This is an automated email from the ASF dual-hosted git repository.
shahar1 pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/airflow.git
The following commit(s) were added to refs/heads/main by this push:
new 5bfc14fcfc2 Validate Google operator templated parameters after
rendering (#70534)
5bfc14fcfc2 is described below
commit 5bfc14fcfc2020e393ef6c1ec7c62a159c6a62ba
Author: Dr Alex Mitre <[email protected]>
AuthorDate: Tue Sep 15 13:55:09 2026 -0600
Validate Google operator templated parameters after rendering (#70534)
CloudSpeechToTextRecognizeSpeechOperator,
CloudTextToSpeechSynthesizeOperator,
and CloudFirestoreExportDatabaseOperator validate template fields inside
_validate_inputs() called from __init__, so the checks ran against
un-rendered Jinja expressions and an empty rendered value was never
caught. The validation now runs at execute time, and the raises are
narrowed from AirflowException to ValueError per the ongoing exception
clean-up.
---
generated/known_airflow_exceptions.txt | 3 ---
.../providers/google/cloud/operators/speech_to_text.py | 7 +++----
.../providers/google/cloud/operators/text_to_speech.py | 5 ++---
.../providers/google/firebase/operators/firestore.py | 5 ++---
.../unit/google/cloud/operators/test_speech_to_text.py | 16 ++++++++++++++++
.../unit/google/cloud/operators/test_text_to_speech.py | 3 +--
.../unit/google/firebase/operators/test_firestore.py | 17 +++++++++++++++++
7 files changed, 41 insertions(+), 15 deletions(-)
diff --git a/generated/known_airflow_exceptions.txt
b/generated/known_airflow_exceptions.txt
index c60aa6a73bd..d1b06ce69e2 100644
--- a/generated/known_airflow_exceptions.txt
+++ b/generated/known_airflow_exceptions.txt
@@ -264,8 +264,6 @@
providers/google/src/airflow/providers/google/cloud/operators/looker.py::1
providers/google/src/airflow/providers/google/cloud/operators/managed_kafka.py::15
providers/google/src/airflow/providers/google/cloud/operators/pubsub.py::1
providers/google/src/airflow/providers/google/cloud/operators/spanner.py::19
-providers/google/src/airflow/providers/google/cloud/operators/speech_to_text.py::2
-providers/google/src/airflow/providers/google/cloud/operators/text_to_speech.py::1
providers/google/src/airflow/providers/google/cloud/operators/translate.py::9
providers/google/src/airflow/providers/google/cloud/operators/translate_speech.py::2
providers/google/src/airflow/providers/google/cloud/operators/vertex_ai/batch_prediction_job.py::1
@@ -311,7 +309,6 @@
providers/google/src/airflow/providers/google/cloud/utils/credentials_provider.p
providers/google/src/airflow/providers/google/common/hooks/base_google.py::7
providers/google/src/airflow/providers/google/common/hooks/operation_helpers.py::2
providers/google/src/airflow/providers/google/firebase/hooks/firestore.py::1
-providers/google/src/airflow/providers/google/firebase/operators/firestore.py::1
providers/google/src/airflow/providers/google/leveldb/hooks/leveldb.py::4
providers/google/src/airflow/providers/google/marketing_platform/hooks/campaign_manager.py::2
providers/google/src/airflow/providers/google/marketing_platform/hooks/search_ads.py::1
diff --git
a/providers/google/src/airflow/providers/google/cloud/operators/speech_to_text.py
b/providers/google/src/airflow/providers/google/cloud/operators/speech_to_text.py
index 5424fba7191..3ac73c54abf 100644
---
a/providers/google/src/airflow/providers/google/cloud/operators/speech_to_text.py
+++
b/providers/google/src/airflow/providers/google/cloud/operators/speech_to_text.py
@@ -25,7 +25,6 @@ from typing import TYPE_CHECKING
from google.api_core.gapic_v1.method import DEFAULT, _MethodDefault
from google.protobuf.json_format import MessageToDict
-from airflow.providers.common.compat.sdk import AirflowException
from airflow.providers.google.cloud.hooks.speech_to_text import
CloudSpeechToTextHook, RecognitionAudio
from airflow.providers.google.cloud.operators.cloud_base import
GoogleCloudBaseOperator
from airflow.providers.google.common.hooks.base_google import
PROVIDE_PROJECT_ID
@@ -99,17 +98,17 @@ class
CloudSpeechToTextRecognizeSpeechOperator(GoogleCloudBaseOperator):
self.gcp_conn_id = gcp_conn_id
self.retry = retry
self.timeout = timeout
- self._validate_inputs()
self.impersonation_chain = impersonation_chain
