| File: | root/firefox-clang/intl/icu/source/i18n/tzfmt.cpp |
| Warning: | line 2118, column 9 Value stored to 'numDigits' is never read |
Press '?' to see keyboard shortcuts
Keyboard shortcuts:
| 1 | // © 2016 and later: Unicode, Inc. and others. |
| 2 | // License & terms of use: http://www.unicode.org/copyright.html |
| 3 | /* |
| 4 | ******************************************************************************* |
| 5 | * Copyright (C) 2011-2015, International Business Machines Corporation and |
| 6 | * others. All Rights Reserved. |
| 7 | ******************************************************************************* |
| 8 | */ |
| 9 | |
| 10 | #include "unicode/utypes.h" |
| 11 | |
| 12 | #if !UCONFIG_NO_FORMATTING0 |
| 13 | |
| 14 | #include "unicode/calendar.h" |
| 15 | #include "unicode/tzfmt.h" |
| 16 | #include "unicode/numsys.h" |
| 17 | #include "unicode/strenum.h" |
| 18 | #include "unicode/uchar.h" |
| 19 | #include "unicode/udat.h" |
| 20 | #include "unicode/ustring.h" |
| 21 | #include "unicode/utf16.h" |
| 22 | #include "charstr.h" |
| 23 | #include "tzgnames.h" |
| 24 | #include "cmemory.h" |
| 25 | #include "cstring.h" |
| 26 | #include "putilimp.h" |
| 27 | #include "uassert.h" |
| 28 | #include "ucln_in.h" |
| 29 | #include "ulocimp.h" |
| 30 | #include "umutex.h" |
| 31 | #include "uresimp.h" |
| 32 | #include "ureslocs.h" |
| 33 | #include "uvector.h" |
| 34 | #include "zonemeta.h" |
| 35 | #include "tznames_impl.h" // TextTrieMap |
| 36 | #include "patternprops.h" |
| 37 | |
| 38 | U_NAMESPACE_BEGINnamespace icu_78 { |
| 39 | |
| 40 | // Bit flags used by the parse method. |
| 41 | // The order must match UTimeZoneFormatStyle enum. |
| 42 | #define ISO_Z_STYLE_FLAG0x0080 0x0080 |
| 43 | #define ISO_LOCAL_STYLE_FLAG0x0100 0x0100 |
| 44 | static const int16_t STYLE_PARSE_FLAGS[] = { |
| 45 | 0x0001, // UTZFMT_STYLE_GENERIC_LOCATION, |
| 46 | 0x0002, // UTZFMT_STYLE_GENERIC_LONG, |
| 47 | 0x0004, // UTZFMT_STYLE_GENERIC_SHORT, |
| 48 | 0x0008, // UTZFMT_STYLE_SPECIFIC_LONG, |
| 49 | 0x0010, // UTZFMT_STYLE_SPECIFIC_SHORT, |
| 50 | 0x0020, // UTZFMT_STYLE_LOCALIZED_GMT, |
| 51 | 0x0040, // UTZFMT_STYLE_LOCALIZED_GMT_SHORT, |
| 52 | ISO_Z_STYLE_FLAG0x0080, // UTZFMT_STYLE_ISO_BASIC_SHORT, |
| 53 | ISO_LOCAL_STYLE_FLAG0x0100, // UTZFMT_STYLE_ISO_BASIC_LOCAL_SHORT, |
| 54 | ISO_Z_STYLE_FLAG0x0080, // UTZFMT_STYLE_ISO_BASIC_FIXED, |
| 55 | ISO_LOCAL_STYLE_FLAG0x0100, // UTZFMT_STYLE_ISO_BASIC_LOCAL_FIXED, |
| 56 | ISO_Z_STYLE_FLAG0x0080, // UTZFMT_STYLE_ISO_BASIC_FULL, |
| 57 | ISO_LOCAL_STYLE_FLAG0x0100, // UTZFMT_STYLE_ISO_BASIC_LOCAL_FULL, |
| 58 | ISO_Z_STYLE_FLAG0x0080, // UTZFMT_STYLE_ISO_EXTENDED_FIXED, |
| 59 | ISO_LOCAL_STYLE_FLAG0x0100, // UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FIXED, |
| 60 | ISO_Z_STYLE_FLAG0x0080, // UTZFMT_STYLE_ISO_EXTENDED_FULL, |
| 61 | ISO_LOCAL_STYLE_FLAG0x0100, // UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FULL, |
| 62 | 0x0200, // UTZFMT_STYLE_ZONE_ID, |
| 63 | 0x0400, // UTZFMT_STYLE_ZONE_ID_SHORT, |
| 64 | 0x0800 // UTZFMT_STYLE_EXEMPLAR_LOCATION |
| 65 | }; |
| 66 | |
| 67 | static const char gZoneStringsTag[] = "zoneStrings"; |
| 68 | static const char gGmtFormatTag[]= "gmtFormat"; |
| 69 | static const char gGmtZeroFormatTag[] = "gmtZeroFormat"; |
| 70 | static const char gHourFormatTag[]= "hourFormat"; |
| 71 | |
| 72 | static const char16_t TZID_GMT[] = {0x0045, 0x0074, 0x0063, 0x002F, 0x0047, 0x004D, 0x0054, 0}; // Etc/GMT |
| 73 | static const char16_t UNKNOWN_ZONE_ID[] = { |
| 74 | 0x0045, 0x0074, 0x0063, 0x002F, 0x0055, 0x006E, 0x006B, 0x006E, 0x006F, 0x0077, 0x006E, 0}; // Etc/Unknown |
| 75 | static const char16_t UNKNOWN_SHORT_ZONE_ID[] = {0x0075, 0x006E, 0x006B, 0}; // unk |
| 76 | static const char16_t UNKNOWN_LOCATION[] = {0x0055, 0x006E, 0x006B, 0x006E, 0x006F, 0x0077, 0x006E, 0}; // Unknown |
| 77 | |
| 78 | static const char16_t DEFAULT_GMT_PATTERN[] = {0x0047, 0x004D, 0x0054, 0x007B, 0x0030, 0x007D, 0}; // GMT{0} |
| 79 | //static const char16_t DEFAULT_GMT_ZERO[] = {0x0047, 0x004D, 0x0054, 0}; // GMT |
| 80 | static const char16_t DEFAULT_GMT_POSITIVE_HM[] = {0x002B, 0x0048, 0x003A, 0x006D, 0x006D, 0}; // +H:mm |
| 81 | static const char16_t DEFAULT_GMT_POSITIVE_HMS[] = {0x002B, 0x0048, 0x003A, 0x006D, 0x006D, 0x003A, 0x0073, 0x0073, 0}; // +H:mm:ss |
| 82 | static const char16_t DEFAULT_GMT_NEGATIVE_HM[] = {0x002D, 0x0048, 0x003A, 0x006D, 0x006D, 0}; // -H:mm |
| 83 | static const char16_t DEFAULT_GMT_NEGATIVE_HMS[] = {0x002D, 0x0048, 0x003A, 0x006D, 0x006D, 0x003A, 0x0073, 0x0073, 0}; // -H:mm:ss |
| 84 | static const char16_t DEFAULT_GMT_POSITIVE_H[] = {0x002B, 0x0048, 0}; // +H |
| 85 | static const char16_t DEFAULT_GMT_NEGATIVE_H[] = {0x002D, 0x0048, 0}; // -H |
| 86 | |
| 87 | static const UChar32 DEFAULT_GMT_DIGITS[] = { |
| 88 | 0x0030, 0x0031, 0x0032, 0x0033, 0x0034, |
| 89 | 0x0035, 0x0036, 0x0037, 0x0038, 0x0039 |
| 90 | }; |
| 91 | |
| 92 | static const char16_t DEFAULT_GMT_OFFSET_SEP = 0x003A; // ':' |
| 93 | |
| 94 | static const char16_t ARG0[] = {0x007B, 0x0030, 0x007D}; // "{0}" |
| 95 | static const int32_t ARG0_LEN = 3; |
| 96 | |
| 97 | static const char16_t DEFAULT_GMT_OFFSET_MINUTE_PATTERN[] = {0x006D, 0x006D, 0}; // "mm" |
| 98 | static const char16_t DEFAULT_GMT_OFFSET_SECOND_PATTERN[] = {0x0073, 0x0073, 0}; // "ss" |
| 99 | |
| 100 | static const char16_t ALT_GMT_STRINGS[][4] = { |
| 101 | {0x0047, 0x004D, 0x0054, 0}, // GMT |
| 102 | {0x0055, 0x0054, 0x0043, 0}, // UTC |
| 103 | {0x0055, 0x0054, 0, 0}, // UT |
| 104 | {0, 0, 0, 0} |
| 105 | }; |
| 106 | |
| 107 | // Order of GMT offset pattern parsing, *_HMS must be evaluated first |
| 108 | // because *_HM is most likely a substring of *_HMS |
| 109 | static const int32_t PARSE_GMT_OFFSET_TYPES[] = { |
| 110 | UTZFMT_PAT_POSITIVE_HMS, |
| 111 | UTZFMT_PAT_NEGATIVE_HMS, |
| 112 | UTZFMT_PAT_POSITIVE_HM, |
| 113 | UTZFMT_PAT_NEGATIVE_HM, |
| 114 | UTZFMT_PAT_POSITIVE_H, |
| 115 | UTZFMT_PAT_NEGATIVE_H, |
| 116 | -1 |
| 117 | }; |
| 118 | |
| 119 | static const char16_t SINGLEQUOTE = 0x0027; |
| 120 | static const char16_t PLUS = 0x002B; |
| 121 | static const char16_t MINUS = 0x002D; |
| 122 | static const char16_t ISO8601_UTC = 0x005A; // 'Z' |
| 123 | static const char16_t ISO8601_SEP = 0x003A; // ':' |
| 124 | |
| 125 | static const int32_t MILLIS_PER_HOUR = 60 * 60 * 1000; |
| 126 | static const int32_t MILLIS_PER_MINUTE = 60 * 1000; |
| 127 | static const int32_t MILLIS_PER_SECOND = 1000; |
| 128 | |
| 129 | // Maximum offset (exclusive) in millisecond supported by offset formats |
| 130 | static int32_t MAX_OFFSET = 24 * MILLIS_PER_HOUR; |
| 131 | |
| 132 | // Maximum values for GMT offset fields |
| 133 | static const int32_t MAX_OFFSET_HOUR = 23; |
| 134 | static const int32_t MAX_OFFSET_MINUTE = 59; |
| 135 | static const int32_t MAX_OFFSET_SECOND = 59; |
| 136 | |
| 137 | static const int32_t UNKNOWN_OFFSET = 0x7FFFFFFF; |
| 138 | |
| 139 | static const int32_t ALL_SIMPLE_NAME_TYPES = UTZNM_LONG_STANDARD | UTZNM_LONG_DAYLIGHT | UTZNM_SHORT_STANDARD | UTZNM_SHORT_DAYLIGHT | UTZNM_EXEMPLAR_LOCATION; |
| 140 | static const int32_t ALL_GENERIC_NAME_TYPES = UTZGNM_LOCATION | UTZGNM_LONG | UTZGNM_SHORT; |
| 141 | |
| 142 | #define DIGIT_VAL(c)(0x0030 <= (c) && (c) <= 0x0039 ? (c) - 0x0030 : -1) (0x0030 <= (c) && (c) <= 0x0039 ? (c) - 0x0030 : -1) |
| 143 | #define MAX_OFFSET_DIGITS6 6 |
| 144 | |
| 145 | // Time Zone ID/Short ID trie |
| 146 | static TextTrieMap *gZoneIdTrie = nullptr; |
| 147 | static icu::UInitOnce gZoneIdTrieInitOnce {}; |
| 148 | |
| 149 | static TextTrieMap *gShortZoneIdTrie = nullptr; |
| 150 | static icu::UInitOnce gShortZoneIdTrieInitOnce {}; |
| 151 | |
| 152 | static UMutex gLock; |
| 153 | |
| 154 | U_CDECL_BEGINextern "C" { |
| 155 | /** |
| 156 | * Cleanup callback func |
| 157 | */ |
| 158 | static UBool U_CALLCONV tzfmt_cleanup() |
| 159 | { |
| 160 | delete gZoneIdTrie; |
| 161 | gZoneIdTrie = nullptr; |
| 162 | gZoneIdTrieInitOnce.reset(); |
| 163 | |
| 164 | delete gShortZoneIdTrie; |
| 165 | gShortZoneIdTrie = nullptr; |
| 166 | gShortZoneIdTrieInitOnce.reset(); |
| 167 | |
| 168 | return true; |
| 169 | } |
| 170 | U_CDECL_END} |
| 171 | |
| 172 | // ------------------------------------------------------------------ |
| 173 | // GMTOffsetField |
| 174 | // |
| 175 | // This class represents a localized GMT offset pattern |
| 176 | // item and used by TimeZoneFormat |
| 177 | // ------------------------------------------------------------------ |
| 178 | class GMTOffsetField : public UMemory { |
| 179 | public: |
| 180 | enum FieldType { |
| 181 | TEXT = 0, |
| 182 | HOUR = 1, |
| 183 | MINUTE = 2, |
| 184 | SECOND = 4 |
| 185 | }; |
| 186 | |
| 187 | virtual ~GMTOffsetField(); |
| 188 | |
| 189 | static GMTOffsetField* createText(const UnicodeString& text, UErrorCode& status); |
| 190 | static GMTOffsetField* createTimeField(FieldType type, uint8_t width, UErrorCode& status); |
| 191 | static UBool isValid(FieldType type, int32_t width); |
| 192 | static FieldType getTypeByLetter(char16_t ch); |
| 193 | |
| 194 | FieldType getType() const; |
| 195 | uint8_t getWidth() const; |
| 196 | const char16_t* getPatternText() const; |
| 197 | |
| 198 | private: |
| 199 | char16_t* fText; |
| 200 | FieldType fType; |
| 201 | uint8_t fWidth; |
| 202 | |
| 203 | GMTOffsetField(); |
| 204 | }; |
| 205 | |
| 206 | GMTOffsetField::GMTOffsetField() |
| 207 | : fText(nullptr), fType(TEXT), fWidth(0) { |
| 208 | } |
| 209 | |
| 210 | GMTOffsetField::~GMTOffsetField() { |
| 211 | if (fText) { |
| 212 | uprv_free(fText); |
| 213 | } |
| 214 | } |
| 215 | |
| 216 | GMTOffsetField* |
| 217 | GMTOffsetField::createText(const UnicodeString& text, UErrorCode& status) { |
| 218 | if (U_FAILURE(status)) { |
| 219 | return nullptr; |
| 220 | } |
| 221 | GMTOffsetField* result = new GMTOffsetField(); |
| 222 | if (result == nullptr) { |
| 223 | status = U_MEMORY_ALLOCATION_ERROR; |
| 224 | return nullptr; |
| 225 | } |
| 226 | |
| 227 | int32_t len = text.length(); |
| 228 | result->fText = static_cast<char16_t*>(uprv_malloc((len + 1) * sizeof(char16_t))); |
| 229 | if (result->fText == nullptr) { |
| 230 | status = U_MEMORY_ALLOCATION_ERROR; |
| 231 | delete result; |
| 232 | return nullptr; |
| 233 | } |
| 234 | u_strncpy(result->fText, text.getBuffer(), len); |
| 235 | result->fText[len] = 0; |
| 236 | result->fType = TEXT; |
| 237 | |
| 238 | return result; |
| 239 | } |
| 240 | |
| 241 | GMTOffsetField* |
| 242 | GMTOffsetField::createTimeField(FieldType type, uint8_t width, UErrorCode& status) { |
| 243 | U_ASSERT(type != TEXT)(static_cast <bool> (type != TEXT) ? void (0) : __assert_fail ("type != TEXT", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 244 | if (U_FAILURE(status)) { |
| 245 | return nullptr; |
| 246 | } |
| 247 | GMTOffsetField* result = new GMTOffsetField(); |
| 248 | if (result == nullptr) { |
| 249 | status = U_MEMORY_ALLOCATION_ERROR; |
| 250 | return nullptr; |
| 251 | } |
| 252 | |
| 253 | result->fType = type; |
| 254 | result->fWidth = width; |
| 255 | |
| 256 | return result; |
| 257 | } |
| 258 | |
| 259 | UBool |
| 260 | GMTOffsetField::isValid(FieldType type, int32_t width) { |
| 261 | switch (type) { |
| 262 | case HOUR: |
| 263 | return (width == 1 || width == 2); |
| 264 | case MINUTE: |
| 265 | case SECOND: |
| 266 | return (width == 2); |
| 267 | default: |
| 268 | UPRV_UNREACHABLE_EXITabort(); |
| 269 | } |
| 270 | return (width > 0); |
| 271 | } |
| 272 | |
| 273 | GMTOffsetField::FieldType |
| 274 | GMTOffsetField::getTypeByLetter(char16_t ch) { |
| 275 | if (ch == 0x0048 /* H */) { |
| 276 | return HOUR; |
| 277 | } else if (ch == 0x006D /* m */) { |
| 278 | return MINUTE; |
| 279 | } else if (ch == 0x0073 /* s */) { |
| 280 | return SECOND; |
| 281 | } |
| 282 | return TEXT; |
| 283 | } |
| 284 | |
| 285 | inline GMTOffsetField::FieldType |
| 286 | GMTOffsetField::getType() const { |
| 287 | return fType; |
| 288 | } |
| 289 | |
| 290 | inline uint8_t |
| 291 | GMTOffsetField::getWidth() const { |
| 292 | return fWidth; |
| 293 | } |
| 294 | |
| 295 | inline const char16_t* |
| 296 | GMTOffsetField::getPatternText() const { |
| 297 | return fText; |
| 298 | } |
| 299 | |
| 300 | |
| 301 | U_CDECL_BEGINextern "C" { |
| 302 | static void U_CALLCONV |
| 303 | deleteGMTOffsetField(void *obj) { |
| 304 | delete static_cast<GMTOffsetField *>(obj); |
| 305 | } |
| 306 | U_CDECL_END} |
| 307 | |
| 308 | |
| 309 | // ------------------------------------------------------------------ |
| 310 | // TimeZoneFormat |
| 311 | // ------------------------------------------------------------------ |
| 312 | UOBJECT_DEFINE_RTTI_IMPLEMENTATION(TimeZoneFormat)UClassID TimeZoneFormat::getStaticClassID() { static char classID = 0; return (UClassID)&classID; } UClassID TimeZoneFormat ::getDynamicClassID() const { return TimeZoneFormat::getStaticClassID (); } |
| 313 | |
| 314 | TimeZoneFormat::TimeZoneFormat(const Locale& locale, UErrorCode& status) |
| 315 | : fLocale(locale), fTimeZoneNames(nullptr), fTimeZoneGenericNames(nullptr), |
| 316 | fDefParseOptionFlags(0), fTZDBTimeZoneNames(nullptr) { |
| 317 | |
| 318 | for (int32_t i = 0; i < UTZFMT_PAT_COUNT; i++) { |
| 319 | fGMTOffsetPatternItems[i] = nullptr; |
| 320 | } |
| 321 | |
| 322 | const char* region = fLocale.getCountry(); |
| 323 | int32_t regionLen = static_cast<int32_t>(uprv_strlen(region):: strlen(region)); |
| 324 | if (regionLen == 0) { |
| 325 | UErrorCode tempStatus = U_ZERO_ERROR; |
| 326 | CharString loc = ulocimp_addLikelySubtags(fLocale.getName(), tempStatus); |
| 327 | |
| 328 | regionLen = uloc_getCountry(loc.data(), fTargetRegion, sizeof(fTargetRegion), &tempStatus); |
| 329 | if (U_SUCCESS(tempStatus)) { |
| 330 | fTargetRegion[regionLen] = 0; |
| 331 | } else { |
| 332 | fTargetRegion[0] = 0; |
| 333 | } |
| 334 | } else if (regionLen < static_cast<int32_t>(sizeof(fTargetRegion))) { |
| 335 | uprv_strcpy(fTargetRegion, region):: strcpy(fTargetRegion, region); |
| 336 | } else { |
| 337 | fTargetRegion[0] = 0; |
| 338 | } |
| 339 | |
| 340 | fTimeZoneNames = TimeZoneNames::createInstance(locale, status); |
| 341 | // fTimeZoneGenericNames is lazily instantiated |
| 342 | if (U_FAILURE(status)) { |
| 343 | return; |
| 344 | } |
| 345 | |
| 346 | const char16_t* gmtPattern = nullptr; |
| 347 | const char16_t* hourFormats = nullptr; |
| 348 | |
| 349 | UResourceBundle *zoneBundle = ures_open(U_ICUDATA_ZONE"icudt" "78" "l" "-" "zone", locale.getName(), &status); |
| 350 | UResourceBundle *zoneStringsArray = ures_getByKeyWithFallback(zoneBundle, gZoneStringsTag, nullptr, &status); |
| 351 | if (U_SUCCESS(status)) { |
| 352 | const char16_t* resStr; |
| 353 | int32_t len; |
| 354 | resStr = ures_getStringByKeyWithFallback(zoneStringsArray, gGmtFormatTag, &len, &status); |
