X-Git-Url: http://git.cyclocoop.org/?a=blobdiff_plain;f=languages%2FLanguage.php;h=463ea19062c43a0c291e8ddd024a52c2ce4cef7b;hb=ddcf8cc660c810f1601b61c90aa33cb36ca60cad;hp=7d4041a8f66caea6da209b7687e3ee0e926c919b;hpb=6e2acfd778db237e2d36f654ffeae55ae2b32922;p=lhc%2Fweb%2Fwiklou.git

diff --git a/languages/Language.php b/languages/Language.php
index 7d4041a8f6..463ea19062 100644
--- a/languages/Language.php
+++ b/languages/Language.php
@@ -29,10 +29,15 @@ if ( function_exists( 'mb_strtoupper' ) ) {
  * @ingroup Language
  */
 class FakeConverter {
+
+	/**
+	 * @var Language
+	 */
 	var $mLang;
 	function __construct( $langobj ) { $this->mLang = $langobj; }
 	function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
 	function convert( $t ) { return $t; }
+	function convertTo( $text, $variant ) { return $text; }
 	function convertTitle( $t ) { return $t->getPrefixedText(); }
 	function getVariants() { return array( $this->mLang->getCode() ); }
 	function getPreferredVariant() { return $this->mLang->getCode(); }
@@ -63,10 +68,11 @@ class Language {
 	var $mMagicExtensions = array(), $mMagicHookDone = false;
 	private $mHtmlCode = null;
 
-	var $mNamespaceIds, $namespaceNames, $namespaceAliases;
 	var $dateFormatStrings = array();
 	var $mExtendedSpecialPageAliases;
 
+	protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
+
 	/**
 	 * ReplacementArray object caches
 	 */
@@ -152,6 +158,7 @@ class Language {
 	/**
 	 * Create a language object for a given language code
 	 * @param $code String
+	 * @throws MWException
 	 * @return Language
 	 */
 	protected static function newFromCode( $code ) {
@@ -329,7 +336,7 @@ class Language {
 	/**
 	 * @return array
 	 */
-	function getNamespaces() {
+	public function getNamespaces() {
 		if ( is_null( $this->namespaceNames ) ) {
 			global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
 
@@ -363,6 +370,14 @@ class Language {
 		return $this->namespaceNames;
 	}
 
+	/**
+	 * Arbitrarily set all of the namespace names at once. Mainly used for testing
+	 * @param $namespaces Array of namespaces (id => name)
+	 */
+	public function setNamespaces( array $namespaces ) {
+		$this->namespaceNames = $namespaces;
+	}
+
 	/**
 	 * A convenience function that returns the same thing as
 	 * getNamespaces() except with the array values changed to ' '
@@ -541,11 +556,10 @@ class Language {
 	 */
 	function getVariantname( $code, $usemsg = true ) {
 		$msg = "variantname-$code";
-		list( $rootCode ) = explode( '-', $code );
 		if ( $usemsg && wfMessage( $msg )->exists() ) {
 			return $this->getMessageFromDB( $msg );
 		}
-		$name = self::getLanguageName( $code );
+		$name = self::fetchLanguageName( $code );
 		if ( $name ) {
 			return $name; # if it's defined as a language name, show that
 		} else {
@@ -638,36 +652,17 @@ class Language {
 	}
 
 	/**
-	 * Get language names, indexed by code.
+	 * Get native language names, indexed by code.
+	 * Only those defined in MediaWiki, no other data like CLDR.
 	 * If $customisedOnly is true, only returns codes with a messages file
 	 *
 	 * @param $customisedOnly bool
 	 *
 	 * @return array
+	 * @deprecated in 1.20, use fetchLanguageNames()
 	 */
 	public static function getLanguageNames( $customisedOnly = false ) {
-		global $wgExtraLanguageNames;
-		static $coreLanguageNames;
-
-		if ( $coreLanguageNames === null ) {
-			include( MWInit::compiledPath( 'languages/Names.php' ) );
-		}
-
-		$allNames = $wgExtraLanguageNames + $coreLanguageNames;
-		if ( !$customisedOnly ) {
-			return $allNames;
-		}
-
-		global $IP;
-		$names = array();
-		// We do this using a foreach over the codes instead of a directory
-		// loop so that messages files in extensions will work correctly.
-		foreach ( $allNames as $code => $value ) {
-			if ( is_readable( self::getMessagesFileName( $code ) ) ) {
-				$names[$code] = $allNames[$code];
-			}
-		}
-		return $names;
+		return self::fetchLanguageNames( null, $customisedOnly ? 'mwfile' : 'mw' );
 	}
 
