Expand description
Append pipeline for TimeSeriesTable.
This module contains the core append implementation plus the public wrappers. It is responsible for:
- loading/deriving segment metadata and logical schema,
- adopting the first schema or normalizing into the registered schema,
- computing segment coverage, detecting overlaps, and writing coverage sidecars,
- optimistic commit to the transaction log and in-memory state update. Keep new append-time invariants here so the flow remains centralized.
Structs§
- Append
Request - One Arrow source plus physical settings for a single append.
Enums§
- Parquet
Compression - Compression used for Parquet segments written by append.
Traits§
- Into
Record Batch Reader - Convert an Arrow batch source into a schema-bearing
RecordBatchReader.