• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /*
4  **********************************************************************
5  *   Copyright (C) 1996-2016, International Business Machines
6  *   Corporation and others.  All Rights Reserved.
7  **********************************************************************
8  *
9  * Provides functionality for mapping between
10  * LCID and Posix IDs or ICU locale to codepage
11  *
12  * Note: All classes and code in this file are
13  *       intended for internal use only.
14  *
15  * Methods of interest:
16  *   unsigned long convertToLCID(const char*);
17  *   const char* convertToPosix(unsigned long);
18  *
19  * Kathleen Wilson, 4/30/96
20  *
21  *  Date        Name        Description
22  *  3/11/97     aliu        Fixed off-by-one bug in assignment operator. Added
23  *                          setId() method and safety check against
24  *                          MAX_ID_LENGTH.
25  * 04/23/99     stephen     Added C wrapper for convertToPosix.
26  * 09/18/00     george      Removed the memory leaks.
27  * 08/23/01     george      Convert to C
28  */
29 
30 #include "locmap.h"
31 #include "charstr.h"
32 #include "cstring.h"
33 #include "cmemory.h"
34 #include "ulocimp.h"
35 #include "unicode/uloc.h"
36 
37 #if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
38 #include <windows.h>
39 #include <winnls.h> // LCIDToLocaleName and LocaleNameToLCID
40 #endif
41 
42 /*
43  * Note:
44  * The mapping from Win32 locale ID numbers to POSIX locale strings should
45  * be the faster one.
46  *
47  * Windows LCIDs are defined at https://msdn.microsoft.com/en-us/library/cc233965.aspx
48  * [MS-LCID] Windows Language Code Identifier (LCID) Reference
49  */
50 
51 namespace {
52 
53 /*
54 ////////////////////////////////////////////////
55 //
56 // Internal Classes for LCID <--> POSIX Mapping
57 //
58 /////////////////////////////////////////////////
59 */
60 
61 typedef struct ILcidPosixElement
62 {
63     const uint32_t hostID;
64     const char * const posixID;
65 } ILcidPosixElement;
66 
67 typedef struct ILcidPosixMap
68 {
69     const uint32_t numRegions;
70     const struct ILcidPosixElement* const regionMaps;
71 } ILcidPosixMap;
72 
73 
74 /*
75 /////////////////////////////////////////////////
76 //
77 // Easy macros to make the LCID <--> POSIX Mapping
78 //
79 /////////////////////////////////////////////////
80 */
81 
82 /**
83  * The standard one language/one country mapping for LCID.
84  * The first element must be the language, and the following
85  * elements are the language with the country.
86  * @param hostID LCID in host format such as 0x044d
87  * @param languageID posix ID of just the language such as 'de'
88  * @param posixID posix ID of the language_TERRITORY such as 'de_CH'
89  */
90 #define ILCID_POSIX_ELEMENT_ARRAY(hostID, languageID, posixID) \
91 constexpr ILcidPosixElement locmap_ ## languageID [] = { \
92     {LANGUAGE_LCID(hostID), #languageID},     /* parent locale */ \
93     {hostID, #posixID}, \
94 };
95 
96 /**
97  * Define a subtable by ID
98  * @param id the POSIX ID, either a language or language_TERRITORY
99  */
100 #define ILCID_POSIX_SUBTABLE(id) \
101 constexpr ILcidPosixElement locmap_ ## id [] =
102 
103 
104 /**
105  * Create the map for the posixID. This macro supposes that the language string
106  * name is the same as the global variable name, and that the first element
107  * in the ILcidPosixElement is just the language.
108  * @param _posixID the full POSIX ID for this entry.
109  */
110 #define ILCID_POSIX_MAP(_posixID) \
111     {UPRV_LENGTHOF(locmap_ ## _posixID), locmap_ ## _posixID}
112 
113 /*
114 ////////////////////////////////////////////
115 //
116 // Create the table of LCID to POSIX Mapping
117 // None of it should be dynamically created.
118 //
119 // Keep static locale variables inside the function so that
120 // it can be created properly during static init.
121 //
122 // Note: This table should be updated periodically. Check the [MS-LCID] Windows Language Code Identifier
123 //       (LCID) Reference defined at https://msdn.microsoft.com/en-us/library/cc233965.aspx
124 //
125 //       Microsoft is moving away from LCID in favor of locale name as of Vista.  This table needs to be
126 //       maintained for support of older Windows version.
127 //       Update: Windows 7 (091130)
128 //
129 // Note: Microsoft assign a different LCID if a locale has a sorting variant. POSIX IDs below may contain
130 //       @collation=XXX, but no other keywords are allowed (at least for now). When uprv_convertToLCID() is
131 //       called from uloc_getLCID(), keywords other than collation are already removed. If we really need
132 //       to support other keywords in this mapping data, we must update the implementation.
133 ////////////////////////////////////////////
134 */
135 
136 // TODO: For Windows ideally this table would be a list of exceptions rather than a complete list as
137 // LocaleNameToLCID and LCIDToLocaleName provide 90% of these.
138 
139 ILCID_POSIX_ELEMENT_ARRAY(0x0436, af, af_ZA)
140 
ILCID_POSIX_SUBTABLE(ar)141 ILCID_POSIX_SUBTABLE(ar) {
142     {0x01,   "ar"},
143     {0x3801, "ar_AE"},
144     {0x3c01, "ar_BH"},
145     {0x1401, "ar_DZ"},
146     {0x0c01, "ar_EG"},
147     {0x0801, "ar_IQ"},
148     {0x2c01, "ar_JO"},
149     {0x3401, "ar_KW"},
150     {0x3001, "ar_LB"},
151     {0x1001, "ar_LY"},
152     {0x1801, "ar_MA"},
153     {0x1801, "ar_MO"},
154     {0x2001, "ar_OM"},
155     {0x4001, "ar_QA"},
156     {0x0401, "ar_SA"},
157     {0x2801, "ar_SY"},
158     {0x1c01, "ar_TN"},
159     {0x2401, "ar_YE"}
160 };
161 
162 ILCID_POSIX_ELEMENT_ARRAY(0x044d, as, as_IN)
163 ILCID_POSIX_ELEMENT_ARRAY(0x045e, am, am_ET)
164 ILCID_POSIX_ELEMENT_ARRAY(0x047a, arn,arn_CL)
165 
ILCID_POSIX_SUBTABLE(az)166 ILCID_POSIX_SUBTABLE(az) {
167     {0x2c,   "az"},
168     {0x082c, "az_Cyrl_AZ"},  /* Cyrillic based */
169     {0x742c, "az_Cyrl"},  /* Cyrillic based */
170     {0x042c, "az_Latn_AZ"}, /* Latin based */
171     {0x782c, "az_Latn"}, /* Latin based */
172     {0x042c, "az_AZ"} /* Latin based */
173 };
174 
175 ILCID_POSIX_ELEMENT_ARRAY(0x046d, ba, ba_RU)
176 ILCID_POSIX_ELEMENT_ARRAY(0x0423, be, be_BY)
177 
178 /*ILCID_POSIX_SUBTABLE(ber) {
179     {0x5f,   "ber"},
180     {0x045f, "ber_Arab_DZ"},
181     {0x045f, "ber_Arab"},
182     {0x085f, "ber_Latn_DZ"},
183     {0x085f, "ber_Latn"}
184 };*/
185 
186 ILCID_POSIX_ELEMENT_ARRAY(0x0402, bg, bg_BG)
187 
ILCID_POSIX_SUBTABLE(bin)188 ILCID_POSIX_SUBTABLE(bin) {
189     {0x66, "bin"},
190     {0x0466, "bin_NG"}
191 };
192 
ILCID_POSIX_SUBTABLE(bn)193 ILCID_POSIX_SUBTABLE(bn) {
194     {0x45,   "bn"},
195     {0x0845, "bn_BD"},
196     {0x0445, "bn_IN"}
197 };
198 
ILCID_POSIX_SUBTABLE(bo)199 ILCID_POSIX_SUBTABLE(bo) {
200     {0x51,   "bo"},
201     {0x0851, "bo_BT"},
202     {0x0451, "bo_CN"},
203     {0x0c51, "dz_BT"}
204 };
205 
206 ILCID_POSIX_ELEMENT_ARRAY(0x047e, br, br_FR)
207 
ILCID_POSIX_SUBTABLE(ca)208 ILCID_POSIX_SUBTABLE(ca) {
209     {0x03,   "ca"},
210     {0x0403, "ca_ES"},
211     {0x0803, "ca_ES_VALENCIA"}
212 };
213 
214 ILCID_POSIX_ELEMENT_ARRAY(0x0483, co, co_FR)
215 
ILCID_POSIX_SUBTABLE(chr)216 ILCID_POSIX_SUBTABLE(chr) {
217     {0x05c,  "chr"},
218     {0x7c5c, "chr_Cher"},
219     {0x045c, "chr_Cher_US"},
220     {0x045c, "chr_US"}
221 };
222 
223 // ICU has chosen different names for these.
