chatbot.py

import os
import uuid
import asyncio
import streamlit as st
from html import escape

import edge_tts
from groq import Groq
from langchain_groq import ChatGroq
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser

# Load API key from environment
GROQ_API_KEY = os.getenv("GROQ_API_KEY")


# ── Core Bot Logic ─────────────────────────────
class CodeAssistantBot:
    def __init__(self):
        self.client = Groq(api_key=GROQ_API_KEY)
        self.model = ChatGroq(model="llama-3.3-70b-versatile", temperature=0.6)

        self.analysis_prompt = ChatPromptTemplate.from_messages([
            ("system", "You are a skilled coding assistant. Keep answers relevant and concise."),
            ("user", "Code: {code}\nOutput: {output}\nError: {error}\n"
                     "Summary: {summary}\nRecent: {recent}\nQuestion: {question}")
        ])

        self.summary_prompt = ChatPromptTemplate.from_messages([
            ("system", "Summarize key technical points from the conversation so far."),
            ("user", "Conversation: {conversation}")
        ])

        self.voice_prompt = ChatPromptTemplate.from_messages([
            ("system", "You are a friendly narrator. Explain the answer clearly and casually."),
            ("user", "Code: {code}\nOutput: {output}\nError: {error}\n"
                     "Conversation so far: {summary}\nAnswer to explain: {answer}")
        ])

    def analyze_code(self, code, output, error, question, summary="", history=None):
        parser = StrOutputParser()
        recent = "\n".join([f"User: {q}\nBot: {a}" for q, a in (history or [])[-4:]])
        return (self.analysis_prompt | self.model | parser).invoke({
            'code': code, 'output': output, 'error': error,
            'summary': summary, 'recent': recent, 'question': question
        })

    def narrate_response(self, code, output, error, answer, summary=""):
        parser = StrOutputParser()
        return (self.voice_prompt | self.model | parser).invoke({
            'code': code, 'output': output, 'error': error,
            'summary': summary, 'answer': answer
        })


# ── Text to Speech ─────────────────────────────
async def text_to_speech(text, filename):
    voice = "fr-FR-VivienneMultilingualNeural"
    communicator = edge_tts.Communicate(text, voice)
    await communicator.save(filename)


# ── Chat UI Logic ──────────────────────────────
def render_chatbot(code, output, error):
    st.markdown("""
    <style>
        .chat-container {
            max-height: 60vh;
            overflow-y: auto;
            padding: 1rem 0.5rem 1rem 1rem;
            border: 1px solid #ddd;
            border-radius: 8px;
            background-color: #f9f9f9;
        }
        .chat-message {
            margin-bottom: 1rem;
            word-wrap: break-word;
        }
        .user-message {
            font-weight: bold;
            color: #1a73e8;
        }
        .bot-message pre {
            background-color: #f0f0f0;
            padding: 0.5rem;
            border-radius: 5px;
            overflow-x: auto;
        }
    </style>
    """, unsafe_allow_html=True)

    # Session setup
    st.session_state.setdefault('conversation', [])
    st.session_state.setdefault('chat_summary', "")
    st.session_state.setdefault('chat_display_count', 5)
    st.session_state.setdefault('narrated_audio', {})

    # Input row
    col1, col2 = st.columns([4, 1])
    with col1:
        question = st.text_input("Ask something about your code...", key="chat_input")
    with col2:
        send = st.button("🚀")

    # Bot response
    if send and question:
        bot = CodeAssistantBot()
        history = st.session_state.conversation[-4:]
        summary = st.session_state.chat_summary
        answer = bot.analyze_code(code, output, error, question, summary, history)
        st.session_state.conversation.append((question, answer))
        st.session_state.chat_display_count = 5

        if len(st.session_state.conversation) >= 3:
            try:
                full_chat = "\n".join([f"User: {q}\nBot: {a}" for q, a in st.session_state.conversation[-10:]])
                summarizer = bot.summary_prompt | bot.model | StrOutputParser()
                st.session_state.chat_summary = summarizer.invoke({'conversation': full_chat})
            except:
                pass

    # Display messages
    visible = list(reversed(st.session_state.conversation[-st.session_state.chat_display_count:]))

    for idx, (q, a) in enumerate(visible):
        st.markdown(f'<div class="chat-message user-message">{escape(q)}</div>', unsafe_allow_html=True)

        def format_response(text):
            parts = text.split("```")
            result = ""
            for i, part in enumerate(parts):
                if i % 2 == 1:
                    lines = part.splitlines()
                    if lines and lines[0].isalpha():
                        lines = lines[1:]
                    result += f'<pre><code>{escape("\\n".join(lines))}</code></pre>'
                else:
                    result += escape(part)
            return result

        st.markdown(f'<div class="chat-message bot-message">{format_response(a)}</div>', unsafe_allow_html=True)

        # Narration logic
        audio_file = st.session_state.narrated_audio.get((q, a))
        if not audio_file:
            if st.button("🔊 Narrate", key=f"narrate_{idx}"):
                status = st.empty()
                status.info("🧠 Generating narration...")
                bot = CodeAssistantBot()
                narration = bot.narrate_response(code, output, error, a, st.session_state.chat_summary)
                status.info("🎙️ Converting to audio...")
                audio_file = f"audio_{uuid.uuid4().hex}.mp3"
                asyncio.run(text_to_speech(narration, audio_file))
                st.session_state.narrated_audio[(q, a)] = audio_file
                status.success("🔊 Narration ready!")
                st.audio(audio_file, format="audio/mp3", autoplay=True)
        else:
            st.audio(audio_file, format="audio/mp3", autoplay=False)

    if len(visible) < len(st.session_state.conversation):
        if st.button("🔽 Show more"):
            st.session_state.chat_display_count += 5
            st.rerun()

    # Auto-scroll & pause others on audio play
    st.markdown("""
    <script>
    const container = window.parent.document.querySelector('.chat-container');
    if (container) container.scrollTop = container.scrollHeight;

    document.querySelectorAll('audio').forEach(audio => {
        audio.addEventListener('play', function () {
            document.querySelectorAll('audio').forEach(a => {
                if (a !== this) a.pause();
            });
        });
    });
    </script>
    """, unsafe_allow_html=True)