sectionId = $tag->getSectionId();
+        if (!$sectionId) {
+            $contents = [...$doc->getBodyChildren()];
+            return new self($contents, false);
+        }
+
+        $section = $doc->getElementById($sectionId);
+        if (!$section) {
+            return new self([], true);
+        }
+
+        $isTopLevel = in_array(strtolower($section->nodeName), static::$topLevelTags);
+        $contents = $isTopLevel ? [$section] : [...$section->childNodes];
+        return new self($contents, !$isTopLevel);
+    }
+
+    public static function fromInlineHtml(string $html): self
+    {
+        if (empty($html)) {
+            return new self([], true);
+        }
+
+        $doc = new HtmlDocument($html);
+
+        return new self([...$doc->getBodyChildren()], true);
+    }
+
+    public function isInline(): bool
+    {
+        return $this->isInline;
+    }
+
+    public function isEmpty(): bool
+    {
+        return empty($this->contents);
+    }
+
+    /**
+     * @return DOMNode[]
+     */
+    public function toDomNodes(): array
+    {
+        return $this->contents;
+    }
+
+    public function toHtml(): string
+    {
+        $html = '';
+
+        foreach ($this->contents as $content) {
+            $html .= $content->ownerDocument->saveHTML($content);
+        }
+
+        return $html;
+    }
+}