app.py

import streamlit as st
from openai import AzureOpenAI, AssistantEventHandler
import os
import re
from typing_extensions import override

# Set up your Azure OpenAI API credentials
# Configuration
api_key = os.environ.get("OPENAI_API_KEY")
azure_endpoint = os.environ.get("OPENAI_AZURE_ENDPOINT")
deployment_name = "gpt-4o"  # Replace with your deployment name
api_version = "2024-05-01-preview"  # Ensure this matches your Azure OpenAI resource

# Initialize the Azure OpenAI client
client = AzureOpenAI(
    api_key=api_key,
    azure_endpoint=azure_endpoint,
    api_version=api_version
)

# Retrieve the assistant
assistant_id = "asst_c54LY1ZIel6lOG4CHsJ5xhaw"  # Replace with your assistant ID
assistant = client.beta.assistants.retrieve(assistant_id)


# Streamlit app setup
st.set_page_config(page_title="ZimmerBot", page_icon=":hospital:", layout="wide")
st.image("https://download.logo.wine/logo/Zimmer_Biomet/Zimmer_Biomet-Logo.wine.png", width=200)


# Initialize session state
if "thread_id" not in st.session_state:
    thread = client.beta.threads.create()
    st.session_state.thread_id = thread.id

if "messages" not in st.session_state:
    st.session_state.messages = []

if "placeholders" not in st.session_state:
    st.session_state.placeholders = []

def clean_content(content):
    # Remove and style citations
    content = re.sub(r'【\d+(?::\d+)?†source】', '', content)
    # Prettify [1][2] style citations to [1, 2]
    content = re.sub(r'\[(\d+)\](?:\[(\d+)\])+', lambda m: f"[{', '.join(sorted(set(m.groups())))}]", content)
    return content.strip()

def display_message(role, content, citations=None):
    with st.chat_message(role):
        st.markdown(content)
        if citations:
            st.markdown("---")
            st.markdown("**Annotations:**")
            for citation in citations:
                st.markdown(f"- {citation}")

def process_message_content(message):
    annotations = []
    content_text = ''
    citations = []
    citation_map = {}

    # Loop over all content blocks in message.content
    for message_content in message.content:
        # Extract text
        text_block = message_content.text.value
        content_text += text_block

        # Extract annotations
        content_annotations = getattr(message_content, 'annotations', [])
        annotations.extend(content_annotations)

    # Now process annotations
    for annotation in annotations:
        annotation_text = annotation.text.value  # Extract the string value
        if hasattr(annotation, 'file_citation'):
            file_citation = annotation.file_citation
            cited_file = client.files.retrieve(file_citation.file_id)
            citation = f"{cited_file.filename}"
            if citation not in citation_map:
                citation_map[citation] = len(citation_map) + 1
            content_text = content_text.replace(annotation_text, f"[{citation_map[citation]}]")
            citations.append(f"[{citation_map[citation]}] {citation}")
        elif hasattr(annotation, 'file_path'):
            file_path = annotation.file_path
            cited_file = client.files.retrieve(file_path.file_id)
            citation = f"{cited_file.filename}"
            if citation not in citation_map:
                citation_map[citation] = len(citation_map) + 1
            content_text = content_text.replace(annotation_text, f"[{citation_map[citation]}]")
            citations.append(f"[{citation_map[citation]}] {citation}")

    cleaned_content = clean_content(content_text)
    return cleaned_content, citations

def submit_message(message):
    # Add user message to session state and display it
    st.session_state.messages.append({"role": "user", "content": message})
    display_message("user", message)

    # Send message to the assistant
    client.beta.threads.messages.create(
        thread_id=st.session_state.thread_id,
        role="user",
        content=message
    )

    # Create a placeholder for the assistant's response
    with st.chat_message("assistant"):
        message_placeholder = st.empty()
        full_response = ""

        # Define EventHandler to handle the streaming events
        class StreamEventHandler(AssistantEventHandler):
            def __init__(self):
                super().__init__()
                self.content = ""

            @override
            def on_text_delta(self, delta, snapshot):
                self.content += delta.value
                # Update the message placeholder
                message_placeholder.markdown(self.content)

        # Create and start the run with streaming enabled
        event_handler = StreamEventHandler()
        with client.beta.threads.runs.stream(
            thread_id=st.session_state.thread_id,
            assistant_id=assistant.id,
            event_handler=event_handler,
        ) as stream_run:
            stream_run.until_done()

        # Store the streamed content
        streamed_content = event_handler.content

    # After streaming is done, process annotations
    messages = client.beta.threads.messages.list(
        thread_id=st.session_state.thread_id
    )

    # Find the last assistant message
    assistant_messages = [msg for msg in messages.data if msg.role == 'assistant']
    if assistant_messages:
        assistant_message = assistant_messages[-1]
        processed_content, citations = process_message_content(assistant_message)

        # Combine streamed content with processed content
        final_content = streamed_content if streamed_content else processed_content

        # Add assistant response to session state
        st.session_state.messages.append({
            "role": "assistant",
            "content": final_content,
            "citations": citations
        })

        # Update the message placeholder with the final content
        message_placeholder.markdown(final_content)
        if citations:
            with message_placeholder.container():
                st.markdown("---")
                st.markdown("**Annotations:**")
                for citation in citations:
                    st.markdown(f"- {citation}")
    else:
        st.error("No assistant response was found.")

# Display chat history
for message in st.session_state.messages:
    display_message(message["role"], message["content"], message.get("citations"))

# User input
user_input = st.chat_input("What would you like to ask?")
if user_input:
    submit_message(user_input)