Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
84.96% |
113 / 133 |
|
0.00% |
0 / 3 |
CRAP | |
0.00% |
0 / 1 |
| CLDRParser | |
84.96% |
113 / 133 |
|
0.00% |
0 / 3 |
66.66 | |
0.00% |
0 / 1 |
| parseMain | |
78.69% |
48 / 61 |
|
0.00% |
0 / 1 |
26.68 | |||
| parseSupplemental | |
92.31% |
24 / 26 |
|
0.00% |
0 / 1 |
11.06 | |||
| parseCurrencySymbols | |
89.13% |
41 / 46 |
|
0.00% |
0 / 1 |
23.68 | |||
| 1 | <?php |
| 2 | |
| 3 | namespace MediaWiki\Extension\CLDR; |
| 4 | |
| 5 | use RuntimeException; |
| 6 | use SimpleXMLElement; |
| 7 | |
| 8 | /** |
| 9 | * Extract data from cldr XML. |
| 10 | * |
| 11 | * @author Niklas Laxström |
| 12 | * @author Ryan Kaldari |
| 13 | * @author Santhosh Thottingal |
| 14 | * @author Sam Reed |
| 15 | * @copyright Copyright © 2007-2015 |
| 16 | * @license GPL-2.0-or-later |
| 17 | */ |
| 18 | class CLDRParser { |
| 19 | |
| 20 | public const LOCALITY_DEFAULT = '!DEFAULT'; |
| 21 | public const LANGUAGE_DEFAULT = '!root'; |
| 22 | public const CURRENCY_DEFAULT = '!DEFAULT'; |
| 23 | |
| 24 | /** |
| 25 | * Read the main/<locale>.xml file from CLDR core and convert to PHP |
| 26 | * |
| 27 | * @param string $inputFile filename |
| 28 | */ |
| 29 | public function parseMain( $inputFile ): array { |
| 30 | $contents = file_get_contents( $inputFile ); |
| 31 | $doc = new SimpleXMLElement( $contents ); |
| 32 | |
| 33 | $data = [ |
| 34 | 'indexCharacters' => [], |
| 35 | 'languageNames' => [], |
| 36 | 'currencyNames' => [], |
| 37 | 'currencySymbols' => [], |
| 38 | 'countryNames' => [], |
| 39 | 'timeUnits' => [], |
| 40 | ]; |
| 41 | |
| 42 | // Take a Unicode Set for an alphabet and extract simple example characters. |
| 43 | // For example, "[aàâ b {ch}]" is extracted as `["a", "b", "ch"]`. |
| 44 | // TODO: Unicode Set allows for more complex syntax, but we support only |
| 45 | // the subset currently used here. Should rely on a library instead. |
| 46 | $indexCharacters = $doc->xpath( '//characters/exemplarCharacters[@type="index"]' ); |
| 47 | if ( $indexCharacters && count( $indexCharacters ) === 1 ) { |
| 48 | [ $characters ] = $indexCharacters; |
| 49 | $splitSequence = preg_split( '/\s/', |
| 50 | trim( (string)$characters, '[]' ) ); |
| 51 | $data['indexCharacters'] = array_map( |
| 52 | static fn ( $letter ) => preg_replace_callback_array( [ |
| 53 | // Convert unicode literals to characters. |
| 54 | '/^\\\\u([\da-f]{4})/i' => static fn ( $m ) => mb_chr( hexdec( $m[1] ) ), |
| 55 | |
| 56 | // Take only the first character from a set like "aàâ". |
| 57 | // When the character is made up of multiple symbols, it |
| 58 | // will be enclosed in curly braces like "{ch}", and in this |
| 59 | // case we want the entire group. It's possible that the |
| 60 | // two cases are combined like "{ch}ç". |
| 61 | '/^(?:([^{])|\{([^}]+)\}).*$/u' => static fn ( $m ) => $m[2] ?? $m[1], |
| 62 | ], $letter ), |
| 63 | $splitSequence |
| 64 | ); |
| 65 | } |
| 66 | |
| 67 | foreach ( $doc->xpath( '//languages/language' ) as $elem ) { |