| 355 | if (len > 0) { |
| 356 | gmtPattern = resStr; |
| 357 | } |
| 358 | resStr = ures_getStringByKeyWithFallback(zoneStringsArray, gGmtZeroFormatTag, &len, &status); |
| 359 | if (len > 0) { |
| 360 | fGMTZeroFormat.setTo(true, resStr, len); |
| 361 | } |
| 362 | resStr = ures_getStringByKeyWithFallback(zoneStringsArray, gHourFormatTag, &len, &status); |
| 363 | if (len > 0) { |
| 364 | hourFormats = resStr; |
| 365 | } |
| 366 | ures_close(zoneStringsArray); |
| 367 | ures_close(zoneBundle); |
| 368 | } |
| 369 | |
| 370 | if (gmtPattern == nullptr) { |
| 371 | gmtPattern = DEFAULT_GMT_PATTERN; |
| 372 | } |
| 373 | initGMTPattern(UnicodeString(true, gmtPattern, -1), status); |
| 374 | |
| 375 | UBool useDefaultOffsetPatterns = true; |
| 376 | if (hourFormats) { |
| 377 | char16_t* sep = u_strchr(hourFormats, static_cast<char16_t>(0x003B) /* ';' */); |
| 378 | if (sep != nullptr) { |
| 379 | UErrorCode tmpStatus = U_ZERO_ERROR; |
| 380 | fGMTOffsetPatterns[UTZFMT_PAT_POSITIVE_HM].setTo(false, hourFormats, static_cast<int32_t>(sep - hourFormats)); |
| 381 | fGMTOffsetPatterns[UTZFMT_PAT_NEGATIVE_HM].setTo(true, sep + 1, -1); |
| 382 | expandOffsetPattern(fGMTOffsetPatterns[UTZFMT_PAT_POSITIVE_HM], fGMTOffsetPatterns[UTZFMT_PAT_POSITIVE_HMS], tmpStatus); |
| 383 | expandOffsetPattern(fGMTOffsetPatterns[UTZFMT_PAT_NEGATIVE_HM], fGMTOffsetPatterns[UTZFMT_PAT_NEGATIVE_HMS], tmpStatus); |
| 384 | truncateOffsetPattern(fGMTOffsetPatterns[UTZFMT_PAT_POSITIVE_HM], fGMTOffsetPatterns[UTZFMT_PAT_POSITIVE_H], tmpStatus); |
| 385 | truncateOffsetPattern(fGMTOffsetPatterns[UTZFMT_PAT_NEGATIVE_HM], fGMTOffsetPatterns[UTZFMT_PAT_NEGATIVE_H], tmpStatus); |
| 386 | if (U_SUCCESS(tmpStatus)) { |
| 387 | useDefaultOffsetPatterns = false; |
| 388 | } |
| 389 | } |
| 390 | } |
| 391 | if (useDefaultOffsetPatterns) { |
| 392 | fGMTOffsetPatterns[UTZFMT_PAT_POSITIVE_H].setTo(true, DEFAULT_GMT_POSITIVE_H, -1); |
| 393 | fGMTOffsetPatterns[UTZFMT_PAT_POSITIVE_HM].setTo(true, DEFAULT_GMT_POSITIVE_HM, -1); |
| 394 | fGMTOffsetPatterns[UTZFMT_PAT_POSITIVE_HMS].setTo(true, DEFAULT_GMT_POSITIVE_HMS, -1); |
| 395 | fGMTOffsetPatterns[UTZFMT_PAT_NEGATIVE_H].setTo(true, DEFAULT_GMT_NEGATIVE_H, -1); |
| 396 | fGMTOffsetPatterns[UTZFMT_PAT_NEGATIVE_HM].setTo(true, DEFAULT_GMT_NEGATIVE_HM, -1); |
| 397 | fGMTOffsetPatterns[UTZFMT_PAT_NEGATIVE_HMS].setTo(true, DEFAULT_GMT_NEGATIVE_HMS, -1); |
| 398 | } |
| 399 | initGMTOffsetPatterns(status); |
| 400 | |
| 401 | NumberingSystem* ns = NumberingSystem::createInstance(locale, status); |
| 402 | UBool useDefDigits = true; |
| 403 | if (ns && !ns->isAlgorithmic()) { |
| 404 | UnicodeString digits = ns->getDescription(); |
| 405 | useDefDigits = !toCodePoints(digits, fGMTOffsetDigits, 10); |
| 406 | } |
| 407 | if (useDefDigits) { |
| 408 | uprv_memcpy(fGMTOffsetDigits, DEFAULT_GMT_DIGITS, sizeof(UChar32) * 10)do { clang diagnostic push
clang diagnostic ignored "-Waddress" (static_cast <bool> (fGMTOffsetDigits != __null) ? void (0) : __assert_fail ("fGMTOffsetDigits != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); ( static_cast <bool> (DEFAULT_GMT_DIGITS != __null) ? void (0) : __assert_fail ("DEFAULT_GMT_DIGITS != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); clang diagnostic pop :: memcpy(fGMTOffsetDigits, DEFAULT_GMT_DIGITS , sizeof(UChar32) * 10); } while (false); |
| 409 | } |
| 410 | delete ns; |
| 411 | } |
| 412 | |
| 413 | TimeZoneFormat::TimeZoneFormat(const TimeZoneFormat& other) |
| 414 | : Format(other), fTimeZoneNames(nullptr), fTimeZoneGenericNames(nullptr), |
| 415 | fTZDBTimeZoneNames(nullptr) { |
| 416 | |
| 417 | for (int32_t i = 0; i < UTZFMT_PAT_COUNT; i++) { |
| 418 | fGMTOffsetPatternItems[i] = nullptr; |
| 419 | } |
| 420 | *this = other; |
| 421 | } |
| 422 | |
| 423 | |
| 424 | TimeZoneFormat::~TimeZoneFormat() { |
| 425 | delete fTimeZoneNames; |
| 426 | delete fTimeZoneGenericNames; |
| 427 | delete fTZDBTimeZoneNames; |
| 428 | for (int32_t i = 0; i < UTZFMT_PAT_COUNT; i++) { |
| 429 | delete fGMTOffsetPatternItems[i]; |
| 430 | } |
| 431 | } |
| 432 | |
| 433 | TimeZoneFormat& |
| 434 | TimeZoneFormat::operator=(const TimeZoneFormat& other) { |
| 435 | if (this == &other) { |
| 436 | return *this; |
| 437 | } |
| 438 | |
| 439 | delete fTimeZoneNames; |
| 440 | delete fTimeZoneGenericNames; |
| 441 | fTimeZoneGenericNames = nullptr; |
| 442 | delete fTZDBTimeZoneNames; |
| 443 | fTZDBTimeZoneNames = nullptr; |
| 444 | |
| 445 | fLocale = other.fLocale; |
| 446 | uprv_memcpy(fTargetRegion, other.fTargetRegion, sizeof(fTargetRegion))do { clang diagnostic push
clang diagnostic ignored "-Waddress" (static_cast <bool> (fTargetRegion != __null) ? void ( 0) : __assert_fail ("fTargetRegion != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); ( static_cast <bool> (other.fTargetRegion != __null) ? void (0) : __assert_fail ("other.fTargetRegion != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); clang diagnostic pop :: memcpy(fTargetRegion, other.fTargetRegion , sizeof(fTargetRegion)); } while (false); |
| 447 | |
| 448 | fTimeZoneNames = other.fTimeZoneNames->clone(); |
| 449 | if (other.fTimeZoneGenericNames) { |
| 450 | // TODO: this test has dubious thread safety. |
| 451 | fTimeZoneGenericNames = other.fTimeZoneGenericNames->clone(); |
| 452 | } |
| 453 | |
| 454 | fGMTPattern = other.fGMTPattern; |
| 455 | fGMTPatternPrefix = other.fGMTPatternPrefix; |
| 456 | fGMTPatternSuffix = other.fGMTPatternSuffix; |
| 457 | |
| 458 | UErrorCode status = U_ZERO_ERROR; |
| 459 | for (int32_t i = 0; i < UTZFMT_PAT_COUNT; i++) { |
| 460 | fGMTOffsetPatterns[i] = other.fGMTOffsetPatterns[i]; |
| 461 | delete fGMTOffsetPatternItems[i]; |
| 462 | fGMTOffsetPatternItems[i] = nullptr; |
| 463 | } |
| 464 | initGMTOffsetPatterns(status); |
| 465 | U_ASSERT(U_SUCCESS(status))(static_cast <bool> (U_SUCCESS(status)) ? void (0) : __assert_fail ("U_SUCCESS(status)", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 466 | |
| 467 | fGMTZeroFormat = other.fGMTZeroFormat; |
| 468 | |
| 469 | uprv_memcpy(fGMTOffsetDigits, other.fGMTOffsetDigits, sizeof(fGMTOffsetDigits))do { clang diagnostic push
clang diagnostic ignored "-Waddress" (static_cast <bool> (fGMTOffsetDigits != __null) ? void (0) : __assert_fail ("fGMTOffsetDigits != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); ( static_cast <bool> (other.fGMTOffsetDigits != __null) ? void (0) : __assert_fail ("other.fGMTOffsetDigits != __null" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); clang diagnostic pop :: memcpy(fGMTOffsetDigits, other.fGMTOffsetDigits , sizeof(fGMTOffsetDigits)); } while (false); |
| 470 | |
| 471 | fDefParseOptionFlags = other.fDefParseOptionFlags; |
| 472 | |
| 473 | return *this; |
| 474 | } |
| 475 | |
| 476 | |
| 477 | bool |
| 478 | TimeZoneFormat::operator==(const Format& other) const { |
| 479 | TimeZoneFormat* tzfmt = (TimeZoneFormat*)&other; |
| 480 | |
| 481 | bool isEqual = |
| 482 | fLocale == tzfmt->fLocale |
| 483 | && fGMTPattern == tzfmt->fGMTPattern |
| 484 | && fGMTZeroFormat == tzfmt->fGMTZeroFormat |
| 485 | && *fTimeZoneNames == *tzfmt->fTimeZoneNames; |
| 486 | |
| 487 | for (int32_t i = 0; i < UTZFMT_PAT_COUNT && isEqual; i++) { |
| 488 | isEqual = fGMTOffsetPatterns[i] == tzfmt->fGMTOffsetPatterns[i]; |
| 489 | } |
| 490 | for (int32_t i = 0; i < 10 && isEqual; i++) { |
| 491 | isEqual = fGMTOffsetDigits[i] == tzfmt->fGMTOffsetDigits[i]; |
| 492 | } |
| 493 | // TODO |
| 494 | // Check fTimeZoneGenericNames. For now, |
| 495 | // if fTimeZoneNames is same, fTimeZoneGenericNames should |
| 496 | // be also equivalent. |
| 497 | return isEqual; |
| 498 | } |
| 499 | |
| 500 | TimeZoneFormat* |
| 501 | TimeZoneFormat::clone() const { |
| 502 | return new TimeZoneFormat(*this); |
| 503 | } |
| 504 | |
| 505 | TimeZoneFormat* U_EXPORT2 |
| 506 | TimeZoneFormat::createInstance(const Locale& locale, UErrorCode& status) { |
| 507 | TimeZoneFormat* tzfmt = new TimeZoneFormat(locale, status); |
| 508 | if (U_SUCCESS(status)) { |
| 509 | return tzfmt; |
| 510 | } |
| 511 | delete tzfmt; |
| 512 | return nullptr; |
| 513 | } |
| 514 | |
| 515 | // ------------------------------------------------------------------ |
| 516 | // Setter and Getter |
| 517 | |
| 518 | const TimeZoneNames* |
| 519 | TimeZoneFormat::getTimeZoneNames() const { |
| 520 | return (const TimeZoneNames*)fTimeZoneNames; |
| 521 | } |
| 522 | |
| 523 | void |
| 524 | TimeZoneFormat::adoptTimeZoneNames(TimeZoneNames *tznames) { |
| 525 | delete fTimeZoneNames; |
| 526 | fTimeZoneNames = tznames; |
| 527 | |
| 528 | // TODO - We should also update fTimeZoneGenericNames |
| 529 | } |
| 530 | |
| 531 | void |
| 532 | TimeZoneFormat::setTimeZoneNames(const TimeZoneNames &tznames) { |
| 533 | delete fTimeZoneNames; |
| 534 | fTimeZoneNames = tznames.clone(); |
| 535 | |
| 536 | // TODO - We should also update fTimeZoneGenericNames |
| 537 | } |
| 538 | |
| 539 | void |
| 540 | TimeZoneFormat::setDefaultParseOptions(uint32_t flags) { |
| 541 | fDefParseOptionFlags = flags; |
| 542 | } |
| 543 | |
| 544 | uint32_t |
| 545 | TimeZoneFormat::getDefaultParseOptions() const { |
| 546 | return fDefParseOptionFlags; |
| 547 | } |
| 548 | |
| 549 | |
| 550 | UnicodeString& |
| 551 | TimeZoneFormat::getGMTPattern(UnicodeString& pattern) const { |
| 552 | return pattern.setTo(fGMTPattern); |
| 553 | } |
| 554 | |
| 555 | void |
| 556 | TimeZoneFormat::setGMTPattern(const UnicodeString& pattern, UErrorCode& status) { |
| 557 | initGMTPattern(pattern, status); |
| 558 | } |
| 559 | |
| 560 | UnicodeString& |
| 561 | TimeZoneFormat::getGMTOffsetPattern(UTimeZoneFormatGMTOffsetPatternType type, UnicodeString& pattern) const { |
| 562 | return pattern.setTo(fGMTOffsetPatterns[type]); |
| 563 | } |
| 564 | |
| 565 | void |
| 566 | TimeZoneFormat::setGMTOffsetPattern(UTimeZoneFormatGMTOffsetPatternType type, const UnicodeString& pattern, UErrorCode& status) { |
| 567 | if (U_FAILURE(status)) { |
| 568 | return; |
| 569 | } |
| 570 | if (pattern == fGMTOffsetPatterns[type]) { |
| 571 | // No need to reset |
| 572 | return; |
| 573 | } |
| 574 | |
| 575 | OffsetFields required = FIELDS_HM; |
| 576 | switch (type) { |
| 577 | case UTZFMT_PAT_POSITIVE_H: |
| 578 | case UTZFMT_PAT_NEGATIVE_H: |
| 579 | required = FIELDS_H; |
| 580 | break; |
| 581 | case UTZFMT_PAT_POSITIVE_HM: |
| 582 | case UTZFMT_PAT_NEGATIVE_HM: |
| 583 | required = FIELDS_HM; |
| 584 | break; |
| 585 | case UTZFMT_PAT_POSITIVE_HMS: |
| 586 | case UTZFMT_PAT_NEGATIVE_HMS: |
| 587 | required = FIELDS_HMS; |
| 588 | break; |
| 589 | default: |
| 590 | UPRV_UNREACHABLE_EXITabort(); |
| 591 | } |
| 592 | |
| 593 | UVector* patternItems = parseOffsetPattern(pattern, required, status); |
| 594 | if (patternItems == nullptr) { |
| 595 | return; |
| 596 | } |
| 597 | |
| 598 | fGMTOffsetPatterns[type].setTo(pattern); |
| 599 | delete fGMTOffsetPatternItems[type]; |
| 600 | fGMTOffsetPatternItems[type] = patternItems; |
| 601 | checkAbuttingHoursAndMinutes(); |
| 602 | } |
| 603 | |
| 604 | UnicodeString& |
| 605 | TimeZoneFormat::getGMTOffsetDigits(UnicodeString& digits) const { |
| 606 | digits.remove(); |
| 607 | for (int32_t i = 0; i < 10; i++) { |
| 608 | digits.append(fGMTOffsetDigits[i]); |
| 609 | } |
| 610 | return digits; |
| 611 | } |
| 612 | |
| 613 | void |
| 614 | TimeZoneFormat::setGMTOffsetDigits(const UnicodeString& digits, UErrorCode& status) { |
| 615 | if (U_FAILURE(status)) { |
| 616 | return; |
| 617 | } |
| 618 | UChar32 digitArray[10]; |
| 619 | if (!toCodePoints(digits, digitArray, 10)) { |
| 620 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 621 | return; |
| 622 | } |
| 623 | uprv_memcpy(fGMTOffsetDigits, digitArray, sizeof(UChar32)*10)do { clang diagnostic push
clang diagnostic ignored "-Waddress" (static_cast <bool> (fGMTOffsetDigits != __null) ? void (0) : __assert_fail ("fGMTOffsetDigits != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); ( static_cast <bool> (digitArray != __null) ? void (0) : __assert_fail ("digitArray != __null", __builtin_FILE (), __builtin_LINE ( ), __extension__ __PRETTY_FUNCTION__)); clang diagnostic pop :: memcpy(fGMTOffsetDigits, digitArray, sizeof(UChar32)*10); } while (false); |
| 624 | } |
| 625 | |
| 626 | UnicodeString& |
| 627 | TimeZoneFormat::getGMTZeroFormat(UnicodeString& gmtZeroFormat) const { |
| 628 | return gmtZeroFormat.setTo(fGMTZeroFormat); |
| 629 | } |
| 630 | |
| 631 | void |
| 632 | TimeZoneFormat::setGMTZeroFormat(const UnicodeString& gmtZeroFormat, UErrorCode& status) { |
| 633 | if (U_SUCCESS(status)) { |
| 634 | if (gmtZeroFormat.isEmpty()) { |
| 635 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 636 | } else if (gmtZeroFormat != fGMTZeroFormat) { |
| 637 | fGMTZeroFormat.setTo(gmtZeroFormat); |
| 638 | } |
| 639 | } |
| 640 | } |
| 641 | |
| 642 | // ------------------------------------------------------------------ |
| 643 | // Format and Parse |
| 644 | |
| 645 | UnicodeString& |
| 646 | TimeZoneFormat::format(UTimeZoneFormatStyle style, const TimeZone& tz, UDate date, |
| 647 | UnicodeString& name, UTimeZoneFormatTimeType* timeType /* = nullptr */) const { |
| 648 | if (timeType) { |
| 649 | *timeType = UTZFMT_TIME_TYPE_UNKNOWN; |
| 650 | } |
| 651 | |
| 652 | UBool noOffsetFormatFallback = false; |
| 653 | |
| 654 | switch (style) { |
| 655 | case UTZFMT_STYLE_GENERIC_LOCATION: |
| 656 | formatGeneric(tz, UTZGNM_LOCATION, date, name); |
| 657 | break; |
| 658 | case UTZFMT_STYLE_GENERIC_LONG: |
| 659 | formatGeneric(tz, UTZGNM_LONG, date, name); |
| 660 | break; |
| 661 | case UTZFMT_STYLE_GENERIC_SHORT: |
| 662 | formatGeneric(tz, UTZGNM_SHORT, date, name); |
| 663 | break; |
| 664 | case UTZFMT_STYLE_SPECIFIC_LONG: |
| 665 | formatSpecific(tz, UTZNM_LONG_STANDARD, UTZNM_LONG_DAYLIGHT, date, name, timeType); |
| 666 | break; |
| 667 | case UTZFMT_STYLE_SPECIFIC_SHORT: |
| 668 | formatSpecific(tz, UTZNM_SHORT_STANDARD, UTZNM_SHORT_DAYLIGHT, date, name, timeType); |
| 669 | break; |
| 670 | |
| 671 | case UTZFMT_STYLE_ZONE_ID: |
| 672 | tz.getID(name); |
| 673 | noOffsetFormatFallback = true; |
| 674 | break; |
| 675 | case UTZFMT_STYLE_ZONE_ID_SHORT: |
| 676 | { |
| 677 | const char16_t* shortID = ZoneMeta::getShortID(tz); |
| 678 | if (shortID == nullptr) { |
| 679 | shortID = UNKNOWN_SHORT_ZONE_ID; |
| 680 | } |
| 681 | name.setTo(shortID, -1); |
| 682 | } |
| 683 | noOffsetFormatFallback = true; |
| 684 | break; |
| 685 | |
| 686 | case UTZFMT_STYLE_EXEMPLAR_LOCATION: |
| 687 | formatExemplarLocation(tz, name); |
| 688 | noOffsetFormatFallback = true; |
| 689 | break; |
| 690 | |
| 691 | default: |
| 692 | // will be handled below |
| 693 | break; |
| 694 | } |
| 695 | |
| 696 | if (name.isEmpty() && !noOffsetFormatFallback) { |
