ashb commented on a change in pull request #14709: URL: https://github.com/apache/airflow/pull/14709#discussion_r592413860
########## File path: airflow/decorators/base.py ########## @@ -0,0 +1,191 @@ +# 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. + +import functools +import inspect +import re +from inspect import signature +from typing import Any, Callable, Dict, Optional, Tuple, TypeVar, cast + +from airflow.exceptions import AirflowException +from airflow.models import BaseOperator +from airflow.models.dag import DAG, DagContext +from airflow.models.xcom_arg import XComArg +from airflow.utils.decorators import apply_defaults +from airflow.utils.task_group import TaskGroup, TaskGroupContext + + +class BaseDecoratedOperator(BaseOperator): + """ + Wraps a Python callable and captures args/kwargs when called for execution. + + :param python_callable: A reference to an object that is callable + :type python_callable: python callable + :param op_kwargs: a dictionary of keyword arguments that will get unpacked + in your function (templated) + :type op_kwargs: dict + :param op_args: a list of positional arguments that will get unpacked when + calling your callable (templated) + :type op_args: list + :param multiple_outputs: if set, function return value will be + unrolled to multiple XCom values. Dict will unroll to xcom values with keys as keys. + Defaults to False. + :type multiple_outputs: bool + """ + + template_fields = ('op_args', 'op_kwargs') + template_fields_renderers = {"op_args": "py", "op_kwargs": "py"} + + # since we won't mutate the arguments, we should just do the shallow copy + # there are some cases we can't deepcopy the objects (e.g protobuf). + shallow_copy_attrs = ('python_callable',) + + @apply_defaults + def __init__( + self, + *, + python_callable: Callable, + task_id: str, + op_args: Tuple[Any], + op_kwargs: Dict[str, Any], + multiple_outputs: bool = False, + **kwargs, + ) -> None: + kwargs['task_id'] = self._get_unique_task_id(task_id, kwargs.get('dag'), kwargs.get('task_group')) + super().__init__(**kwargs) + self.python_callable = python_callable + + # Check that arguments can be binded + signature(python_callable).bind(*op_args, **op_kwargs) + self.multiple_outputs = multiple_outputs + self.op_args = op_args + self.op_kwargs = op_kwargs + + @staticmethod + def _get_unique_task_id( + task_id: str, dag: Optional[DAG] = None, task_group: Optional[TaskGroup] = None + ) -> str: + """ + Generate unique task id given a DAG (or if run in a DAG context) + Ids are generated by appending a unique number to the end of + the original task id. + + Example: + task_id + task_id__1 + task_id__2 + ... + task_id__20 + """ + dag = dag or DagContext.get_current_dag() + if not dag: + return task_id + + # We need to check if we are in the context of TaskGroup as the task_id may + # already be altered + task_group = task_group or TaskGroupContext.get_current_task_group(dag) + tg_task_id = task_group.child_id(task_id) if task_group else task_id + + if tg_task_id not in dag.task_ids: + return task_id + core = re.split(r'__\d+$', task_id)[0] + suffixes = sorted( + [ + int(re.split(r'^.+__', task_id)[1]) + for task_id in dag.task_ids + if re.match(rf'^{core}__\d+$', task_id) + ] + ) + if not suffixes: + return f'{core}__1' + return f'{core}__{suffixes[-1] + 1}' + + @staticmethod + def validate_python_callable(python_callable): + """ + Validate that python callable can be wrapped by operator. + Raises exception if invalid. + + :param python_callable: Python object to be validated + :raises: TypeError, AirflowException + """ + if not callable(python_callable): + raise TypeError('`python_callable` param must be callable') + if 'self' in signature(python_callable).parameters.keys(): + raise AirflowException('@task does not support methods') + + def execute(self, context: Dict): + raise NotImplementedError() + + +T = TypeVar("T", bound=Callable) # pylint: disable=invalid-name + + +def base_task( + python_callable: Optional[Callable] = None, + multiple_outputs: Optional[bool] = None, + impl_class=None, + **kwargs, +) -> Callable[[T], T]: + """ + Python operator decorator. Wraps a function into an Airflow operator. Review comment: This doc string should be updated -- it's still written as if it's for a DAG author, but this is now intended to be used by developers/internal. ########## File path: airflow/decorators/python.py ########## @@ -0,0 +1,112 @@ +# 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. + +import functools +from typing import Callable, Dict, Optional, TypeVar + +from airflow.decorators.base import BaseDecoratedOperator, base_task +from airflow.exceptions import