From c88eb729a499197e8b3ab9d5019b4426a65d3d41 Mon Sep 17 00:00:00 2001
From: Dan Brown <ssddanbrown@googlemail.com>
Date: Fri, 24 Nov 2023 23:39:16 +0000
Subject: [PATCH] Includes: Added block-level handling to new include system
Implements block promoting to body (including position choosing based
upon likely tag position within parent) and block splitting where we're
only a single depth down from the body child.
---
app/Entities/Tools/PageIncludeParser.php | 111 ++++++++++++++++-------
app/Util/HtmlDocument.php | 15 ---
tests/Unit/PageIncludeParserTest.php | 96 +++++++++++++++++++-
3 files changed, 172 insertions(+), 50 deletions(-)
diff --git a/app/Entities/Tools/PageIncludeParser.php b/app/Entities/Tools/PageIncludeParser.php
index 070b0cc11..5ce847d6c 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 DOMDocument;
+use DOMElement;
use DOMNode;
use DOMText;
@@ -22,48 +24,26 @@ class PageIncludeParser
$doc = new HtmlDocument($this->pageHtml);
$tags = $this->locateAndIsolateIncludeTags($doc);
+ $topLevel = [...$doc->getBodyChildren()];
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, $tag->domNode);
+ if (!$content->isInline()) {
+ $isParentTopLevel = in_array($tag->domNode->parentNode, $topLevel, true);
+ if ($isParentTopLevel) {
+ $this->splitNodeAtChildNode($tag->domNode->parentNode, $tag->domNode);
+ } else {
+ $this->promoteTagNodeToBody($tag, $doc->getBody());
}
- $tag->domNode->parentNode->removeChild($tag->domNode);
- continue;
}
- // TODO - Non-inline
+ $this->replaceNodeWithNodes($tag->domNode, $content->toDomNodes());
}
- // TODO:
- // Hunt down the specific text nodes with matches
- // Split out tag text node from rest of content
- // Fetch tag content->
- // If range or top-block: delete tag text node, [Promote to top-block], delete old top-block if empty
- // If inline: Replace current text node with new text or elem
- // !! "Range" or "inline" status should come from tag parser and content fetcher, not guessed direct from content
- // since we could have a range of inline elements
-
- // [Promote to top-block]
- // Tricky operation.
- // Can throw in before or after current top-block depending on relative position
- // Could [Split] top-block but complex past a single level depth.
- // Maybe [Split] if one level depth, otherwise default to before/after block
- // Should work for the vast majority of cases, and not for those which would
- // technically be invalid in-editor anyway.
-
- // [Split]
- // Copy original top-block node type and attrs (apart from ID)
- // Move nodes after promoted tag-node into copy
- // Insert copy after original (after promoted top-block eventually)
-
- // Notes: May want to eventually parse through backwards, which should avoid issues
- // in changes affecting the next tag, where tags may be in the same/adjacent nodes.
-
+ // TODO Notes: May want to eventually parse through backwards, which should avoid issues
+ // in changes affecting the next tag, where tags may be in the same/adjacent nodes.
return $doc->getBodyInnerHtml();
}
@@ -125,4 +105,71 @@ class PageIncludeParser
return $includeTags;
}
+
+ /**
+ * @param DOMNode[] $replacements
+ */
+ protected function replaceNodeWithNodes(DOMNode $toReplace, array $replacements): void
+ {
+ /** @var DOMDocument $targetDoc */
+ $targetDoc = $toReplace->ownerDocument;
+
+ foreach ($replacements as $replacement) {
+ if ($replacement->ownerDocument !== $targetDoc) {
+ $replacement = $targetDoc->adoptNode($replacement);
+ }
+
+ $toReplace->parentNode->insertBefore($replacement, $toReplace);
+ }
+
+ $toReplace->parentNode->removeChild($toReplace);
+ }
+
+ protected function promoteTagNodeToBody(PageIncludeTag $tag, DOMNode $body): void
+ {
+ /** @var DOMNode $topParent */
+ $topParent = $tag->domNode->parentNode;
+ while ($topParent->parentNode !== $body) {
+ $topParent = $topParent->parentNode;
+ }
+
+ $parentText = $topParent->textContent;
+ $tagPos = strpos($parentText, $tag->tagContent);
+ $before = $tagPos < (strlen($parentText) / 2);
+
+ if ($before) {
+ $body->insertBefore($tag->domNode, $topParent);
+ } else {
+ $body->insertBefore($tag->domNode, $topParent->nextSibling);
+ }
+ }
+
+ protected function splitNodeAtChildNode(DOMElement $parentNode, DOMNode $domNode): void
+ {
+ $children = [...$parentNode->childNodes];
+ $splitPos = array_search($domNode, $children, true) ?: count($children);
+ $parentClone = $parentNode->cloneNode();
+ $parentClone->removeAttribute('id');