ILCID_POSIX_SUBTABLE(ckb)224 ILCID_POSIX_SUBTABLE(ckb) {
225     {0x92,   "ckb"},
226     {0x7c92, "ckb_Arab"},
227     {0x0492, "ckb_Arab_IQ"}
228 };
229 
230 /* Declared as cs_CZ to get around compiler errors on z/OS, which defines cs as a function */
231 ILCID_POSIX_ELEMENT_ARRAY(0x0405, cs, cs_CZ)
232 
233 ILCID_POSIX_ELEMENT_ARRAY(0x0452, cy, cy_GB)
234 ILCID_POSIX_ELEMENT_ARRAY(0x0406, da, da_DK)
235 
236 // Windows doesn't know POSIX or BCP47 Unicode phonebook sort names
ILCID_POSIX_SUBTABLE(de)237 ILCID_POSIX_SUBTABLE(de) {
238     {0x07,   "de"},
239     {0x0c07, "de_AT"},
240     {0x0807, "de_CH"},
241     {0x0407, "de_DE"},
242     {0x1407, "de_LI"},
243     {0x1007, "de_LU"},
244     {0x10407,"de_DE@collation=phonebook"},  /*This is really de_DE_PHONEBOOK on Windows*/
245     {0x10407,"de@collation=phonebook"}  /*This is really de_DE_PHONEBOOK on Windows*/
246 };
247 
248 ILCID_POSIX_ELEMENT_ARRAY(0x0465, dv, dv_MV)
249 ILCID_POSIX_ELEMENT_ARRAY(0x0408, el, el_GR)
250 
251 // Windows uses an empty string for 'invariant'
ILCID_POSIX_SUBTABLE(en)252 ILCID_POSIX_SUBTABLE(en) {
253     {0x09,   "en"},
254     {0x0c09, "en_AU"},
255     {0x2809, "en_BZ"},
256     {0x1009, "en_CA"},
257     {0x0809, "en_GB"},
258     {0x3c09, "en_HK"},
259     {0x3809, "en_ID"},
260     {0x1809, "en_IE"},
261     {0x4009, "en_IN"},
262     {0x2009, "en_JM"},
263     {0x4409, "en_MY"},
264     {0x1409, "en_NZ"},
265     {0x3409, "en_PH"},
266     {0x4809, "en_SG"},
267     {0x2C09, "en_TT"},
268     {0x0409, "en_US"},
269     {0x007f, "en_US_POSIX"}, /* duplicate for round-tripping */
270     {0x2409, "en_029"},
271     {0x1c09, "en_ZA"},
272     {0x3009, "en_ZW"},
273     {0x2409, "en_VI"},  /* Virgin Islands AKA Caribbean Islands (en_CB). On Windows8+ This is 0x1000 or dynamically assigned */
274     {0x0409, "en_AS"},  /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
275     {0x0409, "en_GU"},  /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
276     {0x0409, "en_MH"},  /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
277     {0x0409, "en_MP"},  /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
278     {0x0409, "en_UM"}   /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
279 };
280 
ILCID_POSIX_SUBTABLE(en_US_POSIX)281 ILCID_POSIX_SUBTABLE(en_US_POSIX) {
282     {0x007f, "en_US_POSIX"} /* duplicate for roundtripping */
283 };
284 
285 // Windows doesn't know POSIX or BCP47 Unicode traditional sort names
ILCID_POSIX_SUBTABLE(es)286 ILCID_POSIX_SUBTABLE(es) {
287     {0x0a,   "es"},
288     {0x2c0a, "es_AR"},
289     {0x400a, "es_BO"},
290     {0x340a, "es_CL"},
291     {0x240a, "es_CO"},
292     {0x140a, "es_CR"},
293     {0x5c0a, "es_CU"},
294     {0x1c0a, "es_DO"},
295     {0x300a, "es_EC"},
296     {0x0c0a, "es_ES"},      /*Modern sort.*/
297     {0x100a, "es_GT"},
298     {0x480a, "es_HN"},
299     {0x080a, "es_MX"},
300     {0x4c0a, "es_NI"},
301     {0x180a, "es_PA"},
302     {0x280a, "es_PE"},
303     {0x500a, "es_PR"},
304     {0x3c0a, "es_PY"},
305     {0x440a, "es_SV"},
306     {0x540a, "es_US"},
307     {0x380a, "es_UY"},
308     {0x200a, "es_VE"},
309     {0x580a, "es_419"},
310     {0x040a, "es_ES@collation=traditional"},
311     {0x040a, "es@collation=traditional"}        // Windows will treat this as es-ES@collation=traditional
312 };
313 
314 ILCID_POSIX_ELEMENT_ARRAY(0x0425, et, et_EE)
315 ILCID_POSIX_ELEMENT_ARRAY(0x042d, eu, eu_ES)
316 
317 /* ISO-639 doesn't distinguish between Persian and Dari.*/
ILCID_POSIX_SUBTABLE(fa)318 ILCID_POSIX_SUBTABLE(fa) {
319     {0x29,   "fa"},
320     {0x0429, "fa_IR"},  /* Persian/Farsi (Iran) */
321     {0x048c, "fa_AF"}   /* Persian/Dari (Afghanistan) */
322 };
323 
324 
325 /* duplicate for roundtripping */
ILCID_POSIX_SUBTABLE(fa_AF)326 ILCID_POSIX_SUBTABLE(fa_AF) {
327     {0x8c,   "fa_AF"},  /* Persian/Dari (Afghanistan) */
328     {0x048c, "fa_AF"}   /* Persian/Dari (Afghanistan) */
329 };
330 
ILCID_POSIX_SUBTABLE(ff)331 ILCID_POSIX_SUBTABLE(ff) {
332     {0x67,   "ff"},
333     {0x7c67, "ff_Latn"},
334     {0x0867, "ff_Latn_SN"},
335     {0x0467, "ff_NG"}
336 };
337 
338 ILCID_POSIX_ELEMENT_ARRAY(0x040b, fi, fi_FI)
339 ILCID_POSIX_ELEMENT_ARRAY(0x0464, fil,fil_PH)
340 ILCID_POSIX_ELEMENT_ARRAY(0x0438, fo, fo_FO)
341 
ILCID_POSIX_SUBTABLE(fr)342 ILCID_POSIX_SUBTABLE(fr) {
343     {0x0c,   "fr"},
344     {0x080c, "fr_BE"},
345     {0x0c0c, "fr_CA"},
346     {0x240c, "fr_CD"},
347     {0x240c, "fr_CG"},
348     {0x100c, "fr_CH"},
349     {0x300c, "fr_CI"},
350     {0x2c0c, "fr_CM"},
351     {0x040c, "fr_FR"},
352     {0x3c0c, "fr_HT"},
353     {0x140c, "fr_LU"},
354     {0x380c, "fr_MA"},
355     {0x180c, "fr_MC"},
356     {0x340c, "fr_ML"},
357     {0x200c, "fr_RE"},
358     {0x280c, "fr_SN"},
359     {0xe40c, "fr_015"},
360     {0x1c0c, "fr_029"}
361 };
362 
363 ILCID_POSIX_ELEMENT_ARRAY(0x0467, fuv, fuv_NG)
364 
365 ILCID_POSIX_ELEMENT_ARRAY(0x0462, fy, fy_NL)
366 
ILCID_POSIX_SUBTABLE(ga)367 ILCID_POSIX_SUBTABLE(ga) { /* Gaelic (Ireland) */
368     {0x3c,   "ga"},
369     {0x083c, "ga_IE"},
370     {0x043c, "gd_GB"}
371 };
372 
ILCID_POSIX_SUBTABLE(gd)373 ILCID_POSIX_SUBTABLE(gd) { /* Gaelic (Scotland) */
374     {0x91,   "gd"},
375     {0x0491, "gd_GB"}
376 };
377 
378 ILCID_POSIX_ELEMENT_ARRAY(0x0456, gl, gl_ES)
