Spaces:
Sleeping
Sleeping
import streamlit as st | |
import os | |
import time | |
import re | |
import requests | |
from PIL import Image | |
from io import BytesIO | |
from openai import OpenAI | |
# ------------------ App Configuration ------------------ | |
st.set_page_config(page_title="Schlaeger Forrestdale DocAIA", layout="wide") | |
st.title("π Document AI Assistant") | |
st.caption("Chat with an AI Assistant on your construction documents") | |
# ------------------ Load API Key and Assistant ID ------------------ | |
OPENAI_API_KEY = os.environ.get("OPENAI_API_KEY") | |
ASSISTANT_ID = os.environ.get("ASSISTANT_ID") | |
if not OPENAI_API_KEY or not ASSISTANT_ID: | |
st.error("β Missing secrets. Please set both OPENAI_API_KEY and ASSISTANT_ID in Hugging Face Space secrets.") | |
st.stop() | |
client = OpenAI(api_key=OPENAI_API_KEY) | |
# ------------------ Session State Initialization ------------------ | |
if "messages" not in st.session_state: | |
st.session_state.messages = [] | |
if "thread_id" not in st.session_state: | |
st.session_state.thread_id = None | |
if "image_url" not in st.session_state: | |
st.session_state.image_url = None | |
if "image_updated" not in st.session_state: | |
st.session_state.image_updated = False | |
# ------------------ Sidebar Controls ------------------ | |
st.sidebar.header("π§ Settings") | |
if st.sidebar.button("π Clear Chat"): | |
st.session_state.messages = [] | |
st.session_state.thread_id = None | |
st.session_state.image_url = None | |
st.session_state.image_updated = False | |
st.rerun() | |
show_image = st.sidebar.checkbox("π Show Document Image", value=True) | |
# ------------------ Layout: Image + Chat ------------------ | |
col1, col2 = st.columns([1, 2]) | |
# ------------------ Left Panel: Image ------------------ | |
with col1: | |
if show_image and st.session_state.image_url: | |
with st.spinner("πΌοΈ Loading image preview..."): | |
try: | |
response = requests.get(st.session_state.image_url) | |
response.raise_for_status() | |
img = Image.open(BytesIO(response.content)) | |
st.image(img, caption="π Extracted Page", use_container_width=True) | |
st.session_state.image_updated = False | |
except Exception as e: | |
st.warning(f"β οΈ Failed to load image from URL:\n{st.session_state.image_url}\n\nError: {e}") | |
# ------------------ Right Panel: Chat ------------------ | |
with col2: | |
prompt = st.chat_input("Type your question about the document or choose from FAQs below...") | |
# Display chat history in a cleaner bubble layout | |
paired_messages = [] | |
buffer = [] | |
for msg in st.session_state.messages: | |
buffer.append(msg) | |
if msg["role"] == "assistant" and len(buffer) == 2: | |
paired_messages.append(buffer.copy()) | |
buffer.clear() | |
if buffer: | |
paired_messages.append(buffer.copy()) | |
for pair in reversed(paired_messages): | |
for msg in pair: | |
with st.chat_message(msg["role"]): | |
st.markdown(f"{msg['content']}", unsafe_allow_html=True) | |
if prompt: | |
st.session_state.messages.append({"role": "user", "content": prompt}) | |
try: | |
if st.session_state.thread_id is None: | |
thread = client.beta.threads.create() | |
st.session_state.thread_id = thread.id | |
client.beta.threads.messages.create( | |
thread_id=st.session_state.thread_id, | |
role="user", | |
content=prompt | |
) | |
run = client.beta.threads.runs.create( | |
thread_id=st.session_state.thread_id, | |
assistant_id=ASSISTANT_ID | |
) | |
with st.spinner("π€ Assistant is thinking..."): | |
while True: | |
run_status = client.beta.threads.runs.retrieve( | |
thread_id=st.session_state.thread_id, | |
run_id=run.id | |
) | |
if run_status.status == "completed": | |
break | |
time.sleep(1) | |
messages = client.beta.threads.messages.list(thread_id=st.session_state.thread_id) | |
assistant_message = None | |
for message in reversed(messages.data): | |
if message.role == "assistant": | |
assistant_message = message.content[0].text.value | |
break | |
st.session_state.messages.append({"role": "assistant", "content": assistant_message}) | |
match = re.search(r'Document Reference:\s+(.+?),\s+Page\s+(\d+)', assistant_message) | |
if match: | |
doc_name = match.group(1).strip() | |
page = int(match.group(2)) | |
page_str = f"{page:04d}" | |
corrected_url = ( | |
f"https://raw.githubusercontent.com/AndrewLORTech/c2ozschlaegerforrestdale/main/" | |
f"{doc_name}/{doc_name}_page_{page_str}.png" | |
) | |
st.session_state.image_url = corrected_url | |
st.session_state.image_updated = True | |
st.rerun() | |
except Exception as e: | |
st.error(f"β Error: {str(e)}") | |
# ------------------ Suggested FAQs Section ------------------ | |
st.markdown(""" | |
<hr> | |
<b>π§ Suggested FAQs:</b> | |
- What are the mechanical services on page 3? | |
- Summarize the demolition schedule. | |
- List the safety compliance steps. | |
""") | |