[TDB-247] Wait for region leader election before region migration

[CRZbulabula]

Description

After a ConfigNode leader switch, RemoveRegionPeerProcedure can resume before the Region leader cache has been rebuilt. RegionMaintainHandler consults that cache while selecting a replica and transferring a Ratis leader. A missing or explicitly unknown entry can otherwise cause a NullPointerException, terminating the procedure before region removal completes.

Centralize Region leader lookup in a helper that treats a missing or -1 entry as unknown, waits for the existing LoadManager.waitForRegionGroupReady leader-election and priority-update cycle, and then reads the cache again. If the leader is still unavailable after the wait, the helper retains the -1 fallback so the procedure can continue safely without dereferencing null. Both replica selection and Ratis leader transfer use this lookup.

Add regression coverage for DataRegions and SchemaRegions covering missing and unknown leaders, waiting and rereading the cache after election, avoiding an unnecessary wait for a known leader, successful transfer and cache updates, and a cache entry disappearing between lookups.

Validation

• mvn -B -pl iotdb-core/confignode -Dtest=RegionMaintainHandlerTest,RegionMaintainHandlerConsensusPipeTest,RemoveRegionPeerProcedureTest,RegionMigrateProcedureTest -Dsurefire.failIfNoSpecifiedTests=false -DfailIfNoTests=false test (28 tests passed)
• mvn -B spotless:apply -pl iotdb-core/confignode
• mvn -B test-compile -DskipTests -P with-zh-locale

Cluster failover integration testing was not run.

This PR has:

• been self-reviewed.
• added unit tests to cover the affected code paths.

Key changed/added classes

• RegionMaintainHandler
• RegionMaintainHandlerTest