Incorrect hebrew date calculations

Unified diff

I don't have the same code as you, for some reason. That is, case 353 already returns 0

Hrm

That's my point. The array "MonthDays?" is really based at 1, so 353 should return 1. For some reason, the leap years are fine (ie, 4-6), but the non-leap years return 0-2 instead of 1-3.

Here's a second bug, also for non-leap years. Lines 29-31 of the file define the length of each month. When there's a leap year, an extra month is added in the middle of the year (it takes spot number seven). So for non-leap years, month seven should be length zero. In other words, change:

	    [ 0, 30, 29, 29, 29, 30, 29, 30, 29, 30, 29, 30, 29, 0 ],   // 1
	    [ 0, 30, 29, 30, 29, 30, 29, 30, 29, 30, 29, 30, 29, 0 ],   // 2
	    [ 0, 30, 30, 30, 29, 30, 29, 30, 29, 30, 29, 30, 29, 0 ],   // 3

to

	    [ 0, 30, 29, 29, 29, 30, 29, 0, 30, 29, 30, 29, 30, 29 ],   // 1
	    [ 0, 30, 29, 30, 29, 30, 29, 0, 30, 29, 30, 29, 30, 29 ],   // 2
	    [ 0, 30, 30, 30, 29, 30, 29, 0, 30, 29, 30, 29, 30, 29 ],   // 3

Note: I'm already using this changes in production code on a website I maintain at http://www.kehilla.org/. For those interested, I run that site on a web framework I've written against Tango which I (endearingly) call Snoyweb.

What I meant was that the diff file appeared to be inverted :)

Could you please attach an updated patch file, with the array adjustments also? I'll get those changes into Tango

thx

Correction of array offset and repositioning 0-length month in non-leap years

(In [2519]) closes #543 :: Incorrect hebrew date calculations

Thanks to snoyberg