#
# 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 a Google AutoML hook.
.. spelling:word-list::
PredictResponse
"""
from __future__ import annotations
from functools import cached_property
from typing import TYPE_CHECKING, Sequence
from google.api_core.gapic_v1.method import DEFAULT, _MethodDefault
from google.cloud.automl_v1beta1 import (
AutoMlClient,
BatchPredictInputConfig,
BatchPredictOutputConfig,
Dataset,
ExamplePayload,
ImageObjectDetectionModelDeploymentMetadata,
InputConfig,
Model,
PredictionServiceClient,
PredictResponse,
)
from airflow.exceptions import AirflowException
from airflow.providers.google.common.consts import CLIENT_INFO
from airflow.providers.google.common.hooks.base_google import PROVIDE_PROJECT_ID, GoogleBaseHook
if TYPE_CHECKING:
from google.api_core.operation import Operation
from google.api_core.retry import Retry
from google.cloud.automl_v1beta1.services.auto_ml.pagers import (
ListColumnSpecsPager,
ListDatasetsPager,
ListTableSpecsPager,
)
from google.protobuf.field_mask_pb2 import FieldMask
[docs]class CloudAutoMLHook(GoogleBaseHook):
"""
Google Cloud AutoML hook.
All the methods in the hook where project_id is used must be called with
keyword arguments rather than positional.
"""
def __init__(
self,
gcp_conn_id: str = "google_cloud_default",
impersonation_chain: str | Sequence[str] | None = None,
**kwargs,
) -> None:
if kwargs.get("delegate_to") is not None:
raise RuntimeError(
"The `delegate_to` parameter has been deprecated before and finally removed in this version"
" of Google Provider. You MUST convert it to `impersonate_chain`"
)
super().__init__(
gcp_conn_id=gcp_conn_id,
impersonation_chain=impersonation_chain,
)
self._client: AutoMlClient | None = None
@staticmethod
[docs] def get_conn(self) -> AutoMlClient:
"""
Retrieve connection to AutoML.
:return: Google Cloud AutoML client object.
"""
if self._client is None:
self._client = AutoMlClient(credentials=self.get_credentials(), client_info=CLIENT_INFO)
return self._client
[docs] def wait_for_operation(self, operation: Operation, timeout: float | None = None):
"""Wait for long-lasting operation to complete."""
try:
return operation.result(timeout=timeout)
except Exception:
error = operation.exception(timeout=timeout)
raise AirflowException(error)
@cached_property
[docs] def prediction_client(self) -> PredictionServiceClient:
"""
Creates PredictionServiceClient.
:return: Google Cloud AutoML PredictionServiceClient client object.
"""
return PredictionServiceClient(credentials=self.get_credentials(), client_info=CLIENT_INFO)
@GoogleBaseHook.fallback_to_default_project_id
[docs] def create_model(
self,
model: dict | Model,
location: str,
project_id: str = PROVIDE_PROJECT_ID,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
retry: Retry | _MethodDefault = DEFAULT,
) -> Operation:
"""
Create a model_id and returns a Model in the `response` field when it completes.
When you create a model, several model evaluations are created for it:
a global evaluation, and one evaluation for each annotation spec.
:param model: The model_id to create. If a dict is provided, it must be of the same form
as the protobuf message `google.cloud.automl_v1beta1.types.Model`
:param project_id: ID of the Google Cloud project where model will be created if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests
will not be retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete.
Note that if `retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types._OperationFuture` instance
"""
client = self.get_conn()
parent = f"projects/{project_id}/locations/{location}"
return client.create_model(
request={"parent": parent, "model": model},
retry=retry,
timeout=timeout,
metadata=metadata,
)
@GoogleBaseHook.fallback_to_default_project_id
[docs] def batch_predict(
self,
model_id: str,
input_config: dict | BatchPredictInputConfig,
output_config: dict | BatchPredictOutputConfig,
location: str,
project_id: str = PROVIDE_PROJECT_ID,
params: dict[str, str] | None = None,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> Operation:
"""
Perform a batch prediction and returns a long-running operation object.
Unlike the online `Predict`, batch prediction result won't be immediately
available in the response. Instead, a long-running operation object is returned.
:param model_id: Name of the model_id requested to serve the batch prediction.
:param input_config: Required. The input configuration for batch prediction.