 	/**
@@ -677,16 +672,84 @@ class Language {
 	 * @param $code String Language code.
 	 * @return Array language code => language name
 	 * @since 1.18.0
+	 * @deprecated in 1.20, use fetchLanguageNames()
 	 */
 	public static function getTranslatedLanguageNames( $code ) {
+		return self::fetchLanguageNames( $code, 'all' );
+	}
+
+	/**
+	 * Get an array of language names, indexed by code.
+	 * @param $inLanguage null|string: Code of language in which to return the names
+	 *		Use null for autonyms (native names)
+	 * @param $include string:
+	 *		'all' all available languages
+	 *		'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames
+	 *		'mwfile' only if the language is in 'mw' *and* has a message file
+	 * @return array|bool: language code => language name, false if $include is wrong
+	 * @since 1.20
+	 */
+	public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
+		global $wgExtraLanguageNames;
+		static $coreLanguageNames;
+
+		if ( $coreLanguageNames === null ) {
+			include( MWInit::compiledPath( 'languages/Names.php' ) );
+		}
+
 		$names = array();
-		wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $code ) );
 
-		foreach ( self::getLanguageNames() as $code => $name ) {
-			if ( !isset( $names[$code] ) ) $names[$code] = $name;
+		if( $inLanguage ) {
+			# TODO: also include when $inLanguage is null, when this code is more efficient
+			wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
+		}
+
+		$mwNames = $wgExtraLanguageNames + $coreLanguageNames;
+		foreach ( $mwNames as $mwCode => $mwName ) {
+			# - Prefer own MediaWiki native name when not using the hook
+			#	TODO: prefer it always to make it consistent, but casing is different in CLDR
+			# - For other names just add if not added through the hook
+			if ( ( $mwCode === $inLanguage && !$inLanguage ) || !isset( $names[$mwCode] ) ) {
+				$names[$mwCode] = $mwName;
+			}
+		}
+
+		if ( $include === 'all' ) {
+			return $names;
+		}
+
+		$returnMw = array();
+		$coreCodes = array_keys( $mwNames );
+		foreach( $coreCodes as $coreCode ) {
+			$returnMw[$coreCode] = $names[$coreCode];
+		}
+
+		if( $include === 'mw' ) {
+			return $returnMw;
+		} elseif( $include === 'mwfile' ) {
+			$namesMwFile = array();
+			# We do this using a foreach over the codes instead of a directory
+			# loop so that messages files in extensions will work correctly.
+			foreach ( $returnMw as $code => $value ) {
+				if ( is_readable( self::getMessagesFileName( $code ) ) ) {
+					$namesMwFile[$code] = $names[$code];
+				}
+			}
+			return $namesMwFile;
 		}
+		return false;
+	}
 
-		return $names;
+	/**
+	 * @param $code string: The code of the language for which to get the name
+	 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
+	 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
+	 * @return string: Language name or empty
+	 * @since 1.20
+	 */
+	public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
+		$array = self::fetchLanguageNames( $inLanguage, $include );
+		return !array_key_exists( $code, $array ) ? '' : $array[$code];
 	}
 
 	/**
@@ -700,15 +763,14 @@ class Language {
 	}
 
 	/**
+	 * Get the native language name of $code.
+	 * Only if defined in MediaWiki, no other data like CLDR.
 	 * @param $code string
 	 * @return string
+	 * @deprecated in 1.20, use fetchLanguageName()
 	 */
 	function getLanguageName( $code ) {
-		$names = self::getLanguageNames();
-		if ( !array_key_exists( $code, $names ) ) {
-			return '';
-		}
-		return $names[$code];
+		return self::fetchLanguageName( $code );
 	}
 