| 697 | UErrorCode status = U_ZERO_ERROR; |
| 698 | int32_t rawOffset, dstOffset; |
| 699 | tz.getOffset(date, false, rawOffset, dstOffset, status); |
| 700 | int32_t offset = rawOffset + dstOffset; |
| 701 | if (U_SUCCESS(status)) { |
| 702 | switch (style) { |
| 703 | case UTZFMT_STYLE_GENERIC_LOCATION: |
| 704 | case UTZFMT_STYLE_GENERIC_LONG: |
| 705 | case UTZFMT_STYLE_SPECIFIC_LONG: |
| 706 | case UTZFMT_STYLE_LOCALIZED_GMT: |
| 707 | formatOffsetLocalizedGMT(offset, name, status); |
| 708 | break; |
| 709 | |
| 710 | case UTZFMT_STYLE_GENERIC_SHORT: |
| 711 | case UTZFMT_STYLE_SPECIFIC_SHORT: |
| 712 | case UTZFMT_STYLE_LOCALIZED_GMT_SHORT: |
| 713 | formatOffsetShortLocalizedGMT(offset, name, status); |
| 714 | break; |
| 715 | |
| 716 | case UTZFMT_STYLE_ISO_BASIC_SHORT: |
| 717 | formatOffsetISO8601Basic(offset, true, true, true, name, status); |
| 718 | break; |
| 719 | |
| 720 | case UTZFMT_STYLE_ISO_BASIC_LOCAL_SHORT: |
| 721 | formatOffsetISO8601Basic(offset, false, true, true, name, status); |
| 722 | break; |
| 723 | |
| 724 | case UTZFMT_STYLE_ISO_BASIC_FIXED: |
| 725 | formatOffsetISO8601Basic(offset, true, false, true, name, status); |
| 726 | break; |
| 727 | |
| 728 | case UTZFMT_STYLE_ISO_BASIC_LOCAL_FIXED: |
| 729 | formatOffsetISO8601Basic(offset, false, false, true, name, status); |
| 730 | break; |
| 731 | |
| 732 | case UTZFMT_STYLE_ISO_EXTENDED_FIXED: |
| 733 | formatOffsetISO8601Extended(offset, true, false, true, name, status); |
| 734 | break; |
| 735 | |
| 736 | case UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FIXED: |
| 737 | formatOffsetISO8601Extended(offset, false, false, true, name, status); |
| 738 | break; |
| 739 | |
| 740 | case UTZFMT_STYLE_ISO_BASIC_FULL: |
| 741 | formatOffsetISO8601Basic(offset, true, false, false, name, status); |
| 742 | break; |
| 743 | |
| 744 | case UTZFMT_STYLE_ISO_BASIC_LOCAL_FULL: |
| 745 | formatOffsetISO8601Basic(offset, false, false, false, name, status); |
| 746 | break; |
| 747 | |
| 748 | case UTZFMT_STYLE_ISO_EXTENDED_FULL: |
| 749 | formatOffsetISO8601Extended(offset, true, false, false, name, status); |
| 750 | break; |
| 751 | |
| 752 | case UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FULL: |
| 753 | formatOffsetISO8601Extended(offset, false, false, false, name, status); |
| 754 | break; |
| 755 | |
| 756 | default: |
| 757 | // UTZFMT_STYLE_ZONE_ID, UTZFMT_STYLE_ZONE_ID_SHORT, UTZFMT_STYLE_EXEMPLAR_LOCATION |
| 758 | break; |
| 759 | } |
| 760 | |
| 761 | if (timeType) { |
| 762 | *timeType = (dstOffset != 0) ? UTZFMT_TIME_TYPE_DAYLIGHT : UTZFMT_TIME_TYPE_STANDARD; |
| 763 | } |
| 764 | } |
| 765 | } |
| 766 | |
| 767 | return name; |
| 768 | } |
| 769 | |
| 770 | UnicodeString& |
| 771 | TimeZoneFormat::format(const Formattable& obj, UnicodeString& appendTo, |
| 772 | FieldPosition& pos, UErrorCode& status) const { |
| 773 | if (U_FAILURE(status)) { |
| 774 | return appendTo; |
| 775 | } |
| 776 | UDate date = Calendar::getNow(); |
| 777 | if (obj.getType() == Formattable::kObject) { |
| 778 | const UObject* formatObj = obj.getObject(); |
| 779 | const TimeZone* tz = dynamic_cast<const TimeZone*>(formatObj); |
| 780 | if (tz == nullptr) { |
| 781 | const Calendar* cal = dynamic_cast<const Calendar*>(formatObj); |
| 782 | if (cal != nullptr) { |
| 783 | tz = &cal->getTimeZone(); |
| 784 | date = cal->getTime(status); |
| 785 | } |
| 786 | } |
| 787 | if (tz != nullptr) { |
| 788 | int32_t rawOffset, dstOffset; |
| 789 | tz->getOffset(date, false, rawOffset, dstOffset, status); |
| 790 | char16_t buf[ZONE_NAME_U16_MAX128]; |
| 791 | UnicodeString result(buf, 0, UPRV_LENGTHOF(buf)(int32_t)(sizeof(buf)/sizeof((buf)[0]))); |
| 792 | formatOffsetLocalizedGMT(rawOffset + dstOffset, result, status); |
| 793 | if (U_SUCCESS(status)) { |
| 794 | appendTo.append(result); |
| 795 | if (pos.getField() == UDAT_TIMEZONE_FIELD) { |
| 796 | pos.setBeginIndex(0); |
| 797 | pos.setEndIndex(result.length()); |
| 798 | } |
| 799 | } |
| 800 | } |
| 801 | } |
| 802 | return appendTo; |
| 803 | } |
| 804 | |
| 805 | TimeZone* |
| 806 | TimeZoneFormat::parse(UTimeZoneFormatStyle style, const UnicodeString& text, ParsePosition& pos, |
| 807 | UTimeZoneFormatTimeType* timeType /*= nullptr*/) const { |
| 808 | return parse(style, text, pos, getDefaultParseOptions(), timeType); |
| 809 | } |
| 810 | |
| 811 | TimeZone* |
| 812 | TimeZoneFormat::parse(UTimeZoneFormatStyle style, const UnicodeString& text, ParsePosition& pos, |
| 813 | int32_t parseOptions, UTimeZoneFormatTimeType* timeType /* = nullptr */) const { |
| 814 | if (timeType) { |
| 815 | *timeType = UTZFMT_TIME_TYPE_UNKNOWN; |
| 816 | } |
| 817 | |
| 818 | int32_t startIdx = pos.getIndex(); |
| 819 | int32_t maxPos = text.length(); |
| 820 | int32_t offset; |
| 821 | |
| 822 | // Styles using localized GMT format as fallback |
| 823 | UBool fallbackLocalizedGMT = |
| 824 | (style == UTZFMT_STYLE_SPECIFIC_LONG || style == UTZFMT_STYLE_GENERIC_LONG || style == UTZFMT_STYLE_GENERIC_LOCATION); |
| 825 | UBool fallbackShortLocalizedGMT = |
| 826 | (style == UTZFMT_STYLE_SPECIFIC_SHORT || style == UTZFMT_STYLE_GENERIC_SHORT); |
| 827 | |
| 828 | int32_t evaluated = 0; // bit flags representing already evaluated styles |
| 829 | ParsePosition tmpPos(startIdx); |
| 830 | |
| 831 | int32_t parsedOffset = UNKNOWN_OFFSET; // stores successfully parsed offset for later use |
| 832 | int32_t parsedPos = -1; // stores successfully parsed offset position for later use |
| 833 | |
| 834 | // Try localized GMT format first if necessary |
| 835 | if (fallbackLocalizedGMT || fallbackShortLocalizedGMT) { |
| 836 | UBool hasDigitOffset = false; |
| 837 | offset = parseOffsetLocalizedGMT(text, tmpPos, fallbackShortLocalizedGMT, &hasDigitOffset); |
| 838 | if (tmpPos.getErrorIndex() == -1) { |
| 839 | // Even when the input text was successfully parsed as a localized GMT format text, |
| 840 | // we may still need to evaluate the specified style if - |
| 841 | // 1) GMT zero format was used, and |
| 842 | // 2) The input text was not completely processed |
| 843 | if (tmpPos.getIndex() == maxPos || hasDigitOffset) { |
| 844 | pos.setIndex(tmpPos.getIndex()); |
| 845 | return createTimeZoneForOffset(offset); |
| 846 | } |
| 847 | parsedOffset = offset; |
| 848 | parsedPos = tmpPos.getIndex(); |
| 849 | } |
| 850 | // Note: For now, no distinction between long/short localized GMT format in the parser. |
| 851 | // This might be changed in future. |
| 852 | // evaluated |= (fallbackLocalizedGMT ? STYLE_PARSE_FLAGS[UTZFMT_STYLE_LOCALIZED_GMT] : STYLE_PARSE_FLAGS[UTZFMT_STYLE_LOCALIZED_GMT_SHORT]); |
| 853 | evaluated |= STYLE_PARSE_FLAGS[UTZFMT_STYLE_LOCALIZED_GMT] | STYLE_PARSE_FLAGS[UTZFMT_STYLE_LOCALIZED_GMT_SHORT]; |
| 854 | } |
| 855 | |
| 856 | UErrorCode status = U_ZERO_ERROR; |
| 857 | char16_t tzIDBuf[32]; |
| 858 | UnicodeString tzID(tzIDBuf, 0, UPRV_LENGTHOF(tzIDBuf)(int32_t)(sizeof(tzIDBuf)/sizeof((tzIDBuf)[0]))); |
| 859 | |
| 860 | UBool parseTZDBAbbrev = ((parseOptions & UTZFMT_PARSE_OPTION_TZ_DATABASE_ABBREVIATIONS) != 0); |
| 861 | |
| 862 | // Try the specified style |
| 863 | switch (style) { |
| 864 | case UTZFMT_STYLE_LOCALIZED_GMT: |
| 865 | { |
| 866 | tmpPos.setIndex(startIdx); |
| 867 | tmpPos.setErrorIndex(-1); |
| 868 | |
| 869 | offset = parseOffsetLocalizedGMT(text, tmpPos); |
| 870 | if (tmpPos.getErrorIndex() == -1) { |
| 871 | pos.setIndex(tmpPos.getIndex()); |
| 872 | return createTimeZoneForOffset(offset); |
| 873 | } |
| 874 | |
| 875 | // Note: For now, no distinction between long/short localized GMT format in the parser. |
| 876 | // This might be changed in future. |
| 877 | evaluated |= STYLE_PARSE_FLAGS[UTZFMT_STYLE_LOCALIZED_GMT_SHORT]; |
| 878 | |
| 879 | break; |
| 880 | } |
| 881 | case UTZFMT_STYLE_LOCALIZED_GMT_SHORT: |
| 882 | { |
| 883 | tmpPos.setIndex(startIdx); |
| 884 | tmpPos.setErrorIndex(-1); |
| 885 | |
| 886 | offset = parseOffsetShortLocalizedGMT(text, tmpPos); |
| 887 | if (tmpPos.getErrorIndex() == -1) { |
| 888 | pos.setIndex(tmpPos.getIndex()); |
| 889 | return createTimeZoneForOffset(offset); |
| 890 | } |
| 891 | |
| 892 | // Note: For now, no distinction between long/short localized GMT format in the parser. |
| 893 | // This might be changed in future. |
| 894 | evaluated |= STYLE_PARSE_FLAGS[UTZFMT_STYLE_LOCALIZED_GMT]; |
| 895 | |
| 896 | break; |
| 897 | } |
| 898 | case UTZFMT_STYLE_ISO_BASIC_SHORT: |
| 899 | case UTZFMT_STYLE_ISO_BASIC_FIXED: |
| 900 | case UTZFMT_STYLE_ISO_BASIC_FULL: |
| 901 | case UTZFMT_STYLE_ISO_EXTENDED_FIXED: |
| 902 | case UTZFMT_STYLE_ISO_EXTENDED_FULL: |
| 903 | { |
| 904 | tmpPos.setIndex(startIdx); |
| 905 | tmpPos.setErrorIndex(-1); |
| 906 | |
| 907 | offset = parseOffsetISO8601(text, tmpPos); |
| 908 | if (tmpPos.getErrorIndex() == -1) { |
| 909 | pos.setIndex(tmpPos.getIndex()); |
| 910 | return createTimeZoneForOffset(offset); |
| 911 | } |
| 912 | |
| 913 | break; |
| 914 | } |
| 915 | |
| 916 | case UTZFMT_STYLE_ISO_BASIC_LOCAL_SHORT: |
| 917 | case UTZFMT_STYLE_ISO_BASIC_LOCAL_FIXED: |
| 918 | case UTZFMT_STYLE_ISO_BASIC_LOCAL_FULL: |
| 919 | case UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FIXED: |
| 920 | case UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FULL: |
| 921 | { |
| 922 | tmpPos.setIndex(startIdx); |
| 923 | tmpPos.setErrorIndex(-1); |
| 924 | |
| 925 | // Exclude the case of UTC Indicator "Z" here |
| 926 | UBool hasDigitOffset = false; |
| 927 | offset = parseOffsetISO8601(text, tmpPos, false, &hasDigitOffset); |
| 928 | if (tmpPos.getErrorIndex() == -1 && hasDigitOffset) { |
| 929 | pos.setIndex(tmpPos.getIndex()); |
| 930 | return createTimeZoneForOffset(offset); |
| 931 | } |
| 932 | |
| 933 | break; |
| 934 | } |
| 935 | |
| 936 | case UTZFMT_STYLE_SPECIFIC_LONG: |
| 937 | case UTZFMT_STYLE_SPECIFIC_SHORT: |
| 938 | { |
| 939 | // Specific styles |
| 940 | int32_t nameTypes = 0; |
| 941 | if (style == UTZFMT_STYLE_SPECIFIC_LONG) { |
| 942 | nameTypes = (UTZNM_LONG_STANDARD | UTZNM_LONG_DAYLIGHT); |
| 943 | } else { |
| 944 | U_ASSERT(style == UTZFMT_STYLE_SPECIFIC_SHORT)(static_cast <bool> (style == UTZFMT_STYLE_SPECIFIC_SHORT ) ? void (0) : __assert_fail ("style == UTZFMT_STYLE_SPECIFIC_SHORT" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 945 | nameTypes = (UTZNM_SHORT_STANDARD | UTZNM_SHORT_DAYLIGHT); |
| 946 | } |
| 947 | LocalPointer<TimeZoneNames::MatchInfoCollection> specificMatches(fTimeZoneNames->find(text, startIdx, nameTypes, status)); |
| 948 | if (U_FAILURE(status)) { |
| 949 | pos.setErrorIndex(startIdx); |
| 950 | return nullptr; |
| 951 | } |
| 952 | if (!specificMatches.isNull()) { |
| 953 | int32_t matchIdx = -1; |
| 954 | int32_t matchPos = -1; |
| 955 | for (int32_t i = 0; i < specificMatches->size(); i++) { |
| 956 | matchPos = startIdx + specificMatches->getMatchLengthAt(i); |
| 957 | if (matchPos > parsedPos) { |
| 958 | matchIdx = i; |
| 959 | parsedPos = matchPos; |
| 960 | } |
| 961 | } |
| 962 | if (matchIdx >= 0) { |
| 963 | if (timeType) { |
| 964 | *timeType = getTimeType(specificMatches->getNameTypeAt(matchIdx)); |
| 965 | } |
| 966 | pos.setIndex(matchPos); |
| 967 | getTimeZoneID(specificMatches.getAlias(), matchIdx, tzID); |
| 968 | U_ASSERT(!tzID.isEmpty())(static_cast <bool> (!tzID.isEmpty()) ? void (0) : __assert_fail ("!tzID.isEmpty()", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 969 | return TimeZone::createTimeZone(tzID); |
| 970 | } |
| 971 | } |
| 972 | |
| 973 | if (parseTZDBAbbrev && style == UTZFMT_STYLE_SPECIFIC_SHORT) { |
| 974 | U_ASSERT((nameTypes & UTZNM_SHORT_STANDARD) != 0)(static_cast <bool> ((nameTypes & UTZNM_SHORT_STANDARD ) != 0) ? void (0) : __assert_fail ("(nameTypes & UTZNM_SHORT_STANDARD) != 0" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 975 | U_ASSERT((nameTypes & UTZNM_SHORT_DAYLIGHT) != 0)(static_cast <bool> ((nameTypes & UTZNM_SHORT_DAYLIGHT ) != 0) ? void (0) : __assert_fail ("(nameTypes & UTZNM_SHORT_DAYLIGHT) != 0" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 976 | |
| 977 | const TZDBTimeZoneNames *tzdbTimeZoneNames = getTZDBTimeZoneNames(status); |
| 978 | if (U_SUCCESS(status)) { |
| 979 | LocalPointer<TimeZoneNames::MatchInfoCollection> tzdbNameMatches( |
| 980 | tzdbTimeZoneNames->find(text, startIdx, nameTypes, status)); |
| 981 | if (U_FAILURE(status)) { |
| 982 | pos.setErrorIndex(startIdx); |
| 983 | return nullptr; |
| 984 | } |
| 985 | if (!tzdbNameMatches.isNull()) { |
| 986 | int32_t matchIdx = -1; |
| 987 | int32_t matchPos = -1; |
| 988 | for (int32_t i = 0; i < tzdbNameMatches->size(); i++) { |
| 989 | matchPos = startIdx + tzdbNameMatches->getMatchLengthAt(i); |
| 990 | if (matchPos > parsedPos) { |
| 991 | matchIdx = i; |
| 992 | parsedPos = matchPos; |
| 993 | } |
| 994 | } |
| 995 | if (matchIdx >= 0) { |
| 996 | if (timeType) { |
| 997 | *timeType = getTimeType(tzdbNameMatches->getNameTypeAt(matchIdx)); |
| 998 | } |
| 999 | pos.setIndex(matchPos); |
| 1000 | getTimeZoneID(tzdbNameMatches.getAlias(), matchIdx, tzID); |
| 1001 | U_ASSERT(!tzID.isEmpty())(static_cast <bool> (!tzID.isEmpty()) ? void (0) : __assert_fail ("!tzID.isEmpty()", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1002 | return TimeZone::createTimeZone(tzID); |
| 1003 | } |
| 1004 | } |
| 1005 | } |
| 1006 | } |
| 1007 | break; |
| 1008 | } |
| 1009 | case UTZFMT_STYLE_GENERIC_LONG: |
| 1010 | case UTZFMT_STYLE_GENERIC_SHORT: |
| 1011 | case UTZFMT_STYLE_GENERIC_LOCATION: |
| 1012 | { |
| 1013 | int32_t genericNameTypes = 0; |
| 1014 | switch (style) { |
| 1015 | case UTZFMT_STYLE_GENERIC_LOCATION: |
| 1016 | genericNameTypes = UTZGNM_LOCATION; |
| 1017 | break; |
| 1018 | |
| 1019 | case UTZFMT_STYLE_GENERIC_LONG: |
| 1020 | genericNameTypes = UTZGNM_LONG | UTZGNM_LOCATION; |
| 1021 | break; |
| 1022 | |
| 1023 | case UTZFMT_STYLE_GENERIC_SHORT: |
| 1024 | genericNameTypes = UTZGNM_SHORT | UTZGNM_LOCATION; |
| 1025 | break; |
| 1026 | |
| 1027 | default: |
| 1028 | UPRV_UNREACHABLE_EXITabort(); |
| 1029 | } |
| 1030 | |
| 1031 | int32_t len = 0; |
| 1032 | UTimeZoneFormatTimeType tt = UTZFMT_TIME_TYPE_UNKNOWN; |
| 1033 | const TimeZoneGenericNames *gnames = getTimeZoneGenericNames(status); |
| 1034 | if (U_SUCCESS(status)) { |
| 1035 | len = gnames->findBestMatch(text, startIdx, genericNameTypes, tzID, tt, status); |
| 1036 | } |
| 1037 | if (U_FAILURE(status)) { |
| 1038 | pos.setErrorIndex(startIdx); |
| 1039 | return nullptr; |
| 1040 | } |
| 1041 | if (len > 0) { |
| 1042 | // Found a match |
| 1043 | if (timeType) { |
| 1044 | *timeType = tt; |
| 1045 | } |
| 1046 | pos.setIndex(startIdx + len); |
| 1047 | U_ASSERT(!tzID.isEmpty())(static_cast <bool> (!tzID.isEmpty()) ? void (0) : __assert_fail ("!tzID.isEmpty()", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1048 | return TimeZone::createTimeZone(tzID); |
