Create components/upload_files.py

components/upload_files.py

@@ -0,0 +1,43 @@
+# components/upload_files.py
+
+import streamlit as st
+import os
+from langchain.document_loaders import PyPDFLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.vectorstores import Chroma
+from langchain.embeddings import HuggingFaceEmbeddings
+
+def run_upload_files():
+    st.header("📂 Upload Business Documents")
+
+    uploaded_file = st.file_uploader("Upload a PDF", type=["pdf"])
+    
+    if uploaded_file:
+        with st.spinner("Processing document..."):
+            os.makedirs("data/user_docs", exist_ok=True)
+
+            # Save uploaded PDF
+            file_path = os.path.join("data/user_docs", uploaded_file.name)
+            with open(file_path, "wb") as f:
+                f.write(uploaded_file.read())
+
+            # Load PDF text
+            loader = PyPDFLoader(file_path)
+            pages = loader.load_and_split()
+
+            # Split into chunks
+            splitter = RecursiveCharacterTextSplitter(chunk_size=800, chunk_overlap=100)
+            docs = splitter.split_documents(pages)
+
+            # Create embeddings
+            embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
+
+            vectordb = Chroma.from_documents(
+                documents=docs,
+                embedding=embeddings,
+                persist_directory="data/vectorstore"
+            )
+            vectordb.persist()
+            st.session_state.vectorstore = vectordb
+
+            st.success("✅ Document processed and added to your knowledge base.")