main.py

from fastapi import FastAPI, Form
from fastapi.responses import JSONResponse
from pathlib import Path
from fastapi.responses import JSONResponse
from fastapi.middleware.cors import CORSMiddleware
from rag import build_vector_index, query_from_vector, PDF_DIR
from dotenv import load_dotenv

from utils import *

import threading
import os
import traceback
import logging

# ------------------ INIT ------------------
app = FastAPI(title="Virtual Assistant Chatbot API", version="1.0")

# ------------------ CORS ------------------
origins = [
    "https://chabot.demo.viproject.net",
    "chabot.demo.viproject.net"
]

app.add_middleware(
    CORSMiddleware,
    allow_origins=origins,
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

logger = logging.getLogger(__name__)
logging.basicConfig(level=logging.INFO)

os.makedirs(PDF_DIR, exist_ok=True)

OUTPUT_FOLDER = Path("documents")
OUTPUT_FOLDER.mkdir(exist_ok=True)

# ------------------ load env ------------------
load_dotenv()
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
init_openai(OPENAI_API_KEY)


# ------------------ HELPER ------------------        
def clear_documents_folder():
    """Hapus semua file di folder documents"""
    for file in OUTPUT_FOLDER.iterdir():
        if file.is_file():
            file.unlink()

# ------------------ ROUTES ------------------
@app.get("/")
def root():
    """Cek status API"""
    return {"message": "Virtual Assistant API aktif dan siap digunakan!"}


@app.post("/build-knowledge")
async def build_vector_db():
    """Bangun FAISS vector database dari semua PDF"""
    try:
        result = build_vector_index()
        return JSONResponse(result)
    except Exception as e:
        tb = traceback.format_exc()
        logger.error(f"/build_vector_db error: {e}\n{tb}")
        return JSONResponse({"error": str(e), "traceback": tb}, status_code=500)


@app.post("/web-scraping/")
async def scrape_urls(filename: str = Form(...), urls: str = Form(...), location: str = "server"):
    """
    Ambil semua URL, gabungkan teks → bersihkan → satu PDF utuh
    """
    # Hapus semua file lama
    clear_documents_folder()

    # Parse URL
    url_list = [u.strip() for u in urls.split(",") if u.strip()]
    extracted_texts = [None] * len(url_list)
    threads = []

    # STEP 1: Ekstraksi halaman
    def worker_extract(i, url):
        driver = init_driver_local() if location == "local" else init_driver()
        try:
            extracted_texts[i] = fetch_page_text(driver, url)
        finally:
            driver.quit()

    for i, url in enumerate(url_list):
        t = threading.Thread(target=worker_extract, args=(i, url))
        t.start()
        threads.append(t)

    for t in threads:
        t.join()

    # STEP 2: Gabungkan semua teks
    combined_text = ""
    for url, text in zip(url_list, extracted_texts):
        combined_text += f"===== URL: {url} =====\n{text}\n\n"

    # STEP 3: Bersihkan dengan OpenAI
    cleaned_text = clean_text_with_openai(combined_text)

    # STEP 4: Simpan PDF
    output_file = OUTPUT_FOLDER / f"{filename}.pdf"
    save_to_pdf(cleaned_text, output_file)

    return JSONResponse({"success": True, "pdf_file": str(output_file)})
    

@app.post("/ask")
async def ask_question(
    question: str = Form(...),
    session_id: str = Form(None),
):
    """Ajukan pertanyaan ke dokumen yang sudah diindeks dengan session chat"""
    try:
        result = query_from_vector(
            query=question,
            session_id=session_id,
        )
        return JSONResponse(result)
    except Exception as e:
        tb = traceback.format_exc()
        return JSONResponse({"error": str(e), "traceback": tb}, status_code=500)