ocr improved

test_fix_safedistance.py

@@ -0,0 +1,30 @@
+import asyncio
+import sys
+import os
+sys.path.insert(0, 'LightRAG-main')
+from lightrag.document_processor import get_document_processor
+
+async def test_pdf():
+    processor = get_document_processor()
+    result = await processor.process_document('test/safedistance.pdf')
+    print(f"Success: {result.success}")
+    print(f"Content length: {len(result.content)}")
+    print(f"Tables found: {len(result.tables) if result.tables else 0}")
+    if result.tables:
+        for i, table in enumerate(result.tables):
+            print(f"Table {i}: rows={table.get('rows')}, cols={table.get('columns')}")
+            # Print first few rows
+            data = table.get('data', [])
+            for r in range(min(3, len(data))):
+                print(f"  Row {r}: {data[r]}")
+    # Check if OCR was used
+    if result.metadata.get('processed_with_ocr'):
+        print("OCR was used (good)")
+    else:
+        print("OCR was NOT used (maybe text extraction succeeded?)")
+    # Print first 500 chars of content
+    print("\n--- First 500 characters of extracted content ---")
+    print(result.content[:500])
+
+if __name__ == "__main__":
+    asyncio.run(test_pdf())