Changelog

PyPI History

2.1.0 (2020-10-08)

Features

add constants for MONTH and YEAR time partitioning types (#283) (9090e1c)

Bug Fixes

remove unnecessary dependency on libcst (#308) (c055930)

Performance Improvements

remove redundant array deepcopy (#26) (b54f867)

Documentation

samples: add create_table_clustered code snippet (#291) (d1eb8b3)

2.0.0

09-30-2020 14:51 PDT

Implementation Changes

Transition the library to microgenerator. (#278) This is a breaking change that drops support for Python 2.7 and 3.5 and brings a few other changes. See migration guide for more info.

Internal / Testing Changes

Update protoc-generated comments (via synth). (#270)
Add CI secrets manager (via synth). (#271)

1.28.0 (2020-09-22)

Features

add custom cell magic parser to handle complex --params values (#213) (dcfbac2)
add instrumentation to list methods (#239) (fa9f9ca)
add opentelemetry tracing (#215) (a04996c)
expose require_partition_filter for hive_partition (#257) (aa1613c)

Bug Fixes

fix dependency issue in fastavro (#241) (2874abf)
update minimum dependency versions (#263) (1be66ce)
validate job_config.source_format in load_table_from_dataframe (#262) (6160fee)

Documentation

recommend insert_rows_json to avoid call to tables.get (#258) (ae647eb)

1.27.2 (2020-08-18)

Bug Fixes

rationalize platform constraints for ‘pyarrow’ extra (#235) (c9a0567)

1.27.1 (2020-08-18)

Bug Fixes

tweak pyarrow extra to soothe PyPI (#230) (c15efbd)

1.27.0 (2020-08-15)

Features

add support and tests for struct fields (#146) (fee2ba8)
add support for getting and setting table IAM policy (#144) (f59fc9a)
bigquery: add client_options to base class (#216) (478597a)

Bug Fixes

converting to dataframe with out of bounds timestamps (#209) (8209203), closes #168
raise error if inserting rows with unknown fields (#163) (8fe7254)

1.26.1 (2020-07-25)

Documentation

Migrated code samples from https://github.com/GoogleCloudPlatform/python-docs-samples

Bug Fixes

RowIterator.to_arrow() error when BQ Storage client cannot be created (#181) (7afa3d7)

Dependencies

Updated version constraints on grmp dependency in anticipation of 1.0.0 release (#189)

1.26.0 (2020-07-20)

Features

use BigQuery Storage client by default (if dependencies available) (#55) (e75ff82), closes #91
bigquery: add eq method for class PartitionRange and RangePartitioning (#162) (0d2a88d)
bigquery: expose date_as_object parameter to users (#150) (a2d5ce9)
bigquery: expose date_as_object parameter to users (#150) (cbd831e)

Bug Fixes

dry run queries with DB API cursor (#128) (bc33a67)
omit NaN values when uploading from insert_rows_from_dataframe (#170) (f9f2f45)

Documentation

bigquery: add client thread-safety documentation (#132) (fce76b3)
bigquery: add docstring for conflict exception (#171) (9c3409b)
bigquery: consistent use of optional keyword (#153) (79d8c61)
bigquery: fix the broken docs (#139) (3235255)

1.25.0 (2020-06-06)

Features

add BigQuery storage client support to DB API (#36) (ba9b2f8)
bigquery: add create job method (#32) (2abdef8)
bigquery: add support of model for extract job (#71) (4a7a514)
add HOUR support for time partitioning interval (#91) (0dd90b9)
add support for policy tags (#77) (38a5c01)
make AccessEntry objects hashable (#93) (23a173b)
bigquery: expose start index parameter for query result (#121) (be86de3)
bigquery: unit and system test for dataframe with int column with Nan values (#39) (5fd840e)

Bug Fixes

allow partial streaming_buffer statistics (#37) (645f0fd)
distinguish server timeouts from transport timeouts (#43) (a17be5f)
improve cell magic error message on missing query (#58) (6182cf4)
bigquery: fix repr of model reference (#66) (26c6204)
bigquery: fix start index with page size for list rows (#27) (400673b)

