intl/icu/source/i18n/smpdtfmt.cpp@b8a032363ba2
intl/icu/source/i18n/smpdtfmt.cpp
Thu, 22 Jan 2015 13:21:57 +0100
- author
- Michael Schloh von Bennewitz <michael@schloh.com>
- date
- Thu, 22 Jan 2015 13:21:57 +0100
- branch
- TOR_BUG_9701
- changeset 15
- b8a032363ba2
- permissions
- -rw-r--r--
Incorporate requested changes from Mozilla in review:
https://bugzilla.mozilla.org/show_bug.cgi?id=1123480#c6
1 /*
2 *******************************************************************************
3 * Copyright (C) 1997-2013, International Business Machines Corporation and *
4 * others. All Rights Reserved. *
5 *******************************************************************************
6 *
7 * File SMPDTFMT.CPP
8 *
9 * Modification History:
10 *
11 * Date Name Description
12 * 02/19/97 aliu Converted from java.
13 * 03/31/97 aliu Modified extensively to work with 50 locales.
14 * 04/01/97 aliu Added support for centuries.
15 * 07/09/97 helena Made ParsePosition into a class.
16 * 07/21/98 stephen Added initializeDefaultCentury.
17 * Removed getZoneIndex (added in DateFormatSymbols)
18 * Removed subParseLong
19 * Removed chk
20 * 02/22/99 stephen Removed character literals for EBCDIC safety
21 * 10/14/99 aliu Updated 2-digit year parsing so that only "00" thru
22 * "99" are recognized. {j28 4182066}
23 * 11/15/99 weiv Added support for week of year/day of week format
24 ********************************************************************************
25 */
27 #define ZID_KEY_MAX 128
29 #include "unicode/utypes.h"
31 #if !UCONFIG_NO_FORMATTING
33 #include "unicode/smpdtfmt.h"
34 #include "unicode/dtfmtsym.h"
35 #include "unicode/ures.h"
36 #include "unicode/msgfmt.h"
37 #include "unicode/calendar.h"
38 #include "unicode/gregocal.h"
39 #include "unicode/timezone.h"
40 #include "unicode/decimfmt.h"
41 #include "unicode/dcfmtsym.h"
42 #include "unicode/uchar.h"
43 #include "unicode/uniset.h"
44 #include "unicode/ustring.h"
45 #include "unicode/basictz.h"
46 #include "unicode/simpletz.h"
47 #include "unicode/rbtz.h"
48 #include "unicode/tzfmt.h"
49 #include "unicode/utf16.h"
50 #include "unicode/vtzone.h"
51 #include "unicode/udisplaycontext.h"
52 #include "olsontz.h"
53 #include "patternprops.h"
54 #include "fphdlimp.h"
55 #include "gregoimp.h"
56 #include "hebrwcal.h"
57 #include "cstring.h"
58 #include "uassert.h"
59 #include "cmemory.h"
60 #include "umutex.h"
61 #include <float.h>
62 #include "smpdtfst.h"
64 #if defined( U_DEBUG_CALSVC ) || defined (U_DEBUG_CAL)
65 #include <stdio.h>
66 #endif
68 // *****************************************************************************
69 // class SimpleDateFormat
70 // *****************************************************************************
72 U_NAMESPACE_BEGIN
74 static const UChar PATTERN_CHAR_BASE = 0x40;
76 /**
77 * Last-resort string to use for "GMT" when constructing time zone strings.
78 */
79 // For time zones that have no names, use strings GMT+minutes and
80 // GMT-minutes. For instance, in France the time zone is GMT+60.
81 // Also accepted are GMT+H:MM or GMT-H:MM.
82 // Currently not being used
83 //static const UChar gGmt[] = {0x0047, 0x004D, 0x0054, 0x0000}; // "GMT"
84 //static const UChar gGmtPlus[] = {0x0047, 0x004D, 0x0054, 0x002B, 0x0000}; // "GMT+"
85 //static const UChar gGmtMinus[] = {0x0047, 0x004D, 0x0054, 0x002D, 0x0000}; // "GMT-"
86 //static const UChar gDefGmtPat[] = {0x0047, 0x004D, 0x0054, 0x007B, 0x0030, 0x007D, 0x0000}; /* GMT{0} */
87 //static const UChar gDefGmtNegHmsPat[] = {0x002D, 0x0048, 0x0048, 0x003A, 0x006D, 0x006D, 0x003A, 0x0073, 0x0073, 0x0000}; /* -HH:mm:ss */
88 //static const UChar gDefGmtNegHmPat[] = {0x002D, 0x0048, 0x0048, 0x003A, 0x006D, 0x006D, 0x0000}; /* -HH:mm */
89 //static const UChar gDefGmtPosHmsPat[] = {0x002B, 0x0048, 0x0048, 0x003A, 0x006D, 0x006D, 0x003A, 0x0073, 0x0073, 0x0000}; /* +HH:mm:ss */
90 //static const UChar gDefGmtPosHmPat[] = {0x002B, 0x0048, 0x0048, 0x003A, 0x006D, 0x006D, 0x0000}; /* +HH:mm */
91 //static const UChar gUt[] = {0x0055, 0x0054, 0x0000}; // "UT"
92 //static const UChar gUtc[] = {0x0055, 0x0054, 0x0043, 0x0000}; // "UT"
94 typedef enum GmtPatSize {
95 kGmtLen = 3,
96 kGmtPatLen = 6,
97 kNegHmsLen = 9,
98 kNegHmLen = 6,
99 kPosHmsLen = 9,
100 kPosHmLen = 6,
101 kUtLen = 2,
102 kUtcLen = 3
103 } GmtPatSize;
105 // Stuff needed for numbering system overrides
107 typedef enum OvrStrType {
108 kOvrStrDate = 0,
109 kOvrStrTime = 1,
110 kOvrStrBoth = 2
111 } OvrStrType;
113 static const UDateFormatField kDateFields[] = {
114 UDAT_YEAR_FIELD,
115 UDAT_MONTH_FIELD,
116 UDAT_DATE_FIELD,
117 UDAT_DAY_OF_YEAR_FIELD,
118 UDAT_DAY_OF_WEEK_IN_MONTH_FIELD,
119 UDAT_WEEK_OF_YEAR_FIELD,
120 UDAT_WEEK_OF_MONTH_FIELD,
121 UDAT_YEAR_WOY_FIELD,
122 UDAT_EXTENDED_YEAR_FIELD,
123 UDAT_JULIAN_DAY_FIELD,
124 UDAT_STANDALONE_DAY_FIELD,
125 UDAT_STANDALONE_MONTH_FIELD,
126 UDAT_QUARTER_FIELD,
127 UDAT_STANDALONE_QUARTER_FIELD,
128 UDAT_YEAR_NAME_FIELD };
129 static const int8_t kDateFieldsCount = 15;
131 static const UDateFormatField kTimeFields[] = {
132 UDAT_HOUR_OF_DAY1_FIELD,
133 UDAT_HOUR_OF_DAY0_FIELD,
134 UDAT_MINUTE_FIELD,
135 UDAT_SECOND_FIELD,
136 UDAT_FRACTIONAL_SECOND_FIELD,
137 UDAT_HOUR1_FIELD,
138 UDAT_HOUR0_FIELD,
139 UDAT_MILLISECONDS_IN_DAY_FIELD,
140 UDAT_TIMEZONE_RFC_FIELD,
141 UDAT_TIMEZONE_LOCALIZED_GMT_OFFSET_FIELD };
142 static const int8_t kTimeFieldsCount = 10;
145 // This is a pattern-of-last-resort used when we can't load a usable pattern out
146 // of a resource.
147 static const UChar gDefaultPattern[] =
148 {
149 0x79, 0x79, 0x79, 0x79, 0x4D, 0x4D, 0x64, 0x64, 0x20, 0x68, 0x68, 0x3A, 0x6D, 0x6D, 0x20, 0x61, 0
150 }; /* "yyyyMMdd hh:mm a" */
152 // This prefix is designed to NEVER MATCH real text, in order to
153 // suppress the parsing of negative numbers. Adjust as needed (if
154 // this becomes valid Unicode).
155 static const UChar SUPPRESS_NEGATIVE_PREFIX[] = {0xAB00, 0};
157 /**
158 * These are the tags we expect to see in normal resource bundle files associated
159 * with a locale.
160 */
161 static const char gDateTimePatternsTag[]="DateTimePatterns";
163 //static const UChar gEtcUTC[] = {0x45, 0x74, 0x63, 0x2F, 0x55, 0x54, 0x43, 0x00}; // "Etc/UTC"
164 static const UChar QUOTE = 0x27; // Single quote
166 /*
167 * The field range check bias for each UDateFormatField.
168 * The bias is added to the minimum and maximum values
169 * before they are compared to the parsed number.
170 * For example, the calendar stores zero-based month numbers
171 * but the parsed month numbers start at 1, so the bias is 1.
172 *
173 * A value of -1 means that the value is not checked.
174 */
175 static const int32_t gFieldRangeBias[] = {
176 -1, // 'G' - UDAT_ERA_FIELD
177 -1, // 'y' - UDAT_YEAR_FIELD
178 1, // 'M' - UDAT_MONTH_FIELD
179 0, // 'd' - UDAT_DATE_FIELD
180 -1, // 'k' - UDAT_HOUR_OF_DAY1_FIELD
181 -1, // 'H' - UDAT_HOUR_OF_DAY0_FIELD
182 0, // 'm' - UDAT_MINUTE_FIELD
183 0, // 's' - UDAT_SEOND_FIELD
184 -1, // 'S' - UDAT_FRACTIONAL_SECOND_FIELD (0-999?)
185 -1, // 'E' - UDAT_DAY_OF_WEEK_FIELD (1-7?)
186 -1, // 'D' - UDAT_DAY_OF_YEAR_FIELD (1 - 366?)
187 -1, // 'F' - UDAT_DAY_OF_WEEK_IN_MONTH_FIELD (1-5?)
188 -1, // 'w' - UDAT_WEEK_OF_YEAR_FIELD (1-52?)
189 -1, // 'W' - UDAT_WEEK_OF_MONTH_FIELD (1-5?)
190 -1, // 'a' - UDAT_AM_PM_FIELD
191 -1, // 'h' - UDAT_HOUR1_FIELD
192 -1, // 'K' - UDAT_HOUR0_FIELD
193 -1, // 'z' - UDAT_TIMEZONE_FIELD
194 -1, // 'Y' - UDAT_YEAR_WOY_FIELD
195 -1, // 'e' - UDAT_DOW_LOCAL_FIELD
196 -1, // 'u' - UDAT_EXTENDED_YEAR_FIELD
197 -1, // 'g' - UDAT_JULIAN_DAY_FIELD
198 -1, // 'A' - UDAT_MILLISECONDS_IN_DAY_FIELD
199 -1, // 'Z' - UDAT_TIMEZONE_RFC_FIELD
200 -1, // 'v' - UDAT_TIMEZONE_GENERIC_FIELD
201 0, // 'c' - UDAT_STANDALONE_DAY_FIELD
202 1, // 'L' - UDAT_STANDALONE_MONTH_FIELD
203 -1, // 'Q' - UDAT_QUARTER_FIELD (1-4?)
204 -1, // 'q' - UDAT_STANDALONE_QUARTER_FIELD
205 -1 // 'V' - UDAT_TIMEZONE_SPECIAL_FIELD
206 -1, // 'U' - UDAT_YEAR_NAME_FIELD
207 -1, // 'O' - UDAT_TIMEZONE_LOCALIZED_GMT_OFFSET_FIELD
208 -1, // 'X' - UDAT_TIMEZONE_ISO_FIELD
209 -1, // 'x' - UDAT_TIMEZONE_ISO_LOCAL_FIELD
210 };
212 // When calendar uses hebr numbering (i.e. he@calendar=hebrew),
213 // offset the years within the current millenium down to 1-999
214 static const int32_t HEBREW_CAL_CUR_MILLENIUM_START_YEAR = 5000;
215 static const int32_t HEBREW_CAL_CUR_MILLENIUM_END_YEAR = 6000;
217 static UMutex LOCK = U_MUTEX_INITIALIZER;
219 UOBJECT_DEFINE_RTTI_IMPLEMENTATION(SimpleDateFormat)
221 //----------------------------------------------------------------------
223 SimpleDateFormat::~SimpleDateFormat()
224 {
225 delete fSymbols;
226 if (fNumberFormatters) {
227 uprv_free(fNumberFormatters);
228 }
229 if (fTimeZoneFormat) {
230 delete fTimeZoneFormat;
231 }
233 while (fOverrideList) {
234 NSOverride *cur = fOverrideList;
235 fOverrideList = cur->next;
236 delete cur->nf;
237 uprv_free(cur);
238 }
239 }
241 //----------------------------------------------------------------------
243 SimpleDateFormat::SimpleDateFormat(UErrorCode& status)
244 : fLocale(Locale::getDefault()),
245 fSymbols(NULL),
246 fTimeZoneFormat(NULL),
247 fNumberFormatters(NULL),
248 fOverrideList(NULL),
249 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
250 {
251 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
252 construct(kShort, (EStyle) (kShort + kDateOffset), fLocale, status);
253 initializeDefaultCentury();
254 }
256 //----------------------------------------------------------------------
258 SimpleDateFormat::SimpleDateFormat(const UnicodeString& pattern,
259 UErrorCode &status)
260 : fPattern(pattern),
261 fLocale(Locale::getDefault()),
262 fSymbols(NULL),
263 fTimeZoneFormat(NULL),
264 fNumberFormatters(NULL),
265 fOverrideList(NULL),
266 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
267 {
268 fDateOverride.setToBogus();
269 fTimeOverride.setToBogus();
270 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
271 initializeSymbols(fLocale, initializeCalendar(NULL,fLocale,status), status);
272 initialize(fLocale, status);
273 initializeDefaultCentury();
275 }
276 //----------------------------------------------------------------------
278 SimpleDateFormat::SimpleDateFormat(const UnicodeString& pattern,
279 const UnicodeString& override,
280 UErrorCode &status)
281 : fPattern(pattern),
282 fLocale(Locale::getDefault()),
283 fSymbols(NULL),
284 fTimeZoneFormat(NULL),
285 fNumberFormatters(NULL),
286 fOverrideList(NULL),
287 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
288 {
289 fDateOverride.setTo(override);
290 fTimeOverride.setToBogus();
291 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
292 initializeSymbols(fLocale, initializeCalendar(NULL,fLocale,status), status);
293 initialize(fLocale, status);
294 initializeDefaultCentury();
296 processOverrideString(fLocale,override,kOvrStrBoth,status);
298 }
300 //----------------------------------------------------------------------
302 SimpleDateFormat::SimpleDateFormat(const UnicodeString& pattern,
303 const Locale& locale,
304 UErrorCode& status)
305 : fPattern(pattern),
306 fLocale(locale),
307 fTimeZoneFormat(NULL),
308 fNumberFormatters(NULL),
309 fOverrideList(NULL),
310 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
311 {
313 fDateOverride.setToBogus();
314 fTimeOverride.setToBogus();
315 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
317 initializeSymbols(fLocale, initializeCalendar(NULL,fLocale,status), status);
318 initialize(fLocale, status);
319 initializeDefaultCentury();
320 }
322 //----------------------------------------------------------------------
324 SimpleDateFormat::SimpleDateFormat(const UnicodeString& pattern,
325 const UnicodeString& override,
326 const Locale& locale,
327 UErrorCode& status)
328 : fPattern(pattern),
329 fLocale(locale),
330 fTimeZoneFormat(NULL),
331 fNumberFormatters(NULL),
332 fOverrideList(NULL),
333 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
334 {
336 fDateOverride.setTo(override);
337 fTimeOverride.setToBogus();
338 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
340 initializeSymbols(fLocale, initializeCalendar(NULL,fLocale,status), status);
341 initialize(fLocale, status);
342 initializeDefaultCentury();
344 processOverrideString(locale,override,kOvrStrBoth,status);
346 }
348 //----------------------------------------------------------------------
350 SimpleDateFormat::SimpleDateFormat(const UnicodeString& pattern,
351 DateFormatSymbols* symbolsToAdopt,
352 UErrorCode& status)
353 : fPattern(pattern),
354 fLocale(Locale::getDefault()),
355 fSymbols(symbolsToAdopt),
356 fTimeZoneFormat(NULL),
357 fNumberFormatters(NULL),
358 fOverrideList(NULL),
359 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
360 {
362 fDateOverride.setToBogus();
363 fTimeOverride.setToBogus();
364 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
366 initializeCalendar(NULL,fLocale,status);
367 initialize(fLocale, status);
368 initializeDefaultCentury();
369 }
371 //----------------------------------------------------------------------
373 SimpleDateFormat::SimpleDateFormat(const UnicodeString& pattern,
374 const DateFormatSymbols& symbols,
375 UErrorCode& status)
376 : fPattern(pattern),
377 fLocale(Locale::getDefault()),
378 fSymbols(new DateFormatSymbols(symbols)),
379 fTimeZoneFormat(NULL),
380 fNumberFormatters(NULL),
381 fOverrideList(NULL),
382 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
383 {
385 fDateOverride.setToBogus();
386 fTimeOverride.setToBogus();
387 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
389 initializeCalendar(NULL, fLocale, status);
390 initialize(fLocale, status);
391 initializeDefaultCentury();
392 }
394 //----------------------------------------------------------------------
396 // Not for public consumption; used by DateFormat
397 SimpleDateFormat::SimpleDateFormat(EStyle timeStyle,
398 EStyle dateStyle,
399 const Locale& locale,
400 UErrorCode& status)
401 : fLocale(locale),
402 fSymbols(NULL),
403 fTimeZoneFormat(NULL),
404 fNumberFormatters(NULL),
405 fOverrideList(NULL),
406 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
407 {
408 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
409 construct(timeStyle, dateStyle, fLocale, status);
410 if(U_SUCCESS(status)) {
411 initializeDefaultCentury();
412 }
413 }
415 //----------------------------------------------------------------------
417 /**
418 * Not for public consumption; used by DateFormat. This constructor
419 * never fails. If the resource data is not available, it uses the
420 * the last resort symbols.
