Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
Total | |
94.86% |
277 / 292 |
|
81.48% |
22 / 27 |
CRAP | |
0.00% |
0 / 1 |
CommentUtils | |
94.86% |
277 / 292 |
|
81.48% |
22 / 27 |
163.47 | |
0.00% |
0 / 1 |
__construct | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
isBlockElement | |
100.00% |
2 / 2 |
|
100.00% |
1 / 1 |
2 | |||
isRenderingTransparentNode | |
100.00% |
22 / 22 |
|
100.00% |
1 / 1 |
11 | |||
isOurGeneratedNode | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
4 | |||
cantHaveElementChildren | |
100.00% |
18 / 18 |
|
100.00% |
1 / 1 |
7 | |||
isCommentSeparator | |
100.00% |
20 / 20 |
|
100.00% |
1 / 1 |
15 | |||
isCommentContent | |
100.00% |
7 / 7 |
|
100.00% |
1 / 1 |
3 | |||
childIndexOf | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
2 | |||
contains | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
3 | |||
closestElement | |
100.00% |
6 / 6 |
|
100.00% |
1 / 1 |
3 | |||
closestElementWithSibling | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
72 | |||
getTranscludedFromElement | |
100.00% |
14 / 14 |
|
100.00% |
1 / 1 |
10 | |||
getHeadlineNodeAndOffset | |
76.92% |
10 / 13 |
|
0.00% |
0 / 1 |
5.31 | |||
htmlTrim | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
getIndentLevel | |
100.00% |
9 / 9 |
|
100.00% |
1 / 1 |
6 | |||
getCoveredSiblings | |
100.00% |
13 / 13 |
|
100.00% |
1 / 1 |
5 | |||
getFullyCoveredSiblings | |
100.00% |
17 / 17 |
|
100.00% |
1 / 1 |
5 | |||
unwrapParsoidSections | |
100.00% |
6 / 6 |
|
100.00% |
1 / 1 |
3 | |||
getTitleFromUrl | |
100.00% |
22 / 22 |
|
100.00% |
1 / 1 |
6 | |||
linearWalk | |
100.00% |
11 / 11 |
|
100.00% |
1 / 1 |
5 | |||
linearWalkBackwards | |
100.00% |
11 / 11 |
|
100.00% |
1 / 1 |
5 | |||
getRangeFirstNode | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
3 | |||
getRangeLastNode | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
3 | |||
compareRanges | |
88.00% |
22 / 25 |
|
0.00% |
0 / 1 |
23.91 | |||
compareRangesAlmostEqualBoundaries | |
100.00% |
28 / 28 |
|
100.00% |
1 / 1 |
15 | |||
isSingleCommentSignedBy | |
100.00% |
17 / 17 |
|
100.00% |
1 / 1 |
5 | |||
getNewTopicsSubscriptionId | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 |
1 | <?php |
2 | |
3 | namespace MediaWiki\Extension\DiscussionTools; |
4 | |
5 | use LogicException; |
6 | use MediaWiki\Config\Config; |
7 | use MediaWiki\Extension\DiscussionTools\ThreadItem\ContentCommentItem; |
8 | use MediaWiki\Extension\DiscussionTools\ThreadItem\ContentThreadItem; |
9 | use MediaWiki\MainConfigNames; |
10 | use MediaWiki\Title\Title; |
11 | use Wikimedia\Assert\Assert; |
12 | use Wikimedia\Parsoid\DOM\Comment; |
13 | use Wikimedia\Parsoid\DOM\Element; |
14 | use Wikimedia\Parsoid\DOM\Node; |
15 | use Wikimedia\Parsoid\DOM\Text; |
16 | use Wikimedia\Parsoid\Utils\DOMCompat; |
17 | |
18 | class CommentUtils { |
19 | |
20 | private function __construct() { |
21 | } |
22 | |
23 | private const BLOCK_ELEMENT_TYPES = [ |
24 | 'div', 'p', |
25 | // Tables |
26 | 'table', 'tbody', 'thead', 'tfoot', 'caption', 'th', 'tr', 'td', |
27 | // Lists |
28 | 'ul', 'ol', 'li', 'dl', 'dt', 'dd', |
29 | // HTML5 heading content |
30 | 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'hgroup', |
31 | // HTML5 sectioning content |
32 | 'article', 'aside', 'body', 'nav', 'section', 'footer', 'header', 'figure', |
33 | 'figcaption', 'fieldset', 'details', 'blockquote', |
34 | // Other |
35 | 'hr', 'button', 'canvas', 'center', 'col', 'colgroup', 'embed', |
36 | 'map', 'object', 'pre', 'progress', 'video' |
37 | ]; |
38 | |
39 | /** |
40 | * @param Node $node |
41 | * @return bool Node is a block element |
42 | */ |
43 | public static function isBlockElement( Node $node ): bool { |
44 | return $node instanceof Element && |
45 | in_array( strtolower( $node->tagName ), static::BLOCK_ELEMENT_TYPES, true ); |
46 | } |
47 | |
