- added three iwm articles as test data

- added simple webview and FastAPI server

local_rag.py

@@ -5,14 +5,14 @@ Minimal dependencies, simple code
 import os
 from pathlib import Path
 from langchain_community.document_loaders import PyPDFLoader, TextLoader
-from langchain.text_splitter import RecursiveCharacterTextSplitter
-from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+from langchain_huggingface import HuggingFaceEmbeddings
 from langchain_community.vectorstores import FAISS
 from langchain_ollama import ChatOllama
 
 
 class LocalRAG:
-    def __init__(self, vectorstore_path="./vectorstore", ollama_model="llama2"):
+    def __init__(self, vectorstore_path="./vectorstore", ollama_model="mistral:7b"):
         """Initialize local RAG system"""
         self.vectorstore_path = vectorstore_path
         self.ollama_model = ollama_model
@@ -97,6 +97,42 @@ class LocalRAG:
         self.vectorstore.save_local(self.vectorstore_path)
         print(f"Vector store saved to {self.vectorstore_path}")
     
+    def list_documents(self):
+        """List all documents in the vector store"""
+        if self.vectorstore is None:
+            print("No documents in vector store.")
+            return []
+        
+        # Get all documents from the vector store
+        # We'll retrieve a large number to get all documents
+        all_docs = self.vectorstore.similarity_search("", k=10000)  # Large k to get all
+        
+        # Extract unique document sources from metadata
+        documents = {}
+        for doc in all_docs:
+            source = doc.metadata.get('source', 'Unknown')
+            if source not in documents:
+                documents[source] = {
+                    'source': source,
+                    'chunks': 0,
+                    'page': doc.metadata.get('page', None)
+                }
+            documents[source]['chunks'] += 1
+        
+        # Convert to list and sort
+        doc_list = list(documents.values())
+        doc_list.sort(key=lambda x: x['source'])
+        
+        print(f"\nDocuments in vector store ({len(doc_list)} unique documents):")
+        print("-" * 60)
+        for doc_info in doc_list:
+            print(f"  - {doc_info['source']}")
+            print(f"    Chunks: {doc_info['chunks']}")
+            if doc_info['page'] is not None:
+                print(f"    Page: {doc_info['page']}")
+        
+        return doc_list
+    
     def query(self, question, k=4):
         """Query the RAG system"""
         if self.vectorstore is None:
@@ -134,21 +170,25 @@ def main():
     print("=" * 60)
     
     # Initialize
-    rag = LocalRAG(ollama_model="llama2")
+    rag = LocalRAG(ollama_model="mistral:7b")
     
     # Add documents (uncomment and add your file paths)
-    rag.add_documents([
-         "diverses/local_rag/test1.pdf",
-         "diverses/local_rag/test2.txt"
-    ])
+    # rag.add_documents([
+    #     "data/dok1.pdf",
+    #     "data/dok2.pdf",
+    #     "data/dok3.pdf"
+    # ])
+    
+    # List documents
+    rag.list_documents()
     
     # Query
-    # question = "What is this document about?"
-    # answer = rag.query(question)
-    # print(f"\nQuestion: {question}")
-    # print(f"Answer: {answer}")
+    question = "What do the documents say about modality for perceived message perception?"
+    answer = rag.query(question)
+    print(f"\nQuestion: {question}")
+    print(f"Answer: {answer}")
     
-    print("\nSetup complete! Uncomment the code above to add documents and query.")
+    # print("\nSetup complete! Uncomment the code above to add documents and query.")
 
 
 if __name__ == "__main__":