| 68 | if ( (string)$elem['alt'] !== '' ) { |
| 69 | continue; |
| 70 | } |
| 71 | |
| 72 | if ( (string)$elem['type'] === 'root' ) { |
| 73 | continue; |
| 74 | } |
| 75 | |
| 76 | $key = str_replace( '_', '-', strtolower( $elem['type'] ) ); |
| 77 | |
| 78 | $data['languageNames'][$key] = (string)$elem; |
| 79 | } |
| 80 | |
| 81 | foreach ( $doc->xpath( '//currencies/currency' ) as $elem ) { |
| 82 | if ( (string)$elem->displayName[0] === '' ) { |
| 83 | continue; |
| 84 | } |
| 85 | |
| 86 | $data['currencyNames'][(string)$elem['type']] = (string)$elem->displayName[0]; |
| 87 | if ( (string)$elem->symbol[0] !== '' ) { |
| 88 | $data['currencySymbols'][(string)$elem['type']] = (string)$elem->symbol[0]; |
| 89 | } |
| 90 | } |
| 91 | |
| 92 | foreach ( $doc->xpath( '//territories/territory' ) as $elem ) { |
| 93 | if ( (string)$elem['alt'] !== '' && (string)$elem['alt'] !== 'short' ) { |
| 94 | continue; |
| 95 | } |
| 96 | |
| 97 | if ( (string)$elem['type'] === 'ZZ' || |
| 98 | !preg_match( '/^[A-Z][A-Z]$/', $elem['type'] ) |
| 99 | ) { |
| 100 | continue; |
| 101 | } |
| 102 | |
| 103 | $data['countryNames'][(string)$elem['type']] = (string)$elem; |
| 104 | } |
| 105 | foreach ( $doc->xpath( '//units/unitLength' ) as $unitLength ) { |
| 106 | if ( (string)$unitLength['type'] !== 'long' ) { |
| 107 | continue; |
| 108 | } |
| 109 | foreach ( $unitLength->unit as $elem ) { |
| 110 | $type = (string)$elem['type']; |
| 111 | if ( !str_starts_with( $type, 'duration-' ) ) { |
| 112 | continue; |
| 113 | } |
| 114 | $type = substr( $type, strlen( 'duration-' ) ); |
| 115 | foreach ( $elem->unitPattern as $pattern ) { |
| 116 | $data['timeUnits'][$type . '-' . (string)$pattern['count']] = (string)$pattern; |
| 117 | } |
| 118 | } |
| 119 | } |
| 120 | foreach ( $doc->xpath( '//fields/field' ) as $field ) { |
| 121 | $fieldType = (string)$field['type']; |
| 122 | |
| 123 | foreach ( $field->relativeTime as $relative ) { |
| 124 | $type = (string)$relative['type']; |
| 125 | foreach ( $relative->relativeTimePattern as $pattern ) { |
| 126 | $data['timeUnits'][$fieldType . '-' . $type |
| 127 | . '-' . (string)$pattern['count']] = (string)$pattern; |
| 128 | } |
| 129 | } |
| 130 | } |
| 131 | |
| 132 | ksort( $data['timeUnits'] ); |
| 133 | return $data; |
| 134 | } |
| 135 | |
| 136 | /** |
| 137 | * Parse method for the file structure found in common/supplemental/supplementalData.xml |
| 138 | * @param string $inputFile |
| 139 | */ |
| 140 | public function parseSupplemental( $inputFile ): array { |
| 141 | // Open the input file for reading |
| 142 | |
| 143 | $contents = file_get_contents( $inputFile ); |
| 144 | $doc = new SimpleXMLElement( $contents ); |
| 145 | |
| 146 | $data = [ |
| 147 | 'currencyFractions' => [], |
| 148 | 'localeCurrencies' => [], |
| 149 | ]; |
| 150 | |
| 151 | // Pull currency attributes - digits, rounding, and cashRounding. |
| 152 | // This will tell us how many decmal places make sense to use with any currency, |