379 ILCID_POSIX_ELEMENT_ARRAY(0x0447, gu, gu_IN)
380 ILCID_POSIX_ELEMENT_ARRAY(0x0474, gn, gn_PY)
381 ILCID_POSIX_ELEMENT_ARRAY(0x0484, gsw,gsw_FR)
382 
ILCID_POSIX_SUBTABLE(ha)383 ILCID_POSIX_SUBTABLE(ha) {
384     {0x68,   "ha"},
385     {0x7c68, "ha_Latn"},
386     {0x0468, "ha_Latn_NG"},
387 };
388 
389 ILCID_POSIX_ELEMENT_ARRAY(0x0475, haw,haw_US)
390 ILCID_POSIX_ELEMENT_ARRAY(0x040d, he, he_IL)
391 ILCID_POSIX_ELEMENT_ARRAY(0x0439, hi, hi_IN)
392 
393 /* This LCID is really four different locales.*/
ILCID_POSIX_SUBTABLE(hr)394 ILCID_POSIX_SUBTABLE(hr) {
395     {0x1a,   "hr"},
396     {0x141a, "bs_Latn_BA"},  /* Bosnian, Bosnia and Herzegovina */
397     {0x681a, "bs_Latn"},  /* Bosnian, Bosnia and Herzegovina */
398     {0x141a, "bs_BA"},  /* Bosnian, Bosnia and Herzegovina */
399     {0x781a, "bs"},     /* Bosnian */
400     {0x201a, "bs_Cyrl_BA"},  /* Bosnian, Bosnia and Herzegovina */
401     {0x641a, "bs_Cyrl"},  /* Bosnian, Bosnia and Herzegovina */
402     {0x101a, "hr_BA"},  /* Croatian in Bosnia */
403     {0x041a, "hr_HR"},  /* Croatian*/
404     {0x2c1a, "sr_Latn_ME"},
405     {0x241a, "sr_Latn_RS"},
406     {0x181a, "sr_Latn_BA"}, /* Serbo-Croatian in Bosnia */
407     {0x081a, "sr_Latn_CS"}, /* Serbo-Croatian*/
408     {0x701a, "sr_Latn"},    /* It's 0x1a or 0x081a, pick one to make the test program happy. */
409     {0x1c1a, "sr_Cyrl_BA"}, /* Serbo-Croatian in Bosnia */
410     {0x0c1a, "sr_Cyrl_CS"}, /* Serbian*/
411     {0x301a, "sr_Cyrl_ME"},
412     {0x281a, "sr_Cyrl_RS"},
413     {0x6c1a, "sr_Cyrl"},    /* It's 0x1a or 0x0c1a, pick one to make the test program happy. */
414     {0x7c1a, "sr"}          /* In CLDR sr is sr_Cyrl. */
415 };
416 
ILCID_POSIX_SUBTABLE(hsb)417 ILCID_POSIX_SUBTABLE(hsb) {
418     {0x2E,   "hsb"},
419     {0x042E, "hsb_DE"},
420     {0x082E, "dsb_DE"},
421     {0x7C2E, "dsb"},
422 };
423 
424 ILCID_POSIX_ELEMENT_ARRAY(0x040e, hu, hu_HU)
425 ILCID_POSIX_ELEMENT_ARRAY(0x042b, hy, hy_AM)
426 
ILCID_POSIX_SUBTABLE(ibb)427 ILCID_POSIX_SUBTABLE(ibb) {
428     {0x69, "ibb"},
429     {0x0469, "ibb_NG"}
430 };
431 
432 ILCID_POSIX_ELEMENT_ARRAY(0x0421, id, id_ID)
433 ILCID_POSIX_ELEMENT_ARRAY(0x0470, ig, ig_NG)
434 ILCID_POSIX_ELEMENT_ARRAY(0x0478, ii, ii_CN)
435 ILCID_POSIX_ELEMENT_ARRAY(0x040f, is, is_IS)
436 
ILCID_POSIX_SUBTABLE(it)437 ILCID_POSIX_SUBTABLE(it) {
438     {0x10,   "it"},
439     {0x0810, "it_CH"},
440     {0x0410, "it_IT"}
441 };
442 
ILCID_POSIX_SUBTABLE(iu)443 ILCID_POSIX_SUBTABLE(iu) {
444     {0x5d,   "iu"},
445     {0x045d, "iu_Cans_CA"},
446     {0x785d, "iu_Cans"},
447     {0x085d, "iu_Latn_CA"},
448     {0x7c5d, "iu_Latn"}
449 };
450 
451 ILCID_POSIX_ELEMENT_ARRAY(0x040d, iw, iw_IL)    /*Left in for compatibility*/
452 ILCID_POSIX_ELEMENT_ARRAY(0x0411, ja, ja_JP)
453 ILCID_POSIX_ELEMENT_ARRAY(0x0437, ka, ka_GE)
454 ILCID_POSIX_ELEMENT_ARRAY(0x043f, kk, kk_KZ)
455 ILCID_POSIX_ELEMENT_ARRAY(0x046f, kl, kl_GL)
456 ILCID_POSIX_ELEMENT_ARRAY(0x0453, km, km_KH)
457 ILCID_POSIX_ELEMENT_ARRAY(0x044b, kn, kn_IN)
458 
ILCID_POSIX_SUBTABLE(ko)459 ILCID_POSIX_SUBTABLE(ko) {
460     {0x12,   "ko"},
461     {0x0812, "ko_KP"},
462     {0x0412, "ko_KR"}
463 };
464 
465 ILCID_POSIX_ELEMENT_ARRAY(0x0457, kok, kok_IN)
466 ILCID_POSIX_ELEMENT_ARRAY(0x0471, kr,  kr_NG)
467 
ILCID_POSIX_SUBTABLE(ks)468 ILCID_POSIX_SUBTABLE(ks) {         /* We could add PK and CN too */
469     {0x60,   "ks"},
470     {0x0460, "ks_Arab_IN"},
471     {0x0860, "ks_Deva_IN"}
472 };
473 
474 ILCID_POSIX_ELEMENT_ARRAY(0x0440, ky, ky_KG)   /* Kyrgyz is spoken in Kyrgyzstan */
475 
ILCID_POSIX_SUBTABLE(la)476 ILCID_POSIX_SUBTABLE(la) {
477     {0x76,   "la"},
478     {0x0476, "la_001"},
479     {0x0476, "la_IT"}       /*Left in for compatibility*/
480 };
481 
482 ILCID_POSIX_ELEMENT_ARRAY(0x046e, lb, lb_LU)
483 ILCID_POSIX_ELEMENT_ARRAY(0x0454, lo, lo_LA)
484 ILCID_POSIX_ELEMENT_ARRAY(0x0427, lt, lt_LT)
485 ILCID_POSIX_ELEMENT_ARRAY(0x0426, lv, lv_LV)
486 ILCID_POSIX_ELEMENT_ARRAY(0x0481, mi, mi_NZ)
487 ILCID_POSIX_ELEMENT_ARRAY(0x042f, mk, mk_MK)
488 ILCID_POSIX_ELEMENT_ARRAY(0x044c, ml, ml_IN)
489 
ILCID_POSIX_SUBTABLE(mn)490 ILCID_POSIX_SUBTABLE(mn) {
491     {0x50,   "mn"},
492     {0x0450, "mn_MN"},
493     {0x7c50, "mn_Mong"},
494     {0x0850, "mn_Mong_CN"},
495     {0x0850, "mn_CN"},
496     {0x7850, "mn_Cyrl"},
497     {0x0c50, "mn_Mong_MN"}
498 };
499 
500 ILCID_POSIX_ELEMENT_ARRAY(0x0458, mni,mni_IN)
501 ILCID_POSIX_ELEMENT_ARRAY(0x047c, moh,moh_CA)
502 ILCID_POSIX_ELEMENT_ARRAY(0x044e, mr, mr_IN)
503 
ILCID_POSIX_SUBTABLE(ms)504 ILCID_POSIX_SUBTABLE(ms) {
505     {0x3e,   "ms"},
506     {0x083e, "ms_BN"},   /* Brunei Darussalam*/
507     {0x043e, "ms_MY"}    /* Malaysia*/
508 };
509 
510 ILCID_POSIX_ELEMENT_ARRAY(0x043a, mt, mt_MT)
511 ILCID_POSIX_ELEMENT_ARRAY(0x0455, my, my_MM)
512 
ILCID_POSIX_SUBTABLE(ne)513 ILCID_POSIX_SUBTABLE(ne) {
514     {0x61,   "ne"},
515     {0x0861, "ne_IN"},   /* India*/
516     {0x0461, "ne_NP"}    /* Nepal*/
517 };
518 
ILCID_POSIX_SUBTABLE(nl)519 ILCID_POSIX_SUBTABLE(nl) {
520     {0x13,   "nl"},
521     {0x0813, "nl_BE"},
522     {0x0413, "nl_NL"}
523 };
524 
525 /* The "no" locale split into nb and nn.  By default in ICU, "no" is nb.*/
526 // TODO: Not all of these are needed on Windows, but I don't know how ICU treats preferred ones here.
