feat: add migrations, test scripts, and utility tools

- Add database migrations (006-028) for face recognition, identity, file_uuid - Add test scripts for ASR, face, search, processing - Add portal frontend (Tauri) - Add config, benchmark, and monitoring utilities - Add model checkpoints and pretrained model references
2026-04-30 15:11:53 +08:00
parent 4d75b2e251
commit b54c2def30
192 changed files with 46721 additions and 0 deletions
--- a/test_direct_asr.py
+++ b/test_direct_asr.py
@@ -0,0 +1,151 @@
+#!/opt/homebrew/bin/python3.11
+"""Test ASR processor directly (not via subprocess)."""
+
+import sys
+import os
+import tempfile
+import json
+
+# Add scripts directory to path
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+
+# Import the run_asr function
+try:
+    from scripts.asr_processor_debug import run_asr
+
+    print("✓ Imported run_asr from asr_processor_debug")
+except ImportError as e:
+    print(f"✗ Failed to import: {e}")
+    # Try alternative import
+    try:
+        import scripts.asr_processor_debug as asr_module
+
+        run_asr = asr_module.run_asr
+        print("✓ Imported module")
+    except Exception as e2:
+        print(f"✗ Failed: {e2}")
+        sys.exit(1)
+
+
+def test_small_video():
+    """Test on a small video file (should work)."""
+    test_video = "/Users/accusys/test_video/20250209_212949.mp4"
+    if not os.path.exists(test_video):
+        print(f"Test video not found: {test_video}")
+        return
+
+    print(f"Testing on small video: {test_video}")
+
+    # Create temp output file
+    with tempfile.NamedTemporaryFile(suffix=".json", delete=False) as f:
+        output_path = f.name
+
+    try:
+        # Run ASR directly (no subprocess)
+        print("Calling run_asr...")
+        run_asr(
+            video_path=test_video,
+            output_path=output_path,
+            uuid="test_direct",
+            chunk_duration=600,  # 10 minutes
+            max_direct_duration=1200,  # 20 minutes
+            model_size="tiny",
+            compute_type="int8",
+            monitor_interval=60,
+        )
+
+        # Check output
+        if os.path.exists(output_path):
+            with open(output_path, "r") as f:
+                output = json.load(f)
+            print(f"✓ ASR completed successfully")
+            print(f"  Segments: {len(output.get('segments', []))}")
+            print(f"  Language: {output.get('language')}")
+        else:
+            print("✗ Output file not created")
+
+    except Exception as e:
+        print(f"✗ ASR failed: {e}")
+        import traceback
+
+        traceback.print_exc()
+    finally:
+        if os.path.exists(output_path):
+            os.unlink(output_path)
+            print("✓ Cleaned up output file")
+
+
+def test_large_video_chunk():
+    """Test on a large video but with small chunk duration to debug hang."""
+    # Use a large video file
+    test_video = "/Users/accusys/test_video/1636719d-c31f-78ac-f1dd-8ab0b0b36c66.mov"
+    if not os.path.exists(test_video):
+        print(f"Large test video not found: {test_video}")
+        return
+
+    print(f"\nTesting on large video with 60s chunks (debug): {test_video}")
+
+    # Create temp output file
+    with tempfile.NamedTemporaryFile(suffix=".json", delete=False) as f:
+        output_path = f.name
+
+    try:
+        # Run with very short chunk duration to see if we can get past first chunk
+        print("Calling run_asr with 60s chunks...")
+        # We'll run in a separate thread with timeout
+        import threading
+        import time
+
+        result = {"success": False, "error": None, "output": None}
+
+        def run_asr_thread():
+            try:
+                run_asr(
+                    video_path=test_video,
+                    output_path=output_path,
+                    uuid="test_large_debug",
+                    chunk_duration=60,  # 1 minute chunks (smaller for debugging)
+                    max_direct_duration=120,  # 2 minutes
+                    model_size="tiny",
+                    compute_type="int8",
+                    monitor_interval=30,
+                )
+                result["success"] = True
+            except Exception as e:
+                result["error"] = str(e)
+                result["traceback"] = traceback.format_exc()
+
+        thread = threading.Thread(target=run_asr_thread)
+        thread.daemon = True
+        thread.start()
+
+        # Wait for 30 seconds max
+        thread.join(timeout=30)
+
+        if thread.is_alive():
+            print("✗ ASR is hanging (thread still alive after 30 seconds)")
+            result["timeout"] = True
+        else:
+            if result["success"]:
+                print("✓ ASR completed within 30 seconds")
+                if os.path.exists(output_path):
+                    with open(output_path, "r") as f:
+                        output = json.load(f)
+                    print(f"  Segments: {len(output.get('segments', []))}")
+            else:
+                print(f"✗ ASR failed: {result.get('error')}")
+
+    except Exception as e:
+        print(f"✗ Test failed: {e}")
+        import traceback
+
+        traceback.print_exc()
+    finally:
+        if os.path.exists(output_path):
+            os.unlink(output_path)
+            print("✓ Cleaned up output file")
+
+
+if __name__ == "__main__":
+    test_small_video()
+    test_large_video_chunk()