lib/localename.c

   1 /* Determine name of the currently selected locale.
   2    Copyright (C) 1995-2018 Free Software Foundation, Inc.
   3
   4    This program is free software: you can redistribute it and/or modify
   5    it under the terms of the GNU Lesser General Public License as published by
   6    the Free Software Foundation; either version 2.1 of the License, or
   7    (at your option) any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU Lesser General Public License for more details.
  13
  14    You should have received a copy of the GNU Lesser General Public License
  15    along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
  16
  17 /* Written by Ulrich Drepper <drepper@gnu.org>, 1995.  */
  18 /* Native Windows code written by Tor Lillqvist <tml@iki.fi>.  */
  19 /* Mac OS X code written by Bruno Haible <bruno@clisp.org>.  */
  20
  21 #include <config.h>
  22
  23 /* Specification.  */
  24 #ifdef IN_LIBINTL
  25 # include "gettextP.h"
  26 #else
  27 # include "localename.h"
  28 #endif
  29
  30 #include <limits.h>
  31 #include <stddef.h>
  32 #include <stdlib.h>
  33 #include <locale.h>
  34 #include <string.h>
  35
  36 #include "flexmember.h"
  37
  38 #if HAVE_USELOCALE
  39 /* Mac OS X 10.5 defines the locale_t type in <xlocale.h>.  */
  40 # if defined __APPLE__ && defined __MACH__
  41 #  include <xlocale.h>
  42 # endif
  43 # if (__GLIBC__ >= 2 && !defined __UCLIBC__) || (defined __linux__ && HAVE_LANGINFO_H) || defined __CYGWIN__
  44 #  include <langinfo.h>
  45 # endif
  46 # if !defined IN_LIBINTL
  47 #  include "glthread/lock.h"
  48 # endif
  49 # if defined __sun
  50 #  if HAVE_GETLOCALENAME_L
  51 /* Solaris >= 12.  */
  52 extern char * getlocalename_l(int, locale_t);
  53 #  elif HAVE_SOLARIS114_LOCALES
  54 #   include <sys/localedef.h>
  55 #  endif
  56 # endif
  57 # if HAVE_NAMELESS_LOCALES
  58 #  include <errno.h>
  59 #  include "localename-table.h"
  60 # endif
  61 #endif
  62
  63 #if HAVE_CFLOCALECOPYCURRENT || HAVE_CFPREFERENCESCOPYAPPVALUE
  64 # include <CoreFoundation/CFString.h>
  65 # if HAVE_CFLOCALECOPYCURRENT
  66 #  include <CoreFoundation/CFLocale.h>
  67 # elif HAVE_CFPREFERENCESCOPYAPPVALUE
  68 #  include <CoreFoundation/CFPreferences.h>
  69 # endif
  70 #endif
  71
  72 #if defined _WIN32 && !defined __CYGWIN__
  73 # define WINDOWS_NATIVE
  74 # if !defined IN_LIBINTL
  75 #  include "glthread/lock.h"
  76 # endif
  77 #endif
  78
  79 #if defined WINDOWS_NATIVE || defined __CYGWIN__ /* Native Windows or Cygwin */
  80 # define WIN32_LEAN_AND_MEAN
  81 # include <windows.h>
  82 # include <winnls.h>
  83 /* List of language codes, sorted by value:
  84    0x01 LANG_ARABIC
  85    0x02 LANG_BULGARIAN
  86    0x03 LANG_CATALAN
  87    0x04 LANG_CHINESE
  88    0x05 LANG_CZECH
  89    0x06 LANG_DANISH
  90    0x07 LANG_GERMAN
  91    0x08 LANG_GREEK
  92    0x09 LANG_ENGLISH
  93    0x0a LANG_SPANISH
  94    0x0b LANG_FINNISH
  95    0x0c LANG_FRENCH
  96    0x0d LANG_HEBREW
  97    0x0e LANG_HUNGARIAN
  98    0x0f LANG_ICELANDIC
  99    0x10 LANG_ITALIAN
 100    0x11 LANG_JAPANESE
 101    0x12 LANG_KOREAN
 102    0x13 LANG_DUTCH
 103    0x14 LANG_NORWEGIAN
 104    0x15 LANG_POLISH
 105    0x16 LANG_PORTUGUESE
 106    0x17 LANG_ROMANSH
 107    0x18 LANG_ROMANIAN
 108    0x19 LANG_RUSSIAN
 109    0x1a LANG_CROATIAN == LANG_SERBIAN
 110    0x1b LANG_SLOVAK
 111    0x1c LANG_ALBANIAN
 112    0x1d LANG_SWEDISH
 113    0x1e LANG_THAI
 114    0x1f LANG_TURKISH
 115    0x20 LANG_URDU
 116    0x21 LANG_INDONESIAN
 117    0x22 LANG_UKRAINIAN
 118    0x23 LANG_BELARUSIAN
 119    0x24 LANG_SLOVENIAN
 120    0x25 LANG_ESTONIAN
 121    0x26 LANG_LATVIAN
 122    0x27 LANG_LITHUANIAN
 123    0x28 LANG_TAJIK
 124    0x29 LANG_FARSI
 125    0x2a LANG_VIETNAMESE
 126    0x2b LANG_ARMENIAN
 127    0x2c LANG_AZERI
 128    0x2d LANG_BASQUE
 129    0x2e LANG_SORBIAN
 130    0x2f LANG_MACEDONIAN
 131    0x30 LANG_SUTU
 132    0x31 LANG_TSONGA
 133    0x32 LANG_TSWANA
 134    0x33 LANG_VENDA
 135    0x34 LANG_XHOSA
 136    0x35 LANG_ZULU
 137    0x36 LANG_AFRIKAANS
 138    0x37 LANG_GEORGIAN
 139    0x38 LANG_FAEROESE
 140    0x39 LANG_HINDI
 141    0x3a LANG_MALTESE
 142    0x3b LANG_SAMI
 143    0x3c LANG_GAELIC
 144    0x3d LANG_YIDDISH
 145    0x3e LANG_MALAY
 146    0x3f LANG_KAZAK
 147    0x40 LANG_KYRGYZ
 148    0x41 LANG_SWAHILI
 149    0x42 LANG_TURKMEN
 150    0x43 LANG_UZBEK
 151    0x44 LANG_TATAR
 152    0x45 LANG_BENGALI
 153    0x46 LANG_PUNJABI
 154    0x47 LANG_GUJARATI
 155    0x48 LANG_ORIYA
 156    0x49 LANG_TAMIL
 157    0x4a LANG_TELUGU
 158    0x4b LANG_KANNADA
 159    0x4c LANG_MALAYALAM
 160    0x4d LANG_ASSAMESE
 161    0x4e LANG_MARATHI
 162    0x4f LANG_SANSKRIT
 163    0x50 LANG_MONGOLIAN
 164    0x51 LANG_TIBETAN
 165    0x52 LANG_WELSH
 166    0x53 LANG_CAMBODIAN
 167    0x54 LANG_LAO
 168    0x55 LANG_BURMESE
 169    0x56 LANG_GALICIAN
 170    0x57 LANG_KONKANI
 171    0x58 LANG_MANIPURI
 172    0x59 LANG_SINDHI
 173    0x5a LANG_SYRIAC
 174    0x5b LANG_SINHALESE
 175    0x5c LANG_CHEROKEE
 176    0x5d LANG_INUKTITUT
 177    0x5e LANG_AMHARIC
 178    0x5f LANG_TAMAZIGHT
 179    0x60 LANG_KASHMIRI
 180    0x61 LANG_NEPALI
 181    0x62 LANG_FRISIAN
 182    0x63 LANG_PASHTO
 183    0x64 LANG_TAGALOG
 184    0x65 LANG_DIVEHI
 185    0x66 LANG_EDO
 186    0x67 LANG_FULFULDE
 187    0x68 LANG_HAUSA
 188    0x69 LANG_IBIBIO
 189    0x6a LANG_YORUBA
 190    0x6d LANG_BASHKIR
 191    0x6e LANG_LUXEMBOURGISH
 192    0x6f LANG_GREENLANDIC
 193    0x70 LANG_IGBO
 194    0x71 LANG_KANURI
 195    0x72 LANG_OROMO
 196    0x73 LANG_TIGRINYA
 197    0x74 LANG_GUARANI
 198    0x75 LANG_HAWAIIAN
 199    0x76 LANG_LATIN
 200    0x77 LANG_SOMALI
 201    0x78 LANG_YI
 202    0x79 LANG_PAPIAMENTU
 203    0x7a LANG_MAPUDUNGUN
 204    0x7c LANG_MOHAWK
 205    0x7e LANG_BRETON
 206    0x82 LANG_OCCITAN
 207    0x83 LANG_CORSICAN
 208    0x84 LANG_ALSATIAN
 209    0x85 LANG_YAKUT
 210    0x86 LANG_KICHE
 211    0x87 LANG_KINYARWANDA
 212    0x88 LANG_WOLOF
 213    0x8c LANG_DARI
 214    0x91 LANG_SCOTTISH_GAELIC
 215 */
 216 /* Mingw headers don't have latest language and sublanguage codes.  */
 217 # ifndef LANG_AFRIKAANS
 218 # define LANG_AFRIKAANS 0x36
 219 # endif
 220 # ifndef LANG_ALBANIAN
 221 # define LANG_ALBANIAN 0x1c
 222 # endif
 223 # ifndef LANG_ALSATIAN
 224 # define LANG_ALSATIAN 0x84
 225 # endif
 226 # ifndef LANG_AMHARIC
 227 # define LANG_AMHARIC 0x5e
 228 # endif
 229 # ifndef LANG_ARABIC
 230 # define LANG_ARABIC 0x01
 231 # endif
 232 # ifndef LANG_ARMENIAN
 233 # define LANG_ARMENIAN 0x2b
 234 # endif
 235 # ifndef LANG_ASSAMESE
 236 # define LANG_ASSAMESE 0x4d
 237 # endif
 238 # ifndef LANG_AZERI
 239 # define LANG_AZERI 0x2c
 240 # endif
 241 # ifndef LANG_BASHKIR
 242 # define LANG_BASHKIR 0x6d
 243 # endif
 244 # ifndef LANG_BASQUE
 245 # define LANG_BASQUE 0x2d
 246 # endif
 247 # ifndef LANG_BELARUSIAN
 248 # define LANG_BELARUSIAN 0x23
 249 # endif
 250 # ifndef LANG_BENGALI
 251 # define LANG_BENGALI 0x45
 252 # endif
 253 # ifndef LANG_BRETON
 254 # define LANG_BRETON 0x7e
 255 # endif
 256 # ifndef LANG_BURMESE
 257 # define LANG_BURMESE 0x55
 258 # endif
 259 # ifndef LANG_CAMBODIAN
 260 # define LANG_CAMBODIAN 0x53
 261 # endif
 262 # ifndef LANG_CATALAN
 263 # define LANG_CATALAN 0x03
 264 # endif
 265 # ifndef LANG_CHEROKEE
 266 # define LANG_CHEROKEE 0x5c
 267 # endif
 268 # ifndef LANG_CORSICAN
 269 # define LANG_CORSICAN 0x83
 270 # endif
 271 # ifndef LANG_DARI
 272 # define LANG_DARI 0x8c
 273 # endif
 274 # ifndef LANG_DIVEHI
 275 # define LANG_DIVEHI 0x65
 276 # endif
 277 # ifndef LANG_EDO
 278 # define LANG_EDO 0x66
 279 # endif
 280 # ifndef LANG_ESTONIAN
 281 # define LANG_ESTONIAN 0x25
 282 # endif
 283 # ifndef LANG_FAEROESE
 284 # define LANG_FAEROESE 0x38
 285 # endif
 286 # ifndef LANG_FARSI
 287 # define LANG_FARSI 0x29
 288 # endif
 289 # ifndef LANG_FRISIAN
 290 # define LANG_FRISIAN 0x62
 291 # endif
 292 # ifndef LANG_FULFULDE
 293 # define LANG_FULFULDE 0x67
 294 # endif
 295 # ifndef LANG_GAELIC
 296 # define LANG_GAELIC 0x3c
 297 # endif
 298 # ifndef LANG_GALICIAN
 299 # define LANG_GALICIAN 0x56
 300 # endif
 301 # ifndef LANG_GEORGIAN
 302 # define LANG_GEORGIAN 0x37
 303 # endif
 304 # ifndef LANG_GREENLANDIC
 305 # define LANG_GREENLANDIC 0x6f
 306 # endif
 307 # ifndef LANG_GUARANI
 308 # define LANG_GUARANI 0x74
 309 # endif
 310 # ifndef LANG_GUJARATI
 311 # define LANG_GUJARATI 0x47
 312 # endif
 313 # ifndef LANG_HAUSA
 314 # define LANG_HAUSA 0x68
 315 # endif
 316 # ifndef LANG_HAWAIIAN
 317 # define LANG_HAWAIIAN 0x75
 318 # endif
 319 # ifndef LANG_HEBREW
 320 # define LANG_HEBREW 0x0d
 321 # endif
 322 # ifndef LANG_HINDI
 323 # define LANG_HINDI 0x39
 324 # endif
 325 # ifndef LANG_IBIBIO
 326 # define LANG_IBIBIO 0x69
 327 # endif
 328 # ifndef LANG_IGBO
 329 # define LANG_IGBO 0x70
 330 # endif
 331 # ifndef LANG_INDONESIAN
 332 # define LANG_INDONESIAN 0x21
 333 # endif
 334 # ifndef LANG_INUKTITUT
 335 # define LANG_INUKTITUT 0x5d
 336 # endif
 337 # ifndef LANG_KANNADA
 338 # define LANG_KANNADA 0x4b
 339 # endif
 340 # ifndef LANG_KANURI
 341 # define LANG_KANURI 0x71
 342 # endif
 343 # ifndef LANG_KASHMIRI
 344 # define LANG_KASHMIRI 0x60
 345 # endif
 346 # ifndef LANG_KAZAK
 347 # define LANG_KAZAK 0x3f
 348 # endif
 349 # ifndef LANG_KICHE
 350 # define LANG_KICHE 0x86
 351 # endif
 352 # ifndef LANG_KINYARWANDA
 353 # define LANG_KINYARWANDA 0x87
 354 # endif
 355 # ifndef LANG_KONKANI
 356 # define LANG_KONKANI 0x57
 357 # endif
 358 # ifndef LANG_KYRGYZ
 359 # define LANG_KYRGYZ 0x40
 360 # endif
 361 # ifndef LANG_LAO
 362 # define LANG_LAO 0x54
 363 # endif
 364 # ifndef LANG_LATIN
 365 # define LANG_LATIN 0x76
 366 # endif
 367 # ifndef LANG_LATVIAN
 368 # define LANG_LATVIAN 0x26
 369 # endif
 370 # ifndef LANG_LITHUANIAN
 371 # define LANG_LITHUANIAN 0x27
 372 # endif
 373 # ifndef LANG_LUXEMBOURGISH
 374 # define LANG_LUXEMBOURGISH 0x6e
 375 # endif
 376 # ifndef LANG_MACEDONIAN
 377 # define LANG_MACEDONIAN 0x2f
 378 # endif
 379 # ifndef LANG_MALAY
 380 # define LANG_MALAY 0x3e
 381 # endif
 382 # ifndef LANG_MALAYALAM
 383 # define LANG_MALAYALAM 0x4c
 384 # endif
 385 # ifndef LANG_MALTESE
 386 # define LANG_MALTESE 0x3a
 387 # endif
 388 # ifndef LANG_MANIPURI
 389 # define LANG_MANIPURI 0x58
 390 # endif
 391 # ifndef LANG_MAORI
 392 # define LANG_MAORI 0x81
 393 # endif
 394 # ifndef LANG_MAPUDUNGUN
 395 # define LANG_MAPUDUNGUN 0x7a
 396 # endif
 397 # ifndef LANG_MARATHI
 398 # define LANG_MARATHI 0x4e
 399 # endif
 400 # ifndef LANG_MOHAWK
 401 # define LANG_MOHAWK 0x7c
 402 # endif
 403 # ifndef LANG_MONGOLIAN
 404 # define LANG_MONGOLIAN 0x50
 405 # endif
 406 # ifndef LANG_NEPALI
 407 # define LANG_NEPALI 0x61
 408 # endif
 409 # ifndef LANG_OCCITAN
 410 # define LANG_OCCITAN 0x82
 411 # endif
 412 # ifndef LANG_ORIYA
 413 # define LANG_ORIYA 0x48
 414 # endif
 415 # ifndef LANG_OROMO
 416 # define LANG_OROMO 0x72
 417 # endif
 418 # ifndef LANG_PAPIAMENTU
 419 # define LANG_PAPIAMENTU 0x79
 420 # endif
 421 # ifndef LANG_PASHTO
 422 # define LANG_PASHTO 0x63
 423 # endif
