From 75936454cca139d0b226a95ee7a0070bc8702fdc Mon Sep 17 00:00:00 2001
From: Dan Brown <ssddanbrown@googlemail.com>
Date: Thu, 23 Nov 2023 14:29:07 +0000
Subject: [PATCH] Includes: Developed to get new system working with inline
includes
Adds logic for locating and splitting text nodes.
Adds specific classes to offload tag/content specific logic.
---
app/Entities/Tools/PageIncludeContent.php | 68 ++++++++++++++++++
app/Entities/Tools/PageIncludeParser.php | 85 +++++++++++++++++++++--
app/Entities/Tools/PageIncludeTag.php | 30 ++++++++
app/Util/HtmlDocument.php | 15 ++++
tests/Unit/PageIncludeParserTest.php | 2 +-
5 files changed, 192 insertions(+), 8 deletions(-)
create mode 100644 app/Entities/Tools/PageIncludeContent.php
create mode 100644 app/Entities/Tools/PageIncludeTag.php
diff --git a/app/Entities/Tools/PageIncludeContent.php b/app/Entities/Tools/PageIncludeContent.php
new file mode 100644
index 000000000..97c470c68
--- /dev/null
+++ b/app/Entities/Tools/PageIncludeContent.php
@@ -0,0 +1,68 @@
+<?php
+
+namespace BookStack\Entities\Tools;
+
+use BookStack\Util\HtmlDocument;
+use DOMNode;
+
+class PageIncludeContent
+{
+ protected static array $topLevelTags = ['table', 'ul', 'ol', 'pre'];
+
+ /**
+ * @var DOMNode[]
+ */
+ protected array $contents = [];
+
+ protected bool $isTopLevel;
+
+ public function __construct(
+ string $html,
+ PageIncludeTag $tag,
+ ) {
+ $this->parseHtml($html, $tag);
+ }
+
+ protected function parseHtml(string $html, PageIncludeTag $tag): void
+ {
+ if (empty($html)) {
+ return;
+ }
+
+ $doc = new HtmlDocument($html);
+
+ $sectionId = $tag->getSectionId();
+ if (!$sectionId) {
+ $this->contents = [...$doc->getBodyChildren()];
+ $this->isTopLevel = true;
+ return;
+ }
+
+ $section = $doc->getElementById($sectionId);
+ if (!$section) {
+ return;
+ }
+
+ $isTopLevel = in_array(strtolower($section->nodeName), static::$topLevelTags);
+ $this->isTopLevel = $isTopLevel;
+ $this->contents = $isTopLevel ? [$section] : [...$section->childNodes];
+ }
+
+ public function isInline(): bool
+ {
+ return !$this->isTopLevel;
+ }
+
+ public function isEmpty(): bool
+ {
+ return empty($this->contents);
+ }
+
+ /**
+ * @return DOMNode[]
+ */
+ public function toDomNodes(): array
+ {
+ return $this->contents;
+ }
+}
diff --git a/app/Entities/Tools/PageIncludeParser.php b/app/Entities/Tools/PageIncludeParser.php
index 63d3ea8d6..070b0cc11 100644
--- a/app/Entities/Tools/PageIncludeParser.php
+++ b/app/Entities/Tools/PageIncludeParser.php
@@ -4,6 +4,8 @@ namespace BookStack\Entities\Tools;
use BookStack\Util\HtmlDocument;
use Closure;
+use DOMNode;
+use DOMText;
class PageIncludeParser
{
@@ -17,14 +19,25 @@ class PageIncludeParser
public function parse(): string
{
- $html = new HtmlDocument($this->pageHtml);
+ $doc = new HtmlDocument($this->pageHtml);
- $includeHosts = $html->queryXPath("//body//*[contains(text(), '{{@')]");
- $node = $includeHosts->item(0);
+ $tags = $this->locateAndIsolateIncludeTags($doc);
- // One of the direct child textnodes of the "$includeHosts" should be
- // the one with the include tag within.
- $textNode = $node->childNodes->item(0);
+ foreach ($tags as $tag) {
+ $htmlContent = $this->pageContentForId->call($this, $tag->getPageId());
+ $content = new PageIncludeContent($htmlContent, $tag);
+
+ if ($content->isInline()) {
+ $adopted = $doc->adoptNodes($content->toDomNodes());
+ foreach ($adopted as $adoptedContentNode) {
+ $tag->domNode->parentNode->insertBefore($adoptedContentNode, $tag->domNode);
+ }
+ $tag->domNode->parentNode->removeChild($tag->domNode);
+ continue;
+ }
+
+ // TODO - Non-inline
+ }
// TODO:
// Hunt down the specific text nodes with matches
@@ -52,6 +65,64 @@ class PageIncludeParser
// in changes affecting the next tag, where tags may be in the same/adjacent nodes.
