import os
import gradio as gr
from groq import Groq
import tempfile
import requests
import json
import base64
from pathlib import Path
import time
import sys
import platform

# Available English voices for different providers
groq_voices = [
    "Arista-PlayAI", "Atlas-PlayAI", "Basil-PlayAI", "Briggs-PlayAI",
    "Calum-PlayAI", "Celeste-PlayAI", "Cheyenne-PlayAI", "Chip-PlayAI",
    "Cillian-PlayAI", "Deedee-PlayAI", "Fritz-PlayAI", "Gail-PlayAI",
    "Indigo-PlayAI", "Mamaw-PlayAI", "Mason-PlayAI", "Mikail-PlayAI",
    "Mitch-PlayAI", "Quinn-PlayAI", "Thunder-PlayAI"
]

# OpenAI voices (if user has OpenAI API key)
openai_voices = ["alloy", "ash", "coral", "sage", "echo", "fable", "onyx", "nova", "shimmer"]

# Edge TTS voices (free, no API key required)
edge_voices = [
    "en-US-AriaNeural",
    "en-US-AnaNeural",
    "en-US-ChristopherNeural",
    "en-US-EricNeural",
    "en-US-GuyNeural",
    "en-US-JennyNeural",
    "en-US-MichelleNeural",
    "en-US-RogerNeural",
    "en-US-SteffanNeural"
]

def check_internet_connection():
    """Check if internet connection is available"""
    try:
        import socket
        socket.create_connection(("8.8.8.8", 53), timeout=3)
        return True
    except OSError:
        return False

def is_huggingface_space():
    """Check if running on Hugging Face Spaces"""
    return os.environ.get("SPACE_ID") is not None

def chunk_text(text, max_length=4000):
    """Split text into chunks to avoid rate limits"""
    if len(text) <= max_length:
        return [text]
    
    chunks = []
    words = text.split()
    current_chunk = []
    current_length = 0
    
    for word in words:
        if current_length + len(word) + 1 <= max_length:
            current_chunk.append(word)
            current_length += len(word) + 1
        else:
            if current_chunk:
                chunks.append(" ".join(current_chunk))
                current_chunk = [word]
                current_length = len(word)
    
    if current_chunk:
        chunks.append(" ".join(current_chunk))
    
    return chunks

def try_groq_tts(api_key, text, voice):
    """Try Groq TTS with chunking and retry logic"""
    try:
        if not check_internet_connection():
            return None, "❌ No internet connection available for Groq TTS"
            
        client = Groq(api_key=api_key)
        
        # Check if text needs chunking
        chunks = chunk_text(text, 3500)  # Leave some buffer
        
        if len(chunks) == 1:
            # Single chunk - direct call
            response = client.audio.speech.create(
                model="playai-tts",
                voice=voice,
                input=text,
                response_format="wav"
            )
            
            with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_file:
                response.write_to_file(temp_file.name)
                return temp_file.name, "✅ Speech generated successfully with Groq PlayAI!"
        else:
            # Multiple chunks - process separately and inform user
            return None, f"⚠️ Text too long for single request ({len(text)} chars). Try shorter text or use Edge TTS for longer content."
            
    except Exception as e:
        error_msg = str(e)
        if "rate_limit_exceeded" in error_msg or "429" in error_msg:
            return None, "🔄 Groq rate limit reached. Try again in a moment or use Edge TTS..."
        else:
            return None, f"❌ Groq error: {error_msg}"

def try_openai_tts(api_key, text, voice):
    """Try OpenAI TTS as fallback"""
    try:
        if not check_internet_connection():
            return None, "❌ No internet connection available for OpenAI TTS"
            
        import openai
        client = openai.OpenAI(api_key=api_key)
        
        response = client.audio.speech.create(
            model="tts-1",
            voice=voice,
            input=text[:4000]  # OpenAI limit
        )
        
        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_file:
            response.stream_to_file(temp_file.name)
            return temp_file.name, "✅ Speech generated successfully with OpenAI TTS!"
            
