# ============================================================
#  HuggingFace Space — Emotion Classifier Demo
#  File: app.py  (deploy as a Gradio Space)
# ============================================================
# How to deploy:
#   1. Go to https://huggingface.co/spaces → Create new Space
#   2. Choose SDK: Gradio
#   3. Add this file as app.py
#   4. Add requirements.txt (contents below in comment)
#   5. Space will auto-build and launch
#
# requirements.txt contents:
#   transformers>=4.40.0
#   torch>=2.1.0
#   gradio>=4.0.0
# ============================================================

import gradio as gr
from transformers import pipeline

# ── Load model from HuggingFace Hub ──────────────────────────────────────────
MODEL_REPO = "your-hf-username/emotion-classifier-distilbert"   # ← update this

EMOTION_EMOJIS = {
    "sadness":  "😢",
    "anger":    "😠",
    "love":     "❤️",
    "surprise": "😲",
    "fear":     "😨",
    "joy":      "😊",
}

EMOTION_COLORS = {
    "sadness":  "#4a90d9",
    "anger":    "#e74c3c",
    "love":     "#e91e8c",
    "surprise": "#f39c12",
    "fear":     "#8e44ad",
    "joy":      "#27ae60",
}

print(f"Loading model: {MODEL_REPO} …")
classifier = pipeline(
    "text-classification",
    model=MODEL_REPO,
    return_all_scores=True,
)
print("Model loaded ✅")

# ── Inference function ────────────────────────────────────────────────────────
def predict_emotion(text: str):
    if not text or not text.strip():
        return "⚠️ Please enter some text.", {}

    results = classifier(text.strip())[0]  # list of {label, score}
    scores  = {r["label"]: round(r["score"] * 100, 2) for r in results}

    best        = max(results, key=lambda x: x["score"])
    best_label  = best["label"]
    best_score  = best["score"] * 100
    emoji       = EMOTION_EMOJIS.get(best_label, "🤔")

    summary = (
        f"**{emoji} {best_label.capitalize()}**  "
        f"({best_score:.1f}% confidence)"
    )
    return summary, scores

# ── Example sentences ─────────────────────────────────────────────────────────
EXAMPLES = [
    ["I feel so happy and grateful today!"],
    ["I am really angry about what happened."],
    ["I miss my dog so much, it breaks my heart."],
    ["Oh wow, I cannot believe that just happened!"],
    ["I'm terrified of what might come next."],
    ["I love spending time with you."],
]

# ── Gradio UI ─────────────────────────────────────────────────────────────────
with gr.Blocks(
    theme=gr.themes.Soft(),
    title="Emotion Classifier",
    css="""
        .result-box { font-size: 1.4rem; padding: 12px; border-radius: 8px; }
        footer { display: none !important; }
    """,
) as demo:

    gr.Markdown(
        """
        # 🎭 Emotion Classifier
        Detects **6 emotions** in English text using a fine-tuned DistilBERT model.
        """
    )

    with gr.Row():
        with gr.Column(scale=2):
            text_input = gr.Textbox(
                label="Enter text",
                placeholder="Type a sentence and click Analyze…",
                lines=4,
            )
            analyze_btn = gr.Button("Analyze Emotion", variant="primary")

        with gr.Column(scale=2):
            result_label = gr.Markdown(
                label="Top Prediction",
                elem_classes=["result-box"],
            )
            scores_plot = gr.Label(
                label="Confidence Scores (%)",
                num_top_classes=6,
            )

    gr.Examples(
        examples=EXAMPLES,
        inputs=text_input,
        label="Example sentences",
    )

    gr.Markdown(
        """
        ---
        **Labels:** 😢 Sadness · 😠 Anger · ❤️ Love · 😲 Surprise · 😨 Fear · 😊 Joy

        Model: `distilbert-base-uncased` fine-tuned for 10 epochs on ~19 000 labelled sentences.
        """
    )

    analyze_btn.click(
        fn=predict_emotion,
        inputs=text_input,
        outputs=[result_label, scores_plot],
    )
    text_input.submit(
        fn=predict_emotion,
        inputs=text_input,
        outputs=[result_label, scores_plot],
    )

if __name__ == "__main__":
    demo.launch()