apache-airflow-providers-google 10.17.0rc1__py3-none-any.whl → 10.18.0rc1__py3-none-any.whl

airflow/providers/google/cloud/operators/gcs.py

@@ -29,6 +29,8 @@ from typing import TYPE_CHECKING, Sequence
 
 import pendulum
 
+from airflow.providers.google.common.hooks.base_google import PROVIDE_PROJECT_ID
+
 if TYPE_CHECKING:
     from airflow.utils.context import Context
 
@@ -119,7 +121,7 @@ class GCSCreateBucketOperator(GoogleCloudBaseOperator):
         resource: dict | None = None,
         storage_class: str = "MULTI_REGIONAL",
         location: str = "US",
-        project_id: str | None = None,
+        project_id: str = PROVIDE_PROJECT_ID,
         labels: dict | None = None,
         gcp_conn_id: str = "google_cloud_default",
         impersonation_chain: str | Sequence[str] | None = None,
@@ -297,7 +299,7 @@ class GCSDeleteObjectsOperator(GoogleCloudBaseOperator):
         *,
         bucket_name: str,
         objects: list[str] | None = None,
-        prefix: str | None = None,
+        prefix: str | list[str] | None = None,
         gcp_conn_id: str = "google_cloud_default",
         impersonation_chain: str | Sequence[str] | None = None,
         **kwargs,
@@ -309,12 +311,14 @@ class GCSDeleteObjectsOperator(GoogleCloudBaseOperator):
         self.impersonation_chain = impersonation_chain
 
         if objects is None and prefix is None:
-            err_message = "(Task {task_id}) Either object or prefix should be set. Both are None.".format(
+            err_message = "(Task {task_id}) Either objects or prefix should be set. Both are None.".format(
                 **kwargs
             )
             raise ValueError(err_message)
+        if objects is not None and prefix is not None:
+            err_message = "(Task {task_id}) Objects or prefix should be set. Both provided.".format(**kwargs)
+            raise ValueError(err_message)
 
-        self._objects: list[str] = []
         super().__init__(**kwargs)
 
     def execute(self, context: Context) -> None:
@@ -324,15 +328,14 @@ class GCSDeleteObjectsOperator(GoogleCloudBaseOperator):
         )
 
         if self.objects is not None:
-            self._objects = self.objects
+            objects = self.objects
         else:
-            self._objects = hook.list(bucket_name=self.bucket_name, prefix=self.prefix)
-        self.log.info("Deleting %s objects from %s", len(self._objects), self.bucket_name)
-        for object_name in self._objects:
+            objects = hook.list(bucket_name=self.bucket_name, prefix=self.prefix)
+        self.log.info("Deleting %s objects from %s", len(objects), self.bucket_name)
+        for object_name in objects:
             hook.delete(bucket_name=self.bucket_name, object_name=object_name)
 
-    def get_openlineage_facets_on_complete(self, task_instance):
-        """Implement on_complete as execute() resolves object names."""
+    def get_openlineage_facets_on_start(self):
         from openlineage.client.facet import (
             LifecycleStateChange,
             LifecycleStateChangeDatasetFacet,
@@ -342,8 +345,17 @@ class GCSDeleteObjectsOperator(GoogleCloudBaseOperator):
 
         from airflow.providers.openlineage.extractors import OperatorLineage
 
-        if not self._objects:
-            return OperatorLineage()
+        objects = []
+        if self.objects is not None:
+            objects = self.objects
+        elif self.prefix is not None:
+            prefixes = [self.prefix] if isinstance(self.prefix, str) else self.prefix
+            for pref in prefixes:
+                # Use parent if not a file (dot not in name) and not a dir (ends with slash)
+                if "." not in pref.split("/")[-1] and not pref.endswith("/"):
+                    pref = Path(pref).parent.as_posix()
+                pref = "/" if pref in (".", "", "/") else pref.rstrip("/")
+                objects.append(pref)
 
         bucket_url = f"gs://{self.bucket_name}"
         input_datasets = [
@@ -360,7 +372,7 @@ class GCSDeleteObjectsOperator(GoogleCloudBaseOperator):
                     )
                 },
             )
-            for object_name in self._objects
+            for object_name in objects
         ]
 
         return OperatorLineage(inputs=input_datasets)
@@ -774,8 +786,8 @@ class GCSTimeSpanFileTransformOperator(GoogleCloudBaseOperator):
         self.upload_continue_on_fail = upload_continue_on_fail
         self.upload_num_attempts = upload_num_attempts
 
