[fix][managed-ledger] Honor batch-index cap in isCursorDataFullyPersistable

[arimu1]

Fixes #26499

Motivation

ManagedCursorImpl.isCursorDataFullyPersistable() only compared the whole-entry
deletion range count to maxUnackedRangesToPersist. When partial batch ACK state
exceeded maxBatchDeletedIndexToPersist, the method still returned true even
though buildBatchEntryDeletionIndexInfoList() truncates on persist. That
mismatch can leave dispatchers running while some batch ACK records would not
survive a cursor-state write.

Modifications

• Extend isCursorDataFullyPersistable() to return false when
  batchDeletedIndexes exceeds maxBatchDeletedIndexToPersist.
• Add a managed-ledger test reproducing the issue scenario (three partial batch
  ACK records with limit 2).

No change to PIP-299 dispatcher pause policy beyond making the persistability
check accurate.

Verifying this change

• Make sure that the change passes the CI checks.

This change added tests and can be verified as follows:

• ./gradlew :managed-ledger:test --tests org.apache.bookkeeper.mledger.impl.ManagedCursorBatchAckRecoveryTest.testIsCursorDataFullyPersistableReflectsBatchDeletedIndexLimit

Does this pull request potentially affect one of the following parts:

• Dependencies (add or upgrade a dependency)
• The public API
• The schema
• The default values of configurations
• The threading model
• The binary protocol
• The REST endpoints
• The admin CLI options
• The metrics
• Anything that affects deployment

[void-ptr974]

One coordination note for future contributions: I opened #26499 and indicated that I was willing to submit a PR, although I had not opened one yet. In this situation, it would be helpful to leave a short comment before starting work, so contributors can avoid duplicating effort. No problem this time—just please coordinate on the issue first in similar cases.

One change is required: the new org.awaitility.Awaitility import is out of order. I verified that ./gradlew :managed-ledger:checkstyleTest fails with an ImportOrder violation. Please move it after the org.apache... imports.

I did not find another blocker in the current diff.