Source code for airflow.providers.google.cloud.operators.bigquery

#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.
"""This module contains Google BigQuery operators."""

from __future__ import annotations

import enum
import json
import re
import warnings
from functools import cached_property
from typing import TYPE_CHECKING, Any, Iterable, Sequence, SupportsAbs

import attr
from google.api_core.exceptions import Conflict
from google.cloud.bigquery import DEFAULT_RETRY, CopyJob, ExtractJob, LoadJob, QueryJob, Row
from google.cloud.bigquery.table import RowIterator

from airflow.configuration import conf
from airflow.exceptions import AirflowException, AirflowProviderDeprecationWarning, AirflowSkipException
from airflow.models import BaseOperator, BaseOperatorLink
from airflow.models.xcom import XCom
from airflow.providers.common.sql.operators.sql import (
    SQLCheckOperator,
    SQLColumnCheckOperator,
    SQLIntervalCheckOperator,
    SQLTableCheckOperator,
    SQLValueCheckOperator,
    _parse_boolean,
)
from airflow.providers.google.cloud.hooks.bigquery import BigQueryHook, BigQueryJob
from airflow.providers.google.cloud.hooks.gcs import GCSHook, _parse_gcs_url
from airflow.providers.google.cloud.links.bigquery import BigQueryDatasetLink, BigQueryTableLink
from airflow.providers.google.cloud.openlineage.mixins import _BigQueryOpenLineageMixin
from airflow.providers.google.cloud.operators.cloud_base import GoogleCloudBaseOperator
from airflow.providers.google.cloud.triggers.bigquery import (
    BigQueryCheckTrigger,
    BigQueryGetDataTrigger,
    BigQueryInsertJobTrigger,
    BigQueryIntervalCheckTrigger,
    BigQueryValueCheckTrigger,
)
from airflow.providers.google.cloud.utils.bigquery import convert_job_id
from airflow.providers.google.common.deprecated import deprecated
from airflow.providers.google.common.hooks.base_google import PROVIDE_PROJECT_ID
from airflow.utils.helpers import exactly_one

if TYPE_CHECKING:
    from google.api_core.retry import Retry
    from google.cloud.bigquery import UnknownJob

    from airflow.models.taskinstancekey import TaskInstanceKey
    from airflow.utils.context import Context




[docs]LABEL_REGEX = re.compile(r"^[\w-]{0,63}$")
[docs]class BigQueryUIColors(enum.Enum): """Hex colors for BigQuery operators."""
[docs] CHECK = "#C0D7FF"
[docs] QUERY = "#A1BBFF"
[docs] TABLE = "#81A0FF"
[docs] DATASET = "#5F86FF"
[docs]class IfExistAction(enum.Enum): """Action to take if the resource exist."""
[docs] IGNORE = "ignore"
[docs] LOG = "log"
[docs] FAIL = "fail"
[docs] SKIP = "skip"
@attr.s(auto_attribs=True) class _BigQueryDbHookMixin: def get_db_hook(self: BigQueryCheckOperator) -> BigQueryHook: # type:ignore[misc] """Get BigQuery DB Hook.""" return BigQueryHook( gcp_conn_id=self.gcp_conn_id, use_legacy_sql=self.use_legacy_sql, location=self.location, impersonation_chain=self.impersonation_chain, labels=self.labels, ) class _BigQueryOperatorsEncryptionConfigurationMixin: """A class to handle the configuration for BigQueryHook.insert_job method.""" # Note: If you want to add this feature to a new operator you can include the class name in the type # annotation of the `self`. Then you can inherit this class in the target operator. # e.g: BigQueryCheckOperator, BigQueryTableCheckOperator def include_encryption_configuration( # type:ignore[misc] self: BigQueryCheckOperator | BigQueryTableCheckOperator | BigQueryValueCheckOperator | BigQueryColumnCheckOperator | BigQueryGetDataOperator | BigQueryIntervalCheckOperator, configuration: dict, config_key: str, ) -> None: """Add encryption_configuration to destinationEncryptionConfiguration key if it is not None.""" if self.encryption_configuration is not None: configuration[config_key]["destinationEncryptionConfiguration"] = self.encryption_configuration
[docs]class BigQueryCheckOperator( _BigQueryDbHookMixin, SQLCheckOperator, _BigQueryOperatorsEncryptionConfigurationMixin ): """ Performs checks against BigQuery. This operator expects a SQL query that returns a single row. Each value on that row is evaluated using a Python ``bool`` cast. If any of the values is falsy, the check errors out. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryCheckOperator` Note that Python bool casting evals the following as *False*: * ``False`` * ``0`` * Empty string (``""``) * Empty list (``[]``) * Empty dictionary or set (``{}``) Given a query like ``SELECT COUNT(*) FROM foo``, it will fail only if the count equals to zero. You can craft much more complex query that could, for instance, check that the table has the same number of rows as the source table upstream, or that the count of today's partition is greater than yesterday's partition, or that a set of metrics are less than three standard deviation for the 7-day average. This operator can be used as a data quality check in your pipeline. Depending on where you put it in your DAG, you have the choice to stop the critical path, preventing from publishing dubious data, or on the side and receive email alerts without stopping the progress of the DAG. :param sql: SQL to execute. :param gcp_conn_id: Connection ID for Google Cloud. :param use_legacy_sql: Whether to use legacy SQL (true) or standard SQL (false). :param location: The geographic location of the job. See details at: https://cloud.google.com/bigquery/docs/locations#specifying_your_location :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account. (templated) :param labels: a dictionary containing labels for the table, passed to BigQuery. :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } :param deferrable: Run operator in the deferrable mode. :param poll_interval: (Deferrable mode only) polling period in seconds to check for the status of job. :param query_params: a list of dictionary containing query parameter types and values, passed to BigQuery. The structure of dictionary should look like 'queryParameters' in Google BigQuery Jobs API: https://cloud.google.com/bigquery/docs/reference/rest/v2/jobs. For example, [{ 'name': 'corpus', 'parameterType': { 'type': 'STRING' }, 'parameterValue': { 'value': 'romeoandjuliet' } }]. (templated) """
[docs] template_fields: Sequence[str] = ( "sql", "gcp_conn_id", "impersonation_chain", "labels", "query_params", )
[docs] template_ext: Sequence[str] = (".sql",)
[docs] ui_color = BigQueryUIColors.CHECK.value
[docs] conn_id_field = "gcp_conn_id"
def __init__( self, *, sql: str, gcp_conn_id: str = "google_cloud_default", use_legacy_sql: bool = True, location: str | None = None, impersonation_chain: str | Sequence[str] | None = None, labels: dict | None = None, encryption_configuration: dict | None = None, deferrable: bool = conf.getboolean("operators", "default_deferrable", fallback=False), poll_interval: float = 4.0, query_params: list | None = None, **kwargs, ) -> None: super().__init__(sql=sql, **kwargs) self.gcp_conn_id = gcp_conn_id self.use_legacy_sql = use_legacy_sql self.location = location self.impersonation_chain = impersonation_chain self.labels = labels self.encryption_configuration = encryption_configuration self.deferrable = deferrable self.poll_interval = poll_interval self.query_params = query_params def _submit_job( self, hook: BigQueryHook, job_id: str, ) -> BigQueryJob: """Submit a new job and get the job id for polling the status using Trigger.""" configuration = {"query": {"query": self.sql, "useLegacySql": self.use_legacy_sql}} if self.query_params: configuration["query"]["queryParameters"] = self.query_params self.include_encryption_configuration(configuration, "query") return hook.insert_job( configuration=configuration, project_id=hook.project_id, location=self.location, job_id=job_id, nowait=True, )
[docs] def execute(self, context: Context): if not self.deferrable: super().execute(context=context) else: hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ) job = self._submit_job(hook, job_id="") context["ti"].xcom_push(key="job_id", value=job.job_id) if job.running(): self.defer( timeout=self.execution_timeout, trigger=BigQueryCheckTrigger( conn_id=self.gcp_conn_id, job_id=job.job_id, project_id=hook.project_id, location=self.location or hook.location, poll_interval=self.poll_interval, impersonation_chain=self.impersonation_chain, ), method_name="execute_complete", ) self._handle_job_error(job) # job.result() returns a RowIterator. Mypy expects an instance of SupportsNext[Any] for # the next() call which the RowIterator does not resemble to. Hence, ignore the arg-type error. # Row passed to _validate_records is a collection of values only, without column names. self._validate_records(next(iter(job.result()), [])) # type: ignore[arg-type] self.log.info("Current state of job %s is %s", job.job_id, job.state)
@staticmethod def _handle_job_error(job: BigQueryJob | UnknownJob) -> None: if job.error_result: raise AirflowException(f"BigQuery job {job.job_id} failed: {job.error_result}") def _validate_records(self, records) -> None: if not records: raise AirflowException(f"The following query returned zero rows: {self.sql}") elif not all(records): self._raise_exception( # type: ignore[attr-defined] f"Test failed.\nQuery:\n{self.sql}\nResults:\n{records!s}" )
[docs] def execute_complete(self, context: Context, event: dict[str, Any]) -> None: """ Act as a callback for when the trigger fires. This returns immediately. It relies on trigger to throw an exception, otherwise it assumes execution was successful. """ if event["status"] == "error": raise AirflowException(event["message"]) self._validate_records(event["records"]) self.log.info("Record: %s", event["records"]) self.log.info("Success.")