1.24.0

02-03-2020 01:38 PST

Implementation Changes

Fix inserting missing repeated fields. (#10196)
Deprecate client.dataset() in favor of DatasetReference. (#7753)
Use faster to_arrow + to_pandas in to_dataframe() when pyarrow is available. (#10027)
Write pandas datetime[ns] columns to BigQuery TIMESTAMP columns. (#10028)

New Features

Check rows argument type in insert_rows(). (#10174)
Check json_rows arg type in insert_rows_json(). (#10162)
Make RowIterator.to_dataframe_iterable() method public. (#10017)
Add retry parameter to public methods where missing. (#10026)
Add timeout parameter to Client and Job public methods. (#10002)
Add timeout parameter to QueryJob.done() method. (#9875)
Add create_bqstorage_client parameter to to_dataframe() and to_arrow() methods. (#9573)

Dependencies

Fix minimum versions of google-cloud-core and google-resumable-media dependencies. (#10016)

Documentation

Fix a comment typo in job.py. (#10209)
Update code samples of load table file and load table URI. (#10175)
Uncomment Client constructor and imports in samples. (#10058)
Remove unused query code sample. (#10024)
Update code samples to use strings for table and dataset IDs. (#9974)

Internal / Testing Changes

Bump copyright year to 2020, tweak docstring formatting (via synth). #10225
Add tests for concatenating categorical columns. (#10180)
Adjust test assertions to the new default timeout. (#10222)
Use Python 3.6 for the nox blacken session (via synth). (#10012)

1.23.1

12-16-2019 09:39 PST

Implementation Changes

Add iamMember entity type to allowed access classes. (#9973)
Fix typo in import error message (pandas -> pyarrow). (#9955)

Dependencies

Add six as an explicit dependency. (#9979)

Documentation

Add sample to read from query destination table. (#9964)

1.23.0

12-11-2019 13:31 PST

New Features

Add close() method to client for releasing open sockets. (#9894)
Add support of use_avro_logical_types for extract jobs. (#9642)
Add support for hive partitioning options configuration. (#9626)
Add description for routine entities. (#9785)

Documentation

Update code samples to use strings for table and dataset IDs. (#9495)

Internal / Testing Changes

Run unit tests with Python 3.8. (#9880)
Import Mapping from collections.abc not from collections. (#9826)

1.22.0

11-13-2019 12:23 PST

Implementation Changes

Preserve job config passed to Client methods. (#9735)
Use pyarrow fallback for improved schema detection. (#9321)
Add TypeError if wrong job_config type is passed to client job methods. (#9506)
Fix arrow deprecation warning. (#9504)

New Features

Add --destination_table parameter to IPython magic. (#9599)
Allow passing schema as a sequence of dicts. (#9550)
Implement defaultEncryptionConfiguration on datasets. (#9489)
Add range partitioning to tables, load jobs, and query jobs. (#9477)

Dependencies

Pin google-resumable-media to includ 0.5.x. (#9572)

Documentation

Fix link anchors in external config docstrings. (#9627)
Add python 2 sunset banner to documentation. (#9036)
Add table create sample using integer range partitioning. (#9478)
Document how to achieve higher write limit and add tests. (#9574)
Add code sample for scripting. (#9537)
Rewrite docs in Google style, part 2. (#9481)
Use multi-regional key path for CMEK in snippets. (#9523)

Internal / Testing Changes

Fix undelete table system test to use milliseconds in snapshot decorator. (#9649)
Format code with latest version of black. (#9556)
Remove duplicate test dependencies. (#9503)

1.21.0

10-16-2019 10:33 PDT

New Features

add ability to pass in a table ID instead of a query to the %%bigquery magic (#9170)
add support for custom QueryJobConfig in BigQuery.cursor.execute method (#9278)
store QueryJob to destination var on error in %%bigquery magic (#9245)
add script statistics to job resource (#9428)
add support for sheets ranges (#9416)
add support for listing jobs by parent job (#9225)
expose customer managed encryption key for ML models (#9302)
add Dataset.default_partition_expiration_ms and Table.require_partition_filter properties (#9464)

Dependencies

restrict version range of google-resumable-media (#9243)

Documentation

document how to load data as JSON string (#9231)
standardize comments and formatting in existing code samples (#9212)
rewrite docstrings in Google style (#9326)
fix incorrect links to REST API in reference docs (#9436)

