app.py

from flask import Flask, request, jsonify
import joblib
import string
import nltk
import os

# Initialize App
app = Flask(__name__)

# --- ROBUST NLTK SETUP ---
# Set NLTK data path to a local folder to avoid permission issues
nltk_data_dir = os.path.join(os.getcwd(), 'nltk_data')
nltk.data.path.append(nltk_data_dir)

def download_nltk_resources():
    resources = ['stopwords', 'wordnet', 'punkt', 'punkt_tab']
    for res in resources:
        try:
            nltk.data.find(f'corpora/{res}')
        except LookupError:
            try:
                nltk.data.find(f'tokenizers/{res}')
            except LookupError:
                print(f"Downloading {res}...")
                nltk.download(res, download_dir=nltk_data_dir, quiet=True)

download_nltk_resources()
from nltk.stem import WordNetLemmatizer
from nltk.corpus import stopwords
# -------------------------

# Load Model
print("Loading model...")
try:
    model = joblib.load('sentiment_model.pkl')
    vectorizer = joblib.load('tfidf_vectorizer.pkl')
    print("Model loaded successfully.")
except Exception as e:
    print(f"CRITICAL ERROR: Could not load model files. {e}")
    model = None

lemmatizer = WordNetLemmatizer()
stop_words = set(stopwords.words('english'))

def preprocess_text(text):
    if not isinstance(text, str): return ""
    text = text.lower()
    text = text.translate(str.maketrans('', '', string.punctuation))
    tokens = nltk.word_tokenize(text)
    clean_tokens = [lemmatizer.lemmatize(word) for word in tokens if word not in stop_words]
    return " ".join(clean_tokens)

@app.route('/predict', methods=['POST'])
def predict():
    if model is None:
        return jsonify({'error': 'Model not loaded properly.'}), 500
        
    try:
        data = request.get_json()
        if not data or 'review_text' not in data:
            return jsonify({'error': 'No review_text provided'}), 400
            
        text = data['review_text']
        clean_text = preprocess_text(text)
        vectorized_text = vectorizer.transform([clean_text])
        prediction = model.predict(vectorized_text)[0]
        
        return jsonify({
            'review': text, 
            'sentiment': "Positive" if prediction == 1 else "Negative"
        })
    except Exception as e:
        print(f"Prediction Error: {e}")
        return jsonify({'error': str(e)}), 500

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000)