1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/intl/icu/source/common/uloc.cpp Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,2528 @@
1.4 +/*
1.5 +**********************************************************************
1.6 +* Copyright (C) 1997-2013, International Business Machines
1.7 +* Corporation and others. All Rights Reserved.
1.8 +**********************************************************************
1.9 +*
1.10 +* File ULOC.CPP
1.11 +*
1.12 +* Modification History:
1.13 +*
1.14 +* Date Name Description
1.15 +* 04/01/97 aliu Creation.
1.16 +* 08/21/98 stephen JDK 1.2 sync
1.17 +* 12/08/98 rtg New Locale implementation and C API
1.18 +* 03/15/99 damiba overhaul.
1.19 +* 04/06/99 stephen changed setDefault() to realloc and copy
1.20 +* 06/14/99 stephen Changed calls to ures_open for new params
1.21 +* 07/21/99 stephen Modified setDefault() to propagate to C++
1.22 +* 05/14/04 alan 7 years later: refactored, cleaned up, fixed bugs,
1.23 +* brought canonicalization code into line with spec
1.24 +*****************************************************************************/
1.25 +
1.26 +/*
1.27 + POSIX's locale format, from putil.c: [no spaces]
1.28 +
1.29 + ll [ _CC ] [ . MM ] [ @ VV]
1.30 +
1.31 + l = lang, C = ctry, M = charmap, V = variant
1.32 +*/
1.33 +
1.34 +#include "unicode/utypes.h"
1.35 +#include "unicode/ustring.h"
1.36 +#include "unicode/uloc.h"
1.37 +
1.38 +#include "putilimp.h"
1.39 +#include "ustr_imp.h"
1.40 +#include "ulocimp.h"
1.41 +#include "umutex.h"
1.42 +#include "cstring.h"
1.43 +#include "cmemory.h"
1.44 +#include "ucln_cmn.h"
1.45 +#include "locmap.h"
1.46 +#include "uarrsort.h"
1.47 +#include "uenumimp.h"
1.48 +#include "uassert.h"
1.49 +
1.50 +#include <stdio.h> /* for sprintf */
1.51 +
1.52 +/* ### Declarations **************************************************/
1.53 +
1.54 +/* Locale stuff from locid.cpp */
1.55 +U_CFUNC void locale_set_default(const char *id);
1.56 +U_CFUNC const char *locale_get_default(void);
1.57 +U_CFUNC int32_t
1.58 +locale_getKeywords(const char *localeID,
1.59 + char prev,
1.60 + char *keywords, int32_t keywordCapacity,
1.61 + char *values, int32_t valuesCapacity, int32_t *valLen,
1.62 + UBool valuesToo,
1.63 + UErrorCode *status);
1.64 +
1.65 +/* ### Data tables **************************************************/
1.66 +
1.67 +/**
1.68 + * Table of language codes, both 2- and 3-letter, with preference
1.69 + * given to 2-letter codes where possible. Includes 3-letter codes
1.70 + * that lack a 2-letter equivalent.
1.71 + *
1.72 + * This list must be in sorted order. This list is returned directly
1.73 + * to the user by some API.
1.74 + *
1.75 + * This list must be kept in sync with LANGUAGES_3, with corresponding
1.76 + * entries matched.
1.77 + *
1.78 + * This table should be terminated with a NULL entry, followed by a
1.79 + * second list, and another NULL entry. The first list is visible to
1.80 + * user code when this array is returned by API. The second list
1.81 + * contains codes we support, but do not expose through user API.
1.82 + *
1.83 + * Notes
1.84 + *
1.85 + * Tables updated per http://lcweb.loc.gov/standards/iso639-2/ to
1.86 + * include the revisions up to 2001/7/27 *CWB*
1.87 + *
1.88 + * The 3 character codes are the terminology codes like RFC 3066. This
1.89 + * is compatible with prior ICU codes
1.90 + *
1.91 + * "in" "iw" "ji" "jw" & "sh" have been withdrawn but are still in the
1.92 + * table but now at the end of the table because 3 character codes are
1.93 + * duplicates. This avoids bad searches going from 3 to 2 character
1.94 + * codes.
1.95 + *
1.96 + * The range qaa-qtz is reserved for local use
1.97 + */
1.98 +/* Generated using org.unicode.cldr.icu.GenerateISO639LanguageTables */
1.99 +/* ISO639 table version is 20130531 */
1.100 +static const char * const LANGUAGES[] = {
1.101 + "aa", "ab", "ace", "ach", "ada", "ady", "ae", "af",
1.102 + "afa", "afh", "agq", "ain", "ak", "akk", "ale", "alg",
1.103 + "alt", "am", "an", "ang", "anp", "apa", "ar", "arc",
1.104 + "arn", "arp", "art", "arw", "as", "asa", "ast", "ath",
1.105 + "aus", "av", "awa", "ay", "az",
1.106 + "ba", "bad", "bai", "bal", "ban", "bas", "bat", "bax",
1.107 + "bbj", "be", "bej", "bem", "ber", "bez", "bfd", "bg",
1.108 + "bh", "bho", "bi", "bik", "bin", "bkm", "bla", "bm",
1.109 + "bn", "bnt", "bo", "br", "bra", "brx", "bs", "bss",
1.110 + "btk", "bua", "bug", "bum", "byn", "byv",
1.111 + "ca", "cad", "cai", "car", "cau", "cay", "cch", "ce",
1.112 + "ceb", "cel", "cgg", "ch", "chb", "chg", "chk", "chm",
1.113 + "chn", "cho", "chp", "chr", "chy", "ckb", "cmc", "co",
1.114 + "cop", "cpe", "cpf", "cpp", "cr", "crh", "crp", "cs",
1.115 + "csb", "cu", "cus", "cv", "cy",
1.116 + "da", "dak", "dar", "dav", "day", "de", "del", "den",
1.117 + "dgr", "din", "dje", "doi", "dra", "dsb", "dua", "dum",
1.118 + "dv", "dyo", "dyu", "dz", "dzg",
1.119 + "ebu", "ee", "efi", "egy", "eka", "el", "elx", "en",
1.120 + "enm", "eo", "es", "et", "eu", "ewo",
1.121 + "fa", "fan", "fat", "ff", "fi", "fil", "fiu", "fj",
1.122 + "fo", "fon", "fr", "frm", "fro", "frr", "frs", "fur",
1.123 + "fy",
1.124 + "ga", "gaa", "gay", "gba", "gd", "gem", "gez", "gil",
1.125 + "gl", "gmh", "gn", "goh", "gon", "gor", "got", "grb",
1.126 + "grc", "gsw", "gu", "guz", "gv", "gwi",
1.127 + "ha", "hai", "haw", "he", "hi", "hil", "him", "hit",
1.128 + "hmn", "ho", "hr", "hsb", "ht", "hu", "hup", "hy",
1.129 + "hz",
1.130 + "ia", "iba", "ibb", "id", "ie", "ig", "ii", "ijo",
1.131 + "ik", "ilo", "inc", "ine", "inh", "io", "ira", "iro",
1.132 + "is", "it", "iu",
1.133 + "ja", "jbo", "jgo", "jmc", "jpr", "jrb", "jv",
1.134 + "ka", "kaa", "kab", "kac", "kaj", "kam", "kar", "kaw",
1.135 + "kbd", "kbl", "kcg", "kde", "kea", "kfo", "kg", "kha",
1.136 + "khi", "kho", "khq", "ki", "kj", "kk", "kkj", "kl",
1.137 + "kln", "km", "kmb", "kn", "ko", "kok", "kos", "kpe",
1.138 + "kr", "krc", "krl", "kro", "kru", "ks", "ksb", "ksf",
1.139 + "ksh", "ku", "kum", "kut", "kv", "kw", "ky",
1.140 + "la", "lad", "lag", "lah", "lam", "lb", "lez", "lg",
1.141 + "li", "lkt", "ln", "lo", "lol", "loz", "lt", "lu",
1.142 + "lua", "lui", "lun", "luo", "lus", "luy", "lv",
1.143 + "mad", "maf", "mag", "mai", "mak", "man", "map", "mas",
1.144 + "mde", "mdf", "mdr", "men", "mer", "mfe", "mg", "mga",
1.145 + "mgh", "mgo", "mh", "mi", "mic", "min", "mis", "mk",
1.146 + "mkh", "ml", "mn", "mnc", "mni", "mno", "mo", "moh",
1.147 + "mos", "mr", "ms", "mt", "mua", "mul", "mun", "mus",
1.148 + "mwl", "mwr", "my", "mye", "myn", "myv",
1.149 + "na", "nah", "nai", "nap", "naq", "nb", "nd", "nds",
1.150 + "ne", "new", "ng", "nia", "nic", "niu", "nl", "nmg",
1.151 + "nn", "nnh", "no", "nog", "non", "nqo", "nr", "nso",
1.152 + "nub", "nus", "nv", "nwc", "ny", "nym", "nyn", "nyo",
1.153 + "nzi",
1.154 + "oc", "oj", "om", "or", "os", "osa", "ota", "oto",
1.155 + "pa", "paa", "pag", "pal", "pam", "pap", "pau", "peo",
1.156 + "phi", "phn", "pi", "pl", "pon", "pra", "pro", "ps",
1.157 + "pt",
1.158 + "qu",
1.159 + "raj", "rap", "rar", "rm", "rn", "ro", "roa", "rof",
1.160 + "rom", "ru", "rup", "rw", "rwk",
1.161 + "sa", "sad", "sah", "sai", "sal", "sam", "saq", "sas",
1.162 + "sat", "sba", "sbp", "sc", "scn", "sco", "sd", "se",
1.163 + "see", "seh", "sel", "sem", "ses", "sg", "sga", "sgn",
1.164 + "shi", "shn", "shu", "si", "sid", "sio", "sit",
1.165 + "sk", "sl", "sla", "sm", "sma", "smi", "smj", "smn",
1.166 + "sms", "sn", "snk", "so", "sog", "son", "sq", "sr",
1.167 + "srn", "srr", "ss", "ssa", "ssy", "st", "su", "suk",
1.168 + "sus", "sux", "sv", "sw", "swb", "swc", "syc", "syr",
1.169 + "ta", "tai", "te", "tem", "teo", "ter", "tet", "tg",
1.170 + "th", "ti", "tig", "tiv", "tk", "tkl", "tl", "tlh",
1.171 + "tli", "tmh", "tn", "to", "tog", "tpi", "tr", "trv",
1.172 + "ts", "tsi", "tt", "tum", "tup", "tut", "tvl", "tw",
1.173 + "twq", "ty", "tyv", "tzm",
1.174 + "udm", "ug", "uga", "uk", "umb", "und", "ur", "uz",
1.175 + "vai", "ve", "vi", "vo", "vot", "vun",
1.176 + "wa", "wae", "wak", "wal", "war", "was", "wen", "wo",
1.177 + "xal", "xh", "xog",
1.178 + "yao", "yap", "yav", "ybb", "yi", "yo", "ypk", "yue",
1.179 + "za", "zap", "zbl", "zen", "zgh", "zh", "znd", "zu",
1.180 + "zun", "zxx", "zza",
1.181 +NULL,
1.182 + "in", "iw", "ji", "jw", "sh", /* obsolete language codes */
1.183 +NULL
1.184 +};
1.185 +
1.186 +static const char* const DEPRECATED_LANGUAGES[]={
1.187 + "in", "iw", "ji", "jw", NULL, NULL
1.188 +};
1.189 +static const char* const REPLACEMENT_LANGUAGES[]={
1.190 + "id", "he", "yi", "jv", NULL, NULL
1.191 +};
1.192 +
1.193 +/**
1.194 + * Table of 3-letter language codes.
1.195 + *
1.196 + * This is a lookup table used to convert 3-letter language codes to
1.197 + * their 2-letter equivalent, where possible. It must be kept in sync
1.198 + * with LANGUAGES. For all valid i, LANGUAGES[i] must refer to the
1.199 + * same language as LANGUAGES_3[i]. The commented-out lines are
1.200 + * copied from LANGUAGES to make eyeballing this baby easier.
1.201 + *
1.202 + * Where a 3-letter language code has no 2-letter equivalent, the
1.203 + * 3-letter code occupies both LANGUAGES[i] and LANGUAGES_3[i].
1.204 + *
1.205 + * This table should be terminated with a NULL entry, followed by a
1.206 + * second list, and another NULL entry. The two lists correspond to
1.207 + * the two lists in LANGUAGES.
1.208 + */
1.209 +/* Generated using org.unicode.cldr.icu.GenerateISO639LanguageTables */
1.210 +/* ISO639 table version is 20130531 */
1.211 +static const char * const LANGUAGES_3[] = {
1.212 + "aar", "abk", "ace", "ach", "ada", "ady", "ave", "afr",
1.213 + "afa", "afh", "agq", "ain", "aka", "akk", "ale", "alg",
1.214 + "alt", "amh", "arg", "ang", "anp", "apa", "ara", "arc",
1.215 + "arn", "arp", "art", "arw", "asm", "asa", "ast", "ath",
1.216 + "aus", "ava", "awa", "aym", "aze",
1.217 + "bak", "bad", "bai", "bal", "ban", "bas", "bat", "bax",
1.218 + "bbj", "bel", "bej", "bem", "ber", "bez", "bfd", "bul",
1.219 + "bih", "bho", "bis", "bik", "bin", "bkm", "bla", "bam",
1.220 + "ben", "bnt", "bod", "bre", "bra", "brx", "bos", "bss",
1.221 + "btk", "bua", "bug", "bum", "byn", "byv",
1.222 + "cat", "cad", "cai", "car", "cau", "cay", "cch", "che",
1.223 + "ceb", "cel", "cgg", "cha", "chb", "chg", "chk", "chm",
1.224 + "chn", "cho", "chp", "chr", "chy", "ckb", "cmc", "cos",
1.225 + "cop", "cpe", "cpf", "cpp", "cre", "crh", "crp", "ces",
1.226 + "csb", "chu", "cus", "chv", "cym",
1.227 + "dan", "dak", "dar", "dav", "day", "deu", "del", "den",
1.228 + "dgr", "din", "dje", "doi", "dra", "dsb", "dua", "dum",
1.229 + "div", "dyo", "dyu", "dzo", "dzg",
1.230 + "ebu", "ewe", "efi", "egy", "eka", "ell", "elx", "eng",
1.231 + "enm", "epo", "spa", "est", "eus", "ewo",
1.232 + "fas", "fan", "fat", "ful", "fin", "fil", "fiu", "fij",
1.233 + "fao", "fon", "fra", "frm", "fro", "frr", "frs", "fur",
1.234 + "fry",
1.235 + "gle", "gaa", "gay", "gba", "gla", "gem", "gez", "gil",
1.236 + "glg", "gmh", "grn", "goh", "gon", "gor", "got", "grb",
1.237 + "grc", "gsw", "guj", "guz", "glv", "gwi",
1.238 + "hau", "hai", "haw", "heb", "hin", "hil", "him", "hit",
1.239 + "hmn", "hmo", "hrv", "hsb", "hat", "hun", "hup", "hye",
1.240 + "her",
1.241 + "ina", "iba", "ibb", "ind", "ile", "ibo", "iii", "ijo",
1.242 + "ipk", "ilo", "inc", "ine", "inh", "ido", "ira", "iro",
1.243 + "isl", "ita", "iku",
1.244 + "jpn", "jbo", "jgo", "jmc", "jpr", "jrb", "jav",
1.245 + "kat", "kaa", "kab", "kac", "kaj", "kam", "kar", "kaw",
1.246 + "kbd", "kbl", "kcg", "kde", "kea", "kfo", "kon", "kha",
1.247 + "khi", "kho", "khq", "kik", "kua", "kaz", "kkj", "kal",
1.248 + "kln", "khm", "kmb", "kan", "kor", "kok", "kos", "kpe",
1.249 + "kau", "krc", "krl", "kro", "kru", "kas", "ksb", "ksf",
1.250 + "ksh", "kur", "kum", "kut", "kom", "cor", "kir",
1.251 + "lat", "lad", "lag", "lah", "lam", "ltz", "lez", "lug",
1.252 + "lim", "lkt", "lin", "lao", "lol", "loz", "lit", "lub",
1.253 + "lua", "lui", "lun", "luo", "lus", "luy", "lav",
1.254 + "mad", "maf", "mag", "mai", "mak", "man", "map", "mas",
1.255 + "mde", "mdf", "mdr", "men", "mer", "mfe", "mlg", "mga",
1.256 + "mgh", "mgo", "mah", "mri", "mic", "min", "mis", "mkd",
1.257 + "mkh", "mal", "mon", "mnc", "mni", "mno", "mol", "moh",
1.258 + "mos", "mar", "msa", "mlt", "mua", "mul", "mun", "mus",
1.259 + "mwl", "mwr", "mya", "mye", "myn", "myv",
1.260 + "nau", "nah", "nai", "nap", "naq", "nob", "nde", "nds",
1.261 + "nep", "new", "ndo", "nia", "nic", "niu", "nld", "nmg",
1.262 + "nno", "nnh", "nor", "nog", "non", "nqo", "nbl", "nso",
1.263 + "nub", "nus", "nav", "nwc", "nya", "nym", "nyn", "nyo",
1.264 + "nzi",
1.265 + "oci", "oji", "orm", "ori", "oss", "osa", "ota", "oto",
1.266 + "pan", "paa", "pag", "pal", "pam", "pap", "pau", "peo",
1.267 + "phi", "phn", "pli", "pol", "pon", "pra", "pro", "pus",
1.268 + "por",
1.269 + "que",
1.270 + "raj", "rap", "rar", "roh", "run", "ron", "roa", "rof",
1.271 + "rom", "rus", "rup", "kin", "rwk",
1.272 + "san", "sad", "sah", "sai", "sal", "sam", "saq", "sas",
1.273 + "sat", "sba", "sbp", "srd", "scn", "sco", "snd", "sme",
1.274 + "see", "seh", "sel", "sem", "ses", "sag", "sga", "sgn",
1.275 + "shi", "shn", "shu", "sin", "sid", "sio", "sit",
1.276 + "slk", "slv", "sla", "smo", "sma", "smi", "smj", "smn",
1.277 + "sms", "sna", "snk", "som", "sog", "son", "sqi", "srp",
1.278 + "srn", "srr", "ssw", "ssa", "ssy", "sot", "sun", "suk",
1.279 + "sus", "sux", "swe", "swa", "swb", "swc", "syc", "syr",
1.280 + "tam", "tai", "tel", "tem", "teo", "ter", "tet", "tgk",
1.281 + "tha", "tir", "tig", "tiv", "tuk", "tkl", "tgl", "tlh",
1.282 + "tli", "tmh", "tsn", "ton", "tog", "tpi", "tur", "trv",
1.283 + "tso", "tsi", "tat", "tum", "tup", "tut", "tvl", "twi",
1.284 + "twq", "tah", "tyv", "tzm",
1.285 + "udm", "uig", "uga", "ukr", "umb", "und", "urd", "uzb",
1.286 + "vai", "ven", "vie", "vol", "vot", "vun",
1.287 + "wln", "wae", "wak", "wal", "war", "was", "wen", "wol",
1.288 + "xal", "xho", "xog",
1.289 + "yao", "yap", "yav", "ybb", "yid", "yor", "ypk", "yue",
1.290 + "zha", "zap", "zbl", "zen", "zgh", "zho", "znd", "zul",
1.291 + "zun", "zxx", "zza",
1.292 +NULL,
1.293 +/* "in", "iw", "ji", "jw", "sh", */
1.294 + "ind", "heb", "yid", "jaw", "srp",
1.295 +NULL
1.296 +};
1.297 +
1.298 +/**
1.299 + * Table of 2-letter country codes.