Internal / Testing Changes

add code samples to lint check (#9277)
update code samples to use strings for table and dataset IDs (#9136)
simplify scripting system test to reduce flakiness (#9458)

1.20.0

09-13-2019 11:22 PDT

Implementation Changes

Change default endpoint to bigquery.googleapis.com (#9213)
Change the default value of Cursor instances’ arraysize attribute to None (#9199)
Deprecate automatic schema conversion. (#9176)
Fix list_rows() max results with BQ storage client (#9178)

New Features

Add Model.encryption_config. (via synth) (#9214)
Add Client.insert_rows_from_dataframe() method (#9162)
Add support for array parameters to Cursor.execute(). (#9189)
Add support for project IDs with org prefix to Table.from_string() factory. (#9161)
Add --max_results option to Jupyter magics (#9169)
Autofetch table schema on load if not provided. (#9108)
Add max_results parameter to QueryJob.result(). (#9167)

Documentation

Fix doc link. (#9200)

Internal / Testing Changes

Revert “Disable failing snippets test (#9156).” (#9220)

1.19.0

09-03-2019 14:33 PDT

Implementation Changes

Raise when unexpected fields are present in the LoadJobConfig.schema when calling load_table_from_dataframe. (#9096)
Determine the schema in load_table_from_dataframe based on dtypes. (#9049)
Raise helpful error when loading table from dataframe with STRUCT columns. (#9053)
Fix schema recognition of struct field types. (#9001)
Fix deserializing None in QueryJob for queries with parameters. (#9029)

New Features

Include indexes in table written by load_table_from_dataframe, only if fields corresponding to indexes are present in LoadJobConfig.schema. (#9084)
Add client_options to constructor. (#8999)
Add --dry_run option to %%bigquery magic. (#9067)
Add load_table_from_json() method to create a table from a list of dictionaries. (#9076)
Allow subset of schema to be passed into load_table_from_dataframe. (#9064)
Add support for unsetting LoadJobConfig.schema. (#9077)
Add support to Dataset for project IDs containing an org prefix. (#8877)
Add enum with SQL type names allowed to be used in SchemaField. (#9040)

Documentation

Fix the reference URL for Client.create_dataset(). (#9149)
Update code samples to use strings for table names instead of client.dataset(). (#9032)
Remove compatability badges from READMEs. (#9035)
Fix Pandas DataFrame load example under Python 2.7. (#9022)

Internal / Testing Changes

Disable failing snippets test for copying CMEK-protected tables. (#9156)
Fix BigQuery client unit test assertions (#9112)
Replace avro with arrow schemas in test_table.py (#9056)

1.18.0

08-08-2019 12:28 PDT

New Features

Add bqstorage_client param to QueryJob.to_arrow() (#8693)
Include SQL query and job ID in exception messages. (#8748)
Allow using TableListItem to construct a Table object. (#8738)
Add StandardSqlDataTypes enum to BigQuery (#8782)
Add to_standard_sql() method to SchemaField (#8880)
Add debug logging statements to track when BQ Storage API is used. (#8838)
Hide error traceback in BigQuery cell magic (#8808)
Allow choice of compression when loading from dataframe (#8938)
Additional clustering metrics for BQML K-means models (via synth). (#8945)

Documentation

Add compatibility check badges to READMEs. (#8288)
Link to googleapis.dev documentation in READMEs. (#8705)
Remove redundant service account key code sample. (#8891)

Internal / Testing Changes

Fix several pytest “skip if” markers (#8694)
Update tests to support conversion of NaN as NULL in pyarrow 0.14.\*. (#8785)
Mock external calls in one of BigQuery unit tests (#8727)
Set IPython user agent when running queries with IPython cell magic (#8713)
Use configurable bucket name for GCS samples data in systems tests. (#8783)
Move maybe_fail_import() to top level test utils (#8840)
Set BQ Storage client user-agent when in Jupyter cell (#8734)

1.17.0

07-12-2019 07:56 PDT

New Features

Support faster Arrow data format in to_dataframe when using BigQuery Storage API. (#8551)
Add to_arrow to get a pyarrow.Table from query results. (#8609)

Dependencies

Exclude bad 0.14.0 pyarrow release. (#8551)