[docs]class BigQueryValueCheckOperator( _BigQueryDbHookMixin, SQLValueCheckOperator, _BigQueryOperatorsEncryptionConfigurationMixin ): """ Perform a simple value check using sql code. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryValueCheckOperator` :param sql: SQL to execute. :param use_legacy_sql: Whether to use legacy SQL (true) or standard SQL (false). :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param location: The geographic location of the job. See details at: https://cloud.google.com/bigquery/docs/locations#specifying_your_location :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account. (templated) :param labels: a dictionary containing labels for the table, passed to BigQuery. :param deferrable: Run operator in the deferrable mode. :param poll_interval: (Deferrable mode only) polling period in seconds to check for the status of job. """
[docs] template_fields: Sequence[str] = ( "sql", "gcp_conn_id", "pass_value", "impersonation_chain", "labels", )
[docs] template_ext: Sequence[str] = (".sql",)
[docs] ui_color = BigQueryUIColors.CHECK.value
[docs] conn_id_field = "gcp_conn_id"
def __init__( self, *, sql: str, pass_value: Any, tolerance: Any = None, encryption_configuration: dict | None = None, gcp_conn_id: str = "google_cloud_default", use_legacy_sql: bool = True, location: str | None = None, impersonation_chain: str | Sequence[str] | None = None, labels: dict | None = None, deferrable: bool = conf.getboolean("operators", "default_deferrable", fallback=False), poll_interval: float = 4.0, **kwargs, ) -> None: super().__init__(sql=sql, pass_value=pass_value, tolerance=tolerance, **kwargs) self.location = location self.gcp_conn_id = gcp_conn_id self.use_legacy_sql = use_legacy_sql self.encryption_configuration = encryption_configuration self.impersonation_chain = impersonation_chain self.labels = labels self.deferrable = deferrable self.poll_interval = poll_interval def _submit_job( self, hook: BigQueryHook, job_id: str, ) -> BigQueryJob: """Submit a new job and get the job id for polling the status using Triggerer.""" configuration = { "query": { "query": self.sql, "useLegacySql": self.use_legacy_sql, }, } self.include_encryption_configuration(configuration, "query") return hook.insert_job( configuration=configuration, project_id=hook.project_id, location=self.location, job_id=job_id, nowait=True, )
[docs] def execute(self, context: Context) -> None: # type: ignore[override] if not self.deferrable: super().execute(context=context) else: hook = BigQueryHook(gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain) job = self._submit_job(hook, job_id="") context["ti"].xcom_push(key="job_id", value=job.job_id) if job.running(): self.defer( timeout=self.execution_timeout, trigger=BigQueryValueCheckTrigger( conn_id=self.gcp_conn_id, job_id=job.job_id, project_id=hook.project_id, location=self.location or hook.location, sql=self.sql, pass_value=self.pass_value, tolerance=self.tol, poll_interval=self.poll_interval, impersonation_chain=self.impersonation_chain, ), method_name="execute_complete", ) self._handle_job_error(job) # job.result() returns a RowIterator. Mypy expects an instance of SupportsNext[Any] for # the next() call which the RowIterator does not resemble to. Hence, ignore the arg-type error. # Row passed to check_value is a collection of values only, without column names. self.check_value(next(iter(job.result()), [])) # type: ignore[arg-type] self.log.info("Current state of job %s is %s", job.job_id, job.state)
@staticmethod def _handle_job_error(job: BigQueryJob | UnknownJob) -> None: if job.error_result: raise AirflowException(f"BigQuery job {job.job_id} failed: {job.error_result}")
[docs] def execute_complete(self, context: Context, event: dict[str, Any]) -> None: """ Act as a callback for when the trigger fires. This returns immediately. It relies on trigger to throw an exception, otherwise it assumes execution was successful. """ if event["status"] == "error": raise AirflowException(event["message"]) self.log.info( "%s completed with response %s ", self.task_id, event["message"], )
[docs]class BigQueryIntervalCheckOperator( _BigQueryDbHookMixin, SQLIntervalCheckOperator, _BigQueryOperatorsEncryptionConfigurationMixin ): """ Check that the values of metrics given as SQL expressions are within a tolerance of the older ones. This method constructs a query like so :: SELECT {metrics_threshold_dict_key} FROM {table} WHERE {date_filter_column}=<date> .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryIntervalCheckOperator` :param table: the table name :param days_back: number of days between ds and the ds we want to check against. Defaults to 7 days :param metrics_thresholds: a dictionary of ratios indexed by metrics, for example 'COUNT(*)': 1.5 would require a 50 percent or less difference between the current day, and the prior days_back. :param use_legacy_sql: Whether to use legacy SQL (true) or standard SQL (false). :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param location: The geographic location of the job. See details at: https://cloud.google.com/bigquery/docs/locations#specifying_your_location :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). :param labels: a dictionary containing labels for the table, passed to BigQuery :param deferrable: Run operator in the deferrable mode :param poll_interval: (Deferrable mode only) polling period in seconds to check for the status of job. Defaults to 4 seconds. :param project_id: a string represents the BigQuery projectId """
[docs] template_fields: Sequence[str] = ( "table", "gcp_conn_id", "sql1", "sql2", "impersonation_chain", "labels", )
[docs] ui_color = BigQueryUIColors.CHECK.value
[docs] conn_id_field = "gcp_conn_id"
def __init__( self, *, table: str, metrics_thresholds: dict, date_filter_column: str = "ds", days_back: SupportsAbs[int] = -7, gcp_conn_id: str = "google_cloud_default", use_legacy_sql: bool = True, location: str | None = None, encryption_configuration: dict | None = None, impersonation_chain: str | Sequence[str] | None = None, labels: dict | None = None, deferrable: bool = conf.getboolean("operators", "default_deferrable", fallback=False), poll_interval: float = 4.0, project_id: str = PROVIDE_PROJECT_ID, **kwargs, ) -> None: super().__init__( table=table, metrics_thresholds=metrics_thresholds, date_filter_column=date_filter_column, days_back=days_back, **kwargs, ) self.gcp_conn_id = gcp_conn_id self.use_legacy_sql = use_legacy_sql self.location = location self.encryption_configuration = encryption_configuration self.impersonation_chain = impersonation_chain self.labels = labels self.project_id = project_id self.deferrable = deferrable self.poll_interval = poll_interval def _submit_job( self, hook: BigQueryHook, sql: str, job_id: str, ) -> BigQueryJob: """Submit a new job and get the job id for polling the status using Triggerer.""" configuration = {"query": {"query": sql, "useLegacySql": self.use_legacy_sql}} self.include_encryption_configuration(configuration, "query") return hook.insert_job( configuration=configuration, project_id=self.project_id or hook.project_id, location=self.location, job_id=job_id, nowait=True, )
[docs] def execute(self, context: Context): if not self.deferrable: super().execute(context) else: hook = BigQueryHook(gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain) self.log.info("Using ratio formula: %s", self.ratio_formula) self.log.info("Executing SQL check: %s", self.sql1) job_1 = self._submit_job(hook, sql=self.sql1, job_id="") context["ti"].xcom_push(key="job_id", value=job_1.job_id) self.log.info("Executing SQL check: %s", self.sql2) job_2 = self._submit_job(hook, sql=self.sql2, job_id="") self.defer( timeout=self.execution_timeout, trigger=BigQueryIntervalCheckTrigger( conn_id=self.gcp_conn_id, first_job_id=job_1.job_id, second_job_id=job_2.job_id, project_id=hook.project_id, table=self.table, location=self.location or hook.location, metrics_thresholds=self.metrics_thresholds, date_filter_column=self.date_filter_column, days_back=self.days_back, ratio_formula=self.ratio_formula, ignore_zero=self.ignore_zero, poll_interval=self.poll_interval, impersonation_chain=self.impersonation_chain, ), method_name="execute_complete", )
[docs] def execute_complete(self, context: Context, event: dict[str, Any]) -> None: """ Act as a callback for when the trigger fires. This returns immediately. It relies on trigger to throw an exception, otherwise it assumes execution was successful. """ if event["status"] == "error": raise AirflowException(event["message"]) self.log.info( "%s completed with response %s ", self.task_id, event["message"], )
[docs]class BigQueryColumnCheckOperator( _BigQueryDbHookMixin, SQLColumnCheckOperator, _BigQueryOperatorsEncryptionConfigurationMixin ): """ Subclasses the SQLColumnCheckOperator in order to provide a job id for OpenLineage to parse. See base class docstring for usage. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryColumnCheckOperator` :param table: the table name :param column_mapping: a dictionary relating columns to their checks :param partition_clause: a string SQL statement added to a WHERE clause to partition data :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } :param use_legacy_sql: Whether to use legacy SQL (true) or standard SQL (false). :param location: The geographic location of the job. See details at: https://cloud.google.com/bigquery/docs/locations#specifying_your_location :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). :param labels: a dictionary containing labels for the table, passed to BigQuery """
[docs] template_fields: Sequence[str] = tuple(set(SQLColumnCheckOperator.template_fields) | {"gcp_conn_id"})
[docs] conn_id_field = "gcp_conn_id"
def __init__( self, *, table: str, column_mapping: dict, partition_clause: str | None = None, database: str | None = None, accept_none: bool = True, encryption_configuration: dict | None = None, gcp_conn_id: str = "google_cloud_default", use_legacy_sql: bool = True, location: str | None = None, impersonation_chain: str | Sequence[str] | None = None, labels: dict | None = None, **kwargs, ) -> None: super().__init__( table=table, column_mapping=column_mapping, partition_clause=partition_clause, database=database, accept_none=accept_none, **kwargs, ) self.table = table self.column_mapping = column_mapping self.partition_clause = partition_clause self.database = database self.accept_none = accept_none self.gcp_conn_id = gcp_conn_id self.encryption_configuration = encryption_configuration self.use_legacy_sql = use_legacy_sql self.location = location self.impersonation_chain = impersonation_chain self.labels = labels def _submit_job( self, hook: BigQueryHook, job_id: str, ) -> BigQueryJob: """Submit a new job and get the job id for polling the status using Trigger.""" configuration = {"query": {"query": self.sql, "useLegacySql": self.use_legacy_sql}} self.include_encryption_configuration(configuration, "query") return hook.insert_job( configuration=configuration, project_id=hook.project_id, location=self.location, job_id=job_id, nowait=False, )
