Coverage for /pythoncovmergedfiles/medio/medio/src/airflow/airflow/utils/dag_cycle_tester.py: 28%

1# Licensed to the Apache Software Foundation (ASF) under one 

2# or more contributor license agreements. See the NOTICE file 

3# distributed with this work for additional information 

4# regarding copyright ownership. The ASF licenses this file 

5# to you under the Apache License, Version 2.0 (the 

6# "License"); you may not use this file except in compliance 

7# with the License. You may obtain a copy of the License at 

8# 

9# http://www.apache.org/licenses/LICENSE-2.0 

10# 

11# Unless required by applicable law or agreed to in writing, 

12# software distributed under the License is distributed on an 

13# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY 

14# KIND, either express or implied. See the License for the 

15# specific language governing permissions and limitations 

16# under the License. 

17"""DAG Cycle tester.""" 

18 

19from __future__ import annotations 

20 

21from collections import defaultdict, deque 

22from typing import TYPE_CHECKING 

23 

24from airflow.exceptions import AirflowDagCycleException, RemovedInAirflow3Warning 

25 

26if TYPE_CHECKING: 

27 from airflow.models.dag import DAG 

28 

29CYCLE_NEW = 0 

30CYCLE_IN_PROGRESS = 1 

31CYCLE_DONE = 2 

32 

33 

34def test_cycle(dag: DAG) -> None: 

35 """ 

36 A wrapper function of `check_cycle` for backward compatibility purpose. 

37 

38 New code should use `check_cycle` instead since this function name `test_cycle` starts 

39 with 'test_' and will be considered as a unit test by pytest, resulting in failure. 

40 """ 

41 from warnings import warn 

42 

43 warn( 

44 "Deprecated, please use `check_cycle` at the same module instead.", 

45 RemovedInAirflow3Warning, 

46 stacklevel=2, 

47 ) 

48 return check_cycle(dag) 

49 

50 

51def check_cycle(dag: DAG) -> None: 

52 """Check to see if there are any cycles in the DAG. 

53 

54 :raises AirflowDagCycleException: If cycle is found in the DAG. 

55 """ 

56 # default of int is 0 which corresponds to CYCLE_NEW 

57 visited: dict[str, int] = defaultdict(int) 

58 path_stack: deque[str] = deque() 

59 task_dict = dag.task_dict 

60 

61 def _check_adjacent_tasks(task_id, current_task): 

62 """Return first untraversed child task, else None if all tasks traversed.""" 

63 for adjacent_task in current_task.get_direct_relative_ids(): 

64 if visited[adjacent_task] == CYCLE_IN_PROGRESS: 

65 msg = f"Cycle detected in DAG: {dag.dag_id}. Faulty task: {task_id}" 

66 raise AirflowDagCycleException(msg) 

67 elif visited[adjacent_task] == CYCLE_NEW: 

68 return adjacent_task 

69 return None 

70 

71 for dag_task_id in dag.task_dict.keys(): 

72 if visited[dag_task_id] == CYCLE_DONE: 

73 continue 

74 path_stack.append(dag_task_id) 

75 while path_stack: 

76 current_task_id = path_stack[-1] 

77 if visited[current_task_id] == CYCLE_NEW: 

78 visited[current_task_id] = CYCLE_IN_PROGRESS 

79 task = task_dict[current_task_id] 

80 child_to_check = _check_adjacent_tasks(current_task_id, task) 

81 if not child_to_check: 

82 visited[current_task_id] = CYCLE_DONE 

83 path_stack.pop() 

84 else: 

85 path_stack.append(child_to_check)