 	/**
@@ -824,6 +886,7 @@ class Language {
 	 *    xij  j (day number) in Iranian calendar
 	 *    xiF  F (month name) in Iranian calendar
 	 *    xin  n (month number) in Iranian calendar
+	 *    xiy  y (two digit year) in Iranian calendar
 	 *    xiY  Y (full year) in Iranian calendar
 	 *
 	 *    xjj  j (day number) in Hebrew calendar
@@ -1042,7 +1105,7 @@ class Language {
 					if ( !$unix ) {
 						$unix = wfTimestamp( TS_UNIX, $ts );
 					}
-					$num = date( 'o', $unix );
+					$num = gmdate( 'o', $unix );
 					break;
 				case 'Y':
 					$num = substr( $ts, 0, 4 );
@@ -1086,6 +1149,12 @@ class Language {
 				case 'y':
 					$num = substr( $ts, 2, 2 );
 					break;
+				case 'xiy':
+					if ( !$iranian ) {
+						$iranian = self::tsToIranian( $ts );
+					}
+					$num = substr( $iranian[0], -2 );
+					break;
 				case 'a':
 					$s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
 					break;
@@ -1242,7 +1311,7 @@ class Language {
 	 *
 	 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
 	 *
-	 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
+	 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
 	 *
 	 * @param $ts string
 	 *
@@ -1651,7 +1720,7 @@ class Language {
 	 *            get user timecorrection setting)
 	 * @return int
 	 */
-	function userAdjust( $ts, $tz = false )	{
+	function userAdjust( $ts, $tz = false ) {
 		global $wgUser, $wgLocalTZoffset;
 
 		if ( $tz === false ) {
@@ -1855,6 +1924,7 @@ class Language {
 	 *             - true: use user's preference
 	 *             - false: use default preference
 	 *             - string: format to use
+	 * @since 1.19
 	 * @return String
 	 */
 	private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
@@ -1893,6 +1963,7 @@ class Language {
 	 *             - true: use user's preference
 	 *             - false: use default preference
 	 *             - string: format to use
+	 * @since 1.19
 	 * @return String
 	 */
 	public function userDate( $ts, User $user, array $options = array() ) {
@@ -1915,6 +1986,7 @@ class Language {
 	 *             - true: use user's preference
 	 *             - false: use default preference
 	 *             - string: format to use
+	 * @since 1.19
 	 * @return String
 	 */
 	public function userTime( $ts, User $user, array $options = array() ) {
@@ -1937,6 +2009,7 @@ class Language {
 	 *             - true: use user's preference
 	 *             - false: use default preference
 	 *             - string: format to use
+	 * @since 1.19
 	 * @return String
 	 */
 	public function userTimeAndDate( $ts, User $user, array $options = array() ) {
@@ -2467,6 +2540,7 @@ class Language {
 	 * @param $file string
 	 * @param $string string
 	 *
+	 * @throws MWException
 	 * @return string
 	 */
 	function transformUsingPairFile( $file, $string ) {
@@ -2522,16 +2596,35 @@ class Language {
 	}
 
 	/**
-	 * A hidden direction mark (LRM or RLM), depending on the language direction
+	 * A hidden direction mark (LRM or RLM), depending on the language direction.
+	 * Unlike getDirMark(), this function returns the character as an HTML entity.
+	 * This function should be used when the output is guaranteed to be HTML,
+	 * because it makes the output HTML source code more readable. When
+	 * the output is plain text or can be escaped, getDirMark() should be used.
+	 *
+	 * @param $opposite Boolean Get the direction mark opposite to your language
+	 * @return string
+	 */
+	function getDirMarkEntity( $opposite = false ) {
+		if ( $opposite ) { return $this->isRTL() ? '&lrm;' : '&rlm;'; }
+		return $this->isRTL() ? '&rlm;' : '&lrm;';
+	}
+
+	/**
+	 * A hidden direction mark (LRM or RLM), depending on the language direction.
+	 * This function produces them as invisible Unicode characters and
+	 * the output may be hard to read and debug, so it should only be used
+	 * when the output is plain text or can be escaped. When the output is
+	 * HTML, use getDirMarkEntity() instead.
 	 *
 	 * @param $opposite Boolean Get the direction mark opposite to your language
 	 * @return string
 	 */
 	function getDirMark( $opposite = false ) {
-		$rtl = "\xE2\x80\x8F";
-		$ltr = "\xE2\x80\x8E";
-		if ( $opposite ) { return $this->isRTL() ? $ltr : $rtl; }
-		return $this->isRTL() ? $rtl : $ltr;
+		$lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
+		$rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
+		if ( $opposite ) { return $this->isRTL() ? $lrm : $rlm; }
+		return $this->isRTL() ? $rlm : $lrm;
 	}
 