[docs] def execute(self, context=None): """Perform checks on the given columns.""" hook = self.get_db_hook() failed_tests = [] job = self._submit_job(hook, job_id="") context["ti"].xcom_push(key="job_id", value=job.job_id) records = job.result().to_dataframe() if records.empty: raise AirflowException(f"The following query returned zero rows: {self.sql}") records.columns = records.columns.str.lower() self.log.info("Record: %s", records) for row in records.iterrows(): column = row[1].get("col_name") check = row[1].get("check_type") result = row[1].get("check_result") tolerance = self.column_mapping[column][check].get("tolerance") self.column_mapping[column][check]["result"] = result self.column_mapping[column][check]["success"] = self._get_match( self.column_mapping[column][check], result, tolerance ) failed_tests.extend( f"Column: {col}\n\tCheck: {check},\n\tCheck Values: {check_values}\n" for col, checks in self.column_mapping.items() for check, check_values in checks.items() if not check_values["success"] ) if failed_tests: exception_string = ( f"Test failed.\nResults:\n{records!s}\n" f"The following tests have failed:" f"\n{''.join(failed_tests)}" ) self._raise_exception(exception_string) self.log.info("All tests have passed")
[docs]class BigQueryTableCheckOperator( _BigQueryDbHookMixin, SQLTableCheckOperator, _BigQueryOperatorsEncryptionConfigurationMixin ): """ Subclasses the SQLTableCheckOperator in order to provide a job id for OpenLineage to parse. See base class for usage. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryTableCheckOperator` :param table: the table name :param checks: a dictionary of check names and boolean SQL statements :param partition_clause: a string SQL statement added to a WHERE clause to partition data :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param use_legacy_sql: Whether to use legacy SQL (true) or standard SQL (false). :param location: The geographic location of the job. See details at: https://cloud.google.com/bigquery/docs/locations#specifying_your_location :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). :param labels: a dictionary containing labels for the table, passed to BigQuery :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } """
[docs] template_fields: Sequence[str] = tuple(set(SQLTableCheckOperator.template_fields) | {"gcp_conn_id"})
[docs] conn_id_field = "gcp_conn_id"
def __init__( self, *, table: str, checks: dict, partition_clause: str | None = None, gcp_conn_id: str = "google_cloud_default", use_legacy_sql: bool = True, location: str | None = None, impersonation_chain: str | Sequence[str] | None = None, labels: dict | None = None, encryption_configuration: dict | None = None, **kwargs, ) -> None: super().__init__(table=table, checks=checks, partition_clause=partition_clause, **kwargs) self.gcp_conn_id = gcp_conn_id self.use_legacy_sql = use_legacy_sql self.location = location self.impersonation_chain = impersonation_chain self.labels = labels self.encryption_configuration = encryption_configuration def _submit_job( self, hook: BigQueryHook, job_id: str, ) -> BigQueryJob: """Submit a new job and get the job id for polling the status using Trigger.""" configuration = {"query": {"query": self.sql, "useLegacySql": self.use_legacy_sql}} self.include_encryption_configuration(configuration, "query") return hook.insert_job( configuration=configuration, project_id=hook.project_id, location=self.location, job_id=job_id, nowait=False, )
[docs] def execute(self, context=None): """Execute the given checks on the table.""" hook = self.get_db_hook() job = self._submit_job(hook, job_id="") context["ti"].xcom_push(key="job_id", value=job.job_id) records = job.result().to_dataframe() if records.empty: raise AirflowException(f"The following query returned zero rows: {self.sql}") records.columns = records.columns.str.lower() self.log.info("Record:\n%s", records) for row in records.iterrows(): check = row[1].get("check_name") result = row[1].get("check_result") self.checks[check]["success"] = _parse_boolean(str(result)) failed_tests = [ f"\tCheck: {check},\n\tCheck Values: {check_values}\n" for check, check_values in self.checks.items() if not check_values["success"] ] if failed_tests: exception_string = ( f"Test failed.\nQuery:\n{self.sql}\nResults:\n{records!s}\n" f"The following tests have failed:\n{', '.join(failed_tests)}" ) self._raise_exception(exception_string) self.log.info("All tests have passed")
[docs]class BigQueryGetDataOperator(GoogleCloudBaseOperator, _BigQueryOperatorsEncryptionConfigurationMixin): """ Fetch data and return it, either from a BigQuery table, or results of a query job. Data could be narrowed down by specific columns or retrieved as a whole. It is returned in either of the following two formats, based on "as_dict" value: 1. False (Default) - A Python list of lists, with the number of nested lists equal to the number of rows fetched. Each nested list represents a row, where the elements within it correspond to the column values for that particular row. **Example Result**: ``[['Tony', 10], ['Mike', 20]`` 2. True - A Python list of dictionaries, where each dictionary represents a row. In each dictionary, the keys are the column names and the values are the corresponding values for those columns. **Example Result**: ``[{'name': 'Tony', 'age': 10}, {'name': 'Mike', 'age': 20}]`` .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryGetDataOperator` .. note:: If you pass fields to ``selected_fields`` which are in different order than the order of columns already in BQ table/job, the data will still be in the order of BQ table. For example if the BQ table has 3 columns as ``[A,B,C]`` and you pass 'B,A' in the ``selected_fields`` the data would still be of the form ``'A,B'``. .. note:: When utilizing job id not in deferrable mode, the job should be in DONE state. **Example - Retrieve data from BigQuery using table**:: get_data = BigQueryGetDataOperator( task_id="get_data_from_bq", dataset_id="test_dataset", table_id="Transaction_partitions", table_project_id="internal-gcp-project", max_results=100, selected_fields="DATE", gcp_conn_id="airflow-conn-id", ) **Example - Retrieve data from BigQuery using a job id**:: get_data = BigQueryGetDataOperator( job_id="airflow_8999918812727394_86a1cecc69c5e3028d28247affd7563", job_project_id="internal-gcp-project", max_results=100, selected_fields="DATE", gcp_conn_id="airflow-conn-id", ) :param dataset_id: The dataset ID of the requested table. (templated) :param table_id: The table ID of the requested table. Mutually exclusive with job_id. (templated) :param table_project_id: (Optional) The project ID of the requested table. If None, it will be derived from the hook's project ID. (templated) :param job_id: The job ID from which query results are retrieved. Mutually exclusive with table_id. (templated) :param job_project_id: (Optional) Google Cloud Project where the job is running. If None, it will be derived from the hook's project ID. (templated) :param project_id: (Deprecated) (Optional) The name of the project where the data will be returned from. If None, it will be derived from the hook's project ID. (templated) :param max_results: The maximum number of records (rows) to be fetched from the table. (templated) :param selected_fields: List of fields to return (comma-separated). If unspecified, all fields are returned. :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param location: The location used for the operation. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). :param deferrable: Run operator in the deferrable mode :param poll_interval: (Deferrable mode only) polling period in seconds to check for the status of job. Defaults to 4 seconds. :param as_dict: if True returns the result as a list of dictionaries, otherwise as list of lists (default: False). :param use_legacy_sql: Whether to use legacy SQL (true) or standard SQL (false). """
[docs] template_fields: Sequence[str] = ( "dataset_id", "table_id", "table_project_id", "job_id", "job_project_id", "project_id", "max_results", "selected_fields", "impersonation_chain", )
[docs] ui_color = BigQueryUIColors.QUERY.value
def __init__( self, *, dataset_id: str | None = None, table_id: str | None = None, table_project_id: str | None = None, job_id: str | None = None, job_project_id: str | None = None, project_id: str = PROVIDE_PROJECT_ID, max_results: int = 100, selected_fields: str | None = None, gcp_conn_id: str = "google_cloud_default", location: str | None = None, encryption_configuration: dict | None = None, impersonation_chain: str | Sequence[str] | None = None, deferrable: bool = conf.getboolean("operators", "default_deferrable", fallback=False), poll_interval: float = 4.0, as_dict: bool = False, use_legacy_sql: bool = True, **kwargs, ) -> None: super().__init__(**kwargs) self.table_project_id = table_project_id self.dataset_id = dataset_id self.table_id = table_id self.job_project_id = job_project_id self.job_id = job_id self.max_results = max_results self.selected_fields = selected_fields self.gcp_conn_id = gcp_conn_id self.location = location self.impersonation_chain = impersonation_chain self.encryption_configuration = encryption_configuration self.project_id = project_id self.deferrable = deferrable self.poll_interval = poll_interval self.as_dict = as_dict self.use_legacy_sql = use_legacy_sql def _submit_job( self, hook: BigQueryHook, job_id: str, ) -> BigQueryJob: get_query = self.generate_query(hook=hook) configuration = {"query": {"query": get_query, "useLegacySql": self.use_legacy_sql}} self.include_encryption_configuration(configuration, "query") """Submit a new job and get the job id for polling the status using Triggerer.""" return hook.insert_job( configuration=configuration, location=self.location, project_id=self.job_project_id or hook.project_id, job_id=job_id, nowait=True, )
[docs] def generate_query(self, hook: BigQueryHook) -> str: """Generate a SELECT query if for the given dataset and table ID.""" query = "select " if self.selected_fields: query += self.selected_fields else: query += "*" query += ( f" from `{self.table_project_id or hook.project_id}.{self.dataset_id}" f".{self.table_id}` limit {self.max_results}" ) return query
[docs] def execute(self, context: Context): if self.project_id: self.log.warning( "The project_id parameter is deprecated, and will be removed in a future release." " Please use table_project_id instead.", ) if not self.table_project_id: self.table_project_id = self.project_id else: self.log.info("Ignoring 'project_id' parameter, as 'table_project_id' is found.") if not exactly_one(self.job_id, self.table_id): raise AirflowException( "'job_id' and 'table_id' parameters are mutually exclusive, " "ensure that exactly one of them is specified" ) hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, use_legacy_sql=self.use_legacy_sql, ) if not self.deferrable: if not self.job_id: self.log.info( "Fetching Data from %s.%s.%s max results: %s", self.table_project_id or hook.project_id, self.dataset_id, self.table_id, self.max_results, ) if not self.selected_fields: schema: dict[str, list] = hook.get_schema( dataset_id=self.dataset_id, table_id=self.table_id, project_id=self.table_project_id or hook.project_id, ) if "fields" in schema: self.selected_fields = ",".join([field["name"] for field in schema["fields"]]) rows: list[Row] | RowIterator | list[dict[str, Any]] = hook.list_rows( dataset_id=self.dataset_id, table_id=self.table_id, max_results=self.max_results, selected_fields=self.selected_fields, location=self.location, project_id=self.table_project_id or hook.project_id, ) else: self.log.info( "Fetching data from job '%s:%s.%s' max results: %s", self.job_project_id or hook.project_id, self.location, self.job_id, self.max_results, ) rows = hook.get_query_results( job_id=self.job_id, location=self.location, selected_fields=self.selected_fields, max_results=self.max_results, project_id=self.job_project_id or hook.project_id, ) if isinstance(rows, RowIterator): raise TypeError( "BigQueryHook.list_rows() returns iterator when return_iterator is False (default)" ) self.log.info("Total extracted rows: %s", len(rows)) if self.as_dict: table_data = [dict(row) for row in rows] else: table_data = [row.values() if isinstance(row, Row) else list(row.values()) for row in rows] return table_data if not self.job_id: job: BigQueryJob | UnknownJob = self._submit_job(hook, job_id="") else: job = hook.get_job( job_id=self.job_id, project_id=self.job_project_id or hook.project_id, location=self.location ) context["ti"].xcom_push(key="job_id", value=job.job_id) self.defer( timeout=self.execution_timeout, trigger=BigQueryGetDataTrigger( conn_id=self.gcp_conn_id, job_id=job.job_id, dataset_id=self.dataset_id, table_id=self.table_id, project_id=self.job_project_id or hook.project_id, location=self.location or hook.location, poll_interval=self.poll_interval, as_dict=self.as_dict, impersonation_chain=self.impersonation_chain, selected_fields=self.selected_fields, ), method_name="execute_complete", )