    except ImportError:
        return None, "❌ OpenAI library not installed. Install with: pip install openai"
    except Exception as e:
        return None, f"❌ OpenAI error: {str(e)}"

def try_edge_tts(text, voice="en-US-JennyNeural"):
    """Try Microsoft Edge TTS as free fallback"""
    try:
        if not check_internet_connection():
            return None, "❌ No internet connection available for Edge TTS"
            
        import edge_tts
        import asyncio
        
        async def generate_edge_speech():
            communicate = edge_tts.Communicate(text[:10000], voice)  # Edge TTS can handle longer text
            with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as temp_file:
                async for chunk in communicate.stream():
                    if chunk["type"] == "audio":
                        temp_file.write(chunk["data"])
                return temp_file.name
        
        # Run async function
        try:
            loop = asyncio.get_event_loop()
        except RuntimeError:
            loop = asyncio.new_event_loop()
            asyncio.set_event_loop(loop)
            
        audio_file = loop.run_until_complete(generate_edge_speech())
        
        return audio_file, "✅ Speech generated successfully with Microsoft Edge TTS (Free & High Quality)!"
        
    except ImportError:
        return None, "❌ Edge TTS not installed. Install with: pip install edge-tts"
    except Exception as e:
        return None, f"❌ Edge TTS error: {str(e)}"

def try_pyttsx3_fallback(text):
    """Ultimate fallback using pyttsx3 (offline) - Only works in local environments"""
    if is_huggingface_space():
        return None, "❌ Offline TTS not available in Hugging Face Spaces. Use Edge TTS (Free) instead!"
    
    try:
        import pyttsx3
        
        # Initialize the engine
        engine = pyttsx3.init()
        
        # Get available voices and set a good one
        voices = engine.getProperty('voices')
        if voices:
            # Try to find a female English voice first, then any English voice
            english_voice = None
            for voice in voices:
                voice_id = voice.id.lower()
                voice_name = voice.name.lower() if hasattr(voice, 'name') else ""
                
                # Look for English voices
                if any(keyword in voice_id or keyword in voice_name 
                       for keyword in ['english', 'en-us', 'en_us', 'zira', 'hazel', 'eva']):
                    english_voice = voice.id
                    if any(female_keyword in voice_name 
                           for female_keyword in ['zira', 'hazel', 'eva', 'female']):
                        break  # Prefer female voices
            
            if english_voice:
                engine.setProperty('voice', english_voice)
        
        # Set properties for better quality
        engine.setProperty('rate', 180)    # Speed of speech (words per minute)
        engine.setProperty('volume', 0.9)  # Volume level (0.0 to 1.0)
        
        # Create temporary file
        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_file:
            temp_path = temp_file.name
        
        # Limit text length for stability
        limited_text = text[:2000]  # Limit for stability
        
        # Use the save_to_file method correctly
        engine.save_to_file(limited_text, temp_path)
        engine.runAndWait()
        
        # Check if file was created and has content
        if os.path.exists(temp_path) and os.path.getsize(temp_path) > 0:
            return temp_path, "✅ Speech generated with offline TTS! (No internet required)"
        else:
            # If save_to_file didn't work, try alternative method
            return try_alternative_offline_tts(limited_text)
            
    except ImportError:
        return None, "❌ pyttsx3 not installed. Use Edge TTS (Free) instead!"
    except Exception as e:
        # Try alternative offline method if pyttsx3 fails
        return try_alternative_offline_tts(text[:2000])

def try_alternative_offline_tts(text):
    """Alternative offline TTS using system commands - Only for local environments"""
    if is_huggingface_space():
        return None, "❌ System TTS not available in Hugging Face Spaces. Use Edge TTS (Free) instead!"
    
    try:
        system = platform.system().lower()
        
        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_file:
            temp_path = temp_file.name
        
        if system == "windows":
            # Windows SAPI TTS
            try:
                import win32com.client
                speaker = win32com.client.Dispatch("SAPI.SpVoice")
                
                # Save to file
                file_stream = win32com.client.Dispatch("SAPI.SpFileStream")
                file_stream.Open(temp_path, 3)
                speaker.AudioOutputStream = file_stream
                speaker.Speak(text)
                file_stream.Close()
                
                if os.path.exists(temp_path) and os.path.getsize(temp_path) > 0:
                    return temp_path, "✅ Speech generated with Windows SAPI TTS (Offline)!"
            except ImportError:
                pass
                
        elif system == "darwin":  # macOS
            # Use macOS 'say' command
            import subprocess
            try:
                subprocess.run(['say', '-o', temp_path, '--data-format=LEF32@22050', text], 
                             check=True, timeout=30)
                if os.path.exists(temp_path) and os.path.getsize(temp_path) > 0:
                    return temp_path, "✅ Speech generated with macOS TTS (Offline)!"
            except (subprocess.CalledProcessError, subprocess.TimeoutExpired, FileNotFoundError):
                pass
                
