Source code for airflow.timetables.simple

# 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.
from __future__ import annotations

from typing import TYPE_CHECKING, Any, Collection, Sequence

from airflow.assets import AssetAlias, _AssetAliasCondition
from airflow.timetables.base import DagRunInfo, DataInterval, Timetable
from airflow.utils import timezone

if TYPE_CHECKING:
    from pendulum import DateTime
    from sqlalchemy import Session

    from airflow.assets import BaseAsset
    from airflow.models.asset import AssetEvent
    from airflow.timetables.base import TimeRestriction
    from airflow.utils.types import DagRunType


class _TrivialTimetable(Timetable):
    """Some code reuse for "trivial" timetables that has nothing complex."""

    periodic = False
    run_ordering: Sequence[str] = ("execution_date",)

    @classmethod
    def deserialize(cls, data: dict[str, Any]) -> Timetable:
        return cls()

    def __eq__(self, other: Any) -> bool:
        """
        As long as *other* is of the same type.

        This is only for testing purposes and should not be relied on otherwise.
        """
        if not isinstance(other, type(self)):
            return NotImplemented
        return True

    def serialize(self) -> dict[str, Any]:
        return {}

    def infer_manual_data_interval(self, *, run_after: DateTime) -> DataInterval:
        return DataInterval.exact(run_after)


[docs]class NullTimetable(_TrivialTimetable):
    """
    Timetable that never schedules anything.

    This corresponds to ``schedule=None``.
    """

[docs]    can_be_scheduled = False
[docs]    description: str = "Never, external triggers only"

    @property
[docs]    def summary(self) -> str:
        return "None"

[docs]    def next_dagrun_info(
        self,
        *,
        last_automated_data_interval: DataInterval | None,
        restriction: TimeRestriction,
    ) -> DagRunInfo | None:
        return None


[docs]class OnceTimetable(_TrivialTimetable):
    """
    Timetable that schedules the execution once as soon as possible.

    This corresponds to ``schedule="@once"``.
    """

[docs]    description: str = "Once, as soon as possible"

    @property
[docs]    def summary(self) -> str:
        return "@once"

[docs]    def next_dagrun_info(
        self,
        *,
        last_automated_data_interval: DataInterval | None,
        restriction: TimeRestriction,
    ) -> DagRunInfo | None:
        if last_automated_data_interval is not None:
            return None  # Already run, no more scheduling.
        if restriction.earliest is None:  # No start date, won't run.
            return None
        # "@once" always schedule to the start_date determined by the DAG and
        # tasks, regardless of catchup or not. This has been the case since 1.10
        # and we're inheriting it.
        run_after = restriction.earliest
        if restriction.latest is not None and run_after > restriction.latest:
            return None
        return DagRunInfo.exact(run_after)


[docs]class ContinuousTimetable(_TrivialTimetable):
    """
    Timetable that schedules continually, while still respecting start_date and end_date.

    This corresponds to ``schedule="@continuous"``.
    """

[docs]    description: str = "As frequently as possible, but only one run at a time."

[docs]    active_runs_limit = 1  # Continuous DAGRuns should be constrained to one run at a time

    @property
[docs]    def summary(self) -> str:
        return "@continuous"

[docs]    def next_dagrun_info(
        self,
        *,
        last_automated_data_interval: DataInterval | None,
        restriction: TimeRestriction,
    ) -> DagRunInfo | None:
        if restriction.earliest is None:  # No start date, won't run.
            return None
        if last_automated_data_interval is not None:  # has already run once
            start = last_automated_data_interval.end
            end = timezone.coerce_datetime(timezone.utcnow())
        else:  # first run
            start = restriction.earliest
            end = max(
                restriction.earliest, timezone.coerce_datetime(timezone.utcnow())
            )  # won't run any earlier than start_date

        if restriction.latest is not None and end > restriction.latest:
            return None

        return DagRunInfo.interval(start, end)


class AssetTriggeredTimetable(_TrivialTimetable):
    """
    Timetable that never schedules anything.

    This should not be directly used anywhere, but only set if a DAG is triggered by assets.

    :meta private:
    """

    UNRESOLVED_ALIAS_SUMMARY = "Unresolved AssetAlias"

    description: str = "Triggered by assets"

    def __init__(self, assets: BaseAsset) -> None:
        super().__init__()
        self.asset_condition = assets
        if isinstance(self.asset_condition, AssetAlias):
            self.asset_condition = _AssetAliasCondition(self.asset_condition.name)

        if not next(self.asset_condition.iter_assets(), False):
            self._summary = AssetTriggeredTimetable.UNRESOLVED_ALIAS_SUMMARY
        else:
            self._summary = "Asset"

    @classmethod
    def deserialize(cls, data: dict[str, Any]) -> Timetable:
        from airflow.serialization.serialized_objects import decode_asset_condition

        return cls(decode_asset_condition(data["asset_condition"]))

    @property
    def summary(self) -> str:
        return self._summary

    def serialize(self) -> dict[str, Any]:
        from airflow.serialization.serialized_objects import encode_asset_condition

        return {"asset_condition": encode_asset_condition(self.asset_condition)}

    def generate_run_id(
        self,
        *,
        run_type: DagRunType,
        logical_date: DateTime,
        data_interval: DataInterval | None,
        session: Session | None = None,
        events: Collection[AssetEvent] | None = None,
        **extra,
    ) -> str:
        from airflow.models.dagrun import DagRun

        return DagRun.generate_run_id(run_type, logical_date)

    def data_interval_for_events(
        self,
        logical_date: DateTime,
        events: Collection[AssetEvent],
    ) -> DataInterval:
        if not events:
            return DataInterval(logical_date, logical_date)

        start_dates, end_dates = [], []
        for event in events:
            if event.source_dag_run is not None:
                start_dates.append(event.source_dag_run.data_interval_start)
                end_dates.append(event.source_dag_run.data_interval_end)
            else:
                start_dates.append(event.timestamp)
                end_dates.append(event.timestamp)

        start = min(start_dates)
        end = max(end_dates)
        return DataInterval(start, end)

    def next_dagrun_info(
        self,
        *,
        last_automated_data_interval: DataInterval | None,
        restriction: TimeRestriction,
    ) -> DagRunInfo | None:
        return None