import gradio as gr

from src.inference import generate, list_voices

LANGUAGES = [
    "en", "es", "fr", "de", "it", "pt", "pl", "tr", "ru",
    "nl", "cs", "ar", "zh-cn", "ja", "hu", "ko", "hi"
]

VOICES = list_voices()

EXAMPLES = [
    ["Hello! Welcome to FroxAI Flex-Audio.", "en"],
    ["This is a multilingual text to speech demonstration.", "en"],
    ["नमस्ते! यह FroxAI Flex-Audio का डेमो है।", "hi"],
    ["Bonjour ! Bienvenue dans la démonstration FroxAI.", "fr"],
]


def run(text, language, voice):
    if not text or not text.strip():
        return None, "Enter some text first."

    if not voice:
        return None, "No voices found — make sure voice_refs/ is included in the repo."

    try:
        path = generate(text.strip(), language, voice)
        return path, f"Generated with voice '{voice}' in language '{language}'."
    except Exception as e:
        return None, f"Error: {e}"


with gr.Blocks(title="FroxAI Flex-Audio") as demo:
    gr.Markdown(
        """
# 🦊 FroxAI — Flex-Audio

**Fine-tuned XTTS-v2 text-to-speech with voice cloning.**

Enter text, choose a language and voice, then click **Generate**.

> ⚠️ This model inherits the base model's **CPML license (non-commercial use only)**.
"""
    )

    with gr.Row():
        with gr.Column():
            text_in = gr.Textbox(
                label="Text",
                lines=5,
                placeholder="Type something to speak..."
            )

            language_in = gr.Dropdown(
                choices=LANGUAGES,
                value="en",
                label="Language"
            )

            voice_in = gr.Dropdown(
                choices=VOICES,
                value=VOICES[0] if VOICES else None,
                label="Voice"
            )

            submit = gr.Button("🔊 Generate", variant="primary")

        with gr.Column():
            audio_out = gr.Audio(label="Generated Audio", type="filepath")
            status = gr.Textbox(label="Status", interactive=False)

    gr.Markdown("## 🎙️ Demo examples")
    gr.Examples(
        examples=EXAMPLES,
        inputs=[text_in, language_in],
        label="Click an example to load it, then select a voice and press Generate."
    )

    submit.click(
        run,
        inputs=[text_in, language_in, voice_in],
        outputs=[audio_out, status]
    )

    text_in.submit(
        run,
        inputs=[text_in, language_in, voice_in],
        outputs=[audio_out, status]
    )


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