diff options
Diffstat (limited to 'languages/LanguageConverter.php')
| -rw-r--r-- | languages/LanguageConverter.php | 88 |
1 files changed, 54 insertions, 34 deletions
diff --git a/languages/LanguageConverter.php b/languages/LanguageConverter.php index ba89be74..5e4bce8d 100644 --- a/languages/LanguageConverter.php +++ b/languages/LanguageConverter.php @@ -117,7 +117,7 @@ class LanguageConverter { * in this case. Right now this is only used by zh. * * @param $variant String: the language code of the variant - * @return String: The code of the fallback language or the + * @return String|array: The code of the fallback language or the * main code if there is no fallback */ public function getVariantFallbacks( $variant ) { @@ -158,7 +158,7 @@ class LanguageConverter { // not memoized (i.e. there return value is not cached) since // new information might appear during processing after this // is first called. - if ( $req ) { + if ( $this->validateVariant( $req ) ) { return $req; } return $this->mMainLanguageCode; @@ -189,7 +189,7 @@ class LanguageConverter { * @param $variant String: the variant to validate * @return Mixed: returns the variant if it is valid, null otherwise */ - protected function validateVariant( $variant = null ) { + public function validateVariant( $variant = null ) { if ( $variant !== null && in_array( $variant, $this->mVariants ) ) { return $variant; } @@ -322,6 +322,11 @@ class LanguageConverter { } } + if( $this->guessVariant( $text, $toVariant ) ) { + wfProfileOut( __METHOD__ ); + return $text; + } + /* we convert everything except: 1. HTML markups (anything between < and >) 2. HTML entities @@ -368,11 +373,11 @@ class LanguageConverter { $sourceBlob .= substr( $text, $startPos, $elementPos - $startPos ) . "\000"; // Advance to the next position - $startPos = $elementPos + strlen( $element ); + $startPos = $elementPos + strlen( $element ); // Translate any alt or title attributes inside the matched element - if ( $element !== '' && preg_match( '/^(<[^>\s]*)\s([^>]*)(.*)$/', $element, - $elementMatches ) ) + if ( $element !== '' && preg_match( '/^(<[^>\s]*)\s([^>]*)(.*)$/', $element, + $elementMatches ) ) { $attrs = Sanitizer::decodeTagAttributes( $elementMatches[2] ); $changed = false; @@ -385,7 +390,7 @@ class LanguageConverter { if ( !strpos( $attr, '://' ) ) { $attr = $this->translate( $attr, $toVariant ); } - + // Remove HTML tags to avoid disrupting the layout $attr = preg_replace( '/<[^>]+>/', '', $attr ); if ( $attr !== $attrs[$attrName] ) { @@ -394,7 +399,7 @@ class LanguageConverter { } } if ( $changed ) { - $element = $elementMatches[1] . Html::expandAttributes( $attrs ) . + $element = $elementMatches[1] . Html::expandAttributes( $attrs ) . $elementMatches[3]; } } @@ -571,7 +576,7 @@ class LanguageConverter { */ public function convertTo( $text, $variant ) { global $wgDisableLangConversion; - if ( $wgDisableLangConversion ) { + if ( $wgDisableLangConversion || $this->guessVariant( $text, $variant ) ) { return $text; } return $this->recursiveConvertTopLevel( $text, $variant ); @@ -773,6 +778,20 @@ class LanguageConverter { } /** + * Guess if a text is written in a variant. This should be implemented in subclasses. + * + * @param string $text the text to be checked + * @param string $variant language code of the variant to be checked for + * @return bool true if $text appears to be written in $variant, false if not + * + * @author Nikola Smolenski <smolensk@eunet.rs> + * @since 1.19 + */ + public function guessVariant($text, $variant) { + return false; + } + + /** * Load default conversion tables. * This method must be implemented in derived class. * @@ -870,26 +889,26 @@ class LanguageConverter { return array(); } - if ( strpos( $code, '/' ) === false ) { - $txt = MessageCache::singleton()->get( 'Conversiontable', true, $code ); - if ( $txt === false ) { - # @todo FIXME: This method doesn't seem to be expecting - # this possible outcome... - $txt = '<Conversiontable>'; - } + $parsed[$key] = true; + + if ( $subpage === '' ) { + $txt = MessageCache::singleton()->get( 'conversiontable', true, $code ); } else { - $title = Title::makeTitleSafe( - NS_MEDIAWIKI, - "Conversiontable/$code" - ); + $txt = false; + $title = Title::makeTitleSafe( NS_MEDIAWIKI, $key ); if ( $title && $title->exists() ) { - $article = new Article( $title ); - $txt = $article->getContents(); - } else { - $txt = ''; + $revision = Revision::newFromTitle( $title ); + if ( $revision ) { + $txt = $revision->getRawText(); + } } } + # Nothing to parse if there's no text + if ( $txt === false || $txt === null || $txt === '' ) { + return array(); + } + // get all subpage links of the form // [[MediaWiki:Conversiontable/zh-xx/...|...]] $linkhead = $this->mLangObj->getNsText( NS_MEDIAWIKI ) . @@ -938,7 +957,6 @@ class LanguageConverter { $ret[trim( $m[0] )] = trim( $tt[0] ); } } - $parsed[$key] = true; // recursively parse the subpages if ( $recursive ) { @@ -1364,19 +1382,21 @@ class ConverterRule { if ( isset( $this->mVariantFlags[$variant] ) ) { // then convert <text to convert> to current language $this->mRules = $this->mConverter->autoConvert( $this->mRules, - $variant ); + $variant ); } else { // if current variant no in flags, // then we check its fallback variants. $variantFallbacks = $this->mConverter->getVariantFallbacks( $variant ); - foreach ( $variantFallbacks as $variantFallback ) { - // if current variant's fallback exist in flags - if ( isset( $this->mVariantFlags[$variantFallback] ) ) { - // then convert <text to convert> to fallback language - $this->mRules = - $this->mConverter->autoConvert( $this->mRules, - $variantFallback ); - break; + if( is_array( $variantFallbacks ) ) { + foreach ( $variantFallbacks as $variantFallback ) { + // if current variant's fallback exist in flags + if ( isset( $this->mVariantFlags[$variantFallback] ) ) { + // then convert <text to convert> to fallback language + $this->mRules = + $this->mConverter->autoConvert( $this->mRules, + $variantFallback ); + break; + } } } } |