languages/LanguageUtf8.php

   1 <?php
   2 /**
   3   * @package MediaWiki
   4   * @subpackage Language
   5   */
   6
   7 if( defined( "MEDIAWIKI" ) ) {
   8
   9 # This file and LanguageLatin1.php may be included from within functions, so
  10 # we need to have global statements
  11
  12 global $wgInputEncoding, $wgOutputEncoding, $wikiUpperChars, $wikiLowerChars;
  13 global $wgDBname, $wgMemc;
  14
  15 $wgInputEncoding    = "UTF-8";
  16 $wgOutputEncoding       = "UTF-8";
  17
  18 if( function_exists( 'mb_strtoupper' ) ) {
  19         mb_internal_encoding('UTF-8');
  20 } else {
  21         # Hack our own case conversion routines
  22
  23         # Loading serialized arrays is faster than parsing code :P
  24         $wikiUpperChars = $wgMemc->get( $key1 = "$wgDBname:utf8:upper" );
  25         $wikiLowerChars = $wgMemc->get( $key2 = "$wgDBname:utf8:lower" );
  26
  27         if(empty( $wikiUpperChars) || empty($wikiLowerChars )) {
  28                 require_once( "includes/Utf8Case.php" );
  29                 $wgMemc->set( $key1, $wikiUpperChars );
  30                 $wgMemc->set( $key2, $wikiLowerChars );
  31         }
  32 }
  33
  34 /**
  35  * Base stuff useful to all UTF-8 based language files
  36  * @package MediaWiki
  37  */
  38 class LanguageUtf8 extends Language {
  39
  40         # These functions use mbstring library, if it is loaded
  41         # or compiled and character mapping arrays otherwise.
  42         # In case of language-specific character mismatch
  43         # it should be dealt with in Language classes.
  44
  45         function ucfirst( $str ) {
  46                 return $this->uc( $str, true );
  47         }
  48
  49         function uc( $str, $first = false ) {
  50                 if ( function_exists( 'mb_strtoupper' ) )
  51                         return $first ? mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 ) : mb_strtoupper( $str );
  52                 else {
  53                         global $wikiUpperChars;
  54                         $x = $first ? '^' : '';
  55                         return preg_replace(
  56                                 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/e",
  57                                 "strtr( \"\$1\" , \$wikiUpperChars )",
  58                                 $str
  59                         );
  60                 }
  61         }
  62
  63         function lcfirst( $str ) {
  64                 return $this->lc( $str, true );
  65         }
  66
  67         function lc( $str, $first = false ) {
  68                 if ( function_exists( 'mb_strtolower' ) )
  69                         return $first ? mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 ) : mb_strtolower( $str );
  70                 else {
  71                         global $wikiLowerChars;
  72                         $x = $first ? '^' : '';
  73                         return preg_replace(
  74                                 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/e",
  75                                 "strtr( \"\$1\" , \$wikiLowerChars )",
  76                                 $str
  77                         );
  78                 }
  79         }
  80
  81         function stripForSearch( $string ) {
  82                 # MySQL fulltext index doesn't grok utf-8, so we
  83                 # need to fold cases and convert to hex
  84
  85                 # In Language:: it just returns lowercase, maybe
  86                 # all strtolower on stripped output or argument
  87                 # should be removed and all stripForSearch
  88                 # methods adjusted to that.
  89
  90                 wfProfileIn( "LanguageUtf8::stripForSearch" );
  91                 if( function_exists( 'mb_strtolower' ) ) {
  92                         $out = preg_replace(
  93                                 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
  94                                 "'U8' . bin2hex( \"$1\" )",
  95                                 mb_strtolower( $string ) );
  96                 } else {
  97                         global $wikiLowerChars;
  98                         $out = preg_replace(
  99                                 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
 100                                 "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
 101                                 $string );
 102                 }
 103                 wfProfileOut( "LanguageUtf8::stripForSearch" );
 104                 return $out;
 105         }
 106
 107         function fallback8bitEncoding() {
 108                 # Windows codepage 1252 is a superset of iso 8859-1
 109                 # override this to use difference source encoding to
 110                 # translate incoming 8-bit URLs.
 111                 return "windows-1252";
 112         }
 113
 114         function checkTitleEncoding( $s ) {
 115                 global $wgInputEncoding;
 116
 117                 if( is_array( $s ) ) {
 118                         wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
 119                 }
 120                 # Check for non-UTF-8 URLs
 121                 $ishigh = preg_match( '/[\x80-\xff]/', $s);
 122                 if(!$ishigh) return $s;
 123
 124                 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
 125                 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
 126                 if( $isutf8 ) return $s;
 127
 128                 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
 129         }
 130
 131         function firstChar( $s ) {
 132                 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
 133                 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
 134
 135                 return isset( $matches[1] ) ? $matches[1] : "";
 136         }
 137
 138         # Crop a string from the beginning or end to a certain number of bytes.
 139         # (Bytes are used because our storage has limited byte lengths for some
 140         # columns in the database.) Multibyte charsets will need to make sure that
 141         # only whole characters are included!
 142         #
 143         # $length does not include the optional ellipsis.
 144         # If $length is negative, snip from the beginning
 145         function truncate( $string, $length, $ellipsis = "" ) {
 146                 if( $length == 0 ) {
 147                         return $ellipsis;
 148                 }
 149                 if ( strlen( $string ) <= abs( $length ) ) {
 150                         return $string;
 151                 }
 152                 if( $length > 0 ) {
 153                         $string = substr( $string, 0, $length );
 154                         $char = ord( $string[strlen( $string ) - 1] );
 155                         if ($char >= 0xc0) {
 156                                 # We got the first byte only of a multibyte char; remove it.
 157                                 $string = substr( $string, 0, -1 );
 158                         } elseif( $char >= 0x80 &&
 159                                   preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
 160                                               '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
 161                             # We chopped in the middle of a character; remove it
 162                                 $string = $m[1];
 163                         }
 164                         return $string . $ellipsis;
 165                 } else {
 166                         $string = substr( $string, $length );
 167                         $char = ord( $string[0] );
 168                         if( $char >= 0x80 && $char < 0xc0 ) {
 169                                 # We chopped in the middle of a character; remove the whole thing
 170                                 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
 171                         }
 172                         return $ellipsis . $string;
 173                 }
 174         }
 175 }
 176
 177 } # ifdef MEDIAWIKI
 178
 179 ?>