1.16.0

07-01-2019 10:22 PDT

New Features

Add Routines API. (#8491)
Add more stats to Models API, such as optimization_strategy (via synth). (#8344)

Documentation

Add docs job to publish to googleapis.dev. (#8464)
Add sample demonstrating how to create a job. (#8422)
Use autodetected location in code samples. (#8340, #8341)

Internal / Testing Changes

Refactor to_dataframe to deterministicly update progress bar. (#8303)

1.15.0

06-14-2019 10:10 PDT

Implementation Changes

Fix bug where load_table_from_dataframe could not append to REQUIRED fields. (#8230)

New Features

Add page_size parameter to QueryJob.result. (#8206)

1.14.0

06-04-2019 11:11 PDT

New Features

Add maximum_bytes_billed argument and context.default_query_job_config property to magics. (#8179)

Dependencies

Don’t pin google-api-core in libs using google-cloud-core. (#8213)

1.13.0

05-31-2019 10:22 PDT

New Features

Use job_config.schema for data type conversion if specified in load_table_from_dataframe. (#8105)

Internal / Testing Changes

Adds private _connection object to magics context. (#8192)
Fix coverage in ‘types.py’ (via synth). (#8146)

1.12.1

05-21-2019 11:16 PDT

Implementation Changes

Don’t raise error when encountering unknown fields in Models API. (#8083)

Documentation

Use alabaster theme everwhere. (#8021)

Internal / Testing Changes

Add empty lines (via synth). (#8049)

1.12.0

05-16-2019 11:25 PDT

Implementation Changes

Remove duplicates from index on pandas DataFrames returned by to_dataframe(). (#7953)
Prevent error when time partitioning is populated with empty dict (#7904)
Preserve order in to_dataframe with BQ Storage from queries containing ORDER BY (#7793)
Respect progress_bar_type in to_dataframe when used with BQ Storage API (#7697)
Refactor QueryJob.query to read from resource dictionary (#7763)
Close the to_dataframe progress bar when finished. (#7757)
Ensure that KeyboardInterrupt during to_dataframeno longer hangs. (#7698)
Raise ValueError when BQ Storage is required but missing (#7726)
Make total_rows available on RowIterator before iteration (#7622)
Avoid masking auth errors in to_dataframe with BQ Storage API (#7674)

New Features

Add support for passing client_info. (#7849 and (#7806)
Phase 1 for storing schemas for later use. (#7761)
Add destination and related properties to LoadJob. (#7710)
Add clustering_fields property to TableListItem (#7692)
Add created and expires properties to TableListItem (#7684)

Dependencies

Pin google-cloud-core >= 1.0.0, < 2.0dev. (#7993)
Add [all] extras to install all extra dependencies (#7610)

Documentation

Move table and dataset snippets to samples/ directory (#7683)

Internal / Testing Changes

Blacken unit tests. (#7960)
Cleanup client tests with method to create minimal table resource (#7802)

1.11.2

04-05-2019 08:16 PDT

Dependencies

Add dependency on protobuf. (#7668)

1.11.1

04-04-2019 09:19 PDT

Internal / Testing Changes

Increment version number in setup.py.

1.11.0

04-03-2019 19:33 PDT

Implementation Changes

Remove classifier for Python 3.4 for end-of-life. (#7535)

New Features

Enable fastparquet support by using temporary file in load_table_from_dataframe (#7545)
Allow string for copy sources, query destination, and default dataset (#7560)
Add progress_bar_type argument to to_dataframe to use tqdm to display a progress bar (#7552)
Call get_table in list_rows if the schema is not available (#7621)
Fallback to BQ API when there are problems reading from BQ Storage. (#7633)
Add methods for Models API (#7562)
Add option to use BigQuery Storage API from IPython magics (#7640)

Documentation

Remove typo in Table.from_api_repr docstring. (#7509)
Add docs session to nox configuration for BigQuery (#7541)

Internal / Testing Changes

Refactor table() methods into shared implementation. (#7516)
Blacken noxfile and setup file in nox session (#7619)
Actually use the progress_bar_type argument in QueryJob.to_dataframe(). (#7616)