| 153 | // or if the currency is totally non-fractional |
| 154 | foreach ( $doc->xpath( '//currencyData/fractions/info' ) as $elem ) { |
| 155 | $iso4217 = (string)$elem['iso4217']; |
| 156 | if ( $iso4217 === '' ) { |
| 157 | continue; |
| 158 | } |
| 159 | if ( $iso4217 === 'DEFAULT' ) { |
| 160 | $iso4217 = self::CURRENCY_DEFAULT; |
| 161 | } |
| 162 | |
| 163 | $attributes = [ 'digits', 'rounding', 'cashDigits', 'cashRounding' ]; |
| 164 | foreach ( $attributes as $att ) { |
| 165 | if ( (string)$elem[$att] !== '' ) { |
| 166 | $data['currencyFractions'][$iso4217][$att] = (string)$elem[$att]; |
| 167 | } |
| 168 | } |
| 169 | } |
| 170 | |
| 171 | ksort( $data['currencyFractions'] ); |
| 172 | |
| 173 | // Pull a map of regions to currencies in order of preference. |
| 174 | foreach ( $doc->xpath( '//currencyData/region' ) as $elem ) { |
| 175 | if ( (string)$elem['iso3166'] === '' ) { |
| 176 | continue; |
| 177 | } |
| 178 | |
| 179 | $region = (string)$elem['iso3166']; |
| 180 | |
| 181 | foreach ( $elem->currency as $currencynode ) { |
| 182 | if ( (string)$currencynode['to'] === '' && (string)$currencynode['tender'] !== 'false' ) { |
| 183 | $data['localeCurrencies'][$region][] = (string)$currencynode['iso4217']; |
| 184 | } |
| 185 | } |
| 186 | } |
| 187 | |
| 188 | ksort( $data['localeCurrencies'] ); |
| 189 | return $data; |
| 190 | } |
| 191 | |
| 192 | /** |
| 193 | * Parse method for the currency section in the names files. |
| 194 | * This is separate from the regular parse function, because we need all of |
| 195 | * the currency locale information, even if mediawiki doesn't support the language. |
| 196 | * (For instance: en_AU uses '$' for AUD, not USD, but it's not a supported mediawiki locality) |
| 197 | * @param string $inputDir the directory, in which we will parse everything. |
| 198 | */ |
| 199 | public function parseCurrencySymbols( $inputDir ): array { |
| 200 | if ( !file_exists( $inputDir ) ) { |
| 201 | throw new RuntimeException( 'Input directory not found.' ); |
| 202 | } |
| 203 | $files = scandir( $inputDir ); |
| 204 | |
| 205 | $data = [ |
| 206 | 'currencySymbols' => [], |
| 207 | ]; |
| 208 | |
| 209 | // Foreach files! |
| 210 | foreach ( $files as $inputFile ) { |
| 211 | if ( !str_ends_with( $inputFile, '.xml' ) ) { |
| 212 | continue; |
| 213 | } |
| 214 | |
| 215 | $contents = file_get_contents( $inputDir . '/' . $inputFile ); |
| 216 | $doc = new SimpleXMLElement( $contents ); |
| 217 | |
| 218 | // Tags in the <identity> section are guaranteed to appear once |
| 219 | $languages = $doc->xpath( '//identity/language/@type' ); |
| 220 | $language = $languages |
| 221 | ? (string)$languages[0] |
| 222 | : pathinfo( $inputFile, PATHINFO_FILENAME ); |
| 223 | |
| 224 | // The <script> element is optional |
| 225 | $scripts = $doc->xpath( '//identity/script/@type' ); |
| 226 | $script = $scripts ? (string)$scripts[0] : ''; |
| 227 | // expand the language |
| 228 | if ( $script !== '' ) { |