[docs] def execute_complete(self, context: Context, event: dict[str, Any]) -> Any: """ Act as a callback for when the trigger fires. This returns immediately. It relies on trigger to throw an exception, otherwise it assumes execution was successful. """ if event["status"] == "error": raise AirflowException(event["message"]) self.log.info("Total extracted rows: %s", len(event["records"])) return event["records"]
@deprecated( planned_removal_date="November 01, 2024", use_instead="BigQueryInsertJobOperator", category=AirflowProviderDeprecationWarning, )
[docs]class BigQueryExecuteQueryOperator(GoogleCloudBaseOperator): """ Executes BigQuery SQL queries in a specific BigQuery database. This operator is deprecated. Please use :class:`airflow.providers.google.cloud.operators.bigquery.BigQueryInsertJobOperator` instead. This operator does not assert idempotency. :param sql: the SQL code to be executed as a single string, or a list of str (sql statements), or a reference to a template file. Template references are recognized by str ending in '.sql' :param destination_dataset_table: A dotted ``(<project>.|<project>:)<dataset>.<table>`` that, if set, will store the results of the query. (templated) :param write_disposition: Specifies the action that occurs if the destination table already exists. (default: 'WRITE_EMPTY') :param create_disposition: Specifies whether the job is allowed to create new tables. (default: 'CREATE_IF_NEEDED') :param allow_large_results: Whether to allow large results. :param flatten_results: If true and query uses legacy SQL dialect, flattens all nested and repeated fields in the query results. ``allow_large_results`` must be ``true`` if this is set to ``false``. For standard SQL queries, this flag is ignored and results are never flattened. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param udf_config: The User Defined Function configuration for the query. See https://cloud.google.com/bigquery/user-defined-functions for details. :param use_legacy_sql: Whether to use legacy SQL (true) or standard SQL (false). :param maximum_billing_tier: Positive integer that serves as a multiplier of the basic price. Defaults to None, in which case it uses the value set in the project. :param maximum_bytes_billed: Limits the bytes billed for this job. Queries that will have bytes billed beyond this limit will fail (without incurring a charge). If unspecified, this will be set to your project default. :param api_resource_configs: a dictionary that contain params 'configuration' applied for Google BigQuery Jobs API: https://cloud.google.com/bigquery/docs/reference/rest/v2/jobs for example, {'query': {'useQueryCache': False}}. You could use it if you need to provide some params that are not supported by BigQueryOperator like args. :param schema_update_options: Allows the schema of the destination table to be updated as a side effect of the load job. :param query_params: a list of dictionary containing query parameter types and values, passed to BigQuery. The structure of dictionary should look like 'queryParameters' in Google BigQuery Jobs API: https://cloud.google.com/bigquery/docs/reference/rest/v2/jobs. For example, [{ 'name': 'corpus', 'parameterType': { 'type': 'STRING' }, 'parameterValue': { 'value': 'romeoandjuliet' } }]. (templated) :param labels: a dictionary containing labels for the job/query, passed to BigQuery :param priority: Specifies a priority for the query. Possible values include INTERACTIVE and BATCH. The default value is INTERACTIVE. :param time_partitioning: configure optional time partitioning fields i.e. partition by field, type and expiration as per API specifications. :param cluster_fields: Request that the result of this query be stored sorted by one or more columns. BigQuery supports clustering for both partitioned and non-partitioned tables. The order of columns given determines the sort order. :param location: The geographic location of the job. Required except for US and EU. See details at https://cloud.google.com/bigquery/docs/locations#specifying_your_location :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "sql", "destination_dataset_table", "labels", "query_params", "impersonation_chain", )
[docs] template_ext: Sequence[str] = (".sql",)
[docs] template_fields_renderers = {"sql": "sql"}
[docs] ui_color = BigQueryUIColors.QUERY.value
@property def __init__( self, *, sql: str | Iterable[str], destination_dataset_table: str | None = None, write_disposition: str = "WRITE_EMPTY", allow_large_results: bool = False, flatten_results: bool | None = None, gcp_conn_id: str = "google_cloud_default", udf_config: list | None = None, use_legacy_sql: bool = True, maximum_billing_tier: int | None = None, maximum_bytes_billed: float | None = None, create_disposition: str = "CREATE_IF_NEEDED", schema_update_options: list | tuple | set | None = None, query_params: list | None = None, labels: dict | None = None, priority: str = "INTERACTIVE", time_partitioning: dict | None = None, api_resource_configs: dict | None = None, cluster_fields: list[str] | None = None, location: str | None = None, encryption_configuration: dict | None = None, impersonation_chain: str | Sequence[str] | None = None, impersonation_scopes: str | Sequence[str] | None = None, job_id: str | list[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.sql = sql self.destination_dataset_table = destination_dataset_table self.write_disposition = write_disposition self.create_disposition = create_disposition self.allow_large_results = allow_large_results self.flatten_results = flatten_results self.gcp_conn_id = gcp_conn_id self.udf_config = udf_config self.use_legacy_sql = use_legacy_sql self.maximum_billing_tier = maximum_billing_tier self.maximum_bytes_billed = maximum_bytes_billed self.schema_update_options = schema_update_options self.query_params = query_params self.labels = labels self.priority = priority self.time_partitioning = time_partitioning self.api_resource_configs = api_resource_configs self.cluster_fields = cluster_fields self.location = location self.encryption_configuration = encryption_configuration self.hook: BigQueryHook | None = None self.impersonation_chain = impersonation_chain self.impersonation_scopes = impersonation_scopes self.job_id = job_id
[docs] def execute(self, context: Context): if self.hook is None: self.log.info("Executing: %s", self.sql) self.hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, use_legacy_sql=self.use_legacy_sql, location=self.location, impersonation_chain=self.impersonation_chain, impersonation_scopes=self.impersonation_scopes, ) if isinstance(self.sql, str): self.job_id = self.hook.run_query( sql=self.sql, destination_dataset_table=self.destination_dataset_table, write_disposition=self.write_disposition, allow_large_results=self.allow_large_results, flatten_results=self.flatten_results, udf_config=self.udf_config, maximum_billing_tier=self.maximum_billing_tier, maximum_bytes_billed=self.maximum_bytes_billed, create_disposition=self.create_disposition, query_params=self.query_params, labels=self.labels, schema_update_options=self.schema_update_options, priority=self.priority, time_partitioning=self.time_partitioning, api_resource_configs=self.api_resource_configs, cluster_fields=self.cluster_fields, encryption_configuration=self.encryption_configuration, ) elif isinstance(self.sql, Iterable): self.job_id = [ self.hook.run_query( sql=s, destination_dataset_table=self.destination_dataset_table, write_disposition=self.write_disposition, allow_large_results=self.allow_large_results, flatten_results=self.flatten_results, udf_config=self.udf_config, maximum_billing_tier=self.maximum_billing_tier, maximum_bytes_billed=self.maximum_bytes_billed, create_disposition=self.create_disposition, query_params=self.query_params, labels=self.labels, schema_update_options=self.schema_update_options, priority=self.priority, time_partitioning=self.time_partitioning, api_resource_configs=self.api_resource_configs, cluster_fields=self.cluster_fields, encryption_configuration=self.encryption_configuration, ) for s in self.sql ] else: raise AirflowException(f"argument 'sql' of type {type(str)} is neither a string nor an iterable") project_id = self.hook.project_id if project_id: job_id_path = convert_job_id(job_id=self.job_id, project_id=project_id, location=self.location) # type: ignore[arg-type] context["task_instance"].xcom_push(key="job_id_path", value=job_id_path) return self.job_id
[docs] def on_kill(self) -> None: super().on_kill() if self.hook is not None: self.log.info("Cancelling running query") self.hook.cancel_job(self.hook.running_job_id)
[docs]class BigQueryCreateEmptyTableOperator(GoogleCloudBaseOperator): """ Creates a new table in the specified BigQuery dataset, optionally with schema. The schema to be used for the BigQuery table may be specified in one of two ways. You may either directly pass the schema fields in, or you may point the operator to a Google Cloud Storage object name. The object in Google Cloud Storage must be a JSON file with the schema fields in it. You can also create a table without schema. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryCreateEmptyTableOperator` :param project_id: The project to create the table into. (templated) :param dataset_id: The dataset to create the table into. (templated) :param table_id: The Name of the table to be created. (templated) :param table_resource: Table resource as described in documentation: https://cloud.google.com/bigquery/docs/reference/rest/v2/tables#Table If provided all other parameters are ignored. (templated) :param schema_fields: If set, the schema field list as defined here: https://cloud.google.com/bigquery/docs/reference/rest/v2/jobs#configuration.load.schema **Example**:: schema_fields = [ {"name": "emp_name", "type": "STRING", "mode": "REQUIRED"}, {"name": "salary", "type": "INTEGER", "mode": "NULLABLE"}, ] :param gcs_schema_object: Full path to the JSON file containing schema (templated). For example: ``gs://test-bucket/dir1/dir2/employee_schema.json`` :param time_partitioning: configure optional time partitioning fields i.e. partition by field, type and expiration as per API specifications. .. seealso:: https://cloud.google.com/bigquery/docs/reference/rest/v2/tables#timePartitioning :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud and interact with the Bigquery service. :param google_cloud_storage_conn_id: (Optional) The connection ID used to connect to Google Cloud. and interact with the Google Cloud Storage service. :param labels: a dictionary containing labels for the table, passed to BigQuery **Example (with schema JSON in GCS)**:: CreateTable = BigQueryCreateEmptyTableOperator( task_id="BigQueryCreateEmptyTableOperator_task", dataset_id="ODS", table_id="Employees", project_id="internal-gcp-project", gcs_schema_object="gs://schema-bucket/employee_schema.json", gcp_conn_id="airflow-conn-id", google_cloud_storage_conn_id="airflow-conn-id", ) **Corresponding Schema file** (``employee_schema.json``):: [ {"mode": "NULLABLE", "name": "emp_name", "type": "STRING"}, {"mode": "REQUIRED", "name": "salary", "type": "INTEGER"}, ] **Example (with schema in the DAG)**:: CreateTable = BigQueryCreateEmptyTableOperator( task_id="BigQueryCreateEmptyTableOperator_task", dataset_id="ODS", table_id="Employees", project_id="internal-gcp-project", schema_fields=[ {"name": "emp_name", "type": "STRING", "mode": "REQUIRED"}, {"name": "salary", "type": "INTEGER", "mode": "NULLABLE"}, ], gcp_conn_id="airflow-conn-id-account", google_cloud_storage_conn_id="airflow-conn-id", ) :param view: (Optional) A dictionary containing definition for the view. If set, it will create a view instead of a table: .. seealso:: https://cloud.google.com/bigquery/docs/reference/rest/v2/tables#ViewDefinition :param materialized_view: (Optional) The materialized view definition. :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } :param location: The location used for the operation. :param cluster_fields: (Optional) The fields used for clustering. BigQuery supports clustering for both partitioned and non-partitioned tables. .. seealso:: https://cloud.google.com/bigquery/docs/reference/rest/v2/tables#clustering.fields :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). :param if_exists: What should Airflow do if the table exists. If set to `log`, the TI will be passed to success and an error message will be logged. Set to `ignore` to ignore the error, set to `fail` to fail the TI, and set to `skip` to skip it. :param exists_ok: Deprecated - use `if_exists="ignore"` instead. """
[docs] template_fields: Sequence[str] = ( "dataset_id", "table_id", "table_resource", "project_id", "gcs_schema_object", "labels", "view", "materialized_view", "impersonation_chain", )
[docs] template_fields_renderers = {"table_resource": "json", "materialized_view": "json"}
[docs] ui_color = BigQueryUIColors.TABLE.value
def __init__( self, *, dataset_id: str, table_id: str, table_resource: dict[str, Any] | None = None, project_id: str = PROVIDE_PROJECT_ID, schema_fields: list | None = None, gcs_schema_object: str | None = None, time_partitioning: dict | None = None, gcp_conn_id: str = "google_cloud_default", google_cloud_storage_conn_id: str = "google_cloud_default", labels: dict | None = None, view: dict | None = None, materialized_view: dict | None = None, encryption_configuration: dict | None = None, location: str | None = None, cluster_fields: list[str] | None = None, impersonation_chain: str | Sequence[str] | None = None, if_exists: str = "log", bigquery_conn_id: str | None = None, exists_ok: bool | None = None, **kwargs, ) -> None: if bigquery_conn_id: warnings.warn( "The bigquery_conn_id parameter has been deprecated. Use the gcp_conn_id parameter instead.", AirflowProviderDeprecationWarning, stacklevel=2, ) gcp_conn_id = bigquery_conn_id super().__init__(**kwargs) self.project_id = project_id self.dataset_id = dataset_id self.table_id = table_id self.schema_fields = schema_fields self.gcs_schema_object = gcs_schema_object self.gcp_conn_id = gcp_conn_id self.google_cloud_storage_conn_id = google_cloud_storage_conn_id self.time_partitioning = time_partitioning or {} self.labels = labels self.view = view self.materialized_view = materialized_view self.encryption_configuration = encryption_configuration self.location = location self.cluster_fields = cluster_fields self.table_resource = table_resource self.impersonation_chain = impersonation_chain if exists_ok is not None: warnings.warn( "`exists_ok` parameter is deprecated, please use `if_exists`", AirflowProviderDeprecationWarning, stacklevel=2, ) self.if_exists = IfExistAction.IGNORE if exists_ok else IfExistAction.LOG else: self.if_exists = IfExistAction(if_exists)
[docs] def execute(self, context: Context) -> None: bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, location=self.location, impersonation_chain=self.impersonation_chain, ) if not self.schema_fields and self.gcs_schema_object: gcs_bucket, gcs_object = _parse_gcs_url(self.gcs_schema_object) gcs_hook = GCSHook( gcp_conn_id=self.google_cloud_storage_conn_id, impersonation_chain=self.impersonation_chain, ) schema_fields_string = gcs_hook.download_as_byte_array(gcs_bucket, gcs_object).decode("utf-8") schema_fields = json.loads(schema_fields_string) else: schema_fields = self.schema_fields try: self.log.info("Creating table") table = bq_hook.create_empty_table( project_id=self.project_id, dataset_id=self.dataset_id, table_id=self.table_id, schema_fields=schema_fields, time_partitioning=self.time_partitioning, cluster_fields=self.cluster_fields, labels=self.labels, view=self.view, materialized_view=self.materialized_view, encryption_configuration=self.encryption_configuration, table_resource=self.table_resource, exists_ok=self.if_exists == IfExistAction.IGNORE, ) persist_kwargs = { "context": context, "task_instance": self, "project_id": table.to_api_repr()["tableReference"]["projectId"], "dataset_id": table.to_api_repr()["tableReference"]["datasetId"], "table_id": table.to_api_repr()["tableReference"]["tableId"], } self.log.info( "Table %s.%s.%s created successfully", table.project, table.dataset_id, table.table_id ) except Conflict: error_msg = f"Table {self.dataset_id}.{self.table_id} already exists." if self.if_exists == IfExistAction.LOG: self.log.info(error_msg) persist_kwargs = { "context": context, "task_instance": self, "project_id": self.project_id or bq_hook.project_id, "dataset_id": self.dataset_id, "table_id": self.table_id, } elif self.if_exists == IfExistAction.FAIL: raise AirflowException(error_msg) else: raise AirflowSkipException(error_msg) BigQueryTableLink.persist(**persist_kwargs)
[docs]class BigQueryCreateExternalTableOperator(GoogleCloudBaseOperator): """ Create a new external table with data from Google Cloud Storage. The schema to be used for the BigQuery table may be specified in one of two ways. You may either directly pass the schema fields in, or you may point the operator to a Google Cloud Storage object name. The object in Google Cloud Storage must be a JSON file with the schema fields in it. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryCreateExternalTableOperator` :param bucket: The bucket to point the external table to. (templated) :param source_objects: List of Google Cloud Storage URIs to point table to. If source_format is 'DATASTORE_BACKUP', the list must only contain a single URI. :param destination_project_dataset_table: The dotted ``(<project>.)<dataset>.<table>`` BigQuery table to load data into (templated). If ``<project>`` is not included, project will be the project defined in the connection json. :param schema_fields: If set, the schema field list as defined here: https://cloud.google.com/bigquery/docs/reference/rest/v2/jobs#configuration.load.schema **Example**:: schema_fields = [ {"name": "emp_name", "type": "STRING", "mode": "REQUIRED"}, {"name": "salary", "type": "INTEGER", "mode": "NULLABLE"}, ] Should not be set when source_format is 'DATASTORE_BACKUP'. :param table_resource: Table resource as described in documentation: https://cloud.google.com/bigquery/docs/reference/rest/v2/tables#Table If provided all other parameters are ignored. External schema from object will be resolved. :param schema_object: If set, a GCS object path pointing to a .json file that contains the schema for the table. (templated) :param gcs_schema_bucket: GCS bucket name where the schema JSON is stored (templated). The default value is self.bucket. :param source_format: File format of the data. :param autodetect: Try to detect schema and format options automatically. The schema_fields and schema_object options will be honored when specified explicitly. https://cloud.google.com/bigquery/docs/schema-detect#schema_auto-detection_for_external_data_sources :param compression: (Optional) The compression type of the data source. Possible values include GZIP and NONE. The default value is NONE. This setting is ignored for Google Cloud Bigtable, Google Cloud Datastore backups and Avro formats. :param skip_leading_rows: Number of rows to skip when loading from a CSV. :param field_delimiter: The delimiter to use for the CSV. :param max_bad_records: The maximum number of bad records that BigQuery can ignore when running the job. :param quote_character: The value that is used to quote data sections in a CSV file. :param allow_quoted_newlines: Whether to allow quoted newlines (true) or not (false). :param allow_jagged_rows: Accept rows that are missing trailing optional columns. The missing values are treated as nulls. If false, records with missing trailing columns are treated as bad records, and if there are too many bad records, an invalid error is returned in the job result. Only applicable to CSV, ignored for other formats. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud and interact with the Bigquery service. :param google_cloud_storage_conn_id: (Optional) The connection ID used to connect to Google Cloud and interact with the Google Cloud Storage service. :param src_fmt_configs: configure optional fields specific to the source format :param labels: a dictionary containing labels for the table, passed to BigQuery :param encryption_configuration: (Optional) Custom encryption configuration (e.g., Cloud KMS keys). .. code-block:: python encryption_configuration = { "kmsKeyName": "projects/PROJECT/locations/LOCATION/keyRings/KEY_RING/cryptoKeys/KEY", } :param location: The location used for the operation. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "bucket", "source_objects", "schema_object", "gcs_schema_bucket", "destination_project_dataset_table", "labels", "table_resource", "impersonation_chain", )
[docs] template_fields_renderers = {"table_resource": "json"}
[docs] ui_color = BigQueryUIColors.TABLE.value