 424 # ifndef LANG_PUNJABI
 425 # define LANG_PUNJABI 0x46
 426 # endif
 427 # ifndef LANG_QUECHUA
 428 # define LANG_QUECHUA 0x6b
 429 # endif
 430 # ifndef LANG_ROMANSH
 431 # define LANG_ROMANSH 0x17
 432 # endif
 433 # ifndef LANG_SAMI
 434 # define LANG_SAMI 0x3b
 435 # endif
 436 # ifndef LANG_SANSKRIT
 437 # define LANG_SANSKRIT 0x4f
 438 # endif
 439 # ifndef LANG_SCOTTISH_GAELIC
 440 # define LANG_SCOTTISH_GAELIC 0x91
 441 # endif
 442 # ifndef LANG_SERBIAN
 443 # define LANG_SERBIAN 0x1a
 444 # endif
 445 # ifndef LANG_SINDHI
 446 # define LANG_SINDHI 0x59
 447 # endif
 448 # ifndef LANG_SINHALESE
 449 # define LANG_SINHALESE 0x5b
 450 # endif
 451 # ifndef LANG_SLOVAK
 452 # define LANG_SLOVAK 0x1b
 453 # endif
 454 # ifndef LANG_SOMALI
 455 # define LANG_SOMALI 0x77
 456 # endif
 457 # ifndef LANG_SORBIAN
 458 # define LANG_SORBIAN 0x2e
 459 # endif
 460 # ifndef LANG_SOTHO
 461 # define LANG_SOTHO 0x6c
 462 # endif
 463 # ifndef LANG_SUTU
 464 # define LANG_SUTU 0x30
 465 # endif
 466 # ifndef LANG_SWAHILI
 467 # define LANG_SWAHILI 0x41
 468 # endif
 469 # ifndef LANG_SYRIAC
 470 # define LANG_SYRIAC 0x5a
 471 # endif
 472 # ifndef LANG_TAGALOG
 473 # define LANG_TAGALOG 0x64
 474 # endif
 475 # ifndef LANG_TAJIK
 476 # define LANG_TAJIK 0x28
 477 # endif
 478 # ifndef LANG_TAMAZIGHT
 479 # define LANG_TAMAZIGHT 0x5f
 480 # endif
 481 # ifndef LANG_TAMIL
 482 # define LANG_TAMIL 0x49
 483 # endif
 484 # ifndef LANG_TATAR
 485 # define LANG_TATAR 0x44
 486 # endif
 487 # ifndef LANG_TELUGU
 488 # define LANG_TELUGU 0x4a
 489 # endif
 490 # ifndef LANG_THAI
 491 # define LANG_THAI 0x1e
 492 # endif
 493 # ifndef LANG_TIBETAN
 494 # define LANG_TIBETAN 0x51
 495 # endif
 496 # ifndef LANG_TIGRINYA
 497 # define LANG_TIGRINYA 0x73
 498 # endif
 499 # ifndef LANG_TSONGA
 500 # define LANG_TSONGA 0x31
 501 # endif
 502 # ifndef LANG_TSWANA
 503 # define LANG_TSWANA 0x32
 504 # endif
 505 # ifndef LANG_TURKMEN
 506 # define LANG_TURKMEN 0x42
 507 # endif
 508 # ifndef LANG_UIGHUR
 509 # define LANG_UIGHUR 0x80
 510 # endif
 511 # ifndef LANG_UKRAINIAN
 512 # define LANG_UKRAINIAN 0x22
 513 # endif
 514 # ifndef LANG_URDU
 515 # define LANG_URDU 0x20
 516 # endif
 517 # ifndef LANG_UZBEK
 518 # define LANG_UZBEK 0x43
 519 # endif
 520 # ifndef LANG_VENDA
 521 # define LANG_VENDA 0x33
 522 # endif
 523 # ifndef LANG_VIETNAMESE
 524 # define LANG_VIETNAMESE 0x2a
 525 # endif
 526 # ifndef LANG_WELSH
 527 # define LANG_WELSH 0x52
 528 # endif
 529 # ifndef LANG_WOLOF
 530 # define LANG_WOLOF 0x88
 531 # endif
 532 # ifndef LANG_XHOSA
 533 # define LANG_XHOSA 0x34
 534 # endif
 535 # ifndef LANG_YAKUT
 536 # define LANG_YAKUT 0x85
 537 # endif
 538 # ifndef LANG_YI
 539 # define LANG_YI 0x78
 540 # endif
 541 # ifndef LANG_YIDDISH
 542 # define LANG_YIDDISH 0x3d
 543 # endif
 544 # ifndef LANG_YORUBA
 545 # define LANG_YORUBA 0x6a
 546 # endif
 547 # ifndef LANG_ZULU
 548 # define LANG_ZULU 0x35
 549 # endif
 550 # ifndef SUBLANG_AFRIKAANS_SOUTH_AFRICA
 551 # define SUBLANG_AFRIKAANS_SOUTH_AFRICA 0x01
 552 # endif
 553 # ifndef SUBLANG_ALBANIAN_ALBANIA
 554 # define SUBLANG_ALBANIAN_ALBANIA 0x01
 555 # endif
 556 # ifndef SUBLANG_ALSATIAN_FRANCE
 557 # define SUBLANG_ALSATIAN_FRANCE 0x01
 558 # endif
 559 # ifndef SUBLANG_AMHARIC_ETHIOPIA
 560 # define SUBLANG_AMHARIC_ETHIOPIA 0x01
 561 # endif
 562 # ifndef SUBLANG_ARABIC_SAUDI_ARABIA
 563 # define SUBLANG_ARABIC_SAUDI_ARABIA 0x01
 564 # endif
 565 # ifndef SUBLANG_ARABIC_IRAQ
 566 # define SUBLANG_ARABIC_IRAQ 0x02
 567 # endif
 568 # ifndef SUBLANG_ARABIC_EGYPT
 569 # define SUBLANG_ARABIC_EGYPT 0x03
 570 # endif
 571 # ifndef SUBLANG_ARABIC_LIBYA
 572 # define SUBLANG_ARABIC_LIBYA 0x04
 573 # endif
 574 # ifndef SUBLANG_ARABIC_ALGERIA
 575 # define SUBLANG_ARABIC_ALGERIA 0x05
 576 # endif
 577 # ifndef SUBLANG_ARABIC_MOROCCO
 578 # define SUBLANG_ARABIC_MOROCCO 0x06
 579 # endif
 580 # ifndef SUBLANG_ARABIC_TUNISIA
 581 # define SUBLANG_ARABIC_TUNISIA 0x07
 582 # endif
 583 # ifndef SUBLANG_ARABIC_OMAN
 584 # define SUBLANG_ARABIC_OMAN 0x08
 585 # endif
 586 # ifndef SUBLANG_ARABIC_YEMEN
 587 # define SUBLANG_ARABIC_YEMEN 0x09
 588 # endif
 589 # ifndef SUBLANG_ARABIC_SYRIA
 590 # define SUBLANG_ARABIC_SYRIA 0x0a
 591 # endif
 592 # ifndef SUBLANG_ARABIC_JORDAN
 593 # define SUBLANG_ARABIC_JORDAN 0x0b
 594 # endif
 595 # ifndef SUBLANG_ARABIC_LEBANON
 596 # define SUBLANG_ARABIC_LEBANON 0x0c
 597 # endif
 598 # ifndef SUBLANG_ARABIC_KUWAIT
 599 # define SUBLANG_ARABIC_KUWAIT 0x0d
 600 # endif
 601 # ifndef SUBLANG_ARABIC_UAE
 602 # define SUBLANG_ARABIC_UAE 0x0e
 603 # endif
 604 # ifndef SUBLANG_ARABIC_BAHRAIN
 605 # define SUBLANG_ARABIC_BAHRAIN 0x0f
 606 # endif
 607 # ifndef SUBLANG_ARABIC_QATAR
 608 # define SUBLANG_ARABIC_QATAR 0x10
 609 # endif
 610 # ifndef SUBLANG_ARMENIAN_ARMENIA
 611 # define SUBLANG_ARMENIAN_ARMENIA 0x01
 612 # endif
 613 # ifndef SUBLANG_ASSAMESE_INDIA
 614 # define SUBLANG_ASSAMESE_INDIA 0x01
 615 # endif
 616 # ifndef SUBLANG_AZERI_LATIN
 617 # define SUBLANG_AZERI_LATIN 0x01
 618 # endif
 619 # ifndef SUBLANG_AZERI_CYRILLIC
 620 # define SUBLANG_AZERI_CYRILLIC 0x02
 621 # endif
 622 # ifndef SUBLANG_BASHKIR_RUSSIA
 623 # define SUBLANG_BASHKIR_RUSSIA 0x01
 624 # endif
 625 # ifndef SUBLANG_BASQUE_BASQUE
 626 # define SUBLANG_BASQUE_BASQUE 0x01
 627 # endif
 628 # ifndef SUBLANG_BELARUSIAN_BELARUS
 629 # define SUBLANG_BELARUSIAN_BELARUS 0x01
 630 # endif
 631 # ifndef SUBLANG_BENGALI_INDIA
 632 # define SUBLANG_BENGALI_INDIA 0x01
 633 # endif
 634 # ifndef SUBLANG_BENGALI_BANGLADESH
 635 # define SUBLANG_BENGALI_BANGLADESH 0x02
 636 # endif
 637 # ifndef SUBLANG_BOSNIAN_BOSNIA_HERZEGOVINA_LATIN
 638 # define SUBLANG_BOSNIAN_BOSNIA_HERZEGOVINA_LATIN 0x05
 639 # endif
 640 # ifndef SUBLANG_BOSNIAN_BOSNIA_HERZEGOVINA_CYRILLIC
 641 # define SUBLANG_BOSNIAN_BOSNIA_HERZEGOVINA_CYRILLIC 0x08
 642 # endif
 643 # ifndef SUBLANG_BRETON_FRANCE
 644 # define SUBLANG_BRETON_FRANCE 0x01
 645 # endif
 646 # ifndef SUBLANG_BULGARIAN_BULGARIA
 647 # define SUBLANG_BULGARIAN_BULGARIA 0x01
 648 # endif
 649 # ifndef SUBLANG_CAMBODIAN_CAMBODIA
 650 # define SUBLANG_CAMBODIAN_CAMBODIA 0x01
 651 # endif
 652 # ifndef SUBLANG_CATALAN_SPAIN
 653 # define SUBLANG_CATALAN_SPAIN 0x01
 654 # endif
 655 # ifndef SUBLANG_CORSICAN_FRANCE
 656 # define SUBLANG_CORSICAN_FRANCE 0x01
 657 # endif
 658 # ifndef SUBLANG_CROATIAN_CROATIA
 659 # define SUBLANG_CROATIAN_CROATIA 0x01
 660 # endif
 661 # ifndef SUBLANG_CROATIAN_BOSNIA_HERZEGOVINA_LATIN
 662 # define SUBLANG_CROATIAN_BOSNIA_HERZEGOVINA_LATIN 0x04
 663 # endif
 664 # ifndef SUBLANG_CHINESE_MACAU
 665 # define SUBLANG_CHINESE_MACAU 0x05
 666 # endif
 667 # ifndef SUBLANG_CZECH_CZECH_REPUBLIC
 668 # define SUBLANG_CZECH_CZECH_REPUBLIC 0x01
 669 # endif
 670 # ifndef SUBLANG_DANISH_DENMARK
 671 # define SUBLANG_DANISH_DENMARK 0x01
 672 # endif
 673 # ifndef SUBLANG_DARI_AFGHANISTAN
 674 # define SUBLANG_DARI_AFGHANISTAN 0x01
 675 # endif
 676 # ifndef SUBLANG_DIVEHI_MALDIVES
 677 # define SUBLANG_DIVEHI_MALDIVES 0x01
 678 # endif
 679 # ifndef SUBLANG_DUTCH_SURINAM
 680 # define SUBLANG_DUTCH_SURINAM 0x03
 681 # endif
 682 # ifndef SUBLANG_ENGLISH_SOUTH_AFRICA
 683 # define SUBLANG_ENGLISH_SOUTH_AFRICA 0x07
 684 # endif
 685 # ifndef SUBLANG_ENGLISH_JAMAICA
 686 # define SUBLANG_ENGLISH_JAMAICA 0x08
 687 # endif
 688 # ifndef SUBLANG_ENGLISH_CARIBBEAN
 689 # define SUBLANG_ENGLISH_CARIBBEAN 0x09
 690 # endif
 691 # ifndef SUBLANG_ENGLISH_BELIZE
 692 # define SUBLANG_ENGLISH_BELIZE 0x0a
 693 # endif
 694 # ifndef SUBLANG_ENGLISH_TRINIDAD
 695 # define SUBLANG_ENGLISH_TRINIDAD 0x0b
 696 # endif
 697 # ifndef SUBLANG_ENGLISH_ZIMBABWE
 698 # define SUBLANG_ENGLISH_ZIMBABWE 0x0c
 699 # endif
 700 # ifndef SUBLANG_ENGLISH_PHILIPPINES
 701 # define SUBLANG_ENGLISH_PHILIPPINES 0x0d
 702 # endif
 703 # ifndef SUBLANG_ENGLISH_INDONESIA
 704 # define SUBLANG_ENGLISH_INDONESIA 0x0e
 705 # endif
 706 # ifndef SUBLANG_ENGLISH_HONGKONG
 707 # define SUBLANG_ENGLISH_HONGKONG 0x0f
 708 # endif
 709 # ifndef SUBLANG_ENGLISH_INDIA
 710 # define SUBLANG_ENGLISH_INDIA 0x10
 711 # endif
 712 # ifndef SUBLANG_ENGLISH_MALAYSIA
 713 # define SUBLANG_ENGLISH_MALAYSIA 0x11
 714 # endif
 715 # ifndef SUBLANG_ENGLISH_SINGAPORE
 716 # define SUBLANG_ENGLISH_SINGAPORE 0x12
 717 # endif
 718 # ifndef SUBLANG_ESTONIAN_ESTONIA
 719 # define SUBLANG_ESTONIAN_ESTONIA 0x01
 720 # endif
 721 # ifndef SUBLANG_FAEROESE_FAROE_ISLANDS
 722 # define SUBLANG_FAEROESE_FAROE_ISLANDS 0x01
 723 # endif
 724 # ifndef SUBLANG_FARSI_IRAN
 725 # define SUBLANG_FARSI_IRAN 0x01
 726 # endif
 727 # ifndef SUBLANG_FINNISH_FINLAND
 728 # define SUBLANG_FINNISH_FINLAND 0x01
 729 # endif
 730 # ifndef SUBLANG_FRENCH_LUXEMBOURG
 731 # define SUBLANG_FRENCH_LUXEMBOURG 0x05
 732 # endif
 733 # ifndef SUBLANG_FRENCH_MONACO
 734 # define SUBLANG_FRENCH_MONACO 0x06
 735 # endif
 736 # ifndef SUBLANG_FRENCH_WESTINDIES
 737 # define SUBLANG_FRENCH_WESTINDIES 0x07
 738 # endif
 739 # ifndef SUBLANG_FRENCH_REUNION
 740 # define SUBLANG_FRENCH_REUNION 0x08
 741 # endif
 742 # ifndef SUBLANG_FRENCH_CONGO
 743 # define SUBLANG_FRENCH_CONGO 0x09
 744 # endif
 745 # ifndef SUBLANG_FRENCH_SENEGAL
 746 # define SUBLANG_FRENCH_SENEGAL 0x0a
 747 # endif
 748 # ifndef SUBLANG_FRENCH_CAMEROON
 749 # define SUBLANG_FRENCH_CAMEROON 0x0b
 750 # endif
 751 # ifndef SUBLANG_FRENCH_COTEDIVOIRE
 752 # define SUBLANG_FRENCH_COTEDIVOIRE 0x0c
 753 # endif
 754 # ifndef SUBLANG_FRENCH_MALI
 755 # define SUBLANG_FRENCH_MALI 0x0d
 756 # endif
 757 # ifndef SUBLANG_FRENCH_MOROCCO
 758 # define SUBLANG_FRENCH_MOROCCO 0x0e
 759 # endif
 760 # ifndef SUBLANG_FRENCH_HAITI
 761 # define SUBLANG_FRENCH_HAITI 0x0f
 762 # endif
 763 # ifndef SUBLANG_FRISIAN_NETHERLANDS
 764 # define SUBLANG_FRISIAN_NETHERLANDS 0x01
 765 # endif
 766 # ifndef SUBLANG_GALICIAN_SPAIN
 767 # define SUBLANG_GALICIAN_SPAIN 0x01
 768 # endif
 769 # ifndef SUBLANG_GEORGIAN_GEORGIA
 770 # define SUBLANG_GEORGIAN_GEORGIA 0x01
 771 # endif
 772 # ifndef SUBLANG_GERMAN_LUXEMBOURG
 773 # define SUBLANG_GERMAN_LUXEMBOURG 0x04
 774 # endif
 775 # ifndef SUBLANG_GERMAN_LIECHTENSTEIN
 776 # define SUBLANG_GERMAN_LIECHTENSTEIN 0x05
 777 # endif
 778 # ifndef SUBLANG_GREEK_GREECE
 779 # define SUBLANG_GREEK_GREECE 0x01
 780 # endif
 781 # ifndef SUBLANG_GREENLANDIC_GREENLAND
 782 # define SUBLANG_GREENLANDIC_GREENLAND 0x01
 783 # endif
 784 # ifndef SUBLANG_GUJARATI_INDIA
 785 # define SUBLANG_GUJARATI_INDIA 0x01
 786 # endif
 787 # ifndef SUBLANG_HAUSA_NIGERIA_LATIN
 788 # define SUBLANG_HAUSA_NIGERIA_LATIN 0x01
 789 # endif
 790 # ifndef SUBLANG_HEBREW_ISRAEL
 791 # define SUBLANG_HEBREW_ISRAEL 0x01
 792 # endif
 793 # ifndef SUBLANG_HINDI_INDIA
 794 # define SUBLANG_HINDI_INDIA 0x01
 795 # endif
 796 # ifndef SUBLANG_HUNGARIAN_HUNGARY
 797 # define SUBLANG_HUNGARIAN_HUNGARY 0x01
 798 # endif
 799 # ifndef SUBLANG_ICELANDIC_ICELAND
 800 # define SUBLANG_ICELANDIC_ICELAND 0x01
 801 # endif
 802 # ifndef SUBLANG_IGBO_NIGERIA
 803 # define SUBLANG_IGBO_NIGERIA 0x01
 804 # endif
 805 # ifndef SUBLANG_INDONESIAN_INDONESIA
 806 # define SUBLANG_INDONESIAN_INDONESIA 0x01
 807 # endif