1.300 + *
1.301 + * This list must be in sorted order. This list is returned directly
1.302 + * to the user by some API.
1.303 + *
1.304 + * This list must be kept in sync with COUNTRIES_3, with corresponding
1.305 + * entries matched.
1.306 + *
1.307 + * This table should be terminated with a NULL entry, followed by a
1.308 + * second list, and another NULL entry. The first list is visible to
1.309 + * user code when this array is returned by API. The second list
1.310 + * contains codes we support, but do not expose through user API.
1.311 + *
1.312 + * Notes:
1.313 + *
1.314 + * ZR(ZAR) is now CD(COD) and FX(FXX) is PS(PSE) as per
1.315 + * http://www.evertype.com/standards/iso3166/iso3166-1-en.html added
1.316 + * new codes keeping the old ones for compatibility updated to include
1.317 + * 1999/12/03 revisions *CWB*
1.318 + *
1.319 + * RO(ROM) is now RO(ROU) according to
1.320 + * http://www.iso.org/iso/en/prods-services/iso3166ma/03updates-on-iso-3166/nlv3e-rou.html
1.321 + */
1.322 +static const char * const COUNTRIES[] = {
1.323 + "AD", "AE", "AF", "AG", "AI", "AL", "AM",
1.324 + "AO", "AQ", "AR", "AS", "AT", "AU", "AW", "AX", "AZ",
1.325 + "BA", "BB", "BD", "BE", "BF", "BG", "BH", "BI",
1.326 + "BJ", "BL", "BM", "BN", "BO", "BQ", "BR", "BS", "BT", "BV",
1.327 + "BW", "BY", "BZ", "CA", "CC", "CD", "CF", "CG",
1.328 + "CH", "CI", "CK", "CL", "CM", "CN", "CO", "CR",
1.329 + "CU", "CV", "CW", "CX", "CY", "CZ", "DE", "DJ", "DK",
1.330 + "DM", "DO", "DZ", "EC", "EE", "EG", "EH", "ER",
1.331 + "ES", "ET", "FI", "FJ", "FK", "FM", "FO", "FR",
1.332 + "GA", "GB", "GD", "GE", "GF", "GG", "GH", "GI", "GL",
1.333 + "GM", "GN", "GP", "GQ", "GR", "GS", "GT", "GU",
1.334 + "GW", "GY", "HK", "HM", "HN", "HR", "HT", "HU",
1.335 + "ID", "IE", "IL", "IM", "IN", "IO", "IQ", "IR", "IS",
1.336 + "IT", "JE", "JM", "JO", "JP", "KE", "KG", "KH", "KI",
1.337 + "KM", "KN", "KP", "KR", "KW", "KY", "KZ", "LA",
1.338 + "LB", "LC", "LI", "LK", "LR", "LS", "LT", "LU",
1.339 + "LV", "LY", "MA", "MC", "MD", "ME", "MF", "MG", "MH", "MK",
1.340 + "ML", "MM", "MN", "MO", "MP", "MQ", "MR", "MS",
1.341 + "MT", "MU", "MV", "MW", "MX", "MY", "MZ", "NA",
1.342 + "NC", "NE", "NF", "NG", "NI", "NL", "NO", "NP",
1.343 + "NR", "NU", "NZ", "OM", "PA", "PE", "PF", "PG",
1.344 + "PH", "PK", "PL", "PM", "PN", "PR", "PS", "PT",
1.345 + "PW", "PY", "QA", "RE", "RO", "RS", "RU", "RW", "SA",
1.346 + "SB", "SC", "SD", "SE", "SG", "SH", "SI", "SJ",
1.347 + "SK", "SL", "SM", "SN", "SO", "SR", "SS", "ST", "SV",
1.348 + "SX", "SY", "SZ", "TC", "TD", "TF", "TG", "TH", "TJ",
1.349 + "TK", "TL", "TM", "TN", "TO", "TR", "TT", "TV",
1.350 + "TW", "TZ", "UA", "UG", "UM", "US", "UY", "UZ",
1.351 + "VA", "VC", "VE", "VG", "VI", "VN", "VU", "WF",
1.352 + "WS", "YE", "YT", "ZA", "ZM", "ZW",
1.353 +NULL,
1.354 + "AN", "BU", "CS", "FX", "RO", "SU", "TP", "YD", "YU", "ZR", /* obsolete country codes */
1.355 +NULL
1.356 +};
1.357 +
1.358 +static const char* const DEPRECATED_COUNTRIES[] = {
1.359 + "AN", "BU", "CS", "DD", "DY", "FX", "HV", "NH", "RH", "SU", "TP", "UK", "VD", "YD", "YU", "ZR", NULL, NULL /* deprecated country list */
1.360 +};
1.361 +static const char* const REPLACEMENT_COUNTRIES[] = {
1.362 +/* "AN", "BU", "CS", "DD", "DY", "FX", "HV", "NH", "RH", "SU", "TP", "UK", "VD", "YD", "YU", "ZR" */
1.363 + "CW", "MM", "RS", "DE", "BJ", "FR", "BF", "VU", "ZW", "RU", "TL", "GB", "VN", "YE", "RS", "CD", NULL, NULL /* replacement country codes */
1.364 +};
1.365 +
1.366 +/**
1.367 + * Table of 3-letter country codes.
1.368 + *
1.369 + * This is a lookup table used to convert 3-letter country codes to
1.370 + * their 2-letter equivalent. It must be kept in sync with COUNTRIES.
1.371 + * For all valid i, COUNTRIES[i] must refer to the same country as
1.372 + * COUNTRIES_3[i]. The commented-out lines are copied from COUNTRIES
1.373 + * to make eyeballing this baby easier.
1.374 + *
1.375 + * This table should be terminated with a NULL entry, followed by a
1.376 + * second list, and another NULL entry. The two lists correspond to
1.377 + * the two lists in COUNTRIES.
1.378 + */
1.379 +static const char * const COUNTRIES_3[] = {
1.380 +/* "AD", "AE", "AF", "AG", "AI", "AL", "AM", */
1.381 + "AND", "ARE", "AFG", "ATG", "AIA", "ALB", "ARM",
1.382 +/* "AO", "AQ", "AR", "AS", "AT", "AU", "AW", "AX", "AZ", */
1.383 + "AGO", "ATA", "ARG", "ASM", "AUT", "AUS", "ABW", "ALA", "AZE",
1.384 +/* "BA", "BB", "BD", "BE", "BF", "BG", "BH", "BI", */
1.385 + "BIH", "BRB", "BGD", "BEL", "BFA", "BGR", "BHR", "BDI",
1.386 +/* "BJ", "BL", "BM", "BN", "BO", "BQ", "BR", "BS", "BT", "BV", */
1.387 + "BEN", "BLM", "BMU", "BRN", "BOL", "BES", "BRA", "BHS", "BTN", "BVT",
1.388 +/* "BW", "BY", "BZ", "CA", "CC", "CD", "CF", "CG", */
1.389 + "BWA", "BLR", "BLZ", "CAN", "CCK", "COD", "CAF", "COG",
1.390 +/* "CH", "CI", "CK", "CL", "CM", "CN", "CO", "CR", */
1.391 + "CHE", "CIV", "COK", "CHL", "CMR", "CHN", "COL", "CRI",
1.392 +/* "CU", "CV", "CW", "CX", "CY", "CZ", "DE", "DJ", "DK", */
1.393 + "CUB", "CPV", "CUW", "CXR", "CYP", "CZE", "DEU", "DJI", "DNK",
1.394 +/* "DM", "DO", "DZ", "EC", "EE", "EG", "EH", "ER", */
1.395 + "DMA", "DOM", "DZA", "ECU", "EST", "EGY", "ESH", "ERI",
1.396 +/* "ES", "ET", "FI", "FJ", "FK", "FM", "FO", "FR", */
1.397 + "ESP", "ETH", "FIN", "FJI", "FLK", "FSM", "FRO", "FRA",
1.398 +/* "GA", "GB", "GD", "GE", "GF", "GG", "GH", "GI", "GL", */
1.399 + "GAB", "GBR", "GRD", "GEO", "GUF", "GGY", "GHA", "GIB", "GRL",
1.400 +/* "GM", "GN", "GP", "GQ", "GR", "GS", "GT", "GU", */
1.401 + "GMB", "GIN", "GLP", "GNQ", "GRC", "SGS", "GTM", "GUM",
1.402 +/* "GW", "GY", "HK", "HM", "HN", "HR", "HT", "HU", */
1.403 + "GNB", "GUY", "HKG", "HMD", "HND", "HRV", "HTI", "HUN",
1.404 +/* "ID", "IE", "IL", "IM", "IN", "IO", "IQ", "IR", "IS" */
1.405 + "IDN", "IRL", "ISR", "IMN", "IND", "IOT", "IRQ", "IRN", "ISL",
1.406 +/* "IT", "JE", "JM", "JO", "JP", "KE", "KG", "KH", "KI", */
1.407 + "ITA", "JEY", "JAM", "JOR", "JPN", "KEN", "KGZ", "KHM", "KIR",
1.408 +/* "KM", "KN", "KP", "KR", "KW", "KY", "KZ", "LA", */
1.409 + "COM", "KNA", "PRK", "KOR", "KWT", "CYM", "KAZ", "LAO",
1.410 +/* "LB", "LC", "LI", "LK", "LR", "LS", "LT", "LU", */
1.411 + "LBN", "LCA", "LIE", "LKA", "LBR", "LSO", "LTU", "LUX",
1.412 +/* "LV", "LY", "MA", "MC", "MD", "ME", "MF", "MG", "MH", "MK", */
1.413 + "LVA", "LBY", "MAR", "MCO", "MDA", "MNE", "MAF", "MDG", "MHL", "MKD",
1.414 +/* "ML", "MM", "MN", "MO", "MP", "MQ", "MR", "MS", */
1.415 + "MLI", "MMR", "MNG", "MAC", "MNP", "MTQ", "MRT", "MSR",
1.416 +/* "MT", "MU", "MV", "MW", "MX", "MY", "MZ", "NA", */
1.417 + "MLT", "MUS", "MDV", "MWI", "MEX", "MYS", "MOZ", "NAM",
1.418 +/* "NC", "NE", "NF", "NG", "NI", "NL", "NO", "NP", */
1.419 + "NCL", "NER", "NFK", "NGA", "NIC", "NLD", "NOR", "NPL",
1.420 +/* "NR", "NU", "NZ", "OM", "PA", "PE", "PF", "PG", */
1.421 + "NRU", "NIU", "NZL", "OMN", "PAN", "PER", "PYF", "PNG",
1.422 +/* "PH", "PK", "PL", "PM", "PN", "PR", "PS", "PT", */
1.423 + "PHL", "PAK", "POL", "SPM", "PCN", "PRI", "PSE", "PRT",
1.424 +/* "PW", "PY", "QA", "RE", "RO", "RS", "RU", "RW", "SA", */
1.425 + "PLW", "PRY", "QAT", "REU", "ROU", "SRB", "RUS", "RWA", "SAU",
1.426 +/* "SB", "SC", "SD", "SE", "SG", "SH", "SI", "SJ", */
1.427 + "SLB", "SYC", "SDN", "SWE", "SGP", "SHN", "SVN", "SJM",
1.428 +/* "SK", "SL", "SM", "SN", "SO", "SR", "SS", "ST", "SV", */
1.429 + "SVK", "SLE", "SMR", "SEN", "SOM", "SUR", "SSD", "STP", "SLV",
1.430 +/* "SX", "SY", "SZ", "TC", "TD", "TF", "TG", "TH", "TJ", */
1.431 + "SXM", "SYR", "SWZ", "TCA", "TCD", "ATF", "TGO", "THA", "TJK",
1.432 +/* "TK", "TL", "TM", "TN", "TO", "TR", "TT", "TV", */
1.433 + "TKL", "TLS", "TKM", "TUN", "TON", "TUR", "TTO", "TUV",
1.434 +/* "TW", "TZ", "UA", "UG", "UM", "US", "UY", "UZ", */
1.435 + "TWN", "TZA", "UKR", "UGA", "UMI", "USA", "URY", "UZB",
1.436 +/* "VA", "VC", "VE", "VG", "VI", "VN", "VU", "WF", */
1.437 + "VAT", "VCT", "VEN", "VGB", "VIR", "VNM", "VUT", "WLF",
1.438 +/* "WS", "YE", "YT", "ZA", "ZM", "ZW", */
1.439 + "WSM", "YEM", "MYT", "ZAF", "ZMB", "ZWE",
1.440 +NULL,
1.441 +/* "AN", "BU", "CS", "FX", "RO", "SU", "TP", "YD", "YU", "ZR" */
1.442 + "ANT", "BUR", "SCG", "FXX", "ROM", "SUN", "TMP", "YMD", "YUG", "ZAR",
1.443 +NULL
1.444 +};
1.445 +
1.446 +typedef struct CanonicalizationMap {
1.447 + const char *id; /* input ID */
1.448 + const char *canonicalID; /* canonicalized output ID */
1.449 + const char *keyword; /* keyword, or NULL if none */
1.450 + const char *value; /* keyword value, or NULL if kw==NULL */
1.451 +} CanonicalizationMap;
1.452 +
1.453 +/**
1.454 + * A map to canonicalize locale IDs. This handles a variety of
1.455 + * different semantic kinds of transformations.
