openclaw - 💡(How to fix) Fix [Feature Request] openclaw-weixin: 支持发送原生微信语音消息（语音气泡） [2 comments, 2 participants]

TL;DR

To fix the issue, modify the send-media.ts file to add support for audio/voice messages by uploading the audio to the WeChat server and obtaining a media_id before sending the voice message.

Guidance

• Identify the audio format generated by the TTS system and ensure it matches one of the supported formats by WeChat (AMR or MP3 with 16KHz sampling rate).
• Update the send-media.ts file to include a new route for handling audio/voice messages, using the WeChat API to upload the audio and obtain a media_id.
• Modify the sendFileMessageWeixin function to handle audio files differently, using the new audio/voice message route instead of treating them as file attachments.
• Test the updated code with different audio formats and scenarios to ensure compatibility and correct behavior.

Example

// Example of how to upload audio to WeChat server and obtain media_id
const audioBuffer = // audio data generated by TTS
const audioType = 'audio/mp3'; // or 'audio/amr'
const mediaId = await uploadMediaToWeChat(audioBuffer, audioType);
const voiceMessage = {
  msgtype: 'voice',
  voice: {
    media_id: mediaId,
  },
};
// Send the voice message using the WeChat API

Notes

The solution assumes that the TTS system generates audio in a format compatible with WeChat's requirements. Additional error handling and logging may be necessary to ensure robustness.

Recommendation

Apply workaround: Modify the send-media.ts file to add support for audio/voice messages, as this will allow the system to send voice messages correctly without requiring a full upgrade.