Expand description
§faucet-source-iceberg
Apache Iceberg table source for the faucet-stream ecosystem. Reads a table
through a REST / Glue / SQL / HMS catalog (shared with the Iceberg sink via
faucet-common-iceberg) with column projection, a pushed-down row filter,
time travel (snapshot_id / as_of_timestamp), and incremental reads of
the data files added by append snapshots since a bookmarked snapshot.
Re-exports§
pub use config::IcebergSourceConfig;pub use config::OnExpired;pub use config::OnRewrite;pub use config::ReadMode;pub use stream::IcebergSource;
Modules§
- config
- Configuration types for the Apache Iceberg source.
- convert
- Pure conversions: Arrow batches to JSON rows, Iceberg schemas to JSON Schema, discovery descriptors, and the delete-file support check.
- filter
- The
filter:predicate DSL: parsing into anExprtree and typing it against a table schema as an IcebergPredicatefor scan pushdown. - snapshots
- Pure snapshot-lineage logic: time-travel resolution, incremental planning, and the bookmark shape.
- stream
- The Iceberg source: plans a run from the table’s snapshot lineage, scans the selected data files through iceberg-rust, and yields Arrow batches as rows or columnar pages.
Enums§
- Catalog
Config - Iceberg catalog type and its connection settings.
Constants§
- DESCRIPTOR
- The connector’s static descriptor (RFC 0013): its capabilities, answered from config alone.