# app.py
import gradio as gr
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.naive_bayes import MultinomialNB

# Sample documents and labels
documents = [
    "I love data science and machine learning",      # Positive
    "Data science is fun and exciting",              # Positive
    "Machine learning is a part of data science",    # Positive
    "I hate dealing with messy data",                # Negative
    "Data preprocessing is so boring",               # Negative
    "I dislike coding errors and bugs"               # Negative
]
labels = [1, 1, 1, 0, 0, 0]  # 1 = Positive, 0 = Negative

# Train model
vectorizer = TfidfVectorizer()
X = vectorizer.fit_transform(documents)
model = MultinomialNB()
model.fit(X, labels)

# Prediction function
def predict_sentiment(text):
    X_new = vectorizer.transform([text])
    prediction = model.predict(X_new)[0]
    return "Positive 😊" if prediction == 1 else "Negative 😠"

# Gradio interface
interface = gr.Interface(
    fn=predict_sentiment,
    inputs=gr.Textbox(lines=2, placeholder="Enter your sentence here..."),
    outputs="text",
    title="Simple Test Sentiment Classifier (Beta)",
    description="Enter a sentence to classify it as Positive 😊 or Negative 😠."
)

# Run the app
if __name__ == "__main__":
    interface.launch()