ILCID_POSIX_SUBTABLE(no)527 ILCID_POSIX_SUBTABLE(no) {
528     {0x14,   "no"},     /* really nb_NO - actually Windows differentiates between neutral (no region) and specific (with region) */
529     {0x7c14, "nb"},     /* really nb */
530     {0x0414, "nb_NO"},  /* really nb_NO. Keep first in the 414 list. */
531     {0x0414, "no_NO"},  /* really nb_NO */
532     {0x0814, "nn_NO"},  /* really nn_NO. Keep first in the 814 list.  */
533     {0x7814, "nn"},     /* It's 0x14 or 0x814, pick one to make the test program happy. */
534     {0x0814, "no_NO_NY"}/* really nn_NO */
535 };
536 
537 ILCID_POSIX_ELEMENT_ARRAY(0x046c, nso,nso_ZA)   /* TODO: Verify the ISO-639 code */
538 ILCID_POSIX_ELEMENT_ARRAY(0x0482, oc, oc_FR)
539 
ILCID_POSIX_SUBTABLE(om)540 ILCID_POSIX_SUBTABLE(om) { /* TODO: Verify the country */
541     {0x72,   "om"},
542     {0x0472, "om_ET"},
543     {0x0472, "gaz_ET"}
544 };
545 
546 /* Declared as or_IN to get around compiler errors*/
ILCID_POSIX_SUBTABLE(or_IN)547 ILCID_POSIX_SUBTABLE(or_IN) {
548     {0x48,   "or"},
549     {0x0448, "or_IN"},
550 };
551 
ILCID_POSIX_SUBTABLE(pa)552 ILCID_POSIX_SUBTABLE(pa) {
553     {0x46,   "pa"},
554     {0x0446, "pa_IN"},
555     {0x0846, "pa_Arab_PK"},
556     {0x0846, "pa_PK"}
557 };
558 
ILCID_POSIX_SUBTABLE(pap)559 ILCID_POSIX_SUBTABLE(pap) {
560     {0x79, "pap"},
561     {0x0479, "pap_029"},
562     {0x0479, "pap_AN"}     /*Left in for compatibility*/
563 };
564 
565 ILCID_POSIX_ELEMENT_ARRAY(0x0415, pl, pl_PL)
566 ILCID_POSIX_ELEMENT_ARRAY(0x0463, ps, ps_AF)
567 
ILCID_POSIX_SUBTABLE(pt)568 ILCID_POSIX_SUBTABLE(pt) {
569     {0x16,   "pt"},
570     {0x0416, "pt_BR"},
571     {0x0816, "pt_PT"}
572 };
573 
ILCID_POSIX_SUBTABLE(qu)574 ILCID_POSIX_SUBTABLE(qu) {
575     {0x6b,   "qu"},
576     {0x046b, "qu_BO"},
577     {0x086b, "qu_EC"},
578     {0x0C6b, "qu_PE"},
579     {0x046b, "quz_BO"},
580     {0x086b, "quz_EC"},
581     {0x0C6b, "quz_PE"}
582 };
583 
ILCID_POSIX_SUBTABLE(quc)584 ILCID_POSIX_SUBTABLE(quc) {
585     {0x93,   "quc"},
586     {0x0493, "quc_CO"},
587     /*
588         "quc_Latn_GT" is an exceptional case. Language ID of "quc"
589         is 0x93, but LCID of "quc_Latn_GT" is 0x486, which should be
590         under the group of "qut". "qut" is a retired ISO 639-3 language
591         code for West Central Quiche, and merged to "quc".
592         It looks Windows previously reserved "qut" for K'iche', but,
593         decided to use "quc" when adding a locale for K'iche' (Guatemala).
594 
595         This data structure used here assumes language ID bits in
596         LCID is unique for alphabetic language code. But this is not true
597         for "quc_Latn_GT". If we don't have the data below, LCID look up
598         by alphabetic locale ID (POSIX) will fail. The same entry is found
599         under "qut" below, which is required for reverse look up.
600     */
601     {0x0486, "quc_Latn_GT"}
602 };
603 
ILCID_POSIX_SUBTABLE(qut)604 ILCID_POSIX_SUBTABLE(qut) {
605     {0x86,   "qut"},
606     {0x0486, "qut_GT"},
607     /*
608         See the note in "quc" above.
609     */
610     {0x0486, "quc_Latn_GT"}
611 };
612 
613 ILCID_POSIX_ELEMENT_ARRAY(0x0417, rm, rm_CH)
614 
ILCID_POSIX_SUBTABLE(ro)615 ILCID_POSIX_SUBTABLE(ro) {
616     {0x18,   "ro"},
617     {0x0418, "ro_RO"},
618     {0x0818, "ro_MD"}
619 };
620 
621 // TODO: This is almost certainly 'wrong'.  0 in Windows is a synonym for LOCALE_USER_DEFAULT.
622 // More likely this is a similar concept to the Windows 0x7f Invariant locale ""
623 // (Except that it's not invariant in ICU)
ILCID_POSIX_SUBTABLE(root)624 ILCID_POSIX_SUBTABLE(root) {
625     {0x00,   "root"}
626 };
627 
ILCID_POSIX_SUBTABLE(ru)628 ILCID_POSIX_SUBTABLE(ru) {
629     {0x19,   "ru"},
630     {0x0419, "ru_RU"},
631     {0x0819, "ru_MD"}
632 };
633 
634 ILCID_POSIX_ELEMENT_ARRAY(0x0487, rw, rw_RW)
635 ILCID_POSIX_ELEMENT_ARRAY(0x044f, sa, sa_IN)
636 ILCID_POSIX_ELEMENT_ARRAY(0x0485, sah,sah_RU)
637 
ILCID_POSIX_SUBTABLE(sd)638 ILCID_POSIX_SUBTABLE(sd) {
639     {0x59,   "sd"},
640     {0x0459, "sd_Deva_IN"},
641     {0x0459, "sd_IN"},
642     {0x0859, "sd_Arab_PK"},
643     {0x0859, "sd_PK"},
644     {0x7c59, "sd_Arab"}
645 };
646 
ILCID_POSIX_SUBTABLE(se)647 ILCID_POSIX_SUBTABLE(se) {
648     {0x3b,   "se"},
649     {0x0c3b, "se_FI"},
650     {0x043b, "se_NO"},
651     {0x083b, "se_SE"},
652     {0x783b, "sma"},
653     {0x183b, "sma_NO"},
654     {0x1c3b, "sma_SE"},
655     {0x7c3b, "smj"},
656     {0x703b, "smn"},
657     {0x743b, "sms"},
658     {0x103b, "smj_NO"},
659     {0x143b, "smj_SE"},
660     {0x243b, "smn_FI"},
661     {0x203b, "sms_FI"},
662 };
663 
664 ILCID_POSIX_ELEMENT_ARRAY(0x045b, si, si_LK)
665 ILCID_POSIX_ELEMENT_ARRAY(0x041b, sk, sk_SK)
666 ILCID_POSIX_ELEMENT_ARRAY(0x0424, sl, sl_SI)
667 
ILCID_POSIX_SUBTABLE(so)668 ILCID_POSIX_SUBTABLE(so) {
669     {0x77,   "so"},
670     {0x0477, "so_SO"}
671 };
672 
673 ILCID_POSIX_ELEMENT_ARRAY(0x041c, sq, sq_AL)
674 ILCID_POSIX_ELEMENT_ARRAY(0x0430, st, st_ZA)
675 
ILCID_POSIX_SUBTABLE(sv)676 ILCID_POSIX_SUBTABLE(sv) {
677     {0x1d,   "sv"},
678     {0x081d, "sv_FI"},
679     {0x041d, "sv_SE"}
680 };
681 
682 ILCID_POSIX_ELEMENT_ARRAY(0x0441, sw, sw_KE)
683 ILCID_POSIX_ELEMENT_ARRAY(0x045A, syr, syr_SY)
684 
ILCID_POSIX_SUBTABLE(ta)685 ILCID_POSIX_SUBTABLE(ta) {
686     {0x49,   "ta"},
687     {0x0449, "ta_IN"},
688     {0x0849, "ta_LK"}
689 };
690 
691 ILCID_POSIX_ELEMENT_ARRAY(0x044a, te, te_IN)
692 
693 /* Cyrillic based by default */
ILCID_POSIX_SUBTABLE(tg)694 ILCID_POSIX_SUBTABLE(tg) {
695     {0x28,   "tg"},
696     {0x7c28, "tg_Cyrl"},
697     {0x0428, "tg_Cyrl_TJ"}
698 };
699 
700 ILCID_POSIX_ELEMENT_ARRAY(0x041e, th, th_TH)
701 
ILCID_POSIX_SUBTABLE(ti)702 ILCID_POSIX_SUBTABLE(ti) {
703     {0x73,   "ti"},
704     {0x0873, "ti_ER"},
705     {0x0473, "ti_ET"}
706 };
707 
708 ILCID_POSIX_ELEMENT_ARRAY(0x0442, tk, tk_TM)
709 
ILCID_POSIX_SUBTABLE(tn)710 ILCID_POSIX_SUBTABLE(tn) {
711     {0x32,   "tn"},
712     {0x0832, "tn_BW"},
713     {0x0432, "tn_ZA"}
714 };
715 
716 ILCID_POSIX_ELEMENT_ARRAY(0x041f, tr, tr_TR)
717 ILCID_POSIX_ELEMENT_ARRAY(0x0431, ts, ts_ZA)
718 ILCID_POSIX_ELEMENT_ARRAY(0x0444, tt, tt_RU)
719 
ILCID_POSIX_SUBTABLE(tzm)720 ILCID_POSIX_SUBTABLE(tzm) {
721     {0x5f,   "tzm"},
722     {0x7c5f, "tzm_Latn"},
723     {0x085f, "tzm_Latn_DZ"},
724     {0x105f, "tzm_Tfng_MA"},
725     {0x045f, "tzm_Arab_MA"},
726     {0x045f, "tmz"}
727 };
728 
ILCID_POSIX_SUBTABLE(ug)729 ILCID_POSIX_SUBTABLE(ug) {
730     {0x80,   "ug"},
731     {0x0480, "ug_CN"},
732     {0x0480, "ug_Arab_CN"}
733 };
734 
735 ILCID_POSIX_ELEMENT_ARRAY(0x0422, uk, uk_UA)