        elif system == "linux":
            # Try espeak or festival on Linux
            import subprocess
            try:
                # Try espeak first
                subprocess.run(['espeak', '-w', temp_path, text], 
                             check=True, timeout=30)
                if os.path.exists(temp_path) and os.path.getsize(temp_path) > 0:
                    return temp_path, "✅ Speech generated with espeak TTS (Offline)!"
            except (subprocess.CalledProcessError, subprocess.TimeoutExpired, FileNotFoundError):
                try:
                    # Try festival as backup
                    subprocess.run(['text2wave', '-o', temp_path], 
                                 input=text, text=True, check=True, timeout=30)
                    if os.path.exists(temp_path) and os.path.getsize(temp_path) > 0:
                        return temp_path, "✅ Speech generated with Festival TTS (Offline)!"
                except (subprocess.CalledProcessError, subprocess.TimeoutExpired, FileNotFoundError):
                    pass
        
        return None, f"❌ No offline TTS available for {system}. Use Edge TTS (Free) instead!"
        
    except Exception as e:
        return None, f"❌ Alternative offline TTS error: {str(e)}"

# Main function with multiple provider fallback
def generate_speech(groq_api_key, openai_api_key, text, voice_provider, voice):
    if not text:
        return None, "⚠️ Please enter some text to generate speech."

    if len(text) > 10000:
        return None, "🚫 Text input exceeds 10,000 character limit."

    # Check internet connection and environment
    has_internet = check_internet_connection()
    is_hf_space = is_huggingface_space()
    
    internet_status = "🌐 Internet: Connected" if has_internet else "📴 Internet: Offline"
    env_status = " (Hugging Face Spaces)" if is_hf_space else " (Local Environment)"
    
    # Status message
    status_msg = f"🔄 Generating speech...\n{internet_status}{env_status}\n"
    
    # If no internet, inform user about limitations
    if not has_internet:
        return None, status_msg + "❌ No internet connection available. All TTS providers require internet connection in this environment."
    
    # Try providers in order based on selection
    if voice_provider == "Groq (PlayAI)" and groq_api_key:
        status_msg += "🎯 Using Groq PlayAI...\n"
        audio_file, message = try_groq_tts(groq_api_key, text, voice)
        if audio_file:
            return audio_file, status_msg + message
        else:
            status_msg += f"❌ Groq failed: {message}\n"
    
    # Try OpenAI if available
    if openai_api_key and voice_provider in ["OpenAI TTS", "Auto (Try All)"]:
        status_msg += "🔄 Trying OpenAI TTS...\n"
        openai_voice = voice if voice in openai_voices else "alloy"
        audio_file, message = try_openai_tts(openai_api_key, text, openai_voice)
        if audio_file:
            return audio_file, status_msg + message
        else:
            status_msg += f"❌ OpenAI failed: {message}\n"
    
    # Try Edge TTS (Free) - This should be the main fallback for HF Spaces
    if voice_provider in ["Edge TTS (Free)", "Auto (Try All)", "Groq (PlayAI)", "OpenAI TTS"]:
        status_msg += "🔄 Using Edge TTS (Free & High Quality)...\n"
        edge_voice = voice if voice in edge_voices else "en-US-JennyNeural"
        audio_file, message = try_edge_tts(text, edge_voice)
        if audio_file:
            return audio_file, status_msg + message
        else:
            status_msg += f"❌ Edge TTS failed: {message}\n"
    
    # Only try offline TTS if specifically requested and not in HF Spaces
    if voice_provider == "Offline TTS":
        if is_hf_space:
            return None, status_msg + "❌ Offline TTS is not available in Hugging Face Spaces. Please use 'Edge TTS (Free)' instead!"
        else:
            status_msg += "🔄 Using offline TTS (works without internet)...\n"
            audio_file, message = try_pyttsx3_fallback(text)
            if audio_file:
                return audio_file, status_msg + message
            else:
                status_msg += f"❌ Offline TTS failed: {message}\n"
    
    # Final fallback message
    if is_hf_space:
        return None, status_msg + "❌ All online TTS providers failed. Please check your API keys or try again later. Edge TTS (Free) is recommended for HF Spaces!"
    else:
        return None, status_msg + "❌ All TTS providers failed. Please check your setup or try shorter text."

