google和百度的TTS我都不太好用，因为我的homeassistant是docker版的，无法调用宿主声卡，所以我曲线救国

一、设置一个swith command_line来修改一个文本标志

• 修改configuration.yaml
• /config 是docker的加载目录

switch:
  - platform: command_line
    switches:
      egg:
        command_on: sed -i 's/stop/on/' /config/egg.txt
        command_off: sed -i 's/stop/off/' /config/egg.txt

二、一个循环脚本

while [ i==0 ]
do
msg=$(sed -n '1,1p' /data/docker/hass_config/egg.txt)
echo "查看egg.txt: ... ... ... " $msg
if [ $msg = "on" ];then
	        echo "on"
		cvlc --no-interact --play-and-exit file:///opt/say/egg_on.mp3
		sleep 7;
		sudo sed -i 's/on/stop/' /data/docker/hass_config/egg.txt
	elif [ $msg = "off" ];then
                echo "off"
            cvlc --no-interact --play-and-exit file:///opt/say/egg_off.mp3
            sleep 7;
            sudo sed -i 's/off/stop/' /data/docker/hass_config/egg.txt
        else
                    echo "stop"
    fi
    sleep 2;
done

三、利用文本转语音工具制做.mp3

四、在启动raspberrypi时