 808 # ifndef SUBLANG_INUKTITUT_CANADA
 809 # define SUBLANG_INUKTITUT_CANADA 0x01
 810 # endif
 811 # undef SUBLANG_INUKTITUT_CANADA_LATIN
 812 # define SUBLANG_INUKTITUT_CANADA_LATIN 0x02
 813 # undef SUBLANG_IRISH_IRELAND
 814 # define SUBLANG_IRISH_IRELAND 0x02
 815 # ifndef SUBLANG_JAPANESE_JAPAN
 816 # define SUBLANG_JAPANESE_JAPAN 0x01
 817 # endif
 818 # ifndef SUBLANG_KANNADA_INDIA
 819 # define SUBLANG_KANNADA_INDIA 0x01
 820 # endif
 821 # ifndef SUBLANG_KASHMIRI_INDIA
 822 # define SUBLANG_KASHMIRI_INDIA 0x02
 823 # endif
 824 # ifndef SUBLANG_KAZAK_KAZAKHSTAN
 825 # define SUBLANG_KAZAK_KAZAKHSTAN 0x01
 826 # endif
 827 # ifndef SUBLANG_KICHE_GUATEMALA
 828 # define SUBLANG_KICHE_GUATEMALA 0x01
 829 # endif
 830 # ifndef SUBLANG_KINYARWANDA_RWANDA
 831 # define SUBLANG_KINYARWANDA_RWANDA 0x01
 832 # endif
 833 # ifndef SUBLANG_KONKANI_INDIA
 834 # define SUBLANG_KONKANI_INDIA 0x01
 835 # endif
 836 # ifndef SUBLANG_KYRGYZ_KYRGYZSTAN
 837 # define SUBLANG_KYRGYZ_KYRGYZSTAN 0x01
 838 # endif
 839 # ifndef SUBLANG_LAO_LAOS
 840 # define SUBLANG_LAO_LAOS 0x01
 841 # endif
 842 # ifndef SUBLANG_LATVIAN_LATVIA
 843 # define SUBLANG_LATVIAN_LATVIA 0x01
 844 # endif
 845 # ifndef SUBLANG_LITHUANIAN_LITHUANIA
 846 # define SUBLANG_LITHUANIAN_LITHUANIA 0x01
 847 # endif
 848 # undef SUBLANG_LOWER_SORBIAN_GERMANY
 849 # define SUBLANG_LOWER_SORBIAN_GERMANY 0x02
 850 # ifndef SUBLANG_LUXEMBOURGISH_LUXEMBOURG
 851 # define SUBLANG_LUXEMBOURGISH_LUXEMBOURG 0x01
 852 # endif
 853 # ifndef SUBLANG_MACEDONIAN_MACEDONIA
 854 # define SUBLANG_MACEDONIAN_MACEDONIA 0x01
 855 # endif
 856 # ifndef SUBLANG_MALAY_MALAYSIA
 857 # define SUBLANG_MALAY_MALAYSIA 0x01
 858 # endif
 859 # ifndef SUBLANG_MALAY_BRUNEI_DARUSSALAM
 860 # define SUBLANG_MALAY_BRUNEI_DARUSSALAM 0x02
 861 # endif
 862 # ifndef SUBLANG_MALAYALAM_INDIA
 863 # define SUBLANG_MALAYALAM_INDIA 0x01
 864 # endif
 865 # ifndef SUBLANG_MALTESE_MALTA
 866 # define SUBLANG_MALTESE_MALTA 0x01
 867 # endif
 868 # ifndef SUBLANG_MAORI_NEW_ZEALAND
 869 # define SUBLANG_MAORI_NEW_ZEALAND 0x01
 870 # endif
 871 # ifndef SUBLANG_MAPUDUNGUN_CHILE
 872 # define SUBLANG_MAPUDUNGUN_CHILE 0x01
 873 # endif
 874 # ifndef SUBLANG_MARATHI_INDIA
 875 # define SUBLANG_MARATHI_INDIA 0x01
 876 # endif
 877 # ifndef SUBLANG_MOHAWK_CANADA
 878 # define SUBLANG_MOHAWK_CANADA 0x01
 879 # endif
 880 # ifndef SUBLANG_MONGOLIAN_CYRILLIC_MONGOLIA
 881 # define SUBLANG_MONGOLIAN_CYRILLIC_MONGOLIA 0x01
 882 # endif
 883 # ifndef SUBLANG_MONGOLIAN_PRC
 884 # define SUBLANG_MONGOLIAN_PRC 0x02
 885 # endif
 886 # ifndef SUBLANG_NEPALI_NEPAL
 887 # define SUBLANG_NEPALI_NEPAL 0x01
 888 # endif
 889 # ifndef SUBLANG_NEPALI_INDIA
 890 # define SUBLANG_NEPALI_INDIA 0x02
 891 # endif
 892 # ifndef SUBLANG_OCCITAN_FRANCE
 893 # define SUBLANG_OCCITAN_FRANCE 0x01
 894 # endif
 895 # ifndef SUBLANG_ORIYA_INDIA
 896 # define SUBLANG_ORIYA_INDIA 0x01
 897 # endif
 898 # ifndef SUBLANG_PASHTO_AFGHANISTAN
 899 # define SUBLANG_PASHTO_AFGHANISTAN 0x01
 900 # endif
 901 # ifndef SUBLANG_POLISH_POLAND
 902 # define SUBLANG_POLISH_POLAND 0x01
 903 # endif
 904 # ifndef SUBLANG_PUNJABI_INDIA
 905 # define SUBLANG_PUNJABI_INDIA 0x01
 906 # endif
 907 # ifndef SUBLANG_PUNJABI_PAKISTAN
 908 # define SUBLANG_PUNJABI_PAKISTAN 0x02
 909 # endif
 910 # ifndef SUBLANG_QUECHUA_BOLIVIA
 911 # define SUBLANG_QUECHUA_BOLIVIA 0x01
 912 # endif
 913 # ifndef SUBLANG_QUECHUA_ECUADOR
 914 # define SUBLANG_QUECHUA_ECUADOR 0x02
 915 # endif
 916 # ifndef SUBLANG_QUECHUA_PERU
 917 # define SUBLANG_QUECHUA_PERU 0x03
 918 # endif
 919 # ifndef SUBLANG_ROMANIAN_ROMANIA
 920 # define SUBLANG_ROMANIAN_ROMANIA 0x01
 921 # endif
 922 # ifndef SUBLANG_ROMANIAN_MOLDOVA
 923 # define SUBLANG_ROMANIAN_MOLDOVA 0x02
 924 # endif
 925 # ifndef SUBLANG_ROMANSH_SWITZERLAND
 926 # define SUBLANG_ROMANSH_SWITZERLAND 0x01
 927 # endif
 928 # ifndef SUBLANG_RUSSIAN_RUSSIA
 929 # define SUBLANG_RUSSIAN_RUSSIA 0x01
 930 # endif
 931 # ifndef SUBLANG_RUSSIAN_MOLDAVIA
 932 # define SUBLANG_RUSSIAN_MOLDAVIA 0x02
 933 # endif
 934 # ifndef SUBLANG_SAMI_NORTHERN_NORWAY
 935 # define SUBLANG_SAMI_NORTHERN_NORWAY 0x01
 936 # endif
 937 # ifndef SUBLANG_SAMI_NORTHERN_SWEDEN
 938 # define SUBLANG_SAMI_NORTHERN_SWEDEN 0x02
 939 # endif
 940 # ifndef SUBLANG_SAMI_NORTHERN_FINLAND
 941 # define SUBLANG_SAMI_NORTHERN_FINLAND 0x03
 942 # endif
 943 # ifndef SUBLANG_SAMI_LULE_NORWAY
 944 # define SUBLANG_SAMI_LULE_NORWAY 0x04
 945 # endif
 946 # ifndef SUBLANG_SAMI_LULE_SWEDEN
 947 # define SUBLANG_SAMI_LULE_SWEDEN 0x05
 948 # endif
 949 # ifndef SUBLANG_SAMI_SOUTHERN_NORWAY
 950 # define SUBLANG_SAMI_SOUTHERN_NORWAY 0x06
 951 # endif
 952 # ifndef SUBLANG_SAMI_SOUTHERN_SWEDEN
 953 # define SUBLANG_SAMI_SOUTHERN_SWEDEN 0x07
 954 # endif
 955 # undef SUBLANG_SAMI_SKOLT_FINLAND
 956 # define SUBLANG_SAMI_SKOLT_FINLAND 0x08
 957 # undef SUBLANG_SAMI_INARI_FINLAND
 958 # define SUBLANG_SAMI_INARI_FINLAND 0x09
 959 # ifndef SUBLANG_SANSKRIT_INDIA
 960 # define SUBLANG_SANSKRIT_INDIA 0x01
 961 # endif
 962 # ifndef SUBLANG_SERBIAN_LATIN
 963 # define SUBLANG_SERBIAN_LATIN 0x02
 964 # endif
 965 # ifndef SUBLANG_SERBIAN_CYRILLIC
 966 # define SUBLANG_SERBIAN_CYRILLIC 0x03
 967 # endif
 968 # ifndef SUBLANG_SINDHI_INDIA
 969 # define SUBLANG_SINDHI_INDIA 0x01
 970 # endif
 971 # undef SUBLANG_SINDHI_PAKISTAN
 972 # define SUBLANG_SINDHI_PAKISTAN 0x02
 973 # ifndef SUBLANG_SINDHI_AFGHANISTAN
 974 # define SUBLANG_SINDHI_AFGHANISTAN 0x02
 975 # endif
 976 # ifndef SUBLANG_SINHALESE_SRI_LANKA
 977 # define SUBLANG_SINHALESE_SRI_LANKA 0x01
 978 # endif
 979 # ifndef SUBLANG_SLOVAK_SLOVAKIA
 980 # define SUBLANG_SLOVAK_SLOVAKIA 0x01
 981 # endif
 982 # ifndef SUBLANG_SLOVENIAN_SLOVENIA
 983 # define SUBLANG_SLOVENIAN_SLOVENIA 0x01
 984 # endif
 985 # ifndef SUBLANG_SOTHO_SOUTH_AFRICA
 986 # define SUBLANG_SOTHO_SOUTH_AFRICA 0x01
 987 # endif
 988 # ifndef SUBLANG_SPANISH_GUATEMALA
 989 # define SUBLANG_SPANISH_GUATEMALA 0x04
 990 # endif
 991 # ifndef SUBLANG_SPANISH_COSTA_RICA
 992 # define SUBLANG_SPANISH_COSTA_RICA 0x05
 993 # endif
 994 # ifndef SUBLANG_SPANISH_PANAMA
 995 # define SUBLANG_SPANISH_PANAMA 0x06
 996 # endif
 997 # ifndef SUBLANG_SPANISH_DOMINICAN_REPUBLIC
 998 # define SUBLANG_SPANISH_DOMINICAN_REPUBLIC 0x07
 999 # endif
1000 # ifndef SUBLANG_SPANISH_VENEZUELA
1001 # define SUBLANG_SPANISH_VENEZUELA 0x08
1002 # endif
1003 # ifndef SUBLANG_SPANISH_COLOMBIA
1004 # define SUBLANG_SPANISH_COLOMBIA 0x09
1005 # endif
1006 # ifndef SUBLANG_SPANISH_PERU
1007 # define SUBLANG_SPANISH_PERU 0x0a
1008 # endif
1009 # ifndef SUBLANG_SPANISH_ARGENTINA
1010 # define SUBLANG_SPANISH_ARGENTINA 0x0b
1011 # endif
1012 # ifndef SUBLANG_SPANISH_ECUADOR
1013 # define SUBLANG_SPANISH_ECUADOR 0x0c
1014 # endif
1015 # ifndef SUBLANG_SPANISH_CHILE
1016 # define SUBLANG_SPANISH_CHILE 0x0d
1017 # endif
1018 # ifndef SUBLANG_SPANISH_URUGUAY
1019 # define SUBLANG_SPANISH_URUGUAY 0x0e
1020 # endif
1021 # ifndef SUBLANG_SPANISH_PARAGUAY
1022 # define SUBLANG_SPANISH_PARAGUAY 0x0f
1023 # endif
1024 # ifndef SUBLANG_SPANISH_BOLIVIA
1025 # define SUBLANG_SPANISH_BOLIVIA 0x10
1026 # endif
1027 # ifndef SUBLANG_SPANISH_EL_SALVADOR
1028 # define SUBLANG_SPANISH_EL_SALVADOR 0x11
1029 # endif
1030 # ifndef SUBLANG_SPANISH_HONDURAS
1031 # define SUBLANG_SPANISH_HONDURAS 0x12
1032 # endif
1033 # ifndef SUBLANG_SPANISH_NICARAGUA
1034 # define SUBLANG_SPANISH_NICARAGUA 0x13
1035 # endif
1036 # ifndef SUBLANG_SPANISH_PUERTO_RICO
1037 # define SUBLANG_SPANISH_PUERTO_RICO 0x14
1038 # endif
1039 # ifndef SUBLANG_SPANISH_US
1040 # define SUBLANG_SPANISH_US 0x15
1041 # endif
1042 # ifndef SUBLANG_SWAHILI_KENYA
1043 # define SUBLANG_SWAHILI_KENYA 0x01
1044 # endif
1045 # ifndef SUBLANG_SWEDISH_SWEDEN
1046 # define SUBLANG_SWEDISH_SWEDEN 0x01
1047 # endif
1048 # ifndef SUBLANG_SWEDISH_FINLAND
1049 # define SUBLANG_SWEDISH_FINLAND 0x02
1050 # endif
1051 # ifndef SUBLANG_SYRIAC_SYRIA
1052 # define SUBLANG_SYRIAC_SYRIA 0x01
1053 # endif
1054 # ifndef SUBLANG_TAGALOG_PHILIPPINES
1055 # define SUBLANG_TAGALOG_PHILIPPINES 0x01
1056 # endif
1057 # ifndef SUBLANG_TAJIK_TAJIKISTAN
1058 # define SUBLANG_TAJIK_TAJIKISTAN 0x01
1059 # endif
1060 # ifndef SUBLANG_TAMAZIGHT_ARABIC
1061 # define SUBLANG_TAMAZIGHT_ARABIC 0x01
1062 # endif
1063 # ifndef SUBLANG_TAMAZIGHT_ALGERIA_LATIN
1064 # define SUBLANG_TAMAZIGHT_ALGERIA_LATIN 0x02
1065 # endif
1066 # ifndef SUBLANG_TAMIL_INDIA
1067 # define SUBLANG_TAMIL_INDIA 0x01
1068 # endif
1069 # ifndef SUBLANG_TATAR_RUSSIA
1070 # define SUBLANG_TATAR_RUSSIA 0x01
1071 # endif
1072 # ifndef SUBLANG_TELUGU_INDIA
1073 # define SUBLANG_TELUGU_INDIA 0x01
1074 # endif
1075 # ifndef SUBLANG_THAI_THAILAND
1076 # define SUBLANG_THAI_THAILAND 0x01
1077 # endif
1078 # ifndef SUBLANG_TIBETAN_PRC
1079 # define SUBLANG_TIBETAN_PRC 0x01
1080 # endif
1081 # undef SUBLANG_TIBETAN_BHUTAN
1082 # define SUBLANG_TIBETAN_BHUTAN 0x02
1083 # ifndef SUBLANG_TIGRINYA_ETHIOPIA
1084 # define SUBLANG_TIGRINYA_ETHIOPIA 0x01
1085 # endif
1086 # ifndef SUBLANG_TIGRINYA_ERITREA
1087 # define SUBLANG_TIGRINYA_ERITREA 0x02
1088 # endif
1089 # ifndef SUBLANG_TSWANA_SOUTH_AFRICA
1090 # define SUBLANG_TSWANA_SOUTH_AFRICA 0x01
1091 # endif
1092 # ifndef SUBLANG_TURKISH_TURKEY
1093 # define SUBLANG_TURKISH_TURKEY 0x01
1094 # endif
1095 # ifndef SUBLANG_TURKMEN_TURKMENISTAN
1096 # define SUBLANG_TURKMEN_TURKMENISTAN 0x01
1097 # endif
1098 # ifndef SUBLANG_UIGHUR_PRC
1099 # define SUBLANG_UIGHUR_PRC 0x01
1100 # endif
1101 # ifndef SUBLANG_UKRAINIAN_UKRAINE
1102 # define SUBLANG_UKRAINIAN_UKRAINE 0x01
1103 # endif
1104 # ifndef SUBLANG_UPPER_SORBIAN_GERMANY
1105 # define SUBLANG_UPPER_SORBIAN_GERMANY 0x01
1106 # endif
1107 # ifndef SUBLANG_URDU_PAKISTAN
1108 # define SUBLANG_URDU_PAKISTAN 0x01
1109 # endif
1110 # ifndef SUBLANG_URDU_INDIA
1111 # define SUBLANG_URDU_INDIA 0x02
1112 # endif
1113 # ifndef SUBLANG_UZBEK_LATIN
1114 # define SUBLANG_UZBEK_LATIN 0x01
1115 # endif
1116 # ifndef SUBLANG_UZBEK_CYRILLIC
1117 # define SUBLANG_UZBEK_CYRILLIC 0x02
1118 # endif
1119 # ifndef SUBLANG_VIETNAMESE_VIETNAM
1120 # define SUBLANG_VIETNAMESE_VIETNAM 0x01
1121 # endif
1122 # ifndef SUBLANG_WELSH_UNITED_KINGDOM
1123 # define SUBLANG_WELSH_UNITED_KINGDOM 0x01
1124 # endif
1125 # ifndef SUBLANG_WOLOF_SENEGAL
1126 # define SUBLANG_WOLOF_SENEGAL 0x01
1127 # endif
1128 # ifndef SUBLANG_XHOSA_SOUTH_AFRICA
1129 # define SUBLANG_XHOSA_SOUTH_AFRICA 0x01
1130 # endif
1131 # ifndef SUBLANG_YAKUT_RUSSIA
1132 # define SUBLANG_YAKUT_RUSSIA 0x01
1133 # endif
1134 # ifndef SUBLANG_YI_PRC
1135 # define SUBLANG_YI_PRC 0x01
1136 # endif
1137 # ifndef SUBLANG_YORUBA_NIGERIA
1138 # define SUBLANG_YORUBA_NIGERIA 0x01
1139 # endif
1140 # ifndef SUBLANG_ZULU_SOUTH_AFRICA
1141 # define SUBLANG_ZULU_SOUTH_AFRICA 0x01
1142 # endif
1143 /* GetLocaleInfoA operations.  */
1144 # ifndef LOCALE_SNAME
1145 # define LOCALE_SNAME 0x5c
1146 # endif
1147 # ifndef LOCALE_NAME_MAX_LENGTH
1148 # define LOCALE_NAME_MAX_LENGTH 85
1149 # endif
1150 #endif
1151
1152
1153 #if HAVE_CFLOCALECOPYCURRENT || HAVE_CFPREFERENCESCOPYAPPVALUE
1154 /* Mac OS X 10.4 or newer */
1155
1156 /* Canonicalize a Mac OS X locale name to a Unix locale name.
