* Note that this only works in terms of sequences of digits, and the behavior for decimal fractions
* or pretty-formatted numbers may be unexpected.
*
+ * Digits will be based on the wiki's content language settings. If
+ * you change the content langauge of a wiki you will need to run
+ * updateCollation.php --force. Only English (ASCII 0-9) and the
+ * localized version will be counted. Localized digits from other languages
+ * or weird unicode digit equivalents (e.g. 4, 𝟜, ⓸ , ⁴, etc) will not count.
+ *
* @since 1.28
*/
class NumericUppercaseCollation extends UppercaseCollation {
+
+ /**
+ * @var $digitTransformLang Language How to convert digits (usually $wgContLang)
+ */
+ private $digitTransformLang;
+
+ /**
+ * Constructor
+ *
+ * @param $lang Language How to convert digits.
+ * For example, if given language "my" than ၇ is treated like 7.
+ *
+ * It is expected that usually this is given $wgContLang.
+ */
+ public function __construct( Language $lang ) {
+ $this->digitTransformLang = $lang;
+ parent::__construct();
+ }
+
public function getSortKey( $string ) {
$sortkey = parent::getSortKey( $string );
-
+ $sortkey = $this->convertDigits( $sortkey );
// For each sequence of digits, insert the digit '0' and then the length of the sequence
// (encoded in two bytes) before it. That's all folks, it sorts correctly now! The '0' ensures
// correct position (where digits would normally sort), then the length will be compared putting
return $sortkey;
}
+ /**
+ * Convert localized digits to english digits.
+ *
+ * based on Language::parseFormattedNumber but without commas.
+ *
+ * @param $string String sortkey to unlocalize digits of
+ * @return String Sortkey with all localized digits replaced with ASCII digits.
+ */
+ private function convertDigits( $string ) {
+ $table = $this->digitTransformLang->digitTransformTable();
+ if ( $table ) {
+ $table = array_filter( $table );
+ $flipped = array_flip( $table );
+ // Some languages seem to also have commas in this table.
+ $flipped = array_filter( $flipped, 'is_numeric' );
+ $string = strtr( $string, $flipped );
+ }
+ return $string;
+ }
+
public function getFirstLetter( $string ) {
- if ( preg_match( '/^\d/', $string ) ) {
- // Note that we pass 0 and 9 as normal params, not numParams(). This only works for 0-9
- // and not localised digits, so we don't want them to be converted.
- return wfMessage( 'category-header-numerals' )->params( 0, 9 )->text();
+ $convertedString = $this->convertDigits( $string );
+
+ if ( preg_match( '/^\d/', $convertedString ) ) {
+ return wfMessage( 'category-header-numerals' )
+ ->numParams( 0, 9 )
+ ->text();
} else {
return parent::getFirstLetter( $string );
}