def transcribe_audio_file(bot, update, path):
chat_id = get_chat_id(update)
lang = TBDB.get_chat_lang(chat_id)
message_id = get_message_id(update)
is_group = chat_id < 0
api_key = config.get_config_prop("wit").get(lang, None)
if api_key is None:
logger.error("Language not found in wit.json %s", lang)
message = bot.send_message(
chat_id=chat_id,
text=R.get_string_resource("unknown_api_key",
lang).format(language=lang) + "\n",
reply_to_message_id=message_id,
parse_mode="html",
is_group=is_group).result()
return
logger.debug("Using key %s for lang %s", api_key, lang)
message = bot.send_message(
chat_id=chat_id,
text=R.get_string_resource("transcribing", lang) + "\n",
reply_to_message_id=message_id,
parse_mode="html",
is_group=is_group).result()
TranscriberBot.get().start_thread(message_id)
logger.debug("Starting thread %d", message_id)
keyboard = InlineKeyboardMarkup(
[[InlineKeyboardButton("Stop", callback_data=message_id)]])
text = ""
if is_group:
text = R.get_string_resource("transcription_text", lang) + "\n"
success = False
for speech in audiotools.transcribe(path, api_key):
logger.debug("Thread %d running: %r", message_id,
TranscriberBot.get().thread_running(message_id))
if TranscriberBot.get().thread_running(message_id) is False:
TranscriberBot.get().del_thread(message_id)
return
retry = True
retry_num = 0
while retry and TranscriberBot.get().thread_running(message_id):
try:
if len(text + " " + speech) >= 4000:
text = R.get_string_resource("transcription_continues",
lang) + "\n"
message = bot.send_message(
chat_id=chat_id,
text=text + " " + speech + " <b>[...]</b>",
reply_to_message_id=message.message_id,
parse_mode="html",
is_group=is_group,
reply_markup=keyboard).result()
else:
message = bot.edit_message_text(
text=text + " " + speech + " <b>[...]</b>",
chat_id=chat_id,
message_id=message.message_id,
parse_mode="html",
is_group=is_group,
reply_markup=keyboard).result()
text += " " + speech
retry = False
success = True
except telegram.error.TimedOut as t:
logger.error("Timeout error %s", traceback.format_exc())
retry_num += 1
if retry_num >= 3:
retry = False
except telegram.error.RetryAfter as r:
logger.warning("Retrying after %d", r.retry_after)
time.sleep(r.retry_after)
except telegram.error.TelegramError as te:
logger.error("Telegram error %s", traceback.format_exc())
retry = False
except Exception as e:
logger.error("Exception %s", traceback.format_exc())
retry = False
retry = True
retry_num = 0
while retry and TranscriberBot.get().thread_running(message_id):
try:
if success:
bot.edit_message_text(text=text,
chat_id=chat_id,
message_id=message.message_id,
parse_mode="html",
is_group=is_group)
else:
bot.edit_message_text(R.get_string_resource(
"transcription_failed", lang),
chat_id=chat_id,
message_id=message.message_id,
parse_mode="html",
is_group=is_group)
retry = False
except telegram.error.TimedOut as t:
logger.error("Timeout error %s", traceback.format_exc())
retry_num += 1
if retry_num >= 3:
retry = False
except telegram.error.RetryAfter as r:
logger.warning("Retrying after %d", r.retry_after)
time.sleep(r.retry_after)
except telegram.error.TelegramError as te:
logger.error("Telegram error %s", traceback.format_exc())
retry = False
except Exception as e:
logger.error("Exception %s", traceback.format_exc())
retry = False
TranscriberBot.get().del_thread(message_id)
def run(self):
path = self.path
lang = "en"
chat_id = get_chat_id(self.update)
message_id = get_message_id(self.update)
is_group = chat_id < 0
print("Started Processing. ID: " + str(message_id))
keyboard = InlineKeyboardMarkup(
[[InlineKeyboardButton("Stop", callback_data=message_id)]])
message = self.bot.send_message(chat_id=chat_id,
text="Transcribing..\n",
reply_to_message_id=message_id,
parse_mode="html",
is_group=is_group)
text = "<b>Text:</b>" + "\n"
success = False
for speech in audiotools.transcribe(path, lang):
retry = True
retry_num = 0
while retry:
try:
if len(text + speech) >= 4080:
text = "<b>[continues]:</b>" + "\n"
message = self.bot.send_message(
chat_id=chat_id,
text=text + speech + " <b>[..]</b>",
reply_to_message_id=message.message_id,
parse_mode="html",
is_group=is_group,
reply_markup=keyboard)
else:
message = self.bot.edit_message_text(
text=text + speech + " <b>[..]</b>",
chat_id=chat_id,
message_id=message.message_id,
parse_mode="html",
is_group=is_group,
reply_markup=keyboard)
text += ' ' + speech
retry = False
success = True
except telegram.error.TimedOut as t:
logger.error("Timeout error %s", traceback.format_exc())
retry_num += 1
if retry_num >= 3:
retry = False
except telegram.error.RetryAfter as r:
logger.warning("Retrying after %d", r.retry_after)
time.sleep(r.retry_after)
except telegram.error.TelegramError as te:
logger.error("Telegram error %s", traceback.format_exc())
retry = False
except Exception as e:
logger.error("Exception %s", traceback.format_exc())
retry = False
retry = True
retry_num = 0
while retry:
try:
if success:
self.bot.edit_message_text(text=text,
chat_id=chat_id,
message_id=message.message_id,
parse_mode="html",
is_group=is_group)
else:
self.bot.edit_message_text("Could not transcribe audio\n",
chat_id=chat_id,
message_id=message.message_id,
parse_mode="html",
is_group=is_group)
retry = False
except telegram.error.TimedOut as t:
logger.error("Timeout error %s", traceback.format_exc())
retry_num += 1
if retry_num >= 3:
retry = False
except telegram.error.RetryAfter as r:
logger.warning("Retrying after %d", r.retry_after)
time.sleep(r.retry_after)
except telegram.error.TelegramError as te:
logger.error("Telegram error %s", traceback.format_exc())
retry = False
except Exception as e:
logger.error("Exception %s", traceback.format_exc())
retry = False
finally:
if os.path.exists(path):
os.remove(path)
print("Completed. ID: " + str(message_id))