1157    NAME is a sufficiently large buffer.
1158    On input, it contains the Mac OS X locale name.
1159    On output, it contains the Unix locale name.  */
1160 # if !defined IN_LIBINTL
1161 static
1162 # endif
1163 void
1164 gl_locale_name_canonicalize (char *name)
1165 {
1166   /* This conversion is based on a posting by
1167      Deborah GoldSmith <goldsmit@apple.com> on 2005-03-08,
1168      https://lists.apple.com/archives/carbon-dev/2005/Mar/msg00293.html */
1169
1170   /* Convert legacy (NeXTstep inherited) English names to Unix (ISO 639 and
1171      ISO 3166) names.  Prior to Mac OS X 10.3, there is no API for doing this.
1172      Therefore we do it ourselves, using a table based on the results of the
1173      Mac OS X 10.3.8 function
1174      CFLocaleCreateCanonicalLocaleIdentifierFromString().  */
1175   typedef struct { const char legacy[21+1]; const char unixy[5+1]; }
1176           legacy_entry;
1177   static const legacy_entry legacy_table[] = {
1178     { "Afrikaans",             "af" },
1179     { "Albanian",              "sq" },
1180     { "Amharic",               "am" },
1181     { "Arabic",                "ar" },
1182     { "Armenian",              "hy" },
1183     { "Assamese",              "as" },
1184     { "Aymara",                "ay" },
1185     { "Azerbaijani",           "az" },
1186     { "Basque",                "eu" },
1187     { "Belarusian",            "be" },
1188     { "Belorussian",           "be" },
1189     { "Bengali",               "bn" },
1190     { "Brazilian Portugese",   "pt_BR" },
1191     { "Brazilian Portuguese",  "pt_BR" },
1192     { "Breton",                "br" },
1193     { "Bulgarian",             "bg" },
1194     { "Burmese",               "my" },
1195     { "Byelorussian",          "be" },
1196     { "Catalan",               "ca" },
1197     { "Chewa",                 "ny" },
1198     { "Chichewa",              "ny" },
1199     { "Chinese",               "zh" },
1200     { "Chinese, Simplified",   "zh_CN" },
1201     { "Chinese, Traditional",  "zh_TW" },
1202     { "Chinese, Tradtional",   "zh_TW" },
1203     { "Croatian",              "hr" },
1204     { "Czech",                 "cs" },
1205     { "Danish",                "da" },
1206     { "Dutch",                 "nl" },
1207     { "Dzongkha",              "dz" },
1208     { "English",               "en" },
1209     { "Esperanto",             "eo" },
1210     { "Estonian",              "et" },
1211     { "Faroese",               "fo" },
1212     { "Farsi",                 "fa" },
1213     { "Finnish",               "fi" },
1214     { "Flemish",               "nl_BE" },
1215     { "French",                "fr" },
1216     { "Galician",              "gl" },
1217     { "Gallegan",              "gl" },
1218     { "Georgian",              "ka" },
1219     { "German",                "de" },
1220     { "Greek",                 "el" },
1221     { "Greenlandic",           "kl" },
1222     { "Guarani",               "gn" },
1223     { "Gujarati",              "gu" },
1224     { "Hawaiian",              "haw" }, /* Yes, "haw", not "cpe".  */
1225     { "Hebrew",                "he" },
1226     { "Hindi",                 "hi" },
1227     { "Hungarian",             "hu" },
1228     { "Icelandic",             "is" },
1229     { "Indonesian",            "id" },
1230     { "Inuktitut",             "iu" },
1231     { "Irish",                 "ga" },
1232     { "Italian",               "it" },
1233     { "Japanese",              "ja" },
1234     { "Javanese",              "jv" },
1235     { "Kalaallisut",           "kl" },
1236     { "Kannada",               "kn" },
1237     { "Kashmiri",              "ks" },
1238     { "Kazakh",                "kk" },
1239     { "Khmer",                 "km" },
1240     { "Kinyarwanda",           "rw" },
1241     { "Kirghiz",               "ky" },
1242     { "Korean",                "ko" },
1243     { "Kurdish",               "ku" },
1244     { "Latin",                 "la" },
1245     { "Latvian",               "lv" },
1246     { "Lithuanian",            "lt" },
1247     { "Macedonian",            "mk" },
1248     { "Malagasy",              "mg" },
1249     { "Malay",                 "ms" },
1250     { "Malayalam",             "ml" },
1251     { "Maltese",               "mt" },
1252     { "Manx",                  "gv" },
1253     { "Marathi",               "mr" },
1254     { "Moldavian",             "mo" },
1255     { "Mongolian",             "mn" },
1256     { "Nepali",                "ne" },
1257     { "Norwegian",             "nb" }, /* Yes, "nb", not the obsolete "no".  */
1258     { "Nyanja",                "ny" },
1259     { "Nynorsk",               "nn" },
1260     { "Oriya",                 "or" },
1261     { "Oromo",                 "om" },
1262     { "Panjabi",               "pa" },
1263     { "Pashto",                "ps" },
1264     { "Persian",               "fa" },
1265     { "Polish",                "pl" },
1266     { "Portuguese",            "pt" },
1267     { "Portuguese, Brazilian", "pt_BR" },
1268     { "Punjabi",               "pa" },
1269     { "Pushto",                "ps" },
1270     { "Quechua",               "qu" },
1271     { "Romanian",              "ro" },
1272     { "Ruanda",                "rw" },
1273     { "Rundi",                 "rn" },
1274     { "Russian",               "ru" },
1275     { "Sami",                  "se_NO" }, /* Not just "se".  */
1276     { "Sanskrit",              "sa" },
1277     { "Scottish",              "gd" },
1278     { "Serbian",               "sr" },
1279     { "Simplified Chinese",    "zh_CN" },
1280     { "Sindhi",                "sd" },
1281     { "Sinhalese",             "si" },
1282     { "Slovak",                "sk" },
1283     { "Slovenian",             "sl" },
1284     { "Somali",                "so" },
1285     { "Spanish",               "es" },
1286     { "Sundanese",             "su" },
1287     { "Swahili",               "sw" },
1288     { "Swedish",               "sv" },
1289     { "Tagalog",               "tl" },
1290     { "Tajik",                 "tg" },
1291     { "Tajiki",                "tg" },
1292     { "Tamil",                 "ta" },
1293     { "Tatar",                 "tt" },
1294     { "Telugu",                "te" },
1295     { "Thai",                  "th" },
1296     { "Tibetan",               "bo" },
1297     { "Tigrinya",              "ti" },
1298     { "Tongan",                "to" },
1299     { "Traditional Chinese",   "zh_TW" },
1300     { "Turkish",               "tr" },
1301     { "Turkmen",               "tk" },
1302     { "Uighur",                "ug" },
1303     { "Ukrainian",             "uk" },
1304     { "Urdu",                  "ur" },
1305     { "Uzbek",                 "uz" },
1306     { "Vietnamese",            "vi" },
1307     { "Welsh",                 "cy" },
1308     { "Yiddish",               "yi" }
1309   };
1310
1311   /* Convert new-style locale names with language tags (ISO 639 and ISO 15924)
1312      to Unix (ISO 639 and ISO 3166) names.  */
1313   typedef struct { const char langtag[7+1]; const char unixy[12+1]; }
1314           langtag_entry;
1315   static const langtag_entry langtag_table[] = {
1316     /* Mac OS X has "az-Arab", "az-Cyrl", "az-Latn".
1317        The default script for az on Unix is Latin.  */
1318     { "az-Latn", "az" },
1319     /* Mac OS X has "bs-Cyrl", "bs-Latn".
1320        The default script for bs on Unix is Latin.  */
1321     { "bs-Latn", "bs" },
1322     /* Mac OS X has "ga-dots".  Does not yet exist on Unix.  */
1323     { "ga-dots", "ga" },
1324     /* Mac OS X has "kk-Cyrl".
1325        The default script for kk on Unix is Cyrillic.  */
1326     { "kk-Cyrl", "kk" },
1327     /* Mac OS X has "mn-Cyrl", "mn-Mong".
1328        The default script for mn on Unix is Cyrillic.  */
1329     { "mn-Cyrl", "mn" },
1330     /* Mac OS X has "ms-Arab", "ms-Latn".
1331        The default script for ms on Unix is Latin.  */
1332     { "ms-Latn", "ms" },
1333     /* Mac OS X has "pa-Arab", "pa-Guru".
1334        Country codes are used to distinguish these on Unix.  */
1335     { "pa-Arab", "pa_PK" },
1336     { "pa-Guru", "pa_IN" },
1337     /* Mac OS X has "shi-Latn", "shi-Tfng".  Does not yet exist on Unix.  */
1338     /* Mac OS X has "sr-Cyrl", "sr-Latn".
1339        The default script for sr on Unix is Cyrillic.  */
1340     { "sr-Cyrl", "sr" },
1341     /* Mac OS X has "tg-Cyrl".
1342        The default script for tg on Unix is Cyrillic.  */
1343     { "tg-Cyrl", "tg" },
1344     /* Mac OS X has "tk-Cyrl".
1345        The default script for tk on Unix is Cyrillic.  */
1346     { "tk-Cyrl", "tk" },
1347     /* Mac OS X has "tt-Cyrl".
1348        The default script for tt on Unix is Cyrillic.  */
1349     { "tt-Cyrl", "tt" },
1350     /* Mac OS X has "uz-Arab", "uz-Cyrl", "uz-Latn".
1351        The default script for uz on Unix is Latin.  */
1352     { "uz-Latn", "uz" },
1353     /* Mac OS X has "vai-Latn", "vai-Vaii".  Does not yet exist on Unix.  */
1354     /* Mac OS X has "yue-Hans", "yue-Hant".
1355        The default script for yue on Unix is Simplified Han.  */
1356     { "yue-Hans", "yue" },
1357     /* Mac OS X has "zh-Hans", "zh-Hant".
1358        Country codes are used to distinguish these on Unix.  */
1359     { "zh-Hans", "zh_CN" },
1360     { "zh-Hant", "zh_TW" }
1361   };
1362
1363   /* Convert script names (ISO 15924) to Unix conventions.
1364      See https://www.unicode.org/iso15924/iso15924-codes.html  */
1365   typedef struct { const char script[4+1]; const char unixy[9+1]; }
1366           script_entry;
1367   static const script_entry script_table[] = {
1368     { "Arab", "arabic" },
1369     { "Cyrl", "cyrillic" },
1370     { "Latn", "latin" },
1371     { "Mong", "mongolian" }
1372   };
1373
1374   /* Step 1: Convert using legacy_table.  */
1375   if (name[0] >= 'A' && name[0] <= 'Z')
1376     {
1377       unsigned int i1, i2;
1378       i1 = 0;
1379       i2 = sizeof (legacy_table) / sizeof (legacy_entry);
1380       while (i2 - i1 > 1)
1381         {
1382           /* At this point we know that if name occurs in legacy_table,
1383              its index must be >= i1 and < i2.  */
1384           unsigned int i = (i1 + i2) >> 1;
1385           const legacy_entry *p = &legacy_table[i];
1386           if (strcmp (name, p->legacy) < 0)
1387             i2 = i;
1388           else
1389             i1 = i;
1390         }
1391       if (strcmp (name, legacy_table[i1].legacy) == 0)
1392         {
1393           strcpy (name, legacy_table[i1].unixy);
1394           return;
1395         }
1396     }
1397
1398   /* Step 2: Convert using langtag_table and script_table.  */
1399   if (strlen (name) == 7 && name[2] == '-')
1400     {
1401       unsigned int i1, i2;
1402       i1 = 0;
1403       i2 = sizeof (langtag_table) / sizeof (langtag_entry);
1404       while (i2 - i1 > 1)
1405         {
1406           /* At this point we know that if name occurs in langtag_table,
1407              its index must be >= i1 and < i2.  */
1408           unsigned int i = (i1 + i2) >> 1;
1409           const langtag_entry *p = &langtag_table[i];
1410           if (strcmp (name, p->langtag) < 0)
1411             i2 = i;
1412           else
1413             i1 = i;
1414         }
1415       if (strcmp (name, langtag_table[i1].langtag) == 0)
1416         {
1417           strcpy (name, langtag_table[i1].unixy);
1418           return;
1419         }
1420
1421       i1 = 0;
1422       i2 = sizeof (script_table) / sizeof (script_entry);
1423       while (i2 - i1 > 1)
1424         {
1425           /* At this point we know that if (name + 3) occurs in script_table,
1426              its index must be >= i1 and < i2.  */
1427           unsigned int i = (i1 + i2) >> 1;
1428           const script_entry *p = &script_table[i];
1429           if (strcmp (name + 3, p->script) < 0)
1430             i2 = i;
1431           else
1432             i1 = i;
1433         }
1434       if (strcmp (name + 3, script_table[i1].script) == 0)
1435         {
1436           name[2] = '@';
1437           strcpy (name + 3, script_table[i1].unixy);
1438           return;
1439         }
1440     }
1441
1442   /* Step 3: Convert new-style dash to Unix underscore. */
1443   {
1444     char *p;
1445     for (p = name; *p != '\0'; p++)
1446       if (*p == '-')
1447         *p = '_';
1448   }
1449 }
1450
1451 #endif
1452
1453
1454 #if defined WINDOWS_NATIVE || defined __CYGWIN__ /* Native Windows or Cygwin */
1455
1456 /* Canonicalize a Windows native locale name to a Unix locale name.
1457    NAME is a sufficiently large buffer.
1458    On input, it contains the Windows locale name.
1459    On output, it contains the Unix locale name.  */
1460 # if !defined IN_LIBINTL
1461 static
1462 # endif
1463 void
1464 gl_locale_name_canonicalize (char *name)
1465 {
1466   /* FIXME: This is probably incomplete: it does not handle "zh-Hans" and
1467      "zh-Hant".  */
1468   char *p;
1469
1470   for (p = name; *p != '\0'; p++)
1471     if (*p == '-')
1472       {
1473         *p = '_';
1474         p++;
1475         for (; *p != '\0'; p++)
1476           {
1477             if (*p >= 'a' && *p <= 'z')
1478               *p += 'A' - 'a';
1479             if (*p == '-')
1480               {
1481                 *p = '\0';
1482                 return;
1483               }
1484           }
1485         return;
1486       }
1487 }
1488
1489 # if !defined IN_LIBINTL
1490 static
1491 # endif
1492 const char *
1493 gl_locale_name_from_win32_LANGID (LANGID langid)
1494 {
1495   /* Activate the new code only when the GETTEXT_MUI environment variable is
1496      set, for the time being, since the new code is not well tested.  */
1497   if (getenv ("GETTEXT_MUI") != NULL)
1498     {
1499       static char namebuf[256];
1500
1501       /* Query the system's notion of locale name.
1502          On Windows95/98/ME, GetLocaleInfoA returns some incorrect results.
1503          But we don't need to support systems that are so old.  */
1504       if (GetLocaleInfoA (MAKELCID (langid, SORT_DEFAULT), LOCALE_SNAME,
1505                           namebuf, sizeof (namebuf) - 1))
1506         {
1507           /* Convert it to a Unix locale name.  */
1508           gl_locale_name_canonicalize (namebuf);
1509           return namebuf;
1510         }
1511     }
1512   /* Internet Explorer has an LCID to RFC3066 name mapping stored in
1513      HKEY_CLASSES_ROOT\Mime\Database\Rfc1766.  But we better don't use that
1514      since IE's i18n subsystem is known to be inconsistent with the native
1515      Windows base (e.g. they have different character conversion facilities
1516      that produce different results).  */
1517   /* Use our own table.  */
1518   {
1519     int primary, sub;
1520
1521     /* Split into language and territory part.  */
1522     primary = PRIMARYLANGID (langid);
1523     sub = SUBLANGID (langid);
1524
1525     /* Dispatch on language.