1.10.0

03-06-2019 15:20 PST

Implementation Changes

Harden ‘ArrayQueryParameter.from_api_repr’ against missing ‘parameterValue’. (#7311)
Allow nested records w/ null values. (#7297)

New Features

Add exists_ok and not_found_ok options to ignore errors when creating/deleting datasets/tables. (#7491)
Accept a string in Table and Dataset constructors. (#7483)

Documentation

Update docstring of RowIterator’s to_dataframe (#7306)
Updated client library documentation URLs. (#7307)

Internal / Testing Changes

Fix lint. (#7383)

1.9.0

02-04-2019 13:28 PST

New Features

Add arguments to select dtypes and use BQ Storage API to QueryJob.to_dataframe(). (#7241)

Documentation

Add sample for fetching total_rows from query results. (#7217)

1.8.1

12-17-2018 17:53 PST

Documentation

Document Python 2 deprecation (#6910)
Normalize docs for ‘page_size’ / ‘max_results’ / ‘page_token’ (#6842)

1.8.0

12-10-2018 12:39 PST

Implementation Changes

Add option to use BQ Storage API with to_dataframe (#6854)
Fix exception type in comment (#6847)
Add to_bqstorage to convert from Table[Reference] google-cloud-bigquery-storage reference (#6840)
Import iam.policy from google.api_core. (#6741)
Add avro logical type control for load jobs. (#6827)
Allow setting partition expiration to ‘None’. (#6823)
Add retry argument to _AsyncJob.result. (#6302)

Dependencies

Update dependency to google-cloud-core (#6835)

Documentation

Add avro load samples (#6832)

Internal / Testing Changes

Blacken libraries (#6794)
Fix copy/paste typos in noxfile comments (#6831)

1.7.0

11-05-2018 16:41 PST

Implementation Changes

Add destination table properties to LoadJobConfig. (#6202)
Allow strings or references in create_dataset and create_table (#6199)
Fix swallowed error message (#6168)

New Features

Add --params option to %%bigquery magic (#6277)
Expose to_api_repr method for jobs. (#6176)
Allow string in addition to DatasetReference / TableReference in Client methods. (#6164)
Add keyword arguments to job config constructors for setting properties (#6397)

Documentation

Update README service links in quickstart guides. (#6322)
Move usage guides to their own docs. (#6238)
Normalize use of support level badges (#6159)

Internal / Testing Changes

Deprecation cleanups (#6304)
Use _get_sub_prop helper so missing load stats don’t raise. (#6269)
Use new Nox (#6175)
Harden snippets against transient GCS errors. (#6184)

1.6.0

New Features

Add support for GEOGRAPHY type (#6147)
Add default QueryJobConfig to Client (#6088)

Documentation

Remove unused “append” samples (#6100)

Internal / Testing Changes

Address dataset leaks, conflicts in systests (#6099)
Harden bucket teardown against 429 Too Many Requests. (#6101)

1.5.1

Implementation Changes

Retry ‘502 Bad Gateway’ errors by default. (#5930)
Avoid pulling entire result set into memory when constructing dataframe. (#5870)
Add support for retrying unstructured 429 / 500 / 502 responses. (#6011)
Populate the jobReference from the API response. (#6044)

Documentation

Prepare documentation for repo split (#5955)
Fix leakage of bigquery/spanner sections into sidebar menu. (#5986)

Internal / Testing Changes

Test pandas support under Python 3.7. (#5857)
Nox: use inplace installs (#5865)
Update system test to use test data in bigquery-public-data. (#5965)

1.5.0

Implementation Changes

Make ‘Table.location’ read-only. (#5687)

New Features

Add ‘clustering_fields’ properties. (#5630)
Add support for job labels (#5654)
Add ‘QueryJob.estimated_bytes_processed’ property (#5655)
Add support/tests for loading tables from ‘gzip.GzipFile’. (#5711)
Add ‘ExternalSourceFormat’ enum. (#5674)
Add default location to client (#5678)

Documentation

Fix typo in CopyJob sources docstring (#5690)

Internal / Testing Changes

Add/refactor snippets for managing BigQuery jobs (#5631)
Reenable systests for ‘dataset.update’/’table.update’. (#5732)

1.4.0

Implementation Changes

Add ‘internalError’ to retryable error reasons. (#5599)
Don’t raise exception if viewing CREATE VIEW DDL results (#5602)

