refact: add log

test_paddleocr_vl_integration.py

@@ -0,0 +1,53 @@
+"""Test script for PaddleOCR-VL integration in MineruOCRService."""
+
+import cv2
+import numpy as np
+from app.services.ocr_service import MineruOCRService
+from app.services.converter import Converter
+from app.services.image_processor import ImageProcessor
+
+def test_paddleocr_vl_integration():
+    """Test that PaddleOCR-VL is called when image references are found."""
+    
+    # Create a simple test image (white background with black text)
+    test_image = np.ones((100, 300, 3), dtype=np.uint8) * 255
+    cv2.putText(test_image, "x^2 + y^2 = 1", (50, 50), 
+                cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 0), 2)
+    
+    # Initialize service
+    service = MineruOCRService(
+        api_url="http://127.0.0.1:8000/file_parse",
+        converter=Converter(),
+        image_processor=ImageProcessor(),
+        paddleocr_vl_url="http://localhost:8000/v1"  # Your PaddleOCR-VL server
+    )
+    
+    # Simulate markdown with image reference (this is what Mineru returns)
+    test_markdown = "![](images/af7f211f671f16f57d346e8e17611e68e0f4671bd1ae52ed59013c10eecef589.jpg)"
+    
+    print("Testing formula extraction...")
+    result = service._extract_and_recognize_formulas(test_markdown, test_image)
+    
+    print(f"\nOriginal markdown: {test_markdown}")
+    print(f"Processed markdown: {result}")
+    
+    # Check if the image reference was replaced
+    if "![](images/" in result:
+        print("\n❌ FAILED: Image reference was not replaced")
+    else:
+        print("\n✅ SUCCESS: Image reference was replaced with formula")
+
+if __name__ == "__main__":
+    print("=" * 60)
+    print("PaddleOCR-VL Integration Test")
+    print("=" * 60)
+    print("\nMake sure your PaddleOCR-VL server is running at:")
+    print("http://localhost:8000/v1")
+    print("\n" + "=" * 60 + "\n")
+    
+    try:
+        test_paddleocr_vl_integration()
+    except Exception as e:
+        print(f"\n❌ Test failed with error: {e}")
+        import traceback
+        traceback.print_exc()