X-Git-Url: https://git.heureux-cyclage.org/?a=blobdiff_plain;f=includes%2FCollation.php;h=87afc103a6b8f7f67751f3ff577a505fac96a7ec;hb=30ba4c00ebfa163b4ae2304f1db83d6dc56e1598;hp=3cc7902855d1210962cc926308d9bf80b4c07863;hpb=29ab088e2fb938b0b3330bc4d4891a155e9420d2;p=lhc%2Fweb%2Fwiklou.git

diff --git a/includes/Collation.php b/includes/Collation.php
index 3cc7902855..87afc103a6 100644
--- a/includes/Collation.php
+++ b/includes/Collation.php
@@ -48,8 +48,12 @@ abstract class Collation {
 			case 'uca-default':
 				return new IcuCollation( 'root' );
 			default:
-				# Provide a mechanism for extensions to hook in.
+				$match = array();
+				if ( preg_match( '/^uca-([a-z-]+)$/', $collationName, $match ) ) {
+					return new IcuCollation( $match[1] );
+				}
 
+				# Provide a mechanism for extensions to hook in.
 				$collationObject = null;
 				wfRunHooks( 'Collation::factory', array( $collationName, &$collationObject ) );
 
@@ -176,6 +180,87 @@ class IcuCollation extends Collation {
 		array( 0x2F800, 0x2FA1F ), // CJK Compatibility Ideographs Supplement
 	);
 
+	/**
+	 * Additional characters (or character groups) to be considered first-letters
+	 *
+	 * Generated based on the primary level of Unicode collation tailorings
+	 * available at http://developer.mimer.com/charts/tailorings.htm .
+	 *
+	 * Empty arrays are intended; this signifies that the data for the language is
+	 * available and that there are, in fact, no additional letters to consider.
+	 */
+	static $tailoringFirstLetters = array(
+		// Verified by native speakers
+		'pl' => array( "Ą", "Ć", "Ę", "Ł", "Ń", "Ó", "Ś", "Ź", "Ż" ),
+		'fi' => array( "Å", "Ä", "Ö" ),
+		// Not verified, but likely correct
+		'af' => array(),
+		'ast' => array( "CH", "LL", "Ñ" ),
+		'az' => array( "Ç", "Ə", "Ğ", "İ", "Ö", "Ş", "Ü" ),
+		'be' => array( "Ё" ),
+		'bg' => array(),
+		'br' => array( "CH", "C'H" ),
+		'bs' => array( "Č", "Ć", "DŽ", "Đ", "LJ", "NJ", "Š", "Ž" ),
+		'ca' => array(),
+		'co' => array(),
+		'cs' => array( "Č", "CH", "Ř", "Š", "Ž" ),
+		'cy' => array( "CH", "DD", "FF", "NG", "LL", "PH", "RH", "TH" ),
+		'da' => array( "Æ", "Ø", "Å" ),
+		'de' => array(),
+		'dsb' => array( "Č", "Ć", "DŹ", "Ě", "CH", "Ł", "Ń", "Ŕ", "Š", "Ś", "Ž", "Ź" ),
+		'el' => array(),
+		'en' => array(),
+		'eo' => array( "Ĉ", "Ĝ", "Ĥ", "Ĵ", "Ŝ", "Ŭ" ),
+		'es' => array( "Ñ" ),
+		'et' => array( "Š", "Ž", "Õ", "Ä", "Ö", "Ü" ),
+		'eu' => array( "Ñ" ),
+		'fo' => array( "Á", "Ð", "Í", "Ó", "Ú", "Ý", "Æ", "Ø", "Å" ),
+		'fr' => array(),
+		'fur' => array( "À", "Á", "Â", "È", "Ì", "Ò", "Ù" ),
+		'fy' => array(),
+		'ga' => array(),
+		'gd' => array(),
+		'gl' => array( "CH", "LL", "Ñ" ),
+		'hr' => array( "Č", "Ć", "DŽ", "Đ", "LJ", "NJ", "Š", "Ž" ),
+		'hsb' => array( "Č", "DŹ", "Ě", "CH", "Ł", "Ń", "Ř", "Š", "Ć", "Ž" ),
+		'hu' => array( "CS", "DZ", "DZS", "GY", "LY", "NY", "Ö", "SZ", "TY", "Ü", "ZS" ),
+		'is' => array( "Á", "Ð", "É", "Í", "Ó", "Ú", "Ý", "Þ", "Æ", "Ö", "Å" ),
+		'it' => array(),
+		'kk' => array( "Ү", "І" ),
+		'kl' => array( "Æ", "Ø", "Å" ),
+		'ku' => array( "Ç", "Ê", "Î", "Ş", "Û" ),
+		'ky' => array( "Ё" ),
+		'la' => array(),
+		'lb' => array(),
+		'lt' => array( "Č", "Š", "Ž" ),
+		'lv' => array( "Č", "Ģ", "Ķ", "Ļ", "Ņ", "Š", "Ž" ),
+		'mk' => array(),
+		'mo' => array( "Ă", "Â", "Î", "Ş", "Ţ" ),
+		'mt' => array( "Ċ", "Ġ", "GĦ", "Ħ", "Ż" ),
+		'nl' => array(),
+		'no' => array( "Æ", "Ø", "Å" ),
+		'oc' => array(),
+		'pt' => array(),
+		'rm' => array(),
+		'ro' => array( "Ă", "Â", "Î", "Ş", "Ţ" ),
+		'ru' => array(),
+		'rup' => array( "Ă", "Â", "Î", "Ľ", "Ń", "Ş", "Ţ" ),
+		'sco' => array(),
+		'sk' => array( "Ä", "Č", "CH", "Ô", "Š", "Ž" ),
+		'sl' => array( "Č", "Š", "Ž" ),
+		'smn' => array( "Á", "Č", "Đ", "Ŋ", "Š", "Ŧ", "Ž", "Æ", "Ø", "Å", "Ä", "Ö" ),
+		'sq' => array( "Ç", "DH", "Ë", "GJ", "LL", "NJ", "RR", "SH", "TH", "XH", "ZH" ),
+		'sr' => array(),
+		'sv' => array( "Å", "Ä", "Ö" ),
+		'tk' => array( "Ç", "Ä", "Ž", "Ň", "Ö", "Ş", "Ü", "Ý" ),
+		'tl' => array( "Ñ", "NG" ), /* 'fil' in the data source */
+		'tr' => array( "Ç", "Ğ", "İ", "Ö", "Ş", "Ü" ),
+		'tt' => array( "Ә", "Ө", "Ү", "Җ", "Ң", "Һ" ),
+		'uk' => array( "Ґ", "Ь" ),
+		'uz' => array( "CH", "G'", "NG", "O'", "SH" ),
+		'vi' => array( "Ă", "Â", "Đ", "Ê", "Ô", "Ơ", "Ư" ),
+	);
+
 	const RECORD_LENGTH = 14;
 