736 
ILCID_POSIX_SUBTABLE(ur)737 ILCID_POSIX_SUBTABLE(ur) {
738     {0x20,   "ur"},
739     {0x0820, "ur_IN"},
740     {0x0420, "ur_PK"}
741 };
742 
ILCID_POSIX_SUBTABLE(uz)743 ILCID_POSIX_SUBTABLE(uz) {
744     {0x43,   "uz"},
745     {0x0843, "uz_Cyrl_UZ"},  /* Cyrillic based */
746     {0x7843, "uz_Cyrl"},  /* Cyrillic based */
747     {0x0843, "uz_UZ"},  /* Cyrillic based */
748     {0x0443, "uz_Latn_UZ"}, /* Latin based */
749     {0x7c43, "uz_Latn"} /* Latin based */
750 };
751 
ILCID_POSIX_SUBTABLE(ve)752 ILCID_POSIX_SUBTABLE(ve) { /* TODO: Verify the country */
753     {0x33,   "ve"},
754     {0x0433, "ve_ZA"},
755     {0x0433, "ven_ZA"}
756 };
757 
758 ILCID_POSIX_ELEMENT_ARRAY(0x042a, vi, vi_VN)
759 ILCID_POSIX_ELEMENT_ARRAY(0x0488, wo, wo_SN)
760 ILCID_POSIX_ELEMENT_ARRAY(0x0434, xh, xh_ZA)
761 
ILCID_POSIX_SUBTABLE(yi)762 ILCID_POSIX_SUBTABLE(yi) {
763     {0x003d, "yi"},
764     {0x043d, "yi_001"}
765 };
766 
767 ILCID_POSIX_ELEMENT_ARRAY(0x046a, yo, yo_NG)
768 
769 // Windows & ICU tend to different names for some of these
770 // TODO: Windows probably does not need all of these entries, but I don't know how the precedence works.
ILCID_POSIX_SUBTABLE(zh)771 ILCID_POSIX_SUBTABLE(zh) {
772     {0x0004, "zh_Hans"},
773     {0x7804, "zh"},
774     {0x0804, "zh_CN"},
775     {0x0804, "zh_Hans_CN"},
776     {0x0c04, "zh_Hant_HK"},
777     {0x0c04, "zh_HK"},
778     {0x1404, "zh_Hant_MO"},
779     {0x1404, "zh_MO"},
780     {0x1004, "zh_Hans_SG"},
781     {0x1004, "zh_SG"},
782     {0x0404, "zh_Hant_TW"},
783     {0x7c04, "zh_Hant"},
784     {0x0404, "zh_TW"},
785     {0x30404,"zh_Hant_TW"},     /* Bopomofo order */
786     {0x30404,"zh_TW"},          /* Bopomofo order */
787     {0x20004,"zh@collation=stroke"},
788     {0x20404,"zh_Hant@collation=stroke"},
789     {0x20404,"zh_Hant_TW@collation=stroke"},
790     {0x20404,"zh_TW@collation=stroke"},
791     {0x20804,"zh_Hans@collation=stroke"},
792     {0x20804,"zh_Hans_CN@collation=stroke"},
793     {0x20804,"zh_CN@collation=stroke"}
794     // TODO: Alternate collations for other LCIDs are missing, eg: 0x50804
795 };
796 
797 ILCID_POSIX_ELEMENT_ARRAY(0x0435, zu, zu_ZA)
798 
799 /* This must be static and grouped by LCID. */
800 constexpr ILcidPosixMap gPosixIDmap[] = {
801     ILCID_POSIX_MAP(af),    /*  af  Afrikaans                 0x36 */
802     ILCID_POSIX_MAP(am),    /*  am  Amharic                   0x5e */
803     ILCID_POSIX_MAP(ar),    /*  ar  Arabic                    0x01 */
804     ILCID_POSIX_MAP(arn),   /*  arn Araucanian/Mapudungun     0x7a */
805     ILCID_POSIX_MAP(as),    /*  as  Assamese                  0x4d */
806     ILCID_POSIX_MAP(az),    /*  az  Azerbaijani               0x2c */
807     ILCID_POSIX_MAP(ba),    /*  ba  Bashkir                   0x6d */
808     ILCID_POSIX_MAP(be),    /*  be  Belarusian                0x23 */
809 /*    ILCID_POSIX_MAP(ber),     ber Berber/Tamazight          0x5f */
810     ILCID_POSIX_MAP(bg),    /*  bg  Bulgarian                 0x02 */
811     ILCID_POSIX_MAP(bin),   /*  bin Edo                       0x66 */
812     ILCID_POSIX_MAP(bn),    /*  bn  Bengali; Bangla           0x45 */
813     ILCID_POSIX_MAP(bo),    /*  bo  Tibetan                   0x51 */
814     ILCID_POSIX_MAP(br),    /*  br  Breton                    0x7e */
815     ILCID_POSIX_MAP(ca),    /*  ca  Catalan                   0x03 */
816     ILCID_POSIX_MAP(chr),   /*  chr Cherokee                  0x5c */
817     ILCID_POSIX_MAP(ckb),   /*  ckb Sorani (Central Kurdish)  0x92 */
818     ILCID_POSIX_MAP(co),    /*  co  Corsican                  0x83 */
819     ILCID_POSIX_MAP(cs),    /*  cs  Czech                     0x05 */
820     ILCID_POSIX_MAP(cy),    /*  cy  Welsh                     0x52 */
821     ILCID_POSIX_MAP(da),    /*  da  Danish                    0x06 */
822     ILCID_POSIX_MAP(de),    /*  de  German                    0x07 */
823     ILCID_POSIX_MAP(dv),    /*  dv  Divehi                    0x65 */
824     ILCID_POSIX_MAP(el),    /*  el  Greek                     0x08 */
825     ILCID_POSIX_MAP(en),    /*  en  English                   0x09 */
826     ILCID_POSIX_MAP(en_US_POSIX), /*    invariant             0x7f */
827     ILCID_POSIX_MAP(es),    /*  es  Spanish                   0x0a */
828     ILCID_POSIX_MAP(et),    /*  et  Estonian                  0x25 */
829     ILCID_POSIX_MAP(eu),    /*  eu  Basque                    0x2d */
830     ILCID_POSIX_MAP(fa),    /*  fa  Persian/Farsi             0x29 */
831     ILCID_POSIX_MAP(fa_AF), /*  fa  Persian/Dari              0x8c */
832     ILCID_POSIX_MAP(ff),    /*  ff  Fula                      0x67 */
833     ILCID_POSIX_MAP(fi),    /*  fi  Finnish                   0x0b */
834     ILCID_POSIX_MAP(fil),   /*  fil Filipino                  0x64 */
835     ILCID_POSIX_MAP(fo),    /*  fo  Faroese                   0x38 */
836     ILCID_POSIX_MAP(fr),    /*  fr  French                    0x0c */
837     ILCID_POSIX_MAP(fuv),   /*  fuv Fulfulde - Nigeria        0x67 */
838     ILCID_POSIX_MAP(fy),    /*  fy  Frisian                   0x62 */
839     ILCID_POSIX_MAP(ga),    /*  *   Gaelic (Ireland,Scotland) 0x3c */
840     ILCID_POSIX_MAP(gd),    /*  gd  Gaelic (United Kingdom)   0x91 */
841     ILCID_POSIX_MAP(gl),    /*  gl  Galician                  0x56 */
842     ILCID_POSIX_MAP(gn),    /*  gn  Guarani                   0x74 */
843     ILCID_POSIX_MAP(gsw),   /*  gsw Alemanic/Alsatian/Swiss German 0x84 */
844     ILCID_POSIX_MAP(gu),    /*  gu  Gujarati                  0x47 */
845     ILCID_POSIX_MAP(ha),    /*  ha  Hausa                     0x68 */
846     ILCID_POSIX_MAP(haw),   /*  haw Hawaiian                  0x75 */
847     ILCID_POSIX_MAP(he),    /*  he  Hebrew (formerly iw)      0x0d */
848     ILCID_POSIX_MAP(hi),    /*  hi  Hindi                     0x39 */
849     ILCID_POSIX_MAP(hr),    /*  *   Croatian and others       0x1a */
850     ILCID_POSIX_MAP(hsb),   /*  hsb Upper Sorbian             0x2e */
851     ILCID_POSIX_MAP(hu),    /*  hu  Hungarian                 0x0e */
852     ILCID_POSIX_MAP(hy),    /*  hy  Armenian                  0x2b */
853     ILCID_POSIX_MAP(ibb),   /*  ibb Ibibio - Nigeria          0x69 */
854     ILCID_POSIX_MAP(id),    /*  id  Indonesian (formerly in)  0x21 */
855     ILCID_POSIX_MAP(ig),    /*  ig  Igbo                      0x70 */
856     ILCID_POSIX_MAP(ii),    /*  ii  Sichuan Yi                0x78 */
857     ILCID_POSIX_MAP(is),    /*  is  Icelandic                 0x0f */
858     ILCID_POSIX_MAP(it),    /*  it  Italian                   0x10 */
859     ILCID_POSIX_MAP(iu),    /*  iu  Inuktitut                 0x5d */
