maintenance/language/generateNormalizerDataAr.php

   1 <?php
   2 /**
   3  * Generates the normalizer data file for Arabic.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup MaintenanceLanguage
  22  */
  23
  24 require_once __DIR__ . '/../Maintenance.php';
  25
  26 /**
  27  * Generates the normalizer data file for Arabic.
  28  *
  29  * This data file is used after normalizing to NFC.
  30  *
  31  * Example usage:
  32  *
  33  *    curl 'https://unicode.org/Public/6.0.0/ucd/UnicodeData.txt' > /tmp/UnicodeData.txt
  34  *    php generateNormalizerDataAr.php --unicode-data-file /tmp/UnicodeData.txt
  35  *
  36  * @ingroup MaintenanceLanguage
  37  */
  38 class GenerateNormalizerDataAr extends Maintenance {
  39         public function __construct() {
  40                 parent::__construct();
  41                 $this->addDescription( 'Generate the normalizer data file for Arabic' );
  42                 $this->addOption( 'unicode-data-file', 'The local location of the data file ' .
  43                         'from https://unicode.org/Public/6.0.0/ucd/UnicodeData.txt', false, true );
  44         }
  45
  46         public function getDbType() {
  47                 return Maintenance::DB_NONE;
  48         }
  49
  50         public function execute() {
  51                 if ( !$this->hasOption( 'unicode-data-file' ) ) {
  52                         $dataFile = 'UnicodeData.txt';
  53                         if ( !file_exists( $dataFile ) ) {
  54                                 $this->fatalError( "Unable to find UnicodeData.txt. Please specify " .
  55                                         "its location with --unicode-data-file=<FILE>" );
  56                         }
  57                 } else {
  58                         $dataFile = $this->getOption( 'unicode-data-file' );
  59                         if ( !file_exists( $dataFile ) ) {
  60                                 $this->fatalError( 'Unable to find the specified data file.' );
  61                         }
  62                 }
  63
  64                 $file = fopen( $dataFile, 'r' );
  65                 if ( !$file ) {
  66                         $this->fatalError( 'Unable to open the data file.' );
  67                 }
  68
  69                 // For the file format, see https://www.unicode.org/reports/tr44/
  70                 $fieldNames = [
  71                         'Code',
  72                         'Name',
  73                         'General_Category',
  74                         'Canonical_Combining_Class',
  75                         'Bidi_Class',
  76                         'Decomposition_Type_Mapping',
  77                         'Numeric_Type_Value_6',
  78                         'Numeric_Type_Value_7',
  79                         'Numeric_Type_Value_8',
  80                         'Bidi_Mirrored',
  81                         'Unicode_1_Name',
  82                         'ISO_Comment',
  83                         'Simple_Uppercase_Mapping',
  84                         'Simple_Lowercase_Mapping',
  85                         'Simple_Titlecase_Mapping'
  86                 ];
  87
  88                 $pairs = [];
  89
  90                 $lineNum = 0;
  91                 while ( false !== ( $line = fgets( $file ) ) ) {
  92                         ++$lineNum;
  93
  94                         # Strip comments
  95                         $line = trim( substr( $line, 0, strcspn( $line, '#' ) ) );
  96                         if ( $line === '' ) {
  97                                 continue;
  98                         }
  99
 100                         # Split fields
 101                         $numberedData = explode( ';', $line );
 102                         $data = [];
 103                         foreach ( $fieldNames as $number => $name ) {
 104                                 $data[$name] = $numberedData[$number];
 105                         }
 106
 107                         $code = base_convert( $data['Code'], 16, 10 );
 108                         if ( ( $code >= 0xFB50 && $code <= 0xFDFF ) # Arabic presentation forms A
 109                                 || ( $code >= 0xFE70 && $code <= 0xFEFF ) # Arabic presentation forms B
 110                         ) {
 111                                 if ( $data['Decomposition_Type_Mapping'] === '' ) {
 112                                         // No decomposition
 113                                         continue;
 114                                 }
 115                                 if ( !preg_match( '/^ *(<\w*>) +([0-9A-F ]*)$/',
 116                                         $data['Decomposition_Type_Mapping'], $m )
 117                                 ) {
 118                                         $this->error( "Can't parse Decomposition_Type/Mapping on line $lineNum" );
 119                                         $this->error( $line );
 120                                         continue;
 121                                 }
 122
 123                                 $source = UtfNormal\Utils::hexSequenceToUtf8( $data['Code'] );
 124                                 $dest = UtfNormal\Utils::hexSequenceToUtf8( $m[2] );
 125                                 $pairs[$source] = $dest;
 126                         }
 127                 }
 128
 129                 global $IP;
 130                 $writer = new StaticArrayWriter();
 131                 file_put_contents( "$IP/languages/data/normalize-ar.php", $writer->create(
 132                         $pairs,
 133                         'File created by generateNormalizerDataAr.php'
 134                 ) );
 135
 136                 echo "ar: " . count( $pairs ) . " pairs written.\n";
 137         }
 138 }
 139
 140 $maintClass = GenerateNormalizerDataAr::class;
 141 require_once RUN_MAINTENANCE_IF_MAIN;