matrixhasyou/domovoy_bot
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix: Update gemini_transcribe.py to use Service Account auth and SOCKS5 proxy

Browse source
This commit is contained in:
Admin 2026-07-05 22:10:52 +04:00
parent 6374a50717
commit 3a050af68a
1 changed files with 37 additions and 19 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

52
gemini_transcribe.py
View file

@ -2,11 +2,26 @@
import os
import sys
import time
from google.oauth2 import service_account
import google.generativeai as genai
# Config API key
API_KEY = os.getenv("GOOGLE_API_KEY", "AIzaSyD09NxPpyTk4RZLGqE6dFzJwNxhawRAegc")
genai.configure(api_key=API_KEY)
# Setup proxies for Google API routing through SOCKS5 proxy (RKN bypass)
os.environ["https_proxy"] = "socks5h://127.0.0.1:10808"
os.environ["http_proxy"] = "socks5h://127.0.0.1:10808"
os.environ["all_proxy"] = "socks5h://127.0.0.1:10808"
# Auth using Service Account (avoids invalid GCP API key errors)
SERVICE_ACCOUNT_FILE = "/home/matrixhasyou/domovoy_google_creds.json"
try:
creds = service_account.Credentials.from_service_account_file(
SERVICE_ACCOUNT_FILE,
scopes=["https://www.googleapis.com/auth/generative-language", "https://www.googleapis.com/auth/drive"]
)
genai.configure(credentials=creds)
print("Authenticated successfully using Google Service Account.", flush=True)
except Exception as e:
print(f"Authentication failed: {e}", flush=True)
sys.exit(1)
BOT_TOKEN = "8725618164:AAH1tGalq-pw1l0t4P5c0sdLkCVJdh7IE7M"
CHAT_ID = "197957361"
@ -16,31 +31,33 @@ def send_tg(msg):
import urllib.request
try:
url = f"https://api.telegram.org/bot{BOT_TOKEN}/sendMessage"
data = urllib.parse.urlencode({"chat_id": CHAT_ID, "text": msg, "parse_mode": "markdown"}).encode("utf-8")
# Split message if it's too long for TG limit (4096 chars)
for chunk in [msg[i:i+4000] for i in range(0, len(msg), 4000)]:
data = urllib.parse.urlencode({"chat_id": CHAT_ID, "text": chunk, "parse_mode": "markdown"}).encode("utf-8")
req = urllib.request.Request(url, data=data)
urllib.request.urlopen(req, timeout=15)
except Exception as e:
print(f"Failed to send Telegram message: {e}")
print(f"Failed to send Telegram message: {e}", flush=True)
def transcribe_audio(audio_path, output_prefix, model_name="gemini-1.5-pro"):
if not os.path.exists(audio_path):
print(f"Error: File not found: {audio_path}")
print(f"Error: File not found: {audio_path}", flush=True)
sys.exit(1)
size_mb = os.path.getsize(audio_path) / (1024 * 1024)
print(f"File size: {size_mb:.2f} MB")
print(f"File size: {size_mb:.2f} MB", flush=True)
send_tg(f"🎙 **Транскрибация аудио через Gemini:**\nФайл: `{os.path.basename(audio_path)}` ({size_mb:.1f} MB)\nМодель: `{model_name}`\n\n*Запуск выгрузки в Google API...*")
# 1. Upload file
print(f"Uploading {audio_path} to Google Gemini File API...")
print(f"Uploading {audio_path} to Google Gemini File API...", flush=True)
audio_file = genai.upload_file(path=audio_path)
print(f"File uploaded successfully! API Name: {audio_file.name}")
print(f"File uploaded successfully! API Name: {audio_file.name}", flush=True)
# 2. Wait for processing
start_time = time.time()
while audio_file.state.name == "PROCESSING":
print("File is processing on Google servers... waiting 10s")
print("File is processing on Google servers... waiting 10s", flush=True)
time.sleep(10)
audio_file = genai.get_file(name=audio_file.name)
@ -48,8 +65,8 @@ def transcribe_audio(audio_path, output_prefix, model_name="gemini-1.5-pro"):
send_tg("❌ **Ошибка**: Обработка аудиофайла на сервере Google завершилась неудачей.")
raise ValueError("Audio processing failed on Google servers.")
print("Audio file is active and ready!")
send_tg("⚡️ **Аудиофайл готов к обработке.** Запускаю генерацию текста в Gemini (это может занять до 2-3 минут)...")
print("Audio file is active and ready!", flush=True)
send_tg("⚡️ **Аудиофайл готов.** Запускаю расшифровку речи в Gemini (это может занять до 2-3 минут)...")
# 3. Transcribe with model
model = genai.GenerativeModel(model_name)
@ -75,19 +92,20 @@ def transcribe_audio(audio_path, output_prefix, model_name="gemini-1.5-pro"):
f.write(transcript_text)
elapsed = time.time() - start_time
print(f"Success! Unpacked in {elapsed:.1f}s")
print(f"Success! Unpacked in {elapsed:.1f}s", flush=True)
send_tg(f"✅ **Расшифровка завершена!**\nПродолжительность операции: {elapsed/60:.1f} мин\n\nРезультаты сохранены в:\n- `{md_file}`\n- `{txt_file}`")
# Send transcript to Telegram
send_tg(f"✅ **Расшифровка завершена!** ({elapsed/60:.1f} мин)\nРезультаты сохранены в:\n- `{md_file}`\n- `{txt_file}`\n\n📝 **Текст расшифровки:**\n\n{transcript_text}")
except Exception as e:
print(f"Error during transcription generation: {e}")
print(f"Error during transcription generation: {e}", flush=True)
send_tg(f"❌ **Ошибка во время генерации текста:**\n`{e}`")
finally:
# Delete remote file from Google storage
print("Cleaning up file from Google storage...")
print("Cleaning up file from Google storage...", flush=True)
genai.delete_file(name=audio_file.name)
print("Cleanup done.")
print("Cleanup done.", flush=True)
if __name__ == "__main__":
if len(sys.argv) < 3: