🤖 I have created a releasebeepboop

1.27.1 (2022-01-04)

⚠ BREAKING CHANGES

• 2.22.1
• 1.27.2
• 1.27.1

Features

• accept DatasetListItem where DatasetReference is accepted (#597) (c8b5581)
• accept job object as argument toget_job andcancel_job (#617) (f75dcdf)
• accept TableListItem where TableReference is accepted (#1016) (fe16adc)
• addAvroOptions to configure AVRO external data (#994) (1a9431d)
• addClient.delete_job_metadata method to remove job metadata (#610) (0abb566)
• addExternalConfig.connection_id property to connect to external sources (#560) (d93986e)
• addLoadJobConfig.projection_fields to select DATASTORE_BACKUP fields (#736) (c45a738)
• addmax_queue_size argument toRowIterator.to_dataframe_iterable (#575) (f95f415)
• addQueryJob.schema property for dry run queries (#1014) (2937fa1)
• addreload argument to*Job.done() functions (#341) (e51fd45)
• addTableReference.__str__ to get table ID in standard SQL (#405) (53dff2a)
• add always_use_jwt_access (#714) (92fbd4a)
• add BIGNUMERIC support (#527) (cc3394f)
• add constants for MONTH and YEAR time partitioning types (#283) (9090e1c)
• add context manager support to client (#540) (d5c7e11)
• add custom cell magic parser to handle complex--params values (#213) (dcfbac2)
• add determinism level for javascript UDFs (#522) (edd3328)
• add instrumentation to list methods (#239) (fa9f9ca)
• add max_results parameter to some of the QueryJob methods (#698) (2a9618f)
• add method api_repr for table list item (#299) (07c70f0)
• add mtls support to client (#492) (1823cad)
• add opentelemetry tracing (#215) (a04996c)
• add progress bar for magics (#396) (04d0273)
• add progress bar toQueryJob.to_dataframe andto_arrow (#352) (dc78edd)
• Add py.typed for PEP 561 compliance (#976) (96e6bee)
• add Samples section to CONTRIBUTING.rst (#785) (e587029)
• add session and connection properties to QueryJobConfig (#1024) (e4c94f4)
• add standard sql table type, update scalar type enums (#777) (b8b5433)
• add support and tests for struct fields (#146) (fee2ba8)
• add support for decimal target types (#735) (7d2d3e9)
• add support for getting and setting table IAM policy (#144) (f59fc9a)
• add support for INTERVAL data type tolist_rows (#840) (e37380a)
• add support for listing arima, automl, boosted tree, DNN, and matrix factorization models (#328) (502a092)
• add support for materialized views (#408) (57ffc66)
• add support for more detailed DML stats (#758) (36fe86f)
• add support for Parquet options (#679) (d792ce0)
• add support for table snapshots (#740) (ba86b2a)
• add support for transaction statistics (#849) (7f7b1a8)
• add support for unrecognized model types (#401) (168f035)
• add support for user defined Table View Functions (#724) (8c7b839)
• add timeout paramter to load_table_from_file and it dependent methods (#327) (b0dd892)
• add to_api_repr method to Model (#326) (fb401bd)
• add type hints for public methods (#613) (f8d4aaa)
• add with_name() to ScalarQueryParameterType (#644) (6cc6876)
• allow cell magic body to be a $variable (#1053) (3a681e0)
• allow client options to be set in magics context (#322) (5178b55)
• allow queryJob.result() to be called on a dryRun (#1015) (685f06a)
• allow routine references (#378) (f9480dc)
• bigquery: add client_options to base class (#216) (478597a)
• convertBIGNUMERIC values to decimal objects (#414) (d472d2d)
• DB API cursors are now iterable (#618) (e0b373d)
• detect obsolete BQ Storage extra at runtime (#666) (bd7dbda)
• enable unsetting policy tags on schema fields (#703) (18bb443)
• expose require_partition_filter for hive_partition (#257) (aa1613c)
• expose reservation usage stats on jobs (#524) (4ffb4e0)
• Extended DB API parameter syntax to optionally provide parameter types (#626) (8bcf397)
• include key metadata in Job representation (#964) (acca1cb)
• list_tables, list_projects, list_datasets, list_models, list_routines, and list_jobs now accept a page_size parameter to control page size (#686) (1f1c4b7)
• make it easier to disable best-effort deduplication with streaming inserts (#734) (1246da8)
• make QueryJob.done() method more performant (#544) (a3ab9ef)
• make the sameTable* instances equal to each other (#867) (c1a3d44)
• pass retry from Job.result() to Job.done() (#41) (284e17a)
• promoteRowIterator.to_arrow_iterable to public method (#1073) (21cd710)
• raise error for unknown properties in job config (#446) (1526e39)
• retry failed query jobs inresult() (#837) (519d99c)
• retry google.auth TransportError by default (#624) (34ecc3f)
• set the X-Server-Timeout header when timeout is set (#927) (ba02f24)
• supportScalarQueryParameterType fortype_ argument inScalarQueryParameter constructor (#850) (93d15e2)
• support authorized dataset entity (#1075) (c098cd0)
• support CSV format inload_table_from_dataframe pandas connector (#399) (0046742)
• Support parameterized NUMERIC, BIGNUMERIC, STRING, and BYTES types (#673) (45421e7)
• Support passing struct data to the DB API (#718) (38b3ef9)
• support Python 3.10 (#1043) (5bbb832)
• support script options in query job config (#690) (1259e16)
• Support using GeoPandas for GEOGRAPHY columns (#848) (16f65e6)
• Update proto definitions for bigquery/v2 to support new proto fields for BQML. (#817) (fe7a902)
• use pyarrow stream compression, if available (#593) (dde9dc5)