| 1049 | } |
| 1050 | |
| 1051 | break; |
| 1052 | } |
| 1053 | case UTZFMT_STYLE_ZONE_ID: |
| 1054 | { |
| 1055 | tmpPos.setIndex(startIdx); |
| 1056 | tmpPos.setErrorIndex(-1); |
| 1057 | |
| 1058 | parseZoneID(text, tmpPos, tzID); |
| 1059 | if (tmpPos.getErrorIndex() == -1) { |
| 1060 | pos.setIndex(tmpPos.getIndex()); |
| 1061 | return TimeZone::createTimeZone(tzID); |
| 1062 | } |
| 1063 | break; |
| 1064 | } |
| 1065 | case UTZFMT_STYLE_ZONE_ID_SHORT: |
| 1066 | { |
| 1067 | tmpPos.setIndex(startIdx); |
| 1068 | tmpPos.setErrorIndex(-1); |
| 1069 | |
| 1070 | parseShortZoneID(text, tmpPos, tzID); |
| 1071 | if (tmpPos.getErrorIndex() == -1) { |
| 1072 | pos.setIndex(tmpPos.getIndex()); |
| 1073 | return TimeZone::createTimeZone(tzID); |
| 1074 | } |
| 1075 | break; |
| 1076 | } |
| 1077 | case UTZFMT_STYLE_EXEMPLAR_LOCATION: |
| 1078 | { |
| 1079 | tmpPos.setIndex(startIdx); |
| 1080 | tmpPos.setErrorIndex(-1); |
| 1081 | |
| 1082 | parseExemplarLocation(text, tmpPos, tzID); |
| 1083 | if (tmpPos.getErrorIndex() == -1) { |
| 1084 | pos.setIndex(tmpPos.getIndex()); |
| 1085 | return TimeZone::createTimeZone(tzID); |
| 1086 | } |
| 1087 | break; |
| 1088 | } |
| 1089 | } |
| 1090 | evaluated |= STYLE_PARSE_FLAGS[style]; |
| 1091 | |
| 1092 | |
| 1093 | if (parsedPos > startIdx) { |
| 1094 | // When the specified style is one of SPECIFIC_XXX or GENERIC_XXX, we tried to parse the input |
| 1095 | // as localized GMT format earlier. If parsedOffset is positive, it means it was successfully |
| 1096 | // parsed as localized GMT format, but offset digits were not detected (more specifically, GMT |
| 1097 | // zero format). Then, it tried to find a match within the set of display names, but could not |
| 1098 | // find a match. At this point, we can safely assume the input text contains the localized |
| 1099 | // GMT format. |
| 1100 | U_ASSERT(parsedOffset != UNKNOWN_OFFSET)(static_cast <bool> (parsedOffset != UNKNOWN_OFFSET) ? void (0) : __assert_fail ("parsedOffset != UNKNOWN_OFFSET", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1101 | pos.setIndex(parsedPos); |
| 1102 | return createTimeZoneForOffset(parsedOffset); |
| 1103 | } |
| 1104 | |
| 1105 | // Failed to parse the input text as the time zone format in the specified style. |
| 1106 | // Check the longest match among other styles below. |
| 1107 | char16_t parsedIDBuf[32]; |
| 1108 | UnicodeString parsedID(parsedIDBuf, 0, UPRV_LENGTHOF(parsedIDBuf)(int32_t)(sizeof(parsedIDBuf)/sizeof((parsedIDBuf)[0]))); |
| 1109 | UTimeZoneFormatTimeType parsedTimeType = UTZFMT_TIME_TYPE_UNKNOWN; |
| 1110 | |
| 1111 | U_ASSERT(parsedPos < 0)(static_cast <bool> (parsedPos < 0) ? void (0) : __assert_fail ("parsedPos < 0", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1112 | U_ASSERT(parsedOffset == UNKNOWN_OFFSET)(static_cast <bool> (parsedOffset == UNKNOWN_OFFSET) ? void (0) : __assert_fail ("parsedOffset == UNKNOWN_OFFSET", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1113 | |
| 1114 | // ISO 8601 |
| 1115 | if (parsedPos < maxPos && |
| 1116 | ((evaluated & ISO_Z_STYLE_FLAG0x0080) == 0 || (evaluated & ISO_LOCAL_STYLE_FLAG0x0100) == 0)) { |
| 1117 | tmpPos.setIndex(startIdx); |
| 1118 | tmpPos.setErrorIndex(-1); |
| 1119 | |
| 1120 | UBool hasDigitOffset = false; |
| 1121 | offset = parseOffsetISO8601(text, tmpPos, false, &hasDigitOffset); |
| 1122 | if (tmpPos.getErrorIndex() == -1) { |
| 1123 | if (tmpPos.getIndex() == maxPos || hasDigitOffset) { |
| 1124 | pos.setIndex(tmpPos.getIndex()); |
| 1125 | return createTimeZoneForOffset(offset); |
| 1126 | } |
| 1127 | // Note: When ISO 8601 format contains offset digits, it should not |
| 1128 | // collide with other formats. However, ISO 8601 UTC format "Z" (single letter) |
| 1129 | // may collide with other names. In this case, we need to evaluate other names. |
| 1130 | if (parsedPos < tmpPos.getIndex()) { |
| 1131 | parsedOffset = offset; |
| 1132 | parsedID.setToBogus(); |
| 1133 | parsedTimeType = UTZFMT_TIME_TYPE_UNKNOWN; |
| 1134 | parsedPos = tmpPos.getIndex(); |
| 1135 | U_ASSERT(parsedPos == startIdx + 1)(static_cast <bool> (parsedPos == startIdx + 1) ? void ( 0) : __assert_fail ("parsedPos == startIdx + 1", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); // only when "Z" is used |
| 1136 | } |
| 1137 | } |
| 1138 | } |
| 1139 | |
| 1140 | // Localized GMT format |
| 1141 | if (parsedPos < maxPos && |
| 1142 | (evaluated & STYLE_PARSE_FLAGS[UTZFMT_STYLE_LOCALIZED_GMT]) == 0) { |
| 1143 | tmpPos.setIndex(startIdx); |
| 1144 | tmpPos.setErrorIndex(-1); |
| 1145 | |
| 1146 | UBool hasDigitOffset = false; |
| 1147 | offset = parseOffsetLocalizedGMT(text, tmpPos, false, &hasDigitOffset); |
| 1148 | if (tmpPos.getErrorIndex() == -1) { |
| 1149 | if (tmpPos.getIndex() == maxPos || hasDigitOffset) { |
| 1150 | pos.setIndex(tmpPos.getIndex()); |
| 1151 | return createTimeZoneForOffset(offset); |
| 1152 | } |
| 1153 | // Evaluate other names - see the comment earlier in this method. |
| 1154 | if (parsedPos < tmpPos.getIndex()) { |
| 1155 | parsedOffset = offset; |
| 1156 | parsedID.setToBogus(); |
| 1157 | parsedTimeType = UTZFMT_TIME_TYPE_UNKNOWN; |
| 1158 | parsedPos = tmpPos.getIndex(); |
| 1159 | } |
| 1160 | } |
| 1161 | } |
| 1162 | |
| 1163 | if (parsedPos < maxPos && |
| 1164 | (evaluated & STYLE_PARSE_FLAGS[UTZFMT_STYLE_LOCALIZED_GMT_SHORT]) == 0) { |
| 1165 | tmpPos.setIndex(startIdx); |
| 1166 | tmpPos.setErrorIndex(-1); |
| 1167 | |
| 1168 | UBool hasDigitOffset = false; |
| 1169 | offset = parseOffsetLocalizedGMT(text, tmpPos, true, &hasDigitOffset); |
| 1170 | if (tmpPos.getErrorIndex() == -1) { |
| 1171 | if (tmpPos.getIndex() == maxPos || hasDigitOffset) { |
| 1172 | pos.setIndex(tmpPos.getIndex()); |
| 1173 | return createTimeZoneForOffset(offset); |
| 1174 | } |
| 1175 | // Evaluate other names - see the comment earlier in this method. |
| 1176 | if (parsedPos < tmpPos.getIndex()) { |
| 1177 | parsedOffset = offset; |
| 1178 | parsedID.setToBogus(); |
| 1179 | parsedTimeType = UTZFMT_TIME_TYPE_UNKNOWN; |
| 1180 | parsedPos = tmpPos.getIndex(); |
| 1181 | } |
| 1182 | } |
| 1183 | } |
| 1184 | |
| 1185 | // When ParseOption.ALL_STYLES is available, we also try to look all possible display names and IDs. |
| 1186 | // For example, when style is GENERIC_LONG, "EST" (SPECIFIC_SHORT) is never |
| 1187 | // used for America/New_York. With parseAllStyles true, this code parses "EST" |
| 1188 | // as America/New_York. |
| 1189 | |
| 1190 | // Note: Adding all possible names into the trie used by the implementation is quite heavy operation, |
| 1191 | // which we want to avoid normally (note that we cache the trie, so this is applicable to the |
| 1192 | // first time only as long as the cache does not expire). |
| 1193 | |
| 1194 | if (parseOptions & UTZFMT_PARSE_OPTION_ALL_STYLES) { |
| 1195 | // Try all specific names and exemplar location names |
| 1196 | if (parsedPos < maxPos) { |
| 1197 | LocalPointer<TimeZoneNames::MatchInfoCollection> specificMatches(fTimeZoneNames->find(text, startIdx, ALL_SIMPLE_NAME_TYPES, status)); |
| 1198 | if (U_FAILURE(status)) { |
| 1199 | pos.setErrorIndex(startIdx); |
| 1200 | return nullptr; |
| 1201 | } |
| 1202 | int32_t specificMatchIdx = -1; |
| 1203 | int32_t matchPos = -1; |
| 1204 | if (!specificMatches.isNull()) { |
| 1205 | for (int32_t i = 0; i < specificMatches->size(); i++) { |
| 1206 | if (startIdx + specificMatches->getMatchLengthAt(i) > matchPos) { |
| 1207 | specificMatchIdx = i; |
| 1208 | matchPos = startIdx + specificMatches->getMatchLengthAt(i); |
| 1209 | } |
| 1210 | } |
| 1211 | } |
| 1212 | if (parsedPos < matchPos) { |
| 1213 | U_ASSERT(specificMatchIdx >= 0)(static_cast <bool> (specificMatchIdx >= 0) ? void ( 0) : __assert_fail ("specificMatchIdx >= 0", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1214 | parsedPos = matchPos; |
| 1215 | getTimeZoneID(specificMatches.getAlias(), specificMatchIdx, parsedID); |
| 1216 | parsedTimeType = getTimeType(specificMatches->getNameTypeAt(specificMatchIdx)); |
| 1217 | parsedOffset = UNKNOWN_OFFSET; |
| 1218 | } |
| 1219 | } |
| 1220 | if (parseTZDBAbbrev && parsedPos < maxPos && (evaluated & STYLE_PARSE_FLAGS[UTZFMT_STYLE_SPECIFIC_SHORT]) == 0) { |
| 1221 | const TZDBTimeZoneNames *tzdbTimeZoneNames = getTZDBTimeZoneNames(status); |
| 1222 | if (U_SUCCESS(status)) { |
| 1223 | LocalPointer<TimeZoneNames::MatchInfoCollection> tzdbNameMatches( |
| 1224 | tzdbTimeZoneNames->find(text, startIdx, ALL_SIMPLE_NAME_TYPES, status)); |
| 1225 | if (U_FAILURE(status)) { |
| 1226 | pos.setErrorIndex(startIdx); |
| 1227 | return nullptr; |
| 1228 | } |
| 1229 | int32_t tzdbNameMatchIdx = -1; |
| 1230 | int32_t matchPos = -1; |
| 1231 | if (!tzdbNameMatches.isNull()) { |
| 1232 | for (int32_t i = 0; i < tzdbNameMatches->size(); i++) { |
| 1233 | if (startIdx + tzdbNameMatches->getMatchLengthAt(i) > matchPos) { |
| 1234 | tzdbNameMatchIdx = i; |
| 1235 | matchPos = startIdx + tzdbNameMatches->getMatchLengthAt(i); |
| 1236 | } |
| 1237 | } |
| 1238 | } |
| 1239 | if (parsedPos < matchPos) { |
| 1240 | U_ASSERT(tzdbNameMatchIdx >= 0)(static_cast <bool> (tzdbNameMatchIdx >= 0) ? void ( 0) : __assert_fail ("tzdbNameMatchIdx >= 0", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1241 | parsedPos = matchPos; |
| 1242 | getTimeZoneID(tzdbNameMatches.getAlias(), tzdbNameMatchIdx, parsedID); |
| 1243 | parsedTimeType = getTimeType(tzdbNameMatches->getNameTypeAt(tzdbNameMatchIdx)); |
| 1244 | parsedOffset = UNKNOWN_OFFSET; |
| 1245 | } |
| 1246 | } |
| 1247 | } |
| 1248 | // Try generic names |
| 1249 | if (parsedPos < maxPos) { |
| 1250 | int32_t genMatchLen = -1; |
| 1251 | UTimeZoneFormatTimeType tt = UTZFMT_TIME_TYPE_UNKNOWN; |
| 1252 | |
| 1253 | const TimeZoneGenericNames *gnames = getTimeZoneGenericNames(status); |
| 1254 | if (U_SUCCESS(status)) { |
| 1255 | genMatchLen = gnames->findBestMatch(text, startIdx, ALL_GENERIC_NAME_TYPES, tzID, tt, status); |
| 1256 | } |
| 1257 | if (U_FAILURE(status)) { |
| 1258 | pos.setErrorIndex(startIdx); |
| 1259 | return nullptr; |
| 1260 | } |
| 1261 | |
| 1262 | if (genMatchLen > 0 && parsedPos < startIdx + genMatchLen) { |
| 1263 | parsedPos = startIdx + genMatchLen; |
| 1264 | parsedID.setTo(tzID); |
| 1265 | parsedTimeType = tt; |
| 1266 | parsedOffset = UNKNOWN_OFFSET; |
| 1267 | } |
| 1268 | } |
| 1269 | |
| 1270 | // Try time zone ID |
| 1271 | if (parsedPos < maxPos && (evaluated & STYLE_PARSE_FLAGS[UTZFMT_STYLE_ZONE_ID]) == 0) { |
| 1272 | tmpPos.setIndex(startIdx); |
| 1273 | tmpPos.setErrorIndex(-1); |
| 1274 | |
| 1275 | parseZoneID(text, tmpPos, tzID); |
| 1276 | if (tmpPos.getErrorIndex() == -1 && parsedPos < tmpPos.getIndex()) { |
| 1277 | parsedPos = tmpPos.getIndex(); |
| 1278 | parsedID.setTo(tzID); |
| 1279 | parsedTimeType = UTZFMT_TIME_TYPE_UNKNOWN; |
| 1280 | parsedOffset = UNKNOWN_OFFSET; |
| 1281 | } |
| 1282 | } |
| 1283 | // Try short time zone ID |
| 1284 | if (parsedPos < maxPos && (evaluated & STYLE_PARSE_FLAGS[UTZFMT_STYLE_ZONE_ID]) == 0) { |
| 1285 | tmpPos.setIndex(startIdx); |
| 1286 | tmpPos.setErrorIndex(-1); |
| 1287 | |
| 1288 | parseShortZoneID(text, tmpPos, tzID); |
| 1289 | if (tmpPos.getErrorIndex() == -1 && parsedPos < tmpPos.getIndex()) { |
| 1290 | parsedPos = tmpPos.getIndex(); |
| 1291 | parsedID.setTo(tzID); |
| 1292 | parsedTimeType = UTZFMT_TIME_TYPE_UNKNOWN; |
| 1293 | parsedOffset = UNKNOWN_OFFSET; |
| 1294 | } |
| 1295 | } |
| 1296 | } |
| 1297 | |
| 1298 | if (parsedPos > startIdx) { |
| 1299 | // Parsed successfully |
| 1300 | TimeZone* parsedTZ; |
| 1301 | if (parsedID.length() > 0) { |
| 1302 | parsedTZ = TimeZone::createTimeZone(parsedID); |
| 1303 | } else { |
| 1304 | U_ASSERT(parsedOffset != UNKNOWN_OFFSET)(static_cast <bool> (parsedOffset != UNKNOWN_OFFSET) ? void (0) : __assert_fail ("parsedOffset != UNKNOWN_OFFSET", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1305 | parsedTZ = createTimeZoneForOffset(parsedOffset); |
| 1306 | } |
| 1307 | if (timeType) { |
| 1308 | *timeType = parsedTimeType; |
| 1309 | } |
| 1310 | pos.setIndex(parsedPos); |
| 1311 | return parsedTZ; |
| 1312 | } |
| 1313 | |
| 1314 | pos.setErrorIndex(startIdx); |
| 1315 | return nullptr; |
| 1316 | } |
| 1317 | |
| 1318 | void |
| 1319 | TimeZoneFormat::parseObject(const UnicodeString& source, Formattable& result, |
| 1320 | ParsePosition& parse_pos) const { |
| 1321 | result.adoptObject(parse(UTZFMT_STYLE_GENERIC_LOCATION, source, parse_pos, UTZFMT_PARSE_OPTION_ALL_STYLES)); |
| 1322 | } |
| 1323 | |
| 1324 | |
| 1325 | // ------------------------------------------------------------------ |
| 1326 | // Private zone name format/parse implementation |
| 1327 | |
| 1328 | UnicodeString& |
| 1329 | TimeZoneFormat::formatGeneric(const TimeZone& tz, int32_t genType, UDate date, UnicodeString& name) const { |
| 1330 | UErrorCode status = U_ZERO_ERROR; |
| 1331 | const TimeZoneGenericNames* gnames = getTimeZoneGenericNames(status); |
| 1332 | if (U_FAILURE(status)) { |
| 1333 | name.setToBogus(); |
| 1334 | return name; |
| 1335 | } |
| 1336 | |
| 1337 | if (genType == UTZGNM_LOCATION) { |
| 1338 | const char16_t* canonicalID = ZoneMeta::getCanonicalCLDRID(tz); |
| 1339 | if (canonicalID == nullptr) { |
| 1340 | name.setToBogus(); |
| 1341 | return name; |
| 1342 | } |
| 1343 | return gnames->getGenericLocationName(UnicodeString(true, canonicalID, -1), name); |
| 1344 | } |
| 1345 | return gnames->getDisplayName(tz, static_cast<UTimeZoneGenericNameType>(genType), date, name); |
| 1346 | } |
| 1347 | |
| 1348 | UnicodeString& |
| 1349 | TimeZoneFormat::formatSpecific(const TimeZone& tz, UTimeZoneNameType stdType, UTimeZoneNameType dstType, |
| 1350 | UDate date, UnicodeString& name, UTimeZoneFormatTimeType *timeType) const { |
| 1351 | if (fTimeZoneNames == nullptr) { |
| 1352 | name.setToBogus(); |
| 1353 | return name; |
| 1354 | } |
| 1355 | |
| 1356 | UErrorCode status = U_ZERO_ERROR; |
| 1357 | UBool isDaylight = tz.inDaylightTime(date, status); |
| 1358 | const char16_t* canonicalID = ZoneMeta::getCanonicalCLDRID(tz); |
| 1359 | |
| 1360 | if (U_FAILURE(status) || canonicalID == nullptr) { |
| 1361 | name.setToBogus(); |
| 1362 | return name; |
| 1363 | } |
| 1364 | |
| 1365 | if (isDaylight) { |
| 1366 | fTimeZoneNames->getDisplayName(UnicodeString(true, canonicalID, -1), dstType, date, name); |
| 1367 | } else { |
| 1368 | fTimeZoneNames->getDisplayName(UnicodeString(true, canonicalID, -1), stdType, date, name); |
| 1369 | } |
| 1370 | |
| 1371 | if (timeType && !name.isEmpty()) { |
| 1372 | *timeType = isDaylight ? UTZFMT_TIME_TYPE_DAYLIGHT : UTZFMT_TIME_TYPE_STANDARD; |
| 1373 | } |
| 1374 | return name; |
| 1375 | } |
| 1376 | |
| 1377 | const TimeZoneGenericNames* |