New Features

Add Orc source format support and samples (#5500)
Move ‘DEFAULT_RETRY’ (w/ its predicate) to a new public ‘retry’ module. (#5552)
Allow listing rows on an empty table. (#5584)

Documentation

Add load_table_from_dataframe() to usage docs and changelog and dedents snippets in usage page (#5501)
Add samples for query external data sources (GCS & Sheets) (#5491)
Add BigQuery authorized view samples (#5515)
Update docs to show pyarrow as the only dependency of load_table_from_dataframe() (#5582)

Internal / Testing Changes

Add missing explict coverage for ‘_helpers’ (#5550)
Skip update_table and update_dataset tests until etag issue is resolved. (#5590)

1.3.0

New Features

NUMERIC type support (#5331)
Add timeline and top-level slot-millis to query statistics. (#5312)
Add additional statistics to query plan stages. (#5307)
Add client.load_table_from_dataframe() (#5387)

Documentation

Use autosummary to split up API reference docs (#5340)
Fix typo in Client docstrings (#5342)

Internal / Testing Changes

Prune systests identified as reduntant to snippets. (#5365)
Modify system tests to use prerelease versions of grpcio (#5304)
Improve system test performance (#5319)

1.2.0

Implementation Changes

Switch list_partitions helper to a direct metatable read (#5273)
Fix typo in Encoding.ISO_8859_1 enum value (#5211)

New Features

Add UnknownJob type for redacted jobs. (#5281)
Add project parameter to list_datasets and list_jobs (#5217)
Add from_string factory methods to Dataset and Table (#5255)
Add column based time partitioning (#5267)

Documentation

Standardize docstrings for constants (#5289)
Fix docstring / impl of ExtractJob.destination_uri_file_counts. (#5245)

Internal / Testing Changes

Add testing support for Python 3.7; remove testing support for Python 3.4. (#5295)

1.1.0

New Features

Add client.get_service_account_email (#5203)

Documentation

Update samples and standardize region tags (#5195)

Internal / Testing Changes

Fix trove classifier to be Production/Stable
Don’t suppress ‘dots’ output on test (#5202)

1.0.0

Implementation Changes

Remove deprecated Client methods (#5182)

0.32.0

⚠️ Interface changes

Use job.configuration resource for XXXJobConfig classes (#5036)

Interface additions

Add page_size parameter for list_rows and use in DB-API for arraysize (#4931)
Add IPython magics for running queries (#4983)

Documentation

Add job string constant parameters in init and snippets documentation (#4987)

Internal / Testing changes

Specify IPython version 5.5 when running Python 2.7 tests (#5145)
Move all Dataset property conversion logic into properties (#5130)
Remove unnecessary _Table class from test_job.py (#5126)
Use explicit bytes to initialize ‘BytesIO’. (#5116)
Make SchemaField be able to include description via from_api_repr method (#5114)
Remove _ApiResourceProperty class (#5107)
Add dev version for 0.32.0 release (#5105)
StringIO to BytesIO (#5101)
Shorten snippets test name (#5091)
Don’t use selected_fields for listing query result rows (#5072)
Add location property to job classes. (#5071)
Use autospec for Connection in tests. (#5066)
Add Parquet SourceFormat and samples (#5057)
Remove test_load_table_from_uri_w_autodetect_schema_then_get_job because of duplicate test in snippets (#5004)
Fix encoding variable and strings UTF-8 and ISO-8859-1 difference documentation (#4990)

0.31.0

Interface additions

Add support for EncryptionConfiguration (#4845)

Implementation changes

Allow listing/getting jobs even when there is an “invalid” job. (#4786)

Dependencies

The minimum version for google-api-core has been updated to version 1.0.0. This may cause some incompatibility with older google-cloud libraries, you will need to update those libraries if you have a dependency conflict. (#4944, #4946)

Documentation

Update format in Table.full_table_id and TableListItem.full_table_id docstrings. (#4906)

Testing and internal changes

Install local dependencies when running lint (#4936)
Re-enable lint for tests, remove usage of pylint (#4921)
Normalize all setup.py files (#4909)
Remove unnecessary debug print from tests (#4907)
Use constant strings for job properties in tests (#4833)

0.30.0

This is the release candidate for v1.0.0.