def __init__( self, *, bucket: str | None = None, source_objects: list[str] | None = None, destination_project_dataset_table: str | None = None, table_resource: dict[str, Any] | None = None, schema_fields: list | None = None, schema_object: str | None = None, gcs_schema_bucket: str | None = None, source_format: str | None = None, autodetect: bool = False, compression: str | None = None, skip_leading_rows: int | None = None, field_delimiter: str | None = None, max_bad_records: int = 0, quote_character: str | None = None, allow_quoted_newlines: bool = False, allow_jagged_rows: bool = False, gcp_conn_id: str = "google_cloud_default", google_cloud_storage_conn_id: str = "google_cloud_default", src_fmt_configs: dict | None = None, labels: dict | None = None, encryption_configuration: dict | None = None, location: str | None = None, impersonation_chain: str | Sequence[str] | None = None, bigquery_conn_id: str | None = None, **kwargs, ) -> None: if bigquery_conn_id: warnings.warn( "The bigquery_conn_id parameter has been deprecated. Use the gcp_conn_id parameter instead.", AirflowProviderDeprecationWarning, stacklevel=2, ) gcp_conn_id = bigquery_conn_id super().__init__(**kwargs) self.table_resource = table_resource self.bucket = bucket or "" self.source_objects = source_objects or [] self.schema_object = schema_object or None self.gcs_schema_bucket = gcs_schema_bucket or "" self.destination_project_dataset_table = destination_project_dataset_table or "" # BQ config kwargs_passed = any( [ destination_project_dataset_table, schema_fields, source_format, compression, skip_leading_rows, field_delimiter, max_bad_records, autodetect, quote_character, allow_quoted_newlines, allow_jagged_rows, src_fmt_configs, labels, encryption_configuration, ] ) if not table_resource: warnings.warn( "Passing table parameters via keywords arguments will be deprecated. " "Please provide table definition using `table_resource` parameter.", AirflowProviderDeprecationWarning, stacklevel=2, ) if not bucket: raise ValueError("`bucket` is required when not using `table_resource`.") if not gcs_schema_bucket: gcs_schema_bucket = bucket if not source_objects: raise ValueError("`source_objects` is required when not using `table_resource`.") if not source_format: source_format = "CSV" if not compression: compression = "NONE" if not skip_leading_rows: skip_leading_rows = 0 if not field_delimiter: field_delimiter = "," if not destination_project_dataset_table: raise ValueError( "`destination_project_dataset_table` is required when not using `table_resource`." ) self.bucket = bucket self.source_objects = source_objects self.schema_object = schema_object self.gcs_schema_bucket = gcs_schema_bucket self.destination_project_dataset_table = destination_project_dataset_table self.schema_fields = schema_fields self.source_format = source_format self.compression = compression self.skip_leading_rows = skip_leading_rows self.field_delimiter = field_delimiter self.table_resource = None else: pass if table_resource and kwargs_passed: raise ValueError("You provided both `table_resource` and exclusive keywords arguments.") self.max_bad_records = max_bad_records self.quote_character = quote_character self.allow_quoted_newlines = allow_quoted_newlines self.allow_jagged_rows = allow_jagged_rows self.gcp_conn_id = gcp_conn_id self.google_cloud_storage_conn_id = google_cloud_storage_conn_id self.autodetect = autodetect self.src_fmt_configs = src_fmt_configs or {} self.labels = labels self.encryption_configuration = encryption_configuration self.location = location self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, location=self.location, impersonation_chain=self.impersonation_chain, ) if self.table_resource: table = bq_hook.create_empty_table( table_resource=self.table_resource, ) BigQueryTableLink.persist( context=context, task_instance=self, dataset_id=table.to_api_repr()["tableReference"]["datasetId"], project_id=table.to_api_repr()["tableReference"]["projectId"], table_id=table.to_api_repr()["tableReference"]["tableId"], ) return if not self.schema_fields and self.schema_object and self.source_format != "DATASTORE_BACKUP": gcs_hook = GCSHook( gcp_conn_id=self.google_cloud_storage_conn_id, impersonation_chain=self.impersonation_chain, ) schema_fields = json.loads( gcs_hook.download(self.gcs_schema_bucket, self.schema_object).decode("utf-8") ) else: schema_fields = self.schema_fields source_uris = [f"gs://{self.bucket}/{source_object}" for source_object in self.source_objects] project_id, dataset_id, table_id = bq_hook.split_tablename( table_input=self.destination_project_dataset_table, default_project_id=bq_hook.project_id or "", ) external_data_configuration = { "source_uris": source_uris, "source_format": self.source_format, "autodetect": self.autodetect, "compression": self.compression, "maxBadRecords": self.max_bad_records, } if self.source_format == "CSV": external_data_configuration["csvOptions"] = { "fieldDelimiter": self.field_delimiter, "skipLeadingRows": self.skip_leading_rows, "quote": self.quote_character, "allowQuotedNewlines": self.allow_quoted_newlines, "allowJaggedRows": self.allow_jagged_rows, } table_resource = { "tableReference": { "projectId": project_id, "datasetId": dataset_id, "tableId": table_id, }, "labels": self.labels, "schema": {"fields": schema_fields}, "externalDataConfiguration": external_data_configuration, "location": self.location, "encryptionConfiguration": self.encryption_configuration, } table = bq_hook.create_empty_table( table_resource=table_resource, ) BigQueryTableLink.persist( context=context, task_instance=self, dataset_id=table.to_api_repr()["tableReference"]["datasetId"], project_id=table.to_api_repr()["tableReference"]["projectId"], table_id=table.to_api_repr()["tableReference"]["tableId"], )
[docs]class BigQueryDeleteDatasetOperator(GoogleCloudBaseOperator): """ Delete an existing dataset from your Project in BigQuery. https://cloud.google.com/bigquery/docs/reference/rest/v2/datasets/delete .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryDeleteDatasetOperator` :param project_id: The project id of the dataset. :param dataset_id: The dataset to be deleted. :param delete_contents: (Optional) Whether to force the deletion even if the dataset is not empty. Will delete all tables (if any) in the dataset if set to True. Will raise HttpError 400: "{dataset_id} is still in use" if set to False and dataset is not empty. The default value is False. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). **Example**:: delete_temp_data = BigQueryDeleteDatasetOperator( dataset_id="temp-dataset", project_id="temp-project", delete_contents=True, # Force the deletion of the dataset as well as its tables (if any). gcp_conn_id="_my_gcp_conn_", task_id="Deletetemp", dag=dag, ) """
[docs] template_fields: Sequence[str] = ( "dataset_id", "project_id", "impersonation_chain", )
[docs] ui_color = BigQueryUIColors.DATASET.value
def __init__( self, *, dataset_id: str, project_id: str = PROVIDE_PROJECT_ID, delete_contents: bool = False, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: self.dataset_id = dataset_id self.project_id = project_id self.delete_contents = delete_contents self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain super().__init__(**kwargs)
[docs] def execute(self, context: Context) -> None: self.log.info("Dataset id: %s Project id: %s", self.dataset_id, self.project_id) bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ) bq_hook.delete_dataset( project_id=self.project_id, dataset_id=self.dataset_id, delete_contents=self.delete_contents )
[docs]class BigQueryCreateEmptyDatasetOperator(GoogleCloudBaseOperator): """ Create a new dataset for your Project in BigQuery. https://cloud.google.com/bigquery/docs/reference/rest/v2/datasets#resource .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryCreateEmptyDatasetOperator` :param project_id: The name of the project where we want to create the dataset. :param dataset_id: The id of dataset. Don't need to provide, if datasetId in dataset_reference. :param location: The geographic location where the dataset should reside. :param dataset_reference: Dataset reference that could be provided with request body. More info: https://cloud.google.com/bigquery/docs/reference/rest/v2/datasets#resource :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). :param if_exists: What should Airflow do if the dataset exists. If set to `log`, the TI will be passed to success and an error message will be logged. Set to `ignore` to ignore the error, set to `fail` to fail the TI, and set to `skip` to skip it. **Example**:: create_new_dataset = BigQueryCreateEmptyDatasetOperator( dataset_id='new-dataset', project_id='my-project', dataset_reference={"friendlyName": "New Dataset"} gcp_conn_id='_my_gcp_conn_', task_id='newDatasetCreator', dag=dag) :param exists_ok: Deprecated - use `if_exists="ignore"` instead. """
[docs] template_fields: Sequence[str] = ( "dataset_id", "project_id", "dataset_reference", "impersonation_chain", )
[docs] template_fields_renderers = {"dataset_reference": "json"}
[docs] ui_color = BigQueryUIColors.DATASET.value
def __init__( self, *, dataset_id: str | None = None, project_id: str = PROVIDE_PROJECT_ID, dataset_reference: dict | None = None, location: str | None = None, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, if_exists: str = "log", exists_ok: bool | None = None, **kwargs, ) -> None: self.dataset_id = dataset_id self.project_id = project_id self.location = location self.gcp_conn_id = gcp_conn_id self.dataset_reference = dataset_reference or {} self.impersonation_chain = impersonation_chain if exists_ok is not None: warnings.warn( "`exists_ok` parameter is deprecated, please use `if_exists`", AirflowProviderDeprecationWarning, stacklevel=2, ) self.if_exists = IfExistAction.IGNORE if exists_ok else IfExistAction.LOG else: self.if_exists = IfExistAction(if_exists) super().__init__(**kwargs)
[docs] def execute(self, context: Context) -> None: bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, location=self.location, impersonation_chain=self.impersonation_chain, ) try: dataset = bq_hook.create_empty_dataset( project_id=self.project_id, dataset_id=self.dataset_id, dataset_reference=self.dataset_reference, location=self.location, exists_ok=self.if_exists == IfExistAction.IGNORE, ) persist_kwargs = { "context": context, "task_instance": self, "project_id": dataset["datasetReference"]["projectId"], "dataset_id": dataset["datasetReference"]["datasetId"], } except Conflict: dataset_id = self.dataset_reference.get("datasetReference", {}).get("datasetId", self.dataset_id) project_id = self.dataset_reference.get("datasetReference", {}).get( "projectId", self.project_id or bq_hook.project_id ) persist_kwargs = { "context": context, "task_instance": self, "project_id": project_id, "dataset_id": dataset_id, } error_msg = f"Dataset {dataset_id} already exists." if self.if_exists == IfExistAction.LOG: self.log.info(error_msg) elif self.if_exists == IfExistAction.FAIL: raise AirflowException(error_msg) else: raise AirflowSkipException(error_msg) BigQueryDatasetLink.persist(**persist_kwargs)
[docs]class BigQueryGetDatasetOperator(GoogleCloudBaseOperator): """ Get the dataset specified by ID. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryGetDatasetOperator` :param dataset_id: The id of dataset. Don't need to provide, if datasetId in dataset_reference. :param project_id: The name of the project where we want to create the dataset. Don't need to provide, if projectId in dataset_reference. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "dataset_id", "project_id", "impersonation_chain", )
[docs] ui_color = BigQueryUIColors.DATASET.value
def __init__( self, *, dataset_id: str, project_id: str = PROVIDE_PROJECT_ID, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: self.dataset_id = dataset_id self.project_id = project_id self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain super().__init__(**kwargs)
[docs] def execute(self, context: Context): bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ) self.log.info("Start getting dataset: %s:%s", self.project_id, self.dataset_id) dataset = bq_hook.get_dataset(dataset_id=self.dataset_id, project_id=self.project_id) dataset_api_repr = dataset.to_api_repr() BigQueryDatasetLink.persist( context=context, task_instance=self, dataset_id=dataset_api_repr["datasetReference"]["datasetId"], project_id=dataset_api_repr["datasetReference"]["projectId"], ) return dataset_api_repr