1.456 + */
1.457 +static const CanonicalizationMap CANONICALIZE_MAP[] = {
1.458 + { "", "en_US_POSIX", NULL, NULL }, /* .NET name */
1.459 + { "c", "en_US_POSIX", NULL, NULL }, /* POSIX name */
1.460 + { "posix", "en_US_POSIX", NULL, NULL }, /* POSIX name (alias of C) */
1.461 + { "art_LOJBAN", "jbo", NULL, NULL }, /* registered name */
1.462 + { "az_AZ_CYRL", "az_Cyrl_AZ", NULL, NULL }, /* .NET name */
1.463 + { "az_AZ_LATN", "az_Latn_AZ", NULL, NULL }, /* .NET name */
1.464 + { "ca_ES_PREEURO", "ca_ES", "currency", "ESP" },
1.465 + { "de__PHONEBOOK", "de", "collation", "phonebook" }, /* Old ICU name */
1.466 + { "de_AT_PREEURO", "de_AT", "currency", "ATS" },
1.467 + { "de_DE_PREEURO", "de_DE", "currency", "DEM" },
1.468 + { "de_LU_PREEURO", "de_LU", "currency", "LUF" },
1.469 + { "el_GR_PREEURO", "el_GR", "currency", "GRD" },
1.470 + { "en_BE_PREEURO", "en_BE", "currency", "BEF" },
1.471 + { "en_IE_PREEURO", "en_IE", "currency", "IEP" },
1.472 + { "es__TRADITIONAL", "es", "collation", "traditional" }, /* Old ICU name */
1.473 + { "es_ES_PREEURO", "es_ES", "currency", "ESP" },
1.474 + { "eu_ES_PREEURO", "eu_ES", "currency", "ESP" },
1.475 + { "fi_FI_PREEURO", "fi_FI", "currency", "FIM" },
1.476 + { "fr_BE_PREEURO", "fr_BE", "currency", "BEF" },
1.477 + { "fr_FR_PREEURO", "fr_FR", "currency", "FRF" },
1.478 + { "fr_LU_PREEURO", "fr_LU", "currency", "LUF" },
1.479 + { "ga_IE_PREEURO", "ga_IE", "currency", "IEP" },
1.480 + { "gl_ES_PREEURO", "gl_ES", "currency", "ESP" },
1.481 + { "hi__DIRECT", "hi", "collation", "direct" }, /* Old ICU name */
1.482 + { "it_IT_PREEURO", "it_IT", "currency", "ITL" },
1.483 + { "ja_JP_TRADITIONAL", "ja_JP", "calendar", "japanese" }, /* Old ICU name */
1.484 + { "nb_NO_NY", "nn_NO", NULL, NULL }, /* "markus said this was ok" :-) */
1.485 + { "nl_BE_PREEURO", "nl_BE", "currency", "BEF" },
1.486 + { "nl_NL_PREEURO", "nl_NL", "currency", "NLG" },
1.487 + { "pt_PT_PREEURO", "pt_PT", "currency", "PTE" },
1.488 + { "sr_SP_CYRL", "sr_Cyrl_RS", NULL, NULL }, /* .NET name */
1.489 + { "sr_SP_LATN", "sr_Latn_RS", NULL, NULL }, /* .NET name */
1.490 + { "sr_YU_CYRILLIC", "sr_Cyrl_RS", NULL, NULL }, /* Linux name */
1.491 + { "th_TH_TRADITIONAL", "th_TH", "calendar", "buddhist" }, /* Old ICU name */
1.492 + { "uz_UZ_CYRILLIC", "uz_Cyrl_UZ", NULL, NULL }, /* Linux name */
1.493 + { "uz_UZ_CYRL", "uz_Cyrl_UZ", NULL, NULL }, /* .NET name */
1.494 + { "uz_UZ_LATN", "uz_Latn_UZ", NULL, NULL }, /* .NET name */
1.495 + { "zh_CHS", "zh_Hans", NULL, NULL }, /* .NET name */
1.496 + { "zh_CHT", "zh_Hant", NULL, NULL }, /* .NET name */
1.497 + { "zh_GAN", "gan", NULL, NULL }, /* registered name */
1.498 + { "zh_GUOYU", "zh", NULL, NULL }, /* registered name */
1.499 + { "zh_HAKKA", "hak", NULL, NULL }, /* registered name */
1.500 + { "zh_MIN_NAN", "nan", NULL, NULL }, /* registered name */
1.501 + { "zh_WUU", "wuu", NULL, NULL }, /* registered name */
1.502 + { "zh_XIANG", "hsn", NULL, NULL }, /* registered name */
1.503 + { "zh_YUE", "yue", NULL, NULL }, /* registered name */
1.504 +};
1.505 +
1.506 +typedef struct VariantMap {
1.507 + const char *variant; /* input ID */
1.508 + const char *keyword; /* keyword, or NULL if none */
1.509 + const char *value; /* keyword value, or NULL if kw==NULL */
1.510 +} VariantMap;
1.511 +
1.512 +static const VariantMap VARIANT_MAP[] = {
1.513 + { "EURO", "currency", "EUR" },
1.514 + { "PINYIN", "collation", "pinyin" }, /* Solaris variant */
1.515 + { "STROKE", "collation", "stroke" } /* Solaris variant */
1.516 +};
1.517 +
1.518 +/* ### BCP47 Conversion *******************************************/
1.519 +/* Test if the locale id has BCP47 u extension and does not have '@' */
1.520 +#define _hasBCP47Extension(id) (id && uprv_strstr(id, "@") == NULL && getShortestSubtagLength(localeID) == 1)
1.521 +/* Converts the BCP47 id to Unicode id. Does nothing to id if conversion fails */
1.522 +#define _ConvertBCP47(finalID, id, buffer, length,err) \
1.523 + if (uloc_forLanguageTag(id, buffer, length, NULL, err) <= 0 || U_FAILURE(*err)) { \
1.524 + finalID=id; \
1.525 + } else { \
1.526 + finalID=buffer; \
1.527 + }
1.528 +/* Gets the size of the shortest subtag in the given localeID. */
1.529 +static int32_t getShortestSubtagLength(const char *localeID) {
1.530 + int32_t localeIDLength = uprv_strlen(localeID);
1.531 + int32_t length = localeIDLength;
1.532 + int32_t tmpLength = 0;
1.533 + int32_t i;
1.534 + UBool reset = TRUE;
1.535 +
1.536 + for (i = 0; i < localeIDLength; i++) {
1.537 + if (localeID[i] != '_' && localeID[i] != '-') {
1.538 + if (reset) {
1.539 + tmpLength = 0;
1.540 + reset = FALSE;
1.541 + }
1.542 + tmpLength++;
1.543 + } else {
1.544 + if (tmpLength != 0 && tmpLength < length) {
1.545 + length = tmpLength;
1.546 + }
1.547 + reset = TRUE;
1.548 + }
1.549 + }
1.550 +
1.551 + return length;
1.552 +}
1.553 +
1.554 +/* ### Keywords **************************************************/
1.555 +
1.556 +#define ULOC_KEYWORD_BUFFER_LEN 25
1.557 +#define ULOC_MAX_NO_KEYWORDS 25
1.558 +
1.559 +U_CAPI const char * U_EXPORT2
1.560 +locale_getKeywordsStart(const char *localeID) {
1.561 + const char *result = NULL;
1.562 + if((result = uprv_strchr(localeID, '@')) != NULL) {
1.563 + return result;
1.564 + }
1.565 +#if (U_CHARSET_FAMILY == U_EBCDIC_FAMILY)
1.566 + else {
1.567 + /* We do this because the @ sign is variant, and the @ sign used on one
1.568 + EBCDIC machine won't be compiled the same way on other EBCDIC based
1.569 + machines. */
1.570 + static const uint8_t ebcdicSigns[] = { 0x7C, 0x44, 0x66, 0x80, 0xAC, 0xAE, 0xAF, 0xB5, 0xEC, 0xEF, 0x00 };
1.571 + const uint8_t *charToFind = ebcdicSigns;
1.572 + while(*charToFind) {
1.573 + if((result = uprv_strchr(localeID, *charToFind)) != NULL) {
1.574 + return result;
1.575 + }
1.576 + charToFind++;
1.577 + }
1.578 + }
1.579 +#endif
1.580 + return NULL;
1.581 +}
1.582 +
1.583 +/**
1.584 + * @param buf buffer of size [ULOC_KEYWORD_BUFFER_LEN]
1.585 + * @param keywordName incoming name to be canonicalized
1.586 + * @param status return status (keyword too long)
1.587 + * @return length of the keyword name
1.588 + */
1.589 +static int32_t locale_canonKeywordName(char *buf, const char *keywordName, UErrorCode *status)
1.590 +{
1.591 + int32_t i;
1.592 + int32_t keywordNameLen = (int32_t)uprv_strlen(keywordName);
1.593 +
1.594 + if(keywordNameLen >= ULOC_KEYWORD_BUFFER_LEN) {
1.595 + /* keyword name too long for internal buffer */
1.596 + *status = U_INTERNAL_PROGRAM_ERROR;
1.597 + return 0;
1.598 + }
1.599 +
1.600 + /* normalize the keyword name */
1.601 + for(i = 0; i < keywordNameLen; i++) {
1.602 + buf[i] = uprv_tolower(keywordName[i]);
1.603 + }
1.604 + buf[i] = 0;
1.605 +
1.606 + return keywordNameLen;
1.607 +}
1.608 +
1.609 +typedef struct {
1.610 + char keyword[ULOC_KEYWORD_BUFFER_LEN];
1.611 + int32_t keywordLen;
1.612 + const char *valueStart;
1.613 + int32_t valueLen;
1.614 +} KeywordStruct;
1.615 +
1.616 +static int32_t U_CALLCONV
1.617 +compareKeywordStructs(const void * /*context*/, const void *left, const void *right) {
1.618 + const char* leftString = ((const KeywordStruct *)left)->keyword;
1.619 + const char* rightString = ((const KeywordStruct *)right)->keyword;
1.620 + return uprv_strcmp(leftString, rightString);
1.621 +}
1.622 +
1.623 +/**
1.624 + * Both addKeyword and addValue must already be in canonical form.
1.625 + * Either both addKeyword and addValue are NULL, or neither is NULL.
1.626 + * If they are not NULL they must be zero terminated.
1.627 + * If addKeyword is not NULL is must have length small enough to fit in KeywordStruct.keyword.
1.628 + */
1.629 +static int32_t
1.630 +_getKeywords(const char *localeID,
1.631 + char prev,
1.632 + char *keywords, int32_t keywordCapacity,
1.633 + char *values, int32_t valuesCapacity, int32_t *valLen,
1.634 + UBool valuesToo,
1.635 + const char* addKeyword,
1.636 + const char* addValue,
1.637 + UErrorCode *status)
1.638 +{
1.639 + KeywordStruct keywordList[ULOC_MAX_NO_KEYWORDS];
1.640 +
1.641 + int32_t maxKeywords = ULOC_MAX_NO_KEYWORDS;
1.642 + int32_t numKeywords = 0;
1.643 + const char* pos = localeID;
1.644 + const char* equalSign = NULL;
1.645 + const char* semicolon = NULL;
1.646 + int32_t i = 0, j, n;
1.647 + int32_t keywordsLen = 0;
1.648 + int32_t valuesLen = 0;
1.649 +
1.650 + if(prev == '@') { /* start of keyword definition */
1.651 + /* we will grab pairs, trim spaces, lowercase keywords, sort and return */
1.652 + do {
1.653 + UBool duplicate = FALSE;
1.654 + /* skip leading spaces */
1.655 + while(*pos == ' ') {
1.656 + pos++;
1.657 + }
1.658 + if (!*pos) { /* handle trailing "; " */
1.659 + break;
1.660 + }
1.661 + if(numKeywords == maxKeywords) {
1.662 + *status = U_INTERNAL_PROGRAM_ERROR;
1.663 + return 0;
1.664 + }
1.665 + equalSign = uprv_strchr(pos, '=');
1.666 + semicolon = uprv_strchr(pos, ';');
1.667 + /* lack of '=' [foo@currency] is illegal */
1.668 + /* ';' before '=' [foo@currency;collation=pinyin] is illegal */
1.669 + if(!equalSign || (semicolon && semicolon<equalSign)) {
1.670 + *status = U_INVALID_FORMAT_ERROR;
1.671 + return 0;
1.672 + }
1.673 + /* need to normalize both keyword and keyword name */
1.674 + if(equalSign - pos >= ULOC_KEYWORD_BUFFER_LEN) {
1.675 + /* keyword name too long for internal buffer */
1.676 + *status = U_INTERNAL_PROGRAM_ERROR;
1.677 + return 0;
1.678 + }
1.679 + for(i = 0, n = 0; i < equalSign - pos; ++i) {
1.680 + if (pos[i] != ' ') {
1.681 + keywordList[numKeywords].keyword[n++] = uprv_tolower(pos[i]);
1.682 + }
1.683 + }
1.684 +
1.685 + /* zero-length keyword is an error. */
1.686 + if (n == 0) {
1.687 + *status = U_INVALID_FORMAT_ERROR;
1.688 + return 0;
1.689 + }
1.690 +
1.691 + keywordList[numKeywords].keyword[n] = 0;
1.692 + keywordList[numKeywords].keywordLen = n;
1.693 + /* now grab the value part. First we skip the '=' */
1.694 + equalSign++;
1.695 + /* then we leading spaces */
1.696 + while(*equalSign == ' ') {
1.697 + equalSign++;
1.698 + }
1.699 +
1.700 + /* Premature end or zero-length value */
1.701 + if (!equalSign || equalSign == semicolon) {
1.702 + *status = U_INVALID_FORMAT_ERROR;
1.703 + return 0;
1.704 + }
1.705 +
1.706 + keywordList[numKeywords].valueStart = equalSign;
1.707 +
1.708 + pos = semicolon;
1.709 + i = 0;
1.710 + if(pos) {
1.711 + while(*(pos - i - 1) == ' ') {
1.712 + i++;
1.713 + }
1.714 + keywordList[numKeywords].valueLen = (int32_t)(pos - equalSign - i);
1.715 + pos++;
1.716 + } else {
1.717 + i = (int32_t)uprv_strlen(equalSign);
1.718 + while(i && equalSign[i-1] == ' ') {
1.719 + i--;
1.720 + }
1.721 + keywordList[numKeywords].valueLen = i;
1.722 + }
1.723 + /* If this is a duplicate keyword, then ignore it */
1.724 + for (j=0; j<numKeywords; ++j) {
1.725 + if (uprv_strcmp(keywordList[j].keyword, keywordList[numKeywords].keyword) == 0) {
1.726 + duplicate = TRUE;
1.727 + break;
1.728 + }
1.729 + }
1.730 + if (!duplicate) {
1.731 + ++numKeywords;
1.732 + }
1.733 + } while(pos);
1.734 +
1.735 + /* Handle addKeyword/addValue. */
1.736 + if (addKeyword != NULL) {
1.737 + UBool duplicate = FALSE;
1.738 + U_ASSERT(addValue != NULL);
1.739 + /* Search for duplicate; if found, do nothing. Explicit keyword
1.740 + overrides addKeyword. */
1.741 + for (j=0; j<numKeywords; ++j) {
1.742 + if (uprv_strcmp(keywordList[j].keyword, addKeyword) == 0) {
1.743 + duplicate = TRUE;
1.744 + break;
1.745 + }
1.746 + }
1.747 + if (!duplicate) {
1.748 + if (numKeywords == maxKeywords) {
1.749 + *status = U_INTERNAL_PROGRAM_ERROR;
1.750 + return 0;
1.751 + }
1.752 + uprv_strcpy(keywordList[numKeywords].keyword, addKeyword);
1.753 + keywordList[numKeywords].keywordLen = (int32_t)uprv_strlen(addKeyword);
1.754 + keywordList[numKeywords].valueStart = addValue;
1.755 + keywordList[numKeywords].valueLen = (int32_t)uprv_strlen(addValue);
1.756 + ++numKeywords;
1.757 + }
1.758 + } else {
1.759 + U_ASSERT(addValue == NULL);
1.760 + }
1.761 +
1.762 + /* now we have a list of keywords */
1.763 + /* we need to sort it */
1.764 + uprv_sortArray(keywordList, numKeywords, sizeof(KeywordStruct), compareKeywordStructs, NULL, FALSE, status);
1.765 +
1.766 + /* Now construct the keyword part */
1.767 + for(i = 0; i < numKeywords; i++) {
1.768 + if(keywordsLen + keywordList[i].keywordLen + 1< keywordCapacity) {
1.769 + uprv_strcpy(keywords+keywordsLen, keywordList[i].keyword);
1.770 + if(valuesToo) {
1.771 + keywords[keywordsLen + keywordList[i].keywordLen] = '=';
1.772 + } else {
1.773 + keywords[keywordsLen + keywordList[i].keywordLen] = 0;
1.774 + }
1.775 + }
1.776 + keywordsLen += keywordList[i].keywordLen + 1;
1.777 + if(valuesToo) {
1.778 + if(keywordsLen + keywordList[i].valueLen < keywordCapacity) {
1.779 + uprv_strncpy(keywords+keywordsLen, keywordList[i].valueStart, keywordList[i].valueLen);
1.780 + }
1.781 + keywordsLen += keywordList[i].valueLen;
1.782 +
1.783 + if(i < numKeywords - 1) {
1.784 + if(keywordsLen < keywordCapacity) {
