Merge "Use Remex in Sanitizer::stripAllTags()"

autoload.php

@@ -1219,6 +1219,7 @@ $wgAutoloadLocalClasses = [
 	'RefreshLinks' => __DIR__ . '/maintenance/refreshLinks.php',
 	'RefreshLinksJob' => __DIR__ . '/includes/jobqueue/jobs/RefreshLinksJob.php',
 	'RegexlikeReplacer' => __DIR__ . '/includes/libs/replacers/RegexlikeReplacer.php',
+	'RemexStripTagHandler' => __DIR__ . '/includes/parser/RemexStripTagHandler.php',
 	'RemoveInvalidEmails' => __DIR__ . '/maintenance/removeInvalidEmails.php',
 	'RemoveUnusedAccounts' => __DIR__ . '/maintenance/removeUnusedAccounts.php',
 	'RenameDbPrefix' => __DIR__ . '/maintenance/renameDbPrefix.php',

includes/parser/RemexStripTagHandler.php

@@ -0,0 +1,40 @@
+<?php
+
+use RemexHtml\Tokenizer\Attributes;
+use RemexHtml\Tokenizer\TokenHandler;
+use RemexHtml\Tokenizer\Tokenizer;
+
+/**
+ * @internal
+ */
+class RemexStripTagHandler implements TokenHandler {
+	private $text = '';
+	public function getResult() {
+		return $this->text;
+	}
+
+	function startDocument( Tokenizer $t, $fns, $fn ) {
+		// Do nothing.
+	}
+	function endDocument( $pos ) {
+		// Do nothing.
+	}
+	function error( $text, $pos ) {
+		// Do nothing.
+	}
+	function characters( $text, $start, $length, $sourceStart, $sourceLength ) {
+		$this->text .= substr( $text, $start, $length );
+	}
+	function startTag( $name, Attributes $attrs, $selfClose, $sourceStart, $sourceLength ) {
+		// Do nothing.
+	}
+	function endTag( $name, $sourceStart, $sourceLength ) {
+		// Do nothing.
+	}
+	function doctype( $name, $public, $system, $quirks, $sourceStart, $sourceLength ) {
+		// Do nothing.
+	}
+	function comment( $text, $sourceStart, $sourceLength ) {
+		// Do nothing.
+	}
+}

includes/parser/Sanitizer.php

@@ -1967,17 +1967,22 @@ class Sanitizer {
 	 * Warning: this return value must be further escaped for literal
 	 * inclusion in HTML output as of 1.10!
 	 *
-	 * @param string $text HTML fragment
+	 * @param string $html HTML fragment
 	 * @return string
 	 */
-	static function stripAllTags( $text ) {
-		# Actual <tags>
-		$text = StringUtils::delimiterReplace( '<', '>', '', $text );
+	static function stripAllTags( $html ) {
+		// Use RemexHtml to tokenize $html and extract the text
+		$handler = new RemexStripTagHandler;
+		$tokenizer = new RemexHtml\Tokenizer\Tokenizer( $handler, $html, [
+			'ignoreErrors' => true,
+			// don't ignore char refs, we want them to be decoded
+			'ignoreNulls' => true,
+			'skipPreprocess' => true,
+		] );
+		$tokenizer->execute();
+		$text = $handler->getResult();
 
-		# Normalize &entities and whitespace
-		$text = self::decodeCharReferences( $text );
 		$text = self::normalizeWhitespace( $text );
-
 		return $text;
 	}
 

tests/phpunit/includes/parser/SanitizerTest.php

@@ -530,11 +530,10 @@ class SanitizerTest extends MediaWikiTestCase {
 			[ '<p id="one">Foo</p><p id="two">Bar</p>', 'FooBar' ],
 			[ "<p>Foo</p>\n<p>Bar</p>", 'Foo Bar' ],
 			[ '<p>Hello &lt;strong&gt; wor&#x6c;&#100; caf&eacute;</p>', 'Hello <strong> world café' ],
-			// This one is broken, see T179978
-			//[
-			//	'<p><small data-foo=\'bar"&lt;baz>quux\'><a href="./Foo">Bar</a></small> Whee!</p>',
-			//	'Bar Whee!'
-			//],
+			[
+				'<p><small data-foo=\'bar"&lt;baz>quux\'><a href="./Foo">Bar</a></small> Whee!</p>',
+				'Bar Whee!'
+			],
 			[ '1<span class="<?php">2</span>3', '123' ],
 			[ '1<span class="<?">2</span>3', '123' ],
 		];