860     ILCID_POSIX_MAP(iw),    /*  iw  Hebrew                    0x0d */
861     ILCID_POSIX_MAP(ja),    /*  ja  Japanese                  0x11 */
862     ILCID_POSIX_MAP(ka),    /*  ka  Georgian                  0x37 */
863     ILCID_POSIX_MAP(kk),    /*  kk  Kazakh                    0x3f */
864     ILCID_POSIX_MAP(kl),    /*  kl  Kalaallisut               0x6f */
865     ILCID_POSIX_MAP(km),    /*  km  Khmer                     0x53 */
866     ILCID_POSIX_MAP(kn),    /*  kn  Kannada                   0x4b */
867     ILCID_POSIX_MAP(ko),    /*  ko  Korean                    0x12 */
868     ILCID_POSIX_MAP(kok),   /*  kok Konkani                   0x57 */
869     ILCID_POSIX_MAP(kr),    /*  kr  Kanuri                    0x71 */
870     ILCID_POSIX_MAP(ks),    /*  ks  Kashmiri                  0x60 */
871     ILCID_POSIX_MAP(ky),    /*  ky  Kyrgyz                    0x40 */
872     ILCID_POSIX_MAP(lb),    /*  lb  Luxembourgish             0x6e */
873     ILCID_POSIX_MAP(la),    /*  la  Latin                     0x76 */
874     ILCID_POSIX_MAP(lo),    /*  lo  Lao                       0x54 */
875     ILCID_POSIX_MAP(lt),    /*  lt  Lithuanian                0x27 */
876     ILCID_POSIX_MAP(lv),    /*  lv  Latvian, Lettish          0x26 */
877     ILCID_POSIX_MAP(mi),    /*  mi  Maori                     0x81 */
878     ILCID_POSIX_MAP(mk),    /*  mk  Macedonian                0x2f */
879     ILCID_POSIX_MAP(ml),    /*  ml  Malayalam                 0x4c */
880     ILCID_POSIX_MAP(mn),    /*  mn  Mongolian                 0x50 */
881     ILCID_POSIX_MAP(mni),   /*  mni Manipuri                  0x58 */
882     ILCID_POSIX_MAP(moh),   /*  moh Mohawk                    0x7c */
883     ILCID_POSIX_MAP(mr),    /*  mr  Marathi                   0x4e */
884     ILCID_POSIX_MAP(ms),    /*  ms  Malay                     0x3e */
885     ILCID_POSIX_MAP(mt),    /*  mt  Maltese                   0x3a */
886     ILCID_POSIX_MAP(my),    /*  my  Burmese                   0x55 */
887 /*    ILCID_POSIX_MAP(nb),    //  no  Norwegian                 0x14 */
888     ILCID_POSIX_MAP(ne),    /*  ne  Nepali                    0x61 */
889     ILCID_POSIX_MAP(nl),    /*  nl  Dutch                     0x13 */
890 /*    ILCID_POSIX_MAP(nn),    //  no  Norwegian                 0x14 */
891     ILCID_POSIX_MAP(no),    /*  *   Norwegian                 0x14 */
892     ILCID_POSIX_MAP(nso),   /*  nso Sotho, Northern (Sepedi dialect) 0x6c */
893     ILCID_POSIX_MAP(oc),    /*  oc  Occitan                   0x82 */
894     ILCID_POSIX_MAP(om),    /*  om  Oromo                     0x72 */
895     ILCID_POSIX_MAP(or_IN), /*  or  Oriya                     0x48 */
896     ILCID_POSIX_MAP(pa),    /*  pa  Punjabi                   0x46 */
897     ILCID_POSIX_MAP(pap),   /*  pap Papiamentu                0x79 */
898     ILCID_POSIX_MAP(pl),    /*  pl  Polish                    0x15 */
899     ILCID_POSIX_MAP(ps),    /*  ps  Pashto                    0x63 */
900     ILCID_POSIX_MAP(pt),    /*  pt  Portuguese                0x16 */
901     ILCID_POSIX_MAP(qu),    /*  qu  Quechua                   0x6B */
902     ILCID_POSIX_MAP(quc),   /*  quc K'iche                    0x93 */
903     ILCID_POSIX_MAP(qut),   /*  qut K'iche                    0x86 */
904     ILCID_POSIX_MAP(rm),    /*  rm  Raeto-Romance/Romansh     0x17 */
905     ILCID_POSIX_MAP(ro),    /*  ro  Romanian                  0x18 */
906     ILCID_POSIX_MAP(root),  /*  root                          0x00 */
907     ILCID_POSIX_MAP(ru),    /*  ru  Russian                   0x19 */
908     ILCID_POSIX_MAP(rw),    /*  rw  Kinyarwanda               0x87 */
909     ILCID_POSIX_MAP(sa),    /*  sa  Sanskrit                  0x4f */
910     ILCID_POSIX_MAP(sah),   /*  sah Yakut                     0x85 */
911     ILCID_POSIX_MAP(sd),    /*  sd  Sindhi                    0x59 */
912     ILCID_POSIX_MAP(se),    /*  se  Sami                      0x3b */
913 /*    ILCID_POSIX_MAP(sh),    //  sh  Serbo-Croatian            0x1a */
914     ILCID_POSIX_MAP(si),    /*  si  Sinhalese                 0x5b */
915     ILCID_POSIX_MAP(sk),    /*  sk  Slovak                    0x1b */
916     ILCID_POSIX_MAP(sl),    /*  sl  Slovenian                 0x24 */
917     ILCID_POSIX_MAP(so),    /*  so  Somali                    0x77 */
918     ILCID_POSIX_MAP(sq),    /*  sq  Albanian                  0x1c */
919 /*    ILCID_POSIX_MAP(sr),    //  sr  Serbian                   0x1a */
920     ILCID_POSIX_MAP(st),    /*  st  Sutu                      0x30 */
921     ILCID_POSIX_MAP(sv),    /*  sv  Swedish                   0x1d */
922     ILCID_POSIX_MAP(sw),    /*  sw  Swahili                   0x41 */
923     ILCID_POSIX_MAP(syr),   /*  syr Syriac                    0x5A */
924     ILCID_POSIX_MAP(ta),    /*  ta  Tamil                     0x49 */
925     ILCID_POSIX_MAP(te),    /*  te  Telugu                    0x4a */
926     ILCID_POSIX_MAP(tg),    /*  tg  Tajik                     0x28 */
927     ILCID_POSIX_MAP(th),    /*  th  Thai                      0x1e */
928     ILCID_POSIX_MAP(ti),    /*  ti  Tigrigna                  0x73 */
929     ILCID_POSIX_MAP(tk),    /*  tk  Turkmen                   0x42 */
930     ILCID_POSIX_MAP(tn),    /*  tn  Tswana                    0x32 */
931     ILCID_POSIX_MAP(tr),    /*  tr  Turkish                   0x1f */
932     ILCID_POSIX_MAP(ts),    /*  ts  Tsonga                    0x31 */
933     ILCID_POSIX_MAP(tt),    /*  tt  Tatar                     0x44 */
934     ILCID_POSIX_MAP(tzm),   /*  tzm Tamazight                 0x5f */
935     ILCID_POSIX_MAP(ug),    /*  ug  Uighur                    0x80 */
936     ILCID_POSIX_MAP(uk),    /*  uk  Ukrainian                 0x22 */
937     ILCID_POSIX_MAP(ur),    /*  ur  Urdu                      0x20 */
938     ILCID_POSIX_MAP(uz),    /*  uz  Uzbek                     0x43 */
939     ILCID_POSIX_MAP(ve),    /*  ve  Venda                     0x33 */
940     ILCID_POSIX_MAP(vi),    /*  vi  Vietnamese                0x2a */
941     ILCID_POSIX_MAP(wo),    /*  wo  Wolof                     0x88 */
942     ILCID_POSIX_MAP(xh),    /*  xh  Xhosa                     0x34 */
943     ILCID_POSIX_MAP(yi),    /*  yi  Yiddish                   0x3d */
944     ILCID_POSIX_MAP(yo),    /*  yo  Yoruba                    0x6a */
945     ILCID_POSIX_MAP(zh),    /*  zh  Chinese                   0x04 */
946     ILCID_POSIX_MAP(zu),    /*  zu  Zulu                      0x35 */
947 };
948 
949 constexpr uint32_t gLocaleCount = UPRV_LENGTHOF(gPosixIDmap);
950 
951 /**
952  * Do not call this function. It is called by hostID.