Interface changes / additions

Add delete_contents to delete_dataset. (#4724)

Bugfixes

Add handling of missing properties in SchemaField.from_api_repr(). (#4754)
Fix missing return value in LoadJobConfig.from_api_repr. (#4727)

Documentation

Minor documentation and typo fixes. (#4782, #4718, #4784, #4835, #4836)

0.29.0

Interface changes / additions

Add to_dataframe() method to row iterators. When Pandas is installed this method returns a DataFrame containing the query’s or table’s rows. (#4354)
Iterate over a QueryJob to wait for and get the query results. (#4350)
Add Table.reference and Dataset.reference properties to get the TableReference or DatasetReference corresponding to that Table or Dataset, respectively. (#4405)
Add Row.keys(), Row.items(), and Row.get(). This makes Row act more like a built-in dictionary. (#4393, #4413)

Interface changes / breaking changes

Add Client.insert_rows() and Client.insert_rows_json(), deprecate Client.create_rows() and Client.create_rows_json(). (#4657)
Add Client.list_tables, deprecate Client.list_dataset_tables. (#4653)
Client.list_tables returns an iterators of TableListItem. The API only returns a subset of properties of a table when listing. (#4427)
Remove QueryJob.query_results(). Use QueryJob.result() instead. (#4652)
Remove Client.query_rows(). Use Client.query() instead. (#4429)
Client.list_datasets returns an iterator of DatasetListItem. The API only returns a subset of properties of a dataset when listing. (#4439)

0.28.0

0.28.0 significantly changes the interface for this package. For examples of the differences between 0.28.0 and previous versions, see Migrating to the BigQuery Python client library 0.28. These changes can be summarized as follows:

Query and view operations default to the standard SQL dialect. (#4192)
Client functions related to jobs, like running queries, immediately start the job.
Functions to create, get, update, delete datasets and tables moved to the client class.

Fixes

Populate timeout parameter correctly for queries (#4209)
Automatically retry idempotent RPCs (#4148, #4178)
Parse timestamps in query parameters using canonical format (#3945)
Parse array parameters that contain a struct type. (#4040)
Support Sub Second Datetimes in row data (#3901, #3915, #3926), h/t @page1

Interface changes / additions

Support external table configuration (#4182) in query jobs (#4191) and tables (#4193).
New Row class allows for access by integer index like a tuple, string index like a dictionary, or attribute access like an object. (#4149)
Add option for job ID generation with user-supplied prefix (#4198)
Add support for update of dataset access entries (#4197)
Add support for atomic read-modify-write of a dataset using etag (#4052)
Add support for labels to Dataset (#4026)
Add support for labels to Table (#4207)
Add Table.streaming_buffer property (#4161)
Add TableReference class (#3942)
Add DatasetReference class (#3938, #3942, #3993)
Add ExtractJob.destination_uri_file_counts property. (#3803)
Add client.create_rows_json() to bypass conversions on streaming writes. (#4189)
Add client.get_job() to get arbitrary jobs. (#3804, #4213)
Add filter to client.list_datasets() (#4205)
Add QueryJob.undeclared_query_parameters property. (#3802)
Add QueryJob.referenced_tables property. (#3801)
Add new scalar statistics properties to QueryJob (#3800)
Add QueryJob.query_plan property. (#3799)

Interface changes / breaking changes

Remove client.run_async_query(), use client.query() instead. (#4130)
Remove client.run_sync_query(), use client.query_rows() instead. (#4065, #4248)
Make QueryResults read-only. (#4094, #4144)
Make get_query_results private. Return rows for QueryJob.result() (#3883)
Move \*QueryParameter and UDFResource classes to query module (also exposed in bigquery module). (#4156)

Changes to tables

Remove client from Table class (#4159)
Remove table.exists() (#4145)
Move table.list_parations to client.list_partitions (#4146)
Move table.upload_from_file to client.load_table_from_file (#4136)
Move table.update() and table.patch() to client.update_table() (#4076)
Move table.insert_data() to client.create_rows(). Automatically generates row IDs if not supplied. (#4151, #4173)
Move table.fetch_data() to client.list_rows() (#4119, #4143)
Move table.delete() to client.delete_table() (#4066)
Move table.create() to client.create_table() (#4038, #4043)
Move table.reload() to client.get_table() (#4004)
Rename Table.name attribute to Table.table_id (#3959)
Table constructor takes a TableReference as parameter (#3997)

