languages/LanguageUtf8.php

   1 <?php
   2 if( defined( "MEDIAWIKI" ) ) {
   3
   4 $wgInputEncoding    = "utf-8";
   5 $wgOutputEncoding       = "utf-8";
   6
   7 $wikiUpperChars = $wgMemc->get( $key1 = "$wgDBname:utf8:upper" );
   8 $wikiLowerChars = $wgMemc->get( $key2 = "$wgDBname:utf8:lower" );
   9
  10 if(empty( $wikiUpperChars) || empty($wikiLowerChars )) {
  11         require_once( "includes/Utf8Case.php" );
  12         $wgMemc->set( $key1, $wikiUpperChars );
  13         $wgMemc->set( $key2, $wikiLowerChars );
  14 }
  15
  16 # Base stuff useful to all UTF-8 based language files
  17 class LanguageUtf8 extends Language {
  18
  19         function ucfirst( $string ) {
  20                 # For most languages, this is a wrapper for ucfirst()
  21                 # But that doesn't work right in a UTF-8 locale
  22                 global $wikiUpperChars, $wikiLowerChars;
  23                 return preg_replace (
  24                 "/^([\\x00-\\x7f]|[\\xc0-\\xff][\\x80-\\xbf]*)/e",
  25                 "strtr ( \"\$1\" , \$wikiUpperChars )",
  26                 $string );
  27         }
  28
  29         function lcfirst( $string ) {
  30                 global $wikiUpperChars, $wikiLowerChars;
  31                 return preg_replace (
  32                 "/^([\\x00-\\x7f]|[\\xc0-\\xff][\\x80-\\xbf]*)/e",
  33                 "strtr ( \"\$1\" , \$wikiLowerChars )",
  34                 $string );
  35         }
  36
  37         function stripForSearch( $string ) {
  38                 # MySQL fulltext index doesn't grok utf-8, so we
  39                 # need to fold cases and convert to hex
  40                 global $wikiLowerChars;
  41                 return preg_replace(
  42                   "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
  43                   "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
  44                   $string );
  45         }
  46
  47         function fallback8bitEncoding() {
  48                 # Windows codepage 1252 is a superset of iso 8859-1
  49                 # override this to use difference source encoding to
  50                 # translate incoming 8-bit URLs.
  51                 return "windows-1252";
  52         }
  53
  54         function checkTitleEncoding( $s ) {
  55                 global $wgInputEncoding;
  56
  57                 # Check for non-UTF-8 URLs
  58                 $ishigh = preg_match( '/[\x80-\xff]/', $s);
  59                 if(!$ishigh) return $s;
  60
  61                 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
  62                 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
  63                 if( $isutf8 ) return $s;
  64
  65                 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
  66         }
  67 }
  68
  69 } # ifdef MEDIAWIKI
  70
  71 ?>