AirflowException +from airflow.utils.decorators import apply_defaults + +PYTHON_OPERATOR_UI_COLOR = '#ffefeb' + + +class _PythonDecoratedOperator(BaseDecoratedOperator): + """ + Wraps a Python callable and captures args/kwargs when called for execution. + + :param python_callable: A reference to an object that is callable + :type python_callable: python callable + :param op_kwargs: a dictionary of keyword arguments that will get unpacked + in your function (templated) + :type op_kwargs: dict + :param op_args: a list of positional arguments that will get unpacked when + calling your callable (templated) + :type op_args: list + :param multiple_outputs: if set, function return value will be + unrolled to multiple XCom values. Dict will unroll to xcom values with keys as keys. + Defaults to False. + :type multiple_outputs: bool + """ + + template_fields = ('op_args', 'op_kwargs') + template_fields_renderers = {"op_args": "py", "op_kwargs": "py"} + + ui_color = PYTHON_OPERATOR_UI_COLOR + + # since we won't mutate the arguments, we should just do the shallow copy + # there are some cases we can't deepcopy the objects (e.g protobuf). + shallow_copy_attrs = ('python_callable',) + + @apply_defaults + def __init__( + self, + **kwargs, + ) -> None: + super().__init__(**kwargs) + + def execute(self, context: Dict): + return_value = self.python_callable(*self.op_args, **self.op_kwargs) + self.log.debug("Done. Returned value was: %s", return_value) + if not self.multiple_outputs: + return return_value + if isinstance(return_value, dict): + for key in return_value.keys(): + if not isinstance(key, str): + raise AirflowException( + 'Returned dictionary keys must be strings when using ' + f'multiple_outputs, found {key} ({type(key)}) instead' + ) + for key, value in return_value.items(): + self.xcom_push(context, key, value) + else: + raise AirflowException( + f'Returned output was type {type(return_value)} expected dictionary ' 'for multiple_outputs' + ) + return return_value + + +T = TypeVar("T", bound=Callable) # pylint: disable=invalid-name + + +def task(python_callable: Optional[Callable] = None, multiple_outputs: Optional[bool] = None, **kwargs): + """ + Python operator decorator. Wraps a function into an Airflow operator. + Accepts kwargs for operator kwarg. Can be reused in a single DAG. + :param python_callable: Function to decorate + :type python_callable: Optional[Callable] + :param multiple_outputs: if set, function return value will be + unrolled to multiple XCom values. List/Tuples will unroll to xcom values + with index as key. Dict will unroll to xcom values with keys as XCom keys. + Defaults to False. + :type multiple_outputs: bool + """ + + @functools.wraps(python_callable) + def wrapper(f): + return base_task( + python_callable=f, + multiple_outputs=multiple_outputs, + impl_class=_PythonDecoratedOperator, + **kwargs, + ) + + if callable(python_callable): + return wrapper(python_callable) + elif python_callable is not None: + raise AirflowException('No args allowed while using @task, use kwargs instead') + return wrapper Review comment: Hmm, I wonder if we could avoid the extra wrapper here? (Not critical, just having an extra pointless wrapper would be nice) ########## File path: airflow/operators/python.py ########## @@ -15,32 +15,32 @@ # KIND, either express or implied. See the License for the # specific language governing permissions and limitations # under the License. -import functools import inspect import os import pickle -import re import sys import types import warnings -from inspect import signature from tempfile import TemporaryDirectory from textwrap import dedent -from typing import Any, Callable, Dict, Iterable, List, Optional, Tuple, TypeVar, Union, cast +from typing import Any, Callable, Dict, Iterable, List, Optional, Union import dill +# To maintain backwards compatibility, we import the task object into this file +# This prevents breakages in dags that use `from airflow.operators.python import task` +# pylint: disable=unused-import Review comment: This disables it for the whole file, not just the next line. ########## File path: airflow/decorators/python.py ########## @@ -0,0 +1,112 @@ +# 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. + +import functools +from typing import Callable, Dict, Optional, TypeVar + +from airflow.decorators.base import BaseDecoratedOperator, base_task +from airflow.exceptions import AirflowException +from airflow.utils.decorators import apply_defaults + +PYTHON_OPERATOR_UI_COLOR = '#ffefeb' + + +class _PythonDecoratedOperator(BaseDecoratedOperator): + """ + Wraps a Python callable and captures args/kwargs when called for execution. + + :param