Nanobot needs a voice!!!

[mxnbf]

we can talk to it, and receive its text responses .. it would be great to hear it reply ..

This is for nanobot v0.2.2. Be aware that tts will add a delay while the text is synthesized and that this, its bad code (not properly integrated etc) because im hoping to give a push towards the functionality .. hopefully im not alone here?

please use this only as guide, it will not add audio responses everywhere, instead it will add audio responses too whatsapp, in place of text responses.

config.json:

"whatsapp": {
  "tts": true,
  "tts_url": "http://10.0.0.11:8001/api/tts",
  "tts_speaker": "p364"
}

nanobot/channels/whatsapp.py
find this line:
lid_mappings: dict[str, str] = Field(default_factory=dict)
on the next line, add:
tts: bool = True
tts_url: str = "http://10.0.0.11:8001/api/tts"
tts_speaker: str = "p364"

find this line:
chat_id = msg.chat_id
add this section immediately after:

   tts = self.config.tts == True
    tts_audio_path = "/tmp/tmp.mp3"


    # Download TTS audio before sending if requested
    if tts and msg.content:
        try:
            from urllib.parse import quote_plus

            # For your curl command
            encoded_text123 = quote_plus(msg.content, safe='=')

            subprocess.run(
                ["curl", "-s", f"{self.config.tts_url}?text={encoded_text123}&speaker_id={self.config.tts_speaker}&language_id=en",  "--output", tts_audio_path], check=True, timeout=240
            )
            self.logger.info("Downloaded TTS audio for message...")

            mime, _ = mimetypes.guess_type(tts_audio_path)
            payload = {
                "type": "send_media",
                "to": chat_id,
                "filePath": tts_audio_path,
                "mimetype": mime or "application/octet-stream",
                "fileName": tts_audio_path.rsplit("/", 1)[-1]
            }
            await self._ws.send(json.dumps(payload, ensure_ascii=False))

            if os.path.exists(tts_audio_path):
                self.logger.info(f"TTS audio created: {tts_audio_path.rsplit("/", 1)[-1]} {mime}")
            else:
                self.logger.error("TTS audio file not created")
        except Exception as e:
            self.logger.error(f"TTS download failed: {tts_audio_path}", e)
            raise

find this line (its immediately after ^):
if msg.content:
replace with:
if msg.content and not tts:

or not if you want both text and audio..