mik-laj commented on a change in pull request #9597: URL: https://github.com/apache/airflow/pull/9597#discussion_r502515545
########## File path: airflow/api_connexion/endpoints/task_instance_endpoint.py ########## @@ -14,39 +14,234 @@ # KIND, either express or implied. See the License for the # specific language governing permissions and limitations # under the License. +from typing import Any, List, Optional, Tuple + +from flask import request +from marshmallow import ValidationError +from sqlalchemy import and_, func, or_ -# TODO(mik-laj): We have to implement it. -# Do you want to help? Please look at: https://github.com/apache/airflow/issues/8132 from airflow.api_connexion import security +from airflow.api_connexion.exceptions import NotFound, BadRequest +from airflow.api_connexion.parameters import format_datetime, format_parameters +from airflow.api_connexion.schemas.task_instance_schema import ( + clear_task_instance_form, + TaskInstanceCollection, + task_instance_collection_schema, + task_instance_schema, + task_instance_batch_form, + task_instance_reference_collection_schema, + TaskInstanceReferenceCollection, +) +from airflow.configuration import conf +from airflow.models.dagrun import DagRun as DR +from airflow.models.taskinstance import clear_task_instances, TaskInstance as TI +from airflow.models.dag import DagModel +from airflow.models.dagbag import DagBag +from airflow.models import SlaMiss +from airflow.utils.state import State +from airflow.utils.session import provide_session @security.requires_authentication -def get_task_instance(): - """ - Get a task instance - """ - raise NotImplementedError("Not implemented yet.") +@provide_session +def get_task_instance(dag_id: str, dag_run_id: str, task_id: str, session=None): + """Get task instance""" + query = ( + session.query(TI) + .filter(TI.dag_id == dag_id) + .join(DR, and_(TI.dag_id == DR.dag_id, TI.execution_date == DR.execution_date)) + .filter(DR.run_id == dag_run_id) + .filter(TI.task_id == task_id) + .join( + SlaMiss, + and_( + SlaMiss.dag_id == TI.dag_id, + SlaMiss.execution_date == TI.execution_date, + SlaMiss.task_id == TI.task_id, + ), + isouter=True, + ) + .add_entity(SlaMiss) + ) + task_instance = query.one_or_none() + + if task_instance is None: + raise NotFound("Task instance not found") + + return task_instance_schema.dump(task_instance) + + +def _apply_array_filter(query, key, values): + if values is not None: + query = query.filter(or_(*[key == v for v in values])) + return query + + +def _apply_range_filter(query, key, value_range: Tuple[Any, Any]): + gte_value, lte_value = value_range + if gte_value is not None: + query = query.filter(key >= gte_value) + if lte_value is not None: + query = query.filter(key <= lte_value) + return query +@format_parameters( + { + "execution_date_gte": format_datetime, + "execution_date_lte": format_datetime, + "start_date_gte": format_datetime, + "start_date_lte": format_datetime, + "end_date_gte": format_datetime, + "end_date_lte": format_datetime, + } +) @security.requires_authentication -def get_task_instances(): +@provide_session +def get_task_instances( + limit: int, + dag_id: Optional[str] = None, + dag_run_id: Optional[str] = None, + execution_date_gte: Optional[str] = None, + execution_date_lte: Optional[str] = None, + start_date_gte: Optional[str] = None, + start_date_lte: Optional[str] = None, + end_date_gte: Optional[str] = None, + end_date_lte: Optional[str] = None, + duration_gte: Optional[float] = None, + duration_lte: Optional[float] = None, + state: Optional[str] = None, + pool: Optional[List[str]] = None, + queue: Optional[List[str]] = None, + offset: Optional[int] = None, + session=None, +): # pylint: disable=too-many-arguments """ - Get list of task instances of DAG. + Get list of task instances. """ - raise NotImplementedError("Not implemented yet.") + query = session.query(TI) + + if dag_id != "~": + query = query.filter(TI.dag_id == dag_id) + if dag_run_id != "~": + query = query.join(DR, and_(TI.dag_id == DR.dag_id, TI.execution_date == DR.execution_date)) + query = query.filter(DR.run_id == dag_run_id) + query = _apply_range_filter( + query, key=DR.execution_date, value_range=(execution_date_gte, execution_date_lte), + ) + query = _apply_range_filter(query, key=TI.start_date, value_range=(start_date_gte, start_date_lte)) + query = _apply_range_filter(query, key=TI.end_date, value_range=(end_date_gte, end_date_lte)) + query = _apply_range_filter(query, key=TI.duration, value_range=(duration_gte, duration_lte)) + query = _apply_array_filter(query, key=TI.state, values=state) + query = _apply_array_filter(query, key=TI.pool, values=pool) + query = _apply_array_filter(query, key=TI.queue, values=queue) + + query = query.join( + SlaMiss, + and_( + SlaMiss.dag_id == TI.dag_id, + SlaMiss.task_id == TI.task_id, + SlaMiss.execution_date == TI.execution_date, + ), + isouter=True, + ).add_entity(SlaMiss) + task_instances = query.offset(offset).limit(limit).all() + total_entries = session.query( + func.count(func.distinct(TI.dag_id, TI.execution_date, TI.task_id)) + ).scalar() + return task_instance_collection_schema.dump( + TaskInstanceCollection(task_instances=task_instances, total_entries=total_entries) + ) @security.requires_authentication -def get_task_instances_batch(): +@provide_session +def get_task_instances_batch(session=None): """ Get list of task instances. """ - raise NotImplementedError("Not implemented yet.") + body = request.get_json() + try: + data = task_instance_batch_form.load(body) + except ValidationError as err: + raise BadRequest(detail=str(err.messages)) + query = session.query(TI) + + query = _apply_array_filter(query, key=TI.dag_id, values=data["dag_ids"]) + query = _apply_range_filter( + query, key=TI.execution_date, value_range=(data["execution_date_gte"], data["execution_date_lte"]), + ) + query = _apply_range_filter( + query, key=TI.start_date, value_range=(data["start_date_gte"], data["start_date_lte"]), + ) + query = _apply_range_filter( + query, key=TI.end_date, value_range=(data["end_date_gte"], data["end_date_lte"]) + ) + query = _apply_range_filter( + query, key=TI.duration, value_range=(data["duration_gte"], data["duration_lte"]) + ) + query = _apply_array_filter(query, key=TI.state, values=data["state"]) + query = _apply_array_filter(query, key=TI.pool, values=data["pool"]) + query = _apply_array_filter(query, key=TI.queue, values=data["queue"]) + + query = query.join( + SlaMiss, + and_( + SlaMiss.dag_id == TI.dag_id, + SlaMiss.task_id == TI.task_id, + SlaMiss.execution_date == TI.execution_date, + ), + isouter=True, + ).add_entity(SlaMiss) + task_instances = query.all() + total_entries = session.query( + func.count(func.distinct(TI.dag_id, TI.execution_date, TI.task_id)) + ).scalar() + + return task_instance_collection_schema.dump( + TaskInstanceCollection(task_instances=task_instances, total_entries=total_entries) + ) @security.requires_authentication -def post_clear_task_instances(): +@provide_session +def post_clear_task_instances(dag_id: str, session=None): Review comment: Unfortunately not. I will add them in a moment. ---------------------------------------------------------------- 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]