+
+ /** @var DOMNode $child */
+ for ($i = 0; $i < $splitPos; $i++) {
+ $child = $children[0];
+ $parentClone->appendChild($child);
+ }
+
+ if ($parentClone->hasChildNodes()) {
+ $parentNode->parentNode->insertBefore($parentClone, $parentNode);
+ }
+
+ $parentNode->parentNode->insertBefore($domNode, $parentNode);
+
+ $parentClone->normalize();
+ $parentNode->normalize();
+ if (!$parentNode->hasChildNodes()) {
+ $parentNode->remove();
+ }
+ if (!$parentClone->hasChildNodes()) {
+ $parentClone->remove();
+ }
+ }
}
diff --git a/app/Util/HtmlDocument.php b/app/Util/HtmlDocument.php
index ad5dacd82..b8c53d439 100644
--- a/app/Util/HtmlDocument.php
+++ b/app/Util/HtmlDocument.php
@@ -149,19 +149,4 @@ 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 d1912270e..e0bd69e93 100644
--- a/tests/Unit/PageIncludeParserTest.php
+++ b/tests/Unit/PageIncludeParserTest.php
@@ -7,7 +7,7 @@ use Tests\TestCase;
class PageIncludeParserTest extends TestCase
{
- public function test_include_simple_inline_text()
+ public function test_simple_inline_text()
{
$this->runParserTest(
'<p>{{@45#content}}</p>',
@@ -16,7 +16,7 @@ class PageIncludeParserTest extends TestCase
);
}
- public function test_include_simple_inline_text_with_existing_siblings()
+ public function test_simple_inline_text_with_existing_siblings()
{
$this->runParserTest(
'<p>{{@45#content}} <strong>Hi</strong>there!</p>',
@@ -25,7 +25,7 @@ class PageIncludeParserTest extends TestCase
);
}
- public function test_include_simple_inline_text_within_other_text()
+ public function test_simple_inline_text_within_other_text()
{
$this->runParserTest(
'<p>Hello {{@45#content}}there!</p>',
@@ -34,6 +34,96 @@ class PageIncludeParserTest extends TestCase
);
}
+ public function test_block_content_types()
+ {
+ $inputs = [
+ '<table id="content"><td>Text</td></table>',
+ '<ul id="content"><li>Item A</li></ul>',
+ '<ol id="content"><li>Item A</li></ol>',
+ '<pre id="content">Code</pre>',
+ ];
+
+ foreach ($inputs as $input) {
+ $this->runParserTest(
+ '<p>A{{@45#content}}B</p>',
+ ['45' => $input],
+ '<p>A</p>' . $input . '<p>B</p>',
+ );
+ }
+ }
+
+ public function test_block_content_nested_origin_gets_placed_before()
+ {
+ $this->runParserTest(
+ '<p><strong>A {{@45#content}} there!</strong></p>',
+ ['45' => '<pre id="content">Testing</pre>'],
+ '<pre id="content">Testing</pre><p><strong>A there!</strong></p>',
+ );
+ }
+
+ public function test_block_content_nested_origin_gets_placed_after()
+ {
+ $this->runParserTest(
+ '<p><strong>Some really good {{@45#content}} there!</strong></p>',
+ ['45' => '<pre id="content">Testing</pre>'],
+ '<p><strong>Some really good there!</strong></p><pre id="content">Testing</pre>',
+ );
+ }
+
+ public function test_block_content_in_shallow_origin_gets_split()
+ {
+ $this->runParserTest(
+ '<p>Some really good {{@45#content}} there!</p>',
+ ['45' => '<pre id="content">doggos</pre>'],
+ '<p>Some really good </p><pre id="content">doggos</pre><p> there!</p>',
+ );
+ }
+
+ public function test_block_content_in_shallow_origin_split_does_not_duplicate_id()
+ {
+ $this->runParserTest(
+ '<p id="test" title="Hi">Some really good {{@45#content}} there!</p>',
+ ['45' => '<pre id="content">doggos</pre>'],
+ '<p title="Hi">Some really good </p><pre id="content">doggos</pre><p id="test" title="Hi"> there!</p>',
+ );
+ }
+
+ public function test_block_content_in_shallow_origin_does_not_leave_empty_nodes()
+ {
+ $this->runParserTest(
+ '<p>{{@45#content}}</p>',
+ ['45' => '<pre id="content">doggos</pre>'],
+ '<pre id="content">doggos</pre>',
+ );
+ }
+
+ public function test_simple_whole_document()
+ {
+ $this->runParserTest(
+ '<p>{{@45}}</p>',
+ ['45' => '<p id="content">Testing</p>'],
+ '<p id="content">Testing</p>',
+ );
+ }
+
+ public function test_multi_source_elem_whole_document()
+ {
+ $this->runParserTest(
+ '<p>{{@45}}</p>',
+ ['45' => '<p>Testing</p><blockquote>This</blockquote>'],
+ '<p>Testing</p><blockquote>This</blockquote>',
+ );
+ }
+
+ public function test_multi_source_elem_whole_document_with_shared_content_origin()
+ {
+ $this->runParserTest(
+ '<p>This is {{@45}} some text</p>',
+ ['45' => '<p>Testing</p><blockquote>This</blockquote>'],
+ '<p>This is </p><p>Testing</p><blockquote>This</blockquote><p> some text</p>',
+ );
+ }
+
protected function runParserTest(string $html, array $contentById, string $expected)
{
$parser = new PageIncludeParser($html, function (int $id) use ($contentById) {