1.785 + keywords[keywordsLen] = ';';
1.786 + }
1.787 + keywordsLen++;
1.788 + }
1.789 + }
1.790 + if(values) {
1.791 + if(valuesLen + keywordList[i].valueLen + 1< valuesCapacity) {
1.792 + uprv_strcpy(values+valuesLen, keywordList[i].valueStart);
1.793 + values[valuesLen + keywordList[i].valueLen] = 0;
1.794 + }
1.795 + valuesLen += keywordList[i].valueLen + 1;
1.796 + }
1.797 + }
1.798 + if(values) {
1.799 + values[valuesLen] = 0;
1.800 + if(valLen) {
1.801 + *valLen = valuesLen;
1.802 + }
1.803 + }
1.804 + return u_terminateChars(keywords, keywordCapacity, keywordsLen, status);
1.805 + } else {
1.806 + return 0;
1.807 + }
1.808 +}
1.809 +
1.810 +U_CFUNC int32_t
1.811 +locale_getKeywords(const char *localeID,
1.812 + char prev,
1.813 + char *keywords, int32_t keywordCapacity,
1.814 + char *values, int32_t valuesCapacity, int32_t *valLen,
1.815 + UBool valuesToo,
1.816 + UErrorCode *status) {
1.817 + return _getKeywords(localeID, prev, keywords, keywordCapacity,
1.818 + values, valuesCapacity, valLen, valuesToo,
1.819 + NULL, NULL, status);
1.820 +}
1.821 +
1.822 +U_CAPI int32_t U_EXPORT2
1.823 +uloc_getKeywordValue(const char* localeID,
1.824 + const char* keywordName,
1.825 + char* buffer, int32_t bufferCapacity,
1.826 + UErrorCode* status)
1.827 +{
1.828 + const char* startSearchHere = NULL;
1.829 + const char* nextSeparator = NULL;
1.830 + char keywordNameBuffer[ULOC_KEYWORD_BUFFER_LEN];
1.831 + char localeKeywordNameBuffer[ULOC_KEYWORD_BUFFER_LEN];
1.832 + int32_t i = 0;
1.833 + int32_t result = 0;
1.834 +
1.835 + if(status && U_SUCCESS(*status) && localeID) {
1.836 + char tempBuffer[ULOC_FULLNAME_CAPACITY];
1.837 + const char* tmpLocaleID;
1.838 +
1.839 + if (_hasBCP47Extension(localeID)) {
1.840 + _ConvertBCP47(tmpLocaleID, localeID, tempBuffer, sizeof(tempBuffer), status);
1.841 + } else {
1.842 + tmpLocaleID=localeID;
1.843 + }
1.844 +
1.845 + startSearchHere = uprv_strchr(tmpLocaleID, '@'); /* TODO: REVISIT: shouldn't this be locale_getKeywordsStart ? */
1.846 + if(startSearchHere == NULL) {
1.847 + /* no keywords, return at once */
1.848 + return 0;
1.849 + }
1.850 +
1.851 + locale_canonKeywordName(keywordNameBuffer, keywordName, status);
1.852 + if(U_FAILURE(*status)) {
1.853 + return 0;
1.854 + }
1.855 +
1.856 + /* find the first keyword */
1.857 + while(startSearchHere) {
1.858 + startSearchHere++;
1.859 + /* skip leading spaces (allowed?) */
1.860 + while(*startSearchHere == ' ') {
1.861 + startSearchHere++;
1.862 + }
1.863 + nextSeparator = uprv_strchr(startSearchHere, '=');
1.864 + /* need to normalize both keyword and keyword name */
1.865 + if(!nextSeparator) {
1.866 + break;
1.867 + }
1.868 + if(nextSeparator - startSearchHere >= ULOC_KEYWORD_BUFFER_LEN) {
1.869 + /* keyword name too long for internal buffer */
1.870 + *status = U_INTERNAL_PROGRAM_ERROR;
1.871 + return 0;
1.872 + }
1.873 + for(i = 0; i < nextSeparator - startSearchHere; i++) {
1.874 + localeKeywordNameBuffer[i] = uprv_tolower(startSearchHere[i]);
1.875 + }
1.876 + /* trim trailing spaces */
1.877 + while(startSearchHere[i-1] == ' ') {
1.878 + i--;
1.879 + U_ASSERT(i>=0);
1.880 + }
1.881 + localeKeywordNameBuffer[i] = 0;
1.882 +
1.883 + startSearchHere = uprv_strchr(nextSeparator, ';');
1.884 +
1.885 + if(uprv_strcmp(keywordNameBuffer, localeKeywordNameBuffer) == 0) {
1.886 + nextSeparator++;
1.887 + while(*nextSeparator == ' ') {
1.888 + nextSeparator++;
1.889 + }
1.890 + /* we actually found the keyword. Copy the value */
1.891 + if(startSearchHere && startSearchHere - nextSeparator < bufferCapacity) {
1.892 + while(*(startSearchHere-1) == ' ') {
1.893 + startSearchHere--;
1.894 + }
1.895 + uprv_strncpy(buffer, nextSeparator, startSearchHere - nextSeparator);
1.896 + result = u_terminateChars(buffer, bufferCapacity, (int32_t)(startSearchHere - nextSeparator), status);
1.897 + } else if(!startSearchHere && (int32_t)uprv_strlen(nextSeparator) < bufferCapacity) { /* last item in string */
1.898 + i = (int32_t)uprv_strlen(nextSeparator);
1.899 + while(nextSeparator[i - 1] == ' ') {
1.900 + i--;
1.901 + }
1.902 + uprv_strncpy(buffer, nextSeparator, i);
1.903 + result = u_terminateChars(buffer, bufferCapacity, i, status);
1.904 + } else {
1.905 + /* give a bigger buffer, please */
1.906 + *status = U_BUFFER_OVERFLOW_ERROR;
1.907 + if(startSearchHere) {
1.908 + result = (int32_t)(startSearchHere - nextSeparator);
1.909 + } else {
1.910 + result = (int32_t)uprv_strlen(nextSeparator);
1.911 + }
1.912 + }
1.913 + return result;
1.914 + }
1.915 + }
1.916 + }
1.917 + return 0;
1.918 +}
1.919 +
1.920 +U_CAPI int32_t U_EXPORT2
1.921 +uloc_setKeywordValue(const char* keywordName,
1.922 + const char* keywordValue,
1.923 + char* buffer, int32_t bufferCapacity,
1.924 + UErrorCode* status)
1.925 +{
1.926 + /* TODO: sorting. removal. */
1.927 + int32_t keywordNameLen;
1.928 + int32_t keywordValueLen;
1.929 + int32_t bufLen;
1.930 + int32_t needLen = 0;
1.931 + int32_t foundValueLen;
1.932 + int32_t keywordAtEnd = 0; /* is the keyword at the end of the string? */
1.933 + char keywordNameBuffer[ULOC_KEYWORD_BUFFER_LEN];
1.934 + char localeKeywordNameBuffer[ULOC_KEYWORD_BUFFER_LEN];
1.935 + int32_t i = 0;
1.936 + int32_t rc;
1.937 + char* nextSeparator = NULL;
1.938 + char* nextEqualsign = NULL;
1.939 + char* startSearchHere = NULL;
1.940 + char* keywordStart = NULL;
1.941 + char *insertHere = NULL;
1.942 + if(U_FAILURE(*status)) {
1.943 + return -1;
1.944 + }
1.945 + if(bufferCapacity>1) {
1.946 + bufLen = (int32_t)uprv_strlen(buffer);
1.947 + } else {
1.948 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.949 + return 0;
1.950 + }
1.951 + if(bufferCapacity<bufLen) {
1.952 + /* The capacity is less than the length?! Is this NULL terminated? */
1.953 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.954 + return 0;
1.955 + }
1.956 + if(keywordValue && !*keywordValue) {
1.957 + keywordValue = NULL;
1.958 + }
1.959 + if(keywordValue) {
1.960 + keywordValueLen = (int32_t)uprv_strlen(keywordValue);
1.961 + } else {
1.962 + keywordValueLen = 0;
1.963 + }
1.964 + keywordNameLen = locale_canonKeywordName(keywordNameBuffer, keywordName, status);
1.965 + if(U_FAILURE(*status)) {
1.966 + return 0;
1.967 + }
1.968 + startSearchHere = (char*)locale_getKeywordsStart(buffer);
1.969 + if(startSearchHere == NULL || (startSearchHere[1]==0)) {
1.970 + if(!keywordValue) { /* no keywords = nothing to remove */
1.971 + return bufLen;
1.972 + }
1.973 +
1.974 + needLen = bufLen+1+keywordNameLen+1+keywordValueLen;
1.975 + if(startSearchHere) { /* had a single @ */
1.976 + needLen--; /* already had the @ */
1.977 + /* startSearchHere points at the @ */
1.978 + } else {
1.979 + startSearchHere=buffer+bufLen;
1.980 + }
1.981 + if(needLen >= bufferCapacity) {
1.982 + *status = U_BUFFER_OVERFLOW_ERROR;
1.983 + return needLen; /* no change */
1.984 + }
1.985 + *startSearchHere = '@';
1.986 + startSearchHere++;
1.987 + uprv_strcpy(startSearchHere, keywordNameBuffer);
1.988 + startSearchHere += keywordNameLen;
1.989 + *startSearchHere = '=';
1.990 + startSearchHere++;
1.991 + uprv_strcpy(startSearchHere, keywordValue);
1.992 + startSearchHere+=keywordValueLen;
1.993 + return needLen;
1.994 + } /* end shortcut - no @ */
1.995 +
1.996 + keywordStart = startSearchHere;
1.997 + /* search for keyword */
1.998 + while(keywordStart) {
1.999 + keywordStart++;
1.1000 + /* skip leading spaces (allowed?) */
1.1001 + while(*keywordStart == ' ') {
1.1002 + keywordStart++;
1.1003 + }
1.1004 + nextEqualsign = uprv_strchr(keywordStart, '=');
1.1005 + /* need to normalize both keyword and keyword name */
1.1006 + if(!nextEqualsign) {
1.1007 + break;
1.1008 + }
1.1009 + if(nextEqualsign - keywordStart >= ULOC_KEYWORD_BUFFER_LEN) {
1.1010 + /* keyword name too long for internal buffer */
1.1011 + *status = U_INTERNAL_PROGRAM_ERROR;
1.1012 + return 0;
1.1013 + }
1.1014 + for(i = 0; i < nextEqualsign - keywordStart; i++) {
1.1015 + localeKeywordNameBuffer[i] = uprv_tolower(keywordStart[i]);
1.1016 + }
1.1017 + /* trim trailing spaces */
1.1018 + while(keywordStart[i-1] == ' ') {
1.1019 + i--;
1.1020 + }
1.1021 + U_ASSERT(i>=0 && i<ULOC_KEYWORD_BUFFER_LEN);
1.1022 + localeKeywordNameBuffer[i] = 0;
1.1023 +
1.1024 + nextSeparator = uprv_strchr(nextEqualsign, ';');
1.1025 + rc = uprv_strcmp(keywordNameBuffer, localeKeywordNameBuffer);
1.1026 + if(rc == 0) {
1.1027 + nextEqualsign++;
1.1028 + while(*nextEqualsign == ' ') {
1.1029 + nextEqualsign++;
1.1030 + }
1.1031 + /* we actually found the keyword. Change the value */
1.1032 + if (nextSeparator) {
1.1033 + keywordAtEnd = 0;
1.1034 + foundValueLen = (int32_t)(nextSeparator - nextEqualsign);
1.1035 + } else {
1.1036 + keywordAtEnd = 1;
1.1037 + foundValueLen = (int32_t)uprv_strlen(nextEqualsign);
1.1038 + }
1.1039 + if(keywordValue) { /* adding a value - not removing */
1.1040 + if(foundValueLen == keywordValueLen) {
1.1041 + uprv_strncpy(nextEqualsign, keywordValue, keywordValueLen);
1.1042 + return bufLen; /* no change in size */
1.1043 + } else if(foundValueLen > keywordValueLen) {
1.1044 + int32_t delta = foundValueLen - keywordValueLen;
1.1045 + if(nextSeparator) { /* RH side */
1.1046 + uprv_memmove(nextSeparator - delta, nextSeparator, bufLen-(nextSeparator-buffer));
1.1047 + }
1.1048 + uprv_strncpy(nextEqualsign, keywordValue, keywordValueLen);
1.1049 + bufLen -= delta;
1.1050 + buffer[bufLen]=0;
1.1051 + return bufLen;
1.1052 + } else { /* FVL < KVL */
1.1053 + int32_t delta = keywordValueLen - foundValueLen;
1.1054 + if((bufLen+delta) >= bufferCapacity) {
1.1055 + *status = U_BUFFER_OVERFLOW_ERROR;
1.1056 + return bufLen+delta;
1.1057 + }
1.1058 + if(nextSeparator) { /* RH side */
1.1059 + uprv_memmove(nextSeparator+delta,nextSeparator, bufLen-(nextSeparator-buffer));
1.1060 + }
1.1061 + uprv_strncpy(nextEqualsign, keywordValue, keywordValueLen);
1.1062 + bufLen += delta;
1.1063 + buffer[bufLen]=0;
1.1064 + return bufLen;
1.1065 + }
1.1066 + } else { /* removing a keyword */
1.1067 + if(keywordAtEnd) {
1.1068 + /* zero out the ';' or '@' just before startSearchhere */
1.1069 + keywordStart[-1] = 0;
1.1070 + return (int32_t)((keywordStart-buffer)-1); /* (string length without keyword) minus separator */
1.1071 + } else {
1.1072 + uprv_memmove(keywordStart, nextSeparator+1, bufLen-((nextSeparator+1)-buffer));
1.1073 + keywordStart[bufLen-((nextSeparator+1)-buffer)]=0;
1.1074 + return (int32_t)(bufLen-((nextSeparator+1)-keywordStart));
1.1075 + }
1.1076 + }
1.1077 + } else if(rc<0){ /* end match keyword */
1.1078 + /* could insert at this location. */
1.1079 + insertHere = keywordStart;
1.1080 + }
1.1081 + keywordStart = nextSeparator;
1.1082 + } /* end loop searching */
1.1083 +
1.1084 + if(!keywordValue) {
1.1085 + return bufLen; /* removal of non-extant keyword - no change */
1.1086 + }
1.1087 +
1.1088 + /* we know there is at least one keyword. */
1.1089 + needLen = bufLen+1+keywordNameLen+1+keywordValueLen;
1.1090 + if(needLen >= bufferCapacity) {
1.1091 + *status = U_BUFFER_OVERFLOW_ERROR;
1.1092 + return needLen; /* no change */
1.1093 + }
1.1094 +
1.1095 + if(insertHere) {
1.1096 + uprv_memmove(insertHere+(1+keywordNameLen+1+keywordValueLen), insertHere, bufLen-(insertHere-buffer));
1.1097 + keywordStart = insertHere;
1.1098 + } else {
1.1099 + keywordStart = buffer+bufLen;
1.1100 + *keywordStart = ';';
1.1101 + keywordStart++;
1.1102 + }
1.1103 + uprv_strncpy(keywordStart, keywordNameBuffer, keywordNameLen);
1.1104 + keywordStart += keywordNameLen;
1.1105 + *keywordStart = '=';
1.1106 + keywordStart++;
1.1107 + uprv_strncpy(keywordStart, keywordValue, keywordValueLen); /* terminates. */
1.1108 + keywordStart+=keywordValueLen;
1.1109 + if(insertHere) {
1.1110 + *keywordStart = ';';
1.1111 + keywordStart++;
1.1112 + }
1.1113 + buffer[needLen]=0;
1.1114 + return needLen;
1.1115 +}
1.1116 +
1.1117 +/* ### ID parsing implementation **************************************************/
1.1118 +
1.1119 +#define _isPrefixLetter(a) ((a=='x')||(a=='X')||(a=='i')||(a=='I'))
1.1120 +
1.1121 +/*returns TRUE if one of the special prefixes is here (s=string)
1.1122 + 'x-' or 'i-' */
1.1123 +#define _isIDPrefix(s) (_isPrefixLetter(s[0])&&_isIDSeparator(s[1]))
1.1124 +
1.1125 +/* Dot terminates it because of POSIX form where dot precedes the codepage
1.1126 + * except for variant
1.1127 + */
1.1128 +#define _isTerminator(a) ((a==0)||(a=='.')||(a=='@'))
1.1129 +
1.1130 +static char* _strnchr(const char* str, int32_t len, char c) {
1.1131 + U_ASSERT(str != 0 && len >= 0);
1.1132 + while (len-- != 0) {
1.1133 + char d = *str;
1.1134 + if (d == c) {
1.1135 + return (char*) str;
1.1136 + } else if (d == 0) {
1.1137 + break;
1.1138 + }
1.1139 + ++str;
1.1140 + }
1.1141 + return NULL;
1.1142 +}
1.1143 +
1.1144 +/**
1.1145 + * Lookup 'key' in the array 'list'. The array 'list' should contain
1.1146 + * a NULL entry, followed by more entries, and a second NULL entry.