421 */
422 SimpleDateFormat::SimpleDateFormat(const Locale& locale,
423 UErrorCode& status)
424 : fPattern(gDefaultPattern),
425 fLocale(locale),
426 fSymbols(NULL),
427 fTimeZoneFormat(NULL),
428 fNumberFormatters(NULL),
429 fOverrideList(NULL),
430 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
431 {
432 if (U_FAILURE(status)) return;
433 initializeSymbols(fLocale, initializeCalendar(NULL, fLocale, status),status);
434 if (U_FAILURE(status))
435 {
436 status = U_ZERO_ERROR;
437 delete fSymbols;
438 // This constructor doesn't fail; it uses last resort data
439 fSymbols = new DateFormatSymbols(status);
440 /* test for NULL */
441 if (fSymbols == 0) {
442 status = U_MEMORY_ALLOCATION_ERROR;
443 return;
444 }
445 }
447 fDateOverride.setToBogus();
448 fTimeOverride.setToBogus();
449 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
451 initialize(fLocale, status);
452 if(U_SUCCESS(status)) {
453 initializeDefaultCentury();
454 }
455 }
457 //----------------------------------------------------------------------
459 SimpleDateFormat::SimpleDateFormat(const SimpleDateFormat& other)
460 : DateFormat(other),
461 fLocale(other.fLocale),
462 fSymbols(NULL),
463 fTimeZoneFormat(NULL),
464 fNumberFormatters(NULL),
465 fOverrideList(NULL),
466 fCapitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
467 {
468 UErrorCode status = U_ZERO_ERROR;
469 setBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, true, status).setBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, true, status);
470 *this = other;
471 }
473 //----------------------------------------------------------------------
475 SimpleDateFormat& SimpleDateFormat::operator=(const SimpleDateFormat& other)
476 {
477 if (this == &other) {
478 return *this;
479 }
480 DateFormat::operator=(other);
482 delete fSymbols;
483 fSymbols = NULL;
485 if (other.fSymbols)
486 fSymbols = new DateFormatSymbols(*other.fSymbols);
488 fDefaultCenturyStart = other.fDefaultCenturyStart;
489 fDefaultCenturyStartYear = other.fDefaultCenturyStartYear;
490 fHaveDefaultCentury = other.fHaveDefaultCentury;
492 fPattern = other.fPattern;
494 // TimeZoneFormat in ICU4C only depends on a locale for now
495 if (fLocale != other.fLocale) {
496 delete fTimeZoneFormat;
497 fTimeZoneFormat = NULL; // forces lazy instantiation with the other locale
498 fLocale = other.fLocale;
499 }
501 fCapitalizationContext = other.fCapitalizationContext;
503 return *this;
504 }
506 //----------------------------------------------------------------------
508 Format*
509 SimpleDateFormat::clone() const
510 {
511 return new SimpleDateFormat(*this);
512 }
514 //----------------------------------------------------------------------
516 UBool
517 SimpleDateFormat::operator==(const Format& other) const
518 {
519 if (DateFormat::operator==(other)) {
520 // DateFormat::operator== guarantees following cast is safe
521 SimpleDateFormat* that = (SimpleDateFormat*)&other;
522 return (fPattern == that->fPattern &&
523 fSymbols != NULL && // Check for pathological object
524 that->fSymbols != NULL && // Check for pathological object
525 *fSymbols == *that->fSymbols &&
526 fHaveDefaultCentury == that->fHaveDefaultCentury &&
527 fDefaultCenturyStart == that->fDefaultCenturyStart &&
528 fCapitalizationContext == that->fCapitalizationContext);
529 }
530 return FALSE;
531 }
533 //----------------------------------------------------------------------
535 void SimpleDateFormat::construct(EStyle timeStyle,
536 EStyle dateStyle,
537 const Locale& locale,
538 UErrorCode& status)
539 {
540 // called by several constructors to load pattern data from the resources
541 if (U_FAILURE(status)) return;
543 // We will need the calendar to know what type of symbols to load.
544 initializeCalendar(NULL, locale, status);
545 if (U_FAILURE(status)) return;
547 CalendarData calData(locale, fCalendar?fCalendar->getType():NULL, status);
548 UResourceBundle *dateTimePatterns = calData.getByKey(gDateTimePatternsTag, status);
549 UResourceBundle *currentBundle;
551 if (U_FAILURE(status)) return;
553 if (ures_getSize(dateTimePatterns) <= kDateTime)
554 {
555 status = U_INVALID_FORMAT_ERROR;
556 return;
557 }
559 setLocaleIDs(ures_getLocaleByType(dateTimePatterns, ULOC_VALID_LOCALE, &status),
560 ures_getLocaleByType(dateTimePatterns, ULOC_ACTUAL_LOCALE, &status));
562 // create a symbols object from the locale
563 initializeSymbols(locale,fCalendar, status);
564 if (U_FAILURE(status)) return;
565 /* test for NULL */
566 if (fSymbols == 0) {
567 status = U_MEMORY_ALLOCATION_ERROR;
568 return;
569 }
571 const UChar *resStr,*ovrStr;
572 int32_t resStrLen,ovrStrLen = 0;
573 fDateOverride.setToBogus();
574 fTimeOverride.setToBogus();
576 // if the pattern should include both date and time information, use the date/time
577 // pattern string as a guide to tell use how to glue together the appropriate date
578 // and time pattern strings. The actual gluing-together is handled by a convenience
579 // method on MessageFormat.
580 if ((timeStyle != kNone) && (dateStyle != kNone))
581 {
582 Formattable timeDateArray[2];
584 // use Formattable::adoptString() so that we can use fastCopyFrom()
585 // instead of Formattable::setString()'s unaware, safe, deep string clone
586 // see Jitterbug 2296
588 currentBundle = ures_getByIndex(dateTimePatterns, (int32_t)timeStyle, NULL, &status);
589 if (U_FAILURE(status)) {
590 status = U_INVALID_FORMAT_ERROR;
591 return;
592 }
593 switch (ures_getType(currentBundle)) {
594 case URES_STRING: {
595 resStr = ures_getString(currentBundle, &resStrLen, &status);
596 break;
597 }
598 case URES_ARRAY: {
599 resStr = ures_getStringByIndex(currentBundle, 0, &resStrLen, &status);
600 ovrStr = ures_getStringByIndex(currentBundle, 1, &ovrStrLen, &status);
601 fTimeOverride.setTo(TRUE, ovrStr, ovrStrLen);
602 break;
603 }
604 default: {
605 status = U_INVALID_FORMAT_ERROR;
606 ures_close(currentBundle);
607 return;
608 }
609 }
610 ures_close(currentBundle);
612 UnicodeString *tempus1 = new UnicodeString(TRUE, resStr, resStrLen);
613 // NULL pointer check
614 if (tempus1 == NULL) {
615 status = U_MEMORY_ALLOCATION_ERROR;
616 return;
617 }
618 timeDateArray[0].adoptString(tempus1);
620 currentBundle = ures_getByIndex(dateTimePatterns, (int32_t)dateStyle, NULL, &status);
621 if (U_FAILURE(status)) {
622 status = U_INVALID_FORMAT_ERROR;
623 return;
624 }
625 switch (ures_getType(currentBundle)) {
626 case URES_STRING: {
627 resStr = ures_getString(currentBundle, &resStrLen, &status);
628 break;
629 }
630 case URES_ARRAY: {
631 resStr = ures_getStringByIndex(currentBundle, 0, &resStrLen, &status);
632 ovrStr = ures_getStringByIndex(currentBundle, 1, &ovrStrLen, &status);
633 fDateOverride.setTo(TRUE, ovrStr, ovrStrLen);
634 break;
635 }
636 default: {
637 status = U_INVALID_FORMAT_ERROR;
638 ures_close(currentBundle);
639 return;
640 }
641 }
642 ures_close(currentBundle);
644 UnicodeString *tempus2 = new UnicodeString(TRUE, resStr, resStrLen);
645 // Null pointer check
646 if (tempus2 == NULL) {
647 status = U_MEMORY_ALLOCATION_ERROR;
648 return;
649 }
650 timeDateArray[1].adoptString(tempus2);
652 int32_t glueIndex = kDateTime;
653 int32_t patternsSize = ures_getSize(dateTimePatterns);
654 if (patternsSize >= (kDateTimeOffset + kShort + 1)) {
655 // Get proper date time format
656 glueIndex = (int32_t)(kDateTimeOffset + (dateStyle - kDateOffset));
657 }
659 resStr = ures_getStringByIndex(dateTimePatterns, glueIndex, &resStrLen, &status);
660 MessageFormat::format(UnicodeString(TRUE, resStr, resStrLen), timeDateArray, 2, fPattern, status);
661 }
662 // if the pattern includes just time data or just date date, load the appropriate
663 // pattern string from the resources
664 // setTo() - see DateFormatSymbols::assignArray comments
665 else if (timeStyle != kNone) {
666 currentBundle = ures_getByIndex(dateTimePatterns, (int32_t)timeStyle, NULL, &status);
667 if (U_FAILURE(status)) {
668 status = U_INVALID_FORMAT_ERROR;
669 return;
670 }
671 switch (ures_getType(currentBundle)) {
672 case URES_STRING: {
673 resStr = ures_getString(currentBundle, &resStrLen, &status);
674 break;
675 }
676 case URES_ARRAY: {
677 resStr = ures_getStringByIndex(currentBundle, 0, &resStrLen, &status);
678 ovrStr = ures_getStringByIndex(currentBundle, 1, &ovrStrLen, &status);
679 fDateOverride.setTo(TRUE, ovrStr, ovrStrLen);
680 break;
681 }
682 default: {
683 status = U_INVALID_FORMAT_ERROR;
684 ures_close(currentBundle);
685 return;
686 }
687 }
688 fPattern.setTo(TRUE, resStr, resStrLen);
689 ures_close(currentBundle);
690 }
691 else if (dateStyle != kNone) {
692 currentBundle = ures_getByIndex(dateTimePatterns, (int32_t)dateStyle, NULL, &status);
693 if (U_FAILURE(status)) {
694 status = U_INVALID_FORMAT_ERROR;
695 return;
696 }
697 switch (ures_getType(currentBundle)) {
698 case URES_STRING: {
699 resStr = ures_getString(currentBundle, &resStrLen, &status);
700 break;
701 }
702 case URES_ARRAY: {
703 resStr = ures_getStringByIndex(currentBundle, 0, &resStrLen, &status);
704 ovrStr = ures_getStringByIndex(currentBundle, 1, &ovrStrLen, &status);
705 fDateOverride.setTo(TRUE, ovrStr, ovrStrLen);
706 break;
707 }
708 default: {
709 status = U_INVALID_FORMAT_ERROR;
710 ures_close(currentBundle);
711 return;
712 }
713 }
714 fPattern.setTo(TRUE, resStr, resStrLen);
715 ures_close(currentBundle);
716 }
718 // and if it includes _neither_, that's an error
719 else
720 status = U_INVALID_FORMAT_ERROR;
722 // finally, finish initializing by creating a Calendar and a NumberFormat
723 initialize(locale, status);
724 }
726 //----------------------------------------------------------------------
728 Calendar*
729 SimpleDateFormat::initializeCalendar(TimeZone* adoptZone, const Locale& locale, UErrorCode& status)
730 {
731 if(!U_FAILURE(status)) {
732 fCalendar = Calendar::createInstance(adoptZone?adoptZone:TimeZone::createDefault(), locale, status);
733 }
734 if (U_SUCCESS(status) && fCalendar == NULL) {
735 status = U_MEMORY_ALLOCATION_ERROR;
736 }
737 return fCalendar;
738 }
740 void
741 SimpleDateFormat::initializeSymbols(const Locale& locale, Calendar* calendar, UErrorCode& status)
742 {
743 if(U_FAILURE(status)) {
744 fSymbols = NULL;
745 } else {
746 // pass in calendar type - use NULL (default) if no calendar set (or err).
747 fSymbols = new DateFormatSymbols(locale, calendar?calendar->getType() :NULL , status);
748 // Null pointer check
749 if (fSymbols == NULL) {
750 status = U_MEMORY_ALLOCATION_ERROR;
751 return;
752 }
753 }
754 }
756 void
757 SimpleDateFormat::initialize(const Locale& locale,
758 UErrorCode& status)
759 {
760 if (U_FAILURE(status)) return;
762 // We don't need to check that the row count is >= 1, since all 2d arrays have at
763 // least one row
764 fNumberFormat = NumberFormat::createInstance(locale, status);
765 if (fNumberFormat != NULL && U_SUCCESS(status))
766 {
767 // no matter what the locale's default number format looked like, we want
768 // to modify it so that it doesn't use thousands separators, doesn't always
769 // show the decimal point, and recognizes integers only when parsing
771 fNumberFormat->setGroupingUsed(FALSE);
772 DecimalFormat* decfmt = dynamic_cast<DecimalFormat*>(fNumberFormat);
773 if (decfmt != NULL) {
774 decfmt->setDecimalSeparatorAlwaysShown(FALSE);
775 }
776 fNumberFormat->setParseIntegerOnly(TRUE);
777 fNumberFormat->setMinimumFractionDigits(0); // To prevent "Jan 1.00, 1997.00"
779 //fNumberFormat->setLenient(TRUE); // Java uses a custom DateNumberFormat to format/parse
781 initNumberFormatters(locale,status);
783 }
784 else if (U_SUCCESS(status))
785 {
786 status = U_MISSING_RESOURCE_ERROR;
787 }
788 }
790 /* Initialize the fields we use to disambiguate ambiguous years. Separate
791 * so we can call it from readObject().
792 */
793 void SimpleDateFormat::initializeDefaultCentury()
794 {
795 if(fCalendar) {
796 fHaveDefaultCentury = fCalendar->haveDefaultCentury();
797 if(fHaveDefaultCentury) {
798 fDefaultCenturyStart = fCalendar->defaultCenturyStart();
799 fDefaultCenturyStartYear = fCalendar->defaultCenturyStartYear();
800 } else {
801 fDefaultCenturyStart = DBL_MIN;
802 fDefaultCenturyStartYear = -1;
803 }
804 }
805 }
807 /* Define one-century window into which to disambiguate dates using
808 * two-digit years. Make public in JDK 1.2.
809 */
810 void SimpleDateFormat::parseAmbiguousDatesAsAfter(UDate startDate, UErrorCode& status)
811 {
812 if(U_FAILURE(status)) {
813 return;
814 }
815 if(!fCalendar) {
816 status = U_ILLEGAL_ARGUMENT_ERROR;
817 return;
818 }
820 fCalendar->setTime(startDate, status);
821 if(U_SUCCESS(status)) {
822 fHaveDefaultCentury = TRUE;
823 fDefaultCenturyStart = startDate;
824 fDefaultCenturyStartYear = fCalendar->get(UCAL_YEAR, status);
825 }
826 }
828 //----------------------------------------------------------------------
830 UnicodeString&
831 SimpleDateFormat::format(Calendar& cal, UnicodeString& appendTo, FieldPosition& pos) const
832 {
833 UErrorCode status = U_ZERO_ERROR;
834 FieldPositionOnlyHandler handler(pos);
835 return _format(cal, appendTo, handler, status);
836 }
838 //----------------------------------------------------------------------
840 UnicodeString&
841 SimpleDateFormat::format(Calendar& cal, UnicodeString& appendTo,
842 FieldPositionIterator* posIter, UErrorCode& status) const
843 {
844 FieldPositionIteratorHandler handler(posIter, status);
845 return _format(cal, appendTo, handler, status);
846 }
848 //----------------------------------------------------------------------
850 UnicodeString&
851 SimpleDateFormat::_format(Calendar& cal, UnicodeString& appendTo,
852 FieldPositionHandler& handler, UErrorCode& status) const
853 {
854 if ( U_FAILURE(status) ) {
855 return appendTo;
856 }
857 Calendar* workCal = &cal;
858 Calendar* calClone = NULL;
859 if (&cal != fCalendar && uprv_strcmp(cal.getType(), fCalendar->getType()) != 0) {
860 // Different calendar type
861 // We use the time and time zone from the input calendar, but
862 // do not use the input calendar for field calculation.
863 calClone = fCalendar->clone();
864 if (calClone != NULL) {
865 UDate t = cal.getTime(status);
866 calClone->setTime(t, status);
867 calClone->setTimeZone(cal.getTimeZone());
868 workCal = calClone;
869 } else {
870 status = U_MEMORY_ALLOCATION_ERROR;
871 return appendTo;
872 }
873 }
875 UBool inQuote = FALSE;
876 UChar prevCh = 0;
877 int32_t count = 0;
878 int32_t fieldNum = 0;
880 // loop through the pattern string character by character
881 for (int32_t i = 0; i < fPattern.length() && U_SUCCESS(status); ++i) {
882 UChar ch = fPattern[i];
884 // Use subFormat() to format a repeated pattern character
885 // when a different pattern or non-pattern character is seen
886 if (ch != prevCh && count > 0) {
887 subFormat(appendTo, prevCh, count, fCapitalizationContext, fieldNum++, handler, *workCal, status);
888 count = 0;
889 }
890 if (ch == QUOTE) {
891 // Consecutive single quotes are a single quote literal,
892 // either outside of quotes or between quotes
893 if ((i+1) < fPattern.length() && fPattern[i+1] == QUOTE) {
894 appendTo += (UChar)QUOTE;
895 ++i;
896 } else {
897 inQuote = ! inQuote;
898 }
899 }
900 else if ( ! inQuote && ((ch >= 0x0061 /*'a'*/ && ch <= 0x007A /*'z'*/)
901 || (ch >= 0x0041 /*'A'*/ && ch <= 0x005A /*'Z'*/))) {
902 // ch is a date-time pattern character to be interpreted
903 // by subFormat(); count the number of times it is repeated
904 prevCh = ch;
905 ++count;
906 }
907 else {
908 // Append quoted characters and unquoted non-pattern characters
909 appendTo += ch;
910 }
911 }
913 // Format the last item in the pattern, if any
914 if (count > 0) {
915 subFormat(appendTo, prevCh, count, fCapitalizationContext, fieldNum++, handler, *workCal, status);
916 }
918 if (calClone != NULL) {
919 delete calClone;
920 }
922 return appendTo;
923 }
925 //----------------------------------------------------------------------
927 /* Map calendar field into calendar field level.
928 * the larger the level, the smaller the field unit.
929 * For example, UCAL_ERA level is 0, UCAL_YEAR level is 10,
930 * UCAL_MONTH level is 20.
931 * NOTE: if new fields adds in, the table needs to update.
932 */
933 const int32_t
934 SimpleDateFormat::fgCalendarFieldToLevel[] =
935 {
936 /*GyM*/ 0, 10, 20,
937 /*wW*/ 20, 30,
938 /*dDEF*/ 30, 20, 30, 30,
939 /*ahHm*/ 40, 50, 50, 60,
940 /*sS..*/ 70, 80,
941 /*z?Y*/ 0, 0, 10,
942 /*eug*/ 30, 10, 0,
943 /*A*/ 40
944 };
947 /* Map calendar field LETTER into calendar field level.
948 * the larger the level, the smaller the field unit.
949 * NOTE: if new fields adds in, the table needs to update.