-        self._source_object_names: list[str] = []
-        self._destination_object_names: list[str] = []
+        self._source_prefix_interp: str | None = None
+        self._destination_prefix_interp: str | None = None
 
     def execute(self, context: Context) -> list[str]:
         # Define intervals and prefixes.
@@ -803,11 +815,11 @@ class GCSTimeSpanFileTransformOperator(GoogleCloudBaseOperator):
         timespan_start = timespan_start.in_timezone(timezone.utc)
         timespan_end = timespan_end.in_timezone(timezone.utc)
 
-        source_prefix_interp = GCSTimeSpanFileTransformOperator.interpolate_prefix(
+        self._source_prefix_interp = GCSTimeSpanFileTransformOperator.interpolate_prefix(
             self.source_prefix,
             timespan_start,
         )
-        destination_prefix_interp = GCSTimeSpanFileTransformOperator.interpolate_prefix(
+        self._destination_prefix_interp = GCSTimeSpanFileTransformOperator.interpolate_prefix(
             self.destination_prefix,
             timespan_start,
         )
@@ -828,9 +840,9 @@ class GCSTimeSpanFileTransformOperator(GoogleCloudBaseOperator):
         )
 
         # Fetch list of files.
-        self._source_object_names = source_hook.list_by_timespan(
+        blobs_to_transform = source_hook.list_by_timespan(
             bucket_name=self.source_bucket,
-            prefix=source_prefix_interp,
+            prefix=self._source_prefix_interp,
             timespan_start=timespan_start,
             timespan_end=timespan_end,
         )
@@ -840,7 +852,7 @@ class GCSTimeSpanFileTransformOperator(GoogleCloudBaseOperator):
             temp_output_dir_path = Path(temp_output_dir)
 
             # TODO: download in parallel.
-            for blob_to_transform in self._source_object_names:
+            for blob_to_transform in blobs_to_transform:
                 destination_file = temp_input_dir_path / blob_to_transform
                 destination_file.parent.mkdir(parents=True, exist_ok=True)
                 try:
@@ -877,6 +889,8 @@ class GCSTimeSpanFileTransformOperator(GoogleCloudBaseOperator):
 
             self.log.info("Transformation succeeded. Output temporarily located at %s", temp_output_dir_path)
 
+            files_uploaded = []
+
             # TODO: upload in parallel.
             for upload_file in temp_output_dir_path.glob("**/*"):
                 if upload_file.is_dir():
@@ -884,8 +898,8 @@ class GCSTimeSpanFileTransformOperator(GoogleCloudBaseOperator):
 
                 upload_file_name = str(upload_file.relative_to(temp_output_dir_path))
 
-                if self.destination_prefix is not None:
-                    upload_file_name = f"{destination_prefix_interp}/{upload_file_name}"
+                if self._destination_prefix_interp is not None:
+                    upload_file_name = f"{self._destination_prefix_interp.rstrip('/')}/{upload_file_name}"
 
                 self.log.info("Uploading file %s to %s", upload_file, upload_file_name)
 
@@ -897,35 +911,46 @@ class GCSTimeSpanFileTransformOperator(GoogleCloudBaseOperator):
                         chunk_size=self.chunk_size,
                         num_max_attempts=self.upload_num_attempts,
                     )
-                    self._destination_object_names.append(str(upload_file_name))
+                    files_uploaded.append(str(upload_file_name))
                 except GoogleCloudError:
                     if not self.upload_continue_on_fail:
                         raise
 