1.1147 + *
1.1148 + * The 'list' param should be LANGUAGES, LANGUAGES_3, COUNTRIES, or
1.1149 + * COUNTRIES_3.
1.1150 + */
1.1151 +static int16_t _findIndex(const char* const* list, const char* key)
1.1152 +{
1.1153 + const char* const* anchor = list;
1.1154 + int32_t pass = 0;
1.1155 +
1.1156 + /* Make two passes through two NULL-terminated arrays at 'list' */
1.1157 + while (pass++ < 2) {
1.1158 + while (*list) {
1.1159 + if (uprv_strcmp(key, *list) == 0) {
1.1160 + return (int16_t)(list - anchor);
1.1161 + }
1.1162 + list++;
1.1163 + }
1.1164 + ++list; /* skip final NULL *CWB*/
1.1165 + }
1.1166 + return -1;
1.1167 +}
1.1168 +
1.1169 +/* count the length of src while copying it to dest; return strlen(src) */
1.1170 +static inline int32_t
1.1171 +_copyCount(char *dest, int32_t destCapacity, const char *src) {
1.1172 + const char *anchor;
1.1173 + char c;
1.1174 +
1.1175 + anchor=src;
1.1176 + for(;;) {
1.1177 + if((c=*src)==0) {
1.1178 + return (int32_t)(src-anchor);
1.1179 + }
1.1180 + if(destCapacity<=0) {
1.1181 + return (int32_t)((src-anchor)+uprv_strlen(src));
1.1182 + }
1.1183 + ++src;
1.1184 + *dest++=c;
1.1185 + --destCapacity;
1.1186 + }
1.1187 +}
1.1188 +
1.1189 +U_CFUNC const char*
1.1190 +uloc_getCurrentCountryID(const char* oldID){
1.1191 + int32_t offset = _findIndex(DEPRECATED_COUNTRIES, oldID);
1.1192 + if (offset >= 0) {
1.1193 + return REPLACEMENT_COUNTRIES[offset];
1.1194 + }
1.1195 + return oldID;
1.1196 +}
1.1197 +U_CFUNC const char*
1.1198 +uloc_getCurrentLanguageID(const char* oldID){
1.1199 + int32_t offset = _findIndex(DEPRECATED_LANGUAGES, oldID);
1.1200 + if (offset >= 0) {
1.1201 + return REPLACEMENT_LANGUAGES[offset];
1.1202 + }
1.1203 + return oldID;
1.1204 +}
1.1205 +/*
1.1206 + * the internal functions _getLanguage(), _getCountry(), _getVariant()
1.1207 + * avoid duplicating code to handle the earlier locale ID pieces
1.1208 + * in the functions for the later ones by
1.1209 + * setting the *pEnd pointer to where they stopped parsing
1.1210 + *
1.1211 + * TODO try to use this in Locale
1.1212 + */
1.1213 +U_CFUNC int32_t
1.1214 +ulocimp_getLanguage(const char *localeID,
1.1215 + char *language, int32_t languageCapacity,
1.1216 + const char **pEnd) {
1.1217 + int32_t i=0;
1.1218 + int32_t offset;
1.1219 + char lang[4]={ 0, 0, 0, 0 }; /* temporary buffer to hold language code for searching */
1.1220 +
1.1221 + /* if it starts with i- or x- then copy that prefix */
1.1222 + if(_isIDPrefix(localeID)) {
1.1223 + if(i<languageCapacity) {
1.1224 + language[i]=(char)uprv_tolower(*localeID);
1.1225 + }
1.1226 + if(i<languageCapacity) {
1.1227 + language[i+1]='-';
1.1228 + }
1.1229 + i+=2;
1.1230 + localeID+=2;
1.1231 + }
1.1232 +
1.1233 + /* copy the language as far as possible and count its length */
1.1234 + while(!_isTerminator(*localeID) && !_isIDSeparator(*localeID)) {
1.1235 + if(i<languageCapacity) {
1.1236 + language[i]=(char)uprv_tolower(*localeID);
1.1237 + }
1.1238 + if(i<3) {
1.1239 + U_ASSERT(i>=0);
1.1240 + lang[i]=(char)uprv_tolower(*localeID);
1.1241 + }
1.1242 + i++;
1.1243 + localeID++;
1.1244 + }
1.1245 +
1.1246 + if(i==3) {
1.1247 + /* convert 3 character code to 2 character code if possible *CWB*/
1.1248 + offset=_findIndex(LANGUAGES_3, lang);
1.1249 + if(offset>=0) {
1.1250 + i=_copyCount(language, languageCapacity, LANGUAGES[offset]);
1.1251 + }
1.1252 + }
1.1253 +
1.1254 + if(pEnd!=NULL) {
1.1255 + *pEnd=localeID;
1.1256 + }
1.1257 + return i;
1.1258 +}
1.1259 +
1.1260 +U_CFUNC int32_t
1.1261 +ulocimp_getScript(const char *localeID,
1.1262 + char *script, int32_t scriptCapacity,
1.1263 + const char **pEnd)
1.1264 +{
1.1265 + int32_t idLen = 0;
1.1266 +
1.1267 + if (pEnd != NULL) {
1.1268 + *pEnd = localeID;
1.1269 + }
1.1270 +
1.1271 + /* copy the second item as far as possible and count its length */
1.1272 + while(!_isTerminator(localeID[idLen]) && !_isIDSeparator(localeID[idLen])
1.1273 + && uprv_isASCIILetter(localeID[idLen])) {
1.1274 + idLen++;
1.1275 + }
1.1276 +
1.1277 + /* If it's exactly 4 characters long, then it's a script and not a country. */
1.1278 + if (idLen == 4) {
1.1279 + int32_t i;
1.1280 + if (pEnd != NULL) {
1.1281 + *pEnd = localeID+idLen;
1.1282 + }
1.1283 + if(idLen > scriptCapacity) {
1.1284 + idLen = scriptCapacity;
1.1285 + }
1.1286 + if (idLen >= 1) {
1.1287 + script[0]=(char)uprv_toupper(*(localeID++));
1.1288 + }
1.1289 + for (i = 1; i < idLen; i++) {
1.1290 + script[i]=(char)uprv_tolower(*(localeID++));
1.1291 + }
1.1292 + }
1.1293 + else {
1.1294 + idLen = 0;
1.1295 + }
1.1296 + return idLen;
1.1297 +}
1.1298 +
1.1299 +U_CFUNC int32_t
1.1300 +ulocimp_getCountry(const char *localeID,
1.1301 + char *country, int32_t countryCapacity,
1.1302 + const char **pEnd)
1.1303 +{
1.1304 + int32_t idLen=0;
1.1305 + char cnty[ULOC_COUNTRY_CAPACITY]={ 0, 0, 0, 0 };
1.1306 + int32_t offset;
1.1307 +
1.1308 + /* copy the country as far as possible and count its length */
1.1309 + while(!_isTerminator(localeID[idLen]) && !_isIDSeparator(localeID[idLen])) {
1.1310 + if(idLen<(ULOC_COUNTRY_CAPACITY-1)) { /*CWB*/
1.1311 + cnty[idLen]=(char)uprv_toupper(localeID[idLen]);
1.1312 + }
1.1313 + idLen++;
1.1314 + }
1.1315 +
1.1316 + /* the country should be either length 2 or 3 */
1.1317 + if (idLen == 2 || idLen == 3) {
1.1318 + UBool gotCountry = FALSE;
1.1319 + /* convert 3 character code to 2 character code if possible *CWB*/
1.1320 + if(idLen==3) {
1.1321 + offset=_findIndex(COUNTRIES_3, cnty);
1.1322 + if(offset>=0) {
1.1323 + idLen=_copyCount(country, countryCapacity, COUNTRIES[offset]);
1.1324 + gotCountry = TRUE;
1.1325 + }
1.1326 + }
1.1327 + if (!gotCountry) {
1.1328 + int32_t i = 0;
1.1329 + for (i = 0; i < idLen; i++) {
1.1330 + if (i < countryCapacity) {
1.1331 + country[i]=(char)uprv_toupper(localeID[i]);
1.1332 + }
1.1333 + }
1.1334 + }
1.1335 + localeID+=idLen;
1.1336 + } else {
1.1337 + idLen = 0;
1.1338 + }
1.1339 +
1.1340 + if(pEnd!=NULL) {
1.1341 + *pEnd=localeID;
1.1342 + }
1.1343 +
1.1344 + return idLen;
1.1345 +}
1.1346 +
1.1347 +/**
1.1348 + * @param needSeparator if true, then add leading '_' if any variants
1.1349 + * are added to 'variant'
1.1350 + */
1.1351 +static int32_t
1.1352 +_getVariantEx(const char *localeID,
1.1353 + char prev,
1.1354 + char *variant, int32_t variantCapacity,
1.1355 + UBool needSeparator) {
1.1356 + int32_t i=0;
1.1357 +
1.1358 + /* get one or more variant tags and separate them with '_' */
1.1359 + if(_isIDSeparator(prev)) {
1.1360 + /* get a variant string after a '-' or '_' */
1.1361 + while(!_isTerminator(*localeID)) {
1.1362 + if (needSeparator) {
1.1363 + if (i<variantCapacity) {
1.1364 + variant[i] = '_';
1.1365 + }
1.1366 + ++i;
1.1367 + needSeparator = FALSE;
1.1368 + }
1.1369 + if(i<variantCapacity) {
1.1370 + variant[i]=(char)uprv_toupper(*localeID);
1.1371 + if(variant[i]=='-') {
1.1372 + variant[i]='_';
1.1373 + }
1.1374 + }
1.1375 + i++;
1.1376 + localeID++;
1.1377 + }
1.1378 + }
1.1379 +
1.1380 + /* if there is no variant tag after a '-' or '_' then look for '@' */
1.1381 + if(i==0) {
1.1382 + if(prev=='@') {
1.1383 + /* keep localeID */
1.1384 + } else if((localeID=locale_getKeywordsStart(localeID))!=NULL) {
1.1385 + ++localeID; /* point after the '@' */
1.1386 + } else {
1.1387 + return 0;
1.1388 + }
1.1389 + while(!_isTerminator(*localeID)) {
1.1390 + if (needSeparator) {
1.1391 + if (i<variantCapacity) {
1.1392 + variant[i] = '_';
1.1393 + }
1.1394 + ++i;
1.1395 + needSeparator = FALSE;
1.1396 + }
1.1397 + if(i<variantCapacity) {
1.1398 + variant[i]=(char)uprv_toupper(*localeID);
1.1399 + if(variant[i]=='-' || variant[i]==',') {
1.1400 + variant[i]='_';
1.1401 + }
1.1402 + }
1.1403 + i++;
1.1404 + localeID++;
1.1405 + }
1.1406 + }
1.1407 +
1.1408 + return i;
1.1409 +}
1.1410 +
1.1411 +static int32_t
1.1412 +_getVariant(const char *localeID,
1.1413 + char prev,
1.1414 + char *variant, int32_t variantCapacity) {
1.1415 + return _getVariantEx(localeID, prev, variant, variantCapacity, FALSE);
1.1416 +}
1.1417 +
1.1418 +/**
1.1419 + * Delete ALL instances of a variant from the given list of one or
1.1420 + * more variants. Example: "FOO_EURO_BAR_EURO" => "FOO_BAR".
1.1421 + * @param variants the source string of one or more variants,
1.1422 + * separated by '_'. This will be MODIFIED IN PLACE. Not zero
1.1423 + * terminated; if it is, trailing zero will NOT be maintained.
