/**
* Unicode directional formatting characters, for embedBidi()
*/
- static private $lre = "\xE2\x80\xAA"; // U+202A LEFT-TO-RIGHT EMBEDDING
- static private $rle = "\xE2\x80\xAB"; // U+202B RIGHT-TO-LEFT EMBEDDING
- static private $pdf = "\xE2\x80\xAC"; // U+202C POP DIRECTIONAL FORMATTING
+ static private $lre = "\u{202A}"; // U+202A LEFT-TO-RIGHT EMBEDDING
+ static private $rle = "\u{202B}"; // U+202B RIGHT-TO-LEFT EMBEDDING
+ static private $pdf = "\u{202C}"; // U+202C POP DIRECTIONAL FORMATTING
/**
* Directionality test regex for embedBidi(). Matches the first strong directionality codepoint:
}
// get the language object to process
- $langObj = isset( self::$mLangObjCache[$code] )
- ? self::$mLangObjCache[$code]
- : self::newFromCode( $code );
+ $langObj = self::$mLangObjCache[$code] ?? self::newFromCode( $code );
// merge the language object in to get it up front in the cache
self::$mLangObjCache = array_merge( [ $code => $langObj ], self::$mLangObjCache );
// Check if there is a language class for the code
$class = self::classFromCode( $code, $fallback );
- if ( class_exists( $class ) ) {
+ // LanguageCode does not inherit Language
+ if ( class_exists( $class ) && is_a( $class, 'Language', true ) ) {
$lang = new $class;
return $lang;
}
* language, script or variant codes actually exist in the repositories.
*
* Based on regexes by Mark Davis of the Unicode Consortium:
- * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
+ * https://www.unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
*
* @param string $code
* @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
*/
public function getNsText( $index ) {
$ns = $this->getNamespaces();
- return isset( $ns[$index] ) ? $ns[$index] : false;
+ return $ns[$index] ?? false;
}
/**
$ns = $wgExtraGenderNamespaces +
(array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
- return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
+ return $ns[$index][$gender] ?? $this->getNsText( $index );
}
/**
function getLocalNsIndex( $text ) {
$lctext = $this->lc( $text );
$ids = $this->getNamespaceIds();
- return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
+ return $ids[$lctext] ?? false;
}
/**
return $ns;
}
$ids = $this->getNamespaceIds();
- return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
+ return $ids[$lctext] ?? false;
}
/**
/**
* Get an array of language names, indexed by code.
* @param null|string $inLanguage Code of language in which to return the names
- * Use null for autonyms (native names)
+ * Use null for autonyms (native names)
* @param string $include One of:
- * 'all' all available languages
- * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
- * 'mwfile' only if the language is in 'mw' *and* has a message file
- * @return array Language code => language name
+ * 'all' all available languages
+ * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
+ * 'mwfile' only if the language is in 'mw' *and* has a message file
+ * @return array Language code => language name (sorted by key)
* @since 1.20
*/
public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
* 'all' all available languages
* 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
* 'mwfile' only if the language is in 'mw' *and* has a message file
- * @return array Language code => language name
+ * @return array Language code => language name (sorted by key)
*/
private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
global $wgExtraLanguageNames, $wgUsePigLatinVariant;
* @param string $ts 14-character timestamp
* YYYYMMDDHHMMSS
* 01234567890123
- * @param DateTimeZone $zone Timezone of $ts
+ * @param DateTimeZone|null $zone Timezone of $ts
* @param int &$ttl The amount of time (in seconds) the output may be cached for.
* Only makes sense if $ts is the current time.
* @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
# Add 543 years to the Gregorian calendar
# Months and days are identical
$gy_offset = $gy + 543;
+ # fix for dates between 1912 and 1941
+ # https://en.wikipedia.org/?oldid=836596673#New_year
+ if ( $gy >= 1912 && $gy <= 1940 ) {
+ if ( $gm <= 3 ) {
+ $gy_offset--;
+ }
+ $gm = ( $gm - 3 ) % 12;
+ }
} elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
# Minguo dates
# Deduct 1911 years from the Gregorian calendar
* Gets directionality of the first strongly directional codepoint, for embedBidi()
*
* This is the rule the BIDI algorithm uses to determine the directionality of
- * paragraphs ( http://unicode.org/reports/tr9/#The_Paragraph_Level ) and
- * FSI isolates ( http://unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
+ * paragraphs ( https://www.unicode.org/reports/tr9/#The_Paragraph_Level ) and
+ * FSI isolates ( https://www.unicode.org/reports/tr9/#Explicit_Directional_Isolates ).
*
* TODO: Does not handle BIDI control characters inside the text.
* TODO: Does not handle unallocated characters.