- return $html->getBodyInnerHtml();
+ return $doc->getBodyInnerHtml();
+ }
+
+ /**
+ * Locate include tags within the given document, isolating them to their
+ * own nodes in the DOM for future targeted manipulation.
+ * @return PageIncludeTag[]
+ */
+ protected function locateAndIsolateIncludeTags(HtmlDocument $doc): array
+ {
+ $includeHosts = $doc->queryXPath("//body//*[contains(text(), '{{@')]");
+ $includeTags = [];
+
+ /** @var DOMNode $node */
+ /** @var DOMNode $childNode */
+ foreach ($includeHosts as $node) {
+ foreach ($node->childNodes as $childNode) {
+ if ($childNode->nodeName === '#text') {
+ array_push($includeTags, ...$this->splitTextNodesAtTags($childNode));
+ }
+ }
+ }
+
+ return $includeTags;
+ }
+
+ /**
+ * Takes a text DOMNode and splits its text content at include tags
+ * into multiple text nodes within the original parent.
+ * Returns found PageIncludeTag references.
+ * @return PageIncludeTag[]
+ */
+ protected function splitTextNodesAtTags(DOMNode $textNode): array
+ {
+ $includeTags = [];
+ $text = $textNode->textContent;
+ preg_match_all(static::$includeTagRegex, $text, $matches, PREG_OFFSET_CAPTURE);
+
+ $currentOffset = 0;
+ foreach ($matches[0] as $index => $fullTagMatch) {
+ $tagOuterContent = $fullTagMatch[0];
+ $tagInnerContent = $matches[1][$index][0];
+ $tagStartOffset = $fullTagMatch[1];
+
+ if ($currentOffset < $tagStartOffset) {
+ $previousText = substr($text, $currentOffset, $tagStartOffset - $currentOffset);
+ $textNode->parentNode->insertBefore(new DOMText($previousText), $textNode);
+ }
+
+ $node = $textNode->parentNode->insertBefore(new DOMText($tagOuterContent), $textNode);
+ $includeTags[] = new PageIncludeTag($tagInnerContent, $node);
+ $currentOffset = $tagStartOffset + strlen($tagOuterContent);
+ }
+
+ if ($currentOffset > 0) {
+ $textNode->textContent = substr($text, $currentOffset);
+ }
+
+ return $includeTags;
}
}
diff --git a/app/Entities/Tools/PageIncludeTag.php b/app/Entities/Tools/PageIncludeTag.php
new file mode 100644
index 000000000..05a532fb2
--- /dev/null
+++ b/app/Entities/Tools/PageIncludeTag.php
@@ -0,0 +1,30 @@
+<?php
+
+namespace BookStack\Entities\Tools;
+
+use DOMNode;
+
+class PageIncludeTag
+{
+ public function __construct(
+ public string $tagContent,
+ public DOMNode $domNode,
+ ) {
+ }
+
+ /**
+ * Get the page ID that this tag references.
+ */
+ public function getPageId(): int
+ {
+ return intval(trim(explode('#', $this->tagContent, 2)[0]));
+ }
+
+ /**
+ * Get the section ID that this tag references (if any)
+ */
+ public function getSectionId(): string
+ {
+ return trim(explode('#', $this->tagContent, 2)[1] ?? '');
+ }
+}
diff --git a/app/Util/HtmlDocument.php b/app/Util/HtmlDocument.php
index b8c53d439..ad5dacd82 100644
--- a/app/Util/HtmlDocument.php
+++ b/app/Util/HtmlDocument.php
@@ -149,4 +149,19 @@ class HtmlDocument
{
return $this->document->saveHTML($node);
}
+
+ /**
+ * Adopt the given nodes into this document.
+ * @param DOMNode[] $nodes
+ * @return DOMNode[]
+ */
+ public function adoptNodes(array $nodes): array
+ {
+ $adopted = [];
+ foreach ($nodes as $node) {
+ $adopted[] = $this->document->importNode($node, true);
+ }
+
+ return $adopted;
+ }
}
diff --git a/tests/Unit/PageIncludeParserTest.php b/tests/Unit/PageIncludeParserTest.php
index de31504ff..d1912270e 100644
--- a/tests/Unit/PageIncludeParserTest.php
+++ b/tests/Unit/PageIncludeParserTest.php
@@ -37,7 +37,7 @@ class PageIncludeParserTest extends TestCase
protected function runParserTest(string $html, array $contentById, string $expected)
{
$parser = new PageIncludeParser($html, function (int $id) use ($contentById) {
- return $contentById[strval($id)] ?? null;
+ return $contentById[strval($id)] ?? '';
});
$result = $parser->parse();