-            return self._destination_object_names
+            return files_uploaded
 
     def get_openlineage_facets_on_complete(self, task_instance):
-        """Implement on_complete as execute() resolves object names."""
+        """Implement on_complete as execute() resolves object prefixes."""
         from openlineage.client.run import Dataset
 
         from airflow.providers.openlineage.extractors import OperatorLineage
 
-        input_datasets = [
-            Dataset(
-                namespace=f"gs://{self.source_bucket}",
-                name=object_name,
-            )
-            for object_name in self._source_object_names
-        ]
-        output_datasets = [
-            Dataset(
-                namespace=f"gs://{self.destination_bucket}",
-                name=object_name,
-            )
-            for object_name in self._destination_object_names
-        ]
-
-        return OperatorLineage(inputs=input_datasets, outputs=output_datasets)
+        def _parse_prefix(pref):
+            # Use parent if not a file (dot not in name) and not a dir (ends with slash)
+            if "." not in pref.split("/")[-1] and not pref.endswith("/"):
+                pref = Path(pref).parent.as_posix()
+            return "/" if pref in (".", "/", "") else pref.rstrip("/")
+
+        input_prefix, output_prefix = "/", "/"
+        if self._source_prefix_interp is not None:
+            input_prefix = _parse_prefix(self._source_prefix_interp)
+
+        if self._destination_prefix_interp is not None:
+            output_prefix = _parse_prefix(self._destination_prefix_interp)
+
+        return OperatorLineage(
+            inputs=[
+                Dataset(
+                    namespace=f"gs://{self.source_bucket}",
+                    name=input_prefix,
+                )
+            ],
+            outputs=[
+                Dataset(
+                    namespace=f"gs://{self.destination_bucket}",
+                    name=output_prefix,
+                )
+            ],
+        )
 
 
 class GCSDeleteBucketOperator(GoogleCloudBaseOperator):

airflow/providers/google/cloud/operators/kubernetes_engine.py

@@ -29,13 +29,13 @@ import yaml
 from deprecated import deprecated
 from google.api_core.exceptions import AlreadyExists
 from google.cloud.container_v1.types import Cluster
-from kubernetes.client import V1JobList
+from kubernetes.client import V1JobList, models as k8s
 from kubernetes.utils.create_from_yaml import FailToCreateError
 from packaging.version import parse as parse_version
 
 from airflow.configuration import conf
 from airflow.exceptions import AirflowException, AirflowProviderDeprecationWarning
-from airflow.providers.cncf.kubernetes.operators.job import KubernetesDeleteJobOperator, KubernetesJobOperator
+from airflow.providers.cncf.kubernetes.operators.job import KubernetesJobOperator
 from airflow.providers.cncf.kubernetes.operators.pod import KubernetesPodOperator
 from airflow.providers.cncf.kubernetes.operators.resource import (
     KubernetesCreateResourceOperator,
@@ -47,6 +47,7 @@ from airflow.providers.google.cloud.hooks.kubernetes_engine import (
     GKEDeploymentHook,
     GKEHook,
     GKEJobHook,
+    GKEKubernetesHook,
     GKEPodHook,
 )
 from airflow.providers.google.cloud.links.kubernetes_engine import (
@@ -61,9 +62,20 @@ from airflow.providers.google.cloud.triggers.kubernetes_engine import (
     GKEOperationTrigger,
     GKEStartPodTrigger,
 )
+from airflow.providers.google.common.hooks.base_google import PROVIDE_PROJECT_ID
 from airflow.providers_manager import ProvidersManager
 from airflow.utils.timezone import utcnow
 
+try:
+    from airflow.providers.cncf.kubernetes.operators.job import KubernetesDeleteJobOperator
+except ImportError:
+    from airflow.exceptions import AirflowOptionalProviderFeatureException
+
+    raise AirflowOptionalProviderFeatureException(
+        "Failed to import KubernetesDeleteJobOperator. This operator is only available in cncf-kubernetes "
+        "provider version >=8.1.0"
+    )
+
 if TYPE_CHECKING:
     from kubernetes.client.models import V1Job, V1Pod
 