def update_voice_options(provider):
    """Update voice dropdown based on selected provider"""
    if provider == "Groq (PlayAI)":
        return gr.Dropdown(choices=groq_voices, value="Fritz-PlayAI", visible=True)
    elif provider == "OpenAI TTS":
        return gr.Dropdown(choices=openai_voices, value="alloy", visible=True)
    elif provider == "Edge TTS (Free)":
        return gr.Dropdown(choices=edge_voices, value="en-US-JennyNeural", visible=True)
    elif provider == "Offline TTS":
        return gr.Dropdown(choices=["Default System Voice"], value="Default System Voice", visible=True)
    else:  # Auto
        return gr.Dropdown(choices=groq_voices, value="Fritz-PlayAI", visible=True, label="🎭 Voice (Auto mode will try best match)")

# Custom CSS (keeping your original beautiful design)
custom_css = """
@import url('https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600;700&display=swap');
* {
    box-sizing: border-box;
}
:root {
    --primary-gradient: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
    --secondary-gradient: linear-gradient(135deg, #f093fb 0%, #f5576c 100%);
    --accent-gradient: linear-gradient(135deg, #4facfe 0%, #00f2fe 100%);
    --dark-bg: linear-gradient(135deg, #0c0c0c 0%, #1a1a2e 50%, #16213e 100%);
    --glass-bg: rgba(255, 255, 255, 0.08);
    --glass-border: rgba(255, 255, 255, 0.15);
    --text-primary: #ffffff;
    --text-secondary: rgba(255, 255, 255, 0.7);
    --shadow-primary: 0 8px 32px rgba(0, 0, 0, 0.4);
    --shadow-hover: 0 12px 48px rgba(0, 0, 0, 0.6);
    --border-radius: 16px;
}

body, .gradio-container {
    background: var(--dark-bg) !important;
    font-family: 'Inter', -apple-system, BlinkMacSystemFont, sans-serif !important;
    color: var(--text-primary) !important;
    overflow-x: hidden;
}

.gradio-container {
    max-width: 1400px !important;
    margin: 40px auto !important;
    padding: 0 20px !important;
    min-height: 100vh;
}

.gradio-container > div {
    background: var(--glass-bg) !important;
    backdrop-filter: blur(20px) !important;
    border: 1px solid var(--glass-border) !important;
    border-radius: var(--border-radius) !important;
    padding: 40px !important;
    box-shadow: var(--shadow-primary) !important;
    position: relative;
    overflow: hidden;
}

h1 {
    text-align: center !important;
    font-size: clamp(2.5rem, 5vw, 4rem) !important;
    font-weight: 700 !important;
    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%) !important;
    -webkit-background-clip: text !important;
    -webkit-text-fill-color: transparent !important;
    background-clip: text !important;
    margin: 0 0 50px 0 !important;
    animation: glow 3s ease-in-out infinite alternate;
}

@keyframes glow {
    0% { filter: drop-shadow(0 0 20px rgba(102, 126, 234, 0.3)); }
    100% { filter: drop-shadow(0 0 30px rgba(118, 75, 162, 0.5)); }
}

#generate-btn {
    background: linear-gradient(135deg, #00f2fe 0%, #4facfe 100%) !important;
    border: 2px solid #4facfe !important;
    color: #ffffff !important;
    font-weight: bold !important;
    box-shadow: 0 0 12px rgba(0, 242, 254, 0.5) !important;
    transition: all 0.3s ease !important;
    margin-top: 26px !important;
}

#generate-btn:hover {
    background: linear-gradient(135deg, #4facfe 0%, #00f2fe 100%) !important;
    transform: scale(1.03) !important;
    box-shadow: 0 0 18px rgba(0, 242, 254, 0.7) !important;
}

.gr-textbox, .gr-dropdown {
    background: var(--glass-bg) !important;
    border: 2px solid transparent !important;
    border-radius: 12px !important;
    position: relative;
    transition: all 0.3s cubic-bezier(0.4, 0, 0.2, 1) !important;
    backdrop-filter: blur(10px) !important;
}

.gr-textbox textarea, .gr-textbox input, .gr-dropdown select {
    background: transparent !important;
    color: var(--text-primary) !important;
    border: none !important;
    font-size: 16px !important;
    font-weight: 400 !important;
    padding: 16px !important;
}

