Update app.py

app.py

@@ -1,97 +1,80 @@
 import os
 import gradio as gr
-from rag_tool import RAGTool
+import warnings
+from pathlib import Path
+import shutil
 
-# Initialize the RAG Tool with default settings
-rag_tool = RAGTool(
-    documents_path="./documents",
-    embedding_model="sentence-transformers/all-MiniLM-L6-v2",
-    vector_store_type="faiss",
-    chunk_size=1000,
-    chunk_overlap=200,
-    persist_directory="./vector_store"
-)
+# Suppress LangChain deprecation warnings
+warnings.filterwarnings("ignore", category=DeprecationWarning)
 
-# Function to handle document uploads
-def upload_documents(files, chunk_size, chunk_overlap, embedding_model, vector_store_type):
-    # Create a temporary directory for uploaded files
-    os.makedirs("./uploaded_docs", exist_ok=True)
-    
-    # Save uploaded files
-    for file in files:
-        file_path = os.path.join("./uploaded_docs", os.path.basename(file.name))
-        with open(file_path, "wb") as f:
-            f.write(file.read())
-    
-    # Initialize a new RAG Tool with the uploaded documents
-    global rag_tool
-    rag_tool = RAGTool(
-        documents_path="./uploaded_docs",
-        embedding_model=embedding_model,
-        vector_store_type=vector_store_type,
-        chunk_size=int(chunk_size),
-        chunk_overlap=int(chunk_overlap),
-        persist_directory="./uploaded_vector_store"
-    )
-    
-    return f"Documents uploaded and processed. Vector store created with {embedding_model} model."
+from rag_tool import RAGTool
 
-# Function to handle queries
-def query_documents(query, top_k):
-    global rag_tool
-    return rag_tool(query, top_k=int(top_k))
+# Initialize the RAG Tool
+rag_tool = RAGTool()
 
-# Gradio interface
-with gr.Blocks(title="Advanced RAG Tool") as demo:
-    gr.Markdown("# Advanced RAG Tool")
-    gr.Markdown("Upload documents and query them using semantic search")
-    
-    with gr.Tab("Upload & Configure"):
-        with gr.Row():
-            with gr.Column():
-                files = gr.File(file_count="multiple", label="Upload Documents")
-                chunk_size = gr.Slider(200, 2000, value=1000, step=100, label="Chunk Size")
-                chunk_overlap = gr.Slider(0, 500, value=200, step=50, label="Chunk Overlap")
-            
-            with gr.Column():
-                embedding_models = [
-                    "sentence-transformers/all-MiniLM-L6-v2",
-                    "BAAI/bge-small-en-v1.5",
-                    "BAAI/bge-base-en-v1.5",
-                    "thenlper/gte-small",
-                    "thenlper/gte-base"
-                ]
-                embedding_model = gr.Dropdown(
-                    choices=embedding_models, 
-                    value="sentence-transformers/all-MiniLM-L6-v2",
-                    label="Embedding Model"
-                )
-                vector_store_type = gr.Radio(
-                    choices=["faiss", "chroma"], 
-                    value="faiss",
-                    label="Vector Store Type"
-                )
+# Function to handle document uploads
+def upload_file(file):
+    try:
+        # Create documents directory if it doesn't exist
+        os.makedirs("./documents", exist_ok=True)
+        
+        # Get the file path and name
+        file_path = Path(file.name)
+        destination = Path("./documents") / file_path.name
         
-        upload_button = gr.Button("Upload and Process Documents")
-        upload_result = gr.Textbox(label="Upload Result")
+        # Copy the file to documents directory
+        shutil.copy(file_path, destination)
         
-        upload_button.click(
-            upload_documents,
-            inputs=[files, chunk_size, chunk_overlap, embedding_model, vector_store_type],
-            outputs=upload_result
+        # Configure RAG tool
+        rag_tool.configure(
+            documents_path=str(destination),
+            embedding_model="sentence-transformers/all-MiniLM-L6-v2",
+            persist_directory="./vector_store"
         )
+        
+        return f"File uploaded and processed: {file_path.name}"
+    except Exception as e:
+        return f"Error processing file: {str(e)}"
+
+# Function to query the documents
+def query_document(question):
+    try:
+        if not hasattr(rag_tool, 'vector_store') or rag_tool.vector_store is None:
+            return "Please upload a document first."
+        
+        response = rag_tool(question)
+        return response
+    except Exception as e:
+        return f"Error querying document: {str(e)}"
+
+# Create a simple Gradio interface
+with gr.Blocks(title="RAG Tool") as demo:
+    gr.Markdown("# Document Question Answering System")
+    gr.Markdown("Upload a document (PDF, TXT) and ask questions about it")
     
-    with gr.Tab("Query Documents"):
-        query = gr.Textbox(label="Your Question", placeholder="What information are you looking for?")
-        top_k = gr.Slider(1, 10, value=3, step=1, label="Number of Results")
-        query_button = gr.Button("Search")
-        answer = gr.Textbox(label="Results")
+    with gr.Row():
+        with gr.Column():
+            file_input = gr.File(label="Upload Document")
+            upload_button = gr.Button("Process Document")
+            upload_result = gr.Textbox(label="Upload Status")
         
-        query_button.click(
-            query_documents,
-            inputs=[query, top_k],
-            outputs=answer
-        )
+        with gr.Column():
+            query_input = gr.Textbox(label="Ask a Question", placeholder="What would you like to know?")
+            query_button = gr.Button("Get Answer")
+            response_output = gr.Textbox(label="Answer")
+    
+    # Set up the button click events
+    upload_button.click(
+        upload_file,
+        inputs=file_input,
+        outputs=upload_result
+    )
+    
+    query_button.click(
+        query_document,
+        inputs=query_input,
+        outputs=response_output
+    )
 
 # Launch the app
 if __name__ == "__main__":