Create app.py

app.py

@@ -0,0 +1,68 @@
+import streamlit as st
+import os
+import tempfile
+from pymongo import MongoClient
+from datetime import datetime
+from pathlib import Path
+from document_chunker import DocumentChunker
+from dotenv import load_dotenv
+
+load_dotenv()
+
+# MongoDB connection
+mongo_uri = os.getenv("MONGO_URI")
+db_name = os.getenv("MONGO_DB", "grant_docs")
+client = MongoClient(mongo_uri)
+
+st.set_page_config(page_title="Doc Chunker", layout="wide")
+st.title("📄 Document Chunker & Uploader")
+
+with st.sidebar:
+    st.header("Settings")
+    selected_collection = st.text_input("MongoDB Collection Name", "doc_chunks_cat")
+    is_grant_app = st.toggle("Is this a Grant Application?", value=True)
+    if st.button("Connect to Collection"):
+        collection = client[db_name][selected_collection]
+        st.success(f"Connected to `{selected_collection}` in `{db_name}`")
+
+uploaded_file = st.file_uploader("Upload a DOCX or TXT file", type=["docx", "txt"])
+
+if uploaded_file:
+    # Save file to temp path
+    temp_path = Path(tempfile.gettempdir()) / uploaded_file.name
+    with open(temp_path, "wb") as f:
+        f.write(uploaded_file.getbuffer())
+
+    st.success(f"Uploaded `{uploaded_file.name}`")
+
+    # Check if file already exists in collection
+    modified_time = datetime.now().isoformat()
+    collection = client[db_name][selected_collection]
+    if collection.find_one({"metadata.title": uploaded_file.name}):
+        st.warning("⚠️ This file already exists in the collection. Skipping...")
+    else:
+        st.write("⏳ Processing...")
+        chunker = DocumentChunker()
+        chunks = chunker.process_document(str(temp_path)) if is_grant_app else chunker.process_document(str(temp_path))
+
+        if chunks:
+            for chunk in chunks:
+                chunk['metadata'].update({
+                    "title": uploaded_file.name,
+                    "uploaded_at": modified_time,
+                })
+                collection.insert_one(chunk)
+
+            st.success(f"✅ {len(chunks)} chunks inserted into `{selected_collection}`")
+
+            # Show preview
+            for i, c in enumerate(chunks[:3]):
+                st.subheader(f"Chunk {i+1}: {c['metadata'].get('header') or 'No Header'}")
+                st.markdown(c['text'][:400] + "...")
+                st.caption(f"Topics: {', '.join(c['metadata']['topics'])} | Category: {c['metadata']['category']}")
+                st.progress(c['metadata']['confidence_score'])
+
+            if len(chunks) > 3:
+                st.info(f"... and {len(chunks)-3} more chunks processed.")
+        else:
+            st.warning("⚠️ No chunks were generated.")