.gr-input-label, .gr-output-label {
    color: var(--text-primary) !important;
    font-weight: 600 !important;
    font-size: 14px !important;
    text-transform: uppercase;
    letter-spacing: 0.5px;
    margin-bottom: 12px !important;
    background: var(--accent-gradient);
    -webkit-background-clip: text;
    -webkit-text-fill-color: transparent;
    background-clip: text;
}

.gr-button {
    background: var(--primary-gradient) !important;
    border: none !important;
    color: white !important;
    font-weight: 600 !important;
    font-size: 16px !important;
    padding: 16px 32px !important;
    border-radius: 12px !important;
    transition: all 0.3s cubic-bezier(0.4, 0, 0.2, 1) !important;
    text-transform: uppercase;
    letter-spacing: 0.5px;
    box-shadow: 0 4px 20px rgba(102, 126, 234, 0.3) !important;
    margin-top: 20px !important;
}

.gr-button:hover {
    transform: translateY(-2px) !important;
    box-shadow: var(--shadow-hover) !important;
    filter: brightness(1.1);
}

.gr-audio {
    background: var(--glass-bg) !important;
    border: 1px solid var(--glass-border) !important;
    border-radius: 12px !important;
    padding: 20px !important;
    backdrop-filter: blur(10px) !important;
    transition: all 0.3s ease !important;
}

.info-box {
    background: rgba(0, 242, 254, 0.1) !important;
    border: 1px solid rgba(0, 242, 254, 0.3) !important;
    border-radius: 12px !important;
    padding: 20px !important;
    margin: 20px 0 !important;
    backdrop-filter: blur(10px) !important;
}

.info-box,
.info-box * {
    color: white !important;
}

.warning-box,
.warning-box * {
    color: white !important;
}

.no-api-box {
    background: rgba(46, 204, 113, 0.1) !important;
    border: 1px solid rgba(46, 204, 113, 0.3) !important;
    border-radius: 12px !important;
    padding: 20px !important;
    margin: 20px 0 !important;
    backdrop-filter: blur(10px) !important;
}

.no-api-box,
.no-api-box * {
    color: white !important;
}

.warning-box {
    background: rgba(255, 193, 7, 0.1) !important;
    border: 1px solid rgba(255, 193, 7, 0.3) !important;
    border-radius: 12px !important;
    padding: 15px !important;
    margin: 15px 0 !important;
    backdrop-filter: blur(10px) !important;
}

.hf-space-box {
    background: rgba(255, 87, 34, 0.1) !important;
    border: 1px solid rgba(255, 87, 34, 0.3) !important;
    border-radius: 12px !important;
    padding: 20px !important;
    margin: 20px 0 !important;
    backdrop-filter: blur(10px) !important;
}

.hf-space-box,
.hf-space-box * {
    color: white !important;
}
"""

# Updated HTML for Hugging Face Spaces
def get_custom_html():
    is_hf = is_huggingface_space()
    
    if is_hf:
        return """
        <div style="text-align: center; margin-bottom: 20px;">
            <div style="font-size: 3rem;">🎙️</div>
            <h1>Advanced Multi-Provider TTS Studio</h1>
            <p style="color: rgba(255, 255, 255, 0.7); font-size: 18px; margin: 0; font-weight: 300;">
                Transform your text into natural-sounding speech with multiple AI-powered voice providers
            </p>
            
            <div class="hf-space-box">
                <h3>🚀 Running on Hugging Face Spaces</h3>
                <p><strong>Recommended: Use "Edge TTS (Free)" for best results!</strong></p>
                <p>✅ High-quality voices • ✅ No API key needed • ✅ Up to 10,000 characters • ✅ Multiple voice options</p>
                <p>⚠️ Offline TTS is not available in cloud environments</p>
            </div>
            
            <div class="no-api-box">
                <h3>🆓 No API Key? No Problem!</h3>
                <p><strong>Select "Edge TTS (Free)" for high-quality TTS without any API keys!</strong></p>
                <p>✅ Microsoft's premium voices • ✅ Completely free • ✅ Works great in HF Spaces</p>
            </div>
            
