Spaces:
Sleeping
Sleeping
File size: 4,248 Bytes
25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 8008b94 25b9910 fde7654 25b9910 8008b94 25b9910 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 |
import os
import streamlit as st
import google.generativeai as genai
from dotenv import load_dotenv
from notion_client import Client
# --- CONFIGURATION ---
load_dotenv()
# Configure APIs
try:
genai.configure(api_key=os.getenv("GEMINI_API_KEY"))
notion = Client(auth=os.getenv("NOTION_KEY"))
NOTION_DATABASE_ID = os.getenv("NOTION_DATABASE_ID")
except (AttributeError, TypeError):
st.error("⚠️ API keys or Database ID not found. Please set them in your secrets.")
st.stop()
# --- 1. CONTEXT PROVIDER (Live from Notion) ---
@st.cache_data(ttl=600) # Cache the data for 10 minutes
def fetch_notion_database():
"""Fetches and parses the Notion database."""
try:
response = notion.databases.query(database_id=NOTION_DATABASE_ID)
results = []
for page in response.get("results", []):
properties = page.get("properties", {})
# Extract data from Notion properties
topic_prop = properties.get("Topic", {}).get("title", [])
content_prop = properties.get("Content", {}).get("rich_text", [])
keywords_prop = properties.get("Keywords", {}).get("rich_text", [])
# Safely get the plain text content
topic = topic_prop[0]["plain_text"] if topic_prop else "No Topic"
content = content_prop[0]["plain_text"] if content_prop else ""
keywords_str = keywords_prop[0]["plain_text"] if keywords_prop else ""
# Format into the structure our app expects
results.append({
"id": topic.lower().replace(" ", "-"),
"keywords": [k.strip() for k in keywords_str.split(',')],
"content": content
})
st.success("Successfully connected to Notion!")
return results
except Exception as e:
st.error(f"Failed to connect to Notion: {e}")
return []
# Fetch the data
notion_data = fetch_notion_database()
def get_context(query: str) -> str | None:
"""Finds the most relevant context from the fetched Notion data."""
if not notion_data:
return None
query_words = set(query.lower().split())
best_match = None
max_score = 0
for item in notion_data:
keywords = set(item.get("keywords", []))
score = len(query_words.intersection(keywords))
if score > max_score:
max_score = score
best_match = item
return best_match["content"] if best_match else None
# --- 2. LLM PROVIDER (Gemini - No changes here) ---
model = genai.GenerativeModel('gemini-1.5-flash')
def generate_response(query: str, context: str) -> str:
prompt = f"""
You are a helpful and friendly campus assistant chatbot.
Use the following piece of context to answer the user's question.
If the context doesn't contain the answer, state that you don't have information on that topic.
Context: "{context or 'No context available.'}"
Question: "{query}"
Answer:
"""
try:
response = model.generate_content(prompt)
return response.text
except Exception as e:
return f"Error generating response: {e}"
# --- 3. STREAMLIT UI (No changes here) ---
st.set_page_config(page_title="Campus Helper Bot", page_icon="🤖")
st.title("🤖 Campus Helper Bot")
st.caption("Your AI-powered guide, now connected to Notion!")
if "messages" not in st.session_state:
st.session_state.messages = [
{"role": "assistant", "content": "Hello! I'm now connected to a live Notion database. How can I help?"}
]
for message in st.session_state.messages:
with st.chat_message(message["role"]):
st.markdown(message["content"])
if prompt := st.chat_input("Ask about fee deadlines, scholarships, etc."):
st.session_state.messages.append({"role": "user", "content": prompt})
with st.chat_message("user"):
st.markdown(prompt)
with st.chat_message("assistant"):
with st.spinner("Searching Notion..."):
context = get_context(prompt)
response = generate_response(prompt, context)
st.markdown(response)
st.session_state.messages.append({"role": "assistant", "content": response}) |