 	/**
@@ -2716,6 +2809,9 @@ class Language {
 	 */
 	function commafy( $_ ) {
 		$digitGroupingPattern = $this->digitGroupingPattern();
+		if ( $_ === null ) {
+			return '';
+		}
 
 		if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
 			// default grouping is at thousands,  use the same for ###,###,### pattern too.
@@ -2726,7 +2822,7 @@ class Language {
 			if ( intval( $_ ) < 0 ) {
 				// For negative numbers apply the algorithm like positive number and add sign.
 				$sign =  "-";
-				$_ = substr( $_,1 );
+				$_ = substr( $_, 1 );
 			}
 			$numberpart = array();
 			$decimalpart = array();
@@ -3105,7 +3201,7 @@ class Language {
 	 * (b) clear $tag value
 	 * @param &$tag string Current HTML tag name we are looking at
 	 * @param $tagType int (0-open tag, 1-close tag)
-	 * @param $lastCh char|string Character before the '>' that ended this tag
+	 * @param $lastCh string Character before the '>' that ended this tag
 	 * @param &$openTags array Open tag stack (not accounting for $tag)
 	 */
 	private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
@@ -3137,14 +3233,31 @@ class Language {
 		}
 		return $word;
 	}
-
+	/**
+	 * Get the grammar forms for the content language
+	 * @return array of grammar forms
+	 * @since 1.20
+	 */
+	function getGrammarForms() {
+		global $wgGrammarForms;
+		if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
+			 return $wgGrammarForms[$this->getCode()];
+		}
+		return array();
+	}
 	/**
 	 * Provides an alternative text depending on specified gender.
 	 * Usage {{gender:username|masculine|feminine|neutral}}.
 	 * username is optional, in which case the gender of current user is used,
 	 * but only in (some) interface messages; otherwise default gender is used.
-	 * If second or third parameter are not specified, masculine is used.
-	 * These details may be overriden per language.
+	 *
+	 * If no forms are given, an empty string is returned. If only one form is
+	 * given, it will be returned unconditionally. These details are implied by
+	 * the caller and cannot be overridden in subclasses.
+	 *
+	 * If more than one form is given, the default is to use the neutral one
+	 * if it is specified, and to use the masculine one otherwise. These
+	 * details can be overridden in subclasses.
 	 *
 	 * @param $gender string
 	 * @param $forms array
@@ -3212,7 +3325,7 @@ class Language {
 	 * match up with it.
 	 *
 	 * @param $str String: the validated block duration in English
-	 * @return Somehow translated block duration
+	 * @return string Somehow translated block duration
 	 * @see LanguageFi.php for example implementation
 	 */
 	function translateBlockExpiry( $str ) {
@@ -3261,6 +3374,8 @@ class Language {
 
 	/**
 	 * Return the LanguageConverter used in the Language
+	 *
+	 * @since 1.19
 	 * @return LanguageConverter
 	 */
 	public function getConverter() {
@@ -3308,6 +3423,8 @@ class Language {
 
 	/**
 	 * Check if the language has the specific variant
+	 *
+	 * @since 1.19
 	 * @param $variant string
 	 * @return bool
 	 */
@@ -3484,7 +3601,7 @@ class Language {
 	public function setCode( $code ) {
 		$this->mCode = $code;
 		// Ensure we don't leave an incorrect html code lying around
-		unset( $this->mHtmlCode );
+		$this->mHtmlCode = null;
 	}
 
 	/**
@@ -3492,6 +3609,7 @@ class Language {
 	 * @param $prefix string Prepend this to the filename
 	 * @param $code string Language code
 	 * @param $suffix string Append this to the filename
+	 * @throws MWException
 	 * @return string $prefix . $mangledCode . $suffix
 	 */
 	public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
@@ -3547,7 +3665,7 @@ class Language {
 	 *
 	 * @param $code string
 	 *
-	 * @return false|string
+	 * @return bool|string
 	 */
 	public static function getFallbackFor( $code ) {
 		if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
@@ -3670,6 +3788,8 @@ class Language {
 	/**
 	 * Decode an expiry (block, protection, etc) which has come from the DB
 	 *
+	 * @FIXME: why are we returnings DBMS-dependent strings???
+	 *
 	 * @param $expiry String: Database expiry String
 	 * @param $format Bool|Int true to process using language functions, or TS_ constant
 	 *     to return the expiry in a given timestamp
@@ -3793,57 +3913,67 @@ class Language {
 	}
 