953  * The function is not private because this struct must stay as a C struct,
954  * and this is an internal class.
955  */
956 int32_t
idCmp(const char * id1,const char * id2)957 idCmp(const char* id1, const char* id2)
958 {
959     int32_t diffIdx = 0;
960     while (*id1 == *id2 && *id1 != 0) {
961         diffIdx++;
962         id1++;
963         id2++;
964     }
965     return diffIdx;
966 }
967 
968 /**
969  * Searches for a Windows LCID
970  *
971  * @param posixID the Posix style locale id.
972  * @param status gets set to U_ILLEGAL_ARGUMENT_ERROR when the Posix ID has
973  *               no equivalent Windows LCID.
974  * @return the LCID
975  */
976 uint32_t
getHostID(const ILcidPosixMap * this_0,const char * posixID,UErrorCode & status)977 getHostID(const ILcidPosixMap *this_0, const char* posixID, UErrorCode& status)
978 {
979     if (U_FAILURE(status)) { return locmap_root->hostID; }
980     int32_t bestIdx = 0;
981     int32_t bestIdxDiff = 0;
982     int32_t posixIDlen = (int32_t)uprv_strlen(posixID);
983     uint32_t idx;
984 
985     for (idx = 0; idx < this_0->numRegions; idx++ ) {
986         int32_t sameChars = idCmp(posixID, this_0->regionMaps[idx].posixID);
987         if (sameChars > bestIdxDiff && this_0->regionMaps[idx].posixID[sameChars] == 0) {
988             if (posixIDlen == sameChars) {
989                 /* Exact match */
990                 return this_0->regionMaps[idx].hostID;
991             }
992             bestIdxDiff = sameChars;
993             bestIdx = idx;
994         }
995     }
996     /* We asked for something unusual, like en_ZZ, and we try to return the number for the same language. */
997     /* We also have to make sure that sid and si and similar string subsets don't match. */
998     if ((posixID[bestIdxDiff] == '_' || posixID[bestIdxDiff] == '@')
999         && this_0->regionMaps[bestIdx].posixID[bestIdxDiff] == 0)
1000     {
1001         status = U_USING_FALLBACK_WARNING;
1002         return this_0->regionMaps[bestIdx].hostID;
1003     }
1004 
1005     /*no match found */
1006     status = U_ILLEGAL_ARGUMENT_ERROR;
1007     return locmap_root->hostID;
1008 }
1009 
1010 const char*
getPosixID(const ILcidPosixMap * this_0,uint32_t hostID)1011 getPosixID(const ILcidPosixMap *this_0, uint32_t hostID)
1012 {
1013     uint32_t i;
1014     for (i = 0; i < this_0->numRegions; i++)
1015     {
1016         if (this_0->regionMaps[i].hostID == hostID)
1017         {
1018             return this_0->regionMaps[i].posixID;
1019         }
1020     }
1021 
1022     /* If you get here, then no matching region was found,
1023        so return the language id with the wild card region. */
1024     return this_0->regionMaps[0].posixID;
1025 }
1026 
1027 /*
1028 //////////////////////////////////////
1029 //
1030 // LCID --> POSIX
1031 //
1032 /////////////////////////////////////
1033 */
1034 #if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
1035 /*
1036  * Various language tags needs to be changed:
1037  * quz -> qu
1038  * prs -> fa
1039  */
FIX_LANGUAGE_ID_TAG(char * buffer,int32_t len)1040 void FIX_LANGUAGE_ID_TAG(char* buffer, int32_t len) {
1041     if (len >= 3) {
1042         if (buffer[0] == 'q' && buffer[1] == 'u' && buffer[2] == 'z') {
1043             buffer[2] = 0;
1044             uprv_strcat(buffer, buffer+3);
1045         } else if (buffer[0] == 'p' && buffer[1] == 'r' && buffer[2] == 's') {
1046             buffer[0] = 'f'; buffer[1] = 'a'; buffer[2] = 0;
1047             uprv_strcat(buffer, buffer+3);
1048         }
1049     }
1050 }
1051 #endif
1052 
1053 }  // namespace
1054 
1055 U_CAPI int32_t
uprv_convertToPosix(uint32_t hostid,char * posixID,int32_t posixIDCapacity,UErrorCode * status)1056 uprv_convertToPosix(uint32_t hostid, char *posixID, int32_t posixIDCapacity, UErrorCode* status)
1057 {
1058     uint16_t langID;
1059     uint32_t localeIndex;
1060     UBool bLookup = true;
1061     const char *pPosixID = nullptr;
1062 
1063 #if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
1064     static_assert(ULOC_FULLNAME_CAPACITY > LOCALE_NAME_MAX_LENGTH, "Windows locale names have smaller length than ICU locale names.");
1065 
1066     char locName[LOCALE_NAME_MAX_LENGTH] = {};
1067 
1068     // Note: Windows primary lang ID 0x92 in LCID is used for Central Kurdish and
1069     // GetLocaleInfo() maps such LCID to "ku". However, CLDR uses "ku" for
1070     // Northern Kurdish and "ckb" for Central Kurdish. For this reason, we cannot
1071     // use the Windows API to resolve locale ID for this specific case.
1072     if ((hostid & 0x3FF) != 0x92) {
1073         int32_t tmpLen = 0;
1074         char16_t windowsLocaleName[LOCALE_NAME_MAX_LENGTH] = {};
1075 
1076         // Note: LOCALE_ALLOW_NEUTRAL_NAMES was enabled in Windows7+, prior versions did not handle neutral (no-region) locale names.
1077         tmpLen = LCIDToLocaleName(hostid, (PWSTR)windowsLocaleName, UPRV_LENGTHOF(windowsLocaleName), LOCALE_ALLOW_NEUTRAL_NAMES);
1078         if (tmpLen > 1) {
1079             int32_t i = 0;
1080             // Only need to look up in table if have _, eg for de-de_phoneb type alternate sort.
1081             bLookup = false;
1082             for (i = 0; i < UPRV_LENGTHOF(locName); i++)
1083             {
1084                 locName[i] = (char)(windowsLocaleName[i]);
1085 
1086                 // Windows locale name may contain sorting variant, such as "es-ES_tradnl".
1087                 // In such cases, we need special mapping data found in the hardcoded table
1088                 // in this source file.