48 | private const SOL_TRANSPARENT_LINK_REGEX = |
49 | '/(?:^|\s)mw:PageProp\/(?:Category|redirect|Language)(?=$|\s)/D'; |
50 | |
51 | /** |
52 | * @param Node $node |
53 | * @return bool Node is considered a rendering-transparent node in Parsoid |
54 | */ |
55 | public static function isRenderingTransparentNode( Node $node ): bool { |
56 | $nextSibling = $node->nextSibling; |
57 | return ( |
58 | $node instanceof Comment || |
59 | ( $node instanceof Element && ( |
60 | strtolower( $node->tagName ) === 'meta' || |
61 | ( |
62 | strtolower( $node->tagName ) === 'link' && |
63 | preg_match( static::SOL_TRANSPARENT_LINK_REGEX, $node->getAttribute( 'rel' ) ?? '' ) |
64 | ) || |
65 | // Empty inline templates, e.g. tracking templates. (T269036) |
66 | // But not empty nodes that are just the start of a non-empty template about-group. (T290940) |
67 | ( |
68 | strtolower( $node->tagName ) === 'span' && |
69 | in_array( 'mw:Transclusion', explode( ' ', $node->getAttribute( 'typeof' ) ?? '' ), true ) && |
70 | !static::htmlTrim( DOMCompat::getInnerHTML( $node ) ) && |
71 | ( |
72 | !$nextSibling || !( $nextSibling instanceof Element ) || |
73 | // Maybe we should be checking all of the about-grouped nodes to see if they're empty, |
74 | // but that's prooobably not needed in practice, and it leads to a quadratic worst case. |
75 | $nextSibling->getAttribute( 'about' ) !== $node->getAttribute( 'about' ) |
76 | ) |
77 | ) |
78 | ) ) |
79 | ); |
80 | } |
81 | |
82 | /** |
83 | * @param Node $node |
84 | * @return bool Node was added to the page by DiscussionTools |
85 | */ |
86 | public static function isOurGeneratedNode( Node $node ): bool { |
87 | return $node instanceof Element && ( |
88 | DOMCompat::getClassList( $node )->contains( 'ext-discussiontools-init-replylink-buttons' ) || |
89 | $node->hasAttribute( 'data-mw-comment-start' ) || |
90 | $node->hasAttribute( 'data-mw-comment-end' ) |
91 | ); |
92 | } |
93 | |
94 | /** |
95 | * Elements which can't have element children (but some may have text content). |
96 | */ |
97 | private const NO_ELEMENT_CHILDREN_ELEMENT_TYPES = [ |
98 | // https://html.spec.whatwg.org/multipage/syntax.html#elements-2 |
99 | // Void elements |
100 | 'area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', |
101 | 'link', 'meta', 'param', 'source', 'track', 'wbr', |
102 | // Raw text elements |
103 | 'script', 'style', |
104 | // Escapable raw text elements |
105 | 'textarea', 'title', |
106 | // Foreign elements |
107 | 'math', 'svg', |
108 | // Treated like text when scripting is enabled in the parser |
109 | // https://html.spec.whatwg.org/#the-noscript-element |
110 | 'noscript', |
111 | // Replaced elements (that aren't already included above) |
112 | // https://html.spec.whatwg.org/multipage/rendering.html#replaced-elements |
113 | // They might allow element children, but they aren't rendered on the page. |
114 | 'audio', 'canvas', 'iframe', 'object', 'video', |
115 | ]; |
116 | |
117 | /** |
118 | * @param Node $node |
119 | * @return bool If true, node can't have element children. If false, it's complicated. |
120 | */ |
121 | public static function cantHaveElementChildren( Node $node ): bool { |
122 | return ( |
123 | $node instanceof Comment || |
124 | ( $node instanceof Element && ( |
125 | in_array( strtolower( $node->tagName ), static::NO_ELEMENT_CHILDREN_ELEMENT_TYPES, true ) || |
126 | // Thumbnail wrappers generated by MediaTransformOutput::linkWrap (T301427), |
127 | // for compatibility with TimedMediaHandler. |
128 | // There is no better way to detect them, and we can't insert markers here, |
129 | // because the media DOM CSS depends on specific tag names and their order :( |
130 | // TODO See if we can remove this condition when wgParserEnableLegacyMediaDOM=false |
131 | // is enabled everywhere. |
132 | ( |
133 | in_array( strtolower( $node->tagName ), [ 'a', 'span' ], true ) && |
