Implement BYDAY with offset and limiting behavior

[axunonb]

The following RRULE cases were not implemented (see the diff for newly added unit tests for details):

1. YEARLY + BYMONTH + numeric BYDAY offsets

   • Pattern: FREQ=YEARLY;BYMONTH=6,9;BYDAY=2MO
   • Semantics: treat numeric BYDAY as “nth weekday inside each BYMONTH” (produce the 2nd Monday of each specified month).

2. YEARLY + numeric BYDAY without BYMONTH

   • Pattern: FREQ=YEARLY;BYDAY=20MO
   • Semantics: interpret numeric BYDAY as “nth weekday of the YEAR” (the 20th Monday of the year).

3. YEARLY + BYMONTH + negative numeric BYDAY

   • Pattern: FREQ=YEARLY;BYMONTH=6,9;BYDAY=-1SU
   • Semantics: negative offsets select from the end of the month (e.g. -1SU → last Sunday in each BYMONTH).

4. Implemented the per-month offset expansion and preserved BYWEEKNO / BYMONTH compatibility and negative-offset

These behaviors are closely related and required special handling in RecurrencePatternEvaluator.GetAbsWeekDays and GetAbsWeekDaysYearlyPerMonthOffsets.

[Edit]
Relevant unit tests in RecurrenceTestCases.txt contain a disclaimer regarding RFC5545 interpretation in ical.net

Fixes #782
Closes #888

[codecov]

Codecov Report

❌ Patch coverage is 94.82759% with 3 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
Ical.Net/Evaluation/RecurrencePatternEvaluator.cs	94.6%	1 Missing and 2 partials ⚠️

@@           Coverage Diff           @@
##            main    #887     +/-   ##
=======================================
+ Coverage   68.3%   68.5%   +0.2%     
=======================================
  Files        115     115             
  Lines       4376    4394     +18     
  Branches    1010    1011      +1     
=======================================
+ Hits        2988    3008     +20     
  Misses      1028    1028             
+ Partials     360     358      -2     

Files with missing lines	Coverage Δ
Ical.Net/Evaluation/RecurrenceUtil.cs	100.0% <100.0%> (ø)
Ical.Net/Evaluation/RecurrencePatternEvaluator.cs	92.3% <94.6%> (+1.2%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[maknapp]

FYI, I made several tests related to this for a new recurrence evaluator. I ran those tests against this fix and they all pass.

[axunonb]

@maknapp Thanks for the review and the helpful hint

I made several tests related to this for a new recurrence evaluator. I ran those tests against this fix and they all pass.

So it looks like we're on the right way. As @minichma opened the issue, I'd like have his feedback as well, if he finds the time. He also contributes to libical.

[axunonb]

@minichma Unit test for this PR are now part of RecurrenceTestCases.txt

[minichma]

I started looking at the test cases and from my point of view they look good. Commented some minor thoughts, just nits. Didn't look at the actual code so far. Will try to find time at the weekend.

[minichma]

DTSTART is not in line with the RRULE, which makes the case undefined according to the RFC (similar in other test cases). Unless this test should explicitly cover the undefined case, it would be good to have DTSTART aligned with RRULE, otherwise readers would have to think about whether the deviation is intentional.

The time part could be omitted, which would make the test somewhat more concise.

[minichma]

Very nice! Sorry for the delay!

[sonarqubecloud]

Quality Gate passed

Issues
2 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[axunonb]

@minichma, @maknapp Thanks for your reviews, and happy holidays.