            <div class="info-box">
                <p><strong>🔄 Multiple Providers Available:</strong> Groq PlayAI, OpenAI TTS, Microsoft Edge TTS (Free)</p>
                <p><strong>🛡️ Smart Fallback System:</strong> If one provider fails, automatically tries the next available</p>
                <p><strong>💡 Recommended:</strong> Use "Auto (Try All)" for maximum reliability or "Edge TTS (Free)" for consistent quality</p>
            </div>
        </div>
        """
    else:
        return """
        <div style="text-align: center; margin-bottom: 20px;">
            <div style="font-size: 3rem;">🎙️</div>
            <h1>Advanced Multi-Provider TTS Studio</h1>
            <p style="color: rgba(255, 255, 255, 0.7); font-size: 18px; margin: 0; font-weight: 300;">
                Transform your text into natural-sounding speech with multiple AI-powered voice providers
            </p>
            
            <div class="hf-space-box">
                <h3>💻 Running Locally</h3>
                <p><strong>All TTS options available including offline TTS!</strong></p>
                <p>✅ Edge TTS (Free) • ✅ Offline TTS • ✅ Premium APIs • ✅ Full feature set</p>
            </div>
            
            <div class="no-api-box">
                <h3>🆓 No API Key? Multiple Free Options!</h3>
                <p><strong>Select "Edge TTS (Free)" for online high-quality TTS or "Offline TTS" for no internet!</strong></p>
                <p>✅ Edge TTS: High quality voices with internet • ✅ Offline TTS: Basic quality, works anywhere</p>
            </div>
            
            <div class="info-box">
                <p><strong>🔄 Multiple Providers Available:</strong> Groq PlayAI, OpenAI TTS, Microsoft Edge TTS (Free), Offline TTS</p>
                <p><strong>🛡️ Smart Fallback System:</strong> If one provider fails, automatically tries the next available</p>
                <p><strong>💡 Recommended:</strong> Use "Auto (Try All)" for maximum reliability or specific providers based on your needs</p>
            </div>
        </div>
        """

# Gradio interface
with gr.Blocks(css=custom_css, theme="ocean") as demo:
    gr.HTML(get_custom_html())

    # API Keys section
    with gr.Row():
        with gr.Column():
            groq_api_key = gr.Textbox(
                label="🔐 Groq API Key (Optional)",
                placeholder="Optional: Paste your Groq API key here for PlayAI voices...",
                type="password",
                lines=1
            )
        with gr.Column():
            openai_api_key = gr.Textbox(
                label="🔐 OpenAI API Key (Optional)",
                placeholder="Optional: OpenAI API key for premium TTS voices...",
                type="password",
                lines=1
            )
    
    # Warning box for users without API keys - Updated for HF Spaces
    if is_huggingface_space():
        gr.HTML("""
        <div class="warning-box">
            <h3>⚡ Quick Start for HF Spaces Users</h3>
            <p><strong>Best Option:</strong> Select "Edge TTS (Free)" below - no API key needed!</p>
            <p>💡 <strong>Pro tip:</strong> Edge TTS provides Microsoft's premium voice quality completely free</p>
        </div>
        """)
    else:
        gr.HTML("""
        <div class="warning-box">
            <h3>⚡ Quick Start Options</h3>
            <p><strong>Free Options:</strong> "Edge TTS (Free)" (internet required) or "Offline TTS" (no internet needed)</p>
            <p><strong>Premium Options:</strong> Add your Groq or OpenAI API keys above for additional voice options</p>
        </div>
        """)

    # Main interface
    with gr.Row():
        with gr.Column(scale=2):
            # Provider selection
            provider_choices = ["Auto (Try All)", "Edge TTS (Free)", "Groq (PlayAI)", "OpenAI TTS"]
            if not is_huggingface_space():
                provider_choices.append("Offline TTS")
            
            voice_provider = gr.Dropdown(
                choices=provider_choices,
                value="Edge TTS (Free)" if is_huggingface_space() else "Auto (Try All)",
                label="🎯 TTS Provider",
                info="Edge TTS (Free) is recommended for HF Spaces!" if is_huggingface_space() else "Auto tries all providers in order"
            )
            
            # Voice selection
            voice = gr.Dropdown(
                choices=edge_voices,
                value="en-US-JennyNeural",
                label="🎭 Voice Selection",
                info="Available voices update based on selected provider",
                visible=True
            )
            