950 */
951 const int32_t
952 SimpleDateFormat::fgPatternCharToLevel[] = {
953 // A B C D E F G H I J K L M N O
954 -1, 40, -1, -1, 20, 30, 30, 0, 50, -1, -1, 50, 20, 20, -1, 0,
955 // P Q R S T U V W X Y Z
956 -1, 20, -1, 80, -1, 10, 0, 30, 0, 10, 0, -1, -1, -1, -1, -1,
957 // a b c d e f g h i j k l m n o
958 -1, 40, -1, 30, 30, 30, -1, 0, 50, -1, -1, 50, -1, 60, -1, -1,
959 // p q r s t u v w x y z
960 -1, 20, -1, 70, -1, 10, 0, 20, 0, 10, 0, -1, -1, -1, -1, -1
961 };
964 // Map index into pattern character string to Calendar field number.
965 const UCalendarDateFields
966 SimpleDateFormat::fgPatternIndexToCalendarField[] =
967 {
968 /*GyM*/ UCAL_ERA, UCAL_YEAR, UCAL_MONTH,
969 /*dkH*/ UCAL_DATE, UCAL_HOUR_OF_DAY, UCAL_HOUR_OF_DAY,
970 /*msS*/ UCAL_MINUTE, UCAL_SECOND, UCAL_MILLISECOND,
971 /*EDF*/ UCAL_DAY_OF_WEEK, UCAL_DAY_OF_YEAR, UCAL_DAY_OF_WEEK_IN_MONTH,
972 /*wWa*/ UCAL_WEEK_OF_YEAR, UCAL_WEEK_OF_MONTH, UCAL_AM_PM,
973 /*hKz*/ UCAL_HOUR, UCAL_HOUR, UCAL_ZONE_OFFSET,
974 /*Yeu*/ UCAL_YEAR_WOY, UCAL_DOW_LOCAL, UCAL_EXTENDED_YEAR,
975 /*gAZ*/ UCAL_JULIAN_DAY, UCAL_MILLISECONDS_IN_DAY, UCAL_ZONE_OFFSET,
976 /*v*/ UCAL_ZONE_OFFSET,
977 /*c*/ UCAL_DOW_LOCAL,
978 /*L*/ UCAL_MONTH,
979 /*Q*/ UCAL_MONTH,
980 /*q*/ UCAL_MONTH,
981 /*V*/ UCAL_ZONE_OFFSET,
982 /*U*/ UCAL_YEAR,
983 /*O*/ UCAL_ZONE_OFFSET,
984 /*Xx*/ UCAL_ZONE_OFFSET, UCAL_ZONE_OFFSET,
985 };
987 // Map index into pattern character string to DateFormat field number
988 const UDateFormatField
989 SimpleDateFormat::fgPatternIndexToDateFormatField[] = {
990 /*GyM*/ UDAT_ERA_FIELD, UDAT_YEAR_FIELD, UDAT_MONTH_FIELD,
991 /*dkH*/ UDAT_DATE_FIELD, UDAT_HOUR_OF_DAY1_FIELD, UDAT_HOUR_OF_DAY0_FIELD,
992 /*msS*/ UDAT_MINUTE_FIELD, UDAT_SECOND_FIELD, UDAT_FRACTIONAL_SECOND_FIELD,
993 /*EDF*/ UDAT_DAY_OF_WEEK_FIELD, UDAT_DAY_OF_YEAR_FIELD, UDAT_DAY_OF_WEEK_IN_MONTH_FIELD,
994 /*wWa*/ UDAT_WEEK_OF_YEAR_FIELD, UDAT_WEEK_OF_MONTH_FIELD, UDAT_AM_PM_FIELD,
995 /*hKz*/ UDAT_HOUR1_FIELD, UDAT_HOUR0_FIELD, UDAT_TIMEZONE_FIELD,
996 /*Yeu*/ UDAT_YEAR_WOY_FIELD, UDAT_DOW_LOCAL_FIELD, UDAT_EXTENDED_YEAR_FIELD,
997 /*gAZ*/ UDAT_JULIAN_DAY_FIELD, UDAT_MILLISECONDS_IN_DAY_FIELD, UDAT_TIMEZONE_RFC_FIELD,
998 /*v*/ UDAT_TIMEZONE_GENERIC_FIELD,
999 /*c*/ UDAT_STANDALONE_DAY_FIELD,
1000 /*L*/ UDAT_STANDALONE_MONTH_FIELD,
1001 /*Q*/ UDAT_QUARTER_FIELD,
1002 /*q*/ UDAT_STANDALONE_QUARTER_FIELD,
1003 /*V*/ UDAT_TIMEZONE_SPECIAL_FIELD,
1004 /*U*/ UDAT_YEAR_NAME_FIELD,
1005 /*O*/ UDAT_TIMEZONE_LOCALIZED_GMT_OFFSET_FIELD,
1006 /*Xx*/ UDAT_TIMEZONE_ISO_FIELD, UDAT_TIMEZONE_ISO_LOCAL_FIELD,
1007 };
1009 //----------------------------------------------------------------------
1011 /**
1012 * Append symbols[value] to dst. Make sure the array index is not out
1013 * of bounds.
1014 */
1015 static inline void
1016 _appendSymbol(UnicodeString& dst,
1017 int32_t value,
1018 const UnicodeString* symbols,
1019 int32_t symbolsCount) {
1020 U_ASSERT(0 <= value && value < symbolsCount);
1021 if (0 <= value && value < symbolsCount) {
1022 dst += symbols[value];
1023 }
1024 }
1026 static inline void
1027 _appendSymbolWithMonthPattern(UnicodeString& dst, int32_t value, const UnicodeString* symbols, int32_t symbolsCount,
1028 const UnicodeString* monthPattern, UErrorCode& status) {
1029 U_ASSERT(0 <= value && value < symbolsCount);
1030 if (0 <= value && value < symbolsCount) {
1031 if (monthPattern == NULL) {
1032 dst += symbols[value];
1033 } else {
1034 Formattable monthName((const UnicodeString&)(symbols[value]));
1035 MessageFormat::format(*monthPattern, &monthName, 1, dst, status);
1036 }
1037 }
1038 }
1040 //----------------------------------------------------------------------
1041 void
1042 SimpleDateFormat::initNumberFormatters(const Locale &locale,UErrorCode &status) {
1043 if (U_FAILURE(status)) {
1044 return;
1045 }
1046 if ( fDateOverride.isBogus() && fTimeOverride.isBogus() ) {
1047 return;
1048 }
1049 umtx_lock(&LOCK);
1050 if (fNumberFormatters == NULL) {
1051 fNumberFormatters = (NumberFormat**)uprv_malloc(UDAT_FIELD_COUNT * sizeof(NumberFormat*));
1052 if (fNumberFormatters) {
1053 for (int32_t i = 0; i < UDAT_FIELD_COUNT; i++) {
1054 fNumberFormatters[i] = fNumberFormat;
1055 }
1056 } else {
1057 status = U_MEMORY_ALLOCATION_ERROR;
1058 }
1059 }
1060 umtx_unlock(&LOCK);
1062 processOverrideString(locale,fDateOverride,kOvrStrDate,status);
1063 processOverrideString(locale,fTimeOverride,kOvrStrTime,status);
1065 }
1067 void
1068 SimpleDateFormat::processOverrideString(const Locale &locale, const UnicodeString &str, int8_t type, UErrorCode &status) {
1069 if (str.isBogus()) {
1070 return;
1071 }
1072 int32_t start = 0;
1073 int32_t len;
1074 UnicodeString nsName;
1075 UnicodeString ovrField;
1076 UBool moreToProcess = TRUE;
1078 while (moreToProcess) {
1079 int32_t delimiterPosition = str.indexOf((UChar)ULOC_KEYWORD_ITEM_SEPARATOR_UNICODE,start);
1080 if (delimiterPosition == -1) {
1081 moreToProcess = FALSE;
1082 len = str.length() - start;
1083 } else {
1084 len = delimiterPosition - start;
1085 }
1086 UnicodeString currentString(str,start,len);
1087 int32_t equalSignPosition = currentString.indexOf((UChar)ULOC_KEYWORD_ASSIGN_UNICODE,0);
1088 if (equalSignPosition == -1) { // Simple override string such as "hebrew"
1089 nsName.setTo(currentString);
1090 ovrField.setToBogus();
1091 } else { // Field specific override string such as "y=hebrew"
1092 nsName.setTo(currentString,equalSignPosition+1);
1093 ovrField.setTo(currentString,0,1); // We just need the first character.
1094 }
1096 int32_t nsNameHash = nsName.hashCode();
1097 // See if the numbering system is in the override list, if not, then add it.
1098 NSOverride *cur = fOverrideList;
1099 NumberFormat *nf = NULL;
1100 UBool found = FALSE;
1101 while ( cur && !found ) {
1102 if ( cur->hash == nsNameHash ) {
1103 nf = cur->nf;
1104 found = TRUE;
1105 }
1106 cur = cur->next;
1107 }
1109 if (!found) {
1110 cur = (NSOverride *)uprv_malloc(sizeof(NSOverride));
1111 if (cur) {
1112 char kw[ULOC_KEYWORD_AND_VALUES_CAPACITY];
1113 uprv_strcpy(kw,"numbers=");
1114 nsName.extract(0,len,kw+8,ULOC_KEYWORD_AND_VALUES_CAPACITY-8,US_INV);
1116 Locale ovrLoc(locale.getLanguage(),locale.getCountry(),locale.getVariant(),kw);
1117 nf = NumberFormat::createInstance(ovrLoc,status);
1119 // no matter what the locale's default number format looked like, we want
1120 // to modify it so that it doesn't use thousands separators, doesn't always
1121 // show the decimal point, and recognizes integers only when parsing
1123 if (U_SUCCESS(status)) {
1124 nf->setGroupingUsed(FALSE);
1125 DecimalFormat* decfmt = dynamic_cast<DecimalFormat*>(nf);
1126 if (decfmt != NULL) {
1127 decfmt->setDecimalSeparatorAlwaysShown(FALSE);
1128 }
1129 nf->setParseIntegerOnly(TRUE);
1130 nf->setMinimumFractionDigits(0); // To prevent "Jan 1.00, 1997.00"
1132 cur->nf = nf;
1133 cur->hash = nsNameHash;
1134 cur->next = fOverrideList;
1135 fOverrideList = cur;
1136 }
1137 else {
1138 // clean up before returning
1139 if (cur != NULL) {
1140 uprv_free(cur);
1141 }
1142 return;
1143 }
1145 } else {
1146 status = U_MEMORY_ALLOCATION_ERROR;
1147 return;
1148 }
1149 }
1151 // Now that we have an appropriate number formatter, fill in the appropriate spaces in the
1152 // number formatters table.
1154 if (ovrField.isBogus()) {
1155 switch (type) {
1156 case kOvrStrDate:
1157 case kOvrStrBoth: {
1158 for ( int8_t i=0 ; i<kDateFieldsCount; i++ ) {
1159 fNumberFormatters[kDateFields[i]] = nf;
1160 }
1161 if (type==kOvrStrDate) {
1162 break;
1163 }
1164 }
1165 case kOvrStrTime : {
1166 for ( int8_t i=0 ; i<kTimeFieldsCount; i++ ) {
1167 fNumberFormatters[kTimeFields[i]] = nf;
1168 }
1169 break;
1170 }
1171 }
1172 } else {
1173 // if the pattern character is unrecognized, signal an error and bail out
1174 UDateFormatField patternCharIndex =
1175 DateFormatSymbols::getPatternCharIndex(ovrField.charAt(0));
1176 if (patternCharIndex == UDAT_FIELD_COUNT) {
1177 status = U_INVALID_FORMAT_ERROR;
1178 return;
1179 }
1181 // Set the number formatter in the table
1182 fNumberFormatters[patternCharIndex] = nf;
1183 }
1185 start = delimiterPosition + 1;
1186 }
1187 }
1189 //---------------------------------------------------------------------
1190 void
1191 SimpleDateFormat::subFormat(UnicodeString &appendTo,
1192 UChar ch,
1193 int32_t count,
1194 UDisplayContext capitalizationContext,
1195 int32_t fieldNum,
1196 FieldPositionHandler& handler,
1197 Calendar& cal,
1198 UErrorCode& status) const
1199 {
1200 if (U_FAILURE(status)) {
1201 return;
1202 }
1204 // this function gets called by format() to produce the appropriate substitution
1205 // text for an individual pattern symbol (e.g., "HH" or "yyyy")
1207 UDateFormatField patternCharIndex = DateFormatSymbols::getPatternCharIndex(ch);
1208 const int32_t maxIntCount = 10;
1209 int32_t beginOffset = appendTo.length();
1210 NumberFormat *currentNumberFormat;
1211 DateFormatSymbols::ECapitalizationContextUsageType capContextUsageType = DateFormatSymbols::kCapContextUsageOther;
1213 UBool isHebrewCalendar = (uprv_strcmp(cal.getType(),"hebrew") == 0);
1214 UBool isChineseCalendar = (uprv_strcmp(cal.getType(),"chinese") == 0 || uprv_strcmp(cal.getType(),"dangi") == 0);
1216 // if the pattern character is unrecognized, signal an error and dump out
1217 if (patternCharIndex == UDAT_FIELD_COUNT)
1218 {
1219 if (ch != 0x6C) { // pattern char 'l' (SMALL LETTER L) just gets ignored
1220 status = U_INVALID_FORMAT_ERROR;
1221 }
1222 return;
1223 }
1225 UCalendarDateFields field = fgPatternIndexToCalendarField[patternCharIndex];
1226 int32_t value = cal.get(field, status);
1227 if (U_FAILURE(status)) {
1228 return;
1229 }
1231 currentNumberFormat = getNumberFormatByIndex(patternCharIndex);
1232 UnicodeString hebr("hebr", 4, US_INV);
1234 switch (patternCharIndex) {
1236 // for any "G" symbol, write out the appropriate era string
1237 // "GGGG" is wide era name, "GGGGG" is narrow era name, anything else is abbreviated name
1238 case UDAT_ERA_FIELD:
1239 if (isChineseCalendar) {
1240 zeroPaddingNumber(currentNumberFormat,appendTo, value, 1, 9); // as in ICU4J
1241 } else {
1242 if (count == 5) {
1243 _appendSymbol(appendTo, value, fSymbols->fNarrowEras, fSymbols->fNarrowErasCount);
1244 capContextUsageType = DateFormatSymbols::kCapContextUsageEraNarrow;
1245 } else if (count == 4) {
1246 _appendSymbol(appendTo, value, fSymbols->fEraNames, fSymbols->fEraNamesCount);
1247 capContextUsageType = DateFormatSymbols::kCapContextUsageEraWide;
1248 } else {
1249 _appendSymbol(appendTo, value, fSymbols->fEras, fSymbols->fErasCount);
1250 capContextUsageType = DateFormatSymbols::kCapContextUsageEraAbbrev;
1251 }
1252 }
1253 break;
1255 case UDAT_YEAR_NAME_FIELD:
1256 if (fSymbols->fShortYearNames != NULL && value <= fSymbols->fShortYearNamesCount) {
1257 // the Calendar YEAR field runs 1 through 60 for cyclic years
1258 _appendSymbol(appendTo, value - 1, fSymbols->fShortYearNames, fSymbols->fShortYearNamesCount);
1259 break;
1260 }
1261 // else fall through to numeric year handling, do not break here
1263 // OLD: for "yyyy", write out the whole year; for "yy", write out the last 2 digits
1264 // NEW: UTS#35:
1265 //Year y yy yyy yyyy yyyyy
1266 //AD 1 1 01 001 0001 00001
1267 //AD 12 12 12 012 0012 00012
1268 //AD 123 123 23 123 0123 00123
1269 //AD 1234 1234 34 1234 1234 01234
1270 //AD 12345 12345 45 12345 12345 12345
1271 case UDAT_YEAR_FIELD:
1272 case UDAT_YEAR_WOY_FIELD:
1273 if (fDateOverride.compare(hebr)==0 && value>HEBREW_CAL_CUR_MILLENIUM_START_YEAR && value<HEBREW_CAL_CUR_MILLENIUM_END_YEAR) {
1274 value-=HEBREW_CAL_CUR_MILLENIUM_START_YEAR;
1275 }
1276 if(count == 2)
1277 zeroPaddingNumber(currentNumberFormat, appendTo, value, 2, 2);
1278 else
1279 zeroPaddingNumber(currentNumberFormat, appendTo, value, count, maxIntCount);
1280 break;
1282 // for "MMMM"/"LLLL", write out the whole month name, for "MMM"/"LLL", write out the month
1283 // abbreviation, for "M"/"L" or "MM"/"LL", write out the month as a number with the
1284 // appropriate number of digits
1285 // for "MMMMM"/"LLLLL", use the narrow form
1286 case UDAT_MONTH_FIELD:
1287 case UDAT_STANDALONE_MONTH_FIELD:
1288 if ( isHebrewCalendar ) {
1289 HebrewCalendar *hc = (HebrewCalendar*)&cal;
1290 if (hc->isLeapYear(hc->get(UCAL_YEAR,status)) && value == 6 && count >= 3 )
1291 value = 13; // Show alternate form for Adar II in leap years in Hebrew calendar.
1292 if (!hc->isLeapYear(hc->get(UCAL_YEAR,status)) && value >= 6 && count < 3 )
1293 value--; // Adjust the month number down 1 in Hebrew non-leap years, i.e. Adar is 6, not 7.
1294 }
1295 {
1296 int32_t isLeapMonth = (fSymbols->fLeapMonthPatterns != NULL && fSymbols->fLeapMonthPatternsCount >= DateFormatSymbols::kMonthPatternsCount)?
1297 cal.get(UCAL_IS_LEAP_MONTH, status): 0;
1298 // should consolidate the next section by using arrays of pointers & counts for the right symbols...
1299 if (count == 5) {
1300 if (patternCharIndex == UDAT_MONTH_FIELD) {
1301 _appendSymbolWithMonthPattern(appendTo, value, fSymbols->fNarrowMonths, fSymbols->fNarrowMonthsCount,
1302 (isLeapMonth!=0)? &(fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternFormatNarrow]): NULL, status);
1303 } else {
1304 _appendSymbolWithMonthPattern(appendTo, value, fSymbols->fStandaloneNarrowMonths, fSymbols->fStandaloneNarrowMonthsCount,
1305 (isLeapMonth!=0)? &(fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternStandaloneNarrow]): NULL, status);
1306 }
1307 capContextUsageType = DateFormatSymbols::kCapContextUsageMonthNarrow;
1308 } else if (count == 4) {
1309 if (patternCharIndex == UDAT_MONTH_FIELD) {
1310 _appendSymbolWithMonthPattern(appendTo, value, fSymbols->fMonths, fSymbols->fMonthsCount,
1311 (isLeapMonth!=0)? &(fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternFormatWide]): NULL, status);
1312 capContextUsageType = DateFormatSymbols::kCapContextUsageMonthFormat;
1313 } else {
1314 _appendSymbolWithMonthPattern(appendTo, value, fSymbols->fStandaloneMonths, fSymbols->fStandaloneMonthsCount,
1315 (isLeapMonth!=0)? &(fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternStandaloneWide]): NULL, status);
1316 capContextUsageType = DateFormatSymbols::kCapContextUsageMonthStandalone;
1317 }
1318 } else if (count == 3) {
1319 if (patternCharIndex == UDAT_MONTH_FIELD) {
1320 _appendSymbolWithMonthPattern(appendTo, value, fSymbols->fShortMonths, fSymbols->fShortMonthsCount,
1321 (isLeapMonth!=0)? &(fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternFormatAbbrev]): NULL, status);
1322 capContextUsageType = DateFormatSymbols::kCapContextUsageMonthFormat;
1323 } else {
1324 _appendSymbolWithMonthPattern(appendTo, value, fSymbols->fStandaloneShortMonths, fSymbols->fStandaloneShortMonthsCount,
1325 (isLeapMonth!=0)? &(fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternStandaloneAbbrev]): NULL, status);
1326 capContextUsageType = DateFormatSymbols::kCapContextUsageMonthStandalone;
1327 }
1328 } else {
1329 UnicodeString monthNumber;
1330 zeroPaddingNumber(currentNumberFormat,monthNumber, value + 1, count, maxIntCount);
1331 _appendSymbolWithMonthPattern(appendTo, 0, &monthNumber, 1,
1332 (isLeapMonth!=0)? &(fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternNumeric]): NULL, status);
1333 }
1334 }
1335 break;
1337 // for "k" and "kk", write out the hour, adjusting midnight to appear as "24"
1338 case UDAT_HOUR_OF_DAY1_FIELD:
1339 if (value == 0)
1340 zeroPaddingNumber(currentNumberFormat,appendTo, cal.getMaximum(UCAL_HOUR_OF_DAY) + 1, count, maxIntCount);
1341 else
1342 zeroPaddingNumber(currentNumberFormat,appendTo, value, count, maxIntCount);
1343 break;
1345 case UDAT_FRACTIONAL_SECOND_FIELD:
1346 // Fractional seconds left-justify
1347 {
1348 currentNumberFormat->setMinimumIntegerDigits((count > 3) ? 3 : count);
1349 currentNumberFormat->setMaximumIntegerDigits(maxIntCount);
1350 if (count == 1) {
1351 value /= 100;
1352 } else if (count == 2) {
1353 value /= 10;
1354 }
1355 FieldPosition p(0);
1356 currentNumberFormat->format(value, appendTo, p);
1357 if (count > 3) {
1358 currentNumberFormat->setMinimumIntegerDigits(count - 3);
1359 currentNumberFormat->format((int32_t)0, appendTo, p);
1360 }
1361 }
1362 break;
1364 // for "ee" or "e", use local numeric day-of-the-week
1365 // for "EEEEEE" or "eeeeee", write out the short day-of-the-week name
1366 // for "EEEEE" or "eeeee", write out the narrow day-of-the-week name
1367 // for "EEEE" or "eeee", write out the wide day-of-the-week name
1368 // for "EEE" or "EE" or "E" or "eee", write out the abbreviated day-of-the-week name
1369 case UDAT_DOW_LOCAL_FIELD:
1370 if ( count < 3 ) {
1371 zeroPaddingNumber(currentNumberFormat,appendTo, value, count, maxIntCount);
1372 break;
1373 }
1374 // fall through to EEEEE-EEE handling, but for that we don't want local day-of-week,
1375 // we want standard day-of-week, so first fix value to work for EEEEE-EEE.
1376 value = cal.get(UCAL_DAY_OF_WEEK, status);
1377 if (U_FAILURE(status)) {
1378 return;
1379 }
1380 // fall through, do not break here
1381 case UDAT_DAY_OF_WEEK_FIELD:
1382 if (count == 5) {
1383 _appendSymbol(appendTo, value, fSymbols->fNarrowWeekdays,
1384 fSymbols->fNarrowWeekdaysCount);
1385 capContextUsageType = DateFormatSymbols::kCapContextUsageDayNarrow;
1386 } else if (count == 4) {
1387 _appendSymbol(appendTo, value, fSymbols->fWeekdays,
1388 fSymbols->fWeekdaysCount);
1389 capContextUsageType = DateFormatSymbols::kCapContextUsageDayFormat;
1390 } else if (count == 6) {
1391 _appendSymbol(appendTo, value, fSymbols->fShorterWeekdays,
1392 fSymbols->fShorterWeekdaysCount);
1393 capContextUsageType = DateFormatSymbols::kCapContextUsageDayFormat;
1394 } else {
1395 _appendSymbol(appendTo, value, fSymbols->fShortWeekdays,
1396 fSymbols->fShortWeekdaysCount);
1397 capContextUsageType = DateFormatSymbols::kCapContextUsageDayFormat;
1398 }
1399 break;
1401 // for "ccc", write out the abbreviated day-of-the-week name
1402 // for "cccc", write out the wide day-of-the-week name
1403 // for "ccccc", use the narrow day-of-the-week name
1404 // for "ccccc", use the short day-of-the-week name
1405 case UDAT_STANDALONE_DAY_FIELD:
1406 if ( count < 3 ) {
1407 zeroPaddingNumber(currentNumberFormat,appendTo, value, 1, maxIntCount);
1408 break;
1409 }
1410 // fall through to alpha DOW handling, but for that we don't want local day-of-week,
1411 // we want standard day-of-week, so first fix value.
1412 value = cal.get(UCAL_DAY_OF_WEEK, status);
1413 if (U_FAILURE(status)) {
1414 return;
1415 }
1416 if (count == 5) {
1417 _appendSymbol(appendTo, value, fSymbols->fStandaloneNarrowWeekdays,
1418 fSymbols->fStandaloneNarrowWeekdaysCount);
1419 capContextUsageType = DateFormatSymbols::kCapContextUsageDayNarrow;
1420 } else if (count == 4) {
1421 _appendSymbol(appendTo, value, fSymbols->fStandaloneWeekdays,
1422 fSymbols->fStandaloneWeekdaysCount);
1423 capContextUsageType = DateFormatSymbols::kCapContextUsageDayStandalone;
1424 } else if (count == 6) {
1425 _appendSymbol(appendTo, value, fSymbols->fStandaloneShorterWeekdays,
1426 fSymbols->fStandaloneShorterWeekdaysCount);
1427 capContextUsageType = DateFormatSymbols::kCapContextUsageDayStandalone;
1428 } else { // count == 3
1429 _appendSymbol(appendTo, value, fSymbols->fStandaloneShortWeekdays,
1430 fSymbols->fStandaloneShortWeekdaysCount);
1431 capContextUsageType = DateFormatSymbols::kCapContextUsageDayStandalone;
1432 }
1433 break;
1435 // for and "a" symbol, write out the whole AM/PM string
1436 case UDAT_AM_PM_FIELD:
1437 _appendSymbol(appendTo, value, fSymbols->fAmPms,
1438 fSymbols->fAmPmsCount);
1439 break;
1441 // for "h" and "hh", write out the hour, adjusting noon and midnight to show up
1442 // as "12"
1443 case UDAT_HOUR1_FIELD:
1444 if (value == 0)
1445 zeroPaddingNumber(currentNumberFormat,appendTo, cal.getLeastMaximum(UCAL_HOUR) + 1, count, maxIntCount);
1446 else
1447 zeroPaddingNumber(currentNumberFormat,appendTo, value, count, maxIntCount);
1448 break;
1450 case UDAT_TIMEZONE_FIELD: // 'z'
1451 case UDAT_TIMEZONE_RFC_FIELD: // 'Z'
1452 case UDAT_TIMEZONE_GENERIC_FIELD: // 'v'
1453 case UDAT_TIMEZONE_SPECIAL_FIELD: // 'V'
1454 case UDAT_TIMEZONE_LOCALIZED_GMT_OFFSET_FIELD: // 'O'
1455 case UDAT_TIMEZONE_ISO_FIELD: // 'X'
1456 case UDAT_TIMEZONE_ISO_LOCAL_FIELD: // 'x'
1457 {
1458 UnicodeString zoneString;
1459 const TimeZone& tz = cal.getTimeZone();
1460 UDate date = cal.getTime(status);
1461 if (U_SUCCESS(status)) {
1462 if (patternCharIndex == UDAT_TIMEZONE_FIELD) {
1463 if (count < 4) {
1464 // "z", "zz", "zzz"
1465 tzFormat()->format(UTZFMT_STYLE_SPECIFIC_SHORT, tz, date, zoneString);
1466 capContextUsageType = DateFormatSymbols::kCapContextUsageMetazoneShort;
1467 } else {
1468 // "zzzz" or longer
1469 tzFormat()->format(UTZFMT_STYLE_SPECIFIC_LONG, tz, date, zoneString);
1470 capContextUsageType = DateFormatSymbols::kCapContextUsageMetazoneLong;
1471 }
1472 }
1473 else if (patternCharIndex == UDAT_TIMEZONE_RFC_FIELD) {
1474 if (count < 4) {
1475 // "Z"
1476 tzFormat()->format(UTZFMT_STYLE_ISO_BASIC_LOCAL_FULL, tz, date, zoneString);
1477 } else if (count == 5) {
1478 // "ZZZZZ"
1479 tzFormat()->format(UTZFMT_STYLE_ISO_EXTENDED_FULL, tz, date, zoneString);
1480 } else {
1481 // "ZZ", "ZZZ", "ZZZZ"
1482 tzFormat()->format(UTZFMT_STYLE_LOCALIZED_GMT, tz, date, zoneString);
1483 }
1484 }
1485 else if (patternCharIndex == UDAT_TIMEZONE_GENERIC_FIELD) {
1486 if (count == 1) {
1487 // "v"
1488 tzFormat()->format(UTZFMT_STYLE_GENERIC_SHORT, tz, date, zoneString);
1489 capContextUsageType = DateFormatSymbols::kCapContextUsageMetazoneShort;
1490 } else if (count == 4) {
1491 // "vvvv"
1492 tzFormat()->format(UTZFMT_STYLE_GENERIC_LONG, tz, date, zoneString);
1493 capContextUsageType = DateFormatSymbols::kCapContextUsageMetazoneLong;
1494 }
1495 }
1496 else if (patternCharIndex == UDAT_TIMEZONE_SPECIAL_FIELD) {
1497 if (count == 1) {
1498 // "V"
1499 tzFormat()->format(UTZFMT_STYLE_ZONE_ID_SHORT, tz, date, zoneString);
1500 } else if (count == 2) {
1501 // "VV"
1502 tzFormat()->format(UTZFMT_STYLE_ZONE_ID, tz, date, zoneString);
1503 } else if (count == 3) {
1504 // "VVV"
1505 tzFormat()->format(UTZFMT_STYLE_EXEMPLAR_LOCATION, tz, date, zoneString);
1506 } else if (count == 4) {
1507 // "VVVV"
1508 tzFormat()->format(UTZFMT_STYLE_GENERIC_LOCATION, tz, date, zoneString);
1509 capContextUsageType = DateFormatSymbols::kCapContextUsageZoneLong;
1510 }
1511 }
1512 else if (patternCharIndex == UDAT_TIMEZONE_LOCALIZED_GMT_OFFSET_FIELD) {
1513 if (count == 1) {
1514 // "O"
1515 tzFormat()->format(UTZFMT_STYLE_LOCALIZED_GMT_SHORT, tz, date, zoneString);
1516 } else if (count == 4) {
1517 // "OOOO"
1518 tzFormat()->format(UTZFMT_STYLE_LOCALIZED_GMT, tz, date, zoneString);
1519 }
1520 }
1521 else if (patternCharIndex == UDAT_TIMEZONE_ISO_FIELD) {
1522 if (count == 1) {
1523 // "X"
1524 tzFormat()->format(UTZFMT_STYLE_ISO_BASIC_SHORT, tz, date, zoneString);
1525 } else if (count == 2) {
1526 // "XX"
1527 tzFormat()->format(UTZFMT_STYLE_ISO_BASIC_FIXED, tz, date, zoneString);
1528 } else if (count == 3) {
1529 // "XXX"
1530 tzFormat()->format(UTZFMT_STYLE_ISO_EXTENDED_FIXED, tz, date, zoneString);
1531 } else if (count == 4) {
1532 // "XXXX"
1533 tzFormat()->format(UTZFMT_STYLE_ISO_BASIC_FULL, tz, date, zoneString);
1534 } else if (count == 5) {
1535 // "XXXXX"
1536 tzFormat()->format(UTZFMT_STYLE_ISO_EXTENDED_FULL, tz, date, zoneString);
1537 }
1538 }
1539 else if (patternCharIndex == UDAT_TIMEZONE_ISO_LOCAL_FIELD) {
1540 if (count == 1) {
1541 // "x"
1542 tzFormat()->format(UTZFMT_STYLE_ISO_BASIC_LOCAL_SHORT, tz, date, zoneString);
1543 } else if (count == 2) {
1544 // "xx"
1545 tzFormat()->format(UTZFMT_STYLE_ISO_BASIC_LOCAL_FIXED, tz, date, zoneString);
1546 } else if (count == 3) {
1547 // "xxx"
1548 tzFormat()->format(UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FIXED, tz, date, zoneString);
1549 } else if (count == 4) {
1550 // "xxxx"
1551 tzFormat()->format(UTZFMT_STYLE_ISO_BASIC_LOCAL_FULL, tz, date, zoneString);
1552 } else if (count == 5) {
1553 // "xxxxx"
1554 tzFormat()->format(UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FULL, tz, date, zoneString);
1555 }
1556 }
1557 else {
1558 U_ASSERT(FALSE);
1559 }
1560 }
1561 appendTo += zoneString;
1562 }
1563 break;
1565 case UDAT_QUARTER_FIELD:
1566 if (count >= 4)
1567 _appendSymbol(appendTo, value/3, fSymbols->fQuarters,
1568 fSymbols->fQuartersCount);
1569 else if (count == 3)
1570 _appendSymbol(appendTo, value/3, fSymbols->fShortQuarters,
1571 fSymbols->fShortQuartersCount);
1572 else
1573 zeroPaddingNumber(currentNumberFormat,appendTo, (value/3) + 1, count, maxIntCount);
1574 break;
1576 case UDAT_STANDALONE_QUARTER_FIELD:
1577 if (count >= 4)
1578 _appendSymbol(appendTo, value/3, fSymbols->fStandaloneQuarters,
1579 fSymbols->fStandaloneQuartersCount);
1580 else if (count == 3)
1581 _appendSymbol(appendTo, value/3, fSymbols->fStandaloneShortQuarters,
1582 fSymbols->fStandaloneShortQuartersCount);
1583 else
1584 zeroPaddingNumber(currentNumberFormat,appendTo, (value/3) + 1, count, maxIntCount);
1585 break;
1588 // all of the other pattern symbols can be formatted as simple numbers with
1589 // appropriate zero padding
1590 default:
1591 zeroPaddingNumber(currentNumberFormat,appendTo, value, count, maxIntCount);
1592 break;
1593 }
1594 #if !UCONFIG_NO_BREAK_ITERATION
1595 if (fieldNum == 0) {
1596 // first field, check to see whether we need to titlecase it
1597 UBool titlecase = FALSE;
1598 switch (capitalizationContext) {
1599 case UDISPCTX_CAPITALIZATION_FOR_BEGINNING_OF_SENTENCE:
1600 titlecase = TRUE;
1601 break;
1602 case UDISPCTX_CAPITALIZATION_FOR_UI_LIST_OR_MENU:
1603 titlecase = fSymbols->fCapitalization[capContextUsageType][0];
1604 break;
1605 case UDISPCTX_CAPITALIZATION_FOR_STANDALONE:
1606 titlecase = fSymbols->fCapitalization[capContextUsageType][1];
1607 break;
1608 default:
1609 // titlecase = FALSE;
1610 break;
1611 }
1612 if (titlecase) {
1613 UnicodeString firstField(appendTo, beginOffset);
1614 firstField.toTitle(NULL, fLocale, U_TITLECASE_NO_LOWERCASE | U_TITLECASE_NO_BREAK_ADJUSTMENT);
1615 appendTo.replaceBetween(beginOffset, appendTo.length(), firstField);
1616 }
1617 }
1618 #endif
1620 handler.addAttribute(fgPatternIndexToDateFormatField[patternCharIndex], beginOffset, appendTo.length());
1621 }
1623 //----------------------------------------------------------------------
1625 NumberFormat *
1626 SimpleDateFormat::getNumberFormatByIndex(UDateFormatField index) const {
1627 if (fNumberFormatters != NULL) {
1628 return fNumberFormatters[index];
1629 } else {
1630 return fNumberFormat;
1631 }
1632 }
1634 //----------------------------------------------------------------------
1635 void
1636 SimpleDateFormat::zeroPaddingNumber(NumberFormat *currentNumberFormat,UnicodeString &appendTo,
1637 int32_t value, int32_t minDigits, int32_t maxDigits) const
1638 {
1639 if (currentNumberFormat!=NULL) {
1640 FieldPosition pos(0);
1642 currentNumberFormat->setMinimumIntegerDigits(minDigits);
1643 currentNumberFormat->setMaximumIntegerDigits(maxDigits);
1644 currentNumberFormat->format(value, appendTo, pos); // 3rd arg is there to speed up processing
1645 }
1646 }
1648 //----------------------------------------------------------------------
1650 /**
1651 * Return true if the given format character, occuring count
1652 * times, represents a numeric field.