super().__init__(**kwargs)
def _validate_inputs(self) -> None:
if self.audio == "":
- raise AirflowException("The required parameter 'audio' is empty")
+ raise ValueError("The required parameter 'audio' is empty")
if self.config == "":
- raise AirflowException("The required parameter 'config' is empty")
+ raise ValueError("The required parameter 'config' is empty")
def execute(self, context: Context):
+ self._validate_inputs()
hook = CloudSpeechToTextHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
diff --git
a/providers/google/src/airflow/providers/google/cloud/operators/text_to_speech.py
b/providers/google/src/airflow/providers/google/cloud/operators/text_to_speech.py
index 5deded3ef8b..bfd01409082 100644
---
a/providers/google/src/airflow/providers/google/cloud/operators/text_to_speech.py
+++
b/providers/google/src/airflow/providers/google/cloud/operators/text_to_speech.py
@@ -25,7 +25,6 @@ from typing import TYPE_CHECKING
from google.api_core.gapic_v1.method import DEFAULT, _MethodDefault
-from airflow.providers.common.compat.sdk import AirflowException
from airflow.providers.google.cloud.hooks.gcs import GCSHook
from airflow.providers.google.cloud.hooks.text_to_speech import
CloudTextToSpeechHook
from airflow.providers.google.cloud.operators.cloud_base import
GoogleCloudBaseOperator
@@ -112,7 +111,6 @@ class
CloudTextToSpeechSynthesizeOperator(GoogleCloudBaseOperator):
self.gcp_conn_id = gcp_conn_id
self.retry = retry
self.timeout = timeout
- self._validate_inputs()
self.impersonation_chain = impersonation_chain
super().__init__(**kwargs)
@@ -125,9 +123,10 @@ class
CloudTextToSpeechSynthesizeOperator(GoogleCloudBaseOperator):
"target_filename",
]:
if getattr(self, parameter) == "":
- raise AirflowException(f"The required parameter '{parameter}'
is empty")
+ raise ValueError(f"The required parameter '{parameter}' is
empty")
def execute(self, context: Context) -> None:
+ self._validate_inputs()
hook = CloudTextToSpeechHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
diff --git
a/providers/google/src/airflow/providers/google/firebase/operators/firestore.py
b/providers/google/src/airflow/providers/google/firebase/operators/firestore.py
index 0c35f572923..d39f4bf879b 100644
---
a/providers/google/src/airflow/providers/google/firebase/operators/firestore.py
+++
b/providers/google/src/airflow/providers/google/firebase/operators/firestore.py
@@ -19,7 +19,6 @@ from __future__ import annotations
from collections.abc import Sequence
from typing import TYPE_CHECKING
-from airflow.providers.common.compat.sdk import AirflowException
from airflow.providers.google.common.hooks.base_google import
PROVIDE_PROJECT_ID
from airflow.providers.google.firebase.hooks.firestore import
CloudFirestoreHook
from airflow.providers.google.version_compat import BaseOperator
@@ -78,14 +77,14 @@ class CloudFirestoreExportDatabaseOperator(BaseOperator):
self.project_id = project_id
self.gcp_conn_id = gcp_conn_id
self.api_version = api_version
- self._validate_inputs()
self.impersonation_chain = impersonation_chain
def _validate_inputs(self) -> None:
if not self.body:
- raise AirflowException("The required parameter 'body' is missing")
+ raise ValueError("The required parameter 'body' is missing")
def execute(self, context: Context):
+ self._validate_inputs()
hook = CloudFirestoreHook(
gcp_conn_id=self.gcp_conn_id,
api_version=self.api_version,
diff --git
a/providers/google/tests/unit/google/cloud/operators/test_speech_to_text.py
b/providers/google/tests/unit/google/cloud/operators/test_speech_to_text.py
index 6e515a3d45d..c2e0ad737cd 100644
--- a/providers/google/tests/unit/google/cloud/operators/test_speech_to_text.py
+++ b/providers/google/tests/unit/google/cloud/operators/test_speech_to_text.py
@@ -55,6 +55,22 @@ class TestCloudSpeechToTextRecognizeSpeechOperator:
config=CONFIG, audio=AUDIO, retry=DEFAULT, timeout=None
)
+
@patch("airflow.providers.google.cloud.operators.speech_to_text.CloudSpeechToTextHook")
+ def test_empty_audio_fails_at_execute_time(self, mock_hook):
+ op = CloudSpeechToTextRecognizeSpeechOperator(
+ project_id=PROJECT_ID,
+ gcp_conn_id=GCP_CONN_ID,
+ audio="{{ var.value.audio }}",
+ config=CONFIG,
+ task_id="id",
+ )
+ # Template rendering replaces the Jinja expression with the resolved
value before execute.