1526        See also https://www.unicode.org/unicode/onlinedat/languages.html .
1527        For details about languages, see https://www.ethnologue.com/ .  */
1528     switch (primary)
1529       {
1530       case LANG_AFRIKAANS:
1531         switch (sub)
1532           {
1533           case SUBLANG_AFRIKAANS_SOUTH_AFRICA: return "af_ZA";
1534           }
1535         return "af";
1536       case LANG_ALBANIAN:
1537         switch (sub)
1538           {
1539           case SUBLANG_ALBANIAN_ALBANIA: return "sq_AL";
1540           }
1541         return "sq";
1542       case LANG_ALSATIAN:
1543         switch (sub)
1544           {
1545           case SUBLANG_ALSATIAN_FRANCE: return "gsw_FR";
1546           }
1547         return "gsw";
1548       case LANG_AMHARIC:
1549         switch (sub)
1550           {
1551           case SUBLANG_AMHARIC_ETHIOPIA: return "am_ET";
1552           }
1553         return "am";
1554       case LANG_ARABIC:
1555         switch (sub)
1556           {
1557           case SUBLANG_ARABIC_SAUDI_ARABIA: return "ar_SA";
1558           case SUBLANG_ARABIC_IRAQ: return "ar_IQ";
1559           case SUBLANG_ARABIC_EGYPT: return "ar_EG";
1560           case SUBLANG_ARABIC_LIBYA: return "ar_LY";
1561           case SUBLANG_ARABIC_ALGERIA: return "ar_DZ";
1562           case SUBLANG_ARABIC_MOROCCO: return "ar_MA";
1563           case SUBLANG_ARABIC_TUNISIA: return "ar_TN";
1564           case SUBLANG_ARABIC_OMAN: return "ar_OM";
1565           case SUBLANG_ARABIC_YEMEN: return "ar_YE";
1566           case SUBLANG_ARABIC_SYRIA: return "ar_SY";
1567           case SUBLANG_ARABIC_JORDAN: return "ar_JO";
1568           case SUBLANG_ARABIC_LEBANON: return "ar_LB";
1569           case SUBLANG_ARABIC_KUWAIT: return "ar_KW";
1570           case SUBLANG_ARABIC_UAE: return "ar_AE";
1571           case SUBLANG_ARABIC_BAHRAIN: return "ar_BH";
1572           case SUBLANG_ARABIC_QATAR: return "ar_QA";
1573           }
1574         return "ar";
1575       case LANG_ARMENIAN:
1576         switch (sub)
1577           {
1578           case SUBLANG_ARMENIAN_ARMENIA: return "hy_AM";
1579           }
1580         return "hy";
1581       case LANG_ASSAMESE:
1582         switch (sub)
1583           {
1584           case SUBLANG_ASSAMESE_INDIA: return "as_IN";
1585           }
1586         return "as";
1587       case LANG_AZERI:
1588         switch (sub)
1589           {
1590           /* FIXME: Adjust this when Azerbaijani locales appear on Unix.  */
1591           case 0x1e: return "az@latin";
1592           case SUBLANG_AZERI_LATIN: return "az_AZ@latin";
1593           case 0x1d: return "az@cyrillic";
1594           case SUBLANG_AZERI_CYRILLIC: return "az_AZ@cyrillic";
1595           }
1596         return "az";
1597       case LANG_BASHKIR:
1598         switch (sub)
1599           {
1600           case SUBLANG_BASHKIR_RUSSIA: return "ba_RU";
1601           }
1602         return "ba";
1603       case LANG_BASQUE:
1604         switch (sub)
1605           {
1606           case SUBLANG_BASQUE_BASQUE: return "eu_ES";
1607           }
1608         return "eu"; /* Ambiguous: could be "eu_ES" or "eu_FR".  */
1609       case LANG_BELARUSIAN:
1610         switch (sub)
1611           {
1612           case SUBLANG_BELARUSIAN_BELARUS: return "be_BY";
1613           }
1614         return "be";
1615       case LANG_BENGALI:
1616         switch (sub)
1617           {
1618           case SUBLANG_BENGALI_INDIA: return "bn_IN";
1619           case SUBLANG_BENGALI_BANGLADESH: return "bn_BD";
1620           }
1621         return "bn";
1622       case LANG_BRETON:
1623         switch (sub)
1624           {
1625           case SUBLANG_BRETON_FRANCE: return "br_FR";
1626           }
1627         return "br";
1628       case LANG_BULGARIAN:
1629         switch (sub)
1630           {
1631           case SUBLANG_BULGARIAN_BULGARIA: return "bg_BG";
1632           }
1633         return "bg";
1634       case LANG_BURMESE:
1635         switch (sub)
1636           {
1637           case SUBLANG_DEFAULT: return "my_MM";
1638           }
1639         return "my";
1640       case LANG_CAMBODIAN:
1641         switch (sub)
1642           {
1643           case SUBLANG_CAMBODIAN_CAMBODIA: return "km_KH";
1644           }
1645         return "km";
1646       case LANG_CATALAN:
1647         switch (sub)
1648           {
1649           case SUBLANG_CATALAN_SPAIN: return "ca_ES";
1650           }
1651         return "ca";
1652       case LANG_CHEROKEE:
1653         switch (sub)
1654           {
1655           case SUBLANG_DEFAULT: return "chr_US";
1656           }
1657         return "chr";
1658       case LANG_CHINESE:
1659         switch (sub)
1660           {
1661           case SUBLANG_CHINESE_TRADITIONAL: case 0x1f: return "zh_TW";
1662           case SUBLANG_CHINESE_SIMPLIFIED: case 0x00: return "zh_CN";
1663           case SUBLANG_CHINESE_HONGKONG: return "zh_HK"; /* traditional */
1664           case SUBLANG_CHINESE_SINGAPORE: return "zh_SG"; /* simplified */
1665           case SUBLANG_CHINESE_MACAU: return "zh_MO"; /* traditional */
1666           }
1667         return "zh";
1668       case LANG_CORSICAN:
1669         switch (sub)
1670           {
1671           case SUBLANG_CORSICAN_FRANCE: return "co_FR";
1672           }
1673         return "co";
1674       case LANG_CROATIAN:      /* LANG_CROATIAN == LANG_SERBIAN == LANG_BOSNIAN
1675                                 * What used to be called Serbo-Croatian
1676                                 * should really now be two separate
1677                                 * languages because of political reasons.
1678                                 * (Says tml, who knows nothing about Serbian
1679                                 * or Croatian.)
1680                                 * (I can feel those flames coming already.)
1681                                 */
1682         switch (sub)
1683           {
1684           /* Croatian */
1685           case 0x00: return "hr";
1686           case SUBLANG_CROATIAN_CROATIA: return "hr_HR";
1687           case SUBLANG_CROATIAN_BOSNIA_HERZEGOVINA_LATIN: return "hr_BA";
1688           /* Serbian */
1689           case 0x1f: return "sr";
1690           case 0x1c: return "sr"; /* latin */
1691           case SUBLANG_SERBIAN_LATIN: return "sr_CS"; /* latin */
1692           case 0x09: return "sr_RS"; /* latin */
1693           case 0x0b: return "sr_ME"; /* latin */
1694           case 0x06: return "sr_BA"; /* latin */
1695           case 0x1b: return "sr@cyrillic";
1696           case SUBLANG_SERBIAN_CYRILLIC: return "sr_CS@cyrillic";
1697           case 0x0a: return "sr_RS@cyrillic";
1698           case 0x0c: return "sr_ME@cyrillic";
1699           case 0x07: return "sr_BA@cyrillic";
1700           /* Bosnian */
1701           case 0x1e: return "bs";
1702           case 0x1a: return "bs"; /* latin */
1703           case SUBLANG_BOSNIAN_BOSNIA_HERZEGOVINA_LATIN: return "bs_BA"; /* latin */
1704           case 0x19: return "bs@cyrillic";
1705           case SUBLANG_BOSNIAN_BOSNIA_HERZEGOVINA_CYRILLIC: return "bs_BA@cyrillic";
1706           }
1707         return "hr";
1708       case LANG_CZECH:
1709         switch (sub)
1710           {
1711           case SUBLANG_CZECH_CZECH_REPUBLIC: return "cs_CZ";
1712           }
1713         return "cs";
1714       case LANG_DANISH:
1715         switch (sub)
1716           {
1717           case SUBLANG_DANISH_DENMARK: return "da_DK";
1718           }
1719         return "da";
1720       case LANG_DARI:
1721         /* FIXME: Adjust this when such locales appear on Unix.  */
1722         switch (sub)
1723           {
1724           case SUBLANG_DARI_AFGHANISTAN: return "prs_AF";
1725           }
1726         return "prs";
1727       case LANG_DIVEHI:
1728         switch (sub)
1729           {
1730           case SUBLANG_DIVEHI_MALDIVES: return "dv_MV";
1731           }
1732         return "dv";
1733       case LANG_DUTCH:
1734         switch (sub)
1735           {
1736           case SUBLANG_DUTCH: return "nl_NL";
1737           case SUBLANG_DUTCH_BELGIAN: /* FLEMISH, VLAAMS */ return "nl_BE";
1738           case SUBLANG_DUTCH_SURINAM: return "nl_SR";
1739           }
1740         return "nl";
1741       case LANG_EDO:
1742         switch (sub)
1743           {
1744           case SUBLANG_DEFAULT: return "bin_NG";
1745           }
1746         return "bin";
1747       case LANG_ENGLISH:
1748         switch (sub)
1749           {
1750           /* SUBLANG_ENGLISH_US == SUBLANG_DEFAULT. Heh. I thought
1751            * English was the language spoken in England.
1752            * Oh well.
1753            */
1754           case SUBLANG_ENGLISH_US: return "en_US";
1755           case SUBLANG_ENGLISH_UK: return "en_GB";
1756           case SUBLANG_ENGLISH_AUS: return "en_AU";
1757           case SUBLANG_ENGLISH_CAN: return "en_CA";
1758           case SUBLANG_ENGLISH_NZ: return "en_NZ";
1759           case SUBLANG_ENGLISH_EIRE: return "en_IE";
1760           case SUBLANG_ENGLISH_SOUTH_AFRICA: return "en_ZA";
1761           case SUBLANG_ENGLISH_JAMAICA: return "en_JM";
1762           case SUBLANG_ENGLISH_CARIBBEAN: return "en_GD"; /* Grenada? */
1763           case SUBLANG_ENGLISH_BELIZE: return "en_BZ";
1764           case SUBLANG_ENGLISH_TRINIDAD: return "en_TT";
1765           case SUBLANG_ENGLISH_ZIMBABWE: return "en_ZW";
1766           case SUBLANG_ENGLISH_PHILIPPINES: return "en_PH";
1767           case SUBLANG_ENGLISH_INDONESIA: return "en_ID";
1768           case SUBLANG_ENGLISH_HONGKONG: return "en_HK";
1769           case SUBLANG_ENGLISH_INDIA: return "en_IN";
1770           case SUBLANG_ENGLISH_MALAYSIA: return "en_MY";
1771           case SUBLANG_ENGLISH_SINGAPORE: return "en_SG";
1772           }
1773         return "en";
1774       case LANG_ESTONIAN:
1775         switch (sub)
1776           {
1777           case SUBLANG_ESTONIAN_ESTONIA: return "et_EE";
1778           }
1779         return "et";
1780       case LANG_FAEROESE:
1781         switch (sub)
1782           {
1783           case SUBLANG_FAEROESE_FAROE_ISLANDS: return "fo_FO";
1784           }
1785         return "fo";
1786       case LANG_FARSI:
1787         switch (sub)
1788           {
1789           case SUBLANG_FARSI_IRAN: return "fa_IR";
1790           }
1791         return "fa";
1792       case LANG_FINNISH:
1793         switch (sub)
1794           {
1795           case SUBLANG_FINNISH_FINLAND: return "fi_FI";
1796           }
1797         return "fi";
1798       case LANG_FRENCH:
1799         switch (sub)
1800           {
1801           case SUBLANG_FRENCH: return "fr_FR";
1802           case SUBLANG_FRENCH_BELGIAN: /* WALLOON */ return "fr_BE";
1803           case SUBLANG_FRENCH_CANADIAN: return "fr_CA";
1804           case SUBLANG_FRENCH_SWISS: return "fr_CH";
1805           case SUBLANG_FRENCH_LUXEMBOURG: return "fr_LU";
1806           case SUBLANG_FRENCH_MONACO: return "fr_MC";
1807           case SUBLANG_FRENCH_WESTINDIES: return "fr"; /* Caribbean? */
1808           case SUBLANG_FRENCH_REUNION: return "fr_RE";
1809           case SUBLANG_FRENCH_CONGO: return "fr_CG";
1810           case SUBLANG_FRENCH_SENEGAL: return "fr_SN";
1811           case SUBLANG_FRENCH_CAMEROON: return "fr_CM";
1812           case SUBLANG_FRENCH_COTEDIVOIRE: return "fr_CI";
1813           case SUBLANG_FRENCH_MALI: return "fr_ML";
1814           case SUBLANG_FRENCH_MOROCCO: return "fr_MA";
1815           case SUBLANG_FRENCH_HAITI: return "fr_HT";
1816           }
1817         return "fr";
1818       case LANG_FRISIAN:
1819         switch (sub)
1820           {
1821           case SUBLANG_FRISIAN_NETHERLANDS: return "fy_NL";
1822           }
1823         return "fy";
1824       case LANG_FULFULDE:
1825         /* Spoken in Nigeria, Guinea, Senegal, Mali, Niger, Cameroon, Benin.  */
1826         switch (sub)
1827           {
1828           case SUBLANG_DEFAULT: return "ff_NG";
1829           }
1830         return "ff";
1831       case LANG_GAELIC:
1832         switch (sub)
1833           {
1834           case 0x01: /* SCOTTISH */
1835             /* old, superseded by LANG_SCOTTISH_GAELIC */
1836             return "gd_GB";
1837           case SUBLANG_IRISH_IRELAND: return "ga_IE";
1838           }
1839         return "ga";
1840       case LANG_GALICIAN:
1841         switch (sub)
1842           {
1843           case SUBLANG_GALICIAN_SPAIN: return "gl_ES";
1844           }
1845         return "gl";
1846       case LANG_GEORGIAN:
1847         switch (sub)
1848           {
1849           case SUBLANG_GEORGIAN_GEORGIA: return "ka_GE";
1850           }
1851         return "ka";
1852       case LANG_GERMAN:
1853         switch (sub)
1854           {
1855           case SUBLANG_GERMAN: return "de_DE";
1856           case SUBLANG_GERMAN_SWISS: return "de_CH";
1857           case SUBLANG_GERMAN_AUSTRIAN: return "de_AT";
1858           case SUBLANG_GERMAN_LUXEMBOURG: return "de_LU";
1859           case SUBLANG_GERMAN_LIECHTENSTEIN: return "de_LI";
1860           }
1861         return "de";
1862       case LANG_GREEK:
1863         switch (sub)
1864           {
1865           case SUBLANG_GREEK_GREECE: return "el_GR";
1866           }
1867         return "el";
1868       case LANG_GREENLANDIC:
1869         switch (sub)
1870           {
1871           case SUBLANG_GREENLANDIC_GREENLAND: return "kl_GL";
1872           }
1873         return "kl";
1874       case LANG_GUARANI:
1875         switch (sub)
1876           {
1877           case SUBLANG_DEFAULT: return "gn_PY";
1878           }
1879         return "gn";
1880       case LANG_GUJARATI:
1881         switch (sub)
1882           {
1883           case SUBLANG_GUJARATI_INDIA: return "gu_IN";
1884           }
1885         return "gu";
1886       case LANG_HAUSA:
1887         switch (sub)
1888           {
1889           case 0x1f: return "ha";
1890           case SUBLANG_HAUSA_NIGERIA_LATIN: return "ha_NG";
1891           }
1892         return "ha";
1893       case LANG_HAWAIIAN:
1894         /* FIXME: Do they mean Hawaiian ("haw_US", 1000 speakers)
1895            or Hawaii Creole English ("cpe_US", 600000 speakers)?  */
1896         switch (sub)
1897           {
1898           case SUBLANG_DEFAULT: return "cpe_US";
1899           }
1900         return "cpe";
1901       case LANG_HEBREW:
1902         switch (sub)
1903           {
1904           case SUBLANG_HEBREW_ISRAEL: return "he_IL";
1905           }
1906         return "he";
1907       case LANG_HINDI:
1908         switch (sub)
1909           {
1910           case SUBLANG_HINDI_INDIA: return "hi_IN";
1911           }
1912         return "hi";
1913       case LANG_HUNGARIAN:
1914         switch (sub)
1915           {
1916           case SUBLANG_HUNGARIAN_HUNGARY: return "hu_HU";
1917           }
1918         return "hu";
1919       case LANG_IBIBIO:
1920         switch (sub)
1921           {
1922           case SUBLANG_DEFAULT: return "nic_NG";
1923           }
1924         return "nic";
1925       case LANG_ICELANDIC:
1926         switch (sub)
1927           {
1928           case SUBLANG_ICELANDIC_ICELAND: return "is_IS";
1929           }
1930         return "is";
1931       case LANG_IGBO:
1932         switch (sub)
1933           {
1934           case SUBLANG_IGBO_NIGERIA: return "ig_NG";
1935           }
1936         return "ig";
1937       case LANG_INDONESIAN:
1938         switch (sub)
1939           {
1940           case SUBLANG_INDONESIAN_INDONESIA: return "id_ID";
1941           }
1942         return "id";
1943       case LANG_INUKTITUT:
1944         switch (sub)
1945           {
1946           case 0x1e: return "iu"; /* syllabic */
1947           case SUBLANG_INUKTITUT_CANADA: return "iu_CA"; /* syllabic */
1948           case 0x1f: return "iu@latin";
1949           case SUBLANG_INUKTITUT_CANADA_LATIN: return "iu_CA@latin";
1950           }
1951         return "iu";
1952       case LANG_ITALIAN:
1953         switch (sub)
1954           {
1955           case SUBLANG_ITALIAN: return "it_IT";
1956           case SUBLANG_ITALIAN_SWISS: return "it_CH";
1957           }
1958         return "it";
1959       case LANG_JAPANESE:
1960         switch (sub)
1961           {
1962           case SUBLANG_JAPANESE_JAPAN: return "ja_JP";
1963           }
1964         return "ja";
1965       case LANG_KANNADA:
1966         switch (sub)
1967           {
1968           case SUBLANG_KANNADA_INDIA: return "kn_IN";
1969           }
1970         return "kn";
1971       case LANG_KANURI:
1972         switch (sub)
1973           {
1974           case SUBLANG_DEFAULT: return "kr_NG";
1975           }
1976         return "kr";
1977       case LANG_KASHMIRI:
1978         switch (sub)
1979           {
1980           case SUBLANG_DEFAULT: return "ks_PK";
1981           case SUBLANG_KASHMIRI_INDIA: return "ks_IN";
1982           }
1983         return "ks";
1984       case LANG_KAZAK:
1985         switch (sub)
1986           {
1987           case SUBLANG_KAZAK_KAZAKHSTAN: return "kk_KZ";
1988           }
1989         return "kk";
1990       case LANG_KICHE:
1991         /* FIXME: Adjust this when such locales appear on Unix.  */
1992         switch (sub)
1993           {
1994           case SUBLANG_KICHE_GUATEMALA: return "qut_GT";
1995           }
1996         return "qut";
1997       case LANG_KINYARWANDA:
1998         switch (sub)
1999           {
2000           case SUBLANG_KINYARWANDA_RWANDA: return "rw_RW";
2001           }
2002         return "rw";
2003       case LANG_KONKANI:
2004         /* FIXME: Adjust this when such locales appear on Unix.  */
2005         switch (sub)
2006           {
2007           case SUBLANG_KONKANI_INDIA: return "kok_IN";
2008           }
2009         return "kok";
2010       case LANG_KOREAN:
2011         switch (sub)
2012           {
2013           case SUBLANG_DEFAULT: return "ko_KR";
2014           }
2015         return "ko";
2016       case LANG_KYRGYZ:
2017         switch (sub)
2018           {
2019           case SUBLANG_KYRGYZ_KYRGYZSTAN: return "ky_KG";
2020           }
2021         return "ky";
2022       case LANG_LAO:
2023         switch (sub)
2024           {
2025           case SUBLANG_LAO_LAOS: return "lo_LA";
2026           }
2027         return "lo";
2028       case LANG_LATIN:
2029         switch (sub)
2030           {
2031           case SUBLANG_DEFAULT: return "la_VA";
2032           }
2033         return "la";
2034       case LANG_LATVIAN:
2035         switch (sub)
2036           {
2037           case SUBLANG_LATVIAN_LATVIA: return "lv_LV";
2038           }
2039         return "lv";