1653 */
1654 UBool SimpleDateFormat::isNumeric(UChar formatChar, int32_t count) {
1655 return DateFormatSymbols::isNumericPatternChar(formatChar, count);
1656 }
1658 UBool
1659 SimpleDateFormat::isAtNumericField(const UnicodeString &pattern, int32_t patternOffset) {
1660 if (patternOffset >= pattern.length()) {
1661 // not at any field
1662 return FALSE;
1663 }
1664 UChar ch = pattern.charAt(patternOffset);
1665 UDateFormatField f = DateFormatSymbols::getPatternCharIndex(ch);
1666 if (f == UDAT_FIELD_COUNT) {
1667 // not at any field
1668 return FALSE;
1669 }
1670 int32_t i = patternOffset;
1671 while (pattern.charAt(++i) == ch) {}
1672 return DateFormatSymbols::isNumericField(f, i - patternOffset);
1673 }
1675 UBool
1676 SimpleDateFormat::isAfterNonNumericField(const UnicodeString &pattern, int32_t patternOffset) {
1677 if (patternOffset <= 0) {
1678 // not after any field
1679 return FALSE;
1680 }
1681 UChar ch = pattern.charAt(--patternOffset);
1682 UDateFormatField f = DateFormatSymbols::getPatternCharIndex(ch);
1683 if (f == UDAT_FIELD_COUNT) {
1684 // not after any field
1685 return FALSE;
1686 }
1687 int32_t i = patternOffset;
1688 while (pattern.charAt(--i) == ch) {}
1689 return !DateFormatSymbols::isNumericField(f, patternOffset - i);
1690 }
1692 void
1693 SimpleDateFormat::parse(const UnicodeString& text, Calendar& cal, ParsePosition& parsePos) const
1694 {
1695 UErrorCode status = U_ZERO_ERROR;
1696 int32_t pos = parsePos.getIndex();
1697 int32_t start = pos;
1699 UBool ambiguousYear[] = { FALSE };
1700 int32_t saveHebrewMonth = -1;
1701 int32_t count = 0;
1703 // hack, reset tztype, cast away const
1704 ((SimpleDateFormat*)this)->tztype = UTZFMT_TIME_TYPE_UNKNOWN;
1706 // For parsing abutting numeric fields. 'abutPat' is the
1707 // offset into 'pattern' of the first of 2 or more abutting
1708 // numeric fields. 'abutStart' is the offset into 'text'
1709 // where parsing the fields begins. 'abutPass' starts off as 0
1710 // and increments each time we try to parse the fields.
1711 int32_t abutPat = -1; // If >=0, we are in a run of abutting numeric fields
1712 int32_t abutStart = 0;
1713 int32_t abutPass = 0;
1714 UBool inQuote = FALSE;
1716 MessageFormat * numericLeapMonthFormatter = NULL;
1718 Calendar* calClone = NULL;
1719 Calendar *workCal = &cal;
1720 if (&cal != fCalendar && uprv_strcmp(cal.getType(), fCalendar->getType()) != 0) {
1721 // Different calendar type
1722 // We use the time/zone from the input calendar, but
1723 // do not use the input calendar for field calculation.
1724 calClone = fCalendar->clone();
1725 if (calClone != NULL) {
1726 calClone->setTime(cal.getTime(status),status);
1727 if (U_FAILURE(status)) {
1728 goto ExitParse;
1729 }
1730 calClone->setTimeZone(cal.getTimeZone());
1731 workCal = calClone;
1732 } else {
1733 status = U_MEMORY_ALLOCATION_ERROR;
1734 goto ExitParse;
1735 }
1736 }
1738 if (fSymbols->fLeapMonthPatterns != NULL && fSymbols->fLeapMonthPatternsCount >= DateFormatSymbols::kMonthPatternsCount) {
1739 numericLeapMonthFormatter = new MessageFormat(fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternNumeric], fLocale, status);
1740 if (numericLeapMonthFormatter == NULL) {
1741 status = U_MEMORY_ALLOCATION_ERROR;
1742 goto ExitParse;
1743 } else if (U_FAILURE(status)) {
1744 goto ExitParse; // this will delete numericLeapMonthFormatter
1745 }
1746 }
1748 for (int32_t i=0; i<fPattern.length(); ++i) {
1749 UChar ch = fPattern.charAt(i);
1751 // Handle alphabetic field characters.
1752 if (!inQuote && ((ch >= 0x41 && ch <= 0x5A) || (ch >= 0x61 && ch <= 0x7A))) { // [A-Za-z]
1753 int32_t fieldPat = i;
1755 // Count the length of this field specifier
1756 count = 1;
1757 while ((i+1)<fPattern.length() &&
1758 fPattern.charAt(i+1) == ch) {
1759 ++count;
1760 ++i;
1761 }
1763 if (isNumeric(ch, count)) {
1764 if (abutPat < 0) {
1765 // Determine if there is an abutting numeric field.
1766 // Record the start of a set of abutting numeric fields.
1767 if (isAtNumericField(fPattern, i + 1)) {
1768 abutPat = fieldPat;
1769 abutStart = pos;
1770 abutPass = 0;
1771 }
1772 }
1773 } else {
1774 abutPat = -1; // End of any abutting fields
1775 }
1777 // Handle fields within a run of abutting numeric fields. Take
1778 // the pattern "HHmmss" as an example. We will try to parse
1779 // 2/2/2 characters of the input text, then if that fails,
1780 // 1/2/2. We only adjust the width of the leftmost field; the
1781 // others remain fixed. This allows "123456" => 12:34:56, but
1782 // "12345" => 1:23:45. Likewise, for the pattern "yyyyMMdd" we
1783 // try 4/2/2, 3/2/2, 2/2/2, and finally 1/2/2.
1784 if (abutPat >= 0) {
1785 // If we are at the start of a run of abutting fields, then
1786 // shorten this field in each pass. If we can't shorten
1787 // this field any more, then the parse of this set of
1788 // abutting numeric fields has failed.
1789 if (fieldPat == abutPat) {
1790 count -= abutPass++;
1791 if (count == 0) {
1792 status = U_PARSE_ERROR;
1793 goto ExitParse;
1794 }
1795 }
1797 pos = subParse(text, pos, ch, count,
1798 TRUE, FALSE, ambiguousYear, saveHebrewMonth, *workCal, i, numericLeapMonthFormatter);
1800 // If the parse fails anywhere in the run, back up to the
1801 // start of the run and retry.
1802 if (pos < 0) {
1803 i = abutPat - 1;
1804 pos = abutStart;
1805 continue;
1806 }
1807 }
1809 // Handle non-numeric fields and non-abutting numeric
1810 // fields.
1811 else if (ch != 0x6C) { // pattern char 'l' (SMALL LETTER L) just gets ignored
1812 int32_t s = subParse(text, pos, ch, count,
1813 FALSE, TRUE, ambiguousYear, saveHebrewMonth, *workCal, i, numericLeapMonthFormatter);
1815 if (s == -pos-1) {
1816 // era not present, in special cases allow this to continue
1817 // from the position where the era was expected
1818 s = pos;
1820 if (i+1 < fPattern.length()) {
1821 // move to next pattern character
1822 UChar ch = fPattern.charAt(i+1);
1824 // check for whitespace
1825 if (PatternProps::isWhiteSpace(ch)) {
1826 i++;
1827 // Advance over run in pattern
1828 while ((i+1)<fPattern.length() &&
1829 PatternProps::isWhiteSpace(fPattern.charAt(i+1))) {
1830 ++i;
1831 }
1832 }
1833 }
1834 }
1835 else if (s <= 0) {
1836 status = U_PARSE_ERROR;
1837 goto ExitParse;
1838 }
1839 pos = s;
1840 }
1841 }
1843 // Handle literal pattern characters. These are any
1844 // quoted characters and non-alphabetic unquoted
1845 // characters.
1846 else {
1848 abutPat = -1; // End of any abutting fields
1850 if (! matchLiterals(fPattern, i, text, pos, getBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, status))) {
1851 status = U_PARSE_ERROR;
1852 goto ExitParse;
1853 }
1854 }
1855 }
1857 // Special hack for trailing "." after non-numeric field.
1858 if (text.charAt(pos) == 0x2e && getBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, status)) {
1859 // only do if the last field is not numeric
1860 if (isAfterNonNumericField(fPattern, fPattern.length())) {
1861 pos++; // skip the extra "."
1862 }
1863 }
1865 // At this point the fields of Calendar have been set. Calendar
1866 // will fill in default values for missing fields when the time
1867 // is computed.
1869 parsePos.setIndex(pos);
1871 // This part is a problem: When we call parsedDate.after, we compute the time.
1872 // Take the date April 3 2004 at 2:30 am. When this is first set up, the year
1873 // will be wrong if we're parsing a 2-digit year pattern. It will be 1904.
1874 // April 3 1904 is a Sunday (unlike 2004) so it is the DST onset day. 2:30 am
1875 // is therefore an "impossible" time, since the time goes from 1:59 to 3:00 am
1876 // on that day. It is therefore parsed out to fields as 3:30 am. Then we
1877 // add 100 years, and get April 3 2004 at 3:30 am. Note that April 3 2004 is
1878 // a Saturday, so it can have a 2:30 am -- and it should. [LIU]
1879 /*
1880 UDate parsedDate = calendar.getTime();
1881 if( ambiguousYear[0] && !parsedDate.after(fDefaultCenturyStart) ) {
1882 calendar.add(Calendar.YEAR, 100);
1883 parsedDate = calendar.getTime();
1884 }
1885 */
1886 // Because of the above condition, save off the fields in case we need to readjust.
1887 // The procedure we use here is not particularly efficient, but there is no other
1888 // way to do this given the API restrictions present in Calendar. We minimize
1889 // inefficiency by only performing this computation when it might apply, that is,
1890 // when the two-digit year is equal to the start year, and thus might fall at the
1891 // front or the back of the default century. This only works because we adjust
1892 // the year correctly to start with in other cases -- see subParse().
1893 if (ambiguousYear[0] || tztype != UTZFMT_TIME_TYPE_UNKNOWN) // If this is true then the two-digit year == the default start year
1894 {
1895 // We need a copy of the fields, and we need to avoid triggering a call to
1896 // complete(), which will recalculate the fields. Since we can't access
1897 // the fields[] array in Calendar, we clone the entire object. This will
1898 // stop working if Calendar.clone() is ever rewritten to call complete().
1899 Calendar *copy;
1900 if (ambiguousYear[0]) {
1901 copy = cal.clone();
1902 // Check for failed cloning.
1903 if (copy == NULL) {
1904 status = U_MEMORY_ALLOCATION_ERROR;
1905 goto ExitParse;
1906 }
1907 UDate parsedDate = copy->getTime(status);
1908 // {sfb} check internalGetDefaultCenturyStart
1909 if (fHaveDefaultCentury && (parsedDate < fDefaultCenturyStart)) {
1910 // We can't use add here because that does a complete() first.
1911 cal.set(UCAL_YEAR, fDefaultCenturyStartYear + 100);
1912 }
1913 delete copy;
1914 }
1916 if (tztype != UTZFMT_TIME_TYPE_UNKNOWN) {
1917 copy = cal.clone();
1918 // Check for failed cloning.
1919 if (copy == NULL) {
1920 status = U_MEMORY_ALLOCATION_ERROR;
1921 goto ExitParse;
1922 }
1923 const TimeZone & tz = cal.getTimeZone();
1924 BasicTimeZone *btz = NULL;
1926 if (dynamic_cast<const OlsonTimeZone *>(&tz) != NULL
1927 || dynamic_cast<const SimpleTimeZone *>(&tz) != NULL
1928 || dynamic_cast<const RuleBasedTimeZone *>(&tz) != NULL
1929 || dynamic_cast<const VTimeZone *>(&tz) != NULL) {
1930 btz = (BasicTimeZone*)&tz;
1931 }
1933 // Get local millis
1934 copy->set(UCAL_ZONE_OFFSET, 0);
1935 copy->set(UCAL_DST_OFFSET, 0);
1936 UDate localMillis = copy->getTime(status);
1938 // Make sure parsed time zone type (Standard or Daylight)
1939 // matches the rule used by the parsed time zone.
1940 int32_t raw, dst;
1941 if (btz != NULL) {
1942 if (tztype == UTZFMT_TIME_TYPE_STANDARD) {
1943 btz->getOffsetFromLocal(localMillis,
1944 BasicTimeZone::kStandard, BasicTimeZone::kStandard, raw, dst, status);
1945 } else {
1946 btz->getOffsetFromLocal(localMillis,
1947 BasicTimeZone::kDaylight, BasicTimeZone::kDaylight, raw, dst, status);
1948 }
1949 } else {
1950 // No good way to resolve ambiguous time at transition,
1951 // but following code work in most case.
1952 tz.getOffset(localMillis, TRUE, raw, dst, status);
1953 }
1955 // Now, compare the results with parsed type, either standard or daylight saving time
1956 int32_t resolvedSavings = dst;
1957 if (tztype == UTZFMT_TIME_TYPE_STANDARD) {
1958 if (dst != 0) {
1959 // Override DST_OFFSET = 0 in the result calendar
1960 resolvedSavings = 0;
1961 }
1962 } else { // tztype == TZTYPE_DST
1963 if (dst == 0) {
1964 if (btz != NULL) {
1965 UDate time = localMillis + raw;
1966 // We use the nearest daylight saving time rule.
1967 TimeZoneTransition beforeTrs, afterTrs;
1968 UDate beforeT = time, afterT = time;
1969 int32_t beforeSav = 0, afterSav = 0;
1970 UBool beforeTrsAvail, afterTrsAvail;
1972 // Search for DST rule before or on the time
1973 while (TRUE) {
1974 beforeTrsAvail = btz->getPreviousTransition(beforeT, TRUE, beforeTrs);
1975 if (!beforeTrsAvail) {
1976 break;
1977 }
1978 beforeT = beforeTrs.getTime() - 1;
1979 beforeSav = beforeTrs.getFrom()->getDSTSavings();
1980 if (beforeSav != 0) {
1981 break;
1982 }
1983 }
1985 // Search for DST rule after the time
1986 while (TRUE) {
1987 afterTrsAvail = btz->getNextTransition(afterT, FALSE, afterTrs);
1988 if (!afterTrsAvail) {
1989 break;
1990 }
1991 afterT = afterTrs.getTime();
1992 afterSav = afterTrs.getTo()->getDSTSavings();
1993 if (afterSav != 0) {
1994 break;
1995 }
1996 }
1998 if (beforeTrsAvail && afterTrsAvail) {
1999 if (time - beforeT > afterT - time) {
2000 resolvedSavings = afterSav;
2001 } else {
2002 resolvedSavings = beforeSav;
2003 }
2004 } else if (beforeTrsAvail && beforeSav != 0) {
2005 resolvedSavings = beforeSav;
2006 } else if (afterTrsAvail && afterSav != 0) {
2007 resolvedSavings = afterSav;
2008 } else {
2009 resolvedSavings = btz->getDSTSavings();
2010 }
2011 } else {
2012 resolvedSavings = tz.getDSTSavings();
2013 }
2014 if (resolvedSavings == 0) {
2015 // final fallback
2016 resolvedSavings = U_MILLIS_PER_HOUR;
2017 }
2018 }
2019 }
2020 cal.set(UCAL_ZONE_OFFSET, raw);
2021 cal.set(UCAL_DST_OFFSET, resolvedSavings);
2022 delete copy;
2023 }
2024 }
2025 ExitParse:
2026 // Set the parsed result if local calendar is used
2027 // instead of the input calendar
2028 if (U_SUCCESS(status) && workCal != &cal) {
2029 cal.setTimeZone(workCal->getTimeZone());
2030 cal.setTime(workCal->getTime(status), status);
2031 }
2033 if (numericLeapMonthFormatter != NULL) {
2034 delete numericLeapMonthFormatter;
2035 }
2036 if (calClone != NULL) {
2037 delete calClone;
2038 }
2040 // If any Calendar calls failed, we pretend that we
2041 // couldn't parse the string, when in reality this isn't quite accurate--
2042 // we did parse it; the Calendar calls just failed.
2043 if (U_FAILURE(status)) {
2044 parsePos.setErrorIndex(pos);
2045 parsePos.setIndex(start);
2046 }
2047 }
2049 //----------------------------------------------------------------------
2051 static UBool
2052 newBestMatchWithOptionalDot(const UnicodeString &lcaseText,
2053 const UnicodeString &data,
2054 UnicodeString &bestMatchName,
2055 int32_t &bestMatchLength);
2057 int32_t SimpleDateFormat::matchQuarterString(const UnicodeString& text,
2058 int32_t start,
2059 UCalendarDateFields field,
2060 const UnicodeString* data,
2061 int32_t dataCount,
2062 Calendar& cal) const
2063 {
2064 int32_t i = 0;
2065 int32_t count = dataCount;
2067 // There may be multiple strings in the data[] array which begin with
2068 // the same prefix (e.g., Cerven and Cervenec (June and July) in Czech).
2069 // We keep track of the longest match, and return that. Note that this
2070 // unfortunately requires us to test all array elements.
2071 int32_t bestMatchLength = 0, bestMatch = -1;
2072 UnicodeString bestMatchName;
2074 // {sfb} kludge to support case-insensitive comparison
2075 // {markus 2002oct11} do not just use caseCompareBetween because we do not know
2076 // the length of the match after case folding
2077 // {alan 20040607} don't case change the whole string, since the length
2078 // can change
2079 // TODO we need a case-insensitive startsWith function
2080 UnicodeString lcaseText;
2081 text.extract(start, INT32_MAX, lcaseText);
2082 lcaseText.foldCase();
2084 for (; i < count; ++i)
2085 {
2086 // Always compare if we have no match yet; otherwise only compare
2087 // against potentially better matches (longer strings).
2089 if (newBestMatchWithOptionalDot(lcaseText, data[i], bestMatchName, bestMatchLength)) {
2090 bestMatch = i;
2091 }
2092 }
2093 if (bestMatch >= 0)
2094 {
2095 cal.set(field, bestMatch * 3);
2097 // Once we have a match, we have to determine the length of the
2098 // original source string. This will usually be == the length of
2099 // the case folded string, but it may differ (e.g. sharp s).
2101 // Most of the time, the length will be the same as the length
2102 // of the string from the locale data. Sometimes it will be
2103 // different, in which case we will have to figure it out by
2104 // adding a character at a time, until we have a match. We do
2105 // this all in one loop, where we try 'len' first (at index
2106 // i==0).
2107 int32_t len = bestMatchName.length(); // 99+% of the time
2108 int32_t n = text.length() - start;
2109 for (i=0; i<=n; ++i) {
2110 int32_t j=i;
2111 if (i == 0) {
2112 j = len;
2113 } else if (i == len) {
2114 continue; // already tried this when i was 0
2115 }
2116 text.extract(start, j, lcaseText);
2117 lcaseText.foldCase();
2118 if (bestMatchName == lcaseText) {
2119 return start + j;
2120 }
2121 }
2122 }
2124 return -start;
2125 }
2127 //----------------------------------------------------------------------
2128 UBool SimpleDateFormat::matchLiterals(const UnicodeString &pattern,
2129 int32_t &patternOffset,
2130 const UnicodeString &text,
2131 int32_t &textOffset,
2132 UBool lenient)
2133 {
2134 UBool inQuote = FALSE;
2135 UnicodeString literal;
2136 int32_t i = patternOffset;
2138 // scan pattern looking for contiguous literal characters
2139 for ( ; i < pattern.length(); i += 1) {
2140 UChar ch = pattern.charAt(i);
2142 if (!inQuote && ((ch >= 0x41 && ch <= 0x5A) || (ch >= 0x61 && ch <= 0x7A))) { // unquoted [A-Za-z]
2143 break;
2144 }
2146 if (ch == QUOTE) {
2147 // Match a quote literal ('') inside OR outside of quotes
2148 if ((i + 1) < pattern.length() && pattern.charAt(i + 1) == QUOTE) {
2149 i += 1;
2150 } else {
2151 inQuote = !inQuote;
2152 continue;
2153 }
2154 }
2156 literal += ch;
2157 }
2159 // at this point, literal contains the literal text
2160 // and i is the index of the next non-literal pattern character.