| 1378 | TimeZoneFormat::getTimeZoneGenericNames(UErrorCode& status) const { |
| 1379 | if (U_FAILURE(status)) { |
| 1380 | return nullptr; |
| 1381 | } |
| 1382 | |
| 1383 | umtx_lock(&gLock); |
| 1384 | if (fTimeZoneGenericNames == nullptr) { |
| 1385 | TimeZoneFormat *nonConstThis = const_cast<TimeZoneFormat *>(this); |
| 1386 | nonConstThis->fTimeZoneGenericNames = TimeZoneGenericNames::createInstance(fLocale, status); |
| 1387 | } |
| 1388 | umtx_unlock(&gLock); |
| 1389 | |
| 1390 | return fTimeZoneGenericNames; |
| 1391 | } |
| 1392 | |
| 1393 | const TZDBTimeZoneNames* |
| 1394 | TimeZoneFormat::getTZDBTimeZoneNames(UErrorCode& status) const { |
| 1395 | if (U_FAILURE(status)) { |
| 1396 | return nullptr; |
| 1397 | } |
| 1398 | |
| 1399 | umtx_lock(&gLock); |
| 1400 | if (fTZDBTimeZoneNames == nullptr) { |
| 1401 | TZDBTimeZoneNames *tzdbNames = new TZDBTimeZoneNames(fLocale); |
| 1402 | if (tzdbNames == nullptr) { |
| 1403 | status = U_MEMORY_ALLOCATION_ERROR; |
| 1404 | } else { |
| 1405 | TimeZoneFormat *nonConstThis = const_cast<TimeZoneFormat *>(this); |
| 1406 | nonConstThis->fTZDBTimeZoneNames = tzdbNames; |
| 1407 | } |
| 1408 | } |
| 1409 | umtx_unlock(&gLock); |
| 1410 | |
| 1411 | return fTZDBTimeZoneNames; |
| 1412 | } |
| 1413 | |
| 1414 | UnicodeString& |
| 1415 | TimeZoneFormat::formatExemplarLocation(const TimeZone& tz, UnicodeString& name) const { |
| 1416 | char16_t locationBuf[ZONE_NAME_U16_MAX128]; |
| 1417 | UnicodeString location(locationBuf, 0, UPRV_LENGTHOF(locationBuf)(int32_t)(sizeof(locationBuf)/sizeof((locationBuf)[0]))); |
| 1418 | const char16_t* canonicalID = ZoneMeta::getCanonicalCLDRID(tz); |
| 1419 | |
| 1420 | if (canonicalID) { |
| 1421 | fTimeZoneNames->getExemplarLocationName(UnicodeString(true, canonicalID, -1), location); |
| 1422 | } |
| 1423 | if (location.length() > 0) { |
| 1424 | name.setTo(location); |
| 1425 | } else { |
| 1426 | // Use "unknown" location |
| 1427 | fTimeZoneNames->getExemplarLocationName(UnicodeString(true, UNKNOWN_ZONE_ID, -1), location); |
| 1428 | if (location.length() > 0) { |
| 1429 | name.setTo(location); |
| 1430 | } else { |
| 1431 | // last resort |
| 1432 | name.setTo(UNKNOWN_LOCATION, -1); |
| 1433 | } |
| 1434 | } |
| 1435 | return name; |
| 1436 | } |
| 1437 | |
| 1438 | |
| 1439 | // ------------------------------------------------------------------ |
| 1440 | // Zone offset format and parse |
| 1441 | |
| 1442 | UnicodeString& |
| 1443 | TimeZoneFormat::formatOffsetISO8601Basic(int32_t offset, UBool useUtcIndicator, UBool isShort, UBool ignoreSeconds, |
| 1444 | UnicodeString& result, UErrorCode& status) const { |
| 1445 | return formatOffsetISO8601(offset, true, useUtcIndicator, isShort, ignoreSeconds, result, status); |
| 1446 | } |
| 1447 | |
| 1448 | UnicodeString& |
| 1449 | TimeZoneFormat::formatOffsetISO8601Extended(int32_t offset, UBool useUtcIndicator, UBool isShort, UBool ignoreSeconds, |
| 1450 | UnicodeString& result, UErrorCode& status) const { |
| 1451 | return formatOffsetISO8601(offset, false, useUtcIndicator, isShort, ignoreSeconds, result, status); |
| 1452 | } |
| 1453 | |
| 1454 | UnicodeString& |
| 1455 | TimeZoneFormat::formatOffsetLocalizedGMT(int32_t offset, UnicodeString& result, UErrorCode& status) const { |
| 1456 | return formatOffsetLocalizedGMT(offset, false, result, status); |
| 1457 | } |
| 1458 | |
| 1459 | UnicodeString& |
| 1460 | TimeZoneFormat::formatOffsetShortLocalizedGMT(int32_t offset, UnicodeString& result, UErrorCode& status) const { |
| 1461 | return formatOffsetLocalizedGMT(offset, true, result, status); |
| 1462 | } |
| 1463 | |
| 1464 | int32_t |
| 1465 | TimeZoneFormat::parseOffsetISO8601(const UnicodeString& text, ParsePosition& pos) const { |
| 1466 | return parseOffsetISO8601(text, pos, false); |
| 1467 | } |
| 1468 | |
| 1469 | int32_t |
| 1470 | TimeZoneFormat::parseOffsetLocalizedGMT(const UnicodeString& text, ParsePosition& pos) const { |
| 1471 | return parseOffsetLocalizedGMT(text, pos, false, nullptr); |
| 1472 | } |
| 1473 | |
| 1474 | int32_t |
| 1475 | TimeZoneFormat::parseOffsetShortLocalizedGMT(const UnicodeString& text, ParsePosition& pos) const { |
| 1476 | return parseOffsetLocalizedGMT(text, pos, true, nullptr); |
| 1477 | } |
| 1478 | |
| 1479 | // ------------------------------------------------------------------ |
| 1480 | // Private zone offset format/parse implementation |
| 1481 | |
| 1482 | UnicodeString& |
| 1483 | TimeZoneFormat::formatOffsetISO8601(int32_t offset, UBool isBasic, UBool useUtcIndicator, |
| 1484 | UBool isShort, UBool ignoreSeconds, UnicodeString& result, UErrorCode& status) const { |
| 1485 | if (U_FAILURE(status)) { |
| 1486 | result.setToBogus(); |
| 1487 | return result; |
| 1488 | } |
| 1489 | int32_t absOffset = offset < 0 ? -offset : offset; |
| 1490 | if (useUtcIndicator && (absOffset < MILLIS_PER_SECOND || (ignoreSeconds && absOffset < MILLIS_PER_MINUTE))) { |
| 1491 | result.setTo(ISO8601_UTC); |
| 1492 | return result; |
| 1493 | } |
| 1494 | |
| 1495 | OffsetFields minFields = isShort ? FIELDS_H : FIELDS_HM; |
| 1496 | OffsetFields maxFields = ignoreSeconds ? FIELDS_HM : FIELDS_HMS; |
| 1497 | char16_t sep = isBasic ? 0 : ISO8601_SEP; |
| 1498 | |
| 1499 | // Note: FIELDS_HMS as maxFields is a CLDR/ICU extension. ISO 8601 specification does |
| 1500 | // not support seconds field. |
| 1501 | |
| 1502 | if (absOffset >= MAX_OFFSET) { |
| 1503 | result.setToBogus(); |
| 1504 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 1505 | return result; |
| 1506 | } |
| 1507 | |
| 1508 | int fields[3]; |
| 1509 | fields[0] = absOffset / MILLIS_PER_HOUR; |
| 1510 | absOffset = absOffset % MILLIS_PER_HOUR; |
| 1511 | fields[1] = absOffset / MILLIS_PER_MINUTE; |
| 1512 | absOffset = absOffset % MILLIS_PER_MINUTE; |
| 1513 | fields[2] = absOffset / MILLIS_PER_SECOND; |
| 1514 | |
| 1515 | U_ASSERT(fields[0] >= 0 && fields[0] <= MAX_OFFSET_HOUR)(static_cast <bool> (fields[0] >= 0 && fields [0] <= MAX_OFFSET_HOUR) ? void (0) : __assert_fail ("fields[0] >= 0 && fields[0] <= MAX_OFFSET_HOUR" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1516 | U_ASSERT(fields[1] >= 0 && fields[1] <= MAX_OFFSET_MINUTE)(static_cast <bool> (fields[1] >= 0 && fields [1] <= MAX_OFFSET_MINUTE) ? void (0) : __assert_fail ("fields[1] >= 0 && fields[1] <= MAX_OFFSET_MINUTE" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1517 | U_ASSERT(fields[2] >= 0 && fields[2] <= MAX_OFFSET_SECOND)(static_cast <bool> (fields[2] >= 0 && fields [2] <= MAX_OFFSET_SECOND) ? void (0) : __assert_fail ("fields[2] >= 0 && fields[2] <= MAX_OFFSET_SECOND" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1518 | |
| 1519 | int32_t lastIdx = maxFields; |
| 1520 | while (lastIdx > minFields) { |
| 1521 | if (fields[lastIdx] != 0) { |
| 1522 | break; |
| 1523 | } |
| 1524 | lastIdx--; |
| 1525 | } |
| 1526 | |
| 1527 | char16_t sign = PLUS; |
| 1528 | if (offset < 0) { |
| 1529 | // if all output fields are 0s, do not use negative sign |
| 1530 | for (int32_t idx = 0; idx <= lastIdx; idx++) { |
| 1531 | if (fields[idx] != 0) { |
| 1532 | sign = MINUS; |
| 1533 | break; |
| 1534 | } |
| 1535 | } |
| 1536 | } |
| 1537 | result.setTo(sign); |
| 1538 | |
| 1539 | for (int32_t idx = 0; idx <= lastIdx; idx++) { |
| 1540 | if (sep && idx != 0) { |
| 1541 | result.append(sep); |
| 1542 | } |
| 1543 | result.append(static_cast<char16_t>(0x0030 + fields[idx] / 10)); |
| 1544 | result.append(static_cast<char16_t>(0x0030 + fields[idx] % 10)); |
| 1545 | } |
| 1546 | |
| 1547 | return result; |
| 1548 | } |
| 1549 | |
| 1550 | UnicodeString& |
| 1551 | TimeZoneFormat::formatOffsetLocalizedGMT(int32_t offset, UBool isShort, UnicodeString& result, UErrorCode& status) const { |
| 1552 | if (U_FAILURE(status)) { |
| 1553 | result.setToBogus(); |
| 1554 | return result; |
| 1555 | } |
| 1556 | if (offset <= -MAX_OFFSET || offset >= MAX_OFFSET) { |
| 1557 | result.setToBogus(); |
| 1558 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 1559 | return result; |
| 1560 | } |
| 1561 | |
| 1562 | UBool positive = true; |
| 1563 | if (offset < 0) { |
| 1564 | offset = -offset; |
| 1565 | positive = false; |
| 1566 | } |
| 1567 | |
| 1568 | int32_t offsetH = offset / MILLIS_PER_HOUR; |
| 1569 | offset = offset % MILLIS_PER_HOUR; |
| 1570 | int32_t offsetM = offset / MILLIS_PER_MINUTE; |
| 1571 | offset = offset % MILLIS_PER_MINUTE; |
| 1572 | int32_t offsetS = offset / MILLIS_PER_SECOND; |
| 1573 | |
| 1574 | U_ASSERT(offsetH <= MAX_OFFSET_HOUR && offsetM <= MAX_OFFSET_MINUTE && offsetS <= MAX_OFFSET_SECOND)(static_cast <bool> (offsetH <= MAX_OFFSET_HOUR && offsetM <= MAX_OFFSET_MINUTE && offsetS <= MAX_OFFSET_SECOND ) ? void (0) : __assert_fail ("offsetH <= MAX_OFFSET_HOUR && offsetM <= MAX_OFFSET_MINUTE && offsetS <= MAX_OFFSET_SECOND" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1575 | |
| 1576 | const UVector* offsetPatternItems = nullptr; |
| 1577 | if (positive) { |
| 1578 | if (offsetS != 0) { |
| 1579 | offsetPatternItems = fGMTOffsetPatternItems[UTZFMT_PAT_POSITIVE_HMS]; |
| 1580 | } else if (offsetM != 0 || !isShort) { |
| 1581 | offsetPatternItems = fGMTOffsetPatternItems[UTZFMT_PAT_POSITIVE_HM]; |
| 1582 | } else { |
| 1583 | offsetPatternItems = fGMTOffsetPatternItems[UTZFMT_PAT_POSITIVE_H]; |
| 1584 | } |
| 1585 | } else { |
| 1586 | if (offsetS != 0) { |
| 1587 | offsetPatternItems = fGMTOffsetPatternItems[UTZFMT_PAT_NEGATIVE_HMS]; |
| 1588 | } else if (offsetM != 0 || !isShort) { |
| 1589 | offsetPatternItems = fGMTOffsetPatternItems[UTZFMT_PAT_NEGATIVE_HM]; |
| 1590 | } else { |
| 1591 | offsetPatternItems = fGMTOffsetPatternItems[UTZFMT_PAT_NEGATIVE_H]; |
| 1592 | } |
| 1593 | } |
| 1594 | |
| 1595 | U_ASSERT(offsetPatternItems != nullptr)(static_cast <bool> (offsetPatternItems != nullptr) ? void (0) : __assert_fail ("offsetPatternItems != nullptr", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1596 | |
| 1597 | // Building the GMT format string |
| 1598 | result.setTo(fGMTPatternPrefix); |
| 1599 | |
| 1600 | for (int32_t i = 0; i < offsetPatternItems->size(); i++) { |
| 1601 | const GMTOffsetField* item = static_cast<GMTOffsetField*>(offsetPatternItems->elementAt(i)); |
| 1602 | GMTOffsetField::FieldType type = item->getType(); |
| 1603 | |
| 1604 | switch (type) { |
| 1605 | case GMTOffsetField::TEXT: |
| 1606 | result.append(item->getPatternText(), -1); |
| 1607 | break; |
| 1608 | |
| 1609 | case GMTOffsetField::HOUR: |
| 1610 | appendOffsetDigits(result, offsetH, (isShort ? 1 : 2)); |
| 1611 | break; |
| 1612 | |
| 1613 | case GMTOffsetField::MINUTE: |
| 1614 | appendOffsetDigits(result, offsetM, 2); |
| 1615 | break; |
| 1616 | |
| 1617 | case GMTOffsetField::SECOND: |
| 1618 | appendOffsetDigits(result, offsetS, 2); |
| 1619 | break; |
| 1620 | } |
| 1621 | } |
| 1622 | |
| 1623 | result.append(fGMTPatternSuffix); |
| 1624 | return result; |
| 1625 | } |
| 1626 | |
| 1627 | int32_t |
| 1628 | TimeZoneFormat::parseOffsetISO8601(const UnicodeString& text, ParsePosition& pos, UBool extendedOnly, UBool* hasDigitOffset /* = nullptr */) const { |
| 1629 | if (hasDigitOffset) { |
| 1630 | *hasDigitOffset = false; |
| 1631 | } |
| 1632 | int32_t start = pos.getIndex(); |
| 1633 | if (start >= text.length()) { |
| 1634 | pos.setErrorIndex(start); |
| 1635 | return 0; |
| 1636 | } |
| 1637 | |
| 1638 | char16_t firstChar = text.charAt(start); |
| 1639 | if (firstChar == ISO8601_UTC || firstChar == static_cast<char16_t>(ISO8601_UTC + 0x20)) { |
| 1640 | // "Z" (or "z") - indicates UTC |
| 1641 | pos.setIndex(start + 1); |
| 1642 | return 0; |
| 1643 | } |
| 1644 | |
| 1645 | int32_t sign = 1; |
| 1646 | if (firstChar == PLUS) { |
| 1647 | sign = 1; |
| 1648 | } else if (firstChar == MINUS) { |
| 1649 | sign = -1; |
| 1650 | } else { |
| 1651 | // Not an ISO 8601 offset string |
| 1652 | pos.setErrorIndex(start); |
| 1653 | return 0; |
| 1654 | } |
| 1655 | ParsePosition posOffset(start + 1); |
| 1656 | int32_t offset = parseAsciiOffsetFields(text, posOffset, ISO8601_SEP, FIELDS_H, FIELDS_HMS); |
| 1657 | if (posOffset.getErrorIndex() == -1 && !extendedOnly && (posOffset.getIndex() - start <= 3)) { |
| 1658 | // If the text is successfully parsed as extended format with the options above, it can be also parsed |
| 1659 | // as basic format. For example, "0230" can be parsed as offset 2:00 (only first digits are valid for |
| 1660 | // extended format), but it can be parsed as offset 2:30 with basic format. We use longer result. |
| 1661 | ParsePosition posBasic(start + 1); |
| 1662 | int32_t tmpOffset = parseAbuttingAsciiOffsetFields(text, posBasic, FIELDS_H, FIELDS_HMS, false); |
| 1663 | if (posBasic.getErrorIndex() == -1 && posBasic.getIndex() > posOffset.getIndex()) { |
| 1664 | offset = tmpOffset; |
| 1665 | posOffset.setIndex(posBasic.getIndex()); |
| 1666 | } |
| 1667 | } |
| 1668 | |
| 1669 | if (posOffset.getErrorIndex() != -1) { |
| 1670 | pos.setErrorIndex(start); |
| 1671 | return 0; |
| 1672 | } |
| 1673 | |
| 1674 | pos.setIndex(posOffset.getIndex()); |
| 1675 | if (hasDigitOffset) { |
| 1676 | *hasDigitOffset = true; |
| 1677 | } |
| 1678 | return sign * offset; |
| 1679 | } |
| 1680 | |
| 1681 | int32_t |
| 1682 | TimeZoneFormat::parseOffsetLocalizedGMT(const UnicodeString& text, ParsePosition& pos, UBool isShort, UBool* hasDigitOffset) const { |
| 1683 | int32_t start = pos.getIndex(); |
| 1684 | int32_t offset = 0; |
| 1685 | int32_t parsedLength = 0; |
| 1686 | |
| 1687 | if (hasDigitOffset) { |
| 1688 | *hasDigitOffset = false; |
| 1689 | } |
| 1690 | |
| 1691 | offset = parseOffsetLocalizedGMTPattern(text, start, isShort, parsedLength); |
| 1692 | |
| 1693 | // For now, parseOffsetLocalizedGMTPattern handles both long and short |
| 1694 | // formats, no matter isShort is true or false. This might be changed in future |
| 1695 | // when strict parsing is necessary, or different set of patterns are used for |
| 1696 | // short/long formats. |
| 1697 | #if 0 |
| 1698 | if (parsedLength == 0) { |
| 1699 | offset = parseOffsetLocalizedGMTPattern(text, start, !isShort, parsedLength); |
| 1700 | } |
| 1701 | #endif |
| 1702 | |
| 1703 | if (parsedLength > 0) { |
| 1704 | if (hasDigitOffset) { |
| 1705 | *hasDigitOffset = true; |
| 1706 | } |
| 1707 | pos.setIndex(start + parsedLength); |
| 1708 | return offset; |
| 1709 | } |
| 1710 | |
| 1711 | // Try the default patterns |
| 1712 | offset = parseOffsetDefaultLocalizedGMT(text, start, parsedLength); |
| 1713 | if (parsedLength > 0) { |
| 1714 | if (hasDigitOffset) { |
| 1715 | *hasDigitOffset = true; |
| 1716 | } |
| 1717 | pos.setIndex(start + parsedLength); |
| 1718 | return offset; |
| 1719 | } |
| 1720 | |
| 1721 | // Check if this is a GMT zero format |
| 1722 | if (text.caseCompare(start, fGMTZeroFormat.length(), fGMTZeroFormat, 0) == 0) { |
