gradio_app.py

import gradio as gr
import edge_tts
import asyncio
import os

# 定义生成音频的异步函数
async def generate_audio(text, voice, rate, volume, pitch):
    try:
        # 创建输出文件路径
        output_path = "output.mp3"
        
        # 使用edge-tts生成音频
        communicate = edge_tts.Communicate(text, voice, rate=rate, volume=volume, pitch=pitch)
        await communicate.save(output_path)
        
        return output_path
    except Exception as e:
        print(f"生成音频失败: {e}")
        raise gr.Error(f"生成音频失败: {str(e)}")

# 定义Gradio界面的主函数
def text_to_speech(text, voice, rate, volume, pitch):
    try:
        # 将rate和volume参数转换为edge-tts需要的格式
        rate_str = f"+{int(rate)}%" if rate >= 0 else f"{int(rate)}%"
        volume_str = f"+{int(volume)}%" if volume >= 0 else f"{int(volume)}%"
        pitch_str = f"+{int(pitch)}Hz" if pitch >= 0 else f"{int(pitch)}Hz"
        
        # 调用异步函数生成音频
        loop = asyncio.new_event_loop()
        asyncio.set_event_loop(loop)
        output_path = loop.run_until_complete(generate_audio(text, voice, rate_str, volume_str, pitch_str))
        loop.close()
        
        return output_path
    except Exception as e:
        print(f"文本转语音失败: {e}")
        raise gr.Error(f"文本转语音失败: {str(e)}")

# 异步获取所有可用语音
def get_voices():
    try:
        loop = asyncio.new_event_loop()
        asyncio.set_event_loop(loop)
        voices = loop.run_until_complete(edge_tts.list_voices())
        loop.close()
        
        # 过滤出支持的语音，并按语言分组
        voice_options = []
        for voice in voices:
            short_name = voice.get("ShortName", "")
            friendly_name = voice.get("FriendlyName", short_name)
            locale = voice.get("Locale", "")
            voice_options.append((f"{friendly_name} ({locale})", short_name))
        
        # 按友好名称排序
        voice_options.sort(key=lambda x: x[0])
        
        return voice_options
    except Exception as e:
        # 如果获取语音列表失败，返回默认的语音列表
        print(f"获取语音列表失败: {e}")
        return [
            ("中文女声 (zh-CN)", "zh-CN-YunxiNeural"),
            ("中文男声 (zh-CN)", "zh-CN-YunxiNeural"),
            ("英文女声 (en-US)", "en-US-JennyNeural"),
            ("英文男声 (en-US)", "en-US-GuyNeural")
        ]

# 获取所有可用语音
voice_options = get_voices()

# 创建Gradio界面
with gr.Blocks(title="Edge TTS 配音工具") as demo:
    gr.Markdown("# Edge TTS 配音工具")
    gr.Markdown("输入文本，选择语音，调节语速、音量和语调，然后点击合成按钮生成音频。")
    
    with gr.Row():
        with gr.Column(scale=2):
            # 文本输入框
            text_input = gr.Textbox(
                label="输入文本",
                placeholder="请输入要转换为语音的文本...",
                lines=5,
                max_lines=10
            )
            
            # 语音选择下拉菜单
            voice_dropdown = gr.Dropdown(
                label="选择语音",
                choices=voice_options,
                value="zh-CN-YunxiNeural"
            )
            
            # 语速滑块
            rate_slider = gr.Slider(
                label="语速",
                minimum=-50,
                maximum=50,
                step=5,
                value=0,
                info="-50% 到 +50%"
            )
            
            # 音量滑块
            volume_slider = gr.Slider(
                label="音量",
                minimum=-50,
                maximum=50,
                step=5,
                value=0,
                info="-50% 到 +50%"
            )
            
            # 语调滑块
            pitch_slider = gr.Slider(
                label="语调",
                minimum=-50,
                maximum=50,
                step=5,
                value=0,
                info="-50Hz 到 +50Hz"
            )
            
            # 合成按钮
            generate_button = gr.Button("合成语音", variant="primary")
        
        with gr.Column(scale=1):
            # 音频输出
            audio_output = gr.Audio(
                label="合成音频",
                type="filepath"
            )
    
    # 设置按钮点击事件
    generate_button.click(
        fn=text_to_speech,
        inputs=[text_input, voice_dropdown, rate_slider, volume_slider, pitch_slider],
        outputs=audio_output
    )

# 启动Gradio应用
if __name__ == "__main__":
    demo.launch(server_name="0.0.0.0", server_port=7860)