@@ -166,7 +178,7 @@ class GKEDeleteClusterOperator(GoogleCloudBaseOperator):
         *,
         name: str,
         location: str,
-        project_id: str | None = None,
+        project_id: str = PROVIDE_PROJECT_ID,
         gcp_conn_id: str = "google_cloud_default",
         api_version: str = "v2",
         impersonation_chain: str | Sequence[str] | None = None,
@@ -310,7 +322,7 @@ class GKECreateClusterOperator(GoogleCloudBaseOperator):
         *,
         location: str,
         body: dict | Cluster,
-        project_id: str | None = None,
+        project_id: str = PROVIDE_PROJECT_ID,
         gcp_conn_id: str = "google_cloud_default",
         api_version: str = "v2",
         impersonation_chain: str | Sequence[str] | None = None,
@@ -496,7 +508,7 @@ class GKEStartKueueInsideClusterOperator(GoogleCloudBaseOperator):
         cluster_name: str,
         kueue_version: str,
         use_internal_ip: bool = False,
-        project_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,
@@ -658,7 +670,7 @@ class GKEStartPodOperator(KubernetesPodOperator):
         location: str,
         cluster_name: str,
         use_internal_ip: bool = False,
-        project_id: str | None = None,
+        project_id: str = PROVIDE_PROJECT_ID,
         gcp_conn_id: str = "google_cloud_default",
         impersonation_chain: str | Sequence[str] | None = None,
         regional: bool | None = None,
@@ -856,7 +868,7 @@ class GKEStartJobOperator(KubernetesJobOperator):
         location: str,
         cluster_name: str,
         use_internal_ip: bool = False,
-        project_id: str | None = None,
+        project_id: str = PROVIDE_PROJECT_ID,
         gcp_conn_id: str = "google_cloud_default",
         impersonation_chain: str | Sequence[str] | None = None,
         deferrable: bool = conf.getboolean("operators", "default_deferrable", fallback=False),
@@ -993,7 +1005,7 @@ class GKEDescribeJobOperator(GoogleCloudBaseOperator):
         location: str,
         namespace: str,
         cluster_name: str,
-        project_id: str | None = None,
+        project_id: str = PROVIDE_PROJECT_ID,
         use_internal_ip: bool = False,
         gcp_conn_id: str = "google_cloud_default",
         impersonation_chain: str | Sequence[str] | None = None,
@@ -1094,7 +1106,7 @@ class GKEListJobsOperator(GoogleCloudBaseOperator):
         location: str,
         cluster_name: str,
         namespace: str | None = None,
-        project_id: str | None = None,
+        project_id: str = PROVIDE_PROJECT_ID,
         use_internal_ip: bool = False,
         do_xcom_push: bool = True,
         gcp_conn_id: str = "google_cloud_default",
@@ -1194,7 +1206,7 @@ class GKECreateCustomResourceOperator(KubernetesCreateResourceOperator):
         location: str,
         cluster_name: str,
         use_internal_ip: bool = False,
-        project_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,
@@ -1296,7 +1308,7 @@ class GKEDeleteCustomResourceOperator(KubernetesDeleteResourceOperator):
         location: str,
         cluster_name: str,
         use_internal_ip: bool = False,
-        project_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,
@@ -1435,7 +1447,7 @@ class GKEDeleteJobOperator(KubernetesDeleteJobOperator):
         location: str,
         cluster_name: str,
         use_internal_ip: bool = False,
-        project_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,
@@ -1494,3 +1506,211 @@ class GKEDeleteJobOperator(KubernetesDeleteJobOperator):
         ).fetch_cluster_info()
 
