File size: 4,910 Bytes
8c4492e
13919c8
 
 
2943948
 
 
13919c8
b386f62
8c4492e
 
 
3e6c499
c1043ca
3e6c499
8c4492e
 
c1043ca
8c4492e
2943948
8c4492e
c1043ca
8c4492e
c1043ca
8c4492e
 
 
 
 
 
 
 
 
c1043ca
8c01614
8c4492e
 
 
 
 
 
 
c1043ca
8c4492e
 
3e6c499
f534be4
9c9251a
3e6c499
9c9251a
 
2943948
 
 
 
 
 
 
 
9c9251a
3e6c499
9c9251a
74c6fff
 
3e6c499
eeb4027
 
 
 
 
 
 
 
 
 
3e6c499
 
 
 
f534be4
74c6fff
9c9251a
 
 
3e6c499
9c9251a
 
 
 
2943948
53fcb59
74c6fff
53fcb59
 
 
 
3e6c499
53fcb59
74c6fff
53fcb59
 
 
3e6c499
6e52561
9c9251a
53fcb59
74c6fff
53fcb59
 
9c9251a
 
 
 
3e6c499
74c6fff
9c9251a
 
 
 
3bbf4ab
 
9c9251a
bcaf273
2943948
bcaf273
3e6c499
bcaf273
 
 
 
 
f534be4
3e6c499
57d0c38
8c4492e
9c9251a
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
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
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="Document AI Assistant", layout="wide")
st.title("πŸ“„ Document AI Assistant")
st.caption("Chat with an AI Assistant on your medical/pathology or 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("❌ Missing secrets. Please set both OPENAI_API_KEY and ASSISTANT_ID in your 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("πŸ”§ Settings")
if st.sidebar.button("πŸ”„ 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("πŸ“– 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:
        try:
            response = requests.get(st.session_state.image_url)
            response.raise_for_status()
            img = Image.open(BytesIO(response.content))
            st.image(img, caption="πŸ“‘ Extracted Page", use_container_width=True)
            st.session_state.image_updated = False
        except Exception as e:
            st.warning(f"⚠️ 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...")

    # Display chat history
    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.write(msg["content"])

    if prompt:
        st.session_state.messages.append({"role": "user", "content": prompt})

        try:
            # Create thread
            if st.session_state.thread_id is None:
                thread = client.beta.threads.create()
                st.session_state.thread_id = thread.id

            # Send user prompt
            client.beta.threads.messages.create(
                thread_id=st.session_state.thread_id,
                role="user",
                content=prompt
            )

            # Run assistant
            run = client.beta.threads.runs.create(
                thread_id=st.session_state.thread_id,
                assistant_id=ASSISTANT_ID
            )

            # Wait for completion
            with st.spinner("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)

            # Get assistant reply
            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})

            # βœ… Extract GitHub raw image URL
            image_match = re.search(
                r'https://raw\.githubusercontent\.com/AndrewLORTech/[a-zA-Z0-9_\-/]+\.png',
                assistant_message
            )
            if image_match:
                st.session_state.image_url = image_match.group(0)
                st.session_state.image_updated = True

            st.rerun()

        except Exception as e:
            st.error(f"❌ Error: {str(e)}")