2040       case LANG_LITHUANIAN:
2041         switch (sub)
2042           {
2043           case SUBLANG_LITHUANIAN_LITHUANIA: return "lt_LT";
2044           }
2045         return "lt";
2046       case LANG_LUXEMBOURGISH:
2047         switch (sub)
2048           {
2049           case SUBLANG_LUXEMBOURGISH_LUXEMBOURG: return "lb_LU";
2050           }
2051         return "lb";
2052       case LANG_MACEDONIAN:
2053         switch (sub)
2054           {
2055           case SUBLANG_MACEDONIAN_MACEDONIA: return "mk_MK";
2056           }
2057         return "mk";
2058       case LANG_MALAY:
2059         switch (sub)
2060           {
2061           case SUBLANG_MALAY_MALAYSIA: return "ms_MY";
2062           case SUBLANG_MALAY_BRUNEI_DARUSSALAM: return "ms_BN";
2063           }
2064         return "ms";
2065       case LANG_MALAYALAM:
2066         switch (sub)
2067           {
2068           case SUBLANG_MALAYALAM_INDIA: return "ml_IN";
2069           }
2070         return "ml";
2071       case LANG_MALTESE:
2072         switch (sub)
2073           {
2074           case SUBLANG_MALTESE_MALTA: return "mt_MT";
2075           }
2076         return "mt";
2077       case LANG_MANIPURI:
2078         /* FIXME: Adjust this when such locales appear on Unix.  */
2079         switch (sub)
2080           {
2081           case SUBLANG_DEFAULT: return "mni_IN";
2082           }
2083         return "mni";
2084       case LANG_MAORI:
2085         switch (sub)
2086           {
2087           case SUBLANG_MAORI_NEW_ZEALAND: return "mi_NZ";
2088           }
2089         return "mi";
2090       case LANG_MAPUDUNGUN:
2091         switch (sub)
2092           {
2093           case SUBLANG_MAPUDUNGUN_CHILE: return "arn_CL";
2094           }
2095         return "arn";
2096       case LANG_MARATHI:
2097         switch (sub)
2098           {
2099           case SUBLANG_MARATHI_INDIA: return "mr_IN";
2100           }
2101         return "mr";
2102       case LANG_MOHAWK:
2103         switch (sub)
2104           {
2105           case SUBLANG_MOHAWK_CANADA: return "moh_CA";
2106           }
2107         return "moh";
2108       case LANG_MONGOLIAN:
2109         switch (sub)
2110           {
2111           case SUBLANG_MONGOLIAN_CYRILLIC_MONGOLIA: case 0x1e: return "mn_MN";
2112           case SUBLANG_MONGOLIAN_PRC: case 0x1f: return "mn_CN";
2113           }
2114         return "mn"; /* Ambiguous: could be "mn_CN" or "mn_MN".  */
2115       case LANG_NEPALI:
2116         switch (sub)
2117           {
2118           case SUBLANG_NEPALI_NEPAL: return "ne_NP";
2119           case SUBLANG_NEPALI_INDIA: return "ne_IN";
2120           }
2121         return "ne";
2122       case LANG_NORWEGIAN:
2123         switch (sub)
2124           {
2125           case 0x1f: return "nb";
2126           case SUBLANG_NORWEGIAN_BOKMAL: return "nb_NO";
2127           case 0x1e: return "nn";
2128           case SUBLANG_NORWEGIAN_NYNORSK: return "nn_NO";
2129           }
2130         return "no";
2131       case LANG_OCCITAN:
2132         switch (sub)
2133           {
2134           case SUBLANG_OCCITAN_FRANCE: return "oc_FR";
2135           }
2136         return "oc";
2137       case LANG_ORIYA:
2138         switch (sub)
2139           {
2140           case SUBLANG_ORIYA_INDIA: return "or_IN";
2141           }
2142         return "or";
2143       case LANG_OROMO:
2144         switch (sub)
2145           {
2146           case SUBLANG_DEFAULT: return "om_ET";
2147           }
2148         return "om";
2149       case LANG_PAPIAMENTU:
2150         switch (sub)
2151           {
2152           case SUBLANG_DEFAULT: return "pap_AN";
2153           }
2154         return "pap";
2155       case LANG_PASHTO:
2156         switch (sub)
2157           {
2158           case SUBLANG_PASHTO_AFGHANISTAN: return "ps_AF";
2159           }
2160         return "ps"; /* Ambiguous: could be "ps_PK" or "ps_AF".  */
2161       case LANG_POLISH:
2162         switch (sub)
2163           {
2164           case SUBLANG_POLISH_POLAND: return "pl_PL";
2165           }
2166         return "pl";
2167       case LANG_PORTUGUESE:
2168         switch (sub)
2169           {
2170           /* Hmm. SUBLANG_PORTUGUESE_BRAZILIAN == SUBLANG_DEFAULT.
2171              Same phenomenon as SUBLANG_ENGLISH_US == SUBLANG_DEFAULT. */
2172           case SUBLANG_PORTUGUESE_BRAZILIAN: return "pt_BR";
2173           case SUBLANG_PORTUGUESE: return "pt_PT";
2174           }
2175         return "pt";
2176       case LANG_PUNJABI:
2177         switch (sub)
2178           {
2179           case SUBLANG_PUNJABI_INDIA: return "pa_IN"; /* Gurmukhi script */
2180           case SUBLANG_PUNJABI_PAKISTAN: return "pa_PK"; /* Arabic script */
2181           }
2182         return "pa";
2183       case LANG_QUECHUA:
2184         /* Note: Microsoft uses the non-ISO language code "quz".  */
2185         switch (sub)
2186           {
2187           case SUBLANG_QUECHUA_BOLIVIA: return "qu_BO";
2188           case SUBLANG_QUECHUA_ECUADOR: return "qu_EC";
2189           case SUBLANG_QUECHUA_PERU: return "qu_PE";
2190           }
2191         return "qu";
2192       case LANG_ROMANIAN:
2193         switch (sub)
2194           {
2195           case SUBLANG_ROMANIAN_ROMANIA: return "ro_RO";
2196           case SUBLANG_ROMANIAN_MOLDOVA: return "ro_MD";
2197           }
2198         return "ro";
2199       case LANG_ROMANSH:
2200         switch (sub)
2201           {
2202           case SUBLANG_ROMANSH_SWITZERLAND: return "rm_CH";
2203           }
2204         return "rm";
2205       case LANG_RUSSIAN:
2206         switch (sub)
2207           {
2208           case SUBLANG_RUSSIAN_RUSSIA: return "ru_RU";
2209           case SUBLANG_RUSSIAN_MOLDAVIA: return "ru_MD";
2210           }
2211         return "ru"; /* Ambiguous: could be "ru_RU" or "ru_UA" or "ru_MD".  */
2212       case LANG_SAMI:
2213         switch (sub)
2214           {
2215           /* Northern Sami */
2216           case 0x00: return "se";
2217           case SUBLANG_SAMI_NORTHERN_NORWAY: return "se_NO";
2218           case SUBLANG_SAMI_NORTHERN_SWEDEN: return "se_SE";
2219           case SUBLANG_SAMI_NORTHERN_FINLAND: return "se_FI";
2220           /* Lule Sami */
2221           case 0x1f: return "smj";
2222           case SUBLANG_SAMI_LULE_NORWAY: return "smj_NO";
2223           case SUBLANG_SAMI_LULE_SWEDEN: return "smj_SE";
2224           /* Southern Sami */
2225           case 0x1e: return "sma";
2226           case SUBLANG_SAMI_SOUTHERN_NORWAY: return "sma_NO";
2227           case SUBLANG_SAMI_SOUTHERN_SWEDEN: return "sma_SE";
2228           /* Skolt Sami */
2229           case 0x1d: return "sms";
2230           case SUBLANG_SAMI_SKOLT_FINLAND: return "sms_FI";
2231           /* Inari Sami */
2232           case 0x1c: return "smn";
2233           case SUBLANG_SAMI_INARI_FINLAND: return "smn_FI";
2234           }
2235         return "se"; /* or "smi"? */
2236       case LANG_SANSKRIT:
2237         switch (sub)
2238           {
2239           case SUBLANG_SANSKRIT_INDIA: return "sa_IN";
2240           }
2241         return "sa";
2242       case LANG_SCOTTISH_GAELIC:
2243         switch (sub)
2244           {
2245           case SUBLANG_DEFAULT: return "gd_GB";
2246           }
2247         return "gd";
2248       case LANG_SINDHI:
2249         switch (sub)
2250           {
2251           case SUBLANG_SINDHI_INDIA: return "sd_IN";
2252           case SUBLANG_SINDHI_PAKISTAN: return "sd_PK";
2253           /*case SUBLANG_SINDHI_AFGHANISTAN: return "sd_AF";*/
2254           }
2255         return "sd";
2256       case LANG_SINHALESE:
2257         switch (sub)
2258           {
2259           case SUBLANG_SINHALESE_SRI_LANKA: return "si_LK";
2260           }
2261         return "si";
2262       case LANG_SLOVAK:
2263         switch (sub)
2264           {
2265           case SUBLANG_SLOVAK_SLOVAKIA: return "sk_SK";
2266           }
2267         return "sk";
2268       case LANG_SLOVENIAN:
2269         switch (sub)
2270           {
2271           case SUBLANG_SLOVENIAN_SLOVENIA: return "sl_SI";
2272           }
2273         return "sl";
2274       case LANG_SOMALI:
2275         switch (sub)
2276           {
2277           case SUBLANG_DEFAULT: return "so_SO";
2278           }
2279         return "so";
2280       case LANG_SORBIAN:
2281         /* FIXME: Adjust this when such locales appear on Unix.  */
2282         switch (sub)
2283           {
2284           /* Upper Sorbian */
2285           case 0x00: return "hsb";
2286           case SUBLANG_UPPER_SORBIAN_GERMANY: return "hsb_DE";
2287           /* Lower Sorbian */
2288           case 0x1f: return "dsb";
2289           case SUBLANG_LOWER_SORBIAN_GERMANY: return "dsb_DE";
2290           }
2291         return "wen";
2292       case LANG_SOTHO:
2293         /* <https://msdn.microsoft.com/en-us/library/dd318693.aspx> calls
2294            it "Sesotho sa Leboa"; according to
2295            <https://www.ethnologue.com/show_language.asp?code=nso>
2296            <https://www.ethnologue.com/show_language.asp?code=sot>
2297            it's the same as Northern Sotho.  */
2298         switch (sub)
2299           {
2300           case SUBLANG_SOTHO_SOUTH_AFRICA: return "nso_ZA";
2301           }
2302         return "nso";
2303       case LANG_SPANISH:
2304         switch (sub)
2305           {
2306           case SUBLANG_SPANISH: return "es_ES";
2307           case SUBLANG_SPANISH_MEXICAN: return "es_MX";
2308           case SUBLANG_SPANISH_MODERN:
2309             return "es_ES@modern";      /* not seen on Unix */
2310           case SUBLANG_SPANISH_GUATEMALA: return "es_GT";
2311           case SUBLANG_SPANISH_COSTA_RICA: return "es_CR";
2312           case SUBLANG_SPANISH_PANAMA: return "es_PA";
2313           case SUBLANG_SPANISH_DOMINICAN_REPUBLIC: return "es_DO";
2314           case SUBLANG_SPANISH_VENEZUELA: return "es_VE";
2315           case SUBLANG_SPANISH_COLOMBIA: return "es_CO";
2316           case SUBLANG_SPANISH_PERU: return "es_PE";
2317           case SUBLANG_SPANISH_ARGENTINA: return "es_AR";
2318           case SUBLANG_SPANISH_ECUADOR: return "es_EC";
2319           case SUBLANG_SPANISH_CHILE: return "es_CL";
2320           case SUBLANG_SPANISH_URUGUAY: return "es_UY";
2321           case SUBLANG_SPANISH_PARAGUAY: return "es_PY";
2322           case SUBLANG_SPANISH_BOLIVIA: return "es_BO";
2323           case SUBLANG_SPANISH_EL_SALVADOR: return "es_SV";
2324           case SUBLANG_SPANISH_HONDURAS: return "es_HN";
2325           case SUBLANG_SPANISH_NICARAGUA: return "es_NI";
2326           case SUBLANG_SPANISH_PUERTO_RICO: return "es_PR";
2327           case SUBLANG_SPANISH_US: return "es_US";
2328           }
2329         return "es";
2330       case LANG_SUTU:
2331         switch (sub)
2332           {
2333           case SUBLANG_DEFAULT: return "bnt_TZ"; /* or "st_LS" or "nso_ZA"? */
2334           }
2335         return "bnt";
2336       case LANG_SWAHILI:
2337         switch (sub)
2338           {
2339           case SUBLANG_SWAHILI_KENYA: return "sw_KE";
2340           }
2341         return "sw";
2342       case LANG_SWEDISH:
2343         switch (sub)
2344           {
2345           case SUBLANG_SWEDISH_SWEDEN: return "sv_SE";
2346           case SUBLANG_SWEDISH_FINLAND: return "sv_FI";
2347           }
2348         return "sv";
2349       case LANG_SYRIAC:
2350         switch (sub)
2351           {
2352           case SUBLANG_SYRIAC_SYRIA: return "syr_SY"; /* An extinct language.  */
2353           }
2354         return "syr";
2355       case LANG_TAGALOG:
2356         switch (sub)
2357           {
2358           case SUBLANG_TAGALOG_PHILIPPINES: return "tl_PH"; /* or "fil_PH"? */
2359           }
2360         return "tl"; /* or "fil"? */
2361       case LANG_TAJIK:
2362         switch (sub)
2363           {
2364           case 0x1f: return "tg";
2365           case SUBLANG_TAJIK_TAJIKISTAN: return "tg_TJ";
2366           }
2367         return "tg";
2368       case LANG_TAMAZIGHT:
2369         /* Note: Microsoft uses the non-ISO language code "tmz".  */
2370         switch (sub)
2371           {
2372           /* FIXME: Adjust this when Tamazight locales appear on Unix.  */
2373           case SUBLANG_TAMAZIGHT_ARABIC: return "ber_MA@arabic";
2374           case 0x1f: return "ber@latin";
2375           case SUBLANG_TAMAZIGHT_ALGERIA_LATIN: return "ber_DZ@latin";
2376           }
2377         return "ber";
2378       case LANG_TAMIL:
2379         switch (sub)
2380           {
2381           case SUBLANG_TAMIL_INDIA: return "ta_IN";
2382           }
2383         return "ta"; /* Ambiguous: could be "ta_IN" or "ta_LK" or "ta_SG".  */
2384       case LANG_TATAR:
2385         switch (sub)
2386           {
2387           case SUBLANG_TATAR_RUSSIA: return "tt_RU";
2388           }
2389         return "tt";
2390       case LANG_TELUGU:
2391         switch (sub)
2392           {
2393           case SUBLANG_TELUGU_INDIA: return "te_IN";
2394           }
2395         return "te";
2396       case LANG_THAI:
2397         switch (sub)
2398           {
2399           case SUBLANG_THAI_THAILAND: return "th_TH";
2400           }
2401         return "th";
2402       case LANG_TIBETAN:
2403         switch (sub)
2404           {
2405           case SUBLANG_TIBETAN_PRC:
2406             /* Most Tibetans would not like "bo_CN".  But Tibet does not yet
2407                have a country code of its own.  */
2408             return "bo";
2409           case SUBLANG_TIBETAN_BHUTAN: return "bo_BT";
2410           }
2411         return "bo";
2412       case LANG_TIGRINYA:
2413         switch (sub)
2414           {
2415           case SUBLANG_TIGRINYA_ETHIOPIA: return "ti_ET";
2416           case SUBLANG_TIGRINYA_ERITREA: return "ti_ER";
2417           }
2418         return "ti";
2419       case LANG_TSONGA:
2420         switch (sub)
2421           {
2422           case SUBLANG_DEFAULT: return "ts_ZA";
2423           }
2424         return "ts";
2425       case LANG_TSWANA:
2426         /* Spoken in South Africa, Botswana.  */
2427         switch (sub)
2428           {
2429           case SUBLANG_TSWANA_SOUTH_AFRICA: return "tn_ZA";
2430           }
2431         return "tn";
2432       case LANG_TURKISH:
2433         switch (sub)
2434           {
2435           case SUBLANG_TURKISH_TURKEY: return "tr_TR";
2436           }
2437         return "tr";
2438       case LANG_TURKMEN:
2439         switch (sub)
2440           {
2441           case SUBLANG_TURKMEN_TURKMENISTAN: return "tk_TM";
2442           }
2443         return "tk";
2444       case LANG_UIGHUR:
2445         switch (sub)
2446           {
2447           case SUBLANG_UIGHUR_PRC: return "ug_CN";
2448           }
2449         return "ug";
2450       case LANG_UKRAINIAN:
2451         switch (sub)
2452           {
2453           case SUBLANG_UKRAINIAN_UKRAINE: return "uk_UA";
2454           }
2455         return "uk";
2456       case LANG_URDU:
2457         switch (sub)
2458           {
2459           case SUBLANG_URDU_PAKISTAN: return "ur_PK";
2460           case SUBLANG_URDU_INDIA: return "ur_IN";
2461           }
2462         return "ur";
2463       case LANG_UZBEK:
2464         switch (sub)
2465           {
2466           case 0x1f: return "uz";
2467           case SUBLANG_UZBEK_LATIN: return "uz_UZ";
2468           case 0x1e: return "uz@cyrillic";
2469           case SUBLANG_UZBEK_CYRILLIC: return "uz_UZ@cyrillic";
2470           }
2471         return "uz";
2472       case LANG_VENDA:
2473         switch (sub)
2474           {
2475           case SUBLANG_DEFAULT: return "ve_ZA";
2476           }
2477         return "ve";
2478       case LANG_VIETNAMESE:
2479         switch (sub)
2480           {
2481           case SUBLANG_VIETNAMESE_VIETNAM: return "vi_VN";
2482           }
2483         return "vi";
2484       case LANG_WELSH:
2485         switch (sub)
2486           {
2487           case SUBLANG_WELSH_UNITED_KINGDOM: return "cy_GB";
2488           }
2489         return "cy";
2490       case LANG_WOLOF:
2491         switch (sub)
2492           {
2493           case SUBLANG_WOLOF_SENEGAL: return "wo_SN";
2494           }
2495         return "wo";
2496       case LANG_XHOSA:
2497         switch (sub)
2498           {
2499           case SUBLANG_XHOSA_SOUTH_AFRICA: return "xh_ZA";
2500           }
2501         return "xh";
2502       case LANG_YAKUT:
2503         switch (sub)
2504           {
2505           case SUBLANG_YAKUT_RUSSIA: return "sah_RU";
2506           }
2507         return "sah";
2508       case LANG_YI:
2509         switch (sub)
2510           {
2511           case SUBLANG_YI_PRC: return "ii_CN";
2512           }
2513         return "ii";
2514       case LANG_YIDDISH:
2515         switch (sub)
2516           {
2517           case SUBLANG_DEFAULT: return "yi_IL";
2518           }
2519         return "yi";
2520       case LANG_YORUBA:
2521         switch (sub)
2522           {
2523           case SUBLANG_YORUBA_NIGERIA: return "yo_NG";
2524           }
2525         return "yo";
2526       case LANG_ZULU:
2527         switch (sub)
2528           {
2529           case SUBLANG_ZULU_SOUTH_AFRICA: return "zu_ZA";
2530           }
2531         return "zu";
2532       default: return "C";
2533       }
2534   }
2535 }
2536
2537 # if !defined IN_LIBINTL
2538 static
2539 # endif
2540 const char *
2541 gl_locale_name_from_win32_LCID (LCID lcid)
2542 {
2543   LANGID langid;
2544
2545   /* Strip off the sorting rules, keep only the language part.  */
2546   langid = LANGIDFROMLCID (lcid);
2547
2548   return gl_locale_name_from_win32_LANGID (langid);
2549 }
2550
2551 # ifdef WINDOWS_NATIVE
2552
2553 /* Two variables to interface between get_lcid and the EnumLocales
2554    callback function below.  */
2555 static LCID found_lcid;
2556 static char lname[LC_MAX * (LOCALE_NAME_MAX_LENGTH + 1) + 1];
2557
2558 /* Callback function for EnumLocales.  */
2559 static BOOL CALLBACK
2560 enum_locales_fn (LPTSTR locale_num_str)
2561 {
2562   char *endp;
2563   char locval[2 * LOCALE_NAME_MAX_LENGTH + 1 + 1];
2564   LCID try_lcid = strtoul (locale_num_str, &endp, 16);
2565
2566   if (GetLocaleInfo (try_lcid, LOCALE_SENGLANGUAGE,
2567                     locval, LOCALE_NAME_MAX_LENGTH))
2568     {
2569       strcat (locval, "_");
2570       if (GetLocaleInfo (try_lcid, LOCALE_SENGCOUNTRY,
2571                         locval + strlen (locval), LOCALE_NAME_MAX_LENGTH))
2572        {
2573          size_t locval_len = strlen (locval);
2574
2575          if (strncmp (locval, lname, locval_len) == 0
2576              && (lname[locval_len] == '.'
2577                  || lname[locval_len] == '\0'))
2578            {
2579              found_lcid = try_lcid;
2580              return FALSE;
2581            }
2582        }
2583     }
2584   return TRUE;
2585 }
2586
2587 /* This lock protects the get_lcid against multiple simultaneous calls.  */
2588 gl_lock_define_initialized(static, get_lcid_lock)
2589
2590 /* Return the Locale ID (LCID) number given the locale's name, a
2591    string, in LOCALE_NAME.  This works by enumerating all the locales
2592    supported by the system, until we find one whose name matches
2593    LOCALE_NAME.  */
2594 static LCID
2595 get_lcid (const char *locale_name)
2596 {
2597   /* A simple cache.  */
2598   static LCID last_lcid;
2599   static char last_locale[1000];
2600
2601   /* Lock while looking for an LCID, to protect access to static
2602      variables: last_lcid, last_locale, found_lcid, and lname.  */
2603   gl_lock_lock (get_lcid_lock);
2604   if (last_lcid > 0 && strcmp (locale_name, last_locale) == 0)
2605     {
2606       gl_lock_unlock (get_lcid_lock);
2607       return last_lcid;
2608     }
2609   strncpy (lname, locale_name, sizeof (lname) - 1);
2610   lname[sizeof (lname) - 1] = '\0';
2611   found_lcid = 0;
2612   EnumSystemLocales (enum_locales_fn, LCID_SUPPORTED);
2613   if (found_lcid > 0)
2614     {
2615       last_lcid = found_lcid;
2616       strcpy (last_locale, locale_name);
2617     }
2618   gl_lock_unlock (get_lcid_lock);
2619   return found_lcid;
2620 }
2621
2622 # endif
2623 #endif
2624
2625
2626 #if HAVE_USELOCALE /* glibc, Mac OS X, Solaris 11 OpenIndiana, or Solaris >= 11.4  */
2627
2628 /* Simple hash set of strings.  We don't want to drag in lots of hash table
2629    code here.  */
2630
2631 # define SIZE_BITS (sizeof (size_t) * CHAR_BIT)
2632
2633 /* A hash function for NUL-terminated char* strings using
2634    the method described by Bruno Haible.
2635    See https://www.haible.de/bruno/hashfunc.html.  */
2636 static size_t _GL_ATTRIBUTE_PURE
2637 string_hash (const void *x)
2638 {
2639   const char *s = (const char *) x;
2640   size_t h = 0;
2641
2642   for (; *s; s++)
2643     h = *s + ((h << 9) | (h >> (SIZE_BITS - 9)));
2644
2645   return h;
2646 }
2647
2648 /* A hash table of fixed size.  Multiple threads can access it read-only
2649    simultaneously, but only one thread can insert into it at the same time.  */
2650
2651 /* A node in a hash bucket collision list.  */
2652 struct struniq_hash_node
2653   {
2654     struct struniq_hash_node * volatile next;
2655     char contents[FLEXIBLE_ARRAY_MEMBER];
2656   };
2657
2658 # define STRUNIQ_HASH_TABLE_SIZE 257
2659 static struct struniq_hash_node * volatile struniq_hash_table[STRUNIQ_HASH_TABLE_SIZE]
2660   /* = { NULL, ..., NULL } */;
2661
2662 /* This lock protects the struniq_hash_table against multiple simultaneous
2663    insertions.  */
2664 gl_lock_define_initialized(static, struniq_lock)
2665
2666 /* Store a copy of the given string in a string pool with indefinite extent.
2667    Return a pointer to this copy.  */
2668 static const char *
2669 struniq (const char *string)
2670 {
2671   size_t hashcode = string_hash (string);
2672   size_t slot = hashcode % STRUNIQ_HASH_TABLE_SIZE;
2673   size_t size;
2674   struct struniq_hash_node *new_node;
2675   struct struniq_hash_node *p;
2676   for (p = struniq_hash_table[slot]; p != NULL; p = p->next)
2677     if (strcmp (p->contents, string) == 0)
2678       return p->contents;
2679   size = strlen (string) + 1;
2680   new_node =
2681     (struct struniq_hash_node *)
2682     malloc (FLEXSIZEOF (struct struniq_hash_node, contents, size));
2683   if (new_node == NULL)
2684     /* Out of memory.  Return a statically allocated string.  */
2685     return "C";
2686   memcpy (new_node->contents, string, size);
2687   /* Lock while inserting new_node.  */
2688   gl_lock_lock (struniq_lock);
2689   /* Check whether another thread already added the string while we were
2690      waiting on the lock.  */
2691   for (p = struniq_hash_table[slot]; p != NULL; p = p->next)
2692     if (strcmp (p->contents, string) == 0)
2693       {
2694         free (new_node);
2695         new_node = p;
2696         goto done;
2697       }
2698   /* Really insert new_node into the hash table.  Fill new_node entirely first,
2699      because other threads may be iterating over the linked list.  */
2700   new_node->next = struniq_hash_table[slot];
2701   struniq_hash_table[slot] = new_node;
2702  done:
2703   /* Unlock after new_node is inserted.  */
2704   gl_lock_unlock (struniq_lock);
2705   return new_node->contents;
2706 }
2707
2708 #endif
2709
2710
2711 #if HAVE_USELOCALE && HAVE_NAMELESS_LOCALES
2712
2713 /* The 'locale_t' object does not contain the names of the locale categories.
2714    We have to associate them with the object through a hash table.
2715    The hash table is defined in localename-table.[hc].  */
2716
2717 /* Returns the name of a given locale category in a given locale_t object,
2718    allocated as a string with indefinite extent.  */
2719 static const char *
2720 get_locale_t_name (int category, locale_t locale)
2721 {
2722   if (locale == LC_GLOBAL_LOCALE)
2723     {
2724       /* Query the global locale.  */
2725       const char *name = setlocale (category, NULL);
2726       if (name != NULL)
2727         return struniq (name);
2728       else
2729         /* Should normally not happen.  */
2730         return "";
2731     }
2732   else
2733     {
2734       /* Look up the names in the hash table.  */
2735       size_t hashcode = locale_hash_function (locale);
2736       size_t slot = hashcode % LOCALE_HASH_TABLE_SIZE;
2737       /* If the locale was not found in the table, return "".  This can
2738          happen if the application uses the original newlocale()/duplocale()
2739          functions instead of the overridden ones.  */
2740       const char *name = "";
2741       struct locale_hash_node *p;
2742       /* Lock while looking up the hash node.  */
2743       gl_rwlock_rdlock (locale_lock);
2744       for (p = locale_hash_table[slot]; p != NULL; p = p->next)
2745         if (p->locale == locale)
2746           {
2747             name = p->names.category_name[category];
2748             break;
2749           }
2750       gl_rwlock_unlock (locale_lock);
2751       return name;
2752     }
2753 }
2754
2755 # if !(defined newlocale && defined duplocale && defined freelocale)
2756 #  error "newlocale, duplocale, freelocale not being replaced as expected!"
2757 # endif
2758
2759 /* newlocale() override.  */
2760 locale_t
2761 newlocale (int category_mask, const char *name, locale_t base)
2762 #undef newlocale
2763 {
2764   struct locale_categories_names names;
2765   struct locale_hash_node *node;
2766   locale_t result;
2767
2768   /* Make sure name has indefinite extent.  */
2769   if (((LC_CTYPE_MASK | LC_NUMERIC_MASK | LC_TIME_MASK | LC_COLLATE_MASK
2770         | LC_MONETARY_MASK | LC_MESSAGES_MASK)
2771        & category_mask) != 0)
2772     name = struniq (name);
2773
2774   /* Determine the category names of the result.  */
2775   if (((LC_CTYPE_MASK | LC_NUMERIC_MASK | LC_TIME_MASK | LC_COLLATE_MASK
2776         | LC_MONETARY_MASK | LC_MESSAGES_MASK)
2777        & ~category_mask) == 0)
2778     {
2779       /* Use name, ignore base.  */
2780       int category;
2781
2782       name = struniq (name);
2783       for (category = 0; category < 6; category++)
2784         names.category_name[category] = name;
2785     }
2786   else
2787     {
2788       /* Use base, possibly also name.  */
2789       if (base == NULL)
2790         {
2791           int category;
2792
2793           for (category = 0; category < 6; category++)
2794             {
2795               int mask;
2796
2797               switch (category)
2798                 {
2799                 case LC_CTYPE:
2800                   mask = LC_CTYPE_MASK;
2801                   break;
2802                 case LC_NUMERIC:
2803                   mask = LC_NUMERIC_MASK;
2804                   break;
2805                 case LC_TIME:
2806                   mask = LC_TIME_MASK;
2807                   break;
2808                 case LC_COLLATE:
2809                   mask = LC_COLLATE_MASK;
2810                   break;
2811                 case LC_MONETARY:
2812                   mask = LC_MONETARY_MASK;
2813                   break;
2814                 case LC_MESSAGES:
2815                   mask = LC_MESSAGES_MASK;
2816                   break;
2817                 default:
2818                   abort ();
2819                 }
2820               names.category_name[category] =
2821                 ((mask & category_mask) != 0 ? name : "C");
2822             }
2823         }
2824       else if (base == LC_GLOBAL_LOCALE)
2825         {
2826           int category;
2827
2828           for (category = 0; category < 6; category++)
2829             {
2830               int mask;
2831
2832               switch (category)
2833                 {
2834                 case LC_CTYPE:
2835                   mask = LC_CTYPE_MASK;
2836                   break;
2837                 case LC_NUMERIC:
2838                   mask = LC_NUMERIC_MASK;
2839                   break;
2840                 case LC_TIME:
2841                   mask = LC_TIME_MASK;
2842                   break;
2843                 case LC_COLLATE:
2844                   mask = LC_COLLATE_MASK;
2845                   break;
2846                 case LC_MONETARY:
2847                   mask = LC_MONETARY_MASK;
2848                   break;
2849                 case LC_MESSAGES:
2850                   mask = LC_MESSAGES_MASK;
2851                   break;
2852                 default:
2853                   abort ();
2854                 }
2855               names.category_name[category] =
2856                 ((mask & category_mask) != 0
2857                  ? name
2858                  : get_locale_t_name (category, LC_GLOBAL_LOCALE));
2859             }
2860         }
2861       else
2862         {
2863           /* Look up the names of base in the hash table.  Like multiple calls
2864              of get_locale_t_name, but locking only once.  */
2865           struct locale_hash_node *p;
2866           int category;
2867
2868           /* Lock while looking up the hash node.  */
2869           gl_rwlock_rdlock (locale_lock);
2870           for (p = locale_hash_table[locale_hash_function (base) % LOCALE_HASH_TABLE_SIZE];
2871                p != NULL;
2872                p = p->next)
2873             if (p->locale == base)
2874               break;
2875
2876           for (category = 0; category < 6; category++)
2877             {
2878               int mask;
2879
2880               switch (category)
2881                 {
2882                 case LC_CTYPE:
2883                   mask = LC_CTYPE_MASK;
2884                   break;
2885                 case LC_NUMERIC:
2886                   mask = LC_NUMERIC_MASK;
2887                   break;
2888                 case LC_TIME:
2889                   mask = LC_TIME_MASK;
2890                   break;
2891                 case LC_COLLATE:
2892                   mask = LC_COLLATE_MASK;
2893                   break;
2894                 case LC_MONETARY:
2895                   mask = LC_MONETARY_MASK;
2896                   break;
2897                 case LC_MESSAGES:
2898                   mask = LC_MESSAGES_MASK;
2899                   break;
2900                 default:
2901                   abort ();
2902                 }
2903               names.category_name[category] =
2904                 ((mask & category_mask) != 0
2905                  ? name
2906                  : (p != NULL ? p->names.category_name[category] : ""));
2907             }
2908
2909           gl_rwlock_unlock (locale_lock);
2910         }
2911     }
2912
2913   node = (struct locale_hash_node *) malloc (sizeof (struct locale_hash_node));
2914   if (node == NULL)
2915     /* errno is set to ENOMEM.  */
2916     return NULL;
2917
2918   result = newlocale (category_mask, name, base);
2919   if (result == NULL)
2920     {
2921       int saved_errno = errno;
2922       free (node);
2923       errno = saved_errno;
2924       return NULL;
2925     }
2926
2927   /* Fill the hash node.  */
2928   node->locale = result;
2929   node->names = names;
2930
2931   /* Insert it in the hash table.  */
2932   {
2933     size_t hashcode = locale_hash_function (result);
2934     size_t slot = hashcode % LOCALE_HASH_TABLE_SIZE;
2935     struct locale_hash_node *p;
2936
2937     /* Lock while inserting the new node.  */
2938     gl_rwlock_wrlock (locale_lock);
2939     for (p = locale_hash_table[slot]; p != NULL; p = p->next)
2940       if (p->locale == result)
2941         {
2942           /* This can happen if the application uses the original freelocale()
2943              function instead of the overridden one.  */
2944           p->names = node->names;
2945           break;
2946         }
2947     if (p == NULL)
2948       {
2949         node->next = locale_hash_table[slot];
2950         locale_hash_table[slot] = node;
2951       }
2952
2953     gl_rwlock_unlock (locale_lock);
2954
2955     if (p != NULL)
2956       free (node);
2957   }
2958
2959   return result;
2960 }
2961
2962 /* duplocale() override.  */
2963 locale_t
2964 duplocale (locale_t locale)
2965 #undef duplocale
2966 {
2967   struct locale_hash_node *node;
2968   locale_t result;
2969
2970   if (locale == NULL)
2971     /* Invalid argument.  */
2972     abort ();
2973
2974   node = (struct locale_hash_node *) malloc (sizeof (struct locale_hash_node));
2975   if (node == NULL)
2976     /* errno is set to ENOMEM.  */
2977     return NULL;
2978
2979   result = duplocale (locale);
2980   if (result == NULL)
2981     {
2982       int saved_errno = errno;
2983       free (node);
2984       errno = saved_errno;
2985       return NULL;
2986     }
2987
2988   /* Fill the hash node.  */
2989   node->locale = result;
2990   if (locale == LC_GLOBAL_LOCALE)
2991     {
2992       int category;
2993
2994       for (category = 0; category < 6; category++)
2995         node->names.category_name[category] =
2996           get_locale_t_name (category, LC_GLOBAL_LOCALE);
2997
2998       /* Lock before inserting the new node.  */
2999       gl_rwlock_wrlock (locale_lock);
3000     }
3001   else
3002     {
3003       struct locale_hash_node *p;
3004
3005       /* Lock once, for the lookup and the insertion.  */
3006       gl_rwlock_wrlock (locale_lock);
3007
3008       for (p = locale_hash_table[locale_hash_function (locale) % LOCALE_HASH_TABLE_SIZE];
3009            p != NULL;
3010            p = p->next)
3011         if (p->locale == locale)
3012           break;
3013       if (p != NULL)
3014         node->names = p->names;
3015       else
3016         {
3017           /* This can happen if the application uses the original
3018              newlocale()/duplocale() functions instead of the overridden
3019              ones.  */
3020           int category;
3021
3022           for (category = 0; category < 6; category++)
3023             node->names.category_name[category] = "";
3024         }
3025     }
3026
3027   /* Insert it in the hash table.  */
3028   {
3029     size_t hashcode = locale_hash_function (result);
3030     size_t slot = hashcode % LOCALE_HASH_TABLE_SIZE;
3031     struct locale_hash_node *p;
3032
3033     for (p = locale_hash_table[slot]; p != NULL; p = p->next)
3034       if (p->locale == result)
3035         {
3036           /* This can happen if the application uses the original freelocale()
3037              function instead of the overridden one.  */
3038           p->names = node->names;
3039           break;
3040         }
3041     if (p == NULL)
3042       {
3043         node->next = locale_hash_table[slot];
3044         locale_hash_table[slot] = node;
3045       }
3046
3047     gl_rwlock_unlock (locale_lock);
3048
3049     if (p != NULL)
3050       free (node);
3051   }
3052
3053   return result;
3054 }
3055
3056 /* freelocale() override.  */
3057 void
3058 freelocale (locale_t locale)
3059 #undef freelocale
3060 {
3061   if (locale == NULL || locale == LC_GLOBAL_LOCALE)
3062     /* Invalid argument.  */
3063     abort ();
3064
3065   {
3066     size_t hashcode = locale_hash_function (locale);
3067     size_t slot = hashcode % LOCALE_HASH_TABLE_SIZE;
3068     struct locale_hash_node *found;
3069     struct locale_hash_node **p;
3070
3071     found = NULL;
3072     /* Lock while removing the hash node.  */
3073     gl_rwlock_wrlock (locale_lock);
3074     for (p = &locale_hash_table[slot]; *p != NULL; p = &(*p)->next)
3075       if ((*p)->locale == locale)
3076         {
3077           found = *p;
3078           *p = (*p)->next;
3079           break;
3080         }
3081     gl_rwlock_unlock (locale_lock);
3082     free (found);
3083   }
3084
3085   freelocale (locale);
3086 }
3087
3088 #endif
3089
3090
3091 #if defined IN_LIBINTL || HAVE_USELOCALE
3092
3093 /* Like gl_locale_name_thread, except that the result is not in storage of
3094    indefinite extent.  */
3095 # if !defined IN_LIBINTL
3096 static
3097 # endif
3098 const char *
3099 gl_locale_name_thread_unsafe (int category, const char *categoryname)
3100 {
3101 # if HAVE_USELOCALE
3102   {
3103     locale_t thread_locale = uselocale (NULL);
3104     if (thread_locale != LC_GLOBAL_LOCALE)
3105       {
3106 #  if __GLIBC__ >= 2 && !defined __UCLIBC__
3107         /* Work around an incorrect definition of the _NL_LOCALE_NAME macro in
3108            glibc < 2.12.