if ( $code < 0xac00 || 0xd7a4 <= $code ) {
return $matches[1];
} elseif ( $code < 0xb098 ) {
- return "\xe3\x84\xb1";
+ return "\u{3131}";
} elseif ( $code < 0xb2e4 ) {
- return "\xe3\x84\xb4";
+ return "\u{3134}";
} elseif ( $code < 0xb77c ) {
- return "\xe3\x84\xb7";
+ return "\u{3137}";
} elseif ( $code < 0xb9c8 ) {
- return "\xe3\x84\xb9";
+ return "\u{3139}";
} elseif ( $code < 0xbc14 ) {
- return "\xe3\x85\x81";
+ return "\u{3141}";
} elseif ( $code < 0xc0ac ) {
- return "\xe3\x85\x82";
+ return "\u{3142}";
} elseif ( $code < 0xc544 ) {
- return "\xe3\x85\x85";
+ return "\u{3145}";
} elseif ( $code < 0xc790 ) {
- return "\xe3\x85\x87";
+ return "\u{3147}";
} elseif ( $code < 0xcc28 ) {
- return "\xe3\x85\x88";
+ return "\u{3148}";
} elseif ( $code < 0xce74 ) {
- return "\xe3\x85\x8a";
+ return "\u{314A}";
} elseif ( $code < 0xd0c0 ) {
- return "\xe3\x85\x8b";
+ return "\u{314B}";
} elseif ( $code < 0xd30c ) {
- return "\xe3\x85\x8c";
+ return "\u{314C}";
} elseif ( $code < 0xd558 ) {
- return "\xe3\x85\x8d";
+ return "\u{314D}";
} else {
- return "\xe3\x85\x8e";
+ return "\u{314E}";
}
} else {
return '';
* @deprecated No-op since 1.28
*/
function initEncoding() {
+ wfDeprecated( __METHOD__, '1.28' );
// No-op.
}
* @deprecated No-op since 1.28
*/
function recodeForEdit( $s ) {
+ wfDeprecated( __METHOD__, '1.28' );
return $s;
}
* @deprecated No-op since 1.28
*/
function recodeInput( $s ) {
+ wfDeprecated( __METHOD__, '1.28' );
return $s;
}
global $wgAllUnicodeFixes;
$s = UtfNormal\Validator::cleanUp( $s );
if ( $wgAllUnicodeFixes ) {
- $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
- $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
+ $s = $this->transformUsingPairFile( 'normalize-ar.php', $s );
+ $s = $this->transformUsingPairFile( 'normalize-ml.php', $s );
}
return $s;
* @throws MWException
* @return string
*/
- function transformUsingPairFile( $file, $string ) {
+ protected function transformUsingPairFile( $file, $string ) {
if ( !isset( $this->transformData[$file] ) ) {
- $data = wfGetPrecompiledData( $file );
- if ( $data === false ) {
- throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
- }
+ global $IP;
+ $data = require "$IP/languages/data/{$file}";
$this->transformData[$file] = new ReplacementArray( $data );
}
return $this->transformData[$file]->replace( $string );
* @return string
*/
function getDirMark( $opposite = false ) {
- $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
- $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
+ $lrm = "\u{200E}"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
+ $rlm = "\u{200F}"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
if ( $opposite ) {
return $this->isRTL() ? $lrm : $rlm;
}
return;
}
$this->mMagicHookDone = true;
- Hooks::run( 'LanguageGetMagic', [ &$this->mMagicExtensions, $this->getCode() ] );
+ Hooks::run( 'LanguageGetMagic', [ &$this->mMagicExtensions, $this->getCode() ], '1.16' );
}
/**
$this->mExtendedSpecialPageAliases =
self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
Hooks::run( 'LanguageGetSpecialPageAliases',
- [ &$this->mExtendedSpecialPageAliases, $this->getCode() ] );
+ [ &$this->mExtendedSpecialPageAliases, $this->getCode() ], '1.16' );
}
return $this->mExtendedSpecialPageAliases;
* @param int $length Maximum length (including ellipsis)
* @param string $ellipsis String to append to the truncated text
* @param bool $adjustLength Subtract length of ellipsis from $length.
- * $adjustLength was introduced in 1.18, before that behaved as if false.
+ * $adjustLength was introduced in 1.18, before that behaved as if false.
* @return string
*/
function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
* @return string
*/
private function truncateInternal(
- $string, $length, $ellipsis = '...', $adjustLength = true, $measureLength, $getSubstring
+ $string, $length, $ellipsis, $adjustLength, $measureLength, $getSubstring
) {
if ( !is_callable( $measureLength ) || !is_callable( $getSubstring ) ) {
throw new InvalidArgumentException( 'Invalid callback provided' );
if ( $gender === 'female' ) {
return $forms[1];
}
- return isset( $forms[2] ) ? $forms[2] : $forms[0];
+ return $forms[2] ?? $forms[0];
}
/**
* match up with it.
*
* @param string $str The validated block duration in English
- * @param User $user User object to use timezone from or null for $wgUser
+ * @param User|null $user User object to use timezone from or null for $wgUser
* @param int $now Current timestamp, for formatting relative block durations
* @return string Somehow translated block duration
* @see LanguageFi.php for example implementation
* the "raw" tag (-{R| }-) to prevent conversion.
*
* This function is called "markNoConversion" for historical
- * reasons.
+ * reasons *BUT DIFFERS SIGNIFICANTLY* from
+ * LanguageConverter::markNoConversion(), with which it is easily
+ * confused.
*
* @param string $text Text to be used for external link
* @param bool $noParse Wrap it without confirming it's a real URL first
* @return string The tagged text
+ * @deprecated since 1.32, use LanguageConverter::markNoConversion()
+ * instead.
*/
public function markNoConversion( $text, $noParse = false ) {
+ wfDeprecated( __METHOD__, '1.32' );
// Excluding protocal-relative URLs may avoid many false positives.
if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
return $this->mConverter->markNoConversion( $text );
* @return bool
*/
public function equals( Language $lang ) {
- return $lang->getCode() === $this->mCode;
+ return $lang === $this || $lang->getCode() === $this->mCode;
}
/**
* @throws MWException
* @return string $prefix . $mangledCode . $suffix
*/
- public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
+ public static function getFileName( $prefix, $code, $suffix = '.php' ) {
if ( !self::isValidBuiltInCode( $code ) ) {
throw new MWException( "Invalid language code \"$code\"" );
}
dépôts / lhc / web / wiklou.git / commitdiff