MenoData
commented
9 months ago Sorry for late reply but it is also not so easy for me to analyze text partly written in Chinese ;-)
Let's first say three things about the formatter you define above:
a) If you apply a CLDR pattern then you indirectly use text resources defined / overtaken by CLDR group. The actual chinese resources (for the Chinese calendar, not for the gregorian one) are like this for the (lunar) month (without leap symbol) and are actually the same for the output style using the pattern letter L:
M(w)_1=正月 M(w)_2=二月 M(w)_3=三月 M(w)_4=四月 M(w)_5=五月 M(w)_6=六月 M(w)_7=七月 M(w)_8=八月 M(w)_9=九月 M(w)_10=十月 M(w)_11=十一月 M(w)_12=腊月
For example: Given your formatter definition above, the gregorian date 2017-07-23 will produce the output
2017(丁酉) 閏六月 01. (周日) 大暑
The month part here (閏六月) first uses the leap indicator 閏 (for the leap month) and then the text resource from CLDR while the day part just uses common arab digits (0-9). And to complete the analysis for your formatter: It produces for the date 2023-11-05 the output (pay attention the transformation of month and day numbers - explanation later under c)):
2023(癸卯) 九月 22. (周日) 霜降
The last part is the solar term which has only meaning in Chinese calendar chronology. It is NOT the zodiac.
b) Looking at your desired text resources for the month, it seems to me that you really want the month numbers written in Taiwanese style (traditional instead of simplified Chinese using the locale zh-TW corresponding to the constant Locale.TRADITIONAL_CHINESE). Well, CLDR defines the resources for the month in this Taiwanese locale (for the Chinese calendar, not for the gregorian one) like:
M(w)_1=正月 M(w)_2=二月 M(w)_3=三月 M(w)_4=四月 M(w)_5=五月 M(w)_6=六月 M(w)_7=七月 M(w)_8=八月 M(w)_9=九月 M(w)_10=十月 M(w)_11=冬月 M(w)_12=臘月
c) You seem to be confused about the difference between language and chronology. The formatter you define above uses the Chinese language AND the old rural Chinese calendar. It is not gregorian. Therefore, you have different month and day numbers for any given year. Given the gregorian date 2023-11-05, you seem to expect the month 11 and the day 5, but in the Chinese calendar those numbers will be transformed to the 9th month and 22. day. Furthermore, the Chinese calendar chronology knows leap months while the gregorian calendar has no leap months.
Obviously you wish another format with customized text resources so defining a CLDR pattern using standardized text mappings is not necessarily the best way, especially for the day text (the formatter would simply use arab digits). Furthermore: You seem to need a formatter for gregorian dates in Chinese language including elements of Chinese calendar like cyclic year and zodiac (as partial property of cyclic year).
Fortunately there is a way to define custom text resources. Let's look at this formatter which will use the Taiwanese locale, a customized day element, the gregorian chronology and two embedded elements of the Chinese cyclic year and zodiac. It even uses a customized month element because the month text resources defined by CLDR are unfortunately different for the Chinese (as shown above) and the gregorian calendar chronology, given the same locale for Taiwan.
Map<Month, String> monthMap = new EnumMap<>(Month.class); // using net.time4j.Month
monthMap.put(Month.JANUARY, "正月");
// complete the mappings as desired
monthMap.put(Month.FEBRUARY, "");
monthMap.put(Month.MARCH, "");
monthMap.put(Month.APRIL, "");
monthMap.put(Month.MAY, "");
monthMap.put(Month.JUNE, "");
monthMap.put(Month.JULY, "");
monthMap.put(Month.AUGUST, "");
monthMap.put(Month.SEPTEMBER, "");
monthMap.put(Month.OCTOBER, "");
monthMap.put(Month.NOVEMBER, "冬月");
monthMap.put(Month.DECEMBER, "");
Map<Integer, String> dayMap = new HashMap<>();
dayMap.put(1, "初一");
dayMap.put(2, "初二");
dayMap.put(3, "初三");
dayMap.put(4, "初四");
dayMap.put(5, "初五");
// define more mappings until day 31
ChronoPrinter<Integer> yearPrinter =
(Integer gregorianYear, StringBuilder buffer, AttributeQuery attributes) -> {
buffer.append(ChineseCalendar.ofNewYear(gregorianYear).getYear().getDisplayName(Locale.TRADITIONAL_CHINESE));
return Collections.emptySet();
};
ChronoPrinter<PlainDate> zodiacPrinter =
(PlainDate gregorianDate, StringBuilder buffer, AttributeQuery attributes) -> {
buffer.append(gregorianDate.transform(ChineseCalendar.axis()).getYear().getZodiac(Locale.TRADITIONAL_CHINESE));
return Collections.emptySet();
};
ChronoParser<Integer> yearParser =
(CharSequence text, ParseLog status, AttributeQuery attributes) -> {
throw new UnsupportedOperationException("Parsing not used.");
};
ChronoParser<PlainDate> zodiacParser =
(CharSequence text, ParseLog status, AttributeQuery attributes) -> {
throw new UnsupportedOperationException("Parsing not used.");
};
ChronoFormatter<PlainDate> gf = ChronoFormatter.setUp(PlainDate.axis(), Locale.TRADITIONAL_CHINESE)
.addPattern("y(", PatternType.CLDR)
.addCustomized(PlainDate.YEAR, yearPrinter, yearParser)
.addLiteral(")年")
.addText(PlainDate.MONTH_OF_YEAR, monthMap)
.addText(PlainDate.DAY_OF_MONTH, dayMap)
.addLiteral('(')
.addCustomized(PlainDate.COMPONENT, zodiacPrinter, zodiacParser)
.addLiteral(')')
.build();
String s = gf.print(PlainDate.of(2023, 11, 5));
System.out.println(s);
Output (like your expectation): 2023(癸卯)年冬月初五(兔)
and could display Month alias like: 一月（alias: 正月），十一月（alias: 冬月），十二月（alias: 腊月）.
and display Day alias like: 1（alias: 初一），2（alias: 初二），3（alias: 初三），4（alias: 初四），5（alias: 初五）... 21（alias: 廿一）... and so on.
example:
2023-11-05 10:00:05，expect format:2023(癸卯)年冬月初五(兔) 10:00:05and parse it toChineseCalendar:2023(癸卯)年冬月初五(兔) 10:00:05to2023-11-05 10:00:05