If a dict is provided, it must be of the same form as the protobuf message
`google.cloud.automl_v1beta1.types.BatchPredictInputConfig`
:param output_config: Required. The Configuration specifying where output predictions should be
written. If a dict is provided, it must be of the same form as the protobuf message
`google.cloud.automl_v1beta1.types.BatchPredictOutputConfig`
:param params: Additional domain-specific parameters for the predictions, any string must be up to
25000 characters long.
:param project_id: ID of the Google Cloud project where model is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types._OperationFuture` instance
"""
client = self.prediction_client
name = f"projects/{project_id}/locations/{location}/models/{model_id}"
result = client.batch_predict(
request={
"name": name,
"input_config": input_config,
"output_config": output_config,
"params": params,
},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def predict(
self,
model_id: str,
payload: dict | ExamplePayload,
location: str,
project_id: str = PROVIDE_PROJECT_ID,
params: dict[str, str] | None = None,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> PredictResponse:
"""
Perform an online prediction and returns the prediction result in the response.
:param model_id: Name of the model_id requested to serve the prediction.
:param payload: Required. Payload to perform a prediction on. The payload must match the problem type
that the model_id was trained to solve. If a dict is provided, it must be of
the same form as the protobuf message `google.cloud.automl_v1beta1.types.ExamplePayload`
:param params: Additional domain-specific parameters, any string must be up to 25000 characters long.
:param project_id: ID of the Google Cloud project where model is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types.PredictResponse` instance
"""
client = self.prediction_client
name = f"projects/{project_id}/locations/{location}/models/{model_id}"
result = client.predict(
request={"name": name, "payload": payload, "params": params},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def create_dataset(
self,
dataset: dict | Dataset,
location: str,
project_id: str = PROVIDE_PROJECT_ID,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> Dataset:
"""
Create a dataset.
:param dataset: The dataset to create. If a dict is provided, it must be of the
same form as the protobuf message Dataset.
:param project_id: ID of the Google Cloud project where dataset is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types.Dataset` instance.
"""
client = self.get_conn()
parent = f"projects/{project_id}/locations/{location}"
result = client.create_dataset(
request={"parent": parent, "dataset": dataset},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def import_data(
self,
dataset_id: str,
location: str,
input_config: dict | InputConfig,
project_id: str = PROVIDE_PROJECT_ID,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> Operation:
"""
Import data into a dataset. For Tables this method can only be called on an empty Dataset.
:param dataset_id: Name of the AutoML dataset.
:param input_config: The desired input location and its domain specific semantics, if any.
If a dict is provided, it must be of the same form as the protobuf message InputConfig.
:param project_id: ID of the Google Cloud project where dataset is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types._OperationFuture` instance
"""
client = self.get_conn()
name = f"projects/{project_id}/locations/{location}/datasets/{dataset_id}"
result = client.import_data(
request={"name": name, "input_config": input_config},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def list_column_specs(
self,
dataset_id: str,
table_spec_id: str,
location: str,
project_id: str = PROVIDE_PROJECT_ID,
field_mask: dict | FieldMask | None = None,
filter_: str | None = None,
page_size: int | None = None,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> ListColumnSpecsPager:
"""
List column specs in a table spec.
:param dataset_id: Name of the AutoML dataset.
:param table_spec_id: table_spec_id for path builder.
:param field_mask: Mask specifying which fields to read. If a dict is provided, it must be of the same
form as the protobuf message `google.cloud.automl_v1beta1.types.FieldMask`
:param filter_: Filter expression, see go/filtering.
:param page_size: The maximum number of resources contained in the
underlying API response. If page streaming is performed per
resource, this parameter does not affect the return value. If page
streaming is performed per-page, this determines the maximum number
of resources in a page.
:param project_id: ID of the Google Cloud project where dataset is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types.ColumnSpec` instance.
"""
client = self.get_conn()
parent = client.table_spec_path(
project=project_id,
location=location,
dataset=dataset_id,
table_spec=table_spec_id,
)
result = client.list_column_specs(
request={"parent": parent, "field_mask": field_mask, "filter": filter_, "page_size": page_size},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def get_model(
self,
model_id: str,
location: str,
project_id: str = PROVIDE_PROJECT_ID,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> Model:
"""
Get a AutoML model.
:param model_id: Name of the model.
:param project_id: ID of the Google Cloud project where model is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types.Model` instance.