| 1723 | pos.setIndex(start + fGMTZeroFormat.length()); |
| 1724 | return 0; |
| 1725 | } |
| 1726 | |
| 1727 | // Check if this is a default GMT zero format |
| 1728 | for (int32_t i = 0; ALT_GMT_STRINGS[i][0] != 0; i++) { |
| 1729 | const char16_t* defGMTZero = ALT_GMT_STRINGS[i]; |
| 1730 | int32_t defGMTZeroLen = u_strlen(defGMTZero); |
| 1731 | if (text.caseCompare(start, defGMTZeroLen, defGMTZero, 0) == 0) { |
| 1732 | pos.setIndex(start + defGMTZeroLen); |
| 1733 | return 0; |
| 1734 | } |
| 1735 | } |
| 1736 | |
| 1737 | // Nothing matched |
| 1738 | pos.setErrorIndex(start); |
| 1739 | return 0; |
| 1740 | } |
| 1741 | |
| 1742 | int32_t |
| 1743 | TimeZoneFormat::parseOffsetLocalizedGMTPattern(const UnicodeString& text, int32_t start, UBool /*isShort*/, int32_t& parsedLen) const { |
| 1744 | int32_t idx = start; |
| 1745 | int32_t offset = 0; |
| 1746 | UBool parsed = false; |
| 1747 | |
| 1748 | do { |
| 1749 | // Prefix part |
| 1750 | int32_t len = fGMTPatternPrefix.length(); |
| 1751 | if (len > 0 && text.caseCompare(idx, len, fGMTPatternPrefix, 0) != 0) { |
| 1752 | // prefix match failed |
| 1753 | break; |
| 1754 | } |
| 1755 | idx += len; |
| 1756 | |
| 1757 | // Offset part |
| 1758 | offset = parseOffsetFields(text, idx, false, len); |
| 1759 | if (len == 0) { |
| 1760 | // offset field match failed |
| 1761 | break; |
| 1762 | } |
| 1763 | idx += len; |
| 1764 | |
| 1765 | len = fGMTPatternSuffix.length(); |
| 1766 | if (len > 0 && text.caseCompare(idx, len, fGMTPatternSuffix, 0) != 0) { |
| 1767 | // no suffix match |
| 1768 | break; |
| 1769 | } |
| 1770 | idx += len; |
| 1771 | parsed = true; |
| 1772 | } while (false); |
| 1773 | |
| 1774 | parsedLen = parsed ? idx - start : 0; |
| 1775 | return offset; |
| 1776 | } |
| 1777 | |
| 1778 | int32_t |
| 1779 | TimeZoneFormat::parseOffsetFields(const UnicodeString& text, int32_t start, UBool /*isShort*/, int32_t& parsedLen) const { |
| 1780 | int32_t outLen = 0; |
| 1781 | int32_t offset = 0; |
| 1782 | int32_t sign = 1; |
| 1783 | |
| 1784 | parsedLen = 0; |
| 1785 | |
| 1786 | int32_t offsetH, offsetM, offsetS; |
| 1787 | offsetH = offsetM = offsetS = 0; |
| 1788 | |
| 1789 | for (int32_t patidx = 0; PARSE_GMT_OFFSET_TYPES[patidx] >= 0; patidx++) { |
| 1790 | int32_t gmtPatType = PARSE_GMT_OFFSET_TYPES[patidx]; |
| 1791 | UVector* items = fGMTOffsetPatternItems[gmtPatType]; |
| 1792 | U_ASSERT(items != nullptr)(static_cast <bool> (items != nullptr) ? void (0) : __assert_fail ("items != nullptr", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1793 | |
| 1794 | outLen = parseOffsetFieldsWithPattern(text, start, items, false, offsetH, offsetM, offsetS); |
| 1795 | if (outLen > 0) { |
| 1796 | sign = (gmtPatType == UTZFMT_PAT_POSITIVE_H || gmtPatType == UTZFMT_PAT_POSITIVE_HM || gmtPatType == UTZFMT_PAT_POSITIVE_HMS) ? |
| 1797 | 1 : -1; |
| 1798 | break; |
| 1799 | } |
| 1800 | } |
| 1801 | |
| 1802 | if (outLen > 0 && fAbuttingOffsetHoursAndMinutes) { |
| 1803 | // When hours field is sabutting minutes field, |
| 1804 | // the parse result above may not be appropriate. |
| 1805 | // For example, "01020" is parsed as 01:02: above, |
| 1806 | // but it should be parsed as 00:10:20. |
| 1807 | int32_t tmpLen = 0; |
| 1808 | int32_t tmpSign = 1; |
| 1809 | int32_t tmpH = 0; |
| 1810 | int32_t tmpM = 0; |
| 1811 | int32_t tmpS = 0; |
| 1812 | |
| 1813 | for (int32_t patidx = 0; PARSE_GMT_OFFSET_TYPES[patidx] >= 0; patidx++) { |
| 1814 | int32_t gmtPatType = PARSE_GMT_OFFSET_TYPES[patidx]; |
| 1815 | UVector* items = fGMTOffsetPatternItems[gmtPatType]; |
| 1816 | U_ASSERT(items != nullptr)(static_cast <bool> (items != nullptr) ? void (0) : __assert_fail ("items != nullptr", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1817 | |
| 1818 | // forcing parse to use single hour digit |
| 1819 | tmpLen = parseOffsetFieldsWithPattern(text, start, items, true, tmpH, tmpM, tmpS); |
| 1820 | if (tmpLen > 0) { |
| 1821 | tmpSign = (gmtPatType == UTZFMT_PAT_POSITIVE_H || gmtPatType == UTZFMT_PAT_POSITIVE_HM || gmtPatType == UTZFMT_PAT_POSITIVE_HMS) ? |
| 1822 | 1 : -1; |
| 1823 | break; |
| 1824 | } |
| 1825 | } |
| 1826 | if (tmpLen > outLen) { |
| 1827 | // Better parse result with single hour digit |
| 1828 | outLen = tmpLen; |
| 1829 | sign = tmpSign; |
| 1830 | offsetH = tmpH; |
| 1831 | offsetM = tmpM; |
| 1832 | offsetS = tmpS; |
| 1833 | } |
| 1834 | } |
| 1835 | |
| 1836 | if (outLen > 0) { |
| 1837 | offset = ((((offsetH * 60) + offsetM) * 60) + offsetS) * 1000 * sign; |
| 1838 | parsedLen = outLen; |
| 1839 | } |
| 1840 | |
| 1841 | return offset; |
| 1842 | } |
| 1843 | |
| 1844 | int32_t |
| 1845 | TimeZoneFormat::parseOffsetFieldsWithPattern(const UnicodeString& text, int32_t start, |
| 1846 | UVector* patternItems, UBool forceSingleHourDigit, int32_t& hour, int32_t& min, int32_t& sec) const { |
| 1847 | UBool failed = false; |
| 1848 | int32_t offsetH, offsetM, offsetS; |
| 1849 | offsetH = offsetM = offsetS = 0; |
| 1850 | int32_t idx = start; |
| 1851 | |
| 1852 | for (int32_t i = 0; i < patternItems->size(); i++) { |
| 1853 | int32_t len = 0; |
| 1854 | const GMTOffsetField* field = static_cast<const GMTOffsetField*>(patternItems->elementAt(i)); |
| 1855 | GMTOffsetField::FieldType fieldType = field->getType(); |
| 1856 | if (fieldType == GMTOffsetField::TEXT) { |
| 1857 | const char16_t* patStr = field->getPatternText(); |
| 1858 | len = u_strlen(patStr); |
| 1859 | if (i == 0) { |
| 1860 | // When TimeZoneFormat parse() is called from SimpleDateFormat, |
| 1861 | // leading space characters might be truncated. If the first pattern text |
| 1862 | // starts with such character (e.g. Bidi control), then we need to |
| 1863 | // skip the leading space characters. |
| 1864 | if (idx < text.length() && !PatternProps::isWhiteSpace(text.char32At(idx))) { |
| 1865 | while (len > 0) { |
| 1866 | UChar32 ch; |
| 1867 | int32_t chLen; |
| 1868 | U16_GET(patStr, 0, 0, len, ch)do { (ch)=(patStr)[0]; if((((ch)&0xfffff800)==0xd800)) { uint16_t __c2; if((((ch)&0x400)==0)) { if((0)+1!=(len) && (((__c2=(patStr)[(0)+1])&0xfffffc00)==0xdc00)) { (ch)=(( (UChar32)((ch))<<10UL)+(UChar32)(__c2)-((0xd800<< 10UL)+0xdc00-0x10000)); } } else { if((0)>(0) && ( ((__c2=(patStr)[(0)-1])&0xfffffc00)==0xd800)) { (ch)=(((UChar32 )(__c2)<<10UL)+(UChar32)((ch))-((0xd800<<10UL)+0xdc00 -0x10000)); } } } } while (false); |
| 1869 | if (PatternProps::isWhiteSpace(ch)) { |
| 1870 | chLen = U16_LENGTH(ch)((uint32_t)(ch)<=0xffff ? 1 : 2); |
| 1871 | len -= chLen; |
| 1872 | patStr += chLen; |
| 1873 | } |
| 1874 | else { |
| 1875 | break; |
| 1876 | } |
| 1877 | } |
| 1878 | } |
| 1879 | } |
| 1880 | if (text.caseCompare(idx, len, patStr, 0) != 0) { |
| 1881 | failed = true; |
| 1882 | break; |
| 1883 | } |
| 1884 | idx += len; |
| 1885 | } else { |
| 1886 | if (fieldType == GMTOffsetField::HOUR) { |
| 1887 | uint8_t maxDigits = forceSingleHourDigit ? 1 : 2; |
| 1888 | offsetH = parseOffsetFieldWithLocalizedDigits(text, idx, 1, maxDigits, 0, MAX_OFFSET_HOUR, len); |
| 1889 | } else if (fieldType == GMTOffsetField::MINUTE) { |
| 1890 | offsetM = parseOffsetFieldWithLocalizedDigits(text, idx, 2, 2, 0, MAX_OFFSET_MINUTE, len); |
| 1891 | } else if (fieldType == GMTOffsetField::SECOND) { |
| 1892 | offsetS = parseOffsetFieldWithLocalizedDigits(text, idx, 2, 2, 0, MAX_OFFSET_SECOND, len); |
| 1893 | } |
| 1894 | |
| 1895 | if (len == 0) { |
| 1896 | failed = true; |
| 1897 | break; |
| 1898 | } |
| 1899 | idx += len; |
| 1900 | } |
| 1901 | } |
| 1902 | |
| 1903 | if (failed) { |
| 1904 | hour = min = sec = 0; |
| 1905 | return 0; |
| 1906 | } |
| 1907 | |
| 1908 | hour = offsetH; |
| 1909 | min = offsetM; |
| 1910 | sec = offsetS; |
| 1911 | |
| 1912 | return idx - start; |
| 1913 | } |
| 1914 | |
| 1915 | int32_t |
| 1916 | TimeZoneFormat::parseAbuttingOffsetFields(const UnicodeString& text, int32_t start, int32_t& parsedLen) const { |
| 1917 | int32_t digits[MAX_OFFSET_DIGITS6]; |
| 1918 | int32_t parsed[MAX_OFFSET_DIGITS6]; // accumulative offsets |
| 1919 | |
| 1920 | // Parse digits into int[] |
| 1921 | int32_t idx = start; |
| 1922 | int32_t len = 0; |
| 1923 | int32_t numDigits = 0; |
| 1924 | for (int32_t i = 0; i < MAX_OFFSET_DIGITS6; i++) { |
| 1925 | digits[i] = parseSingleLocalizedDigit(text, idx, len); |
| 1926 | if (digits[i] < 0) { |
| 1927 | break; |
| 1928 | } |
| 1929 | idx += len; |
| 1930 | parsed[i] = idx - start; |
| 1931 | numDigits++; |
| 1932 | } |
| 1933 | |
| 1934 | if (numDigits == 0) { |
| 1935 | parsedLen = 0; |
| 1936 | return 0; |
| 1937 | } |
| 1938 | |
| 1939 | int32_t offset = 0; |
| 1940 | while (numDigits > 0) { |
| 1941 | int32_t hour = 0; |
| 1942 | int32_t min = 0; |
| 1943 | int32_t sec = 0; |
| 1944 | |
| 1945 | U_ASSERT(numDigits > 0 && numDigits <= MAX_OFFSET_DIGITS)(static_cast <bool> (numDigits > 0 && numDigits <= 6) ? void (0) : __assert_fail ("numDigits > 0 && numDigits <= 6" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1946 | switch (numDigits) { |
| 1947 | case 1: // H |
| 1948 | hour = digits[0]; |
| 1949 | break; |
| 1950 | case 2: // HH |
| 1951 | hour = digits[0] * 10 + digits[1]; |
| 1952 | break; |
| 1953 | case 3: // Hmm |
| 1954 | hour = digits[0]; |
| 1955 | min = digits[1] * 10 + digits[2]; |
| 1956 | break; |
| 1957 | case 4: // HHmm |
| 1958 | hour = digits[0] * 10 + digits[1]; |
| 1959 | min = digits[2] * 10 + digits[3]; |
| 1960 | break; |
| 1961 | case 5: // Hmmss |
| 1962 | hour = digits[0]; |
| 1963 | min = digits[1] * 10 + digits[2]; |
| 1964 | sec = digits[3] * 10 + digits[4]; |
| 1965 | break; |
| 1966 | case 6: // HHmmss |
| 1967 | hour = digits[0] * 10 + digits[1]; |
| 1968 | min = digits[2] * 10 + digits[3]; |
| 1969 | sec = digits[4] * 10 + digits[5]; |
| 1970 | break; |
| 1971 | } |
| 1972 | if (hour <= MAX_OFFSET_HOUR && min <= MAX_OFFSET_MINUTE && sec <= MAX_OFFSET_SECOND) { |
| 1973 | // found a valid combination |
| 1974 | offset = hour * MILLIS_PER_HOUR + min * MILLIS_PER_MINUTE + sec * MILLIS_PER_SECOND; |
| 1975 | parsedLen = parsed[numDigits - 1]; |
| 1976 | break; |
| 1977 | } |
| 1978 | numDigits--; |
| 1979 | } |
| 1980 | return offset; |
| 1981 | } |
| 1982 | |
| 1983 | int32_t |
| 1984 | TimeZoneFormat::parseOffsetDefaultLocalizedGMT(const UnicodeString& text, int start, int32_t& parsedLen) const { |
| 1985 | int32_t idx = start; |
| 1986 | int32_t offset = 0; |
| 1987 | int32_t parsed = 0; |
| 1988 | |
| 1989 | do { |
| 1990 | // check global default GMT alternatives |
| 1991 | int32_t gmtLen = 0; |
| 1992 | |
| 1993 | for (int32_t i = 0; ALT_GMT_STRINGS[i][0] != 0; i++) { |
| 1994 | const char16_t* gmt = ALT_GMT_STRINGS[i]; |
| 1995 | int32_t len = u_strlen(gmt); |
| 1996 | if (text.caseCompare(start, len, gmt, 0) == 0) { |
| 1997 | gmtLen = len; |
| 1998 | break; |
| 1999 | } |
| 2000 | } |
| 2001 | if (gmtLen == 0) { |
| 2002 | break; |
| 2003 | } |
| 2004 | idx += gmtLen; |
| 2005 | |
| 2006 | // offset needs a sign char and a digit at minimum |
| 2007 | if (idx + 1 >= text.length()) { |
| 2008 | break; |
| 2009 | } |
| 2010 | |
| 2011 | // parse sign |
| 2012 | int32_t sign = 1; |
| 2013 | char16_t c = text.charAt(idx); |
| 2014 | if (c == PLUS) { |
| 2015 | sign = 1; |
| 2016 | } else if (c == MINUS) { |
| 2017 | sign = -1; |
| 2018 | } else { |
| 2019 | break; |
| 2020 | } |
| 2021 | idx++; |
| 2022 | |
| 2023 | // offset part |
| 2024 | // try the default pattern with the separator first |
| 2025 | int32_t lenWithSep = 0; |
| 2026 | int32_t offsetWithSep = parseDefaultOffsetFields(text, idx, DEFAULT_GMT_OFFSET_SEP, lenWithSep); |
| 2027 | if (lenWithSep == text.length() - idx) { |
| 2028 | // maximum match |
| 2029 | offset = offsetWithSep * sign; |
| 2030 | idx += lenWithSep; |
| 2031 | } else { |
| 2032 | // try abutting field pattern |
| 2033 | int32_t lenAbut = 0; |
| 2034 | int32_t offsetAbut = parseAbuttingOffsetFields(text, idx, lenAbut); |
| 2035 | |
| 2036 | if (lenWithSep > lenAbut) { |
| 2037 | offset = offsetWithSep * sign; |
| 2038 | idx += lenWithSep; |
| 2039 | } else { |
| 2040 | offset = offsetAbut * sign; |
| 2041 | idx += lenAbut; |
| 2042 | } |
| 2043 | } |
| 2044 | parsed = idx - start; |
| 2045 | } while (false); |
| 2046 | |
| 2047 | parsedLen = parsed; |
| 2048 | return offset; |
| 2049 | } |
| 2050 | |
| 2051 | int32_t |
| 2052 | TimeZoneFormat::parseDefaultOffsetFields(const UnicodeString& text, int32_t start, char16_t separator, int32_t& parsedLen) const { |
| 2053 | int32_t max = text.length(); |
| 2054 | int32_t idx = start; |
| 2055 | int32_t len = 0; |
| 2056 | int32_t hour = 0, min = 0, sec = 0; |
| 2057 | |
| 2058 | parsedLen = 0; |
| 2059 | |
| 2060 | do { |
| 2061 | hour = parseOffsetFieldWithLocalizedDigits(text, idx, 1, 2, 0, MAX_OFFSET_HOUR, len); |
| 2062 | if (len == 0) { |
| 2063 | break; |
| 2064 | } |
| 2065 | idx += len; |
| 2066 | |
| 2067 | if (idx + 1 < max && text.charAt(idx) == separator) { |
| 2068 | min = parseOffsetFieldWithLocalizedDigits(text, idx + 1, 2, 2, 0, MAX_OFFSET_MINUTE, len); |
| 2069 | if (len == 0) { |
| 2070 | break; |
| 2071 | } |
| 2072 | idx += (1 + len); |
| 2073 | |
| 2074 | if (idx + 1 < max && text.charAt(idx) == separator) { |
| 2075 | sec = parseOffsetFieldWithLocalizedDigits(text, idx + 1, 2, 2, 0, MAX_OFFSET_SECOND, len); |
| 2076 | if (len == 0) { |
| 2077 | break; |
| 2078 | } |
| 2079 | idx += (1 + len); |
| 2080 | } |
| 2081 | } |
| 2082 | } while (false); |
| 2083 | |
| 2084 | if (idx == start) { |
| 2085 | return 0; |
| 2086 | } |
| 2087 | |
| 2088 | parsedLen = idx - start; |
| 2089 | return hour * MILLIS_PER_HOUR + min * MILLIS_PER_MINUTE + sec * MILLIS_PER_SECOND; |
| 2090 | } |
| 2091 | |
| 2092 | int32_t |
| 2093 | TimeZoneFormat::parseOffsetFieldWithLocalizedDigits(const UnicodeString& text, int32_t start, uint8_t minDigits, uint8_t maxDigits, uint16_t minVal, uint16_t maxVal, int32_t& parsedLen) const { |
| 2094 | parsedLen = 0; |
| 2095 | |
| 2096 | int32_t decVal = 0; |
| 2097 | int32_t numDigits = 0; |
| 2098 | int32_t idx = start; |
| 2099 | int32_t digitLen = 0; |
| 2100 | |
| 2101 | while (idx < text.length() && numDigits < maxDigits) { |
| 2102 | int32_t digit = parseSingleLocalizedDigit(text, idx, digitLen); |
| 2103 | if (digit < 0) { |
| 2104 | break; |
| 2105 | } |
| 2106 | int32_t tmpVal = decVal * 10 + digit; |
| 2107 | if (tmpVal > maxVal) { |
| 2108 | break; |
| 2109 | } |
| 2110 | decVal = tmpVal; |
| 2111 | numDigits++; |
| 2112 | idx += digitLen; |
| 2113 | } |
| 2114 | |
| 2115 | // Note: maxVal is checked in the while loop |
| 2116 | if (numDigits < minDigits || decVal < minVal) { |
| 2117 | decVal = -1; |
| 2118 | numDigits = 0; |
Value stored to 'numDigits' is never read | |
| 2119 | } else { |
| 2120 | parsedLen = idx - start; |
| 2121 | } |
| 2122 | |