Bug Fixes

• insert_rows() accepts float column values as strings again (#824) (d9378af)
• add ConnectionError to default retry (#571) (a3edb8b)
• add DECIMAL and BIGDECIMAL as aliases for NUMERIC and BIGNUMERIC (#638) (aa59023)
• add minimum timeout to getQueryResults API requests (#444) (015a73e)
• add missing spaces in opentelemetry log message (#360) (4f326b1)
• address issues with concurrent BigQuery tests [(#3426)](fix: address issues with concurrent BigQuery tests GoogleCloudPlatform/python-docs-samples#3426) (34ad9c6)
• apply timeout to all resumable upload requests (#1070) (3314dfb)
• Arrow extension-type metadata was not set when calling the REST API or when there are no rows (#946) (864383b)
• avoid floating point for timestamp ininsert_rows (#393) (a1949ae)
• avoid overly strict dependency on pyarrow 3.x (#564) (97ee6ec)
• avoid policy tags 403 error inload_table_from_dataframe (#557) (84e646e)
• avoid possible job already exists error (#751) (45b9308)
• consistent percents handling in DB API query (#619) (6502a60)
• converting to dataframe with out of bounds timestamps (#209) (8209203)
• create_job method accepts dictionary arguments (#300) (155bacc)
• dbapi: allow rows to be fetched from scripts (#387) (b899ad1)
• dbapi: avoid running % format with no query parameters (#348) (5dd1a5e)
• disambiguate missing policy tags from explicitly unset policy tags (#983) (f83c00a)
• don't try to close closed cursors (#498) (bf44e7b)
• error if eval()-ing repr(SchemaField) (#1046) (13ac860)
• error inserting DataFrame with REPEATED field (#925) (656d2fa)
• error using empty array of structs parameter (#474) (c1d15f4)
• executemany rowcount only reflected the last execution (#660) (aeadc8c)
• fix dependency issue in fastavro (#241) (2874abf)
• guard imports against unsupported pyarrow versions (#934) (b289076)
• handle null values in array query parameters (#426) (78fde4a)
• increase default retry deadline to 10 minutes (#859) (30770fd)
• inserting non-finite floats with insert_rows() (#728) (d047419)
• invalid conversion of timezone-aware datetime values to JSON (#480) (61b4385)
• issue a warning if buggy pyarrow is detected (#787) (e403721)
• make TimePartitioning repr evaluable (#110) (20f473b)
• make unicode characters working well in load_table_from_json (#865) (ad9c802)
• missing license headers in new test files (#604) (df48cc5)
• no longer raise a warning into_dataframe ifmax_results set (#815) (3c1be14)
• populate defaulttimeout and retry after client-side timeout (#896) (b508809)
• preserve timestamp microsecond precision with rows from REST API (#402) (04510a7)
• QueryJob.exception()returns the errors, not raises them (#467) (d763279)
• raise error if inserting rows with unknown fields (#163) (8fe7254)
• rationalize platform constraints for 'pyarrow' extra (#235) (c9a0567)
• reading the labels attribute on Job instances (#471) (80944f0)
• remove DB-API dependency on pyarrow with decimal query parameters (#551) (1b946ba)
• remove default timeout (#974) (1cef0d4)
• remove py.typed since package fails mypy check (#988) (39030f2)
• remove pytz dependency and require pyarrow>=3.0.0 (#875) (2cb3563)
• remove unnecessary dependency on libcst (#308) (c055930)
• retry ChunkedEncodingError by default (#802) (419d36d)
• RowIterator.to_arrow() error when BQ Storage client cannot be created (#181) (7afa3d7)
• support ARRAY data type when loading from DataFrame with Parquet (#980) (1e59083)
• tests: invalid path to strptime() (#672) (591cdd8)
• The DB API Binary function accepts bytes data (#630) (4396e70)
• tweak pyarrow extra to soothe PyPI (#230) (c15efbd)
• underscores weren't allowed in struct field names when passing parameters to the DB API (#930) (fcb0bc6)
• unsetting clustering fileds on Table is now possible (#622) (33a871f)
• update minimum dependency versions (#263) (1be66ce)
• use a larger chunk size when loading data (#799) (b804373)
• use debug logging level for OpenTelemetry message (#442) (7ea6b7c)
• use explicitly given project over the client's default project for load jobs (#482) (530e1e8)
• use pandas function to check for NaN (#750) (67bc5fb)
• use REST API in cell magic when requested (#892) (1cb3e55)
• use version.py instead of pkg_resources.get_distribution (#307) (b8f502b)
• validate job_config.source_format in load_table_from_dataframe (#262) (6160fee)

Performance Improvements

• add size parameter for load table from dataframe and json methods (#280) (3be78b7)
• avoid extra API calls fromto_dataframe if all rows are cached (#384) (c52b317)
• cache first page ofjobs.getQueryResults rows (#374) (86f6a51)
• don't fetch rows when waiting for query to finish (#400) (730df17)
• remove redundant array deepcopy (#26) (b54f867)
• usegetQueryResults from DB-API (#375) (30de15f)
• usejobs.getQueryResults to download result sets (#363) (0c3476d)

Miscellaneous Chores

• release 2.22.1 (#794) (46e65a6)

This PR was generated withRelease Please. Seedocumentation.