1.1424 + * @param variantsLen length of variants
1.1425 + * @param toDelete variant to delete, without separators, e.g. "EURO"
1.1426 + * or "PREEURO"; not zero terminated
1.1427 + * @param toDeleteLen length of toDelete
1.1428 + * @return number of characters deleted from variants
1.1429 + */
1.1430 +static int32_t
1.1431 +_deleteVariant(char* variants, int32_t variantsLen,
1.1432 + const char* toDelete, int32_t toDeleteLen)
1.1433 +{
1.1434 + int32_t delta = 0; /* number of chars deleted */
1.1435 + for (;;) {
1.1436 + UBool flag = FALSE;
1.1437 + if (variantsLen < toDeleteLen) {
1.1438 + return delta;
1.1439 + }
1.1440 + if (uprv_strncmp(variants, toDelete, toDeleteLen) == 0 &&
1.1441 + (variantsLen == toDeleteLen ||
1.1442 + (flag=(variants[toDeleteLen] == '_'))))
1.1443 + {
1.1444 + int32_t d = toDeleteLen + (flag?1:0);
1.1445 + variantsLen -= d;
1.1446 + delta += d;
1.1447 + if (variantsLen > 0) {
1.1448 + uprv_memmove(variants, variants+d, variantsLen);
1.1449 + }
1.1450 + } else {
1.1451 + char* p = _strnchr(variants, variantsLen, '_');
1.1452 + if (p == NULL) {
1.1453 + return delta;
1.1454 + }
1.1455 + ++p;
1.1456 + variantsLen -= (int32_t)(p - variants);
1.1457 + variants = p;
1.1458 + }
1.1459 + }
1.1460 +}
1.1461 +
1.1462 +/* Keyword enumeration */
1.1463 +
1.1464 +typedef struct UKeywordsContext {
1.1465 + char* keywords;
1.1466 + char* current;
1.1467 +} UKeywordsContext;
1.1468 +
1.1469 +static void U_CALLCONV
1.1470 +uloc_kw_closeKeywords(UEnumeration *enumerator) {
1.1471 + uprv_free(((UKeywordsContext *)enumerator->context)->keywords);
1.1472 + uprv_free(enumerator->context);
1.1473 + uprv_free(enumerator);
1.1474 +}
1.1475 +
1.1476 +static int32_t U_CALLCONV
1.1477 +uloc_kw_countKeywords(UEnumeration *en, UErrorCode * /*status*/) {
1.1478 + char *kw = ((UKeywordsContext *)en->context)->keywords;
1.1479 + int32_t result = 0;
1.1480 + while(*kw) {
1.1481 + result++;
1.1482 + kw += uprv_strlen(kw)+1;
1.1483 + }
1.1484 + return result;
1.1485 +}
1.1486 +
1.1487 +static const char* U_CALLCONV
1.1488 +uloc_kw_nextKeyword(UEnumeration* en,
1.1489 + int32_t* resultLength,
1.1490 + UErrorCode* /*status*/) {
1.1491 + const char* result = ((UKeywordsContext *)en->context)->current;
1.1492 + int32_t len = 0;
1.1493 + if(*result) {
1.1494 + len = (int32_t)uprv_strlen(((UKeywordsContext *)en->context)->current);
1.1495 + ((UKeywordsContext *)en->context)->current += len+1;
1.1496 + } else {
1.1497 + result = NULL;
1.1498 + }
1.1499 + if (resultLength) {
1.1500 + *resultLength = len;
1.1501 + }
1.1502 + return result;
1.1503 +}
1.1504 +
1.1505 +static void U_CALLCONV
1.1506 +uloc_kw_resetKeywords(UEnumeration* en,
1.1507 + UErrorCode* /*status*/) {
1.1508 + ((UKeywordsContext *)en->context)->current = ((UKeywordsContext *)en->context)->keywords;
1.1509 +}
1.1510 +
1.1511 +static const UEnumeration gKeywordsEnum = {
1.1512 + NULL,
1.1513 + NULL,
1.1514 + uloc_kw_closeKeywords,
1.1515 + uloc_kw_countKeywords,
1.1516 + uenum_unextDefault,
1.1517 + uloc_kw_nextKeyword,
1.1518 + uloc_kw_resetKeywords
1.1519 +};
1.1520 +
1.1521 +U_CAPI UEnumeration* U_EXPORT2
1.1522 +uloc_openKeywordList(const char *keywordList, int32_t keywordListSize, UErrorCode* status)
1.1523 +{
1.1524 + UKeywordsContext *myContext = NULL;
1.1525 + UEnumeration *result = NULL;
1.1526 +
1.1527 + if(U_FAILURE(*status)) {
1.1528 + return NULL;
1.1529 + }
1.1530 + result = (UEnumeration *)uprv_malloc(sizeof(UEnumeration));
1.1531 + /* Null pointer test */
1.1532 + if (result == NULL) {
1.1533 + *status = U_MEMORY_ALLOCATION_ERROR;
1.1534 + return NULL;
1.1535 + }
1.1536 + uprv_memcpy(result, &gKeywordsEnum, sizeof(UEnumeration));
1.1537 + myContext = static_cast<UKeywordsContext *>(uprv_malloc(sizeof(UKeywordsContext)));
1.1538 + if (myContext == NULL) {
1.1539 + *status = U_MEMORY_ALLOCATION_ERROR;
1.1540 + uprv_free(result);
1.1541 + return NULL;
1.1542 + }
1.1543 + myContext->keywords = (char *)uprv_malloc(keywordListSize+1);
1.1544 + uprv_memcpy(myContext->keywords, keywordList, keywordListSize);
1.1545 + myContext->keywords[keywordListSize] = 0;
1.1546 + myContext->current = myContext->keywords;
1.1547 + result->context = myContext;
1.1548 + return result;
1.1549 +}
1.1550 +
1.1551 +U_CAPI UEnumeration* U_EXPORT2
1.1552 +uloc_openKeywords(const char* localeID,
1.1553 + UErrorCode* status)
1.1554 +{
1.1555 + int32_t i=0;
1.1556 + char keywords[256];
1.1557 + int32_t keywordsCapacity = 256;
1.1558 + char tempBuffer[ULOC_FULLNAME_CAPACITY];
1.1559 + const char* tmpLocaleID;
1.1560 +
1.1561 + if(status==NULL || U_FAILURE(*status)) {
1.1562 + return 0;
1.1563 + }
1.1564 +
1.1565 + if (_hasBCP47Extension(localeID)) {
1.1566 + _ConvertBCP47(tmpLocaleID, localeID, tempBuffer, sizeof(tempBuffer), status);
1.1567 + } else {
1.1568 + if (localeID==NULL) {
1.1569 + localeID=uloc_getDefault();
1.1570 + }
1.1571 + tmpLocaleID=localeID;
1.1572 + }
1.1573 +
1.1574 + /* Skip the language */
1.1575 + ulocimp_getLanguage(tmpLocaleID, NULL, 0, &tmpLocaleID);
1.1576 + if(_isIDSeparator(*tmpLocaleID)) {
1.1577 + const char *scriptID;
1.1578 + /* Skip the script if available */
1.1579 + ulocimp_getScript(tmpLocaleID+1, NULL, 0, &scriptID);
1.1580 + if(scriptID != tmpLocaleID+1) {
1.1581 + /* Found optional script */
1.1582 + tmpLocaleID = scriptID;
1.1583 + }
1.1584 + /* Skip the Country */
1.1585 + if (_isIDSeparator(*tmpLocaleID)) {
1.1586 + ulocimp_getCountry(tmpLocaleID+1, NULL, 0, &tmpLocaleID);
1.1587 + if(_isIDSeparator(*tmpLocaleID)) {
1.1588 + _getVariant(tmpLocaleID+1, *tmpLocaleID, NULL, 0);
1.1589 + }
1.1590 + }
1.1591 + }
1.1592 +
1.1593 + /* keywords are located after '@' */
1.1594 + if((tmpLocaleID = locale_getKeywordsStart(tmpLocaleID)) != NULL) {
1.1595 + i=locale_getKeywords(tmpLocaleID+1, '@', keywords, keywordsCapacity, NULL, 0, NULL, FALSE, status);
1.1596 + }
1.1597 +
1.1598 + if(i) {
1.1599 + return uloc_openKeywordList(keywords, i, status);
1.1600 + } else {
1.1601 + return NULL;
1.1602 + }
1.1603 +}
1.1604 +
1.1605 +
1.1606 +/* bit-flags for 'options' parameter of _canonicalize */
1.1607 +#define _ULOC_STRIP_KEYWORDS 0x2
1.1608 +#define _ULOC_CANONICALIZE 0x1
1.1609 +
1.1610 +#define OPTION_SET(options, mask) ((options & mask) != 0)
1.1611 +
1.1612 +static const char i_default[] = {'i', '-', 'd', 'e', 'f', 'a', 'u', 'l', 't'};
1.1613 +#define I_DEFAULT_LENGTH (sizeof i_default / sizeof i_default[0])
1.1614 +
1.1615 +/**
1.1616 + * Canonicalize the given localeID, to level 1 or to level 2,
1.1617 + * depending on the options. To specify level 1, pass in options=0.
1.1618 + * To specify level 2, pass in options=_ULOC_CANONICALIZE.
1.1619 + *
1.1620 + * This is the code underlying uloc_getName and uloc_canonicalize.
1.1621 + */
1.1622 +static int32_t
1.1623 +_canonicalize(const char* localeID,
1.1624 + char* result,
1.1625 + int32_t resultCapacity,
1.1626 + uint32_t options,
1.1627 + UErrorCode* err) {
1.1628 + int32_t j, len, fieldCount=0, scriptSize=0, variantSize=0, nameCapacity;
1.1629 + char localeBuffer[ULOC_FULLNAME_CAPACITY];
1.1630 + char tempBuffer[ULOC_FULLNAME_CAPACITY];
1.1631 + const char* origLocaleID;
1.1632 + const char* tmpLocaleID;
1.1633 + const char* keywordAssign = NULL;
1.1634 + const char* separatorIndicator = NULL;
1.1635 + const char* addKeyword = NULL;
1.1636 + const char* addValue = NULL;
1.1637 + char* name;
1.1638 + char* variant = NULL; /* pointer into name, or NULL */
1.1639 +
1.1640 + if (U_FAILURE(*err)) {
1.1641 + return 0;
1.1642 + }
1.1643 +
1.1644 + if (_hasBCP47Extension(localeID)) {
1.1645 + _ConvertBCP47(tmpLocaleID, localeID, tempBuffer, sizeof(tempBuffer), err);
1.1646 + } else {
1.1647 + if (localeID==NULL) {
1.1648 + localeID=uloc_getDefault();
1.1649 + }
1.1650 + tmpLocaleID=localeID;
1.1651 + }
1.1652 +
1.1653 + origLocaleID=tmpLocaleID;
1.1654 +
1.1655 + /* if we are doing a full canonicalization, then put results in
1.1656 + localeBuffer, if necessary; otherwise send them to result. */
1.1657 + if (/*OPTION_SET(options, _ULOC_CANONICALIZE) &&*/
1.1658 + (result == NULL || resultCapacity < (int32_t)sizeof(localeBuffer))) {
1.1659 + name = localeBuffer;
1.1660 + nameCapacity = (int32_t)sizeof(localeBuffer);
1.1661 + } else {
1.1662 + name = result;
1.1663 + nameCapacity = resultCapacity;
1.1664 + }
1.1665 +
1.1666 + /* get all pieces, one after another, and separate with '_' */
1.1667 + len=ulocimp_getLanguage(tmpLocaleID, name, nameCapacity, &tmpLocaleID);
1.1668 +
1.1669 + if(len == I_DEFAULT_LENGTH && uprv_strncmp(origLocaleID, i_default, len) == 0) {
1.1670 + const char *d = uloc_getDefault();
1.1671 +
1.1672 + len = (int32_t)uprv_strlen(d);
1.1673 +
1.1674 + if (name != NULL) {
1.1675 + uprv_strncpy(name, d, len);
1.1676 + }
1.1677 + } else if(_isIDSeparator(*tmpLocaleID)) {
1.1678 + const char *scriptID;
1.1679 +
1.1680 + ++fieldCount;
1.1681 + if(len<nameCapacity) {
1.1682 + name[len]='_';
1.1683 + }
1.1684 + ++len;
1.1685 +
1.1686 + scriptSize=ulocimp_getScript(tmpLocaleID+1,
1.1687 + (len<nameCapacity ? name+len : NULL), nameCapacity-len, &scriptID);
1.1688 + if(scriptSize > 0) {
1.1689 + /* Found optional script */
1.1690 + tmpLocaleID = scriptID;
1.1691 + ++fieldCount;
1.1692 + len+=scriptSize;
1.1693 + if (_isIDSeparator(*tmpLocaleID)) {
1.1694 + /* If there is something else, then we add the _ */
1.1695 + if(len<nameCapacity) {
1.1696 + name[len]='_';
1.1697 + }
1.1698 + ++len;
1.1699 + }
1.1700 + }
1.1701 +
1.1702 + if (_isIDSeparator(*tmpLocaleID)) {
1.1703 + const char *cntryID;
1.1704 + int32_t cntrySize = ulocimp_getCountry(tmpLocaleID+1,
1.1705 + (len<nameCapacity ? name+len : NULL), nameCapacity-len, &cntryID);
1.1706 + if (cntrySize > 0) {
1.1707 + /* Found optional country */
1.1708 + tmpLocaleID = cntryID;
1.1709 + len+=cntrySize;
1.1710 + }
1.1711 + if(_isIDSeparator(*tmpLocaleID)) {
1.1712 + /* If there is something else, then we add the _ if we found country before. */
1.1713 + if (cntrySize >= 0 && ! _isIDSeparator(*(tmpLocaleID+1)) ) {
1.1714 + ++fieldCount;
1.1715 + if(len<nameCapacity) {
1.1716 + name[len]='_';
1.1717 + }
1.1718 + ++len;
1.1719 + }
1.1720 +
1.1721 + variantSize = _getVariant(tmpLocaleID+1, *tmpLocaleID,
1.1722 + (len<nameCapacity ? name+len : NULL), nameCapacity-len);
1.1723 + if (variantSize > 0) {
1.1724 + variant = len<nameCapacity ? name+len : NULL;
1.1725 + len += variantSize;
1.1726 + tmpLocaleID += variantSize + 1; /* skip '_' and variant */
1.1727 + }
1.1728 + }
1.1729 + }
1.1730 + }
1.1731 +
1.1732 + /* Copy POSIX-style charset specifier, if any [mr.utf8] */
1.1733 + if (!OPTION_SET(options, _ULOC_CANONICALIZE) && *tmpLocaleID == '.') {
1.1734 + UBool done = FALSE;
1.1735 + do {
1.1736 + char c = *tmpLocaleID;
1.1737 + switch (c) {
1.1738 + case 0:
1.1739 + case '@':
1.1740 + done = TRUE;
1.1741 + break;
1.1742 + default:
1.1743 + if (len<nameCapacity) {
1.1744 + name[len] = c;
1.1745 + }
1.1746 + ++len;
1.1747 + ++tmpLocaleID;
1.1748 + break;
1.1749 + }
1.1750 + } while (!done);
1.1751 + }
1.1752 +
1.1753 + /* Scan ahead to next '@' and determine if it is followed by '=' and/or ';'
1.1754 + After this, tmpLocaleID either points to '@' or is NULL */
1.1755 + if ((tmpLocaleID=locale_getKeywordsStart(tmpLocaleID))!=NULL) {
1.1756 + keywordAssign = uprv_strchr(tmpLocaleID, '=');
1.1757 + separatorIndicator = uprv_strchr(tmpLocaleID, ';');
1.1758 + }
1.1759 +
1.1760 + /* Copy POSIX-style variant, if any [mr@FOO] */
1.1761 + if (!OPTION_SET(options, _ULOC_CANONICALIZE) &&
1.1762 + tmpLocaleID != NULL && keywordAssign == NULL) {
1.1763 + for (;;) {
1.1764 + char c = *tmpLocaleID;
1.1765 + if (c == 0) {
1.1766 + break;
1.1767 + }
1.1768 + if (len<nameCapacity) {
1.1769 + name[len] = c;
1.1770 + }
1.1771 + ++len;
1.1772 + ++tmpLocaleID;
1.1773 + }
1.1774 + }
1.1775 +
1.1776 + if (OPTION_SET(options, _ULOC_CANONICALIZE)) {
1.1777 + /* Handle @FOO variant if @ is present and not followed by = */
1.1778 + if (tmpLocaleID!=NULL && keywordAssign==NULL) {
1.1779 + int32_t posixVariantSize;
1.1780 + /* Add missing '_' if needed */
1.1781 + if (fieldCount < 2 || (fieldCount < 3 && scriptSize > 0)) {
1.1782 + do {
1.1783 + if(len<nameCapacity) {
1.1784 + name[len]='_';
1.1785 + }
1.1786 + ++len;
1.1787 + ++fieldCount;
1.1788 + } while(fieldCount<2);
1.1789 + }
1.1790 + posixVariantSize = _getVariantEx(tmpLocaleID+1, '@', name+len, nameCapacity-len,
1.1791 + (UBool)(variantSize > 0));
1.1792 + if (posixVariantSize > 0) {
1.1793 + if (variant == NULL) {
1.1794 + variant = name+len;
1.1795 + }
1.1796 + len += posixVariantSize;
1.1797 + variantSize += posixVariantSize;
1.1798 + }
1.1799 + }
1.1800 +
1.1801 + /* Handle generic variants first */