134 | $node->firstChild && |
135 | // We always step inside a child node so this can't be infinite, silly Phan |
136 | // @phan-suppress-next-line PhanInfiniteRecursion |
137 | static::cantHaveElementChildren( $node->firstChild ) |
138 | ) || |
139 | // Do not insert anything inside figures when using wgParserEnableLegacyMediaDOM=false, |
140 | // because their CSS can't handle it (T320285). |
141 | strtolower( $node->tagName ) === 'figure' |
142 | ) ) |
143 | ); |
144 | } |
145 | |
146 | /** |
147 | * Check whether the node is a comment separator (instead of a part of the comment). |
148 | */ |
149 | public static function isCommentSeparator( Node $node ): bool { |
150 | if ( !( $node instanceof Element ) ) { |
151 | return false; |
152 | } |
153 | |
154 | $tagName = strtolower( $node->tagName ); |
155 | if ( $tagName === 'br' || $tagName === 'hr' ) { |
156 | return true; |
157 | } |
158 | |
159 | // TemplateStyles followed by any of the others |
160 | if ( $node->nextSibling && |
161 | ( $tagName === 'link' || $tagName === 'style' ) && |
162 | self::isCommentSeparator( $node->nextSibling ) |
163 | ) { |
164 | return true; |
165 | } |
166 | |
167 | $classList = DOMCompat::getClassList( $node ); |
168 | if ( |
169 | // Anything marked as not containing comments |
170 | $classList->contains( 'mw-notalk' ) || |
171 | // {{outdent}} templates |
172 | $classList->contains( 'outdent-template' ) || |
173 | // {{tracked}} templates (T313097) |
174 | $classList->contains( 'mw-trackedTemplate' ) |
175 | ) { |
176 | return true; |
177 | } |
178 | |
179 | // Wikitext definition list term markup (`;`) when used as a fake heading (T265964) |
180 | if ( $tagName === 'dl' && |
181 | count( $node->childNodes ) === 1 && |
182 | $node->firstChild instanceof Element && |
183 | strtolower( $node->firstChild->nodeName ) === 'dt' |
184 | ) { |
185 | return true; |
186 | } |
187 | |
188 | return false; |
189 | } |
190 | |
191 | /** |
192 | * Check whether the node is a comment content. It's a little vague what this means… |
193 | * |
194 | * @param Node $node Node, should be a leaf node (a node with no children) |
195 | * @return bool |
196 | */ |
197 | public static function isCommentContent( Node $node ): bool { |
198 | return ( |
199 | $node instanceof Text && |
200 | static::htmlTrim( $node->nodeValue ?? '' ) !== '' |
201 | ) || |
202 | ( |
203 | static::cantHaveElementChildren( $node ) |
204 | ); |
205 | } |
206 | |
207 | /** |
208 | * Get the index of $child in its parent |
209 | */ |
210 | public static function childIndexOf( Node $child ): int { |
211 | $i = 0; |
212 | while ( ( $child = $child->previousSibling ) ) { |
213 | $i++; |
214 | } |
215 | return $i; |
216 | } |
217 | |
218 | /** |
219 | * Check whether a Node contains (is an ancestor of) another Node (or is the same node) |
220 | */ |
221 | public static function contains( Node $ancestor, Node $descendant ): bool { |
222 | // TODO can we use Node->compareDocumentPosition() here maybe? |
223 | $node = $descendant; |
224 | while ( $node && $node !== $ancestor ) { |
225 | $node = $node->parentNode; |
226 | } |
227 | return $node === $ancestor; |
228 | } |
229 | |
230 | /** |
231 | * Find closest ancestor element using one of the given tag names. |
232 | * |
233 | * @param Node $node |
234 | * @param string[] $tagNames |
235 | * @return Element|null |
236 | */ |
237 | public static function closestElement( Node $node, array $tagNames ): ?Element { |
238 | do { |
239 | if ( |
240 | $node instanceof Element && |
241 | in_array( strtolower( $node->tagName ), $tagNames, true ) |
242 | ) { |
243 | return $node; |
244 | } |
245 | $node = $node->parentNode; |
246 | } while ( $node ); |
247 | return null; |
248 | } |
249 | |
250 | /** |
251 | * Find closest ancestor element that has sibling nodes |
252 | * |
253 | * @param Node $node |
