Merge "Parser: Add guessSectionNameFromStrippedText() and refactor"

[lhc/web/wiklou.git] / includes / parser / Parser.php

diff --git a/includes/parser/Parser.php b/includes/parser/Parser.php
index 3548da9..2b03a70 100644 (file)
--- a/includes/parser/Parser.php
+++ b/includes/parser/Parser.php
@@ -3944,7 +3944,7 @@ class Parser {
                        $this->mForceTocPosition = true;
 
                        # Set a placeholder. At the end we'll fill it in with the TOC.
-                       $text = $mw->replace( '<!--MWTOC-->', $text, 1 );
+                       $text = $mw->replace( '<!--MWTOC\'"-->', $text, 1 );
 
                        # Only keep the first one.
                        $text = $mw->replace( '', $text );
@@ -4207,7 +4207,7 @@ class Parser {
                        # Decode HTML entities
                        $safeHeadline = Sanitizer::decodeCharReferences( $safeHeadline );
 
-                       $safeHeadline = $this->normalizeSectionName( $safeHeadline );
+                       $safeHeadline = self::normalizeSectionName( $safeHeadline );
 
                        $fallbackHeadline = Sanitizer::escapeIdForAttribute( $safeHeadline, Sanitizer::ID_FALLBACK );
                        $linkAnchor = Sanitizer::escapeIdForLink( $safeHeadline );
@@ -4390,7 +4390,7 @@ class Parser {
                $full .= implode( '', $sections );
 
                if ( $this->mForceTocPosition ) {
-                       return str_replace( '<!--MWTOC-->', $toc, $full );
+                       return str_replace( '<!--MWTOC\'"-->', $toc, $full );
                } else {
                        return $full;
                }
@@ -5756,23 +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 );
-               $text = $this->normalizeSectionName( $text );
-
-               return '#' . Sanitizer::escapeIdForLink( $text );
+               $sectionName = self::getSectionNameFromStrippedText( $text );
+               return self::makeAnchor( $sectionName );
        }
 
        /**
@@ -5782,25 +5801,23 @@ 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 );
-               $text = $this->normalizeSectionName( $text );
-
-               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 );
-               }
+               $sectionName = self::getSectionNameFromStrippedText( $text );
+               return self::makeLegacyAnchor( $sectionName );
+       }
 
-               return "#$id";
+       /**
+        * 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 );
        }
 
        /**
@@ -5809,7 +5826,7 @@ class Parser {
         * @param string $text
         * @return string
         */
-       private function normalizeSectionName( $text ) {
+       private static function normalizeSectionName( $text ) {
                # T90902: ensure the same normalization is applied for IDs as to links
                $titleParser = MediaWikiServices::getInstance()->getTitleParser();
                try {