X-Git-Url: https://git.heureux-cyclage.org/?a=blobdiff_plain;f=includes%2Fparser%2FParser.php;h=87d76fa949514777af85834511162ad3bda4127c;hb=6f07389ef2eba048c17c7e343fb8fbb026a4dbd6;hp=e901f6f311c4dedf24b81354a4ecd52814f50fbf;hpb=655931210678d0090cd60a8463f2bb6d95099b0e;p=lhc%2Fweb%2Fwiklou.git

diff --git a/includes/parser/Parser.php b/includes/parser/Parser.php
index e901f6f311..87d76fa949 100644
--- a/includes/parser/Parser.php
+++ b/includes/parser/Parser.php
@@ -2296,8 +2296,10 @@ class Parser {
 						$this->mOutput->addLanguageLink( $nt->getFullText() );
 					}
 
-					$s = rtrim( $s . $prefix );
-					$s .= trim( $trail, "\n" ) == '' ? '' : $prefix . $trail;
+					/**
+					 * Strip the whitespace interwiki links produce, see T10897
+					 */
+					$s = rtrim( $s . $prefix ) . $trail; # T175416
 					continue;
 				}
 
@@ -2322,7 +2324,10 @@ class Parser {
 						continue;
 					}
 				} elseif ( $ns == NS_CATEGORY ) {
-					$s = rtrim( $s . "\n" ); # T2087
+					/**
+					 * Strip the whitespace Category links produce, see T2087
+					 */
+					$s = rtrim( $s . $prefix ) . $trail; # T2087, T87753
 
 					if ( $wasblank ) {
 						$sortkey = $this->getDefaultSort();
@@ -2334,11 +2339,6 @@ class Parser {
 					$sortkey = $this->getConverterLanguage()->convertCategoryKey( $sortkey );
 					$this->mOutput->addCategory( $nt->getDBkey(), $sortkey );
 
-					/**
-					 * Strip the whitespace Category links produce, see T2087
-					 */
-					$s .= trim( $prefix . $trail, "\n" ) == '' ? '' : $prefix . $trail;
-
 					continue;
 				}
 			}
@@ -2504,10 +2504,10 @@ class Parser {
 				$value = '|';
 				break;
 			case 'currentmonth':
-				$value = $pageLang->formatNum( MWTimestamp::getInstance( $ts )->format( 'm' ) );
+				$value = $pageLang->formatNum( MWTimestamp::getInstance( $ts )->format( 'm' ), true );
 				break;
 			case 'currentmonth1':
-				$value = $pageLang->formatNum( MWTimestamp::getInstance( $ts )->format( 'n' ) );
+				$value = $pageLang->formatNum( MWTimestamp::getInstance( $ts )->format( 'n' ), true );
 				break;
 			case 'currentmonthname':
 				$value = $pageLang->getMonthName( MWTimestamp::getInstance( $ts )->format( 'n' ) );
@@ -2519,16 +2519,16 @@ class Parser {
 				$value = $pageLang->getMonthAbbreviation( MWTimestamp::getInstance( $ts )->format( 'n' ) );
 				break;
 			case 'currentday':
-				$value = $pageLang->formatNum( MWTimestamp::getInstance( $ts )->format( 'j' ) );
+				$value = $pageLang->formatNum( MWTimestamp::getInstance( $ts )->format( 'j' ), true );
 				break;
 			case 'currentday2':
-				$value = $pageLang->formatNum( MWTimestamp::getInstance( $ts )->format( 'd' ) );
+				$value = $pageLang->formatNum( MWTimestamp::getInstance( $ts )->format( 'd' ), true );
 				break;
 			case 'localmonth':
-				$value = $pageLang->formatNum( MWTimestamp::getLocalInstance( $ts )->format( 'm' ) );
+				$value = $pageLang->formatNum( MWTimestamp::getLocalInstance( $ts )->format( 'm' ), true );
 				break;
 			case 'localmonth1':
-				$value = $pageLang->formatNum( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
+				$value = $pageLang->formatNum( MWTimestamp::getLocalInstance( $ts )->format( 'n' ), true );
 				break;
 			case 'localmonthname':
 				$value = $pageLang->getMonthName( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
@@ -2540,10 +2540,10 @@ class Parser {
 				$value = $pageLang->getMonthAbbreviation( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
 				break;
 			case 'localday':
-				$value = $pageLang->formatNum( MWTimestamp::getLocalInstance( $ts )->format( 'j' ) );
+				$value = $pageLang->formatNum( MWTimestamp::getLocalInstance( $ts )->format( 'j' ), true );
 				break;
 			case 'localday2':
-				$value = $pageLang->formatNum( MWTimestamp::getLocalInstance( $ts )->format( 'd' ) );
+				$value = $pageLang->formatNum( MWTimestamp::getLocalInstance( $ts )->format( 'd' ), true );
 				break;
 			case 'pagename':
 				$value = wfEscapeWikiText( $this->mTitle->getText() );
@@ -4206,6 +4206,9 @@ class Parser {
 
