Changelog for BigQuery Driver

v1.14.0 (2026-10-15)

New features:

  • Add bigquery.user_agent option to customize the user agent

  • Add experimental support for optional job creation with Arrow result format

  • Return additional BigQuery table metadata

Fixes:

  • Map REST API 403 access denials to the UNAUTHORIZED status code

  • Correct Arrow extension metadata for ARRAY<GEOGRAPHY> results

  • Refresh impersonated credentials automatically when bigquery.impersonate.lifetime is unset

  • Skip tables that return 404 errors in GetObjects

v1.13.0 (2026-09-10)

New features:

  • Return hidden datasets from GetObjects

  • Add normalized BigQuery job statistics keys to result set schema metadata

  • Skip inaccessible tables in GetObjects

  • Include links to BigQuery jobs in error messages

  • Map Google Cloud API errors to more specific ADBC status codes

  • Add BIGQUERY:query_id to result set schema metadata

  • Support SCRIPT queries that return result sets

Fixes:

  • Exclude hidden datasets from GetStatistics

  • Cancel server-side BigQuery jobs when statement execution is cancelled or result reading stops early

  • Restrict statement cancellation to active execution

  • Apply polling backoff consistently while waiting for jobs

  • Allow empty impersonation delegate and scope values

v1.12.2 (2026-08-25)

Fixes:

  • Bumped Go version to address security vulnerabilities (CVEs)

v1.12.1 (2026-07-10)

Fixes:

  • Bumped Go version to address security vulnerabilities (CVEs)

v1.12.0 (2026-06-22)

New features:

  • Make option names more consistent (allow bigquery.foo in addition to adbc.bigquery.sql.foo)

  • Allow overriding the gRPC endpoint for the BigQuery Storage Read API to enable (in theory) connecting to the BigQuery emulator

  • Add query statistics to the schema metadata of the result

  • Implement GetStatistics

Fixes:

  • Add the SRID to returned GeoArrow metadata

  • Allow binding Arrow time[ns] values by truncating

  • Auto-detect the credential type from JSON credentials

  • Improve error handling around authentication

  • Handle dry-run queries properly instead of erroring

  • Process the connection URI first to avoid potential nondeterminism in connection options

v1.11.2 (2026-03-10)

Fixes:

  • Handle cancellation during bulk ingest better

v1.11.0 (2026-02-19)

New features:

  • Add experimental support for using the BigQuery Storage Write API for bulk ingest

  • Improve read performance by enabling Arrow compression

Fixes:

  • Wait for table existence to propagate in bulk ingest

v1.10.0 (2026-01-07)

Note

The version number was bumped to align with the upstream arrow-adbc release.

New features:

  • Support bigquery:// connection URIs

  • Add BIGQUERY:type field metadata to note the original database type

  • Support setting the quota project for a connection

  • Support bulk ingest of list and struct types

Fixes:

  • Work around hanging due to infinite retries in Google SDKs

  • Always return decimal256(76,38) for BIGDECIMAL columns

v1.0.0 (2025-09-17)

New features:

  • Support service account impersonation

  • Allow specifying the BigQuery location

  • Error gracefully if the user lacks roles/bigquery.readSessionUser to use BigQuery Storage Read API

  • Enable bulk ingest into a specified project/dataset (catalog/schema)

  • Tag GEOGRAPHY columns with GeoArrow extension metadata

Fixes:

  • Put extension type metadata on the child in the case of nested types (like array-of-geography)

  • Export AdbcDriverBigQueryInit for backwards compatibility and ``AdbcDriverBigqueryInit` for driver managers

  • Always return decimal128(38,9) and ignore the declared precision/scale (which is not used for result data)