254 | * @param string $direction Can be 'next', 'previous', or 'either' |
255 | * @return Element|null |
256 | */ |
257 | public static function closestElementWithSibling( Node $node, string $direction ): ?Element { |
258 | do { |
259 | if ( |
260 | $node instanceof Element && ( |
261 | ( $node->nextSibling && ( $direction === 'next' || $direction === 'either' ) ) || |
262 | ( $node->previousSibling && ( $direction === 'previous' || $direction === 'either' ) ) |
263 | ) |
264 | ) { |
265 | return $node; |
266 | } |
267 | $node = $node->parentNode; |
268 | } while ( $node ); |
269 | return null; |
270 | } |
271 | |
272 | /** |
273 | * Find the transclusion node which rendered the current node, if it exists. |
274 | * |
275 | * 1. Find the closest ancestor with an 'about' attribute |
276 | * 2. Find the main node of the about-group (first sibling with the same 'about' attribute) |
277 | * 3. If this is an mw:Transclusion node, return it; otherwise, go to step 1 |
278 | * |
279 | * @param Node $node |
280 | * @return Element|null Transclusion node, null if not found |
281 | */ |
282 | public static function getTranscludedFromElement( Node $node ): ?Element { |
283 | while ( $node ) { |
284 | // 1. |
285 | if ( |
286 | $node instanceof Element && |
287 | $node->getAttribute( 'about' ) && |
288 | preg_match( '/^#mwt\d+$/', $node->getAttribute( 'about' ) ?? '' ) |
289 | ) { |
290 | $about = $node->getAttribute( 'about' ); |
291 | |
292 | // 2. |
293 | while ( |
294 | ( $previousSibling = $node->previousSibling ) && |
295 | $previousSibling instanceof Element && |
296 | $previousSibling->getAttribute( 'about' ) === $about |
297 | ) { |
298 | $node = $previousSibling; |
299 | } |
300 | |
301 | // 3. |
302 | if ( |
303 | $node->getAttribute( 'typeof' ) && |
304 | in_array( 'mw:Transclusion', explode( ' ', $node->getAttribute( 'typeof' ) ?? '' ), true ) |
305 | ) { |
306 | break; |
307 | } |
308 | } |
309 | |
310 | $node = $node->parentNode; |
311 | } |
312 | return $node; |
313 | } |
314 | |
315 | /** |
316 | * Given a heading node, return the node on which the ID attribute is set. |
317 | * |
318 | * Also returns the offset within that node where the heading text starts. |
319 | * |
320 | * @param Element $heading Heading node (`<h1>`-`<h6>`) |
321 | * @return array Array containing a 'node' (Element) and offset (int) |
322 | */ |
323 | public static function getHeadlineNodeAndOffset( Element $heading ): array { |
324 | // This code assumes that $wgFragmentMode is [ 'html5', 'legacy' ] or [ 'html5' ] |
325 | $headline = $heading; |
326 | $offset = 0; |
327 | |
328 | if ( $headline->hasAttribute( 'data-mw-comment-start' ) ) { |
329 | $headline = $headline->parentNode; |
330 | Assert::precondition( $headline !== null, 'data-mw-comment-start was attached to a heading' ); |
331 | } |
332 | |
333 | if ( !$headline->getAttribute( 'id' ) && !$headline->getAttribute( 'data-mw-anchor' ) ) { |
334 | // PHP HTML: Find the child with .mw-headline |
335 | $headline = DOMCompat::querySelector( $headline, '.mw-headline' ); |
336 | if ( !$headline ) { |
337 | $headline = $heading; |
338 | } |
339 | } |
340 | |
341 | return [ |
342 | 'node' => $headline, |
343 | 'offset' => $offset, |
344 | ]; |
345 | } |
346 | |
347 | /** |
348 | * Trim ASCII whitespace, as defined in the HTML spec. |
349 | */ |
350 | public static function htmlTrim( string $str ): string { |
351 | // https://infra.spec.whatwg.org/#ascii-whitespace |
352 | return trim( $str, "\t\n\f\r " ); |
353 | } |
354 | |
355 | /** |
356 | * Get the indent level of $node, relative to $rootNode. |
357 | * |
358 | * The indent level is the number of lists inside of which it is nested. |
359 | */ |
360 | public static function getIndentLevel( Node $node, Element $rootNode ): int { |
361 | $indent = 0; |
362 | while ( $node ) { |
363 | if ( $node === $rootNode ) { |
364 | break; |