1.1802 + if (variant) {
1.1803 + for (j=0; j<(int32_t)(sizeof(VARIANT_MAP)/sizeof(VARIANT_MAP[0])); j++) {
1.1804 + const char* variantToCompare = VARIANT_MAP[j].variant;
1.1805 + int32_t n = (int32_t)uprv_strlen(variantToCompare);
1.1806 + int32_t variantLen = _deleteVariant(variant, uprv_min(variantSize, (nameCapacity-len)), variantToCompare, n);
1.1807 + len -= variantLen;
1.1808 + if (variantLen > 0) {
1.1809 + if (len > 0 && name[len-1] == '_') { /* delete trailing '_' */
1.1810 + --len;
1.1811 + }
1.1812 + addKeyword = VARIANT_MAP[j].keyword;
1.1813 + addValue = VARIANT_MAP[j].value;
1.1814 + break;
1.1815 + }
1.1816 + }
1.1817 + if (len > 0 && len <= nameCapacity && name[len-1] == '_') { /* delete trailing '_' */
1.1818 + --len;
1.1819 + }
1.1820 + }
1.1821 +
1.1822 + /* Look up the ID in the canonicalization map */
1.1823 + for (j=0; j<(int32_t)(sizeof(CANONICALIZE_MAP)/sizeof(CANONICALIZE_MAP[0])); j++) {
1.1824 + const char* id = CANONICALIZE_MAP[j].id;
1.1825 + int32_t n = (int32_t)uprv_strlen(id);
1.1826 + if (len == n && uprv_strncmp(name, id, n) == 0) {
1.1827 + if (n == 0 && tmpLocaleID != NULL) {
1.1828 + break; /* Don't remap "" if keywords present */
1.1829 + }
1.1830 + len = _copyCount(name, nameCapacity, CANONICALIZE_MAP[j].canonicalID);
1.1831 + if (CANONICALIZE_MAP[j].keyword) {
1.1832 + addKeyword = CANONICALIZE_MAP[j].keyword;
1.1833 + addValue = CANONICALIZE_MAP[j].value;
1.1834 + }
1.1835 + break;
1.1836 + }
1.1837 + }
1.1838 + }
1.1839 +
1.1840 + if (!OPTION_SET(options, _ULOC_STRIP_KEYWORDS)) {
1.1841 + if (tmpLocaleID!=NULL && keywordAssign!=NULL &&
1.1842 + (!separatorIndicator || separatorIndicator > keywordAssign)) {
1.1843 + if(len<nameCapacity) {
1.1844 + name[len]='@';
1.1845 + }
1.1846 + ++len;
1.1847 + ++fieldCount;
1.1848 + len += _getKeywords(tmpLocaleID+1, '@', (len<nameCapacity ? name+len : NULL), nameCapacity-len,
1.1849 + NULL, 0, NULL, TRUE, addKeyword, addValue, err);
1.1850 + } else if (addKeyword != NULL) {
1.1851 + U_ASSERT(addValue != NULL && len < nameCapacity);
1.1852 + /* inelegant but works -- later make _getKeywords do this? */
1.1853 + len += _copyCount(name+len, nameCapacity-len, "@");
1.1854 + len += _copyCount(name+len, nameCapacity-len, addKeyword);
1.1855 + len += _copyCount(name+len, nameCapacity-len, "=");
1.1856 + len += _copyCount(name+len, nameCapacity-len, addValue);
1.1857 + }
1.1858 + }
1.1859 +
1.1860 + if (U_SUCCESS(*err) && result != NULL && name == localeBuffer) {
1.1861 + uprv_strncpy(result, localeBuffer, (len > resultCapacity) ? resultCapacity : len);
1.1862 + }
1.1863 +
1.1864 + return u_terminateChars(result, resultCapacity, len, err);
1.1865 +}
1.1866 +
1.1867 +/* ### ID parsing API **************************************************/
1.1868 +
1.1869 +U_CAPI int32_t U_EXPORT2
1.1870 +uloc_getParent(const char* localeID,
1.1871 + char* parent,
1.1872 + int32_t parentCapacity,
1.1873 + UErrorCode* err)
1.1874 +{
1.1875 + const char *lastUnderscore;
1.1876 + int32_t i;
1.1877 +
1.1878 + if (U_FAILURE(*err))
1.1879 + return 0;
1.1880 +
1.1881 + if (localeID == NULL)
1.1882 + localeID = uloc_getDefault();
1.1883 +
1.1884 + lastUnderscore=uprv_strrchr(localeID, '_');
1.1885 + if(lastUnderscore!=NULL) {
1.1886 + i=(int32_t)(lastUnderscore-localeID);
1.1887 + } else {
1.1888 + i=0;
1.1889 + }
1.1890 +
1.1891 + if(i>0 && parent != localeID) {
1.1892 + uprv_memcpy(parent, localeID, uprv_min(i, parentCapacity));
1.1893 + }
1.1894 + return u_terminateChars(parent, parentCapacity, i, err);
1.1895 +}
1.1896 +
1.1897 +U_CAPI int32_t U_EXPORT2
1.1898 +uloc_getLanguage(const char* localeID,
1.1899 + char* language,
1.1900 + int32_t languageCapacity,
1.1901 + UErrorCode* err)
1.1902 +{
1.1903 + /* uloc_getLanguage will return a 2 character iso-639 code if one exists. *CWB*/
1.1904 + int32_t i=0;
1.1905 +
1.1906 + if (err==NULL || U_FAILURE(*err)) {
1.1907 + return 0;
1.1908 + }
1.1909 +
1.1910 + if(localeID==NULL) {
1.1911 + localeID=uloc_getDefault();
1.1912 + }
1.1913 +
1.1914 + i=ulocimp_getLanguage(localeID, language, languageCapacity, NULL);
1.1915 + return u_terminateChars(language, languageCapacity, i, err);
1.1916 +}
1.1917 +
1.1918 +U_CAPI int32_t U_EXPORT2
1.1919 +uloc_getScript(const char* localeID,
1.1920 + char* script,
1.1921 + int32_t scriptCapacity,
1.1922 + UErrorCode* err)
1.1923 +{
1.1924 + int32_t i=0;
1.1925 +
1.1926 + if(err==NULL || U_FAILURE(*err)) {
1.1927 + return 0;
1.1928 + }
1.1929 +
1.1930 + if(localeID==NULL) {
1.1931 + localeID=uloc_getDefault();
1.1932 + }
1.1933 +
1.1934 + /* skip the language */
1.1935 + ulocimp_getLanguage(localeID, NULL, 0, &localeID);
1.1936 + if(_isIDSeparator(*localeID)) {
1.1937 + i=ulocimp_getScript(localeID+1, script, scriptCapacity, NULL);
1.1938 + }
1.1939 + return u_terminateChars(script, scriptCapacity, i, err);
1.1940 +}
1.1941 +
1.1942 +U_CAPI int32_t U_EXPORT2
1.1943 +uloc_getCountry(const char* localeID,
1.1944 + char* country,
1.1945 + int32_t countryCapacity,
1.1946 + UErrorCode* err)
1.1947 +{
1.1948 + int32_t i=0;
1.1949 +
1.1950 + if(err==NULL || U_FAILURE(*err)) {
1.1951 + return 0;
1.1952 + }
1.1953 +
1.1954 + if(localeID==NULL) {
1.1955 + localeID=uloc_getDefault();
1.1956 + }
1.1957 +
1.1958 + /* Skip the language */
1.1959 + ulocimp_getLanguage(localeID, NULL, 0, &localeID);
1.1960 + if(_isIDSeparator(*localeID)) {
1.1961 + const char *scriptID;
1.1962 + /* Skip the script if available */
1.1963 + ulocimp_getScript(localeID+1, NULL, 0, &scriptID);
1.1964 + if(scriptID != localeID+1) {
1.1965 + /* Found optional script */
1.1966 + localeID = scriptID;
1.1967 + }
1.1968 + if(_isIDSeparator(*localeID)) {
1.1969 + i=ulocimp_getCountry(localeID+1, country, countryCapacity, NULL);
1.1970 + }
1.1971 + }
1.1972 + return u_terminateChars(country, countryCapacity, i, err);
1.1973 +}
1.1974 +
1.1975 +U_CAPI int32_t U_EXPORT2
1.1976 +uloc_getVariant(const char* localeID,
1.1977 + char* variant,
1.1978 + int32_t variantCapacity,
1.1979 + UErrorCode* err)
1.1980 +{
1.1981 + char tempBuffer[ULOC_FULLNAME_CAPACITY];
1.1982 + const char* tmpLocaleID;
1.1983 + int32_t i=0;
1.1984 +
1.1985 + if(err==NULL || U_FAILURE(*err)) {
1.1986 + return 0;
1.1987 + }
1.1988 +
1.1989 + if (_hasBCP47Extension(localeID)) {
1.1990 + _ConvertBCP47(tmpLocaleID, localeID, tempBuffer, sizeof(tempBuffer), err);
1.1991 + } else {
1.1992 + if (localeID==NULL) {
1.1993 + localeID=uloc_getDefault();
1.1994 + }
1.1995 + tmpLocaleID=localeID;
1.1996 + }
1.1997 +
1.1998 + /* Skip the language */
1.1999 + ulocimp_getLanguage(tmpLocaleID, NULL, 0, &tmpLocaleID);
1.2000 + if(_isIDSeparator(*tmpLocaleID)) {
1.2001 + const char *scriptID;
1.2002 + /* Skip the script if available */
1.2003 + ulocimp_getScript(tmpLocaleID+1, NULL, 0, &scriptID);
1.2004 + if(scriptID != tmpLocaleID+1) {
1.2005 + /* Found optional script */
1.2006 + tmpLocaleID = scriptID;
1.2007 + }
1.2008 + /* Skip the Country */
1.2009 + if (_isIDSeparator(*tmpLocaleID)) {
1.2010 + const char *cntryID;
1.2011 + ulocimp_getCountry(tmpLocaleID+1, NULL, 0, &cntryID);
1.2012 + if (cntryID != tmpLocaleID+1) {
1.2013 + /* Found optional country */
1.2014 + tmpLocaleID = cntryID;
1.2015 + }
1.2016 + if(_isIDSeparator(*tmpLocaleID)) {
1.2017 + /* If there was no country ID, skip a possible extra IDSeparator */
1.2018 + if (tmpLocaleID != cntryID && _isIDSeparator(tmpLocaleID[1])) {
1.2019 + tmpLocaleID++;
1.2020 + }
1.2021 + i=_getVariant(tmpLocaleID+1, *tmpLocaleID, variant, variantCapacity);
1.2022 + }
1.2023 + }
1.2024 + }
1.2025 +
1.2026 + /* removed by weiv. We don't want to handle POSIX variants anymore. Use canonicalization function */
1.2027 + /* if we do not have a variant tag yet then try a POSIX variant after '@' */
1.2028 +/*
1.2029 + if(!haveVariant && (localeID=uprv_strrchr(localeID, '@'))!=NULL) {
1.2030 + i=_getVariant(localeID+1, '@', variant, variantCapacity);
1.2031 + }
1.2032 +*/
1.2033 + return u_terminateChars(variant, variantCapacity, i, err);
1.2034 +}
1.2035 +
1.2036 +U_CAPI int32_t U_EXPORT2
1.2037 +uloc_getName(const char* localeID,
1.2038 + char* name,
1.2039 + int32_t nameCapacity,
1.2040 + UErrorCode* err)
1.2041 +{
1.2042 + return _canonicalize(localeID, name, nameCapacity, 0, err);
1.2043 +}
1.2044 +
1.2045 +U_CAPI int32_t U_EXPORT2
1.2046 +uloc_getBaseName(const char* localeID,
1.2047 + char* name,
1.2048 + int32_t nameCapacity,
1.2049 + UErrorCode* err)
1.2050 +{
1.2051 + return _canonicalize(localeID, name, nameCapacity, _ULOC_STRIP_KEYWORDS, err);
1.2052 +}
1.2053 +
1.2054 +U_CAPI int32_t U_EXPORT2
1.2055 +uloc_canonicalize(const char* localeID,
1.2056 + char* name,
1.2057 + int32_t nameCapacity,
1.2058 + UErrorCode* err)
1.2059 +{
1.2060 + return _canonicalize(localeID, name, nameCapacity, _ULOC_CANONICALIZE, err);
1.2061 +}
1.2062 +
1.2063 +U_CAPI const char* U_EXPORT2
1.2064 +uloc_getISO3Language(const char* localeID)
1.2065 +{
1.2066 + int16_t offset;
1.2067 + char lang[ULOC_LANG_CAPACITY];
1.2068 + UErrorCode err = U_ZERO_ERROR;
1.2069 +
1.2070 + if (localeID == NULL)
1.2071 + {
1.2072 + localeID = uloc_getDefault();
1.2073 + }
1.2074 + uloc_getLanguage(localeID, lang, ULOC_LANG_CAPACITY, &err);
1.2075 + if (U_FAILURE(err))
1.2076 + return "";
1.2077 + offset = _findIndex(LANGUAGES, lang);
1.2078 + if (offset < 0)
1.2079 + return "";
1.2080 + return LANGUAGES_3[offset];
1.2081 +}
1.2082 +
1.2083 +U_CAPI const char* U_EXPORT2
1.2084 +uloc_getISO3Country(const char* localeID)
1.2085 +{
1.2086 + int16_t offset;
1.2087 + char cntry[ULOC_LANG_CAPACITY];
1.2088 + UErrorCode err = U_ZERO_ERROR;
1.2089 +
1.2090 + if (localeID == NULL)
1.2091 + {
1.2092 + localeID = uloc_getDefault();
1.2093 + }
1.2094 + uloc_getCountry(localeID, cntry, ULOC_LANG_CAPACITY, &err);
1.2095 + if (U_FAILURE(err))
1.2096 + return "";
1.2097 + offset = _findIndex(COUNTRIES, cntry);
1.2098 + if (offset < 0)
1.2099 + return "";
1.2100 +
1.2101 + return COUNTRIES_3[offset];
1.2102 +}
1.2103 +
1.2104 +U_CAPI uint32_t U_EXPORT2
1.2105 +uloc_getLCID(const char* localeID)
1.2106 +{
1.2107 + UErrorCode status = U_ZERO_ERROR;
1.2108 + char langID[ULOC_FULLNAME_CAPACITY];
1.2109 +
1.2110 + uloc_getLanguage(localeID, langID, sizeof(langID), &status);
1.2111 + if (U_FAILURE(status)) {
1.2112 + return 0;
1.2113 + }
1.2114 +
1.2115 + if (uprv_strchr(localeID, '@')) {
1.2116 + // uprv_convertToLCID does not support keywords other than collation.
1.2117 + // Remove all keywords except collation.
1.2118 + int32_t len;
1.2119 + char collVal[ULOC_KEYWORDS_CAPACITY];
1.2120 + char tmpLocaleID[ULOC_FULLNAME_CAPACITY];
1.2121 +
1.2122 + len = uloc_getKeywordValue(localeID, "collation", collVal,
1.2123 + sizeof(collVal)/sizeof(collVal[0]) - 1, &status);
1.2124 +
1.2125 + if (U_SUCCESS(status) && len > 0) {
1.2126 + collVal[len] = 0;
1.2127 +
1.2128 + len = uloc_getBaseName(localeID, tmpLocaleID,
1.2129 + sizeof(tmpLocaleID)/sizeof(tmpLocaleID[0]) - 1, &status);
1.2130 +
1.2131 + if (U_SUCCESS(status)) {
1.2132 + tmpLocaleID[len] = 0;
1.2133 +
1.2134 + len = uloc_setKeywordValue("collation", collVal, tmpLocaleID,
1.2135 + sizeof(tmpLocaleID)/sizeof(tmpLocaleID[0]) - len - 1, &status);
1.2136 +
1.2137 + if (U_SUCCESS(status)) {
1.2138 + tmpLocaleID[len] = 0;
1.2139 + return uprv_convertToLCID(langID, tmpLocaleID, &status);
1.2140 + }
1.2141 + }
1.2142 + }
1.2143 +
1.2144 + // fall through - all keywords are simply ignored
1.2145 + status = U_ZERO_ERROR;
1.2146 + }
1.2147 +
1.2148 + return uprv_convertToLCID(langID, localeID, &status);
1.2149 +}
1.2150 +
1.2151 +U_CAPI int32_t U_EXPORT2
1.2152 +uloc_getLocaleForLCID(uint32_t hostid, char *locale, int32_t localeCapacity,
1.2153 + UErrorCode *status)
1.2154 +{
1.2155 + return uprv_convertToPosix(hostid, locale, localeCapacity, status);
1.2156 +}
1.2157 +
1.2158 +/* ### Default locale **************************************************/
1.2159 +
1.2160 +U_CAPI const char* U_EXPORT2
1.2161 +uloc_getDefault()
1.2162 +{
1.2163 + return locale_get_default();
1.2164 +}
1.2165 +
1.2166 +U_CAPI void U_EXPORT2
1.2167 +uloc_setDefault(const char* newDefaultLocale,
1.2168 + UErrorCode* err)
1.2169 +{
1.2170 + if (U_FAILURE(*err))
1.2171 + return;
1.2172 + /* the error code isn't currently used for anything by this function*/
1.2173 +
1.2174 + /* propagate change to C++ */
1.2175 + locale_set_default(newDefaultLocale);
1.2176 +}
1.2177 +
1.2178 +/**
1.2179 + * Returns a list of all 2-letter language codes defined in ISO 639. This is a pointer
1.2180 + * to an array of pointers to arrays of char. All of these pointers are owned
1.2181 + * by ICU-- do not delete them, and do not write through them. The array is
1.2182 + * terminated with a null pointer.
