import os
import gradio as gr
from groq import Groq
import speech_recognition as sr

# ✅ API Key
GROQ_API_KEY = os.environ.get("GROQ_API_KEY")
if not GROQ_API_KEY:
    raise ValueError("❌ Please add your GROQ_API_KEY in Hugging Face Space Secrets!")

MODEL_NAME = "llama3-8b-8192"

SYSTEM_PROMPT = """
You are Dr Cat, a friendly cat expert.
You help cat owners with advice about cat health, food, behavior and care.
"""

# ✅ Speech recognition
def recognize_speech(audio_file):
    if audio_file is None:
        return gr.update(value="")
    recognizer = sr.Recognizer()
    with sr.AudioFile(audio_file) as source:
        audio = recognizer.record(source)
    try:
        text = recognizer.recognize_google(audio)
        return gr.update(value=text)
    except:
        return gr.update(value="Could not recognize speech.")

# ✅ Show Audio on mic click
def show_audio():
    return gr.update(visible=True)

# ✅ Chat logic — FINAL OUTPUTS: 5
def chat(user_input, image_file, video_file, history):
    if history is None:
        history = []
    context = user_input
    if image_file:
        context += f"\n![Cat Photo]({image_file})"
    if video_file:
        context += f"\n<video controls width='250'><source src='{video_file}' type='video/mp4'></video>"

    messages = [{"role": "system", "content": SYSTEM_PROMPT}] + history
    messages.append({"role": "user", "content": context})

    client = Groq(api_key=GROQ_API_KEY)
    response = client.chat.completions.create(
        model=MODEL_NAME,
        messages=messages,
        temperature=0.7,
        max_tokens=512,
    )

    reply = response.choices[0].message.content
    messages.append({"role": "assistant", "content": reply})

    return "", None, None, messages, messages  # ✅ 5 Outputs!

with gr.Blocks(
    title="🐱 Dr Cat | Your AI Cat Care Companion",
    theme=gr.themes.Soft(primary_hue="pink", secondary_hue="rose"),
    css="""
    body { background: #121212; color: #f5f5f5; }
    .gr-button, .gr-upload-button { background: #ff69b4; color: white; border-radius: 50%; width: 40px; height: 40px; padding: 0; font-size: 20px; }
    .gr-button:hover, .gr-upload-button:hover { background: #ff85c1; }
    .gr-textbox textarea { border-radius: 30px; background: #1e1e1e; color: #f5f5f5; font-size: 16px; }
    .gr-chatbot { border-radius: 16px; background: #1e1e1e; }
    h1 { font-size: 42px; }
    p { font-size: 18px; color: #ccc; }
    """
) as demo:
    gr.Markdown(
        """
        <div style="text-align:center">
            <h1>🐱 Dr Cat</h1>
            <p>Your AI Cat Care Companion — Ask anything about your furry friend 🐾</p>
        </div>
        """
    )

    chatbot = gr.Chatbot(label="🐈 Dr Cat", height=500, type="messages")

    with gr.Row():
        msg = gr.Textbox(placeholder="Ask Dr Cat about your kitty...", scale=8)
        image = gr.UploadButton("📸", file_types=["image"], scale=1)
        video = gr.UploadButton("🎥", file_types=["video"], scale=1)
        mic_btn = gr.Button("🎙️", scale=1)

    audio = gr.Audio(visible=False, type="filepath")

    mic_btn.click(show_audio, None, audio)
    audio.change(recognize_speech, inputs=audio, outputs=msg)

    send_btn = gr.Button("💬 Send")

    state = gr.State([])

    msg.submit(chat, [msg, image, video, state], [msg, image, video, chatbot, state])
    send_btn.click(chat, [msg, image, video, state], [msg, image, video, chatbot, state])

demo.launch()