365 | } |
366 | $tagName = $node instanceof Element ? strtolower( $node->tagName ) : null; |
367 | if ( $tagName === 'li' || $tagName === 'dd' ) { |
368 | $indent++; |
369 | } |
370 | $node = $node->parentNode; |
371 | } |
372 | return $indent; |
373 | } |
374 | |
375 | /** |
376 | * Get an array of sibling nodes that contain parts of the given range. |
377 | * |
378 | * @param ImmutableRange $range |
379 | * @return Node[] |
380 | */ |
381 | public static function getCoveredSiblings( ImmutableRange $range ): array { |
382 | $ancestor = $range->commonAncestorContainer; |
383 | |
384 | // Convert to array early because apparently NodeList acts like a linked list |
385 | // and accessing items by index is slow |
386 | $siblings = iterator_to_array( $ancestor->childNodes ); |
387 | $start = 0; |
388 | $end = count( $siblings ) - 1; |
389 | |
390 | // Find first of the siblings that contains the item |
391 | if ( $ancestor === $range->startContainer ) { |
392 | $start = $range->startOffset; |
393 | } else { |
394 | while ( !static::contains( $siblings[ $start ], $range->startContainer ) ) { |
395 | $start++; |
396 | } |
397 | } |
398 | |
399 | // Find last of the siblings that contains the item |
400 | if ( $ancestor === $range->endContainer ) { |
401 | $end = $range->endOffset - 1; |
402 | } else { |
403 | while ( !static::contains( $siblings[ $end ], $range->endContainer ) ) { |
404 | $end--; |
405 | } |
406 | } |
407 | |
408 | return array_slice( $siblings, $start, $end - $start + 1 ); |
409 | } |
410 | |
411 | /** |
412 | * Get the nodes (if any) that contain the given thread item, and nothing else. |
413 | * |
414 | * @param ContentThreadItem $item |
415 | * @param ?Node $excludedAncestorNode Node that shouldn't be included in the result, even if it |
416 | * contains the item and nothing else. This is intended to avoid traversing outside of a node |
417 | * which is a container for all the thread items. |
418 | * @return Node[]|null |
419 | */ |
420 | public static function getFullyCoveredSiblings( |
421 | ContentThreadItem $item, ?Node $excludedAncestorNode = null |
422 | ): ?array { |
423 | $siblings = static::getCoveredSiblings( $item->getRange() ); |
424 | |
425 | $makeRange = static function ( $siblings ) { |
426 | return new ImmutableRange( |
427 | $siblings[0]->parentNode, |
428 | CommentUtils::childIndexOf( $siblings[0] ), |
429 | end( $siblings )->parentNode, |
430 | CommentUtils::childIndexOf( end( $siblings ) ) + 1 |
431 | ); |
432 | }; |
433 | |
434 | $matches = static::compareRanges( $makeRange( $siblings ), $item->getRange() ) === 'equal'; |
435 | |
436 | if ( $matches ) { |
437 | // If these are all of the children (or the only child), go up one more level |
438 | while ( |
439 | ( $parent = $siblings[ 0 ]->parentNode ) && |
440 | $parent !== $excludedAncestorNode && |
441 | static::compareRanges( $makeRange( [ $parent ] ), $item->getRange() ) === 'equal' |
442 | ) { |
443 | $siblings = [ $parent ]; |
444 | } |
445 | return $siblings; |
446 | } |
447 | return null; |
448 | } |
449 | |
450 | /** |
451 | * Unwrap Parsoid sections |
452 | * |
453 | * @param Element $element Parent element, e.g. document body |
454 | */ |
455 | public static function unwrapParsoidSections( Element $element ): void { |
456 | $sections = DOMCompat::querySelectorAll( $element, 'section[data-mw-section-id]' ); |
457 | foreach ( $sections as $section ) { |
458 | $parent = $section->parentNode; |
459 | while ( $section->firstChild ) { |
460 | $parent->insertBefore( $section->firstChild, $section ); |
461 | } |
462 | $parent->removeChild( $section ); |
463 | } |
464 | } |
465 | |
466 | /** |
467 | * Get a MediaWiki page title from a URL |
468 | * |
469 | * @param string $url Relative URL (from a `href` attribute) |
470 | * @param Config $config Config settings needed to resolve the relative URL |
471 | * @return string|null |