2161 int32_t p;
2162 int32_t t = textOffset;
2164 if (lenient) {
2165 // trim leading, trailing whitespace from
2166 // the literal text
2167 literal.trim();
2169 // ignore any leading whitespace in the text
2170 while (t < text.length() && u_isWhitespace(text.charAt(t))) {
2171 t += 1;
2172 }
2173 }
2175 for (p = 0; p < literal.length() && t < text.length();) {
2176 UBool needWhitespace = FALSE;
2178 while (p < literal.length() && PatternProps::isWhiteSpace(literal.charAt(p))) {
2179 needWhitespace = TRUE;
2180 p += 1;
2181 }
2183 if (needWhitespace) {
2184 int32_t tStart = t;
2186 while (t < text.length()) {
2187 UChar tch = text.charAt(t);
2189 if (!u_isUWhiteSpace(tch) && !PatternProps::isWhiteSpace(tch)) {
2190 break;
2191 }
2193 t += 1;
2194 }
2196 // TODO: should we require internal spaces
2197 // in lenient mode? (There won't be any
2198 // leading or trailing spaces)
2199 if (!lenient && t == tStart) {
2200 // didn't find matching whitespace:
2201 // an error in strict mode
2202 return FALSE;
2203 }
2205 // In strict mode, this run of whitespace
2206 // may have been at the end.
2207 if (p >= literal.length()) {
2208 break;
2209 }
2210 }
2212 if (t >= text.length() || literal.charAt(p) != text.charAt(t)) {
2213 // Ran out of text, or found a non-matching character:
2214 // OK in lenient mode, an error in strict mode.
2215 if (lenient) {
2216 if (t == textOffset && text.charAt(t) == 0x2e &&
2217 isAfterNonNumericField(pattern, patternOffset)) {
2218 // Lenient mode and the literal input text begins with a "." and
2219 // we are after a non-numeric field: We skip the "."
2220 ++t;
2221 continue; // Do not update p.
2222 }
2223 break;
2224 }
2226 return FALSE;
2227 }
2228 ++p;
2229 ++t;
2230 }
2232 // At this point if we're in strict mode we have a complete match.
2233 // If we're in lenient mode we may have a partial match, or no
2234 // match at all.
2235 if (p <= 0) {
2236 // no match. Pretend it matched a run of whitespace
2237 // and ignorables in the text.
2238 const UnicodeSet *ignorables = NULL;
2239 UDateFormatField patternCharIndex = DateFormatSymbols::getPatternCharIndex(pattern.charAt(i));
2240 if (patternCharIndex != UDAT_FIELD_COUNT) {
2241 ignorables = SimpleDateFormatStaticSets::getIgnorables(patternCharIndex);
2242 }
2244 for (t = textOffset; t < text.length(); t += 1) {
2245 UChar ch = text.charAt(t);
2247 if (ignorables == NULL || !ignorables->contains(ch)) {
2248 break;
2249 }
2250 }
2251 }
2253 // if we get here, we've got a complete match.
2254 patternOffset = i - 1;
2255 textOffset = t;
2257 return TRUE;
2258 }
2260 //----------------------------------------------------------------------
2262 int32_t SimpleDateFormat::matchString(const UnicodeString& text,
2263 int32_t start,
2264 UCalendarDateFields field,
2265 const UnicodeString* data,
2266 int32_t dataCount,
2267 const UnicodeString* monthPattern,
2268 Calendar& cal) const
2269 {
2270 int32_t i = 0;
2271 int32_t count = dataCount;
2273 if (field == UCAL_DAY_OF_WEEK) i = 1;
2275 // There may be multiple strings in the data[] array which begin with
2276 // the same prefix (e.g., Cerven and Cervenec (June and July) in Czech).
2277 // We keep track of the longest match, and return that. Note that this
2278 // unfortunately requires us to test all array elements.
2279 int32_t bestMatchLength = 0, bestMatch = -1;
2280 UnicodeString bestMatchName;
2281 int32_t isLeapMonth = 0;
2283 // {sfb} kludge to support case-insensitive comparison
2284 // {markus 2002oct11} do not just use caseCompareBetween because we do not know
2285 // the length of the match after case folding
2286 // {alan 20040607} don't case change the whole string, since the length
2287 // can change
2288 // TODO we need a case-insensitive startsWith function
2289 UnicodeString lcaseText;
2290 text.extract(start, INT32_MAX, lcaseText);
2291 lcaseText.foldCase();
2293 for (; i < count; ++i)
2294 {
2295 // Always compare if we have no match yet; otherwise only compare
2296 // against potentially better matches (longer strings).
2298 if (newBestMatchWithOptionalDot(lcaseText, data[i], bestMatchName, bestMatchLength)) {
2299 bestMatch = i;
2300 isLeapMonth = 0;
2301 }
2303 if (monthPattern != NULL) {
2304 UErrorCode status = U_ZERO_ERROR;
2305 UnicodeString leapMonthName;
2306 Formattable monthName((const UnicodeString&)(data[i]));
2307 MessageFormat::format(*monthPattern, &monthName, 1, leapMonthName, status);
2308 if (U_SUCCESS(status)) {
2309 if (newBestMatchWithOptionalDot(lcaseText, leapMonthName, bestMatchName, bestMatchLength)) {
2310 bestMatch = i;
2311 isLeapMonth = 1;
2312 }
2313 }
2314 }
2315 }
2316 if (bestMatch >= 0)
2317 {
2318 // Adjustment for Hebrew Calendar month Adar II
2319 if (!strcmp(cal.getType(),"hebrew") && field==UCAL_MONTH && bestMatch==13) {
2320 cal.set(field,6);
2321 }
2322 else {
2323 if (field == UCAL_YEAR) {
2324 bestMatch++; // only get here for cyclic year names, which match 1-based years 1-60
2325 }
2326 cal.set(field, bestMatch);
2327 }
2328 if (monthPattern != NULL) {
2329 cal.set(UCAL_IS_LEAP_MONTH, isLeapMonth);
2330 }
2332 // Once we have a match, we have to determine the length of the
2333 // original source string. This will usually be == the length of
2334 // the case folded string, but it may differ (e.g. sharp s).
2336 // Most of the time, the length will be the same as the length
2337 // of the string from the locale data. Sometimes it will be
2338 // different, in which case we will have to figure it out by
2339 // adding a character at a time, until we have a match. We do
2340 // this all in one loop, where we try 'len' first (at index
2341 // i==0).
2342 int32_t len = bestMatchName.length(); // 99+% of the time
2343 int32_t n = text.length() - start;
2344 for (i=0; i<=n; ++i) {
2345 int32_t j=i;
2346 if (i == 0) {
2347 j = len;
2348 } else if (i == len) {
2349 continue; // already tried this when i was 0
2350 }
2351 text.extract(start, j, lcaseText);
2352 lcaseText.foldCase();
2353 if (bestMatchName == lcaseText) {
2354 return start + j;
2355 }
2356 }
2357 }
2359 return -start;
2360 }
2362 static UBool
2363 newBestMatchWithOptionalDot(const UnicodeString &lcaseText,
2364 const UnicodeString &data,
2365 UnicodeString &bestMatchName,
2366 int32_t &bestMatchLength) {
2367 UnicodeString lcase;
2368 lcase.fastCopyFrom(data).foldCase();
2369 int32_t length = lcase.length();
2370 if (length <= bestMatchLength) {
2371 // data cannot provide a better match.
2372 return FALSE;
2373 }
2375 if (lcaseText.compareBetween(0, length, lcase, 0, length) == 0) {
2376 // normal match
2377 bestMatchName = lcase;
2378 bestMatchLength = length;
2379 return TRUE;
2380 }
2381 if (lcase.charAt(--length) == 0x2e) {
2382 if (lcaseText.compareBetween(0, length, lcase, 0, length) == 0) {
2383 // The input text matches the data except for data's trailing dot.
2384 bestMatchName = lcase;
2385 bestMatchName.truncate(length);
2386 bestMatchLength = length;
2387 return TRUE;
2388 }
2389 }
2390 return FALSE;
2391 }
2393 //----------------------------------------------------------------------
2395 void
2396 SimpleDateFormat::set2DigitYearStart(UDate d, UErrorCode& status)
2397 {
2398 parseAmbiguousDatesAsAfter(d, status);
2399 }
2401 /**
2402 * Private member function that converts the parsed date strings into
2403 * timeFields. Returns -start (for ParsePosition) if failed.
2404 * @param text the time text to be parsed.
2405 * @param start where to start parsing.
2406 * @param ch the pattern character for the date field text to be parsed.
2407 * @param count the count of a pattern character.
2408 * @return the new start position if matching succeeded; a negative number
2409 * indicating matching failure, otherwise.
2410 */
2411 int32_t SimpleDateFormat::subParse(const UnicodeString& text, int32_t& start, UChar ch, int32_t count,
2412 UBool obeyCount, UBool allowNegative, UBool ambiguousYear[], int32_t& saveHebrewMonth, Calendar& cal,
2413 int32_t patLoc, MessageFormat * numericLeapMonthFormatter) const
2414 {
2415 Formattable number;
2416 int32_t value = 0;
2417 int32_t i;
2418 int32_t ps = 0;
2419 UErrorCode status = U_ZERO_ERROR;
2420 ParsePosition pos(0);
2421 UDateFormatField patternCharIndex = DateFormatSymbols::getPatternCharIndex(ch);
2422 NumberFormat *currentNumberFormat;
2423 UnicodeString temp;
2424 UBool gotNumber = FALSE;
2426 #if defined (U_DEBUG_CAL)
2427 //fprintf(stderr, "%s:%d - [%c] st=%d \n", __FILE__, __LINE__, (char) ch, start);
2428 #endif
2430 if (patternCharIndex == UDAT_FIELD_COUNT) {
2431 return -start;
2432 }
2434 currentNumberFormat = getNumberFormatByIndex(patternCharIndex);
2435 UCalendarDateFields field = fgPatternIndexToCalendarField[patternCharIndex];
2436 UnicodeString hebr("hebr", 4, US_INV);
2438 if (numericLeapMonthFormatter != NULL) {
2439 numericLeapMonthFormatter->setFormats((const Format **)¤tNumberFormat, 1);
2440 }
2441 UBool isChineseCalendar = (uprv_strcmp(cal.getType(),"chinese") == 0 || uprv_strcmp(cal.getType(),"dangi") == 0);
2443 // If there are any spaces here, skip over them. If we hit the end
2444 // of the string, then fail.
2445 for (;;) {
2446 if (start >= text.length()) {
2447 return -start;
2448 }
2449 UChar32 c = text.char32At(start);
2450 if (!u_isUWhiteSpace(c) /*||*/ && !PatternProps::isWhiteSpace(c)) {
2451 break;
2452 }
2453 start += U16_LENGTH(c);
2454 }
2455 pos.setIndex(start);
2457 // We handle a few special cases here where we need to parse
2458 // a number value. We handle further, more generic cases below. We need
2459 // to handle some of them here because some fields require extra processing on
2460 // the parsed value.
2461 if (patternCharIndex == UDAT_HOUR_OF_DAY1_FIELD || // k
2462 patternCharIndex == UDAT_HOUR_OF_DAY0_FIELD || // H
2463 patternCharIndex == UDAT_HOUR1_FIELD || // h
2464 patternCharIndex == UDAT_HOUR0_FIELD || // K
2465 (patternCharIndex == UDAT_DOW_LOCAL_FIELD && count <= 2) || // e
2466 (patternCharIndex == UDAT_STANDALONE_DAY_FIELD && count <= 2) || // c
2467 (patternCharIndex == UDAT_MONTH_FIELD && count <= 2) || // M
2468 (patternCharIndex == UDAT_STANDALONE_MONTH_FIELD && count <= 2) || // L
2469 (patternCharIndex == UDAT_QUARTER_FIELD && count <= 2) || // Q
2470 (patternCharIndex == UDAT_STANDALONE_QUARTER_FIELD && count <= 2) || // q
2471 patternCharIndex == UDAT_YEAR_FIELD || // y
2472 patternCharIndex == UDAT_YEAR_WOY_FIELD || // Y
2473 patternCharIndex == UDAT_YEAR_NAME_FIELD || // U (falls back to numeric)
2474 (patternCharIndex == UDAT_ERA_FIELD && isChineseCalendar) || // G
2475 patternCharIndex == UDAT_FRACTIONAL_SECOND_FIELD) // S
2476 {
2477 int32_t parseStart = pos.getIndex();
2478 // It would be good to unify this with the obeyCount logic below,
2479 // but that's going to be difficult.
2480 const UnicodeString* src;
2482 UBool parsedNumericLeapMonth = FALSE;
2483 if (numericLeapMonthFormatter != NULL && (patternCharIndex == UDAT_MONTH_FIELD || patternCharIndex == UDAT_STANDALONE_MONTH_FIELD)) {
2484 int32_t argCount;
2485 Formattable * args = numericLeapMonthFormatter->parse(text, pos, argCount);
2486 if (args != NULL && argCount == 1 && pos.getIndex() > parseStart && args[0].isNumeric()) {
2487 parsedNumericLeapMonth = TRUE;
2488 number.setLong(args[0].getLong());
2489 cal.set(UCAL_IS_LEAP_MONTH, 1);
2490 delete[] args;
2491 } else {
2492 pos.setIndex(parseStart);
2493 cal.set(UCAL_IS_LEAP_MONTH, 0);
2494 }
2495 }
2497 if (!parsedNumericLeapMonth) {
2498 if (obeyCount) {
2499 if ((start+count) > text.length()) {
2500 return -start;
2501 }
2503 text.extractBetween(0, start + count, temp);
2504 src = &temp;
2505 } else {
2506 src = &text;
2507 }
2509 parseInt(*src, number, pos, allowNegative,currentNumberFormat);
2510 }
2512 int32_t txtLoc = pos.getIndex();
2514 if (txtLoc > parseStart) {
2515 value = number.getLong();
2516 gotNumber = TRUE;
2518 // suffix processing
2519 if (value < 0 ) {
2520 txtLoc = checkIntSuffix(text, txtLoc, patLoc+1, TRUE);
2521 if (txtLoc != pos.getIndex()) {
2522 value *= -1;
2523 }
2524 }
2525 else {
2526 txtLoc = checkIntSuffix(text, txtLoc, patLoc+1, FALSE);
2527 }
2529 if (!getBooleanAttribute(UDAT_PARSE_ALLOW_WHITESPACE, status)) {
2530 // Check the range of the value
2531 int32_t bias = gFieldRangeBias[patternCharIndex];
2532 if (bias >= 0 && (value > cal.getMaximum(field) + bias || value < cal.getMinimum(field) + bias)) {
2533 return -start;
2534 }
2535 }
2537 pos.setIndex(txtLoc);
2538 }
2539 }
2541 // Make sure that we got a number if
2542 // we want one, and didn't get one
2543 // if we don't want one.
2544 switch (patternCharIndex) {
2545 case UDAT_HOUR_OF_DAY1_FIELD:
2546 case UDAT_HOUR_OF_DAY0_FIELD:
2547 case UDAT_HOUR1_FIELD:
2548 case UDAT_HOUR0_FIELD:
2549 // special range check for hours:
2550 if (value < 0 || value > 24) {
2551 return -start;
2552 }
2554 // fall through to gotNumber check
2556 case UDAT_YEAR_FIELD:
2557 case UDAT_YEAR_WOY_FIELD:
2558 case UDAT_FRACTIONAL_SECOND_FIELD:
2559 // these must be a number
2560 if (! gotNumber) {
2561 return -start;
2562 }
2564 break;
2566 default:
2567 // we check the rest of the fields below.
2568 break;
2569 }
2571 switch (patternCharIndex) {
2572 case UDAT_ERA_FIELD:
2573 if (isChineseCalendar) {
2574 if (!gotNumber) {
2575 return -start;
2576 }
2577 cal.set(UCAL_ERA, value);
2578 return pos.getIndex();
2579 }
2580 if (count == 5) {
2581 ps = matchString(text, start, UCAL_ERA, fSymbols->fNarrowEras, fSymbols->fNarrowErasCount, NULL, cal);
2582 } else if (count == 4) {
2583 ps = matchString(text, start, UCAL_ERA, fSymbols->fEraNames, fSymbols->fEraNamesCount, NULL, cal);
2584 } else {
2585 ps = matchString(text, start, UCAL_ERA, fSymbols->fEras, fSymbols->fErasCount, NULL, cal);
2586 }
2588 // check return position, if it equals -start, then matchString error
2589 // special case the return code so we don't necessarily fail out until we
2590 // verify no year information also
2591 if (ps == -start)
2592 ps--;
2594 return ps;
2596 case UDAT_YEAR_FIELD:
2597 // If there are 3 or more YEAR pattern characters, this indicates
2598 // that the year value is to be treated literally, without any
2599 // two-digit year adjustments (e.g., from "01" to 2001). Otherwise
2600 // we made adjustments to place the 2-digit year in the proper
2601 // century, for parsed strings from "00" to "99". Any other string
2602 // is treated literally: "2250", "-1", "1", "002".
2603 if (fDateOverride.compare(hebr)==0 && value < 1000) {
2604 value += HEBREW_CAL_CUR_MILLENIUM_START_YEAR;
2605 } else if ((pos.getIndex() - start) == 2 && !isChineseCalendar
2606 && u_isdigit(text.charAt(start))
2607 && u_isdigit(text.charAt(start+1)))
2608 {
2609 // only adjust year for patterns less than 3.
2610 if(count < 3) {
2611 // Assume for example that the defaultCenturyStart is 6/18/1903.
2612 // This means that two-digit years will be forced into the range
2613 // 6/18/1903 to 6/17/2003. As a result, years 00, 01, and 02
2614 // correspond to 2000, 2001, and 2002. Years 04, 05, etc. correspond
2615 // to 1904, 1905, etc. If the year is 03, then it is 2003 if the
2616 // other fields specify a date before 6/18, or 1903 if they specify a
2617 // date afterwards. As a result, 03 is an ambiguous year. All other
2618 // two-digit years are unambiguous.
2619 if(fHaveDefaultCentury) { // check if this formatter even has a pivot year
2620 int32_t ambiguousTwoDigitYear = fDefaultCenturyStartYear % 100;
2621 ambiguousYear[0] = (value == ambiguousTwoDigitYear);
2622 value += (fDefaultCenturyStartYear/100)*100 +
2623 (value < ambiguousTwoDigitYear ? 100 : 0);
2624 }
2625 }
2626 }
2627 cal.set(UCAL_YEAR, value);
2629 // Delayed checking for adjustment of Hebrew month numbers in non-leap years.