"""
client = self.get_conn()
name = f"projects/{project_id}/locations/{location}/models/{model_id}"
result = client.get_model(
request={"name": name},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def delete_model(
self,
model_id: str,
location: str,
project_id: str = PROVIDE_PROJECT_ID,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> Operation:
"""
Delete a AutoML model.
:param model_id: Name of the model.
:param project_id: ID of the Google Cloud project where model is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types._OperationFuture` instance.
"""
client = self.get_conn()
name = f"projects/{project_id}/locations/{location}/models/{model_id}"
result = client.delete_model(
request={"name": name},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
[docs] def update_dataset(
self,
dataset: dict | Dataset,
update_mask: dict | FieldMask | None = None,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> Dataset:
"""
Update a dataset.
:param dataset: The dataset which replaces the resource on the server.
If a dict is provided, it must be of the same form as the protobuf message Dataset.
:param update_mask: The update mask applies to the resource. If a dict is provided, it must
be of the same form as the protobuf message FieldMask.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types.Dataset` instance..
"""
client = self.get_conn()
result = client.update_dataset(
request={"dataset": dataset, "update_mask": update_mask},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def deploy_model(
self,
model_id: str,
location: str,
project_id: str = PROVIDE_PROJECT_ID,
image_detection_metadata: ImageObjectDetectionModelDeploymentMetadata | dict | None = None,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> Operation:
"""
Deploys a model.
If a model is already deployed, deploying it with the same parameters
has no effect. Deploying with different parameters (as e.g. changing node_number) will
reset the deployment state without pausing the model_id's availability.
Only applicable for Text Classification, Image Object Detection and Tables; all other
domains manage deployment automatically.
:param model_id: Name of the model requested to serve the prediction.
:param image_detection_metadata: Model deployment metadata specific to Image Object Detection.
If a dict is provided, it must be of the same form as the protobuf message
ImageObjectDetectionModelDeploymentMetadata
:param project_id: ID of the Google Cloud project where model will be created if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types._OperationFuture` instance.
"""
client = self.get_conn()
name = f"projects/{project_id}/locations/{location}/models/{model_id}"
result = client.deploy_model(
request={
"name": name,
"image_object_detection_model_deployment_metadata": image_detection_metadata,
},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
[docs] def list_table_specs(
self,
dataset_id: str,
location: str,
project_id: str | None = None,
filter_: str | None = None,
page_size: int | None = None,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> ListTableSpecsPager:
"""
List table specs in a dataset_id.
:param dataset_id: Name of the dataset.
:param filter_: Filter expression, see go/filtering.
:param page_size: The maximum number of resources contained in the
underlying API response. If page streaming is performed per
resource, this parameter does not affect the return value. If page
streaming is performed per-page, this determines the maximum number
of resources in a page.
:param project_id: ID of the Google Cloud project where dataset is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: A `google.gax.PageIterator` instance. By default, this
is an iterable of `google.cloud.automl_v1beta1.types.TableSpec` instances.
This object can also be configured to iterate over the pages
of the response through the `options` parameter.
"""
client = self.get_conn()
parent = f"projects/{project_id}/locations/{location}/datasets/{dataset_id}"
result = client.list_table_specs(
request={"parent": parent, "filter": filter_, "page_size": page_size},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def list_datasets(
self,
location: str,
project_id: str,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> ListDatasetsPager:
"""
List datasets in a project.
:param project_id: ID of the Google Cloud project where dataset is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: A `google.gax.PageIterator` instance. By default, this
is an iterable of `google.cloud.automl_v1beta1.types.Dataset` instances.
This object can also be configured to iterate over the pages
of the response through the `options` parameter.
"""
client = self.get_conn()
parent = f"projects/{project_id}/locations/{location}"
result = client.list_datasets(
request={"parent": parent},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
@GoogleBaseHook.fallback_to_default_project_id
[docs] def delete_dataset(
self,
dataset_id: str,
location: str,
project_id: str,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
) -> Operation:
"""
Delete a dataset and all of its contents.
:param dataset_id: ID of dataset to be deleted.
:param project_id: ID of the Google Cloud project where dataset is located if None then
default project_id is used.
:param location: The location of the project.
:param retry: A retry object used to retry requests. If `None` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
`retry` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:return: `google.cloud.automl_v1beta1.types._OperationFuture` instance
"""
client = self.get_conn()
name = f"projects/{project_id}/locations/{location}/datasets/{dataset_id}"
result = client.delete_dataset(
request={"name": name},
retry=retry,
timeout=timeout,
metadata=metadata,
)
return result