472 | */ |
473 | public static function getTitleFromUrl( string $url, Config $config ): ?string { |
474 | // Protocol-relative URLs are handled really badly by parse_url() |
475 | if ( str_starts_with( $url, '//' ) ) { |
476 | $url = "http:$url"; |
477 | } |
478 | |
479 | $bits = parse_url( $url ); |
480 | $query = wfCgiToArray( $bits['query'] ?? '' ); |
481 | if ( isset( $query['title'] ) ) { |
482 | return $query['title']; |
483 | } |
484 | |
485 | // TODO: Set the correct base in the document? |
486 | $articlePath = $config->get( MainConfigNames::ArticlePath ); |
487 | if ( str_starts_with( $url, './' ) ) { |
488 | // Assume this is URL in the format used by Parsoid documents |
489 | $url = substr( $url, 2 ); |
490 | $path = str_replace( '$1', $url, $articlePath ); |
491 | } elseif ( !str_contains( $url, '://' ) ) { |
492 | // Assume this is URL in the format used by legacy parser documents |
493 | $path = $url; |
494 | } else { |
495 | // External link |
496 | $path = $bits['path'] ?? ''; |
497 | } |
498 | |
499 | $articlePathRegexp = '/^' . str_replace( |
500 | '\\$1', |
501 | '([^?]*)', |
502 | preg_quote( $articlePath, '/' ) |
503 | ) . '/'; |
504 | $matches = null; |
505 | if ( preg_match( $articlePathRegexp, $path, $matches ) ) { |
506 | return urldecode( $matches[1] ); |
507 | } |
508 | return null; |
509 | } |
510 | |
511 | /** |
512 | * Traverse the document in depth-first order, calling the callback whenever entering and leaving |
513 | * a node. The walk starts before the given node and ends when callback returns a truthy value, or |
514 | * after reaching the end of the document. |
515 | * |
516 | * You might also think about this as processing XML token stream linearly (rather than XML |
517 | * nodes), as if we were parsing the document. |
518 | * |
519 | * @param Node $node Node to start at |
520 | * @param callable $callback Function accepting two arguments: $event ('enter' or 'leave') and |
521 | * $node (Node) |
522 | * @return mixed Final return value of the callback |
523 | */ |
524 | public static function linearWalk( Node $node, callable $callback ) { |
525 | $result = null; |
526 | [ $withinNode, $beforeNode ] = [ $node->parentNode, $node ]; |
527 | |
528 | while ( $beforeNode || $withinNode ) { |
529 | if ( $beforeNode ) { |
530 | $result = $callback( 'enter', $beforeNode ); |
531 | [ $withinNode, $beforeNode ] = [ $beforeNode, $beforeNode->firstChild ]; |
532 | } else { |
533 | $result = $callback( 'leave', $withinNode ); |
534 | [ $withinNode, $beforeNode ] = [ $withinNode->parentNode, $withinNode->nextSibling ]; |
535 | } |
536 | |
537 | if ( $result ) { |
538 | return $result; |
539 | } |
540 | } |
541 | return $result; |
542 | } |
543 | |
544 | /** |
545 | * Like #linearWalk, but it goes backwards. |
546 | * |
547 | * @inheritDoc ::linearWalk() |
548 | */ |
549 | public static function linearWalkBackwards( Node $node, callable $callback ) { |
550 | $result = null; |
551 | [ $withinNode, $beforeNode ] = [ $node->parentNode, $node ]; |
552 | |
553 | while ( $beforeNode || $withinNode ) { |
554 | if ( $beforeNode ) { |
555 | $result = $callback( 'enter', $beforeNode ); |
556 | [ $withinNode, $beforeNode ] = [ $beforeNode, $beforeNode->lastChild ]; |
557 | } else { |
558 | $result = $callback( 'leave', $withinNode ); |
559 | [ $withinNode, $beforeNode ] = [ $withinNode->parentNode, $withinNode->previousSibling ]; |
560 | } |
561 | |
562 | if ( $result ) { |
563 | return $result; |
564 | } |
565 | } |
566 | return $result; |
567 | } |
568 | |
569 | /** |
570 | * @param ImmutableRange $range (must not be collapsed) |
571 | * @return Node |
572 | */ |
573 | public static function getRangeFirstNode( ImmutableRange $range ): Node { |
574 | Assert::precondition( !$range->collapsed, 'Range is not collapsed' ); |
575 | // PHP bug: childNodes can be null |
