Changelog category (leave one):

• Bug Fix (user-visible misbehavior in an official stable release)

Changelog entry (a user-readable short description of the changes that goes to CHANGELOG.md):

Fix incorrect partition pruning for toWeek() that caused queries with WHERE toWeek(date, mode) = N to return empty results for weeks 49-52 on tables partitioned by toYYYYMM(date).

Summary

toWeek() incorrectly claimed monotonicity via its FactorTransform (ToStartOfYearImpl), which caused partition pruning to skip partitions containing December dates when filtering by week numbers 49-52.

The root cause is that toWeek() with certain week modes (e.g. mode 3, ISO weeks) can have week numbers that wrap at year boundaries (week 52 → 1 in late December). The ToStartOfYearImpl factor transform does not detect this wrapping because both December 1 and December 31 are in the same calendar year, so it incorrectly reports the function as monotonic. The optimizer then computes an invalid value range (e.g. [49, 1]) and prunes the partition entirely.

No single FactorTransform can correctly handle all week modes because wrapping behavior depends on the runtime week_mode parameter and different modes wrap at different dates.

Fix: Add a hasMonotonicity() static method to each custom week Transform struct. ToWeekImpl returns false; all other transforms return true. IFunctionCustomWeek delegates to Transform::hasMonotonicity() instead of unconditionally returning true, following the same pattern used by hasPreimage() in DateTimeTransforms.h.

Closes #90240