languages/LanguageUtf8.php

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