2630 if (saveHebrewMonth >= 0) {
2631 HebrewCalendar *hc = (HebrewCalendar*)&cal;
2632 if (!hc->isLeapYear(value) && saveHebrewMonth >= 6) {
2633 cal.set(UCAL_MONTH,saveHebrewMonth);
2634 } else {
2635 cal.set(UCAL_MONTH,saveHebrewMonth-1);
2636 }
2637 saveHebrewMonth = -1;
2638 }
2639 return pos.getIndex();
2641 case UDAT_YEAR_WOY_FIELD:
2642 // Comment is the same as for UDAT_Year_FIELDs - look above
2643 if (fDateOverride.compare(hebr)==0 && value < 1000) {
2644 value += HEBREW_CAL_CUR_MILLENIUM_START_YEAR;
2645 } else if ((pos.getIndex() - start) == 2
2646 && u_isdigit(text.charAt(start))
2647 && u_isdigit(text.charAt(start+1))
2648 && fHaveDefaultCentury )
2649 {
2650 int32_t ambiguousTwoDigitYear = fDefaultCenturyStartYear % 100;
2651 ambiguousYear[0] = (value == ambiguousTwoDigitYear);
2652 value += (fDefaultCenturyStartYear/100)*100 +
2653 (value < ambiguousTwoDigitYear ? 100 : 0);
2654 }
2655 cal.set(UCAL_YEAR_WOY, value);
2656 return pos.getIndex();
2658 case UDAT_YEAR_NAME_FIELD:
2659 if (fSymbols->fShortYearNames != NULL) {
2660 int32_t newStart = matchString(text, start, UCAL_YEAR, fSymbols->fShortYearNames, fSymbols->fShortYearNamesCount, NULL, cal);
2661 if (newStart > 0) {
2662 return newStart;
2663 }
2664 }
2665 if (gotNumber && (getBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC,status) || value > fSymbols->fShortYearNamesCount)) {
2666 cal.set(UCAL_YEAR, value);
2667 return pos.getIndex();
2668 }
2669 return -start;
2671 case UDAT_MONTH_FIELD:
2672 case UDAT_STANDALONE_MONTH_FIELD:
2673 if (gotNumber) // i.e., M or MM.
2674 {
2675 // When parsing month numbers from the Hebrew Calendar, we might need to adjust the month depending on whether
2676 // or not it was a leap year. We may or may not yet know what year it is, so might have to delay checking until
2677 // the year is parsed.
2678 if (!strcmp(cal.getType(),"hebrew")) {
2679 HebrewCalendar *hc = (HebrewCalendar*)&cal;
2680 if (cal.isSet(UCAL_YEAR)) {
2681 UErrorCode status = U_ZERO_ERROR;
2682 if (!hc->isLeapYear(hc->get(UCAL_YEAR,status)) && value >= 6) {
2683 cal.set(UCAL_MONTH, value);
2684 } else {
2685 cal.set(UCAL_MONTH, value - 1);
2686 }
2687 } else {
2688 saveHebrewMonth = value;
2689 }
2690 } else {
2691 // Don't want to parse the month if it is a string
2692 // while pattern uses numeric style: M/MM, L/LL
2693 // [We computed 'value' above.]
2694 cal.set(UCAL_MONTH, value - 1);
2695 }
2696 return pos.getIndex();
2697 } else {
2698 // count >= 3 // i.e., MMM/MMMM, LLL/LLLL
2699 // Want to be able to parse both short and long forms.
2700 // Try count == 4 first:
2701 UnicodeString * wideMonthPat = NULL;
2702 UnicodeString * shortMonthPat = NULL;
2703 if (fSymbols->fLeapMonthPatterns != NULL && fSymbols->fLeapMonthPatternsCount >= DateFormatSymbols::kMonthPatternsCount) {
2704 if (patternCharIndex==UDAT_MONTH_FIELD) {
2705 wideMonthPat = &fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternFormatWide];
2706 shortMonthPat = &fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternFormatAbbrev];
2707 } else {
2708 wideMonthPat = &fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternStandaloneWide];
2709 shortMonthPat = &fSymbols->fLeapMonthPatterns[DateFormatSymbols::kLeapMonthPatternStandaloneAbbrev];
2710 }
2711 }
2712 int32_t newStart = 0;
2713 if (patternCharIndex==UDAT_MONTH_FIELD) {
2714 newStart = matchString(text, start, UCAL_MONTH, fSymbols->fMonths, fSymbols->fMonthsCount, wideMonthPat, cal); // try MMMM
2715 if (newStart > 0) {
2716 return newStart;
2717 }
2718 newStart = matchString(text, start, UCAL_MONTH, fSymbols->fShortMonths, fSymbols->fShortMonthsCount, shortMonthPat, cal); // try MMM
2719 } else {
2720 newStart = matchString(text, start, UCAL_MONTH, fSymbols->fStandaloneMonths, fSymbols->fStandaloneMonthsCount, wideMonthPat, cal); // try LLLL
2721 if (newStart > 0) {
2722 return newStart;
2723 }
2724 newStart = matchString(text, start, UCAL_MONTH, fSymbols->fStandaloneShortMonths, fSymbols->fStandaloneShortMonthsCount, shortMonthPat, cal); // try LLL
2725 }
2726 if (newStart > 0 || !getBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, status)) // currently we do not try to parse MMMMM/LLLLL: #8860
2727 return newStart;
2728 // else we allowing parsing as number, below
2729 }
2730 break;
2732 case UDAT_HOUR_OF_DAY1_FIELD:
2733 // [We computed 'value' above.]
2734 if (value == cal.getMaximum(UCAL_HOUR_OF_DAY) + 1)
2735 value = 0;
2737 // fall through to set field
2739 case UDAT_HOUR_OF_DAY0_FIELD:
2740 cal.set(UCAL_HOUR_OF_DAY, value);
2741 return pos.getIndex();
2743 case UDAT_FRACTIONAL_SECOND_FIELD:
2744 // Fractional seconds left-justify
2745 i = pos.getIndex() - start;
2746 if (i < 3) {
2747 while (i < 3) {
2748 value *= 10;
2749 i++;
2750 }
2751 } else {
2752 int32_t a = 1;
2753 while (i > 3) {
2754 a *= 10;
2755 i--;
2756 }
2757 value /= a;
2758 }
2759 cal.set(UCAL_MILLISECOND, value);
2760 return pos.getIndex();
2762 case UDAT_DOW_LOCAL_FIELD:
2763 if (gotNumber) // i.e., e or ee
2764 {
2765 // [We computed 'value' above.]
2766 cal.set(UCAL_DOW_LOCAL, value);
2767 return pos.getIndex();
2768 }
2769 // else for eee-eeeee fall through to handling of EEE-EEEEE
2770 // fall through, do not break here
2771 case UDAT_DAY_OF_WEEK_FIELD:
2772 {
2773 // Want to be able to parse both short and long forms.
2774 // Try count == 4 (EEEE) wide first:
2775 int32_t newStart = 0;
2776 if ((newStart = matchString(text, start, UCAL_DAY_OF_WEEK,
2777 fSymbols->fWeekdays, fSymbols->fWeekdaysCount, NULL, cal)) > 0)
2778 return newStart;
2779 // EEEE wide failed, now try EEE abbreviated
2780 else if ((newStart = matchString(text, start, UCAL_DAY_OF_WEEK,
2781 fSymbols->fShortWeekdays, fSymbols->fShortWeekdaysCount, NULL, cal)) > 0)
2782 return newStart;
2783 // EEE abbreviated failed, now try EEEEEE short
2784 else if ((newStart = matchString(text, start, UCAL_DAY_OF_WEEK,
2785 fSymbols->fShorterWeekdays, fSymbols->fShorterWeekdaysCount, NULL, cal)) > 0)
2786 return newStart;
2787 // EEEEEE short failed, now try EEEEE narrow
2788 else if ((newStart = matchString(text, start, UCAL_DAY_OF_WEEK,
2789 fSymbols->fNarrowWeekdays, fSymbols->fNarrowWeekdaysCount, NULL, cal)) > 0)
2790 return newStart;
2791 else if (!getBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, status) || patternCharIndex == UDAT_DAY_OF_WEEK_FIELD)
2792 return newStart;
2793 // else we allowing parsing as number, below
2794 }
2795 break;
2797 case UDAT_STANDALONE_DAY_FIELD:
2798 {
2799 if (gotNumber) // c or cc
2800 {
2801 // [We computed 'value' above.]
2802 cal.set(UCAL_DOW_LOCAL, value);
2803 return pos.getIndex();
2804 }
2805 // Want to be able to parse both short and long forms.
2806 // Try count == 4 (cccc) first:
2807 int32_t newStart = 0;
2808 if ((newStart = matchString(text, start, UCAL_DAY_OF_WEEK,
2809 fSymbols->fStandaloneWeekdays, fSymbols->fStandaloneWeekdaysCount, NULL, cal)) > 0)
2810 return newStart;
2811 else if ((newStart = matchString(text, start, UCAL_DAY_OF_WEEK,
2812 fSymbols->fStandaloneShortWeekdays, fSymbols->fStandaloneShortWeekdaysCount, NULL, cal)) > 0)
2813 return newStart;
2814 else if ((newStart = matchString(text, start, UCAL_DAY_OF_WEEK,
2815 fSymbols->fStandaloneShorterWeekdays, fSymbols->fStandaloneShorterWeekdaysCount, NULL, cal)) > 0)
2816 return newStart;
2817 else if (!getBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, status))
2818 return newStart;
2819 // else we allowing parsing as number, below
2820 }
2821 break;
2823 case UDAT_AM_PM_FIELD:
2824 return matchString(text, start, UCAL_AM_PM, fSymbols->fAmPms, fSymbols->fAmPmsCount, NULL, cal);
2826 case UDAT_HOUR1_FIELD:
2827 // [We computed 'value' above.]
2828 if (value == cal.getLeastMaximum(UCAL_HOUR)+1)
2829 value = 0;
2831 // fall through to set field
2833 case UDAT_HOUR0_FIELD:
2834 cal.set(UCAL_HOUR, value);
2835 return pos.getIndex();
2837 case UDAT_QUARTER_FIELD:
2838 if (gotNumber) // i.e., Q or QQ.
2839 {
2840 // Don't want to parse the month if it is a string
2841 // while pattern uses numeric style: Q or QQ.
2842 // [We computed 'value' above.]
2843 cal.set(UCAL_MONTH, (value - 1) * 3);
2844 return pos.getIndex();
2845 } else {
2846 // count >= 3 // i.e., QQQ or QQQQ
2847 // Want to be able to parse both short and long forms.
2848 // Try count == 4 first:
2849 int32_t newStart = 0;
2851 if ((newStart = matchQuarterString(text, start, UCAL_MONTH,
2852 fSymbols->fQuarters, fSymbols->fQuartersCount, cal)) > 0)
2853 return newStart;
2854 else if ((newStart = matchQuarterString(text, start, UCAL_MONTH,
2855 fSymbols->fShortQuarters, fSymbols->fShortQuartersCount, cal)) > 0)
2856 return newStart;
2857 else if (!getBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, status))
2858 return newStart;
2859 // else we allowing parsing as number, below
2860 }
2861 break;
2863 case UDAT_STANDALONE_QUARTER_FIELD:
2864 if (gotNumber) // i.e., q or qq.
2865 {
2866 // Don't want to parse the month if it is a string
2867 // while pattern uses numeric style: q or q.
2868 // [We computed 'value' above.]
2869 cal.set(UCAL_MONTH, (value - 1) * 3);
2870 return pos.getIndex();
2871 } else {
2872 // count >= 3 // i.e., qqq or qqqq
2873 // Want to be able to parse both short and long forms.
2874 // Try count == 4 first:
2875 int32_t newStart = 0;
2877 if ((newStart = matchQuarterString(text, start, UCAL_MONTH,
2878 fSymbols->fStandaloneQuarters, fSymbols->fStandaloneQuartersCount, cal)) > 0)
2879 return newStart;
2880 else if ((newStart = matchQuarterString(text, start, UCAL_MONTH,
2881 fSymbols->fStandaloneShortQuarters, fSymbols->fStandaloneShortQuartersCount, cal)) > 0)
2882 return newStart;
2883 else if (!getBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, status))
2884 return newStart;
2885 // else we allowing parsing as number, below
2886 }
2887 break;
2889 case UDAT_TIMEZONE_FIELD: // 'z'
2890 {
2891 UTimeZoneFormatTimeType tzTimeType = UTZFMT_TIME_TYPE_UNKNOWN;
2892 UTimeZoneFormatStyle style = (count < 4) ? UTZFMT_STYLE_SPECIFIC_SHORT : UTZFMT_STYLE_SPECIFIC_LONG;
2893 TimeZone *tz = tzFormat()->parse(style, text, pos, &tzTimeType);
2894 if (tz != NULL) {
2895 ((SimpleDateFormat*)this)->tztype = tzTimeType;
2896 cal.adoptTimeZone(tz);
2897 return pos.getIndex();
2898 }
2899 }
2900 break;
2901 case UDAT_TIMEZONE_RFC_FIELD: // 'Z'
2902 {
2903 UTimeZoneFormatTimeType tzTimeType = UTZFMT_TIME_TYPE_UNKNOWN;
2904 UTimeZoneFormatStyle style = (count < 4) ?
2905 UTZFMT_STYLE_ISO_BASIC_LOCAL_FULL : ((count == 5) ? UTZFMT_STYLE_ISO_EXTENDED_FULL: UTZFMT_STYLE_LOCALIZED_GMT);
2906 TimeZone *tz = tzFormat()->parse(style, text, pos, &tzTimeType);
2907 if (tz != NULL) {
2908 ((SimpleDateFormat*)this)->tztype = tzTimeType;
2909 cal.adoptTimeZone(tz);
2910 return pos.getIndex();
2911 }
2912 return -start;
2913 }
2914 case UDAT_TIMEZONE_GENERIC_FIELD: // 'v'
2915 {
2916 UTimeZoneFormatTimeType tzTimeType = UTZFMT_TIME_TYPE_UNKNOWN;
2917 UTimeZoneFormatStyle style = (count < 4) ? UTZFMT_STYLE_GENERIC_SHORT : UTZFMT_STYLE_GENERIC_LONG;
2918 TimeZone *tz = tzFormat()->parse(style, text, pos, &tzTimeType);
2919 if (tz != NULL) {
2920 ((SimpleDateFormat*)this)->tztype = tzTimeType;
2921 cal.adoptTimeZone(tz);
2922 return pos.getIndex();
2923 }
2924 return -start;
2925 }
2926 case UDAT_TIMEZONE_SPECIAL_FIELD: // 'V'
2927 {
2928 UTimeZoneFormatTimeType tzTimeType = UTZFMT_TIME_TYPE_UNKNOWN;
2929 UTimeZoneFormatStyle style;
2930 switch (count) {
2931 case 1:
2932 style = UTZFMT_STYLE_ZONE_ID_SHORT;
2933 break;
2934 case 2:
2935 style = UTZFMT_STYLE_ZONE_ID;
2936 break;
2937 case 3:
2938 style = UTZFMT_STYLE_EXEMPLAR_LOCATION;
2939 break;
2940 default:
2941 style = UTZFMT_STYLE_GENERIC_LOCATION;
2942 break;
2943 }
2944 TimeZone *tz = tzFormat()->parse(style, text, pos, &tzTimeType);
2945 if (tz != NULL) {
2946 ((SimpleDateFormat*)this)->tztype = tzTimeType;
2947 cal.adoptTimeZone(tz);
2948 return pos.getIndex();
2949 }
2950 return -start;
2951 }
2952 case UDAT_TIMEZONE_LOCALIZED_GMT_OFFSET_FIELD: // 'O'
2953 {
2954 UTimeZoneFormatTimeType tzTimeType = UTZFMT_TIME_TYPE_UNKNOWN;
2955 UTimeZoneFormatStyle style = (count < 4) ? UTZFMT_STYLE_LOCALIZED_GMT_SHORT : UTZFMT_STYLE_LOCALIZED_GMT;
2956 TimeZone *tz = tzFormat()->parse(style, text, pos, &tzTimeType);
2957 if (tz != NULL) {
2958 ((SimpleDateFormat*)this)->tztype = tzTimeType;
2959 cal.adoptTimeZone(tz);
2960 return pos.getIndex();
2961 }
2962 return -start;
2963 }
2964 case UDAT_TIMEZONE_ISO_FIELD: // 'X'
2965 {
2966 UTimeZoneFormatTimeType tzTimeType = UTZFMT_TIME_TYPE_UNKNOWN;
2967 UTimeZoneFormatStyle style;
2968 switch (count) {
2969 case 1:
2970 style = UTZFMT_STYLE_ISO_BASIC_SHORT;
2971 break;
2972 case 2:
2973 style = UTZFMT_STYLE_ISO_BASIC_FIXED;
2974 break;
2975 case 3:
2976 style = UTZFMT_STYLE_ISO_EXTENDED_FIXED;
2977 break;
2978 case 4:
2979 style = UTZFMT_STYLE_ISO_BASIC_FULL;
2980 break;
2981 default:
2982 style = UTZFMT_STYLE_ISO_EXTENDED_FULL;
2983 break;
2984 }
2985 TimeZone *tz = tzFormat()->parse(style, text, pos, &tzTimeType);
2986 if (tz != NULL) {
2987 ((SimpleDateFormat*)this)->tztype = tzTimeType;
2988 cal.adoptTimeZone(tz);
2989 return pos.getIndex();
2990 }
2991 return -start;
2992 }
2993 case UDAT_TIMEZONE_ISO_LOCAL_FIELD: // 'x'
2994 {
2995 UTimeZoneFormatTimeType tzTimeType = UTZFMT_TIME_TYPE_UNKNOWN;
2996 UTimeZoneFormatStyle style;
2997 switch (count) {
2998 case 1:
2999 style = UTZFMT_STYLE_ISO_BASIC_LOCAL_SHORT;
3000 break;
3001 case 2:
3002 style = UTZFMT_STYLE_ISO_BASIC_LOCAL_FIXED;
3003 break;
3004 case 3:
3005 style = UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FIXED;
3006 break;
3007 case 4:
3008 style = UTZFMT_STYLE_ISO_BASIC_LOCAL_FULL;
3009 break;
3010 default:
3011 style = UTZFMT_STYLE_ISO_EXTENDED_LOCAL_FULL;
3012 break;
3013 }
3014 TimeZone *tz = tzFormat()->parse(style, text, pos, &tzTimeType);
3015 if (tz != NULL) {
3016 ((SimpleDateFormat*)this)->tztype = tzTimeType;
3017 cal.adoptTimeZone(tz);
3018 return pos.getIndex();
3019 }
3020 return -start;
3021 }
3023 default:
3024 // Handle "generic" fields
3025 // this is now handled below, outside the switch block
3026 break;
3027 }
3028 // Handle "generic" fields:
3029 // switch default case now handled here (outside switch block) to allow
3030 // parsing of some string fields as digits for lenient case
3032 int32_t parseStart = pos.getIndex();
3033 const UnicodeString* src;
3034 if (obeyCount) {
3035 if ((start+count) > text.length()) {
3036 return -start;
3037 }
3038 text.extractBetween(0, start + count, temp);
3039 src = &temp;
3040 } else {
3041 src = &text;
3042 }
3043 parseInt(*src, number, pos, allowNegative,currentNumberFormat);
3044 if (pos.getIndex() != parseStart) {
3045 int32_t value = number.getLong();
3047 // Don't need suffix processing here (as in number processing at the beginning of the function);
3048 // the new fields being handled as numeric values (month, weekdays, quarters) should not have suffixes.
