Fix negative week year formatting in FastDatePrinter

[alhudz]

• Read the contribution guidelines for this project.
• Read the ASF Generative Tooling Guidance if you use Artificial Intelligence (AI).
• I used AI to create any part of, or all of, this pull request. Which AI tool was used to create this pull request, and to what extent did it contribute?
• Run a successful build using the default Maven goal with mvn; that's mvn on the command line by itself.
• Write unit tests that match behavioral changes, where the tests fail if the changes to the runtime are not applied. This may not always be possible, but it is a best practice.
• Write a pull request description that is detailed enough to understand what the pull request does, how, and why.
• Each commit in the pull request should have a meaningful subject line and body. Note that a maintainer may squash commits during the merge process.

Repro: FastDateFormat.getInstance("YYYY").format(cal) where cal is 15 July 42 BC, that is a GregorianCalendar with ERA set to GregorianCalendar.BC.
Expected: -0041, what SimpleDateFormat prints for the same date and pattern.
Actual: 000 followed by U+0007. YY gives ,/ against SimpleDateFormat's -41.

1. GregorianCalendar.getWeekYear() is proleptic, so a BC date gives a negative week year (-41 here) and WeekYear hands it to the wrapped NumberRule unchanged.
2. The digit rules build each character arithmetically ((char) (value + '0'), value / 10 + '0'), so a negative value walks below '0' and lands on control characters or unrelated code points rather than digits.

Fix: emit the sign in WeekYear and pass the magnitude down, the same handling TimeZoneNumberRule and Iso8601_Rule already apply to negative UTC offsets. Every other rule takes its value from Calendar.get, which never returns a negative, so getWeekYear is the only source of one; YYY and Y keep the four-digit padding the class Javadoc describes.

Checked YYYY against SimpleDateFormat over 46,667 dates spanning 249 BC to 4188 AD: 2,611 mismatches before this change, 337 of them carrying a control character, and 0 after.

[Copilot]

Pull request overview

Fixes incorrect formatting of negative week-based years in FastDatePrinter when using Y/YY/YYYY… patterns with BC dates (where GregorianCalendar.getWeekYear() can return a negative, proleptic value). The change brings FastDateFormat/FastDatePrinter output in line with SimpleDateFormat and prevents control-character output from the digit rules.

Changes:

• Add sign handling to the WeekYear rule so negative week years are rendered with - and the magnitude is passed to the underlying NumberRule.
• Add a regression test that compares multiple Y-width patterns against SimpleDateFormat for a BC date and asserts exact expected strings.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
src/main/java/org/apache/commons/lang3/time/FastDatePrinter.java	WeekYear now emits - and formats the absolute value to avoid negative-digit rendering.
src/test/java/org/apache/commons/lang3/time/FastDatePrinterTest.java	Adds BC week-year regression coverage and parity assertions vs SimpleDateFormat.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[garydgregory]

Thank you @alhudz , merged 🚀