app.py

import os
import shutil
from fastapi import FastAPI, Request, HTTPException, Depends
from fastapi.responses import JSONResponse
from fastapi.middleware.cors import CORSMiddleware
from sentence_transformers import SentenceTransformer, util
import torch
import requests
from typing import List, Dict, Optional
from pydantic import BaseModel

# Rate limiting
from slowapi import Limiter, _rate_limit_exceeded_handler
from slowapi.util import get_remote_address
from slowapi.errors import RateLimitExceeded

# Configuration
class Config:
    SUPABASE_URL = "https://olbjfxlclotxtnpjvpfj.supabase.co"
    SUPABASE_KEY = "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJzdXBhYmFzZSIsInJlZiI6Im9sYmpmeGxjbG90eHRucGp2cGZqIiwicm9sZSI6ImFub24iLCJpYXQiOjE3NTIyMzYwMDEsImV4cCI6MjA2NzgxMjAwMX0.7q_o5DCFEAAysnWXMChH4MI5qNhIVc4OgpT5JvgYxc0"
    MODEL_NAME = "sentence-transformers/paraphrase-MiniLM-L3-v2"
    SIMILARITY_THRESHOLD = 0.7
    HF_CACHE = "/tmp/hf"
    RATE_LIMIT = "10/minute"

# Initialize FastAPI
app = FastAPI(title="Biruu Chatbot API", version="1.0.0")

# CORS Middleware
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_methods=["*"],
    allow_headers=["*"],
)

# Rate Limiter
limiter = Limiter(key_func=get_remote_address)
app.state.limiter = limiter
app.add_exception_handler(RateLimitExceeded, _rate_limit_exceeded_handler)

# Setup Hugging Face cache
os.makedirs(Config.HF_CACHE, exist_ok=True)
os.environ["TRANSFORMERS_CACHE"] = Config.HF_CACHE
os.environ["HF_HOME"] = Config.HF_CACHE

# Clean up locked cache
lock_file = f"{Config.HF_CACHE}/models--{Config.MODEL_NAME.replace('/', '--')}.lock"
if os.path.exists(lock_file):
    os.remove(lock_file)
model_cache = f"{Config.HF_CACHE}/models--{Config.MODEL_NAME.replace('/', '--')}"
if os.path.exists(model_cache):
    shutil.rmtree(model_cache, ignore_errors=True)

# Initialize model
try:
    model = SentenceTransformer(Config.MODEL_NAME)
except Exception as e:
    raise RuntimeError(f"Failed to load model: {str(e)}")

# Pydantic Models
class ChatMessage(BaseModel):
    admin_id: str
    session_id: str
    is_bot: bool
    is_admin: bool
    message: str

class DeleteRequest(BaseModel):
    id: Optional[str] = None
    uid: Optional[str] = None

# Helper Functions
def make_supabase_request(
    method: str,
    endpoint: str,
    params: Optional[Dict] = None,
    data: Optional[Dict] = None
) -> requests.Response:
    """Generic function to make Supabase API requests"""
    url = f"{Config.SUPABASE_URL}{endpoint}"
    headers = {
        "apikey": Config.SUPABASE_KEY,
        "Authorization": f"Bearer {Config.SUPABASE_KEY}",
        "Content-Type": "application/json"
    }
    
    try:
        if method == "GET":
            response = requests.get(url, headers=headers, params=params)
        elif method == "POST":
            response = requests.post(url, headers=headers, json=data)
        elif method == "DELETE":
            response = requests.delete(url, headers=headers)
        else:
            raise ValueError("Unsupported HTTP method")
        
        response.raise_for_status()
        return response
    except requests.exceptions.RequestException as e:
        raise HTTPException(
            status_code=500,
            detail=f"Supabase request failed: {str(e)}"
        )

def get_faq_from_supabase(admin_id: str) -> List[Dict]:
    """Get FAQ items for a specific admin"""
    try:
        response = make_supabase_request(
            "GET",
            "/rest/v1/faq_items",
            params={"admin_id": f"eq.{admin_id}"}
        )
        return response.json()
    except HTTPException:
        return []

# API Endpoints
@app.post("/predict")
async def predict(request: Request):
    try:
        body = await request.json()
        admin_id = body.get("data", [None, None])[0]
        question = body.get("data", [None, None])[1]

        if not admin_id or not question:
            return JSONResponse(
                {"data": ["Admin ID atau pertanyaan tidak valid"]},
                status_code=400
            )

        # Get FAQs for this admin
        faqs = get_faq_from_supabase(admin_id)
        if not faqs:
            return {"data": ["Maaf, belum ada FAQ yang tersedia."]}

        # Process question
        questions = [f["question"] for f in faqs]
        answers = [f["answer"] for f in faqs]

        # Get embeddings
        embeddings = model.encode(questions, convert_to_tensor=True)
        query_embedding = model.encode(question, convert_to_tensor=True)
        
        # Calculate similarity
        similarity = util.pytorch_cos_sim(query_embedding, embeddings)
        best_idx = torch.argmax(similarity).item()
        best_score = similarity[0][best_idx].item()
        
        # Threshold similarity (minimal 0.3)
        if best_score < 0.3:
            return {"data": ["Maaf, saya tidak mengerti pertanyaan Anda"]}
            
        return {"data": [answers[best_idx]]}

    except Exception as e:
        print(f"Error in prediction: {str(e)}")
        return JSONResponse(
            {"data": ["Terjadi kesalahan saat memproses pertanyaan"]},
            status_code=500
        )


@app.post("/save_chat")
async def save_chat(chat: ChatMessage):
    """Save chat message to database"""
    try:
        response = make_supabase_request(
            "POST",
            "/rest/v1/chat_logs",
            data={
                "admin_id": chat.admin_id,
                "session_id": chat.session_id,
                "is_bot": chat.is_bot,
                "is_admin": chat.is_admin,
                "message": chat.message
            }
        )
        saved_data = response.json()[0]
        return {
            "message": "Pesan berhasil disimpan",
            "id": saved_data["id"]
        }
    except HTTPException as e:
        raise e
    except Exception as e:
        raise HTTPException(
            status_code=500,
            detail=f"Failed to save chat: {str(e)}"
        )

@app.get("/chat_history")
async def get_chat_history(admin_id: str, session_id: str):
    """Get chat history for specific session"""
    try:
        response = make_supabase_request(
            "GET",
            "/rest/v1/chat_logs",
            params={
                "admin_id": f"eq.{admin_id}",
                "or": f"(session_id.eq.{session_id},is_bot.eq.true)",
                "order": "created_at.asc"
            }
        )
        return response.json()
    except HTTPException as e:
        raise e
    except Exception as e:
        raise HTTPException(
            status_code=500,
            detail=f"Failed to get chat history: {str(e)}"
        )

@app.post("/delete_chat")
async def delete_chat(request: DeleteRequest):
    """Delete specific chat message"""
    if not request.id:
        raise HTTPException(
            status_code=400,
            detail="Message ID is required"
        )

    try:
        make_supabase_request(
            "DELETE",
            f"/rest/v1/chat_logs?id=eq.{request.id}"
        )
        return {"message": f"Pesan dengan ID {request.id} berhasil dihapus."}
    except HTTPException as e:
        raise e
    except Exception as e:
        raise HTTPException(
            status_code=500,
            detail=f"Failed to delete message: {str(e)}"
        )

@app.post("/delete_all_by_uid")
async def delete_all_by_uid(request: DeleteRequest):
    """Delete all messages for specific user"""
    if not request.uid:
        raise HTTPException(
            status_code=400,
            detail="UID is required"
        )

    try:
        make_supabase_request(
            "DELETE",
            f"/rest/v1/chat_logs?admin_id=eq.{request.uid}"
        )
        return {"message": f"Semua pesan untuk UID {request.uid} berhasil dihapus."}
    except HTTPException as e:
        raise e
    except Exception as e:
        raise HTTPException(
            status_code=500,
            detail=f"Failed to delete messages: {str(e)}"
        )

@app.get("/health")
async def health_check():
    """Health check endpoint"""
    return {"status": "healthy", "version": "1.0.0"}