1.2183 + */
1.2184 +U_CAPI const char* const* U_EXPORT2
1.2185 +uloc_getISOLanguages()
1.2186 +{
1.2187 + return LANGUAGES;
1.2188 +}
1.2189 +
1.2190 +/**
1.2191 + * Returns a list of all 2-letter country codes defined in ISO 639. This is a
1.2192 + * pointer to an array of pointers to arrays of char. All of these pointers are
1.2193 + * owned by ICU-- do not delete them, and do not write through them. The array is
1.2194 + * terminated with a null pointer.
1.2195 + */
1.2196 +U_CAPI const char* const* U_EXPORT2
1.2197 +uloc_getISOCountries()
1.2198 +{
1.2199 + return COUNTRIES;
1.2200 +}
1.2201 +
1.2202 +
1.2203 +/* this function to be moved into cstring.c later */
1.2204 +static char gDecimal = 0;
1.2205 +
1.2206 +static /* U_CAPI */
1.2207 +double
1.2208 +/* U_EXPORT2 */
1.2209 +_uloc_strtod(const char *start, char **end) {
1.2210 + char *decimal;
1.2211 + char *myEnd;
1.2212 + char buf[30];
1.2213 + double rv;
1.2214 + if (!gDecimal) {
1.2215 + char rep[5];
1.2216 + /* For machines that decide to change the decimal on you,
1.2217 + and try to be too smart with localization.
1.2218 + This normally should be just a '.'. */
1.2219 + sprintf(rep, "%+1.1f", 1.0);
1.2220 + gDecimal = rep[2];
1.2221 + }
1.2222 +
1.2223 + if(gDecimal == '.') {
1.2224 + return uprv_strtod(start, end); /* fall through to OS */
1.2225 + } else {
1.2226 + uprv_strncpy(buf, start, 29);
1.2227 + buf[29]=0;
1.2228 + decimal = uprv_strchr(buf, '.');
1.2229 + if(decimal) {
1.2230 + *decimal = gDecimal;
1.2231 + } else {
1.2232 + return uprv_strtod(start, end); /* no decimal point */
1.2233 + }
1.2234 + rv = uprv_strtod(buf, &myEnd);
1.2235 + if(end) {
1.2236 + *end = (char*)(start+(myEnd-buf)); /* cast away const (to follow uprv_strtod API.) */
1.2237 + }
1.2238 + return rv;
1.2239 + }
1.2240 +}
1.2241 +
1.2242 +typedef struct {
1.2243 + float q;
1.2244 + int32_t dummy; /* to avoid uninitialized memory copy from qsort */
1.2245 + char *locale;
1.2246 +} _acceptLangItem;
1.2247 +
1.2248 +static int32_t U_CALLCONV
1.2249 +uloc_acceptLanguageCompare(const void * /*context*/, const void *a, const void *b)
1.2250 +{
1.2251 + const _acceptLangItem *aa = (const _acceptLangItem*)a;
1.2252 + const _acceptLangItem *bb = (const _acceptLangItem*)b;
1.2253 +
1.2254 + int32_t rc = 0;
1.2255 + if(bb->q < aa->q) {
1.2256 + rc = -1; /* A > B */
1.2257 + } else if(bb->q > aa->q) {
1.2258 + rc = 1; /* A < B */
1.2259 + } else {
1.2260 + rc = 0; /* A = B */
1.2261 + }
1.2262 +
1.2263 + if(rc==0) {
1.2264 + rc = uprv_stricmp(aa->locale, bb->locale);
1.2265 + }
1.2266 +
1.2267 +#if defined(ULOC_DEBUG)
1.2268 + /* fprintf(stderr, "a:[%s:%g], b:[%s:%g] -> %d\n",
1.2269 + aa->locale, aa->q,
1.2270 + bb->locale, bb->q,
1.2271 + rc);*/
1.2272 +#endif
1.2273 +
1.2274 + return rc;
1.2275 +}
1.2276 +
1.2277 +/*
1.2278 +mt-mt, ja;q=0.76, en-us;q=0.95, en;q=0.92, en-gb;q=0.89, fr;q=0.87, iu-ca;q=0.84, iu;q=0.82, ja-jp;q=0.79, mt;q=0.97, de-de;q=0.74, de;q=0.71, es;q=0.68, it-it;q=0.66, it;q=0.63, vi-vn;q=0.61, vi;q=0.58, nl-nl;q=0.55, nl;q=0.53
1.2279 +*/
1.2280 +
1.2281 +U_CAPI int32_t U_EXPORT2
1.2282 +uloc_acceptLanguageFromHTTP(char *result, int32_t resultAvailable, UAcceptResult *outResult,
1.2283 + const char *httpAcceptLanguage,
1.2284 + UEnumeration* availableLocales,
1.2285 + UErrorCode *status)
1.2286 +{
1.2287 + _acceptLangItem *j;
1.2288 + _acceptLangItem smallBuffer[30];
1.2289 + char **strs;
1.2290 + char tmp[ULOC_FULLNAME_CAPACITY +1];
1.2291 + int32_t n = 0;
1.2292 + const char *itemEnd;
1.2293 + const char *paramEnd;
1.2294 + const char *s;
1.2295 + const char *t;
1.2296 + int32_t res;
1.2297 + int32_t i;
1.2298 + int32_t l = (int32_t)uprv_strlen(httpAcceptLanguage);
1.2299 + int32_t jSize;
1.2300 + char *tempstr; /* Use for null pointer check */
1.2301 +
1.2302 + j = smallBuffer;
1.2303 + jSize = sizeof(smallBuffer)/sizeof(smallBuffer[0]);
1.2304 + if(U_FAILURE(*status)) {
1.2305 + return -1;
1.2306 + }
1.2307 +
1.2308 + for(s=httpAcceptLanguage;s&&*s;) {
1.2309 + while(isspace(*s)) /* eat space at the beginning */
1.2310 + s++;
1.2311 + itemEnd=uprv_strchr(s,',');
1.2312 + paramEnd=uprv_strchr(s,';');
1.2313 + if(!itemEnd) {
1.2314 + itemEnd = httpAcceptLanguage+l; /* end of string */
1.2315 + }
1.2316 + if(paramEnd && paramEnd<itemEnd) {
1.2317 + /* semicolon (;) is closer than end (,) */
1.2318 + t = paramEnd+1;
1.2319 + if(*t=='q') {
1.2320 + t++;
1.2321 + }
1.2322 + while(isspace(*t)) {
1.2323 + t++;
1.2324 + }
1.2325 + if(*t=='=') {
1.2326 + t++;
1.2327 + }
1.2328 + while(isspace(*t)) {
1.2329 + t++;
1.2330 + }
1.2331 + j[n].q = (float)_uloc_strtod(t,NULL);
1.2332 + } else {
1.2333 + /* no semicolon - it's 1.0 */
1.2334 + j[n].q = 1.0f;
1.2335 + paramEnd = itemEnd;
1.2336 + }
1.2337 + j[n].dummy=0;
1.2338 + /* eat spaces prior to semi */
1.2339 + for(t=(paramEnd-1);(paramEnd>s)&&isspace(*t);t--)
1.2340 + ;
1.2341 + /* Check for null pointer from uprv_strndup */
1.2342 + tempstr = uprv_strndup(s,(int32_t)((t+1)-s));
1.2343 + if (tempstr == NULL) {
1.2344 + *status = U_MEMORY_ALLOCATION_ERROR;
1.2345 + return -1;
1.2346 + }
1.2347 + j[n].locale = tempstr;
1.2348 + uloc_canonicalize(j[n].locale,tmp,sizeof(tmp)/sizeof(tmp[0]),status);
1.2349 + if(strcmp(j[n].locale,tmp)) {
1.2350 + uprv_free(j[n].locale);
1.2351 + j[n].locale=uprv_strdup(tmp);
1.2352 + }
1.2353 +#if defined(ULOC_DEBUG)
1.2354 + /*fprintf(stderr,"%d: s <%s> q <%g>\n", n, j[n].locale, j[n].q);*/
1.2355 +#endif
1.2356 + n++;
1.2357 + s = itemEnd;
1.2358 + while(*s==',') { /* eat duplicate commas */
1.2359 + s++;
1.2360 + }
1.2361 + if(n>=jSize) {
1.2362 + if(j==smallBuffer) { /* overflowed the small buffer. */
1.2363 + j = static_cast<_acceptLangItem *>(uprv_malloc(sizeof(j[0])*(jSize*2)));
1.2364 + if(j!=NULL) {
1.2365 + uprv_memcpy(j,smallBuffer,sizeof(j[0])*jSize);
1.2366 + }
1.2367 +#if defined(ULOC_DEBUG)
1.2368 + fprintf(stderr,"malloced at size %d\n", jSize);
1.2369 +#endif
1.2370 + } else {
1.2371 + j = static_cast<_acceptLangItem *>(uprv_realloc(j, sizeof(j[0])*jSize*2));
1.2372 +#if defined(ULOC_DEBUG)
1.2373 + fprintf(stderr,"re-alloced at size %d\n", jSize);
1.2374 +#endif
1.2375 + }
1.2376 + jSize *= 2;
1.2377 + if(j==NULL) {
1.2378 + *status = U_MEMORY_ALLOCATION_ERROR;
1.2379 + return -1;
1.2380 + }
1.2381 + }
1.2382 + }
1.2383 + uprv_sortArray(j, n, sizeof(j[0]), uloc_acceptLanguageCompare, NULL, TRUE, status);
1.2384 + if(U_FAILURE(*status)) {
1.2385 + if(j != smallBuffer) {
1.2386 +#if defined(ULOC_DEBUG)
1.2387 + fprintf(stderr,"freeing j %p\n", j);
1.2388 +#endif
1.2389 + uprv_free(j);
1.2390 + }
1.2391 + return -1;
1.2392 + }
1.2393 + strs = static_cast<char **>(uprv_malloc((size_t)(sizeof(strs[0])*n)));
1.2394 + /* Check for null pointer */
1.2395 + if (strs == NULL) {
1.2396 + uprv_free(j); /* Free to avoid memory leak */
1.2397 + *status = U_MEMORY_ALLOCATION_ERROR;
1.2398 + return -1;
1.2399 + }
1.2400 + for(i=0;i<n;i++) {
1.2401 +#if defined(ULOC_DEBUG)
1.2402 + /*fprintf(stderr,"%d: s <%s> q <%g>\n", i, j[i].locale, j[i].q);*/
1.2403 +#endif
1.2404 + strs[i]=j[i].locale;
1.2405 + }
1.2406 + res = uloc_acceptLanguage(result, resultAvailable, outResult,
1.2407 + (const char**)strs, n, availableLocales, status);
1.2408 + for(i=0;i<n;i++) {
1.2409 + uprv_free(strs[i]);
1.2410 + }
1.2411 + uprv_free(strs);
1.2412 + if(j != smallBuffer) {
1.2413 +#if defined(ULOC_DEBUG)
1.2414 + fprintf(stderr,"freeing j %p\n", j);
1.2415 +#endif
1.2416 + uprv_free(j);
1.2417 + }
1.2418 + return res;
1.2419 +}
1.2420 +
1.2421 +
1.2422 +U_CAPI int32_t U_EXPORT2
1.2423 +uloc_acceptLanguage(char *result, int32_t resultAvailable,
1.2424 + UAcceptResult *outResult, const char **acceptList,
1.2425 + int32_t acceptListCount,
1.2426 + UEnumeration* availableLocales,
1.2427 + UErrorCode *status)
1.2428 +{
1.2429 + int32_t i,j;
1.2430 + int32_t len;
1.2431 + int32_t maxLen=0;
1.2432 + char tmp[ULOC_FULLNAME_CAPACITY+1];
1.2433 + const char *l;
1.2434 + char **fallbackList;
1.2435 + if(U_FAILURE(*status)) {
1.2436 + return -1;
1.2437 + }
1.2438 + fallbackList = static_cast<char **>(uprv_malloc((size_t)(sizeof(fallbackList[0])*acceptListCount)));
1.2439 + if(fallbackList==NULL) {
1.2440 + *status = U_MEMORY_ALLOCATION_ERROR;
1.2441 + return -1;
1.2442 + }
1.2443 + for(i=0;i<acceptListCount;i++) {
1.2444 +#if defined(ULOC_DEBUG)
1.2445 + fprintf(stderr,"%02d: %s\n", i, acceptList[i]);
1.2446 +#endif
1.2447 + while((l=uenum_next(availableLocales, NULL, status))) {
1.2448 +#if defined(ULOC_DEBUG)
1.2449 + fprintf(stderr," %s\n", l);
1.2450 +#endif
1.2451 + len = (int32_t)uprv_strlen(l);
1.2452 + if(!uprv_strcmp(acceptList[i], l)) {
1.2453 + if(outResult) {
1.2454 + *outResult = ULOC_ACCEPT_VALID;
1.2455 + }
1.2456 +#if defined(ULOC_DEBUG)
1.2457 + fprintf(stderr, "MATCH! %s\n", l);
1.2458 +#endif
1.2459 + if(len>0) {
1.2460 + uprv_strncpy(result, l, uprv_min(len, resultAvailable));
1.2461 + }
1.2462 + for(j=0;j<i;j++) {
1.2463 + uprv_free(fallbackList[j]);
1.2464 + }
1.2465 + uprv_free(fallbackList);
1.2466 + return u_terminateChars(result, resultAvailable, len, status);
1.2467 + }
1.2468 + if(len>maxLen) {
1.2469 + maxLen = len;
1.2470 + }
1.2471 + }
1.2472 + uenum_reset(availableLocales, status);
1.2473 + /* save off parent info */
1.2474 + if(uloc_getParent(acceptList[i], tmp, sizeof(tmp)/sizeof(tmp[0]), status)!=0) {
1.2475 + fallbackList[i] = uprv_strdup(tmp);
1.2476 + } else {
1.2477 + fallbackList[i]=0;
1.2478 + }
1.2479 + }
1.2480 +
1.2481 + for(maxLen--;maxLen>0;maxLen--) {
1.2482 + for(i=0;i<acceptListCount;i++) {
1.2483 + if(fallbackList[i] && ((int32_t)uprv_strlen(fallbackList[i])==maxLen)) {
1.2484 +#if defined(ULOC_DEBUG)
1.2485 + fprintf(stderr,"Try: [%s]", fallbackList[i]);
1.2486 +#endif
1.2487 + while((l=uenum_next(availableLocales, NULL, status))) {
1.2488 +#if defined(ULOC_DEBUG)
1.2489 + fprintf(stderr," %s\n", l);
1.2490 +#endif
1.2491 + len = (int32_t)uprv_strlen(l);
1.2492 + if(!uprv_strcmp(fallbackList[i], l)) {
1.2493 + if(outResult) {
1.2494 + *outResult = ULOC_ACCEPT_FALLBACK;
1.2495 + }
1.2496 +#if defined(ULOC_DEBUG)
1.2497 + fprintf(stderr, "fallback MATCH! %s\n", l);
1.2498 +#endif
1.2499 + if(len>0) {
1.2500 + uprv_strncpy(result, l, uprv_min(len, resultAvailable));
1.2501 + }
1.2502 + for(j=0;j<acceptListCount;j++) {
1.2503 + uprv_free(fallbackList[j]);
1.2504 + }
1.2505 + uprv_free(fallbackList);
1.2506 + return u_terminateChars(result, resultAvailable, len, status);
1.2507 + }
1.2508 + }
1.2509 + uenum_reset(availableLocales, status);
1.2510 +
1.2511 + if(uloc_getParent(fallbackList[i], tmp, sizeof(tmp)/sizeof(tmp[0]), status)!=0) {
1.2512 + uprv_free(fallbackList[i]);
1.2513 + fallbackList[i] = uprv_strdup(tmp);
1.2514 + } else {
1.2515 + uprv_free(fallbackList[i]);
1.2516 + fallbackList[i]=0;
1.2517 + }
1.2518 + }
1.2519 + }
1.2520 + if(outResult) {
1.2521 + *outResult = ULOC_ACCEPT_FAILED;
1.2522 + }
1.2523 + }
1.2524 + for(i=0;i<acceptListCount;i++) {
1.2525 + uprv_free(fallbackList[i]);
1.2526 + }
1.2527 + uprv_free(fallbackList);
1.2528 + return -1;
1.2529 +}
1.2530 +
1.2531 +/*eof*/
