# app.py
import streamlit as st
import pickle
import re
from nltk.tokenize import word_tokenize
from nltk.corpus import stopwords
import nltk


nltk.download('punkt')
nltk.download('stopwords')

# Load model dan tools
with open("model_sentiment.pkl", "rb") as f:
    model = pickle.load(f)

with open("vectorizer.pkl", "rb") as f:
    vectorizer = pickle.load(f)

with open("label_encoder.pkl", "rb") as f:
    le = pickle.load(f)

stop_words = set(stopwords.words('indonesian'))

# Preprocessing function
def preprocess(text):
    text = text.lower()
    text = re.sub(r'[^a-zA-Z\s]', '', text)
    tokens = word_tokenize(text)
    tokens = [t for t in tokens if t not in stop_words]
    return ' '.join(tokens)

# UI
st.title("🧠 Sentiment Analysis Komentar Bahasa Indonesia")
st.markdown("Masukkan komentar di bawah ini:")

user_input = st.text_area("💬 Komentar")

if st.button("Prediksi Sentimen"):
    if user_input.strip() == "":
        st.warning("Komentar tidak boleh kosong!")
    else:
        cleaned = preprocess(user_input)
        vec = vectorizer.transform([cleaned])
        pred = model.predict(vec)
        label = le.inverse_transform(pred)[0]
        
        if label == "positif":
            st.success(f"Hasil: {label.capitalize()} 👍")
        elif label == "negatif":
            st.error(f"Hasil: {label.capitalize()} 👎")
        else:
            st.info(f"Hasil: {label.capitalize()} 😐")