v3.26.0

Compare Source

Features

• Include LegacyPandasError in init imports (#​2014) (3ab5e95)
• Use bigquery-magics package for the %%bigquery magic (#​1965) (60128a5)

Bug Fixes

• Add docfx to the presubmit configuration and delete docs-presubmit (#​1995) (bd83cfd)
• Add warning when encountering unknown field types (#​1989) (8f5a41d)
• Allow protobuf 5.x; require protobuf >=3.20.2; proto-plus >=1.22.3 (#​1976) (57bf873)
• Do not set job timeout extra property if None (#​1987) (edcb79c)
• Set pyarrow field nullable to False for a BigQuery field in REPEATED mode (#​1999) (5352870)

Dependencies

• Bump min version of google-api-core and google-cloud-core to 2.x (#​1972) (a958732)

Documentation

• Add short mode query sample & test (#​1978) (ba61a8a)
• Improve QueryJobConfig.destination docstring (#​2016) (1b4cca0)

v3.25.0

Compare Source

Features

• Add prefer_bqstorage_client option for Connection (#​1945) (bfdeb3f)
• Support load job option ColumnNameCharacterMap (#​1952) (7e522ee)

Bug Fixes

• Do not overwrite page_size with max_results when start_index is set (#​1956) (7d0fcee)

v3.24.0

Compare Source

Features

• Add default timeout for Client.get_job() (#​1935) (9fbad76)
• Add support for map target type in Parquet options (#​1919) (c3f7b23)

Bug Fixes

• Create query job in job.result() if doesn't exist (#​1944) (8f5b4b7)
• Retry is_job_done on ConnectionError (#​1930) (4f72723)

Performance Improvements

• If page_size or max_results is set on QueryJob.result(), use to download first page of results (#​1942) (3e7a48d)