         return super().execute(context)
+
+
+class GKESuspendJobOperator(GoogleCloudBaseOperator):
+    """
+    Suspend Job by given name.
+
+    .. seealso::
+        For more information on how to use this operator, take a look at the guide:
+        :ref:`howto/operator:GKESuspendJobOperator`
+
+    :param name: The name of the Job to suspend
+    :param project_id: The Google Developers Console project id.
+    :param location: The name of the Google Kubernetes Engine zone or region in which the cluster
+        resides.
+    :param cluster_name: The name of the Google Kubernetes Engine cluster.
+    :param namespace: The name of the Google Kubernetes Engine namespace.
+    :param use_internal_ip: Use the internal IP address as the endpoint.
+    :param gcp_conn_id: The connection ID to use connecting 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).
+    """
+
+    template_fields: Sequence[str] = (
+        "project_id",
+        "gcp_conn_id",
+        "name",
+        "namespace",
+        "cluster_name",
+        "location",
+        "impersonation_chain",
+    )
+    operator_extra_links = (KubernetesEngineJobLink(),)
+
+    def __init__(
+        self,
+        *,
+        name: str,
+        location: str,
+        namespace: str,
+        cluster_name: str,
+        project_id: str = PROVIDE_PROJECT_ID,
+        use_internal_ip: bool = False,
+        gcp_conn_id: str = "google_cloud_default",
+        impersonation_chain: str | Sequence[str] | None = None,
+        **kwargs,
+    ) -> None:
+        super().__init__(**kwargs)
+
+        self.project_id = project_id
+        self.gcp_conn_id = gcp_conn_id
+        self.location = location
+        self.name = name
+        self.namespace = namespace
+        self.cluster_name = cluster_name
+        self.use_internal_ip = use_internal_ip
+        self.impersonation_chain = impersonation_chain
+
+        self.job: V1Job | None = None
+        self._ssl_ca_cert: str
+        self._cluster_url: str
+
+    @cached_property
+    def cluster_hook(self) -> GKEHook:
+        return GKEHook(
+            gcp_conn_id=self.gcp_conn_id,
+            location=self.location,
+            impersonation_chain=self.impersonation_chain,
+        )
+
+    @cached_property
+    def hook(self) -> GKEKubernetesHook:
+        self._cluster_url, self._ssl_ca_cert = GKEClusterAuthDetails(
+            cluster_name=self.cluster_name,
+            project_id=self.project_id,
+            use_internal_ip=self.use_internal_ip,
+            cluster_hook=self.cluster_hook,
+        ).fetch_cluster_info()
+
+        return GKEKubernetesHook(
+            gcp_conn_id=self.gcp_conn_id,
+            cluster_url=self._cluster_url,
+            ssl_ca_cert=self._ssl_ca_cert,
+        )
+
+    def execute(self, context: Context) -> None:
+        self.job = self.hook.patch_namespaced_job(
+            job_name=self.name,
+            namespace=self.namespace,
+            body={"spec": {"suspend": True}},
+        )
+        self.log.info(
+            "Job %s from cluster %s was suspended.",
+            self.name,
+            self.cluster_name,
+        )
+        KubernetesEngineJobLink.persist(context=context, task_instance=self)
+
+        return k8s.V1Job.to_dict(self.job)
+
+
+class GKEResumeJobOperator(GoogleCloudBaseOperator):
+    """
+    Resume Job by given name.
+
+    .. seealso::
+        For more information on how to use this operator, take a look at the guide:
+        :ref:`howto/operator:GKEResumeJobOperator`
+
+    :param name: The name of the Job to resume
+    :param project_id: The Google Developers Console project id.
+    :param location: The name of the Google Kubernetes Engine zone or region in which the cluster
+        resides.
+    :param cluster_name: The name of the Google Kubernetes Engine cluster.
+    :param namespace: The name of the Google Kubernetes Engine namespace.
+    :param use_internal_ip: Use the internal IP address as the endpoint.
+    :param gcp_conn_id: The connection ID to use connecting 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).
+    """
+
+    template_fields: Sequence[str] = (
+        "project_id",
+        "gcp_conn_id",
+        "name",
+        "namespace",
+        "cluster_name",
+        "location",
+        "impersonation_chain",
+    )
+    operator_extra_links = (KubernetesEngineJobLink(),)
+
+    def __init__(
+        self,
+        *,
+        name: str,
+        location: str,
+        namespace: str,
+        cluster_name: str,
+        project_id: str = PROVIDE_PROJECT_ID,
+        use_internal_ip: bool = False,
+        gcp_conn_id: str = "google_cloud_default",
+        impersonation_chain: str | Sequence[str] | None = None,
+        **kwargs,
+    ) -> None:
+        super().__init__(**kwargs)
+
+        self.project_id = project_id
+        self.gcp_conn_id = gcp_conn_id
+        self.location = location
+        self.name = name
+        self.namespace = namespace
+        self.cluster_name = cluster_name
+        self.use_internal_ip = use_internal_ip
+        self.impersonation_chain = impersonation_chain
+
+        self.job: V1Job | None = None
+        self._ssl_ca_cert: str
+        self._cluster_url: str
+
+    @cached_property
+    def cluster_hook(self) -> GKEHook:
+        return GKEHook(
+            gcp_conn_id=self.gcp_conn_id,
+            location=self.location,
+            impersonation_chain=self.impersonation_chain,
+        )
+
+    @cached_property
+    def hook(self) -> GKEKubernetesHook:
+        self._cluster_url, self._ssl_ca_cert = GKEClusterAuthDetails(
+            cluster_name=self.cluster_name,
+            project_id=self.project_id,
+            use_internal_ip=self.use_internal_ip,
+            cluster_hook=self.cluster_hook,
+        ).fetch_cluster_info()
+
+        return GKEKubernetesHook(
+            gcp_conn_id=self.gcp_conn_id,
+            cluster_url=self._cluster_url,
+            ssl_ca_cert=self._ssl_ca_cert,
+        )
+
+    def execute(self, context: Context) -> None:
+        self.job = self.hook.patch_namespaced_job(
+            job_name=self.name,
+            namespace=self.namespace,
+            body={"spec": {"suspend": False}},
+        )
+        self.log.info(
+            "Job %s from cluster %s was resumed.",
+            self.name,
+            self.cluster_name,
+        )
+        KubernetesEngineJobLink.persist(context=context, task_instance=self)
+
+        return k8s.V1Job.to_dict(self.job)