| 2123 | return decVal; |
| 2124 | } |
| 2125 | |
| 2126 | int32_t |
| 2127 | TimeZoneFormat::parseSingleLocalizedDigit(const UnicodeString& text, int32_t start, int32_t& len) const { |
| 2128 | int32_t digit = -1; |
| 2129 | len = 0; |
| 2130 | if (start < text.length()) { |
| 2131 | UChar32 cp = text.char32At(start); |
| 2132 | |
| 2133 | // First, try digits configured for this instance |
| 2134 | for (int32_t i = 0; i < 10; i++) { |
| 2135 | if (cp == fGMTOffsetDigits[i]) { |
| 2136 | digit = i; |
| 2137 | break; |
| 2138 | } |
| 2139 | } |
| 2140 | // If failed, check if this is a Unicode digit |
| 2141 | if (digit < 0) { |
| 2142 | int32_t tmp = u_charDigitValue(cp); |
| 2143 | digit = (tmp >= 0 && tmp <= 9) ? tmp : -1; |
| 2144 | } |
| 2145 | |
| 2146 | if (digit >= 0) { |
| 2147 | int32_t next = text.moveIndex32(start, 1); |
| 2148 | len = next - start; |
| 2149 | } |
| 2150 | } |
| 2151 | return digit; |
| 2152 | } |
| 2153 | |
| 2154 | UnicodeString& |
| 2155 | TimeZoneFormat::formatOffsetWithAsciiDigits(int32_t offset, char16_t sep, OffsetFields minFields, OffsetFields maxFields, UnicodeString& result) { |
| 2156 | U_ASSERT(maxFields >= minFields)(static_cast <bool> (maxFields >= minFields) ? void ( 0) : __assert_fail ("maxFields >= minFields", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 2157 | U_ASSERT(offset > -MAX_OFFSET && offset < MAX_OFFSET)(static_cast <bool> (offset > -MAX_OFFSET && offset < MAX_OFFSET) ? void (0) : __assert_fail ("offset > -MAX_OFFSET && offset < MAX_OFFSET" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 2158 | |
| 2159 | char16_t sign = PLUS; |
| 2160 | if (offset < 0) { |
| 2161 | sign = MINUS; |
| 2162 | offset = -offset; |
| 2163 | } |
| 2164 | result.setTo(sign); |
| 2165 | |
| 2166 | int fields[3]; |
| 2167 | fields[0] = offset / MILLIS_PER_HOUR; |
| 2168 | offset = offset % MILLIS_PER_HOUR; |
| 2169 | fields[1] = offset / MILLIS_PER_MINUTE; |
| 2170 | offset = offset % MILLIS_PER_MINUTE; |
| 2171 | fields[2] = offset / MILLIS_PER_SECOND; |
| 2172 | |
| 2173 | U_ASSERT(fields[0] >= 0 && fields[0] <= MAX_OFFSET_HOUR)(static_cast <bool> (fields[0] >= 0 && fields [0] <= MAX_OFFSET_HOUR) ? void (0) : __assert_fail ("fields[0] >= 0 && fields[0] <= MAX_OFFSET_HOUR" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 2174 | U_ASSERT(fields[1] >= 0 && fields[1] <= MAX_OFFSET_MINUTE)(static_cast <bool> (fields[1] >= 0 && fields [1] <= MAX_OFFSET_MINUTE) ? void (0) : __assert_fail ("fields[1] >= 0 && fields[1] <= MAX_OFFSET_MINUTE" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 2175 | U_ASSERT(fields[2] >= 0 && fields[2] <= MAX_OFFSET_SECOND)(static_cast <bool> (fields[2] >= 0 && fields [2] <= MAX_OFFSET_SECOND) ? void (0) : __assert_fail ("fields[2] >= 0 && fields[2] <= MAX_OFFSET_SECOND" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 2176 | |
| 2177 | int32_t lastIdx = maxFields; |
| 2178 | while (lastIdx > minFields) { |
| 2179 | if (fields[lastIdx] != 0) { |
| 2180 | break; |
| 2181 | } |
| 2182 | lastIdx--; |
| 2183 | } |
| 2184 | |
| 2185 | for (int32_t idx = 0; idx <= lastIdx; idx++) { |
| 2186 | if (sep && idx != 0) { |
| 2187 | result.append(sep); |
| 2188 | } |
| 2189 | result.append(static_cast<char16_t>(0x0030 + fields[idx] / 10)); |
| 2190 | result.append(static_cast<char16_t>(0x0030 + fields[idx] % 10)); |
| 2191 | } |
| 2192 | |
| 2193 | return result; |
| 2194 | } |
| 2195 | |
| 2196 | int32_t |
| 2197 | TimeZoneFormat::parseAbuttingAsciiOffsetFields(const UnicodeString& text, ParsePosition& pos, OffsetFields minFields, OffsetFields maxFields, UBool fixedHourWidth) { |
| 2198 | int32_t start = pos.getIndex(); |
| 2199 | |
| 2200 | int32_t minDigits = 2 * (minFields + 1) - (fixedHourWidth ? 0 : 1); |
| 2201 | int32_t maxDigits = 2 * (maxFields + 1); |
| 2202 | |
| 2203 | U_ASSERT(maxDigits <= MAX_OFFSET_DIGITS)(static_cast <bool> (maxDigits <= 6) ? void (0) : __assert_fail ("maxDigits <= 6", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 2204 | |
| 2205 | int32_t digits[MAX_OFFSET_DIGITS6] = {}; |
| 2206 | int32_t numDigits = 0; |
| 2207 | int32_t idx = start; |
| 2208 | while (numDigits < maxDigits && idx < text.length()) { |
| 2209 | char16_t uch = text.charAt(idx); |
| 2210 | int32_t digit = DIGIT_VAL(uch)(0x0030 <= (uch) && (uch) <= 0x0039 ? (uch) - 0x0030 : -1); |
| 2211 | if (digit < 0) { |
| 2212 | break; |
| 2213 | } |
| 2214 | digits[numDigits] = digit; |
| 2215 | numDigits++; |
| 2216 | idx++; |
| 2217 | } |
| 2218 | |
| 2219 | if (fixedHourWidth && (numDigits & 1)) { |
| 2220 | // Fixed digits, so the number of digits must be even number. Truncating. |
| 2221 | numDigits--; |
| 2222 | } |
| 2223 | |
| 2224 | if (numDigits < minDigits) { |
| 2225 | pos.setErrorIndex(start); |
| 2226 | return 0; |
| 2227 | } |
| 2228 | |
| 2229 | int32_t hour = 0, min = 0, sec = 0; |
| 2230 | UBool bParsed = false; |
| 2231 | while (numDigits >= minDigits) { |
| 2232 | switch (numDigits) { |
| 2233 | case 1: //H |
| 2234 | hour = digits[0]; |
| 2235 | break; |
| 2236 | case 2: //HH |
| 2237 | hour = digits[0] * 10 + digits[1]; |
| 2238 | break; |
| 2239 | case 3: //Hmm |
| 2240 | hour = digits[0]; |
| 2241 | min = digits[1] * 10 + digits[2]; |
| 2242 | break; |
| 2243 | case 4: //HHmm |
| 2244 | hour = digits[0] * 10 + digits[1]; |
| 2245 | min = digits[2] * 10 + digits[3]; |
| 2246 | break; |
| 2247 | case 5: //Hmmss |
| 2248 | hour = digits[0]; |
| 2249 | min = digits[1] * 10 + digits[2]; |
| 2250 | sec = digits[3] * 10 + digits[4]; |
| 2251 | break; |
| 2252 | case 6: //HHmmss |
| 2253 | hour = digits[0] * 10 + digits[1]; |
| 2254 | min = digits[2] * 10 + digits[3]; |
| 2255 | sec = digits[4] * 10 + digits[5]; |
| 2256 | break; |
| 2257 | } |
| 2258 | |
| 2259 | if (hour <= MAX_OFFSET_HOUR && min <= MAX_OFFSET_MINUTE && sec <= MAX_OFFSET_SECOND) { |
| 2260 | // Successfully parsed |
| 2261 | bParsed = true; |
| 2262 | break; |
| 2263 | } |
| 2264 | |
| 2265 | // Truncating |
| 2266 | numDigits -= (fixedHourWidth ? 2 : 1); |
| 2267 | hour = min = sec = 0; |
| 2268 | } |
| 2269 | |
| 2270 | if (!bParsed) { |
| 2271 | pos.setErrorIndex(start); |
| 2272 | return 0; |
| 2273 | } |
| 2274 | pos.setIndex(start + numDigits); |
| 2275 | return ((((hour * 60) + min) * 60) + sec) * 1000; |
| 2276 | } |
| 2277 | |
| 2278 | int32_t |
| 2279 | TimeZoneFormat::parseAsciiOffsetFields(const UnicodeString& text, ParsePosition& pos, char16_t sep, OffsetFields minFields, OffsetFields maxFields) { |
| 2280 | int32_t start = pos.getIndex(); |
| 2281 | int32_t fieldVal[] = {0, 0, 0}; |
| 2282 | int32_t fieldLen[] = {0, -1, -1}; |
| 2283 | for (int32_t idx = start, fieldIdx = 0; idx < text.length() && fieldIdx <= maxFields; idx++) { |
| 2284 | char16_t c = text.charAt(idx); |
| 2285 | if (c == sep) { |
| 2286 | if (fieldIdx == 0) { |
| 2287 | if (fieldLen[0] == 0) { |
| 2288 | // no hours field |
| 2289 | break; |
| 2290 | } |
| 2291 | // 1 digit hour, move to next field |
| 2292 | } else { |
| 2293 | if (fieldLen[fieldIdx] != -1) { |
| 2294 | // premature minute or seconds field |
| 2295 | break; |
| 2296 | } |
| 2297 | fieldLen[fieldIdx] = 0; |
| 2298 | } |
| 2299 | continue; |
| 2300 | } else if (fieldLen[fieldIdx] == -1) { |
| 2301 | // no separator after 2 digit field |
| 2302 | break; |
| 2303 | } |
| 2304 | int32_t digit = DIGIT_VAL(c)(0x0030 <= (c) && (c) <= 0x0039 ? (c) - 0x0030 : -1); |
| 2305 | if (digit < 0) { |
| 2306 | // not a digit |
| 2307 | break; |
| 2308 | } |
| 2309 | fieldVal[fieldIdx] = fieldVal[fieldIdx] * 10 + digit; |
| 2310 | fieldLen[fieldIdx]++; |
| 2311 | if (fieldLen[fieldIdx] >= 2) { |
| 2312 | // parsed 2 digits, move to next field |
| 2313 | fieldIdx++; |
| 2314 | } |
| 2315 | } |
| 2316 | |
| 2317 | int32_t offset = 0; |
| 2318 | int32_t parsedLen = 0; |
| 2319 | int32_t parsedFields = -1; |
| 2320 | do { |
| 2321 | // hour |
| 2322 | if (fieldLen[0] == 0) { |
| 2323 | break; |
| 2324 | } |
| 2325 | if (fieldVal[0] > MAX_OFFSET_HOUR) { |
| 2326 | offset = (fieldVal[0] / 10) * MILLIS_PER_HOUR; |
| 2327 | parsedFields = FIELDS_H; |
| 2328 | parsedLen = 1; |
| 2329 | break; |
| 2330 | } |
| 2331 | offset = fieldVal[0] * MILLIS_PER_HOUR; |
| 2332 | parsedLen = fieldLen[0]; |
| 2333 | parsedFields = FIELDS_H; |
| 2334 | |
| 2335 | // minute |
| 2336 | if (fieldLen[1] != 2 || fieldVal[1] > MAX_OFFSET_MINUTE) { |
| 2337 | break; |
| 2338 | } |
| 2339 | offset += fieldVal[1] * MILLIS_PER_MINUTE; |
| 2340 | parsedLen += (1 + fieldLen[1]); |
| 2341 | parsedFields = FIELDS_HM; |
| 2342 | |
| 2343 | // second |
| 2344 | if (fieldLen[2] != 2 || fieldVal[2] > MAX_OFFSET_SECOND) { |
| 2345 | break; |
| 2346 | } |
| 2347 | offset += fieldVal[2] * MILLIS_PER_SECOND; |
| 2348 | parsedLen += (1 + fieldLen[2]); |
| 2349 | parsedFields = FIELDS_HMS; |
| 2350 | } while (false); |
| 2351 | |
| 2352 | if (parsedFields < minFields) { |
| 2353 | pos.setErrorIndex(start); |
| 2354 | return 0; |
| 2355 | } |
| 2356 | |
| 2357 | pos.setIndex(start + parsedLen); |
| 2358 | return offset; |
| 2359 | } |
| 2360 | |
| 2361 | void |
| 2362 | TimeZoneFormat::appendOffsetDigits(UnicodeString& buf, int32_t n, uint8_t minDigits) const { |
| 2363 | U_ASSERT(n >= 0 && n < 60)(static_cast <bool> (n >= 0 && n < 60) ? void (0) : __assert_fail ("n >= 0 && n < 60", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 2364 | int32_t numDigits = n >= 10 ? 2 : 1; |
| 2365 | for (int32_t i = 0; i < minDigits - numDigits; i++) { |
| 2366 | buf.append(fGMTOffsetDigits[0]); |
| 2367 | } |
| 2368 | if (numDigits == 2) { |
| 2369 | buf.append(fGMTOffsetDigits[n / 10]); |
| 2370 | } |
| 2371 | buf.append(fGMTOffsetDigits[n % 10]); |
| 2372 | } |
| 2373 | |
| 2374 | // ------------------------------------------------------------------ |
| 2375 | // Private misc |
| 2376 | void |
| 2377 | TimeZoneFormat::initGMTPattern(const UnicodeString& gmtPattern, UErrorCode& status) { |
| 2378 | if (U_FAILURE(status)) { |
| 2379 | return; |
| 2380 | } |
| 2381 | // This implementation not perfect, but sufficient practically. |
| 2382 | int32_t idx = gmtPattern.indexOf(ARG0, ARG0_LEN, 0); |
| 2383 | if (idx < 0) { |
| 2384 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2385 | return; |
| 2386 | } |
| 2387 | fGMTPattern.setTo(gmtPattern); |
| 2388 | unquote(gmtPattern.tempSubString(0, idx), fGMTPatternPrefix); |
| 2389 | unquote(gmtPattern.tempSubString(idx + ARG0_LEN), fGMTPatternSuffix); |
| 2390 | } |
| 2391 | |
| 2392 | UnicodeString& |
| 2393 | TimeZoneFormat::unquote(const UnicodeString& pattern, UnicodeString& result) { |
| 2394 | if (pattern.indexOf(SINGLEQUOTE) < 0) { |
| 2395 | result.setTo(pattern); |
| 2396 | return result; |
| 2397 | } |
| 2398 | result.remove(); |
| 2399 | UBool isPrevQuote = false; |
| 2400 | UBool inQuote = false; |
| 2401 | for (int32_t i = 0; i < pattern.length(); i++) { |
| 2402 | char16_t c = pattern.charAt(i); |
| 2403 | if (c == SINGLEQUOTE) { |
| 2404 | if (isPrevQuote) { |
| 2405 | result.append(c); |
| 2406 | isPrevQuote = false; |
| 2407 | } else { |
| 2408 | isPrevQuote = true; |
| 2409 | } |
| 2410 | inQuote = !inQuote; |
| 2411 | } else { |
| 2412 | isPrevQuote = false; |
| 2413 | result.append(c); |
| 2414 | } |
| 2415 | } |
| 2416 | return result; |
| 2417 | } |
| 2418 | |
| 2419 | UVector* |
| 2420 | TimeZoneFormat::parseOffsetPattern(const UnicodeString& pattern, OffsetFields required, UErrorCode& status) { |
| 2421 | if (U_FAILURE(status)) { |
| 2422 | return nullptr; |
| 2423 | } |
| 2424 | UVector* result = new UVector(deleteGMTOffsetField, nullptr, status); |
| 2425 | if (result == nullptr) { |
| 2426 | status = U_MEMORY_ALLOCATION_ERROR; |
| 2427 | return nullptr; |
| 2428 | } |
| 2429 | |
| 2430 | int32_t checkBits = 0; |
| 2431 | UBool isPrevQuote = false; |
| 2432 | UBool inQuote = false; |
| 2433 | char16_t textBuf[32]; |
| 2434 | UnicodeString text(textBuf, 0, UPRV_LENGTHOF(textBuf)(int32_t)(sizeof(textBuf)/sizeof((textBuf)[0]))); |
| 2435 | GMTOffsetField::FieldType itemType = GMTOffsetField::TEXT; |
| 2436 | int32_t itemLength = 1; |
| 2437 | |
| 2438 | for (int32_t i = 0; i < pattern.length(); i++) { |
| 2439 | char16_t ch = pattern.charAt(i); |
| 2440 | if (ch == SINGLEQUOTE) { |
| 2441 | if (isPrevQuote) { |
| 2442 | text.append(SINGLEQUOTE); |
| 2443 | isPrevQuote = false; |
| 2444 | } else { |
| 2445 | isPrevQuote = true; |
| 2446 | if (itemType != GMTOffsetField::TEXT) { |
| 2447 | if (GMTOffsetField::isValid(itemType, itemLength)) { |
| 2448 | GMTOffsetField* fld = GMTOffsetField::createTimeField(itemType, static_cast<uint8_t>(itemLength), status); |
| 2449 | result->adoptElement(fld, status); |
| 2450 | if (U_FAILURE(status)) { |
| 2451 | break; |
| 2452 | } |
| 2453 | } else { |
| 2454 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2455 | break; |
| 2456 | } |
| 2457 | itemType = GMTOffsetField::TEXT; |
| 2458 | } |
| 2459 | } |
| 2460 | inQuote = !inQuote; |
| 2461 | } else { |
| 2462 | isPrevQuote = false; |
| 2463 | if (inQuote) { |
| 2464 | text.append(ch); |
| 2465 | } else { |
| 2466 | GMTOffsetField::FieldType tmpType = GMTOffsetField::getTypeByLetter(ch); |
| 2467 | if (tmpType != GMTOffsetField::TEXT) { |
| 2468 | // an offset time pattern character |
| 2469 | if (tmpType == itemType) { |
| 2470 | itemLength++; |
| 2471 | } else { |
| 2472 | if (itemType == GMTOffsetField::TEXT) { |
| 2473 | if (text.length() > 0) { |
| 2474 | GMTOffsetField* textfld = GMTOffsetField::createText(text, status); |
| 2475 | result->adoptElement(textfld, status); |
| 2476 | if (U_FAILURE(status)) { |
| 2477 | break; |
| 2478 | } |
| 2479 | text.remove(); |
| 2480 | } |
| 2481 | } else { |
| 2482 | if (GMTOffsetField::isValid(itemType, itemLength)) { |
| 2483 | GMTOffsetField* fld = GMTOffsetField::createTimeField(itemType, static_cast<uint8_t>(itemLength), status); |
| 2484 | result->adoptElement(fld, status); |
| 2485 | if (U_FAILURE(status)) { |
| 2486 | break; |
| 2487 | } |
| 2488 | } else { |
| 2489 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2490 | break; |
| 2491 | } |
| 2492 | } |
| 2493 | itemType = tmpType; |
| 2494 | itemLength = 1; |
| 2495 | checkBits |= tmpType; |
| 2496 | } |
| 2497 | } else { |
| 2498 | // a string literal |
| 2499 | if (itemType != GMTOffsetField::TEXT) { |
| 2500 | if (GMTOffsetField::isValid(itemType, itemLength)) { |
| 2501 | GMTOffsetField* fld = GMTOffsetField::createTimeField(itemType, static_cast<uint8_t>(itemLength), status); |
| 2502 | result->adoptElement(fld, status); |
| 2503 | if (U_FAILURE(status)) { |
| 2504 | break; |
| 2505 | } |
| 2506 | } else { |
| 2507 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2508 | break; |
| 2509 | } |
| 2510 | itemType = GMTOffsetField::TEXT; |
| 2511 | } |
| 2512 | text.append(ch); |
| 2513 | } |
| 2514 | } |
| 2515 | } |
| 2516 | } |
| 2517 | // handle last item |
