AldawsariNLP commited on
Commit
3c4f305
·
1 Parent(s): 7b888fe

Improve RAG init & docs new3

Browse files
Files changed (1) hide show
  1. backend/main.py +11 -2
backend/main.py CHANGED
@@ -64,8 +64,17 @@ def initialize_rag_system():
64
  print(f"[RAG Init] ✓ Processed and indexed {num_docs} documents")
65
  rag_ready = True
66
  elif processed_json.exists():
67
- print("[RAG Init] processed_documents.json found. Using existing summaries.")
68
- rag_ready = True
 
 
 
 
 
 
 
 
 
69
  else:
70
  print("[RAG Init] No PDFs or processed_documents.json found. RAG remains uninitialized.")
71
  rag_ready = False
 
64
  print(f"[RAG Init] ✓ Processed and indexed {num_docs} documents")
65
  rag_ready = True
66
  elif processed_json.exists():
67
+ print("[RAG Init] processed_documents.json found. Building vectorstore from existing summaries...")
68
+ # Call process_and_index_documents even without PDFs to build vectorstore from JSON
69
+ # The method will detect no new PDFs and build from existing processed_documents.json
70
+ docs_path = str(docs_folder) if docs_folder.exists() else "documents"
71
+ num_docs = rag_system.process_and_index_documents(docs_path)
72
+ if rag_system.vectorstore is not None:
73
+ print(f"[RAG Init] ✓ Built vectorstore from processed_documents.json")
74
+ rag_ready = True
75
+ else:
76
+ print("[RAG Init] Warning: Could not build vectorstore from processed_documents.json")
77
+ rag_ready = False
78
  else:
79
  print("[RAG Init] No PDFs or processed_documents.json found. RAG remains uninitialized.")
80
  rag_ready = False