python_callable: A reference to an object that is callable + :type python_callable: python callable + :param op_kwargs: a dictionary of keyword arguments that will get unpacked + in your function (templated) + :type op_kwargs: dict + :param op_args: a list of positional arguments that will get unpacked when + calling your callable (templated) + :type op_args: list + :param multiple_outputs: if set, function return value will be + unrolled to multiple XCom values. Dict will unroll to xcom values with keys as keys. + Defaults to False. + :type multiple_outputs: bool + """ + + template_fields = ('op_args', 'op_kwargs') + template_fields_renderers = {"op_args": "py", "op_kwargs": "py"} + + ui_color = PYTHON_OPERATOR_UI_COLOR + + # since we won't mutate the arguments, we should just do the shallow copy + # there are some cases we can't deepcopy the objects (e.g protobuf). + shallow_copy_attrs = ('python_callable',) + + @apply_defaults + def __init__( + self, + **kwargs, + ) -> None: + super().__init__(**kwargs) + + def execute(self, context: Dict): + return_value = self.python_callable(*self.op_args, **self.op_kwargs) + self.log.debug("Done. Returned value was: %s", return_value) + if not self.multiple_outputs: + return return_value + if isinstance(return_value, dict): + for key in return_value.keys(): + if not isinstance(key, str): + raise AirflowException( + 'Returned dictionary keys must be strings when using ' + f'multiple_outputs, found {key} ({type(key)}) instead' + ) + for key, value in return_value.items(): + self.xcom_push(context, key, value) + else: + raise AirflowException( + f'Returned output was type {type(return_value)} expected dictionary ' 'for multiple_outputs' + ) + return return_value + + +T = TypeVar("T", bound=Callable) # pylint: disable=invalid-name + + +def task(python_callable: Optional[Callable] = None, multiple_outputs: Optional[bool] = None, **kwargs): + """ + Python operator decorator. Wraps a function into an Airflow operator. + Accepts kwargs for operator kwarg. Can be reused in a single DAG. + :param python_callable: Function to decorate + :type python_callable: Optional[Callable] + :param multiple_outputs: if set, function return value will be + unrolled to multiple XCom values. List/Tuples will unroll to xcom values + with index as key. Dict will unroll to xcom values with keys as XCom keys. + Defaults to False. + :type multiple_outputs: bool + """ + + @functools.wraps(python_callable) + def wrapper(f): + return base_task( + python_callable=f, + multiple_outputs=multiple_outputs, + impl_class=_PythonDecoratedOperator, + **kwargs, + ) + + if callable(python_callable): + return wrapper(python_callable) + elif python_callable is not None: + raise AirflowException('No args allowed while using @task, use kwargs instead') + return wrapper Review comment: I think this can just be: ``` def task(python_callable: Optional[Callable] = None, multiple_outputs: Optional[bool] = None, **kwargs): return base_task( python_callable=f, multiple_outputs=multiple_outputs, impl_class=_PythonDecoratedOperator, **kwargs, ) ``` ########## File path: airflow/operators/python.py ########## @@ -15,32 +15,32 @@ # KIND, either express or implied. See the License for the # specific language governing permissions and limitations # under the License. -import functools import inspect import os import pickle -import re import sys import types import warnings -from inspect import signature from tempfile import TemporaryDirectory from textwrap import dedent -from typing import Any, Callable, Dict, Iterable, List, Optional, Tuple, TypeVar, Union, cast +from typing import Any, Callable, Dict, Iterable, List, Optional, Union import dill +# To maintain backwards compatibility, we import the task object into this file +# This prevents breakages in dags that use `from airflow.operators.python import task` +# pylint: disable=unused-import +from airflow.decorators.python import task # noqa from airflow.exceptions import AirflowException from airflow.models import BaseOperator -from airflow.models.dag import DAG, DagContext from airflow.models.skipmixin import SkipMixin from airflow.models.taskinstance import _CURRENT_CONTEXT -from airflow.models.xcom_arg import XComArg from airflow.utils.decorators import apply_defaults from airflow.utils.operator_helpers import determine_kwargs from airflow.utils.process_utils import execute_in_subprocess from airflow.utils.python_virtualenv import prepare_virtualenv, write_python_script -from airflow.utils.task_group import TaskGroup, TaskGroupContext + +PYTHON_OPERATOR_UI_COLOR = '#ffefeb' Review comment: ? ########## File path: airflow/decorators/base.py ########## @@ -0,0 +1,191 @@ +# 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. + +import functools +import inspect +import re +from inspect import signature +from typing import Any, Callable, Dict, Optional, Tuple, TypeVar, cast + +from airflow.exceptions import AirflowException +from airflow.models import BaseOperator +from airflow.models.dag import DAG, DagContext +from airflow.models.xcom_arg import XComArg +from airflow.utils.decorators import apply_defaults +from airflow.utils.task_group import TaskGroup, TaskGroupContext + + +class BaseDecoratedOperator(BaseOperator): + """ + Wraps a Python callable and captures args/kwargs when called for execution. + + :param python_callable: A reference to an object that is callable + :type python_callable: python callable + :param op_kwargs: a dictionary of keyword arguments that will get unpacked + in your function (templated) + :type op_kwargs: dict + :param op_args: a list of positional arguments that will get unpacked when + calling your callable (templated) + :type op_args: list + :param multiple_outputs: if set, function return value will be + unrolled to multiple XCom values. Dict will unroll to xcom values with keys as keys. + Defaults to False. + :type multiple_outputs: bool + """ + + template_fields = ('op_args', 'op_kwargs') + template_fields_renderers = {"op_args": "py", "op_kwargs": "py"} + + # since we won't mutate the arguments, we should just do the shallow copy + # there are some cases we can't deepcopy the objects (e.g protobuf). + shallow_copy_attrs = ('python_callable',) + + @apply_defaults + def __init__( + self, + *, + python_callable: Callable, + task_id: str, + op_args: Tuple[Any], + op_kwargs: Dict[str, Any], + multiple_outputs: bool = False, + **kwargs, + ) -> None: + kwargs['task_id'] = self._get_unique_task_id(task_id, kwargs.get('dag'), kwargs.get('task_group')) + super().__init__(**kwargs) + self.python_callable = python_callable + + # Check that arguments can be binded + signature(python_callable).bind(*op_args, **op_kwargs) + self.multiple_outputs = multiple_outputs + self.op_args = op_args + self.op_kwargs = op_kwargs + + @staticmethod + def _get_unique_task_id( + task_id: str, dag: Optional[DAG] = None, task_group: Optional[TaskGroup] = None + ) -> str: + """ + Generate unique task id given a DAG (or if run in a DAG context) + Ids are generated by appending a unique number to the end of + the original task id. + + Example: + task_id + task_id__1 + task_id__2 + ... + task_id__20 + """ + dag = dag or DagContext.get_current_dag() + if not dag: + return task_id + + # We need to check if we are in the context of TaskGroup as the task_id may + # already be altered + task_group = task_group or TaskGroupContext.get_current_task_group(dag) + tg_task_id = task_group.child_id(task_id) if task_group else task_id + + if tg_task_id not in dag.task_ids: + return task_id + core = re.split(r'__\d+$', task_id)[0] + suffixes = sorted( + [ + int(re.split(r'^.+__', task_id)[1]) + for task_id in dag.task_ids + if re.match(rf'^{core}__\d+$', task_id) + ] + ) + if not suffixes: + return f'{core}__1' + return f'{core}__{suffixes[-1] + 1}' + + @staticmethod + def validate_python_callable(python_callable): + """ + Validate that python callable can be wrapped by operator. + Raises exception if invalid. + + :param python_callable: Python object to be validated + :raises: TypeError, AirflowException + """ + if not callable(python_callable): + raise TypeError('`python_callable` param must be callable') + if 'self' in signature(python_callable).parameters.keys(): + raise AirflowException('@task does not support methods') + + def execute(self, context: Dict): + raise NotImplementedError() + + +T = TypeVar("T", bound=Callable) # pylint: disable=invalid-name + + +def base_task( + python_callable: Optional[Callable] = None, + multiple_outputs: Optional[bool] = None, + impl_class=None, + **kwargs, +) -> Callable[[T], T]: + """ + Python operator decorator. Wraps a function into an Airflow operator. Review comment: How about calling this `task_decorator_factory`? I think that make it clearer that this what this fn is used for. ########## File path: airflow/decorators/python.py ########## @@ -0,0 +1,112 @@ +# 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. + +import functools +from typing import Callable, Dict, Optional, TypeVar + +from airflow.decorators.base import BaseDecoratedOperator, base_task +from airflow.exceptions import AirflowException +from airflow.utils.decorators import apply_defaults + +PYTHON_OPERATOR_UI_COLOR = '#ffefeb' Review comment: Why a constant if it's only used once? ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
