43 lines
1.2 KiB
Python
43 lines
1.2 KiB
Python
"""Speech-to-text for the actual query, after the wake word fires.
|
|
|
|
Deepgram, same as desk_client/bolt_desk.py.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import requests
|
|
|
|
from .. import config
|
|
from .mic import pcm_to_wav_bytes
|
|
|
|
|
|
class SttError(Exception):
|
|
pass
|
|
|
|
|
|
def transcribe(pcm) -> str:
|
|
if not config.DEEPGRAM_API_KEY:
|
|
raise SttError("DEEPGRAM_API_KEY is not set")
|
|
try:
|
|
response = requests.post(
|
|
"https://api.deepgram.com/v1/listen",
|
|
params={"model": config.DEEPGRAM_MODEL, "language": "en", "smart_format": "true"},
|
|
headers={
|
|
"Authorization": f"Token {config.DEEPGRAM_API_KEY}",
|
|
"Content-Type": "audio/wav",
|
|
},
|
|
data=pcm_to_wav_bytes(pcm),
|
|
timeout=30,
|
|
)
|
|
response.raise_for_status()
|
|
except Exception as exc:
|
|
raise SttError(f"transcription request failed: {exc}") from exc
|
|
try:
|
|
return (
|
|
response.json()
|
|
.get("results", {}).get("channels", [{}])[0]
|
|
.get("alternatives", [{}])[0].get("transcript", "")
|
|
).strip()
|
|
except Exception as exc:
|
|
raise SttError(f"couldn't parse transcription response: {exc}") from exc
|