3109            See <https://sourceware.org/bugzilla/show_bug.cgi?id=10968>.  */
3110         const char *name =
3111           nl_langinfo (_NL_ITEM ((category), _NL_ITEM_INDEX (-1)));
3112         if (name[0] == '\0')
3113           /* Fallback code for glibc < 2.4, which did not implement
3114              nl_langinfo (_NL_LOCALE_NAME (category)).  */
3115           name = thread_locale->__names[category];
3116         return name;
3117 #  elif defined __linux__ && HAVE_LANGINFO_H && defined NL_LOCALE_NAME
3118         /* musl libc */
3119         return nl_langinfo_l (NL_LOCALE_NAME (category), thread_locale);
3120 #  elif (defined __FreeBSD__ || defined __DragonFly__) || (defined __APPLE__ && defined __MACH__)
3121         /* FreeBSD, Mac OS X */
3122         int mask;
3123
3124         switch (category)
3125           {
3126           case LC_CTYPE:
3127             mask = LC_CTYPE_MASK;
3128             break;
3129           case LC_NUMERIC:
3130             mask = LC_NUMERIC_MASK;
3131             break;
3132           case LC_TIME:
3133             mask = LC_TIME_MASK;
3134             break;
3135           case LC_COLLATE:
3136             mask = LC_COLLATE_MASK;
3137             break;
3138           case LC_MONETARY:
3139             mask = LC_MONETARY_MASK;
3140             break;
3141           case LC_MESSAGES:
3142             mask = LC_MESSAGES_MASK;
3143             break;
3144           default: /* We shouldn't get here.  */
3145             return "";
3146           }
3147         return querylocale (mask, thread_locale);
3148 #  elif defined __sun
3149 #   if HAVE_GETLOCALENAME_L
3150         /* Solaris >= 12.  */
3151         return getlocalename_l (category, thread_locale);
3152 #   elif HAVE_SOLARIS114_LOCALES
3153         /* Solaris >= 11.4.  */
3154         void *lcp = (*thread_locale)->core.data->lcp;
3155         if (lcp != NULL)
3156           switch (category)
3157             {
3158             case LC_CTYPE:
3159             case LC_NUMERIC:
3160             case LC_TIME:
3161             case LC_COLLATE:
3162             case LC_MONETARY:
3163             case LC_MESSAGES:
3164               return ((const char * const *) lcp)[category];
3165             default: /* We shouldn't get here.  */
3166               return "";
3167             }
3168 #   elif HAVE_NAMELESS_LOCALES
3169         return get_locale_t_name (category, thread_locale);
3170 #   else
3171         /* Solaris 11 OpenIndiana.
3172            For the internal structure of locale objects, see
3173            https://github.com/OpenIndiana/illumos-gate/blob/master/usr/src/lib/libc/port/locale/localeimpl.h  */
3174         switch (category)
3175           {
3176           case LC_CTYPE:
3177           case LC_NUMERIC:
3178           case LC_TIME:
3179           case LC_COLLATE:
3180           case LC_MONETARY:
3181           case LC_MESSAGES:
3182             return ((const char * const *) thread_locale)[category];
3183           default: /* We shouldn't get here.  */
3184             return "";
3185           }
3186 #   endif
3187 #  elif defined __CYGWIN__
3188         /* Cygwin < 2.6 lacks uselocale and thread-local locales altogether.
3189            Cygwin <= 2.6.1 lacks NL_LOCALE_NAME, requiring peeking inside
3190            an opaque struct.  */
3191 #   ifdef NL_LOCALE_NAME
3192         return nl_langinfo_l (NL_LOCALE_NAME (category), thread_locale);
3193 #   else
3194         /* FIXME: Remove when we can assume new-enough Cygwin.  */
3195         struct __locale_t {
3196           char categories[7][32];
3197         };
3198         return ((struct __locale_t *) thread_locale)->categories[category];
3199 #   endif
3200 #  elif defined __ANDROID__
3201         return MB_CUR_MAX == 4 ? "C.UTF-8" : "C";
3202 #  endif
3203       }
3204   }
3205 # endif
3206   return NULL;
3207 }
3208
3209 #endif
3210
3211 const char *
3212 gl_locale_name_thread (int category, const char *categoryname)
3213 {
3214 #if HAVE_USELOCALE
3215   const char *name = gl_locale_name_thread_unsafe (category, categoryname);
3216   if (name != NULL)
3217     return struniq (name);
3218 #endif
3219   /* On WINDOWS_NATIVE, don't use GetThreadLocale() here, because when
3220      SetThreadLocale has not been called - which is a very frequent case -
3221      the value of GetThreadLocale() ignores past calls to 'setlocale'.  */
3222   return NULL;
3223 }
3224
3225 /* XPG3 defines the result of 'setlocale (category, NULL)' as:
3226    "Directs 'setlocale()' to query 'category' and return the current
3227     setting of 'local'."
3228    However it does not specify the exact format.  Neither do SUSV2 and
3229    ISO C 99.  So we can use this feature only on selected systems (e.g.
3230    those using GNU C Library).  */
3231 #if defined _LIBC || ((defined __GLIBC__ && __GLIBC__ >= 2) && !defined __UCLIBC__)
3232 # define HAVE_LOCALE_NULL
3233 #endif
3234
3235 const char *
3236 gl_locale_name_posix (int category, const char *categoryname)
3237 {
3238 #if defined WINDOWS_NATIVE
3239   if (LC_MIN <= category && category <= LC_MAX)
3240     {
3241       const char *locname = setlocale (category, NULL);
3242       LCID lcid;
3243
3244       /* If CATEGORY is LC_ALL, the result might be a semi-colon
3245         separated list of locales.  We need only one, so we take the
3246         one corresponding to LC_CTYPE, as the most important for
3247         character translations.  */
3248       if (category == LC_ALL && strchr (locname, ';'))
3249         locname = setlocale (LC_CTYPE, NULL);
3250
3251       /* Convert locale name to LCID.  We don't want to use
3252          LocaleNameToLCID because (a) it is only available since Vista,
3253          and (b) it doesn't accept locale names returned by 'setlocale'.  */
3254       lcid = get_lcid (locname);
3255
3256       if (lcid > 0)
3257         return gl_locale_name_from_win32_LCID (lcid);
3258     }
3259 #endif
3260   {
3261     const char *locname;
3262
3263     /* Use the POSIX methods of looking to 'LC_ALL', 'LC_xxx', and 'LANG'.
3264        On some systems this can be done by the 'setlocale' function itself.  */
3265 #if defined HAVE_SETLOCALE && defined HAVE_LC_MESSAGES && defined HAVE_LOCALE_NULL
3266     locname = setlocale (category, NULL);
3267 #else
3268     /* On other systems we ignore what setlocale reports and instead look at the
3269        environment variables directly.  This is necessary
3270          1. on systems which have a facility for customizing the default locale
3271             (Mac OS X, native Windows, Cygwin) and where the system's setlocale()
3272             function ignores this default locale (Mac OS X, Cygwin), in two cases:
3273             a. when the user missed to use the setlocale() override from libintl
3274                (for example by not including <libintl.h>),
3275             b. when setlocale supports only the "C" locale, such as on Cygwin
3276                1.5.x.  In this case even the override from libintl cannot help.
3277          2. on all systems where setlocale supports only the "C" locale.  */
3278     /* Strictly speaking, it is a POSIX violation to look at the environment
3279        variables regardless whether setlocale has been called or not.  POSIX
3280        says:
3281            "For C-language programs, the POSIX locale shall be the
3282             default locale when the setlocale() function is not called."
3283        But we assume that all programs that use internationalized APIs call
3284        setlocale (LC_ALL, "").  */
3285     locname = gl_locale_name_environ (category, categoryname);
3286 #endif
3287     /* Convert the locale name from the format returned by setlocale() or found
3288        in the environment variables to the XPG syntax.  */
3289 #if defined WINDOWS_NATIVE
3290     {
3291       /* Convert locale name to LCID.  We don't want to use
3292          LocaleNameToLCID because (a) it is only available since Vista,
3293          and (b) it doesn't accept locale names returned by 'setlocale'.  */
3294       LCID lcid = get_lcid (locname);
3295
3296       if (lcid > 0)
3297         return gl_locale_name_from_win32_LCID (lcid);
3298     }
3299 #endif
3300     return locname;
3301   }
3302 }
3303
3304 const char *
3305 gl_locale_name_environ (int category, const char *categoryname)
3306 {
3307   const char *retval;
3308
3309   /* Setting of LC_ALL overrides all other.  */
3310   retval = getenv ("LC_ALL");
3311   if (retval != NULL && retval[0] != '\0')
3312     return retval;
3313   /* Next comes the name of the desired category.  */
3314   retval = getenv (categoryname);
3315   if (retval != NULL && retval[0] != '\0')
3316     return retval;
3317   /* Last possibility is the LANG environment variable.  */
3318   retval = getenv ("LANG");
3319   if (retval != NULL && retval[0] != '\0')
3320     {
3321 #if HAVE_CFLOCALECOPYCURRENT || HAVE_CFPREFERENCESCOPYAPPVALUE
3322       /* Mac OS X 10.2 or newer.
3323          Ignore invalid LANG value set by the Terminal application.  */
3324       if (strcmp (retval, "UTF-8") != 0)
3325 #endif
3326 #if defined __CYGWIN__
3327       /* Cygwin.
3328          Ignore dummy LANG value set by ~/.profile.  */
3329       if (strcmp (retval, "C.UTF-8") != 0)
3330 #endif
3331         return retval;
3332     }
3333
3334   return NULL;
3335 }
3336
3337 const char *
3338 gl_locale_name_default (void)
3339 {
3340   /* POSIX:2001 says:
3341      "All implementations shall define a locale as the default locale, to be
3342       invoked when no environment variables are set, or set to the empty
3343       string.  This default locale can be the POSIX locale or any other
3344       implementation-defined locale.  Some implementations may provide
3345       facilities for local installation administrators to set the default
3346       locale, customizing it for each location.  POSIX:2001 does not require
3347       such a facility.
3348
3349      The systems with such a facility are Mac OS X and Windows: They provide a
3350      GUI that allows the user to choose a locale.
3351        - On Mac OS X, by default, none of LC_* or LANG are set.  Starting with
3352          Mac OS X 10.4 or 10.5, LANG is set for processes launched by the
3353          'Terminal' application (but sometimes to an incorrect value "UTF-8").
3354          When no environment variable is set, setlocale (LC_ALL, "") uses the
3355          "C" locale.
3356        - On native Windows, by default, none of LC_* or LANG are set.
3357          When no environment variable is set, setlocale (LC_ALL, "") uses the
3358          locale chosen by the user.
3359        - On Cygwin 1.5.x, by default, none of LC_* or LANG are set.
3360          When no environment variable is set, setlocale (LC_ALL, "") uses the
3361          "C" locale.
3362        - On Cygwin 1.7, by default, LANG is set to "C.UTF-8" when the default
3363          ~/.profile is executed.
3364          When no environment variable is set, setlocale (LC_ALL, "") uses the
3365          "C.UTF-8" locale, which operates in the same way as the "C" locale.
3366   */
3367
3368 #if !(HAVE_CFLOCALECOPYCURRENT || HAVE_CFPREFERENCESCOPYAPPVALUE || defined WINDOWS_NATIVE || defined __CYGWIN__)
3369
3370   /* The system does not have a way of setting the locale, other than the
3371      POSIX specified environment variables.  We use C as default locale.  */
3372   return "C";
3373
3374 #else
3375
3376   /* Return an XPG style locale name language[_territory][@modifier].
3377      Don't even bother determining the codeset; it's not useful in this
3378      context, because message catalogs are not specific to a single
3379      codeset.  */
3380
3381 # if HAVE_CFLOCALECOPYCURRENT || HAVE_CFPREFERENCESCOPYAPPVALUE
3382   /* Mac OS X 10.4 or newer */
3383   {
3384     /* Cache the locale name, since CoreFoundation calls are expensive.  */
3385     static const char *cached_localename;
3386
3387     if (cached_localename == NULL)
3388       {
3389         char namebuf[256];
3390 #  if HAVE_CFLOCALECOPYCURRENT /* Mac OS X 10.5 or newer */
3391         CFLocaleRef locale = CFLocaleCopyCurrent ();
3392         CFStringRef name = CFLocaleGetIdentifier (locale);
3393 #  elif HAVE_CFPREFERENCESCOPYAPPVALUE /* Mac OS X 10.4 or newer */
3394         CFTypeRef value =
3395           CFPreferencesCopyAppValue (CFSTR ("AppleLocale"),
3396                                      kCFPreferencesCurrentApplication);
3397         if (value != NULL && CFGetTypeID (value) == CFStringGetTypeID ())
3398           {
3399             CFStringRef name = (CFStringRef)value;
3400 #  endif
3401
3402             if (CFStringGetCString (name, namebuf, sizeof (namebuf),
3403                                     kCFStringEncodingASCII))
3404               {
3405                 gl_locale_name_canonicalize (namebuf);
3406                 cached_localename = strdup (namebuf);
3407               }
3408
3409 #  if HAVE_CFLOCALECOPYCURRENT /* Mac OS X 10.5 or newer */
3410         CFRelease (locale);
3411 #  elif HAVE_CFPREFERENCESCOPYAPPVALUE /* Mac OS X 10.4 or newer */
3412           }
3413 #  endif
3414         if (cached_localename == NULL)
3415           cached_localename = "C";
3416       }
3417     return cached_localename;
3418   }
3419
3420 # endif
3421
3422 # if defined WINDOWS_NATIVE || defined __CYGWIN__ /* Native Windows or Cygwin */
3423   {
3424     LCID lcid;
3425
3426     /* Use native Windows API locale ID.  */
3427     lcid = GetThreadLocale ();
3428
3429     return gl_locale_name_from_win32_LCID (lcid);
3430   }
3431 # endif
3432 #endif
3433 }
3434
3435 /* Determine the current locale's name, and canonicalize it into XPG syntax
3436      language[_territory][.codeset][@modifier]
3437    The codeset part in the result is not reliable; the locale_charset()
3438    should be used for codeset information instead.
3439    The result must not be freed; it is statically allocated.  */
3440
3441 const char *
3442 gl_locale_name (int category, const char *categoryname)
3443 {
3444   const char *retval;
3445
3446   retval = gl_locale_name_thread (category, categoryname);
3447   if (retval != NULL)
3448     return retval;
3449
3450   retval = gl_locale_name_posix (category, categoryname);
3451   if (retval != NULL)
3452     return retval;
3453
3454   return gl_locale_name_default ();
3455 }