{
  "context": "Evaluates how the solution uses Apache Airflow's Sqoop provider to build a templated import/export DAG with correct filtering, null handling, dependencies, and safe cancellation for Sqoop subprocesses.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Sqoop import task",
      "description": "Uses SqoopOperator (cmd_type=\"import\") with the provided conn_id and either table or query params to pull source_table rows, sets split_by for mapper parallelism, and writes to a target_dir built from hdfs_import_dir and run context.",
      "max_score": 30
    },
    {
      "name": "Templated filtering",
      "description": "Import command renders its WHERE or query filter with Jinja macros (e.g., ds/ds_nodash) rather than hardcoded dates, and includes the required $CONDITIONS placeholder when using query mode so Sqoop applies mapper predicates correctly.",
      "max_score": 20
    },
    {
      "name": "Export configuration",
      "description": "Uses SqoopOperator (cmd_type=\"export\") pointing at export_dir and target_table, runs in batch mode, and sets input_null_string plus input_null_non_string to apply the expected null sentinels during export.",
      "max_score": 25
    },
    {
      "name": "Task dependency",
      "description": "Defines the export SqoopOperator downstream of the import SqoopOperator using Airflow dependencies so a failed import prevents export execution (default trigger_rule behavior is preserved).",
      "max_score": 10
    },
    {
      "name": "Kill handling",
      "description": "Implements or leverages SqoopOperator.on_kill to terminate the underlying SqoopHook subprocess (popen) when the task is cancelled, avoiding orphaned Sqoop jobs.",
      "max_score": 15
    }
  ]
}