Skip to main content

Module storage_read

Module storage_read 

Source
Available on crate feature arrow only.
Expand description

BigQuery Storage Read API (gRPC) Arrow path (#380). Only compiled with the arrow feature. BigQuery Storage Read API (gRPC) Arrow path (#380).

Reads a table directly as Arrow RecordBatches via the google.cloud.bigquery.storage.v1 gRPC service — no jobs.query, no per-row JSON. Used both by the columnar fast path (BigQuerySource::stream_batches) and, when the sink is not columnar, by the row path (Arrow → JSON) so a read_api source works with any sink.

The gcloud-* stack here (gRPC over tonic 0.14, auth via gcloud-auth) is a deliberately separate family from the REST gcp-bigquery-client used for the query path; it is only compiled with the arrow feature.

Functions§

resolve_table
Resolve dataset.table / project.dataset.table into the Storage Read API resource name projects/{p}/datasets/{d}/tables/{t}. Pure.
stream_batches_arrow
Columnar fast path: yield one ColumnarPage per Arrow batch.
stream_pages_arrow
Row path for a non-columnar sink: decode Arrow batches to JSON and re-frame into StreamPages of batch_size records.