Re: [icu-support] 59 day month in Chinese calendar?

[Andre-John M. <aj...@sy...>]

On 25-Nov-07, at 17:13 , Daniel Noll wrote:

> On Saturday 24 November 2007 10:23:28 Andre-John Mas wrote:
>> Hi,
>>
>> I have written a small program that displays the various calendars
>> alongside one another, in SVG:
>>
>> http://ajmas.dyndns.org/?loc=tests/index
>>
>> One thing that struck me was a 59 day month in the year 25 of the
>> Chinese calendar. Not knowing much on the Chinese calendar, could
>> someone confirm that this is indeed correct?
>
> Yes and no. :-)
>
> The Chinese calendar is luni-solar, so the month follows a lunar  
> cycle but
> every few years there is an intercalary month (or "leap month") to  
> keep the
> year more or less in sync with the sun.  One of these does occur in  
> 26.
>
> It is considered to be a separate month, hence there is no 59 day  
> month.  Your
> plot should be displaying 13 months in that year.
>
> However it has the same number as the preceding month, which could  
> explain the
> discrepancy if you were using the month number to determine when a  
> month
> rolls over instead of using the day of the month reverting to 1.  Or  
> perhaps
> ICU had a bug where it didn't reset to 1 at the start of an  
> intercalary
> month, I don't know.

The bug which I have opened contains the code which I use to identify  
the
new years and new months:

http://bugs.icu-project.org/trac/ticket/6033

Basically I determine whether a new month has occurred if the month
number of the previous day is not the same as the current day. If  
there is a
better way of determining whether this is a new month, then I am open  
to ideas.

Andre