            # Text input
            text_input = gr.Textbox(
                label="📝 Text to Convert to Speech",
                placeholder="Enter your text here... (up to 10,000 characters)",
                lines=8,
                max_lines=15,
                info="💡 Longer texts work better with Edge TTS (Free)"
            )
            
            # Generate button
            generate_btn = gr.Button(
                "🎵 Generate Speech",
                variant="primary",
                elem_id="generate-btn",
                size="lg"
            )

        with gr.Column(scale=1):
            # Audio output
            audio_output = gr.Audio(
                label="🔊 Generated Speech",
                type="filepath",
                interactive=False
            )
            
            # Status/Info output
            status_output = gr.Textbox(
                label="📊 Status & Information",
                lines=8,
                interactive=False,
                info="Real-time status updates and results"
            )

    # Update voice options when provider changes
    voice_provider.change(
        fn=update_voice_options,
        inputs=[voice_provider],
        outputs=[voice]
    )

    # Generate speech button click
    generate_btn.click(
        fn=generate_speech,
        inputs=[groq_api_key, openai_api_key, text_input, voice_provider, voice],
        outputs=[audio_output, status_output]
    )

    # Footer with additional information
    gr.HTML("""
    <div style="text-align: center; margin-top: 40px; padding: 20px; background: rgba(255, 255, 255, 0.05); border-radius: 12px; backdrop-filter: blur(10px);">
        <h3 style="color: rgba(255, 255, 255, 0.9); margin-bottom: 15px;">🎯 Provider Comparison</h3>
        <div style="display: grid; grid-template-columns: repeat(auto-fit, minmax(300px, 1fr)); gap: 15px; text-align: left;">
            <div style="background: rgba(0, 242, 254, 0.1); padding: 15px; border-radius: 8px; border: 1px solid rgba(0, 242, 254, 0.3);">
                <h4 style="color: #00f2fe; margin: 0 0 10px 0;">🆓 Edge TTS (Free)</h4>
                <p style="color: rgba(255, 255, 255, 0.8); margin: 0; font-size: 14px;">✅ No API key required<br>✅ High quality Microsoft voices<br>✅ Up to 10,000 characters<br>✅ Perfect for HF Spaces</p>
            </div>
            <div style="background: rgba(102, 126, 234, 0.1); padding: 15px; border-radius: 8px; border: 1px solid rgba(102, 126, 234, 0.3);">
                <h4 style="color: #667eea; margin: 0 0 10px 0;">🎭 Groq PlayAI</h4>
                <p style="color: rgba(255, 255, 255, 0.8); margin: 0; font-size: 14px;">🔑 Requires API key<br>🎪 Premium entertainment voices<br>⚡ Fast generation<br>📝 ~3,500 char limit</p>
            </div>
            <div style="background: rgba(118, 75, 162, 0.1); padding: 15px; border-radius: 8px; border: 1px solid rgba(118, 75, 162, 0.3);">
                <h4 style="color: #764ba2; margin: 0 0 10px 0;">🤖 OpenAI TTS</h4>
                <p style="color: rgba(255, 255, 255, 0.8); margin: 0; font-size: 14px;">🔑 Requires API key<br>🎯 Professional quality<br>💰 Pay per use<br>📝 4,000 char limit</p>
            </div>
        </div>
        <div style="margin-top: 20px; padding: 15px; background: rgba(46, 204, 113, 0.1); border-radius: 8px; border: 1px solid rgba(46, 204, 113, 0.3);">
            <p style="color: rgba(255, 255, 255, 0.9); margin: 0; font-size: 16px;"><strong>💡 Pro Tip:</strong> Use "Auto (Try All)" to automatically find the best available provider, or stick with "Edge TTS (Free)" for consistent, high-quality results without any setup!</p>
        </div>
    </div>
    """)

# Launch configuration
if __name__ == "__main__":
    # Check if we're in HF Spaces and adjust accordingly
    if is_huggingface_space():
        # HF Spaces configuration
        demo.launch(
            server_name="0.0.0.0",
            server_port=7860,
            share=False,  # HF Spaces handles sharing
            show_error=True,
            # show_tips=True,
            enable_monitoring=False,  # Disable monitoring in HF Spaces
            favicon_path=None,
            app_kwargs={
                "docs_url": None,  # Disable docs endpoint
                "redoc_url": None,  # Disable redoc endpoint
            }
        )
    else:
        # Local development configuration
        demo.launch(
            share=True
        )