+ op.audio = ""
+
+ with pytest.raises(ValueError, match="The required parameter 'audio'
is empty"):
+ op.execute(context={"task_instance": Mock()})
+ mock_hook.assert_not_called()
+
@patch("airflow.providers.google.cloud.operators.speech_to_text.CloudSpeechToTextHook")
def test_missing_config(self, mock_hook):
mock_hook.return_value.recognize_speech.return_value = True
diff --git
a/providers/google/tests/unit/google/cloud/operators/test_text_to_speech.py
b/providers/google/tests/unit/google/cloud/operators/test_text_to_speech.py
index d4c2a597870..76c53c5c3f4 100644
--- a/providers/google/tests/unit/google/cloud/operators/test_text_to_speech.py
+++ b/providers/google/tests/unit/google/cloud/operators/test_text_to_speech.py
@@ -22,7 +22,6 @@ from unittest.mock import ANY, MagicMock, Mock, PropertyMock,
patch
import pytest
from google.api_core.gapic_v1.method import DEFAULT
-from airflow.providers.common.compat.sdk import AirflowException
from airflow.providers.google.cloud.operators.text_to_speech import
CloudTextToSpeechSynthesizeOperator
PROJECT_ID = "project-id"
@@ -98,7 +97,7 @@ class TestGcpTextToSpeech:
):
mocked_context = Mock()
- with pytest.raises(AirflowException) as ctx:
+ with pytest.raises(ValueError, match="is empty") as ctx:
CloudTextToSpeechSynthesizeOperator(
project_id="project-id",
input_data=input_data,
diff --git
a/providers/google/tests/unit/google/firebase/operators/test_firestore.py
b/providers/google/tests/unit/google/firebase/operators/test_firestore.py
index 340058ef6eb..e9fd938b017 100644
--- a/providers/google/tests/unit/google/firebase/operators/test_firestore.py
+++ b/providers/google/tests/unit/google/firebase/operators/test_firestore.py
@@ -18,6 +18,8 @@ from __future__ import annotations
from unittest import mock
+import pytest
+
from airflow.providers.google.firebase.operators.firestore import
CloudFirestoreExportDatabaseOperator
TEST_OUTPUT_URI_PREFIX: str = "gs://example-bucket/path"
@@ -42,3 +44,18 @@ class TestCloudFirestoreExportDatabaseOperator:
mock_firestore_hook.return_value.export_documents.assert_called_once_with(
body=EXPORT_DOCUMENT_BODY, database_id="(default)",
project_id=TEST_PROJECT_ID
)
+
+
@mock.patch("airflow.providers.google.firebase.operators.firestore.CloudFirestoreHook")
+ def test_empty_body_fails_at_execute_time(self, mock_firestore_hook):
+ op = CloudFirestoreExportDatabaseOperator(
+ task_id="test-task",
+ body="{{ var.value.export_body }}",
+ gcp_conn_id="google_cloud_default",
+ project_id=TEST_PROJECT_ID,
+ )
+ # Template rendering replaces the Jinja expression with the resolved
value before execute.
+ op.body = None
+
+ with pytest.raises(ValueError, match="The required parameter 'body' is
missing"):
+ op.execute(mock.MagicMock())
+ mock_firestore_hook.return_value.export_documents.assert_not_called()