 			# Decode HTML entities
 			$safeHeadline = Sanitizer::decodeCharReferences( $safeHeadline );
+
+			$safeHeadline = self::normalizeSectionName( $safeHeadline );
+
 			$fallbackHeadline = Sanitizer::escapeIdForAttribute( $safeHeadline, Sanitizer::ID_FALLBACK );
 			$linkAnchor = Sanitizer::escapeIdForLink( $safeHeadline );
 			$safeHeadline = Sanitizer::escapeIdForAttribute( $safeHeadline, Sanitizer::ID_PRIMARY );
@@ -5753,21 +5756,42 @@ class Parser {
 		return $this->mDefaultSort;
 	}
 
+	private static function getSectionNameFromStrippedText( $text ) {
+		$text = Sanitizer::normalizeSectionNameWhitespace( $text );
+		$text = Sanitizer::decodeCharReferences( $text );
+		$text = self::normalizeSectionName( $text );
+		return $text;
+	}
+
+	private static function makeAnchor( $sectionName ) {
+		return '#' . Sanitizer::escapeIdForLink( $sectionName );
+	}
+
+	private static function makeLegacyAnchor( $sectionName ) {
+		global $wgFragmentMode;
+		if ( isset( $wgFragmentMode[1] ) && $wgFragmentMode[1] === 'legacy' ) {
+			// ForAttribute() and ForLink() are the same for legacy encoding
+			$id = Sanitizer::escapeIdForAttribute( $text, Sanitizer::ID_FALLBACK );
+		} else {
+			$id = Sanitizer::escapeIdForLink( $text );
+		}
+
+		return "#$id";
+	}
+
 	/**
 	 * Try to guess the section anchor name based on a wikitext fragment
 	 * presumably extracted from a heading, for example "Header" from
 	 * "== Header ==".
 	 *
 	 * @param string $text
-	 *
-	 * @return string
+	 * @return string Anchor (starting with '#')
 	 */
 	public function guessSectionNameFromWikiText( $text ) {
 		# Strip out wikitext links(they break the anchor)
 		$text = $this->stripSectionName( $text );
-		$text = Sanitizer::normalizeSectionNameWhitespace( $text );
-		$text = Sanitizer::decodeCharReferences( $text );
-		return '#' . Sanitizer::escapeIdForLink( $text );
+		$sectionName = self::getSectionNameFromStrippedText( $text );
+		return self::makeAnchor( $sectionName );
 	}
 
 	/**
@@ -5777,24 +5801,41 @@ class Parser {
 	 * than UTF-8, resulting in breakage.
 	 *
 	 * @param string $text The section name
-	 * @return string An anchor
+	 * @return string Anchor (starting with '#')
 	 */
 	public function guessLegacySectionNameFromWikiText( $text ) {
-		global $wgFragmentMode;
-
 		# Strip out wikitext links(they break the anchor)
 		$text = $this->stripSectionName( $text );
-		$text = Sanitizer::normalizeSectionNameWhitespace( $text );
-		$text = Sanitizer::decodeCharReferences( $text );
+		$sectionName = self::getSectionNameFromStrippedText( $text );
+		return self::makeLegacyAnchor( $sectionName );
+	}
 
-		if ( isset( $wgFragmentMode[1] ) && $wgFragmentMode[1] === 'legacy' ) {
-			// ForAttribute() and ForLink() are the same for legacy encoding
-			$id = Sanitizer::escapeIdForAttribute( $text, Sanitizer::ID_FALLBACK );
-		} else {
-			$id = Sanitizer::escapeIdForLink( $text );
-		}
+	/**
+	 * Like guessSectionNameFromWikiText(), but takes already-stripped text as input.
+	 * @param string $text Section name (plain text)
+	 * @return string Anchor (starting with '#')
+	 */
+	public static function guessSectionNameFromStrippedText( $text ) {
+		$sectionName = self::getSectionNameFromStrippedText( $text );
+		return self::makeAnchor( $sectionName );
+	}
 
-		return "#$id";
+	/**
+	 * Apply the same normalization as code making links to this section would
+	 *
+	 * @param string $text
+	 * @return string
+	 */
+	private static function normalizeSectionName( $text ) {
+		# T90902: ensure the same normalization is applied for IDs as to links
+		$titleParser = MediaWikiServices::getInstance()->getTitleParser();
+		try {
+
+			$parts = $titleParser->splitTitleString( "#$text" );
+		} catch ( MalformedTitleException $ex ) {
+			return $text;
+		}
+		return $parts['fragment'];
 	}
 
 	/**