Search before asking
Description
LookupResult::to_record_batch and the limit scanner's KV decode do the same thing — [schema_id|value] -> FixedSchemaDecoder -> RecordBatch - with separate decoder caches and a duplicated read_schema_id. Extract one shared helper, the future snapshot scanner will want it too.
Touches the lookup hot path, so do it carefully on its own.
Willingness to contribute
Search before asking
Description
LookupResult::to_record_batch and the limit scanner's KV decode do the same thing — [schema_id|value] -> FixedSchemaDecoder -> RecordBatch - with separate decoder caches and a duplicated read_schema_id. Extract one shared helper, the future snapshot scanner will want it too.
Touches the lookup hot path, so do it carefully on its own.
Willingness to contribute