SIENTIAPDE-1350: Implement file cleanup workflow and activities, including MinIO and local directory cleanup, configuration, and metrics.

scripts/run_cleanup_test.py

@@ -0,0 +1,76 @@
+#!/usr/bin/env python3
+"""Run cleanup_files workflow once for manual testing.
+
+This script starts the Temporal workflow `cleanup_files` a single time,
+using the same Temporal namespace and task queue as the main worker.
+
+It is intended only for local/manual testing; scheduling (cron) must be
+configured separately in Temporal.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import os
+import sys
+from datetime import timedelta
+from typing import Any
+
+from temporalio.client import Client
+
+# Ensure project root is on PYTHONPATH when running directly
+ROOT_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+if ROOT_DIR not in sys.path:
+    sys.path.insert(0, ROOT_DIR)
+
+from model_manager.workflows.cleanup_files import CleanupFiles  # noqa: E402
+
+
+async def main(argv: list[str]) -> None:
+    """Entry point for manual cleanup workflow execution.
+
+    Args:
+        argv: Command-line arguments (excluding program name).
+    """
+
+    # Config from environment / defaults
+    temporal_host = os.getenv('TEMPORAL_HOST', 'localhost:37463')
+    temporal_namespace = os.getenv('TEMPORAL_NAMESPACE', 'model-manager')
+    task_queue = 'cleanup-queue'
+
+    default_bucket = os.getenv('DEFAULT_CLEANUP_BUCKET', 'model-training')
+
+    # Optional CLI: bucket name override
+    bucket_name = default_bucket
+    if argv:
+        bucket_name = argv[0]
+
+    print(f"Connecting to Temporal at {temporal_host} (namespace={temporal_namespace})...")
+    client = await Client.connect(temporal_host, namespace=temporal_namespace)
+
+    input_data: dict[str, Any] = {
+        'bucket_name': bucket_name,
+    }
+
+    workflow_id = f"cleanup-files-manual-{int(asyncio.get_event_loop().time())}"
+
+    print(f"Starting cleanup_files workflow once...\n"
+          f"  workflow_id = {workflow_id}\n"
+          f"  task_queue  = {task_queue}\n"
+          f"  bucket_name = {bucket_name}")
+
+    handle = await client.start_workflow(
+        CleanupFiles.run,
+        input_data,
+        id=workflow_id,
+        task_queue=task_queue,
+        run_timeout=timedelta(minutes=10),
+    )
+
+    print("Workflow started, waiting for completion...")
+    await handle.result()
+    print("cleanup_files workflow completed successfully.")
+
+
+if __name__ == '__main__':  # pragma: no cover - manual utility script
+    asyncio.run(main(sys.argv[1:]))