[docs]class BigQueryGetDatasetTablesOperator(GoogleCloudBaseOperator): """ Retrieve the list of tables in the specified dataset. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryGetDatasetTablesOperator` :param dataset_id: the dataset ID of the requested dataset. :param project_id: (Optional) the project of the requested dataset. If None, self.project_id will be used. :param max_results: (Optional) the maximum number of tables to return. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "dataset_id", "project_id", "impersonation_chain", )
[docs] ui_color = BigQueryUIColors.DATASET.value
def __init__( self, *, dataset_id: str, project_id: str = PROVIDE_PROJECT_ID, max_results: int | None = None, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: self.dataset_id = dataset_id self.project_id = project_id self.max_results = max_results self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain super().__init__(**kwargs)
[docs] def execute(self, context: Context): bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ) return bq_hook.get_dataset_tables( dataset_id=self.dataset_id, project_id=self.project_id, max_results=self.max_results, )
@deprecated( planned_removal_date="November 01, 2024", use_instead="BigQueryUpdateDatasetOperator", category=AirflowProviderDeprecationWarning, )
[docs]class BigQueryPatchDatasetOperator(GoogleCloudBaseOperator): """ Patch a dataset for your Project in BigQuery. This operator is deprecated. Please use :class:`airflow.providers.google.cloud.operators.bigquery.BigQueryUpdateTableOperator` instead. Only replaces fields that are provided in the submitted dataset resource. :param dataset_id: The id of dataset. Don't need to provide, if datasetId in dataset_reference. :param dataset_resource: Dataset resource that will be provided with request body. https://cloud.google.com/bigquery/docs/reference/rest/v2/datasets#resource :param project_id: The name of the project where we want to create the dataset. Don't need to provide, if projectId in dataset_reference. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "dataset_id", "project_id", "impersonation_chain", )
[docs] template_fields_renderers = {"dataset_resource": "json"}
[docs] ui_color = BigQueryUIColors.DATASET.value
def __init__( self, *, dataset_id: str, dataset_resource: dict, project_id: str = PROVIDE_PROJECT_ID, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: self.dataset_id = dataset_id self.project_id = project_id self.gcp_conn_id = gcp_conn_id self.dataset_resource = dataset_resource self.impersonation_chain = impersonation_chain super().__init__(**kwargs)
[docs] def execute(self, context: Context): bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ) return bq_hook.patch_dataset( dataset_id=self.dataset_id, dataset_resource=self.dataset_resource, project_id=self.project_id, )
[docs]class BigQueryUpdateTableOperator(GoogleCloudBaseOperator): """ Update a table for your Project in BigQuery. Use ``fields`` to specify which fields of table to update. If a field is listed in ``fields`` and is ``None`` in table, it will be deleted. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryUpdateTableOperator` :param dataset_id: The id of dataset. Don't need to provide, if datasetId in table_reference. :param table_id: The id of table. Don't need to provide, if tableId in table_reference. :param table_resource: Dataset resource that will be provided with request body. https://cloud.google.com/bigquery/docs/reference/rest/v2/tables#resource :param fields: The fields of ``table`` to change, spelled as the Table properties (e.g. "friendly_name"). :param project_id: The name of the project where we want to create the table. Don't need to provide, if projectId in table_reference. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "dataset_id", "table_id", "project_id", "impersonation_chain", )
[docs] template_fields_renderers = {"table_resource": "json"}
[docs] ui_color = BigQueryUIColors.TABLE.value
def __init__( self, *, table_resource: dict[str, Any], fields: list[str] | None = None, dataset_id: str | None = None, table_id: str | None = None, project_id: str = PROVIDE_PROJECT_ID, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: self.dataset_id = dataset_id self.table_id = table_id self.project_id = project_id self.fields = fields self.gcp_conn_id = gcp_conn_id self.table_resource = table_resource self.impersonation_chain = impersonation_chain super().__init__(**kwargs)
[docs] def execute(self, context: Context): bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ) table = bq_hook.update_table( table_resource=self.table_resource, fields=self.fields, dataset_id=self.dataset_id, table_id=self.table_id, project_id=self.project_id, ) BigQueryTableLink.persist( context=context, task_instance=self, dataset_id=table["tableReference"]["datasetId"], project_id=table["tableReference"]["projectId"], table_id=table["tableReference"]["tableId"], ) return table
[docs]class BigQueryUpdateDatasetOperator(GoogleCloudBaseOperator): """ Update a dataset for your Project in BigQuery. Use ``fields`` to specify which fields of dataset to update. If a field is listed in ``fields`` and is ``None`` in dataset, it will be deleted. If no ``fields`` are provided then all fields of provided ``dataset_resource`` will be used. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryUpdateDatasetOperator` :param dataset_id: The id of dataset. Don't need to provide, if datasetId in dataset_reference. :param dataset_resource: Dataset resource that will be provided with request body. https://cloud.google.com/bigquery/docs/reference/rest/v2/datasets#resource :param fields: The properties of dataset to change (e.g. "friendly_name"). :param project_id: The name of the project where we want to create the dataset. Don't need to provide, if projectId in dataset_reference. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "dataset_id", "project_id", "impersonation_chain", )
[docs] template_fields_renderers = {"dataset_resource": "json"}
[docs] ui_color = BigQueryUIColors.DATASET.value
def __init__( self, *, dataset_resource: dict[str, Any], fields: list[str] | None = None, dataset_id: str | None = None, project_id: str = PROVIDE_PROJECT_ID, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: self.dataset_id = dataset_id self.project_id = project_id self.fields = fields self.gcp_conn_id = gcp_conn_id self.dataset_resource = dataset_resource self.impersonation_chain = impersonation_chain super().__init__(**kwargs)
[docs] def execute(self, context: Context): bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ) fields = self.fields or list(self.dataset_resource.keys()) dataset = bq_hook.update_dataset( dataset_resource=self.dataset_resource, project_id=self.project_id, dataset_id=self.dataset_id, fields=fields, ) dataset_api_repr = dataset.to_api_repr() BigQueryDatasetLink.persist( context=context, task_instance=self, dataset_id=dataset_api_repr["datasetReference"]["datasetId"], project_id=dataset_api_repr["datasetReference"]["projectId"], ) return dataset_api_repr
[docs]class BigQueryDeleteTableOperator(GoogleCloudBaseOperator): """ Delete a BigQuery table. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryDeleteTableOperator` :param deletion_dataset_table: A dotted ``(<project>.|<project>:)<dataset>.<table>`` that indicates which table will be deleted. (templated) :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param ignore_if_missing: if True, then return success even if the requested table does not exist. :param location: The location used for the operation. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "deletion_dataset_table", "impersonation_chain", )
[docs] ui_color = BigQueryUIColors.TABLE.value
def __init__( self, *, deletion_dataset_table: str, gcp_conn_id: str = "google_cloud_default", ignore_if_missing: bool = False, location: str | None = None, impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.deletion_dataset_table = deletion_dataset_table self.gcp_conn_id = gcp_conn_id self.ignore_if_missing = ignore_if_missing self.location = location self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: self.log.info("Deleting: %s", self.deletion_dataset_table) hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, location=self.location, impersonation_chain=self.impersonation_chain, ) hook.delete_table(table_id=self.deletion_dataset_table, not_found_ok=self.ignore_if_missing)
[docs]class BigQueryUpsertTableOperator(GoogleCloudBaseOperator): """ Upsert to a BigQuery table. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryUpsertTableOperator` :param dataset_id: A dotted ``(<project>.|<project>:)<dataset>`` that indicates which dataset will be updated. (templated) :param table_resource: a table resource. see https://cloud.google.com/bigquery/docs/reference/v2/tables#resource :param project_id: The name of the project where we want to update the dataset. Don't need to provide, if projectId in dataset_reference. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param location: The location used for the operation. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "dataset_id", "table_resource", "impersonation_chain", "project_id", )
[docs] template_fields_renderers = {"table_resource": "json"}
[docs] ui_color = BigQueryUIColors.TABLE.value
def __init__( self, *, dataset_id: str, table_resource: dict, project_id: str = PROVIDE_PROJECT_ID, gcp_conn_id: str = "google_cloud_default", location: str | None = None, impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.dataset_id = dataset_id self.table_resource = table_resource self.project_id = project_id self.gcp_conn_id = gcp_conn_id self.location = location self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: self.log.info("Upserting Dataset: %s with table_resource: %s", self.dataset_id, self.table_resource) hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, location=self.location, impersonation_chain=self.impersonation_chain, ) table = hook.run_table_upsert( dataset_id=self.dataset_id, table_resource=self.table_resource, project_id=self.project_id, ) BigQueryTableLink.persist( context=context, task_instance=self, dataset_id=table["tableReference"]["datasetId"], project_id=table["tableReference"]["projectId"], table_id=table["tableReference"]["tableId"], )
[docs]class BigQueryUpdateTableSchemaOperator(GoogleCloudBaseOperator): """ Update BigQuery Table Schema. Updates fields on a table schema based on contents of the supplied schema_fields_updates parameter. The supplied schema does not need to be complete, if the field already exists in the schema you only need to supply keys & values for the items you want to patch, just ensure the "name" key is set. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryUpdateTableSchemaOperator` :param schema_fields_updates: a partial schema resource. see https://cloud.google.com/bigquery/docs/reference/rest/v2/tables#TableSchema .. code-block:: python schema_fields_updates = [ {"name": "emp_name", "description": "Some New Description"}, { "name": "salary", "policyTags": {"names": ["some_new_policy_tag"]}, }, { "name": "departments", "fields": [ {"name": "name", "description": "Some New Description"}, {"name": "type", "description": "Some New Description"}, ], }, ] :param include_policy_tags: (Optional) If set to True policy tags will be included in the update request which requires special permissions even if unchanged (default False) see https://cloud.google.com/bigquery/docs/column-level-security#roles :param dataset_id: A dotted ``(<project>.|<project>:)<dataset>`` that indicates which dataset will be updated. (templated) :param table_id: The table ID of the requested table. (templated) :param project_id: The name of the project where we want to update the dataset. Don't need to provide, if projectId in dataset_reference. :param gcp_conn_id: (Optional) The connection ID used to connect to Google Cloud. :param location: The location used for the operation. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). """
[docs] template_fields: Sequence[str] = ( "schema_fields_updates", "dataset_id", "table_id", "project_id", "impersonation_chain", )
[docs] template_fields_renderers = {"schema_fields_updates": "json"}
[docs] ui_color = BigQueryUIColors.TABLE.value
def __init__( self, *, schema_fields_updates: list[dict[str, Any]], dataset_id: str, table_id: str, include_policy_tags: bool = False, project_id: str = PROVIDE_PROJECT_ID, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, location: str | None = None, **kwargs, ) -> None: self.schema_fields_updates = schema_fields_updates self.include_policy_tags = include_policy_tags self.table_id = table_id self.dataset_id = dataset_id self.project_id = project_id self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain self.location = location super().__init__(**kwargs)
[docs] def execute(self, context: Context): bq_hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, location=self.location ) table = bq_hook.update_table_schema( schema_fields_updates=self.schema_fields_updates, include_policy_tags=self.include_policy_tags, dataset_id=self.dataset_id, table_id=self.table_id, project_id=self.project_id, ) BigQueryTableLink.persist( context=context, task_instance=self, dataset_id=table["tableReference"]["datasetId"], project_id=table["tableReference"]["projectId"], table_id=table["tableReference"]["tableId"], ) return table
[docs]class BigQueryInsertJobOperator(GoogleCloudBaseOperator, _BigQueryOpenLineageMixin): """ Execute a BigQuery job. Waits for the job to complete and returns job id. This operator work in the following way: - it calculates a unique hash of the job using job's configuration or uuid if ``force_rerun`` is True - creates ``job_id`` in form of ``[provided_job_id | airflow_{dag_id}_{task_id}_{exec_date}]_{uniqueness_suffix}`` - submits a BigQuery job using the ``job_id`` - if job with given id already exists then it tries to reattach to the job if its not done and its state is in ``reattach_states``. If the job is done the operator will raise ``AirflowException``. Using ``force_rerun`` will submit a new job every time without attaching to already existing ones. For job definition see here: https://cloud.google.com/bigquery/docs/reference/v2/jobs .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:BigQueryInsertJobOperator` :param configuration: The configuration parameter maps directly to BigQuery's configuration field in the job object. For more details see https://cloud.google.com/bigquery/docs/reference/rest/v2/Job#jobconfiguration :param job_id: The ID of the job. It will be suffixed with hash of job configuration unless ``force_rerun`` is True. The ID must contain only letters (a-z, A-Z), numbers (0-9), underscores (_), or dashes (-). The maximum length is 1,024 characters. If not provided then uuid will be generated. :param force_rerun: If True then operator will use hash of uuid as job id suffix :param reattach_states: Set of BigQuery job's states in case of which we should reattach to the job. Should be other than final states. :param project_id: Google Cloud Project where the job is running :param location: location the job is running :param gcp_conn_id: The connection ID used to connect to Google Cloud. :param impersonation_chain: Optional service account to impersonate using short-term credentials, or chained list of accounts required to get the access_token of the last account in the list, which will be impersonated in the request. If set as a string, the account must grant the originating account the Service Account Token Creator IAM role. If set as a sequence, the identities from the list must grant Service Account Token Creator IAM role to the directly preceding identity, with first account from the list granting this role to the originating account (templated). :param cancel_on_kill: Flag which indicates whether cancel the hook's job or not, when on_kill is called :param result_retry: How to retry the `result` call that retrieves rows :param result_timeout: The number of seconds to wait for `result` method before using `result_retry` :param deferrable: Run operator in the deferrable mode :param poll_interval: (Deferrable mode only) polling period in seconds to check for the status of job. Defaults to 4 seconds. """
[docs] template_fields: Sequence[str] = ( "configuration", "job_id", "impersonation_chain", "project_id", )
[docs] template_ext: Sequence[str] = ( ".json", ".sql", )
[docs] template_fields_renderers = {"configuration": "json", "configuration.query.query": "sql"}
[docs] ui_color = BigQueryUIColors.QUERY.value
def __init__( self, configuration: dict[str, Any], project_id: str = PROVIDE_PROJECT_ID, location: str | None = None, job_id: str | None = None, force_rerun: bool = True, reattach_states: set[str] | None = None, gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, cancel_on_kill: bool = True, result_retry: Retry = DEFAULT_RETRY, result_timeout: float | None = None, deferrable: bool = conf.getboolean("operators", "default_deferrable", fallback=False), poll_interval: float = 4.0, **kwargs, ) -> None: super().__init__(**kwargs) self.configuration = configuration self.location = location self.job_id = job_id self.project_id = project_id self.gcp_conn_id = gcp_conn_id self.force_rerun = force_rerun self.reattach_states: set[str] = reattach_states or set() self.impersonation_chain = impersonation_chain self.cancel_on_kill = cancel_on_kill self.result_retry = result_retry self.result_timeout = result_timeout self.hook: BigQueryHook | None = None self.deferrable = deferrable self.poll_interval = poll_interval @cached_property
[docs] def sql(self) -> str | None: try: return self.configuration["query"]["query"] except KeyError: return None
[docs] def prepare_template(self) -> None: # If .json is passed then we have to read the file if isinstance(self.configuration, str) and self.configuration.endswith(".json"): with open(self.configuration) as file: self.configuration = json.loads(file.read())
def _add_job_labels(self) -> None: dag_label = self.dag_id.lower() task_label = self.task_id.lower() if LABEL_REGEX.match(dag_label) and LABEL_REGEX.match(task_label): automatic_labels = {"airflow-dag": dag_label, "airflow-task": task_label} if isinstance(self.configuration.get("labels"), dict): self.configuration["labels"].update(automatic_labels) elif "labels" not in self.configuration: self.configuration["labels"] = automatic_labels def _submit_job( self, hook: BigQueryHook, job_id: str, ) -> BigQueryJob: # Annotate the job with dag and task id labels self._add_job_labels() # Submit a new job without waiting for it to complete. return hook.insert_job( configuration=self.configuration, project_id=self.project_id, location=self.location, job_id=job_id, timeout=self.result_timeout, retry=self.result_retry, nowait=True, ) @staticmethod def _handle_job_error(job: BigQueryJob | UnknownJob) -> None: if job.error_result: raise AirflowException(f"BigQuery job {job.job_id} failed: {job.error_result}")
[docs] def execute(self, context: Any): hook = BigQueryHook( gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ) self.hook = hook if self.project_id is None: self.project_id = hook.project_id self.job_id = hook.generate_job_id( job_id=self.job_id, dag_id=self.dag_id, task_id=self.task_id, logical_date=context["logical_date"], configuration=self.configuration, force_rerun=self.force_rerun, ) try: self.log.info("Executing: %s'", self.configuration) # Create a job job: BigQueryJob | UnknownJob = self._submit_job(hook, self.job_id) except Conflict: # If the job already exists retrieve it job = hook.get_job( project_id=self.project_id, location=self.location, job_id=self.job_id, ) if job.state not in self.reattach_states: # Same job configuration, so we need force_rerun raise AirflowException( f"Job with id: {self.job_id} already exists and is in {job.state} state. If you " f"want to force rerun it consider setting `force_rerun=True`." f"Or, if you want to reattach in this scenario add {job.state} to `reattach_states`" ) else: # Job already reached state DONE if job.state == "DONE": raise AirflowException("Job is already in state DONE. Can not reattach to this job.") # We are reattaching to a job self.log.info("Reattaching to existing Job in state %s", job.state) self._handle_job_error(job) job_types = { LoadJob._JOB_TYPE: ["sourceTable", "destinationTable"], CopyJob._JOB_TYPE: ["sourceTable", "destinationTable"], ExtractJob._JOB_TYPE: ["sourceTable"], QueryJob._JOB_TYPE: ["destinationTable"], } if self.project_id: for job_type, tables_prop in job_types.items(): job_configuration = job.to_api_repr()["configuration"] if job_type in job_configuration: for table_prop in tables_prop: if table_prop in job_configuration[job_type]: table = job_configuration[job_type][table_prop] persist_kwargs = { "context": context, "task_instance": self, "project_id": self.project_id, "table_id": table, } if not isinstance(table, str): persist_kwargs["table_id"] = table["tableId"] persist_kwargs["dataset_id"] = table["datasetId"] persist_kwargs["project_id"] = table["projectId"] BigQueryTableLink.persist(**persist_kwargs) self.job_id = job.job_id if self.project_id: job_id_path = convert_job_id( job_id=self.job_id, # type: ignore[arg-type] project_id=self.project_id, location=self.location, ) context["ti"].xcom_push(key="job_id_path", value=job_id_path) # Wait for the job to complete if not self.deferrable: job.result(timeout=self.result_timeout, retry=self.result_retry) self._handle_job_error(job) return self.job_id else: if job.running(): self.defer( timeout=self.execution_timeout, trigger=BigQueryInsertJobTrigger( conn_id=self.gcp_conn_id, job_id=self.job_id, project_id=self.project_id, location=self.location or hook.location, poll_interval=self.poll_interval, impersonation_chain=self.impersonation_chain, cancel_on_kill=self.cancel_on_kill, ), method_name="execute_complete", ) self.log.info("Current state of job %s is %s", job.job_id, job.state) self._handle_job_error(job)
[docs] def execute_complete(self, context: Context, event: dict[str, Any]) -> str | None: """ Act as a callback for when the trigger fires. This returns immediately. It relies on trigger to throw an exception, otherwise it assumes execution was successful. """ if event["status"] == "error": raise AirflowException(event["message"]) self.log.info( "%s completed with response %s ", self.task_id, event["message"], ) # Save job_id as an attribute to be later used by listeners self.job_id = event.get("job_id") return self.job_id
[docs] def on_kill(self) -> None: if self.job_id and self.cancel_on_kill: self.hook.cancel_job( # type: ignore[union-attr] job_id=self.job_id, project_id=self.project_id, location=self.location ) else: self.log.info("Skipping to cancel job: %s:%s.%s", self.project_id, self.location, self.job_id)

Was this entry helpful?