Releases: dbt-labs/dbt-bigquery
Releases · dbt-labs/dbt-bigquery
dbt-bigquery v1.5.8
dbt-bigquery 1.5.8 - February 01, 2024
Fixes
- Fix contact validation failure on query data types change (#861)
Contributors
dbt-bigquery v1.7.3
dbt-bigquery v1.7.2
dbt-bigquery v1.6.9
dbt-bigquery 1.6.9 - November 09, 2023
Fixes
- Support agate Integer type, test with empty seed (#1003)
dbt-bigquery v1.7.1
dbt-bigquery 1.7.1 - November 07, 2023
Fixes
- Support agate Integer type, test with empty seed (#1003)
dbt-bigquery v1.7.0
dbt-bigquery 1.7.0 - November 02, 2023
Features
- Add support for materialized views (#dbt-labs/dbt-core#6911)
- Support partition_by and cluster_by on python models when supplied in model configurations (#680)
- Change batch_id to model override (#671)
- Support change monitoring for materialized views, including: autorefresh, clustering, partitioning (#924)
- Support storing test failures as views (#6914)
Fixes
- Serverless Spark to Poll with .GetBatch() instead of using operation.result() (#734)
- Avoid creating twice the temp table in dynamic insert overwrite for sql on_schema_change != 'ignore' or python models (#810)
- Requiring Python >=3.8 enables walrus operator (#849)
- Fix contact validation failure on query data types change (#861)
- Avoid setting lifetime on impersonation tokens. (#769)
- case insensitive check on partition matching (#886)
- Time out queries if user supplies
job_execution_timeout
(#231) - Fix bigquery copy materialization (#820)
- changes expected value types to AnyInteger to take into account changes in core (#915)
- Quote column names in struct data types to fix time ingestion partitioning table creation (#913)
- Fix issue where job labels are not rendered when using macro for query comment (#863)
- Patch for json inline --show (#972)
- Lower bound of
2.11.0
forgoogle-api-core
(#979) - Fix refresh syntax, config comparison with empty labels (#983)
- Assign the correct relation type to materialized views in catalog queries (#995)
Under the Hood
- Update stale workflow to use centralized version (#844)
- Re-organize adapters.sql into more granular files inside of macros/relations (#904)
- Update BigQueryCredentials to support migration off hologram (#906)
- Address type annotation issues and remove protected method ref from impl (#933)
- Fixed a mypy failure by reworking BigQueryAdapter constructor. (#934)
- update SQLQuery to include node_info (#936)
- Add tests for inlined limit + sql-header in dbt show query (#940)
- Support for use of type aliases in contract column data_type (#953)
Dependencies
- Update black requirement from ~=23.3 to ~=23.7 (#815)
- Update pip-tools requirement from ~=6.14 to ~=7.0 (#831)
- Update ddtrace requirement from ~=1.16 to ~=1.17 (#839)
- Update wheel requirement from ~=0.40 to ~=0.41 (#843)
- Update flake8 requirement from ~=6.0 to ~=6.1 (#847)
- Update pip-tools requirement from ~=7.0 to ~=7.2 (#851)
- Update pip-tools requirement from ~=7.2 to ~=7.3 (#870)
- Update tox requirement from ~=4.6 to ~=4.7 (#869)
- Update types-protobuf requirement from ~=4.23 to ~=4.24 (#873)
- Bump mypy from 1.4.1 to 1.5.0 (#874)
- Update tox requirement from ~=4.7 to ~=4.8 (#880)
- Update ddtrace requirement from ~=1.17 to ~=1.18 (#881)
- Bump mypy from 1.5.0 to 1.5.1 (#883)
- Update tox requirement from ~=4.8 to ~=4.9 (#884)
- Update tox requirement from ~=4.9 to ~=4.11 (#907)
- Update pre-commit requirement from ~=3.3 to ~=3.4 (#909)
- Update black requirement from ~=23.7 to ~=23.9 (#920)
- Update ddtrace requirement from ~=1.18 to ~=1.19 (#928)
Contributors
- @Kayrnt (#810, #886, #913)
- @cemsbr (#849)
- @cmc333333 (#769)
- @dementiev27 (#861)
- @gmyrianthous (#979)
- @kalanyuz (#680)
- @kodaho (#863)
- @m-sche (#820)
- @matt-winkler (#972)
- @nickozilla (#671)
- @wazi55 (#734)
dbt-bigquery v1.6.8
dbt-bigquery 1.6.8 - November 01, 2023
Fixes
Contributors
dbt-bigquery v1.7.0rc1
dbt-bigquery 1.7.0-rc1 - October 12, 2023
Features
- Support partition_by and cluster_by on python models when supplied in model configurations (#680)
- Support change monitoring for materialized views, including: autorefresh, clustering, partitioning (#924)
- Support storing test failures as views (#6914)
Fixes
- Serverless Spark to Poll with .GetBatch() instead of using operation.result() (#734)
- Avoid setting lifetime on impersonation tokens. (#769)
- Fix bigquery copy materialization (#820)
- Fix issue where job labels are not rendered when using macro for query comment (#863)
Under the Hood
- Address type annotation issues and remove protected method ref from impl (#933)
- Fixed a mypy failure by reworking BigQueryAdapter constructor. (#934)
- update SQLQuery to include node_info (#936)
- Add tests for inlined limit + sql-header in dbt show query (#940)
- Support for use of type aliases in contract column data_type (#953)
Dependencies
- Update ddtrace requirement from ~=1.18 to ~=1.19 (#928)