3050 if (!getBooleanAttribute(UDAT_PARSE_ALLOW_NUMERIC, status)) {
3051 // Check the range of the value
3052 int32_t bias = gFieldRangeBias[patternCharIndex];
3053 if (bias >= 0 && (value > cal.getMaximum(field) + bias || value < cal.getMinimum(field) + bias)) {
3054 return -start;
3055 }
3056 }
3058 // For the following, need to repeat some of the "if (gotNumber)" code above:
3059 // UDAT_[STANDALONE_]MONTH_FIELD, UDAT_DOW_LOCAL_FIELD, UDAT_STANDALONE_DAY_FIELD,
3060 // UDAT_[STANDALONE_]QUARTER_FIELD
3061 switch (patternCharIndex) {
3062 case UDAT_MONTH_FIELD:
3063 // See notes under UDAT_MONTH_FIELD case above
3064 if (!strcmp(cal.getType(),"hebrew")) {
3065 HebrewCalendar *hc = (HebrewCalendar*)&cal;
3066 if (cal.isSet(UCAL_YEAR)) {
3067 UErrorCode status = U_ZERO_ERROR;
3068 if (!hc->isLeapYear(hc->get(UCAL_YEAR,status)) && value >= 6) {
3069 cal.set(UCAL_MONTH, value);
3070 } else {
3071 cal.set(UCAL_MONTH, value - 1);
3072 }
3073 } else {
3074 saveHebrewMonth = value;
3075 }
3076 } else {
3077 cal.set(UCAL_MONTH, value - 1);
3078 }
3079 break;
3080 case UDAT_STANDALONE_MONTH_FIELD:
3081 cal.set(UCAL_MONTH, value - 1);
3082 break;
3083 case UDAT_DOW_LOCAL_FIELD:
3084 case UDAT_STANDALONE_DAY_FIELD:
3085 cal.set(UCAL_DOW_LOCAL, value);
3086 break;
3087 case UDAT_QUARTER_FIELD:
3088 case UDAT_STANDALONE_QUARTER_FIELD:
3089 cal.set(UCAL_MONTH, (value - 1) * 3);
3090 break;
3091 default:
3092 cal.set(field, value);
3093 break;
3094 }
3095 return pos.getIndex();
3096 }
3097 return -start;
3098 }
3100 /**
3101 * Parse an integer using fNumberFormat. This method is semantically
3102 * const, but actually may modify fNumberFormat.
3103 */
3104 void SimpleDateFormat::parseInt(const UnicodeString& text,
3105 Formattable& number,
3106 ParsePosition& pos,
3107 UBool allowNegative,
3108 NumberFormat *fmt) const {
3109 parseInt(text, number, -1, pos, allowNegative,fmt);
3110 }
3112 /**
3113 * Parse an integer using fNumberFormat up to maxDigits.
3114 */
3115 void SimpleDateFormat::parseInt(const UnicodeString& text,
3116 Formattable& number,
3117 int32_t maxDigits,
3118 ParsePosition& pos,
3119 UBool allowNegative,
3120 NumberFormat *fmt) const {
3121 UnicodeString oldPrefix;
3122 DecimalFormat* df = NULL;
3123 if (!allowNegative && (df = dynamic_cast<DecimalFormat*>(fmt)) != NULL) {
3124 df->getNegativePrefix(oldPrefix);
3125 df->setNegativePrefix(UnicodeString(TRUE, SUPPRESS_NEGATIVE_PREFIX, -1));
3126 }
3127 int32_t oldPos = pos.getIndex();
3128 fmt->parse(text, number, pos);
3129 if (df != NULL) {
3130 df->setNegativePrefix(oldPrefix);
3131 }
3133 if (maxDigits > 0) {
3134 // adjust the result to fit into
3135 // the maxDigits and move the position back
3136 int32_t nDigits = pos.getIndex() - oldPos;
3137 if (nDigits > maxDigits) {
3138 int32_t val = number.getLong();
3139 nDigits -= maxDigits;
3140 while (nDigits > 0) {
3141 val /= 10;
3142 nDigits--;
3143 }
3144 pos.setIndex(oldPos + maxDigits);
3145 number.setLong(val);
3146 }
3147 }
3148 }
3150 //----------------------------------------------------------------------
3152 void SimpleDateFormat::translatePattern(const UnicodeString& originalPattern,
3153 UnicodeString& translatedPattern,
3154 const UnicodeString& from,
3155 const UnicodeString& to,
3156 UErrorCode& status)
3157 {
3158 // run through the pattern and convert any pattern symbols from the version
3159 // in "from" to the corresponding character ion "to". This code takes
3160 // quoted strings into account (it doesn't try to translate them), and it signals
3161 // an error if a particular "pattern character" doesn't appear in "from".
3162 // Depending on the values of "from" and "to" this can convert from generic
3163 // to localized patterns or localized to generic.
3164 if (U_FAILURE(status))
3165 return;
3167 translatedPattern.remove();
3168 UBool inQuote = FALSE;
3169 for (int32_t i = 0; i < originalPattern.length(); ++i) {
3170 UChar c = originalPattern[i];
3171 if (inQuote) {
3172 if (c == QUOTE)
3173 inQuote = FALSE;
3174 }
3175 else {
3176 if (c == QUOTE)
3177 inQuote = TRUE;
3178 else if ((c >= 0x0061 /*'a'*/ && c <= 0x007A) /*'z'*/
3179 || (c >= 0x0041 /*'A'*/ && c <= 0x005A /*'Z'*/)) {
3180 int32_t ci = from.indexOf(c);
3181 if (ci == -1) {
3182 status = U_INVALID_FORMAT_ERROR;
3183 return;
3184 }
3185 c = to[ci];
3186 }
3187 }
3188 translatedPattern += c;
3189 }
3190 if (inQuote) {
3191 status = U_INVALID_FORMAT_ERROR;
3192 return;
3193 }
3194 }
3196 //----------------------------------------------------------------------
3198 UnicodeString&
3199 SimpleDateFormat::toPattern(UnicodeString& result) const
3200 {
3201 result = fPattern;
3202 return result;
3203 }
3205 //----------------------------------------------------------------------
3207 UnicodeString&
3208 SimpleDateFormat::toLocalizedPattern(UnicodeString& result,
3209 UErrorCode& status) const
3210 {
3211 translatePattern(fPattern, result,
3212 UnicodeString(DateFormatSymbols::getPatternUChars()),
3213 fSymbols->fLocalPatternChars, status);
3214 return result;
3215 }
3217 //----------------------------------------------------------------------
3219 void
3220 SimpleDateFormat::applyPattern(const UnicodeString& pattern)
3221 {
3222 fPattern = pattern;
3223 }
3225 //----------------------------------------------------------------------
3227 void
3228 SimpleDateFormat::applyLocalizedPattern(const UnicodeString& pattern,
3229 UErrorCode &status)
3230 {
3231 translatePattern(pattern, fPattern,
3232 fSymbols->fLocalPatternChars,
3233 UnicodeString(DateFormatSymbols::getPatternUChars()), status);
3234 }
3236 //----------------------------------------------------------------------
3238 const DateFormatSymbols*
3239 SimpleDateFormat::getDateFormatSymbols() const
3240 {
3241 return fSymbols;
3242 }
3244 //----------------------------------------------------------------------
3246 void
3247 SimpleDateFormat::adoptDateFormatSymbols(DateFormatSymbols* newFormatSymbols)
3248 {
3249 delete fSymbols;
3250 fSymbols = newFormatSymbols;
3251 }
3253 //----------------------------------------------------------------------
3254 void
3255 SimpleDateFormat::setDateFormatSymbols(const DateFormatSymbols& newFormatSymbols)
3256 {
3257 delete fSymbols;
3258 fSymbols = new DateFormatSymbols(newFormatSymbols);
3259 }
3261 //----------------------------------------------------------------------
3262 const TimeZoneFormat*
3263 SimpleDateFormat::getTimeZoneFormat(void) const {
3264 return (const TimeZoneFormat*)tzFormat();
3265 }
3267 //----------------------------------------------------------------------
3268 void
3269 SimpleDateFormat::adoptTimeZoneFormat(TimeZoneFormat* timeZoneFormatToAdopt)
3270 {
3271 delete fTimeZoneFormat;
3272 fTimeZoneFormat = timeZoneFormatToAdopt;
3273 }
3275 //----------------------------------------------------------------------
3276 void
3277 SimpleDateFormat::setTimeZoneFormat(const TimeZoneFormat& newTimeZoneFormat)
3278 {
3279 delete fTimeZoneFormat;
3280 fTimeZoneFormat = new TimeZoneFormat(newTimeZoneFormat);
3281 }
3283 //----------------------------------------------------------------------
3286 void SimpleDateFormat::adoptCalendar(Calendar* calendarToAdopt)
3287 {
3288 UErrorCode status = U_ZERO_ERROR;
3289 DateFormat::adoptCalendar(calendarToAdopt);
3290 delete fSymbols;
3291 fSymbols=NULL;
3292 initializeSymbols(fLocale, fCalendar, status); // we need new symbols
3293 initializeDefaultCentury(); // we need a new century (possibly)
3294 }
3297 //----------------------------------------------------------------------
3300 void SimpleDateFormat::setContext(UDisplayContext value, UErrorCode& status)
3301 {
3302 if (U_FAILURE(status))
3303 return;
3304 if ( (UDisplayContextType)((uint32_t)value >> 8) == UDISPCTX_TYPE_CAPITALIZATION ) {
3305 fCapitalizationContext = value;
3306 } else {
3307 status = U_ILLEGAL_ARGUMENT_ERROR;
3308 }
3309 }
3312 //----------------------------------------------------------------------
3315 UDisplayContext SimpleDateFormat::getContext(UDisplayContextType type, UErrorCode& status) const
3316 {
3317 if (U_FAILURE(status))
3318 return (UDisplayContext)0;
3319 if (type != UDISPCTX_TYPE_CAPITALIZATION) {
3320 status = U_ILLEGAL_ARGUMENT_ERROR;
3321 return (UDisplayContext)0;
3322 }
3323 return fCapitalizationContext;
3324 }
3327 //----------------------------------------------------------------------
3330 UBool
3331 SimpleDateFormat::isFieldUnitIgnored(UCalendarDateFields field) const {
3332 return isFieldUnitIgnored(fPattern, field);
3333 }
3336 UBool
3337 SimpleDateFormat::isFieldUnitIgnored(const UnicodeString& pattern,
3338 UCalendarDateFields field) {
3339 int32_t fieldLevel = fgCalendarFieldToLevel[field];
3340 int32_t level;
3341 UChar ch;
3342 UBool inQuote = FALSE;
3343 UChar prevCh = 0;
3344 int32_t count = 0;
3346 for (int32_t i = 0; i < pattern.length(); ++i) {
3347 ch = pattern[i];
3348 if (ch != prevCh && count > 0) {
3349 level = fgPatternCharToLevel[prevCh - PATTERN_CHAR_BASE];
3350 // the larger the level, the smaller the field unit.
3351 if ( fieldLevel <= level ) {
3352 return FALSE;
3353 }
3354 count = 0;
3355 }
3356 if (ch == QUOTE) {
3357 if ((i+1) < pattern.length() && pattern[i+1] == QUOTE) {
3358 ++i;
3359 } else {
3360 inQuote = ! inQuote;
3361 }
3362 }
3363 else if ( ! inQuote && ((ch >= 0x0061 /*'a'*/ && ch <= 0x007A /*'z'*/)
3364 || (ch >= 0x0041 /*'A'*/ && ch <= 0x005A /*'Z'*/))) {
3365 prevCh = ch;
3366 ++count;
3367 }
3368 }
3369 if ( count > 0 ) {
3370 // last item
3371 level = fgPatternCharToLevel[prevCh - PATTERN_CHAR_BASE];
3372 if ( fieldLevel <= level ) {
3373 return FALSE;
3374 }
3375 }
3376 return TRUE;
3377 }
3379 //----------------------------------------------------------------------
3381 const Locale&
3382 SimpleDateFormat::getSmpFmtLocale(void) const {
3383 return fLocale;
3384 }
3386 //----------------------------------------------------------------------
3388 int32_t
3389 SimpleDateFormat::checkIntSuffix(const UnicodeString& text, int32_t start,
3390 int32_t patLoc, UBool isNegative) const {
3391 // local variables
3392 UnicodeString suf;
3393 int32_t patternMatch;
3394 int32_t textPreMatch;
3395 int32_t textPostMatch;
3397 // check that we are still in range
3398 if ( (start > text.length()) ||
3399 (start < 0) ||
3400 (patLoc < 0) ||
3401 (patLoc > fPattern.length())) {
3402 // out of range, don't advance location in text
3403 return start;
3404 }
3406 // get the suffix
3407 DecimalFormat* decfmt = dynamic_cast<DecimalFormat*>(fNumberFormat);
3408 if (decfmt != NULL) {
3409 if (isNegative) {
3410 suf = decfmt->getNegativeSuffix(suf);
3411 }
3412 else {
3413 suf = decfmt->getPositiveSuffix(suf);
3414 }
3415 }
3417 // check for suffix
3418 if (suf.length() <= 0) {
3419 return start;
3420 }
3422 // check suffix will be encountered in the pattern
3423 patternMatch = compareSimpleAffix(suf,fPattern,patLoc);
3425 // check if a suffix will be encountered in the text
3426 textPreMatch = compareSimpleAffix(suf,text,start);
3428 // check if a suffix was encountered in the text
3429 textPostMatch = compareSimpleAffix(suf,text,start-suf.length());
3431 // check for suffix match
3432 if ((textPreMatch >= 0) && (patternMatch >= 0) && (textPreMatch == patternMatch)) {
3433 return start;
3434 }
3435 else if ((textPostMatch >= 0) && (patternMatch >= 0) && (textPostMatch == patternMatch)) {
3436 return start - suf.length();
3437 }
3439 // should not get here
3440 return start;
3441 }
3443 //----------------------------------------------------------------------
3445 int32_t
3446 SimpleDateFormat::compareSimpleAffix(const UnicodeString& affix,
3447 const UnicodeString& input,
3448 int32_t pos) const {
3449 int32_t start = pos;
3450 for (int32_t i=0; i<affix.length(); ) {
3451 UChar32 c = affix.char32At(i);
3452 int32_t len = U16_LENGTH(c);
3453 if (PatternProps::isWhiteSpace(c)) {
3454 // We may have a pattern like: \u200F \u0020
3455 // and input text like: \u200F \u0020
3456 // Note that U+200F and U+0020 are Pattern_White_Space but only
3457 // U+0020 is UWhiteSpace. So we have to first do a direct
3458 // match of the run of Pattern_White_Space in the pattern,
3459 // then match any extra characters.
3460 UBool literalMatch = FALSE;
3461 while (pos < input.length() &&
3462 input.char32At(pos) == c) {
3463 literalMatch = TRUE;
3464 i += len;
3465 pos += len;
3466 if (i == affix.length()) {
3467 break;
3468 }
3469 c = affix.char32At(i);
3470 len = U16_LENGTH(c);
3471 if (!PatternProps::isWhiteSpace(c)) {
3472 break;
3473 }
3474 }
3476 // Advance over run in pattern
3477 i = skipPatternWhiteSpace(affix, i);
3479 // Advance over run in input text
3480 // Must see at least one white space char in input,
3481 // unless we've already matched some characters literally.
3482 int32_t s = pos;
3483 pos = skipUWhiteSpace(input, pos);
3484 if (pos == s && !literalMatch) {
3485 return -1;
3486 }
3488 // If we skip UWhiteSpace in the input text, we need to skip it in the pattern.
3489 // Otherwise, the previous lines may have skipped over text (such as U+00A0) that
3490 // is also in the affix.
3491 i = skipUWhiteSpace(affix, i);
3492 } else {
3493 if (pos < input.length() &&
3494 input.char32At(pos) == c) {
3495 i += len;
3496 pos += len;
3497 } else {
3498 return -1;
3499 }
3500 }
3501 }
3502 return pos - start;
3503 }
3505 //----------------------------------------------------------------------
3507 int32_t
3508 SimpleDateFormat::skipPatternWhiteSpace(const UnicodeString& text, int32_t pos) const {
3509 const UChar* s = text.getBuffer();
3510 return (int32_t)(PatternProps::skipWhiteSpace(s + pos, text.length() - pos) - s);
3511 }
3513 //----------------------------------------------------------------------
3515 int32_t
3516 SimpleDateFormat::skipUWhiteSpace(const UnicodeString& text, int32_t pos) const {
3517 while (pos < text.length()) {
3518 UChar32 c = text.char32At(pos);
3519 if (!u_isUWhiteSpace(c)) {
3520 break;
3521 }
3522 pos += U16_LENGTH(c);
3523 }
3524 return pos;
3525 }
3527 //----------------------------------------------------------------------
3529 // Lazy TimeZoneFormat instantiation, semantically const.
3530 TimeZoneFormat *
3531 SimpleDateFormat::tzFormat() const {
3532 if (fTimeZoneFormat == NULL) {
3533 umtx_lock(&LOCK);
3534 {
3535 if (fTimeZoneFormat == NULL) {
3536 UErrorCode status = U_ZERO_ERROR;
3537 TimeZoneFormat *tzfmt = TimeZoneFormat::createInstance(fLocale, status);
3538 if (U_FAILURE(status)) {
3539 return NULL;
3540 }
3542 const_cast<SimpleDateFormat *>(this)->fTimeZoneFormat = tzfmt;
3543 }
3544 }
3545 umtx_unlock(&LOCK);
3546 }
3547 return fTimeZoneFormat;
3548 }
3550 U_NAMESPACE_END
3552 #endif /* #if !UCONFIG_NO_FORMATTING */
3554 //eof
