backend/app.py

import asyncio
from fastapi import FastAPI, WebSocket
from fastapi.staticfiles import StaticFiles
from fastapi.responses import FileResponse
import uvicorn
import json
from transformers import pipeline
from collections import deque
from collections import defaultdict
import math
import sys
import random
import os
from fastapi.middleware.cors import CORSMiddleware


sys.path.append(".")

app = FastAPI()

reset_timer = asyncio.Event()

# app.mount("/", StaticFiles(directory="frontend", html=True), name="frontend")

app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],  # Bisa diubah ke domain spesifik untuk keamanan
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)


emotion_classifier = pipeline(
    "zero-shot-classification",
    model="MarfinF/marfin_emotion",
    force_download=True
)

clients = {}
chat_history = deque(maxlen=4)

mood_to_genre = {
    "senang": {
        "genre": "pop",
        "word": "Poppin pops!"
    },
    "sedih": {
        "genre": "acoustic",
        "word": "Playing acoustic"
    },
    "marah": {
        "genre": "rock",
        "word": "Rock 'n play!!"
    },
    "cinta": {
        "genre": "romantic",
        "word": "Playing romantic"
    },
    "chill": {
        "genre": "chill",
        "word": "Chill!"
    }
}

genre_to_song = {
    "pop": ["https://sounds.pond5.com/inspirational-motivational-uplifting-acoustic-positive-music-102770115_nw_prev.m4a"],
    "acoustic": ["https://sounds.pond5.com/sad-piano-music-063450274_nw_prev.m4a"],
    "rock": ["https://sounds.pond5.com/powerful-gritty-action-extreme-rock-music-136693652_nw_prev.m4a"],
    "romantic": ["https://sounds.pond5.com/bloom-sweet-tender-delicate-romantic-music-158563013_nw_prev.m4a"],
    "chill": ["https://sounds.pond5.com/fall-chill-music-088328584_nw_prev.m4a"]
}

def detect_emotion(text):
    labels = ["marah", "sedih", "senang", "cinta"]
    result = emotion_classifier(text, candidate_labels=labels)
    top_emotion = result['labels'][0]
    top_scores = result['scores'][0]
    return top_emotion, top_scores

def get_recommendations_by_mood(genre):
    songs = genre_to_song.get(genre, [])

    random.shuffle(songs)
    return songs[:3]  # Return top 3 shuffled songs

def softmax(scores):
    exp_scores = [math.exp(score) for score in scores]
    total = sum(exp_scores)
    return [exp_score / total for exp_score in exp_scores]

# 🔹 Broadcast User List
async def broadcast_user_list():
    user_list = list(clients.keys())
    message = json.dumps({
        "type": "user_list",
        "users": user_list
    })
    for client in clients.values():
        await client.send_text(message)

# 🔹 Periodic Music Recommender every 30 seconds
async def periodic_recommendation():
    while True:
        user_list = list(clients.keys())
        if len(user_list) >= 1:
            sleep_task = asyncio.create_task(asyncio.sleep(60))  # Start sleep

            try:
                await asyncio.wait_for(sleep_task, timeout=60)  # Wait 60s unless interrupted
            except asyncio.TimeoutError:
                pass  # Timer completed without reset
            except asyncio.CancelledError:
                continue  # Timer was reset, restart loop

            if reset_timer.is_set():  # If reset happened, restart countdown
                continue  

            if clients:  # Only run if someone is connected
                if len(chat_history) > 0:
                    # 1. Detect emotion dan ambil (label, score)
                    print("chat history")
                    print(chat_history)
                    emotions = [detect_emotion(msg) for msg in chat_history]
                    print("Detected Emotions:", emotions)

                    # 2. Group by emotion + sum score
                    emotion_score_sum = defaultdict(float)
                    for label, score in emotions:
                        emotion_score_sum[label] += score

                    # 3. Softmax
                    labels = list(emotion_score_sum.keys())
                    scores = list(emotion_score_sum.values())
                    softmax_scores = softmax(scores)

                    # 4. Pair label + softmax_score
                    softmax_result = list(zip(labels, softmax_scores))
                    print("Softmax Result:", softmax_result)

                    # 5. Dominant emotion
                    most_common_emotion = max(softmax_result, key=lambda x: x[1])[0]
                    print("Dominant Emotion:", most_common_emotion)
                    music = mood_to_genre.get(most_common_emotion, mood_to_genre["chill"])
                    music_recommendations = get_recommendations_by_mood(music["genre"])
                    word = music["word"]
                else:
                    music_recommendations = get_recommendations_by_mood("chill")
                    word = "Chill!"

                recommendation_response = {
                    "recommendations": music_recommendations,
                    "genre": word
                }

                for client in clients.values():
                    await client.send_text(json.dumps(recommendation_response))
        else:
            await asyncio.sleep(2)
            await broadcast_user_list()
            await reset_recommendation_timer()
            

async def reset_recommendation_timer():
    """Call this function when you want to reset the timer to 60 seconds."""
    if reset_timer.is_set():  # Check if the timer is running before resetting
        reset_timer.clear()  # Clear the event before setting it again
        reset_timer.set()  # Trigger the reset

@app.on_event("startup")
async def start_recommender():
    asyncio.create_task(periodic_recommendation())

@app.websocket("/chat/{username}")
async def chat_endpoint(websocket: WebSocket, username: str):
    await websocket.accept()
    clients[username] = websocket
    print(f"{username} joined")

    await broadcast_user_list()

    try:
        while True:
            data = await websocket.receive_text()
            message_data = json.loads(data)

            chat_history.append(message_data["message"])
            response = {
                "username": message_data["username"],
                "message": message_data["message"]
            }

            # Broadcast message to all clients
            for client in clients.values():
                await client.send_text(json.dumps(response))

    except Exception as e:
        print(f"{username} disconnected: {e}")
        del clients[username]
        await broadcast_user_list()

@app.get("/mp")
def read_root():
    print("frontend")
    return FileResponse("frontend/index.html")

if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=7860)