airflow/providers/google/cloud/operators/life_sciences.py

@@ -27,6 +27,7 @@ from airflow.exceptions import AirflowException, AirflowProviderDeprecationWarni
 from airflow.providers.google.cloud.hooks.life_sciences import LifeSciencesHook
 from airflow.providers.google.cloud.links.life_sciences import LifeSciencesLink
 from airflow.providers.google.cloud.operators.cloud_base import GoogleCloudBaseOperator
+from airflow.providers.google.common.hooks.base_google import PROVIDE_PROJECT_ID
 
 if TYPE_CHECKING:
     from airflow.utils.context import Context
@@ -82,7 +83,7 @@ class LifeSciencesRunPipelineOperator(GoogleCloudBaseOperator):
         *,
         body: dict,
         location: str,
-        project_id: str | None = None,
+        project_id: str = PROVIDE_PROJECT_ID,
         gcp_conn_id: str = "google_cloud_default",
         api_version: str = "v2beta",
         impersonation_chain: str | Sequence[str] | None = None,

airflow/providers/google/cloud/operators/mlengine.py

@@ -39,6 +39,7 @@ from airflow.providers.google.cloud.links.mlengine import (
 )
 from airflow.providers.google.cloud.operators.cloud_base import GoogleCloudBaseOperator
 from airflow.providers.google.cloud.triggers.mlengine import MLEngineStartTrainingJobTrigger
+from airflow.providers.google.common.hooks.base_google import PROVIDE_PROJECT_ID
 
 if TYPE_CHECKING:
     from airflow.utils.context import Context
@@ -345,7 +346,7 @@ class MLEngineManageModelOperator(GoogleCloudBaseOperator):
         *,
         model: dict,
         operation: str = "create",
-        project_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,
@@ -417,7 +418,7 @@ class MLEngineCreateModelOperator(GoogleCloudBaseOperator):
         self,
         *,
         model: dict,
-        project_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,
@@ -493,7 +494,7 @@ class MLEngineGetModelOperator(GoogleCloudBaseOperator):
         self,
         *,
         model_name: str,
-        project_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,
@@ -573,7 +574,7 @@ class MLEngineDeleteModelOperator(GoogleCloudBaseOperator):
         *,
         model_name: str,
         delete_contents: bool = False,
-        project_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,
@@ -680,7 +681,7 @@ class MLEngineManageVersionOperator(GoogleCloudBaseOperator):
         version_name: str | None = None,
         version: dict | None = None,
         operation: str = "create",
-        project_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,
@@ -773,7 +774,7 @@ class MLEngineCreateVersionOperator(GoogleCloudBaseOperator):
         *,
         model_name: str,
         version: dict,
-        project_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,
@@ -866,7 +867,7 @@ class MLEngineSetDefaultVersionOperator(GoogleCloudBaseOperator):
         *,
         model_name: str,
         version_name: str,
-        project_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,
@@ -956,7 +957,7 @@ class MLEngineListVersionsOperator(GoogleCloudBaseOperator):
         self,
         *,
         model_name: str,
-        project_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,
@@ -1045,7 +1046,7 @@ class MLEngineDeleteVersionOperator(GoogleCloudBaseOperator):
         *,
         model_name: str,
         version_name: str,
-        project_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,
@@ -1467,7 +1468,7 @@ class MLEngineTrainingCancelJobOperator(GoogleCloudBaseOperator):
         self,
         *,
         job_id: str,
-        project_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,