576 | return $range->startContainer->childNodes && $range->startContainer->childNodes->length ? |
577 | $range->startContainer->childNodes[ $range->startOffset ] : |
578 | $range->startContainer; |
579 | } |
580 | |
581 | /** |
582 | * @param ImmutableRange $range (must not be collapsed) |
583 | * @return Node |
584 | */ |
585 | public static function getRangeLastNode( ImmutableRange $range ): Node { |
586 | Assert::precondition( !$range->collapsed, 'Range is not collapsed' ); |
587 | // PHP bug: childNodes can be null |
588 | return $range->endContainer->childNodes && $range->endContainer->childNodes->length ? |
589 | $range->endContainer->childNodes[ $range->endOffset - 1 ] : |
590 | $range->endContainer; |
591 | } |
592 | |
593 | /** |
594 | * Check whether two ranges overlap, and how. |
595 | * |
596 | * Includes a hack to check for "almost equal" ranges (whose start/end boundaries only differ by |
597 | * "uninteresting" nodes that we ignore when detecting comments), and treat them as equal. |
598 | * |
599 | * Illustration of return values: |
600 | * [ equal ] |
601 | * |[ contained ]| |
602 | * [ | contains | ] |
603 | * [overlap|start] | |
604 | * | [overlap|end] |
605 | * [before] | | |
606 | * | | [after] |
607 | * |
608 | * @param ImmutableRange $a |
609 | * @param ImmutableRange $b |
610 | * @return string One of: |
611 | * - 'equal': Ranges A and B are equal |
612 | * - 'contains': Range A contains range B |
613 | * - 'contained': Range A is contained within range B |
614 | * - 'after': Range A is before range B |
615 | * - 'before': Range A is after range B |
616 | * - 'overlapstart': Start of range A overlaps range B |
617 | * - 'overlapend': End of range A overlaps range B |
618 | */ |
619 | public static function compareRanges( ImmutableRange $a, ImmutableRange $b ): string { |
620 | // Compare the positions of: start of A to start of B, start of A to end of B, and so on. |
621 | // Watch out, the constant names are the opposite of what they should be. |
622 | $startToStart = $a->compareBoundaryPoints( ImmutableRange::START_TO_START, $b ); |
623 | $startToEnd = $a->compareBoundaryPoints( ImmutableRange::END_TO_START, $b ); |
624 | $endToStart = $a->compareBoundaryPoints( ImmutableRange::START_TO_END, $b ); |
625 | $endToEnd = $a->compareBoundaryPoints( ImmutableRange::END_TO_END, $b ); |
626 | |
627 | // Handle almost equal ranges: When start or end boundary points of the two ranges are different, |
628 | // but only differ by "uninteresting" nodes, treat them as equal instead. |
629 | if ( |
630 | ( $startToStart < 0 && static::compareRangesAlmostEqualBoundaries( $a, $b, 'start' ) ) || |
631 | ( $startToStart > 0 && static::compareRangesAlmostEqualBoundaries( $b, $a, 'start' ) ) |
632 | ) { |
633 | $startToStart = 0; |
634 | } |
635 | if ( |
636 | ( $endToEnd < 0 && static::compareRangesAlmostEqualBoundaries( $a, $b, 'end' ) ) || |
637 | ( $endToEnd > 0 && static::compareRangesAlmostEqualBoundaries( $b, $a, 'end' ) ) |
638 | ) { |
639 | $endToEnd = 0; |
640 | } |
641 | |
642 | if ( $startToStart === 0 && $endToEnd === 0 ) { |
643 | return 'equal'; |
644 | } |
645 | if ( $startToStart <= 0 && $endToEnd >= 0 ) { |
646 | return 'contains'; |
647 | } |
648 | if ( $startToStart >= 0 && $endToEnd <= 0 ) { |
649 | return 'contained'; |
650 | } |
651 | if ( $startToEnd >= 0 ) { |
652 | return 'after'; |
653 | } |
654 | if ( $endToStart <= 0 ) { |
655 | return 'before'; |
656 | } |
657 | if ( $startToStart > 0 && $startToEnd < 0 && $endToEnd >= 0 ) { |
658 | return 'overlapstart'; |
659 | } |
660 | if ( $endToEnd < 0 && $endToStart > 0 && $startToStart <= 0 ) { |
661 | return 'overlapend'; |
662 | } |
663 | |
664 | throw new LogicException( 'Unreachable' ); |
665 | } |
666 | |
667 | /** |
668 | * Check if the given boundary points of ranges A and B are almost equal (only differing by |
669 | * uninteresting nodes). |