 	/**
+	 * Format a bitrate for output, using an appropriate
+	 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
+	 *
+	 * This use base 1000. For base 1024 use formatSize(), for another base
+	 * see formatComputingNumbers()
+	 *
 	 * @param $bps int
 	 * @return string
 	 */
 	function formatBitrate( $bps ) {
-		$units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
-		if ( $bps <= 0 ) {
-			return $this->formatNum( $bps ) . $units[0];
-		}
-		$unitIndex = (int)floor( log10( $bps ) / 3 );
-		$mantissa = $bps / pow( 1000, $unitIndex );
-		if ( $mantissa < 10 ) {
-			$mantissa = round( $mantissa, 1 );
-		} else {
-			$mantissa = round( $mantissa );
-		}
-		return $this->formatNum( $mantissa ) . $units[$unitIndex];
+		return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
 	}
 
 	/**
-	 * Format a size in bytes for output, using an appropriate
-	 * unit (B, KB, MB or GB) according to the magnitude in question
-	 *
-	 * @param $size int Size to format
-	 * @return string Plain text (not HTML)
+	 * @param $size int Size of the unit
+	 * @param $boundary int Size boundary (1000, or 1024 in most cases)
+	 * @param $messageKey string Message key to be uesd
+	 * @return string
 	 */
-	function formatSize( $size ) {
+	function formatComputingNumbers( $size, $boundary, $messageKey ) {
+		if ( $size <= 0 ) {
+			return str_replace( '$1', $this->formatNum( $size ),
+				$this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
+			);
+		}
+		$sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
+		$index = 0;
+
+		$maxIndex = count( $sizes ) - 1;
+		while ( $size >= $boundary && $index < $maxIndex ) {
+			$index++;
+			$size /= $boundary;
+		}
+
 		// For small sizes no decimal places necessary
 		$round = 0;
-		if ( $size > 1024 ) {
-			$size = $size / 1024;
-			if ( $size > 1024 ) {
-				$size = $size / 1024;
-				// For MB and bigger two decimal places are smarter
-				$round = 2;
-				if ( $size > 1024 ) {
-					$size = $size / 1024;
-					$msg = 'size-gigabytes';
-				} else {
-					$msg = 'size-megabytes';
-				}
-			} else {
-				$msg = 'size-kilobytes';
-			}
-		} else {
-			$msg = 'size-bytes';
+		if ( $index > 1 ) {
+			// For MB and bigger two decimal places are smarter
+			$round = 2;
 		}
+		$msg = str_replace( '$1', $sizes[$index], $messageKey );
+
 		$size = round( $size, $round );
 		$text = $this->getMessageFromDB( $msg );
 		return str_replace( '$1', $this->formatNum( $size ), $text );
 	}
 
+	/**
+	 * Format a size in bytes for output, using an appropriate
+	 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
+	 *
+	 * This method use base 1024. For base 1000 use formatBitrate(), for
+	 * another base see formatComputingNumbers()
+	 *
+	 * @param $size int Size to format
+	 * @return string Plain text (not HTML)
+	 */
+	function formatSize( $size ) {
+		return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
+	}
+
 	/**
 	 * Make a list item, used by various special pages
 	 *
@@ -3876,7 +4006,7 @@ class Language {
 
 		# Make 'previous' link
 		$prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
-		if( $offset > 0 ) {
+		if ( $offset > 0 ) {
 			$plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
 				$query, $prev, 'prevn-title', 'mw-prevlink' );
 		} else {
@@ -3885,7 +4015,7 @@ class Language {
 
 		# Make 'next' link
 		$next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
-		if( $atend ) {
+		if ( $atend ) {
 			$nlink = htmlspecialchars( $next );
 		} else {
 			$nlink = $this->numLink( $title, $offset + $limit, $limit,
@@ -3894,7 +4024,7 @@ class Language {
 
 		# Make links to set number of items per page
 		$numLinks = array();
-		foreach( array( 20, 50, 100, 250, 500 ) as $num ) {
+		foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
 			$numLinks[] = $this->numLink( $title, $offset, $num,
 				$query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
 		}