 	function __construct( $locale ) {
@@ -256,10 +341,15 @@ class IcuCollation extends Collation {
 
 		// Generate data from serialized data file
 
-		$letters = wfGetPrecompiledData( "first-letters-{$this->locale}.ser" );
-		if ( $letters === false ) {
-			throw new MWException( "MediaWiki does not support ICU locale " .
-				"\"{$this->locale}\"" );
+		if ( isset ( self::$tailoringFirstLetters[$this->locale] ) ) {
+			$letters = wfGetPrecompiledData( "first-letters-root.ser" );
+			$letters = array_merge( $letters, self::$tailoringFirstLetters[$this->locale] );
+		} else {
+			$letters = wfGetPrecompiledData( "first-letters-{$this->locale}.ser" );
+			if ( $letters === false ) {
+				throw new MWException( "MediaWiki does not support ICU locale " .
+					"\"{$this->locale}\"" );
+			}
 		}
 
 		// Sort the letters.
@@ -374,5 +464,55 @@ class IcuCollation extends Collation {
 		}
 		return false;
 	}
-}
 
+	/**
+	 * Return the version of ICU library used by PHP's intl extension,
+	 * or false when the extension is not installed of the version
+	 * can't be determined.
+	 *
+	 * The constant INTL_ICU_VERSION this function refers to isn't really
+	 * documented. It is available since PHP 5.3.7 (see PHP bug 54561).
+	 * This function will return false on older PHPs.
+	 *
+	 * @since 1.21
+	 * @return string|false
+	 */
+	static function getICUVersion() {
+		return defined( 'INTL_ICU_VERSION' ) ? INTL_ICU_VERSION : false;
+	}
+
+	/**
+	 * Return the version of Unicode appropriate for the version of ICU library
+	 * currently in use, or false when it can't be determined.
+	 *
+	 * @since 1.21
+	 * @return string|false
+	 */
+	static function getUnicodeVersionForICU() {
+		$icuVersion = IcuCollation::getICUVersion();
+		if ( !$icuVersion ) {
+			return false;
+		}
+
+		$versionPrefix = substr( $icuVersion, 0, 3 );
+		// Source: http://site.icu-project.org/download
+		$map = array(
+			'50.' => '6.2',
+			'49.' => '6.1',
+			'4.8' => '6.0',
+			'4.6' => '6.0',
+			'4.4' => '5.2',
+			'4.2' => '5.1',
+			'4.0' => '5.1',
+			'3.8' => '5.0',
+			'3.6' => '5.0',
+			'3.4' => '4.1',
+		);
+
+		if ( isset( $map[$versionPrefix] ) ) {
+			return $map[$versionPrefix];
+		} else {
+			return false;
+		}
+	}
+}