Write Airflow Code | Dataford Interview Questions

Can you write Airflow code?

Asked in the Round 1 stage. Implement execution_order(tasks, dependencies) to validate an Airflow-style DAG. tasks is an array of unique task-name strings. Each dependency is [upstream, downstream]. Return all task names in a valid deterministic execution order, choosing the lexicographically smallest available task first. Return an empty array if dependencies reference an unknown task or contain a cycle.

Examples: tasks = ["extract", "transform", "load"], dependencies = [["extract", "transform"], ["transform", "load"]] returns ["extract", "transform", "load"]. A cyclic dependency returns [].

Constraints: Up to 1,000 tasks and 5,000 dependencies; task names are non-empty.

Constraints

1 <= tasks.length <= 1,000
0 <= dependencies.length <= 5,000
Task names are unique, non-empty strings
Each dependency contains exactly two task names
Return an empty array for unknown task references or cycles

Function Signature

def execution_order(tasks, dependencies):

Your question is Write Airflow Code. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem