Update tantivy to latest (04beab3b29) (#6289)

Notable changes:
- Fix deduplicate doc counts in term aggregation for multi-valued fields
- Improve union performance for non-score unions
- Use BinaryHeap for score-based top-K collection
- Optimized intersection count using a bitset when the first leg is dense
- Make BucketEntries::iter, PercentileValuesVecEntry fields, and TopNComputer::threshold public
- Performance improvement for nested cardinality aggregation

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>