| 2518 | if (U_SUCCESS(status)) { |
| 2519 | if (itemType == GMTOffsetField::TEXT) { |
| 2520 | if (text.length() > 0) { |
| 2521 | GMTOffsetField* tfld = GMTOffsetField::createText(text, status); |
| 2522 | result->adoptElement(tfld, status); |
| 2523 | } |
| 2524 | } else { |
| 2525 | if (GMTOffsetField::isValid(itemType, itemLength)) { |
| 2526 | GMTOffsetField* fld = GMTOffsetField::createTimeField(itemType, static_cast<uint8_t>(itemLength), status); |
| 2527 | result->adoptElement(fld, status); |
| 2528 | } else { |
| 2529 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2530 | } |
| 2531 | } |
| 2532 | |
| 2533 | // Check all required fields are set |
| 2534 | if (U_SUCCESS(status)) { |
| 2535 | int32_t reqBits = 0; |
| 2536 | switch (required) { |
| 2537 | case FIELDS_H: |
| 2538 | reqBits = GMTOffsetField::HOUR; |
| 2539 | break; |
| 2540 | case FIELDS_HM: |
| 2541 | reqBits = GMTOffsetField::HOUR | GMTOffsetField::MINUTE; |
| 2542 | break; |
| 2543 | case FIELDS_HMS: |
| 2544 | reqBits = GMTOffsetField::HOUR | GMTOffsetField::MINUTE | GMTOffsetField::SECOND; |
| 2545 | break; |
| 2546 | } |
| 2547 | if (checkBits == reqBits) { |
| 2548 | // all required fields are set, no extra fields |
| 2549 | return result; |
| 2550 | } |
| 2551 | } |
| 2552 | } |
| 2553 | |
| 2554 | // error |
| 2555 | delete result; |
| 2556 | return nullptr; |
| 2557 | } |
| 2558 | |
| 2559 | UnicodeString& |
| 2560 | TimeZoneFormat::expandOffsetPattern(const UnicodeString& offsetHM, UnicodeString& result, UErrorCode& status) { |
| 2561 | result.setToBogus(); |
| 2562 | if (U_FAILURE(status)) { |
| 2563 | return result; |
| 2564 | } |
| 2565 | U_ASSERT(u_strlen(DEFAULT_GMT_OFFSET_MINUTE_PATTERN) == 2)(static_cast <bool> (u_strlen(DEFAULT_GMT_OFFSET_MINUTE_PATTERN ) == 2) ? void (0) : __assert_fail ("u_strlen(DEFAULT_GMT_OFFSET_MINUTE_PATTERN) == 2" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 2566 | |
| 2567 | int32_t idx_mm = offsetHM.indexOf(DEFAULT_GMT_OFFSET_MINUTE_PATTERN, 2, 0); |
| 2568 | if (idx_mm < 0) { |
| 2569 | // Bad time zone hour pattern data |
| 2570 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2571 | return result; |
| 2572 | } |
| 2573 | |
| 2574 | UnicodeString sep; |
| 2575 | int32_t idx_H = offsetHM.tempSubString(0, idx_mm).lastIndexOf(static_cast<char16_t>(0x0048) /* H */); |
| 2576 | if (idx_H >= 0) { |
| 2577 | sep = offsetHM.tempSubString(idx_H + 1, idx_mm - (idx_H + 1)); |
| 2578 | } |
| 2579 | result.setTo(offsetHM.tempSubString(0, idx_mm + 2)); |
| 2580 | result.append(sep); |
| 2581 | result.append(DEFAULT_GMT_OFFSET_SECOND_PATTERN, -1); |
| 2582 | result.append(offsetHM.tempSubString(idx_mm + 2)); |
| 2583 | return result; |
| 2584 | } |
| 2585 | |
| 2586 | UnicodeString& |
| 2587 | TimeZoneFormat::truncateOffsetPattern(const UnicodeString& offsetHM, UnicodeString& result, UErrorCode& status) { |
| 2588 | result.setToBogus(); |
| 2589 | if (U_FAILURE(status)) { |
| 2590 | return result; |
| 2591 | } |
| 2592 | U_ASSERT(u_strlen(DEFAULT_GMT_OFFSET_MINUTE_PATTERN) == 2)(static_cast <bool> (u_strlen(DEFAULT_GMT_OFFSET_MINUTE_PATTERN ) == 2) ? void (0) : __assert_fail ("u_strlen(DEFAULT_GMT_OFFSET_MINUTE_PATTERN) == 2" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 2593 | |
| 2594 | int32_t idx_mm = offsetHM.indexOf(DEFAULT_GMT_OFFSET_MINUTE_PATTERN, 2, 0); |
| 2595 | if (idx_mm < 0) { |
| 2596 | // Bad time zone hour pattern data |
| 2597 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2598 | return result; |
| 2599 | } |
| 2600 | char16_t HH[] = {0x0048, 0x0048}; |
| 2601 | int32_t idx_HH = offsetHM.tempSubString(0, idx_mm).lastIndexOf(HH, 2, 0); |
| 2602 | if (idx_HH >= 0) { |
| 2603 | return result.setTo(offsetHM.tempSubString(0, idx_HH + 2)); |
| 2604 | } |
| 2605 | int32_t idx_H = offsetHM.tempSubString(0, idx_mm).lastIndexOf(static_cast<char16_t>(0x0048), 0); |
| 2606 | if (idx_H >= 0) { |
| 2607 | return result.setTo(offsetHM.tempSubString(0, idx_H + 1)); |
| 2608 | } |
| 2609 | // Bad time zone hour pattern data |
| 2610 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2611 | return result; |
| 2612 | } |
| 2613 | |
| 2614 | void |
| 2615 | TimeZoneFormat::initGMTOffsetPatterns(UErrorCode& status) { |
| 2616 | for (int32_t type = 0; type < UTZFMT_PAT_COUNT; type++) { |
| 2617 | switch (type) { |
| 2618 | case UTZFMT_PAT_POSITIVE_H: |
| 2619 | case UTZFMT_PAT_NEGATIVE_H: |
| 2620 | fGMTOffsetPatternItems[type] = parseOffsetPattern(fGMTOffsetPatterns[type], FIELDS_H, status); |
| 2621 | break; |
| 2622 | case UTZFMT_PAT_POSITIVE_HM: |
| 2623 | case UTZFMT_PAT_NEGATIVE_HM: |
| 2624 | fGMTOffsetPatternItems[type] = parseOffsetPattern(fGMTOffsetPatterns[type], FIELDS_HM, status); |
| 2625 | break; |
| 2626 | case UTZFMT_PAT_POSITIVE_HMS: |
| 2627 | case UTZFMT_PAT_NEGATIVE_HMS: |
| 2628 | fGMTOffsetPatternItems[type] = parseOffsetPattern(fGMTOffsetPatterns[type], FIELDS_HMS, status); |
| 2629 | break; |
| 2630 | } |
| 2631 | } |
| 2632 | if (U_FAILURE(status)) { |
| 2633 | return; |
| 2634 | } |
| 2635 | checkAbuttingHoursAndMinutes(); |
| 2636 | } |
| 2637 | |
| 2638 | void |
| 2639 | TimeZoneFormat::checkAbuttingHoursAndMinutes() { |
| 2640 | fAbuttingOffsetHoursAndMinutes= false; |
| 2641 | for (int32_t type = 0; type < UTZFMT_PAT_COUNT; type++) { |
| 2642 | UBool afterH = false; |
| 2643 | UVector *items = fGMTOffsetPatternItems[type]; |
| 2644 | for (int32_t i = 0; i < items->size(); i++) { |
| 2645 | const GMTOffsetField* item = static_cast<GMTOffsetField*>(items->elementAt(i)); |
| 2646 | GMTOffsetField::FieldType fieldType = item->getType(); |
| 2647 | if (fieldType != GMTOffsetField::TEXT) { |
| 2648 | if (afterH) { |
| 2649 | fAbuttingOffsetHoursAndMinutes = true; |
| 2650 | break; |
| 2651 | } else if (fieldType == GMTOffsetField::HOUR) { |
| 2652 | afterH = true; |
| 2653 | } |
| 2654 | } else if (afterH) { |
| 2655 | break; |
| 2656 | } |
| 2657 | } |
| 2658 | if (fAbuttingOffsetHoursAndMinutes) { |
| 2659 | break; |
| 2660 | } |
| 2661 | } |
| 2662 | } |
| 2663 | |
| 2664 | UBool |
| 2665 | TimeZoneFormat::toCodePoints(const UnicodeString& str, UChar32* codeArray, int32_t size) { |
| 2666 | int32_t count = str.countChar32(); |
| 2667 | if (count != size) { |
| 2668 | return false; |
| 2669 | } |
| 2670 | |
| 2671 | for (int32_t idx = 0, start = 0; idx < size; idx++) { |
| 2672 | codeArray[idx] = str.char32At(start); |
| 2673 | start = str.moveIndex32(start, 1); |
| 2674 | } |
| 2675 | |
| 2676 | return true; |
| 2677 | } |
| 2678 | |
| 2679 | TimeZone* |
| 2680 | TimeZoneFormat::createTimeZoneForOffset(int32_t offset) const { |
| 2681 | if (offset == 0) { |
| 2682 | // when offset is 0, we should use "Etc/GMT" |
| 2683 | return TimeZone::createTimeZone(UnicodeString(true, TZID_GMT, -1)); |
| 2684 | } |
| 2685 | return ZoneMeta::createCustomTimeZone(offset); |
| 2686 | } |
| 2687 | |
| 2688 | UTimeZoneFormatTimeType |
| 2689 | TimeZoneFormat::getTimeType(UTimeZoneNameType nameType) { |
| 2690 | switch (nameType) { |
| 2691 | case UTZNM_LONG_STANDARD: |
| 2692 | case UTZNM_SHORT_STANDARD: |
| 2693 | return UTZFMT_TIME_TYPE_STANDARD; |
| 2694 | |
| 2695 | case UTZNM_LONG_DAYLIGHT: |
| 2696 | case UTZNM_SHORT_DAYLIGHT: |
| 2697 | return UTZFMT_TIME_TYPE_DAYLIGHT; |
| 2698 | |
| 2699 | default: |
| 2700 | return UTZFMT_TIME_TYPE_UNKNOWN; |
| 2701 | } |
| 2702 | } |
| 2703 | |
| 2704 | UnicodeString& |
| 2705 | TimeZoneFormat::getTimeZoneID(const TimeZoneNames::MatchInfoCollection* matches, int32_t idx, UnicodeString& tzID) const { |
| 2706 | if (!matches->getTimeZoneIDAt(idx, tzID)) { |
| 2707 | char16_t mzIDBuf[32]; |
| 2708 | UnicodeString mzID(mzIDBuf, 0, UPRV_LENGTHOF(mzIDBuf)(int32_t)(sizeof(mzIDBuf)/sizeof((mzIDBuf)[0]))); |
| 2709 | if (matches->getMetaZoneIDAt(idx, mzID)) { |
| 2710 | fTimeZoneNames->getReferenceZoneID(mzID, fTargetRegion, tzID); |
| 2711 | } |
| 2712 | } |
| 2713 | return tzID; |
| 2714 | } |
| 2715 | |
| 2716 | |
| 2717 | class ZoneIdMatchHandler : public TextTrieMapSearchResultHandler { |
| 2718 | public: |
| 2719 | ZoneIdMatchHandler(); |
| 2720 | virtual ~ZoneIdMatchHandler(); |
| 2721 | |
| 2722 | UBool handleMatch(int32_t matchLength, const CharacterNode *node, UErrorCode &status) override; |
| 2723 | const char16_t* getID(); |
| 2724 | int32_t getMatchLen(); |
| 2725 | private: |
| 2726 | int32_t fLen; |
| 2727 | const char16_t* fID; |
| 2728 | }; |
| 2729 | |
| 2730 | ZoneIdMatchHandler::ZoneIdMatchHandler() |
| 2731 | : fLen(0), fID(nullptr) { |
| 2732 | } |
| 2733 | |
| 2734 | ZoneIdMatchHandler::~ZoneIdMatchHandler() { |
| 2735 | } |
| 2736 | |
| 2737 | UBool |
| 2738 | ZoneIdMatchHandler::handleMatch(int32_t matchLength, const CharacterNode *node, UErrorCode &status) { |
| 2739 | if (U_FAILURE(status)) { |
| 2740 | return false; |
| 2741 | } |
| 2742 | if (node->hasValues()) { |
| 2743 | const char16_t* id = static_cast<const char16_t*>(node->getValue(0)); |
| 2744 | if (id != nullptr) { |
| 2745 | if (fLen < matchLength) { |
| 2746 | fID = id; |
| 2747 | fLen = matchLength; |
| 2748 | } |
| 2749 | } |
| 2750 | } |
| 2751 | return true; |
| 2752 | } |
| 2753 | |
| 2754 | const char16_t* |
| 2755 | ZoneIdMatchHandler::getID() { |
| 2756 | return fID; |
| 2757 | } |
| 2758 | |
| 2759 | int32_t |
| 2760 | ZoneIdMatchHandler::getMatchLen() { |
| 2761 | return fLen; |
| 2762 | } |
| 2763 | |
| 2764 | |
| 2765 | static void U_CALLCONV initZoneIdTrie(UErrorCode &status) { |
| 2766 | U_ASSERT(gZoneIdTrie == nullptr)(static_cast <bool> (gZoneIdTrie == nullptr) ? void (0) : __assert_fail ("gZoneIdTrie == nullptr", __builtin_FILE () , __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 2767 | ucln_i18n_registerCleanup(UCLN_I18N_TIMEZONEFORMAT, tzfmt_cleanup); |
| 2768 | gZoneIdTrie = new TextTrieMap(true, nullptr); // No deleter, because values are pooled by ZoneMeta |
| 2769 | if (gZoneIdTrie == nullptr) { |
| 2770 | status = U_MEMORY_ALLOCATION_ERROR; |
| 2771 | return; |
| 2772 | } |
| 2773 | StringEnumeration *tzenum = TimeZone::createEnumeration(status); |
| 2774 | if (U_SUCCESS(status)) { |
| 2775 | const UnicodeString *id; |
| 2776 | while ((id = tzenum->snext(status)) != nullptr) { |
| 2777 | const char16_t* uid = ZoneMeta::findTimeZoneID(*id); |
| 2778 | if (uid) { |
| 2779 | gZoneIdTrie->put(uid, const_cast<char16_t *>(uid), status); |
| 2780 | } |
| 2781 | } |
| 2782 | delete tzenum; |
| 2783 | } |
| 2784 | } |
| 2785 | |
| 2786 | |
| 2787 | UnicodeString& |
| 2788 | TimeZoneFormat::parseZoneID(const UnicodeString& text, ParsePosition& pos, UnicodeString& tzID) const { |
| 2789 | UErrorCode status = U_ZERO_ERROR; |
| 2790 | umtx_initOnce(gZoneIdTrieInitOnce, &initZoneIdTrie, status); |
| 2791 | |
| 2792 | int32_t start = pos.getIndex(); |
| 2793 | int32_t len = 0; |
| 2794 | tzID.setToBogus(); |
| 2795 | |
| 2796 | if (U_SUCCESS(status)) { |
| 2797 | LocalPointer<ZoneIdMatchHandler> handler(new ZoneIdMatchHandler()); |
| 2798 | gZoneIdTrie->search(text, start, handler.getAlias(), status); |
| 2799 | len = handler->getMatchLen(); |
| 2800 | if (len > 0) { |
| 2801 | tzID.setTo(handler->getID(), -1); |
| 2802 | } |
| 2803 | } |
| 2804 | |
| 2805 | if (len > 0) { |
| 2806 | pos.setIndex(start + len); |
| 2807 | } else { |
| 2808 | pos.setErrorIndex(start); |
| 2809 | } |
| 2810 | |
| 2811 | return tzID; |
| 2812 | } |
| 2813 | |
| 2814 | static void U_CALLCONV initShortZoneIdTrie(UErrorCode &status) { |
| 2815 | U_ASSERT(gShortZoneIdTrie == nullptr)(static_cast <bool> (gShortZoneIdTrie == nullptr) ? void (0) : __assert_fail ("gShortZoneIdTrie == nullptr", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 2816 | ucln_i18n_registerCleanup(UCLN_I18N_TIMEZONEFORMAT, tzfmt_cleanup); |
| 2817 | StringEnumeration *tzenum = TimeZone::createTimeZoneIDEnumeration(UCAL_ZONE_TYPE_CANONICAL, nullptr, nullptr, status); |
| 2818 | if (U_SUCCESS(status)) { |
| 2819 | gShortZoneIdTrie = new TextTrieMap(true, nullptr); // No deleter, because values are pooled by ZoneMeta |
| 2820 | if (gShortZoneIdTrie == nullptr) { |
| 2821 | status = U_MEMORY_ALLOCATION_ERROR; |
| 2822 | } else { |
| 2823 | const UnicodeString *id; |
| 2824 | while ((id = tzenum->snext(status)) != nullptr) { |
| 2825 | const char16_t* uID = ZoneMeta::findTimeZoneID(*id); |
| 2826 | const char16_t* shortID = ZoneMeta::getShortID(*id); |
| 2827 | if (shortID && uID) { |
| 2828 | gShortZoneIdTrie->put(shortID, const_cast<char16_t *>(uID), status); |
| 2829 | } |
| 2830 | } |
| 2831 | } |
| 2832 | } |
| 2833 | delete tzenum; |
| 2834 | } |
| 2835 | |
| 2836 | |
| 2837 | UnicodeString& |
| 2838 | TimeZoneFormat::parseShortZoneID(const UnicodeString& text, ParsePosition& pos, UnicodeString& tzID) const { |
| 2839 | UErrorCode status = U_ZERO_ERROR; |
| 2840 | umtx_initOnce(gShortZoneIdTrieInitOnce, &initShortZoneIdTrie, status); |
| 2841 | |
| 2842 | int32_t start = pos.getIndex(); |
| 2843 | int32_t len = 0; |
| 2844 | tzID.setToBogus(); |
| 2845 | |
| 2846 | if (U_SUCCESS(status)) { |
| 2847 | LocalPointer<ZoneIdMatchHandler> handler(new ZoneIdMatchHandler()); |
| 2848 | gShortZoneIdTrie->search(text, start, handler.getAlias(), status); |
| 2849 | len = handler->getMatchLen(); |
| 2850 | if (len > 0) { |
| 2851 | tzID.setTo(handler->getID(), -1); |
| 2852 | } |
| 2853 | } |
| 2854 | |
| 2855 | if (len > 0) { |
| 2856 | pos.setIndex(start + len); |
| 2857 | } else { |
| 2858 | pos.setErrorIndex(start); |
| 2859 | } |
| 2860 | |
| 2861 | return tzID; |
| 2862 | } |
| 2863 | |
| 2864 | |
| 2865 | UnicodeString& |
| 2866 | TimeZoneFormat::parseExemplarLocation(const UnicodeString& text, ParsePosition& pos, UnicodeString& tzID) const { |
| 2867 | int32_t startIdx = pos.getIndex(); |
| 2868 | int32_t parsedPos = -1; |
| 2869 | tzID.setToBogus(); |
| 2870 | |
| 2871 | UErrorCode status = U_ZERO_ERROR; |
| 2872 | LocalPointer<TimeZoneNames::MatchInfoCollection> exemplarMatches(fTimeZoneNames->find(text, startIdx, UTZNM_EXEMPLAR_LOCATION, status)); |
| 2873 | if (U_FAILURE(status)) { |
| 2874 | pos.setErrorIndex(startIdx); |
| 2875 | return tzID; |
| 2876 | } |
| 2877 | int32_t matchIdx = -1; |
| 2878 | if (!exemplarMatches.isNull()) { |
| 2879 | for (int32_t i = 0; i < exemplarMatches->size(); i++) { |
| 2880 | if (startIdx + exemplarMatches->getMatchLengthAt(i) > parsedPos) { |
| 2881 | matchIdx = i; |
| 2882 | parsedPos = startIdx + exemplarMatches->getMatchLengthAt(i); |
| 2883 | } |
| 2884 | } |
| 2885 | if (parsedPos > 0) { |
| 2886 | pos.setIndex(parsedPos); |
| 2887 | getTimeZoneID(exemplarMatches.getAlias(), matchIdx, tzID); |
| 2888 | } |
| 2889 | } |
| 2890 | |
| 2891 | if (tzID.length() == 0) { |
| 2892 | pos.setErrorIndex(startIdx); |
| 2893 | } |
| 2894 | |
| 2895 | return tzID; |
| 2896 | } |
| 2897 | |
| 2898 | U_NAMESPACE_END} |
| 2899 | |
| 2900 | #endif |