670 | * |
671 | * Boundary of A must be before the boundary of B in the tree. |
672 | * |
673 | * @param ImmutableRange $a |
674 | * @param ImmutableRange $b |
675 | * @param string $boundary 'start' or 'end' |
676 | * @return bool |
677 | */ |
678 | private static function compareRangesAlmostEqualBoundaries( |
679 | ImmutableRange $a, ImmutableRange $b, string $boundary |
680 | ): bool { |
681 | // This code is awful, but several attempts to rewrite it made it even worse. |
682 | // You're welcome to give it a try. |
683 | |
684 | $from = $boundary === 'end' ? static::getRangeLastNode( $a ) : static::getRangeFirstNode( $a ); |
685 | $to = $boundary === 'end' ? static::getRangeLastNode( $b ) : static::getRangeFirstNode( $b ); |
686 | |
687 | $skipNode = null; |
688 | if ( $boundary === 'end' ) { |
689 | $skipNode = $from; |
690 | } |
691 | |
692 | $foundContent = false; |
693 | static::linearWalk( |
694 | $from, |
695 | static function ( string $event, Node $n ) use ( |
696 | $from, $to, $boundary, &$skipNode, &$foundContent |
697 | ) { |
698 | if ( $n === $to && $event === ( $boundary === 'end' ? 'leave' : 'enter' ) ) { |
699 | return true; |
700 | } |
701 | if ( $skipNode ) { |
702 | if ( $n === $skipNode && $event === 'leave' ) { |
703 | $skipNode = null; |
704 | } |
705 | return; |
706 | } |
707 | |
708 | if ( $event === 'enter' ) { |
709 | if ( |
710 | CommentUtils::isCommentSeparator( $n ) || |
711 | CommentUtils::isRenderingTransparentNode( $n ) || |
712 | CommentUtils::isOurGeneratedNode( $n ) |
713 | ) { |
714 | $skipNode = $n; |
715 | |
716 | } elseif ( |
717 | CommentUtils::isCommentContent( $n ) |
718 | ) { |
719 | $foundContent = true; |
720 | return true; |
721 | } |
722 | } |
723 | } |
724 | ); |
725 | |
726 | return !$foundContent; |
727 | } |
728 | |
729 | /** |
730 | * Assuming that the thread item set contains exactly one comment (or multiple comments with |
731 | * identical signatures, plus optional heading), check whether that comment is properly signed by |
732 | * the expected author (that is: there is a signature, and either there's nothing following the |
733 | * signature, or there's some text within the same paragraph that was detected as part of the same |
734 | * comment). |
735 | */ |
736 | public static function isSingleCommentSignedBy( |
737 | ContentThreadItemSet $itemSet, |
738 | string $author, |
739 | Element $rootNode |
740 | ): bool { |
741 | $items = $itemSet->getThreadItems(); |
742 | |
743 | if ( $items ) { |
744 | $lastItem = end( $items ); |
745 | // Check that we've detected a comment first, not just headings (T304377) |
746 | if ( !( $lastItem instanceof ContentCommentItem && $lastItem->getAuthor() === $author ) ) { |
747 | return false; |
748 | } |
749 | |
750 | // Range covering all of the detected items (to account for a heading, and for multiple |
751 | // signatures resulting in multiple comments) |
752 | $commentsRange = new ImmutableRange( |
753 | $items[0]->getRange()->startContainer, |
754 | $items[0]->getRange()->startOffset, |
755 | $lastItem->getRange()->endContainer, |
756 | $lastItem->getRange()->endOffset |
757 | ); |
758 | $bodyRange = new ImmutableRange( |
759 | $rootNode, 0, $rootNode, count( $rootNode->childNodes ) |
760 | ); |
761 | |
762 | if ( static::compareRanges( $commentsRange, $bodyRange ) === 'equal' ) { |
763 | // New comment includes a signature in the proper place |
764 | return true; |
765 | } |
766 | } |
767 | |
768 | return false; |
769 | } |
770 | |
771 | /** |
772 | * Get the ID for a new topics subscription from a page title |
773 | * |
774 | * @param Title $title Page title |
775 | * @return string ID for a new topics subscription |
776 | */ |
777 | public static function getNewTopicsSubscriptionId( Title $title ) { |
778 | return "p-topics-{$title->getNamespace()}:{$title->getDBkey()}"; |
779 | } |
780 | } |