1089                 if (windowsLocaleName[i] == L'_')
1090                 {
1091                     // Keep the base locale, without variant
1092                     // TODO: Should these be mapped from _phoneb to @collation=phonebook, etc.?
1093                     locName[i] = '\0';
1094                     tmpLen = i;
1095                     bLookup = true;
1096                     break;
1097                 }
1098                 else if (windowsLocaleName[i] == L'-')
1099                 {
1100                     // Windows names use -, ICU uses _
1101                     locName[i] = '_';
1102                 }
1103                 else if (windowsLocaleName[i] == L'\0')
1104                 {
1105                     // No point in doing more work than necessary
1106                     break;
1107                 }
1108             }
1109             // TODO: Need to understand this better, why isn't it an alias?
1110             FIX_LANGUAGE_ID_TAG(locName, tmpLen);
1111             pPosixID = locName;
1112         }
1113     }
1114 #endif
1115 
1116     if (bLookup) {
1117         const char *pCandidate = nullptr;
1118         langID = LANGUAGE_LCID(hostid);
1119 
1120         for (localeIndex = 0; localeIndex < gLocaleCount; localeIndex++) {
1121             if (langID == gPosixIDmap[localeIndex].regionMaps->hostID) {
1122                 pCandidate = getPosixID(&gPosixIDmap[localeIndex], hostid);
1123                 break;
1124             }
1125         }
1126 
1127         /* On Windows, when locale name has a variant, we still look up the hardcoded table.
1128            If a match in the hardcoded table is longer than the Windows locale name without
1129            variant, we use the one as the result */
1130         if (pCandidate && (pPosixID == nullptr || uprv_strlen(pCandidate) > uprv_strlen(pPosixID))) {
1131             pPosixID = pCandidate;
1132         }
1133     }
1134 
1135     if (pPosixID) {
1136         int32_t resLen = static_cast<int32_t>(uprv_strlen(pPosixID));
1137         int32_t copyLen = resLen <= posixIDCapacity ? resLen : posixIDCapacity;
1138         uprv_memcpy(posixID, pPosixID, copyLen);
1139         if (resLen < posixIDCapacity) {
1140             posixID[resLen] = 0;
1141             if (*status == U_STRING_NOT_TERMINATED_WARNING) {
1142                 *status = U_ZERO_ERROR;
1143             }
1144         } else if (resLen == posixIDCapacity) {
1145             *status = U_STRING_NOT_TERMINATED_WARNING;
1146         } else {
1147             *status = U_BUFFER_OVERFLOW_ERROR;
1148         }
1149         return resLen;
1150     }
1151 
1152     /* no match found */
1153     *status = U_ILLEGAL_ARGUMENT_ERROR;
1154     return 0;
1155 }
1156 
1157 /*
1158 //////////////////////////////////////
1159 //
1160 // POSIX --> LCID
1161 // This should only be called from uloc_getLCID.
1162 // The locale ID must be in canonical form.
1163 //
1164 /////////////////////////////////////
1165 */
1166 U_CAPI uint32_t
uprv_convertToLCIDPlatform(const char * localeID,UErrorCode * status)1167 uprv_convertToLCIDPlatform(const char* localeID, UErrorCode* status)
1168 {
1169     if (U_FAILURE(*status)) {
1170         return 0;
1171     }
1172 
1173     // The purpose of this function is to leverage the Windows platform name->lcid
1174     // conversion functionality when available.
1175 #if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
1176     int32_t len;
1177     icu::CharString baseName;
1178     const char * mylocaleID = localeID;
1179 
1180     // Check any for keywords.
1181     if (uprv_strchr(localeID, '@'))
1182     {
1183         icu::CharString collVal = ulocimp_getKeywordValue(localeID, "collation", *status);
1184         if (U_SUCCESS(*status) && !collVal.isEmpty())
1185         {
1186             // If it contains the keyword collation, return 0 so that the LCID lookup table will be used.
1187             return 0;
1188         }
1189         else
1190         {
1191             // If the locale ID contains keywords other than collation, just use the base name.
1192             baseName = ulocimp_getBaseName(localeID, *status);
1193             if (U_SUCCESS(*status) && !baseName.isEmpty())
1194             {
1195                 mylocaleID = baseName.data();
1196             }
1197         }
1198     }
1199 
1200     // this will change it from de_DE@collation=phonebook to de-DE-u-co-phonebk form
1201     icu::CharString asciiBCP47Tag = ulocimp_toLanguageTag(mylocaleID, false, *status);
1202 
1203     if (U_SUCCESS(*status))
1204     {
1205         // Need it to be UTF-16, not 8-bit
1206         wchar_t bcp47Tag[LOCALE_NAME_MAX_LENGTH] = {};
1207         int32_t i;
1208         for (i = 0; i < UPRV_LENGTHOF(bcp47Tag); i++)
1209         {
1210             if (asciiBCP47Tag[i] == '\0')
1211             {
1212                 break;
1213             }
1214             else
1215             {
1216                 // Copy the character
1217                 bcp47Tag[i] = static_cast<wchar_t>(asciiBCP47Tag[i]);
1218             }
1219         }
1220 
1221         if (i < (UPRV_LENGTHOF(bcp47Tag) - 1))
1222         {
1223             // Ensure it's null terminated
1224             bcp47Tag[i] = L'\0';
1225             LCID lcid = LocaleNameToLCID(bcp47Tag, LOCALE_ALLOW_NEUTRAL_NAMES);
1226             if (lcid > 0)
1227             {
1228                 // Found LCID from windows, return that one, unless its completely ambiguous
1229                 // LOCALE_USER_DEFAULT and transients are OK because they will round trip
1230                 // for this process.
1231                 if (lcid != LOCALE_CUSTOM_UNSPECIFIED)
1232                 {
1233                     return lcid;
1234                 }
1235             }
1236         }
1237     }
1238 #else
1239     (void) localeID; // Suppress unused variable warning.
1240 #endif
1241 
1242     // Nothing found, or not implemented.
1243     return 0;
1244 }
1245 
1246 U_CAPI uint32_t
uprv_convertToLCID(const char * langID,const char * posixID,UErrorCode * status)1247 uprv_convertToLCID(const char *langID, const char* posixID, UErrorCode* status)
1248 {
1249     if (U_FAILURE(*status) ||
1250             langID == nullptr ||
1251             posixID == nullptr ||
1252             uprv_strlen(langID) < 2 ||
1253             uprv_strlen(posixID) < 2) {
1254         return locmap_root->hostID;
1255     }
1256 
1257     // This function does the table lookup when native platform name->lcid conversion isn't available,
1258     // or for locales that don't follow patterns the platform expects.
1259     uint32_t   low    = 0;
1260     uint32_t   high   = gLocaleCount;
1261     uint32_t   mid;
1262     uint32_t   oldmid = 0;
1263     int32_t    compVal;
1264 
1265     uint32_t   value         = 0;
1266     uint32_t   fallbackValue = (uint32_t)-1;
1267     UErrorCode myStatus;
1268     uint32_t   idx;
1269 
1270     /*Binary search for the map entry for normal cases */
1271 
1272     while (high > low)  /*binary search*/{
1273 
1274         mid = (high+low) >> 1; /*Finds median*/
1275 
1276         if (mid == oldmid)
1277             break;
1278 
1279         compVal = uprv_strcmp(langID, gPosixIDmap[mid].regionMaps->posixID);
1280         if (compVal < 0){
1281             high = mid;
1282         }
1283         else if (compVal > 0){
1284             low = mid;
1285         }
1286         else /*we found it*/{
1287             return getHostID(&gPosixIDmap[mid], posixID, *status);
1288         }
1289         oldmid = mid;
1290     }
1291 
1292     /*
1293      * Sometimes we can't do a binary search on posixID because some LCIDs
1294      * go to different locales.  We hit one of those special cases.
1295      */
1296     for (idx = 0; idx < gLocaleCount; idx++ ) {
1297         myStatus = U_ZERO_ERROR;
1298         value = getHostID(&gPosixIDmap[idx], posixID, myStatus);
1299         if (myStatus == U_ZERO_ERROR) {
1300             return value;
1301         }
1302         else if (myStatus == U_USING_FALLBACK_WARNING) {
1303             fallbackValue = value;
1304         }
1305     }
1306 
1307     if (fallbackValue != (uint32_t)-1) {
1308         *status = U_USING_FALLBACK_WARNING;
1309         return fallbackValue;
1310     }
1311 
1312     /* no match found */
1313     *status = U_ILLEGAL_ARGUMENT_ERROR;
1314     return locmap_root->hostID;   /* return international (root) */
1315 }
1316