Jalali `isLeap` and `leapsLength` use different leap rules

[hosni]

In calendars/cjs/jalali.js, leap years are decided with two incompatible algorithms:

• isLeap(year) uses the 33-year cycle: ((year + 12) % 33) % 4 === 1
• leapsLength(year) uses the Birashk 2820-year formula

getAllDays() adds leapsLength(year), while month lengths / leap day validity use isLeap(year). Those disagree for most years, so day counts drift relative to the leap table the calendar claims to use.

Reproduction

const jalali = require("./calendars/cjs/jalali");

function countLeapsFromIsLeap(year) {
  let n = 0;
  for (let y = 1; y < year; y++) if (jalali.isLeap(y)) n++;
  return n;
}

console.log(jalali.leapsLength(1403), countLeapsFromIsLeap(1403));
// -> 339, 340

let mismatches = 0;
for (let y = 1; y <= 2000; y++) {
  if (jalali.leapsLength(y) !== countLeapsFromIsLeap(y)) mismatches++;
}
console.log(mismatches);
// -> 1793

Expected

leapsLength(year) should equal the number of leap years in [1, year).
It must use the same rule as isLeap.

Suggested fix

Make leapsLength count with the 33-year cycle (same as isLeap), e.g.:

leapsLength(year) {
  // Number of leap years in [1, year), using the same 33-year rule as isLeap.
  // Closed form of: years y where ((y + 12) % 33) % 4 === 1.
  const n = year - 1;
  return Math.floor(n / 33) * 8 + [0, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 6, 6, 6, 6, 7, 7, 7, 7, 8, 8, 8, 8][n % 33];
}

(Or any equivalent closed form / loop that only calls isLeap.)

This also brings Jalali day-of-epoch counts in line with ICU's Persian calendar for the modern range when paired with the existing epoch.

Impact

Without this, conversions that depend on getAllDays / Julian day can be off by a day versus isLeap-based month length, and versus ICU / civil Persian dates around leap boundaries.

After this fix, Jalali day numbers for modern dates match civil JDN (e.g. 1403-01-01 → 2460390) while Gregorian is still one low until the companion epoch fix.

I think #18 is related to this one!