Changes to datasets

Remove client from Dataset class (#4018)
Remove dataset.exists() (#3996)
Move dataset.list_tables() to client.list_dataset_tables() (#4013)
Move dataset.delete() to client.delete_dataset() (#4012)
Move dataset.patch() and dataset.update() to client.update_dataset() (#4003)
Move dataset.create() to client.create_dataset() (#3982)
Move dataset.reload() to client.get_dataset() (#3973)
Rename Dataset.name attribute to Dataset.dataset_id (#3955)
client.dataset() returns a DatasetReference instead of Dataset. (#3944)
Rename class: dataset.AccessGrant -> dataset.AccessEntry. (#3798)
dataset.table() returns a TableReference instead of a Table (#4014)
Dataset constructor takes a DatasetReference (#4036)

Changes to jobs

Make job.begin() method private. (#4242)
Add LoadJobConfig class and modify LoadJob (#4103, #4137)
Add CopyJobConfig class and modify CopyJob (#4051, #4059)
Type of Job’s and Query’s default_dataset changed from Dataset to DatasetReference (#4037)
Rename client.load_table_from_storage() to client.load_table_from_uri() (#4235)
Rename client.extract_table_to_storage to client.extract_table(). Method starts the extract job immediately. (#3991, #4177)
Rename XJob.name to XJob.job_id. (#3962)
Rename job classes. LoadTableFromStorageJob -> LoadJob and ExtractTableToStorageJob -> jobs.ExtractJob (#3797)

Dependencies

Updating to google-cloud-core ~= 0.28, in particular, the google-api-core package has been moved out of google-cloud-core. (#4221)

PyPI: https://pypi.org/project/google-cloud-bigquery/0.28.0/

0.27.0

Remove client-side enum validation. (#3735)
Add Table.row_from_mapping helper. (#3425)
Move google.cloud.future to google.api.core (#3764)
Fix __eq__ and __ne__. (#3765)
Move google.cloud.iterator to google.api.core.page_iterator (#3770)
nullMarker support for BigQuery Load Jobs (#3777), h/t @leondealmeida
Allow job_id to be explicitly specified in DB-API. (#3779)
Add support for a custom null marker. (#3776)
Add SchemaField serialization and deserialization. (#3786)
Add get_query_results method to the client. (#3838)
Poll for query completion via getQueryResults method. (#3844)
Allow fetching more than the first page when max_results is set. (#3845)

PyPI: https://pypi.org/project/google-cloud-bigquery/0.27.0/

0.26.0

Notable implementation changes

Using the requests transport attached to a Client for for resumable media (i.e. downloads and uploads) (#3705) (this relates to the httplib2 to requests switch)

Interface changes / additions

Adding autodetect property on LoadTableFromStorageJob to enable schema autodetection. (#3648)
Implementing the Python Futures interface for Jobs. Call job.result() to wait for jobs to complete instead of polling manually on the job status. (#3626)
Adding is_nullable property on SchemaField. Can be used to check if a column is nullable. (#3620)
job_name argument added to Table.upload_from_file for setting the job ID. (#3605)
Adding google.cloud.bigquery.dbapi package, which implements PEP-249 DB-API specification. (#2921)
Adding Table.view_use_legacy_sql property. Can be used to create views with legacy or standard SQL. (#3514)

Interface changes / breaking changes

Removing results() method from the QueryJob class. Use query_results() instead. (#3661)
SchemaField is now immutable. It is also hashable so that it can be used in sets. (#3601)

Dependencies

Updating to google-cloud-core ~= 0.26, in particular, the underlying HTTP transport switched from httplib2 to requests (#3654, #3674)
Adding dependency on google-resumable-media for loading BigQuery tables from local files. (#3555)

Packaging

Fix inclusion of tests (vs. unit_tests) in MANIFEST.in (#3552)
Updating author_email in setup.py to googleapis-publisher@google.com. (#3598)

PyPI: https://pypi.org/project/google-cloud-bigquery/0.26.0/