| 229 | $language .= '-' . strtolower( $script ); |
| 230 | } |
| 231 | |
| 232 | // The <territory> element is optional |
| 233 | $territories = $doc->xpath( '//identity/territory/@type' ); |
| 234 | $territory = $territories ? (string)$territories[0] : self::LOCALITY_DEFAULT; |
| 235 | |
| 236 | if ( $language === 'root' ) { |
| 237 | $language = self::LANGUAGE_DEFAULT; |
| 238 | } |
| 239 | |
| 240 | foreach ( $doc->xpath( '//currencies/currency' ) as $elem ) { |
| 241 | if ( (string)$elem->symbol[0] !== '' ) { |
| 242 | $data['currencySymbols'][(string)$elem['type']][$language][$territory] = |
| 243 | (string)$elem->symbol[0]; |
| 244 | } |
| 245 | } |
| 246 | } |
| 247 | |
| 248 | // now massage the data somewhat. It's pretty blown up at this point. |
| 249 | |
| 250 | /** |
| 251 | * Part 1: Stop blowing up on defaults. |
| 252 | * Defaults apparently come in many forms. Listed below in order of scope |
| 253 | * (widest to narrowest) |
| 254 | * 1) The ISO code itself, in the absence of any other defaults |
| 255 | * 2) The 'root' language file definition |
| 256 | * 3) Language with no locality - locality will come in as 'DEFAULT' |
| 257 | * |
| 258 | * Intended behavior: |
| 259 | * From narrowest scope to widest, collapse the defaults |
| 260 | */ |
| 261 | foreach ( $data['currencySymbols'] as $currency => $language ) { |
| 262 | // get the currency default symbol. This will either be defined in the |
| 263 | // 'root' language file, or taken from the ISO code. |
| 264 | $default = $language[self::LANGUAGE_DEFAULT][self::LOCALITY_DEFAULT] ?? $currency; |
| 265 | |
| 266 | foreach ( $language as $lang => $territories ) { |
| 267 | if ( is_array( $territories ) ) { |
| 268 | // Collapse a language (no locality) array if it's just the default. One value will do fine. |
| 269 | if ( count( $territories ) === 1 && array_key_exists( self::LOCALITY_DEFAULT, $territories ) ) { |
| 270 | $data['currencySymbols'][$currency][$lang] = $territories[self::LOCALITY_DEFAULT]; |
| 271 | if ( $territories[self::LOCALITY_DEFAULT] === $default |
| 272 | && $lang !== self::LANGUAGE_DEFAULT |
| 273 | ) { |
| 274 | unset( $data['currencySymbols'][$currency][$lang] ); |
| 275 | } |
| 276 | } else { |
| 277 | // Collapse a language (with locality) array if it's default is just the default |
| 278 | if ( !array_key_exists( self::LOCALITY_DEFAULT, $territories ) |
| 279 | || ( $territories[self::LOCALITY_DEFAULT] === $default |
| 280 | && $lang !== self::LANGUAGE_DEFAULT ) |
| 281 | ) { |
| 282 | foreach ( $territories as $territory => $symbol ) { |
| 283 | if ( $symbol === $default ) { |
| 284 | unset( $data['currencySymbols'][$currency][$lang][$territory] ); |
| 285 | } |
| 286 | } |
| 287 | } |
| 288 | ksort( $data['currencySymbols'][$currency][$lang] ); |
| 289 | } |
| 290 | } |
| 291 | } |
| 292 | |
| 293 | ksort( $data['currencySymbols'][$currency] ); |
| 294 | } |
| 295 | |
| 296 | ksort( $data['currencySymbols'] ); |
| 297 | return $data; |
| 298 | } |
| 299 | |
| 300 | } |