Fix RangeError when using stepping with year field wildcard

[Copilot]

*/2 in the year field throws RangeError: CronPattern: Invalid value for year: 0 (supported range: 1-9999) because handleStepping starts the loop at index 0, which is invalid for years (valid range 1–9999).

The fix keeps the loop start at 0 (consistent with all other fields) but skips invalid year index 0, so */2 for years produces even years (2024, 2026, 2028, ...) — matching the even-tick behavior of other fields like seconds (*/2 → 0, 2, 4, ...).

• src/pattern.ts: Skip year index 0 in the handleStepping loop to avoid the RangeError while preserving even-tick alignment
• test/new-edge-cases.test.ts: Update test to assert */2 works correctly (produces even years) rather than expecting a throw

// Previously threw RangeError, now works
const job = new Cron('* * * * * * */2');

[Copilot]

Pull request overview

This PR fixes year-field stepping with wildcards (e.g., */2) so it no longer attempts to start from year index 0 (invalid under the supported year range 1–9999), and updates the edge-case test to validate the corrected behavior.

Changes:

• Adjust handleStepping to start wildcard stepping at year 1 when the part type is "year".
• Update the edge-case test to assert */2 in the year field produces the expected next run years instead of throwing.

Reviewed changes

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

File	Description
src/pattern.ts	Starts wildcard year stepping at 1 to avoid invalid year 0.
test/new-edge-cases.test.ts	Replaces the previous “should throw” assertion with expectations for odd-year stepping results.

---

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

[Hexagon]

@copilot If other steppings use even ticks for */2, i believe years should too. Is this covered by ocps? If not, change implememtation to handle even year without breaking anything. Keep changes minimal.

[Copilot]

@copilot If other steppings use even ticks for */2, i believe years should too. Is this covered by ocps? If not, change implememtation to handle even year without breaking anything. Ke...

OCPS doesn't specify this behavior. Changed the implementation in 85a8293 — now */2 for years produces even years (2026, 2028, 2030) matching the even-tick behavior of other fields. Instead of starting at 1, the loop starts at 0 (like all other fields) and skips invalid year index 0 with a continue.

[Hexagon]

@copilot Make an unreleased changelog entry and check if docs needs a note somewhere.

[Copilot]

@copilot Make an unreleased changelog entry and check if docs needs a note somewhere.

Added unreleased changelog entry in 05503c5. Checked the docs — year field stepping is already documented in the pattern table (docs/src/usage/pattern.md, line 54: Year | Optional | 1-9999 | * , - /), so no additional doc note is needed since this is a bugfix for existing documented behavior.

[Hexagon]

This needs to be clarified in ocps before merging, pausing.