Necessary information for compute node to access data in the external database.
Compute node will use this information to connect to the external database and scan the table.
When there is no primary key specified while creating source, will use
the message key as primary key in BYTEA type with this name.
Note: the field has version to track, please refer to default_key_column_name_version_mapping
Iceberg snapshot summary property containing the highest RisingWave epoch committed by the
sink. Append-only Iceberg engine batch reads use it to exclude committed rows from the sink
log store.
Get distribution key start index in pk, and return None if dist_key_in_pk_indices is not empty
or continuous.
Note that dist_key_in_pk_indices may be shuffled, the start index should be the
minimum value.
FIXME: Perhaps we should use sth like ColumnIdGenerator::new_alter,
However, the SourceVersion is problematic: It doesn’t contain next_col_id.
(But for now this isn’t a large problem, since drop column is not allowed for source yet..)