From c1cf78cda48384c35db4522a3fad69ce24b82d4d Mon Sep 17 00:00:00 2001 From: misyaguziya <53165965+misyaguziya@users.noreply.github.com> Date: Thu, 9 Oct 2025 13:34:05 +0900 Subject: [PATCH] =?UTF-8?q?[=E6=94=B9=E5=96=84]=20=E5=9E=8B=E6=B3=A8?= =?UTF-8?q?=E9=87=88=E3=81=AE=E8=BF=BD=E5=8A=A0=E3=81=A8=E3=83=89=E3=82=AD?= =?UTF-8?q?=E3=83=A5=E3=83=A1=E3=83=B3=E3=83=88=E3=81=AE=E6=9B=B4=E6=96=B0?= =?UTF-8?q?=20-=20config.py,=20controller.py,=20model.py,=20mainloop.py,?= =?UTF-8?q?=20utils.py=20=E3=81=A7=E3=81=AE=E5=9E=8B=E6=B3=A8=E9=87=88?= =?UTF-8?q?=E3=81=AE=E8=BF=BD=E5=8A=A0=20-=20CODING=5FRULES.md=20=E3=81=A8?= =?UTF-8?q?=20api.md=20=E3=81=AE=E3=83=89=E3=82=AD=E3=83=A5=E3=83=A1?= =?UTF-8?q?=E3=83=B3=E3=83=88=E3=82=92=E6=9B=B4=E6=96=B0=20-=20=E4=B8=8D?= =?UTF-8?q?=E8=A6=81=E3=81=AA=E3=82=B3=E3=83=BC=E3=83=89=E3=81=AE=E5=89=8A?= =?UTF-8?q?=E9=99=A4=E3=81=A8=E3=83=AA=E3=83=95=E3=82=A1=E3=82=AF=E3=82=BF?= =?UTF-8?q?=E3=83=AA=E3=83=B3=E3=82=B0?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- src-python/config.py | 5 +- src-python/controller.py | 40 ++++++++----- src-python/docs/CODING_RULES.md | 2 - src-python/docs/api.md | 3 - src-python/mainloop.py | 5 +- src-python/model.py | 56 +++++++++++-------- src-python/models/__init__.py | 5 ++ src-python/models/overlay/__init__.py | 5 ++ .../transcription/transcription_recorder.py | 1 - src-python/scripts/find_doc_tokens.py | 1 - src-python/utils.py | 8 +-- 11 files changed, 79 insertions(+), 52 deletions(-) create mode 100644 src-python/models/__init__.py create mode 100644 src-python/models/overlay/__init__.py diff --git a/src-python/config.py b/src-python/config.py index 9500b8fe..8f32f4af 100644 --- a/src-python/config.py +++ b/src-python/config.py @@ -5,6 +5,7 @@ from os import path as os_path, makedirs as os_makedirs from json import load as json_load from json import dump as json_dump import threading +from typing import Optional, Dict, Any import torch from device_manager import device_manager from models.translation.translation_languages import translation_lang @@ -22,8 +23,8 @@ def json_serializable(var_name): class Config: _instance = None - _config_data = {} - _timer = None + _config_data: Dict[str, Any] = {} + _timer: Optional[threading.Timer] = None _debounce_time = 2 def __new__(cls): diff --git a/src-python/controller.py b/src-python/controller.py index d5322134..c7d62402 100644 --- a/src-python/controller.py +++ b/src-python/controller.py @@ -1,5 +1,5 @@ import copy -from typing import Callable, Any +from typing import Callable, Any, List, Optional from time import sleep from subprocess import Popen from threading import Thread @@ -11,10 +11,14 @@ from utils import removeLog, printLog, errorLogging, isConnectedNetwork, isValid class Controller: def __init__(self) -> None: - self.init_mapping = {} - self.run_mapping = {} - self.run = None - self.device_access_status = True + # typed attributes to satisfy static type checkers + self.init_mapping: dict = {} + self.run_mapping: dict = {} + # initialize with a no-op callable so callers can safely call self.run + def _noop_run(status: int, endpoint: str, payload: Any = None) -> None: + return None + self.run: Callable[[int, str, Any], None] = _noop_run + self.device_access_status: bool = True def setInitMapping(self, init_mapping:dict) -> None: self.init_mapping = init_mapping @@ -251,7 +255,7 @@ class Controller: elif isinstance(message, str) and len(message) > 0: translation = [] - transliteration_message = [] + transliteration_message: List[Any] = [] transliteration_translation = [] if model.checkKeywords(message): self.run( @@ -407,7 +411,7 @@ class Controller: ) elif isinstance(message, str) and len(message) > 0: translation = [] - transliteration_message = [] + transliteration_message: List[Any] = [] transliteration_translation = [] if model.checkKeywords(message): self.run( @@ -566,12 +570,12 @@ class Controller: translation_text = f" ({'/'.join(translation)})" if translation else "" model.logger.info(f"[RECEIVED] {message}{translation_text}") - def chatMessage(self, data) -> None: + def chatMessage(self, data) -> dict: id = data["id"] message = data["message"] if len(message) > 0: translation = [] - transliteration_message = [] + transliteration_message: List[Any] = [] transliteration_translation = [] if config.ENABLE_TRANSLATION is False: pass @@ -739,6 +743,7 @@ class Controller: self.run_mapping["software_update_info"], software_update_info, ) + return {"status":200, "result": software_update_info} @staticmethod def getComputeMode(*args, **kwargs) -> dict: @@ -800,11 +805,15 @@ class Controller: if is_vram_error: # Defaultのデバイス設定に戻す printLog("VRAM error detected, reverting device setting") + self.run( + 400, + self.run_mapping["error_translation_enable_vram_overflow"], + { + "message":"VRAM out of memory enabling translation", + "data": error_message + }, + ) self.setDisableTranslation() - config.SELECTED_TRANSLATION_COMPUTE_DEVICE = copy.deepcopy(config.SELECTABLE_COMPUTE_DEVICE_LIST[0]) - config.SELECTED_TRANSLATION_COMPUTE_TYPE = "auto" - self.run(200, self.run_mapping["selected_translation_compute_device"], config.SELECTED_TRANSLATION_COMPUTE_DEVICE) - self.run(200, self.run_mapping["selected_translation_compute_type"], config.SELECTED_TRANSLATION_COMPUTE_TYPE) self.run( 400, self.run_mapping["enable_translation"], @@ -2234,13 +2243,13 @@ class Controller: th_stopCheckSpeakerEnergy.join() @staticmethod - def startThreadingDownloadCtranslate2Weight(weight_type:str, callback:Callable[[float], None], end_callback:Callable[[float], None]) -> None: + def startThreadingDownloadCtranslate2Weight(weight_type:str, callback:Callable[[float], None], end_callback:Optional[Callable[..., None]] = None) -> None: th_download = Thread(target=model.downloadCTranslate2ModelWeight, args=(weight_type, callback, end_callback)) th_download.daemon = True th_download.start() @staticmethod - def startThreadingDownloadWhisperWeight(weight_type:str, callback:Callable[[float], None], end_callback:Callable[[float], None]) -> None: + def startThreadingDownloadWhisperWeight(weight_type:str, callback:Callable[[float], None], end_callback:Optional[Callable[..., None]] = None) -> None: th_download = Thread(target=model.downloadWhisperModelWeight, args=(weight_type, callback, end_callback)) th_download.daemon = True th_download.start() @@ -2258,6 +2267,7 @@ class Controller: @staticmethod def setWatchdogCallback(callback) -> dict: model.setWatchdogCallback(callback) + return {"status":200, "result":True} @staticmethod def stopWatchdog(*args, **kwargs) -> dict: diff --git a/src-python/docs/CODING_RULES.md b/src-python/docs/CODING_RULES.md index 19a626b0..4d6d2dd2 100644 --- a/src-python/docs/CODING_RULES.md +++ b/src-python/docs/CODING_RULES.md @@ -29,8 +29,6 @@ - 定数: UPPER_SNAKE_CASE（`config.py` の定数に合わせる）。 - run_mapping のキー: 現在は短い key（例: `transcription_mic`）を内部で使い `run_mapping` に `/run/...` を置いている。この慣習は維持する。Controller 内で `self.run_mapping[...]` を直接参照する実装は許容される。 -例: `selected_translation_compute_device` は内部 key、`/run/selected_translation_compute_device` が外部イベント名である点を区別して使う。 - ## モジュール・パッケージ構成 - 各サブ領域（ocr, overlay, transcription, translation, websocket 等）は `models/` 下に整理済みのため、同様の粒度で新機能は追加する。 - パッケージは必ず `__init__.py` を置く（static analysis / mypy のため）。空の `__init__.py` でも可。これにより相対インポートが安定する。 diff --git a/src-python/docs/api.md b/src-python/docs/api.md index 373e29d4..f512a191 100644 --- a/src-python/docs/api.md +++ b/src-python/docs/api.md @@ -130,9 +130,6 @@ run イベント `/run/selected_transcription_compute_type` (200) - payload: string -`/run/selected_translation_compute_device` (200) - - payload: device descriptor (e.g. {"name":"cuda:0","type":"gpu"}) - `/run/selected_translation_engines` (200) - payload: config.SELECTED_TRANSLATION_ENGINES (list/dict per tab) diff --git a/src-python/mainloop.py b/src-python/mainloop.py index 7fd9fcc2..9315ab56 100644 --- a/src-python/mainloop.py +++ b/src-python/mainloop.py @@ -1,7 +1,7 @@ import sys import json import time -from typing import Any +from typing import Any, Tuple from threading import Thread from queue import Queue import logging @@ -359,7 +359,8 @@ controller.setInitMapping(init_mapping) class Main: def __init__(self, controller_instance, mapping_data) -> None: - self.queue = Queue() + # queue holds tuples of (endpoint, data) + self.queue: Queue[Tuple[str, Any]] = Queue() self.main_loop = True self.controller = controller_instance self.mapping = mapping_data diff --git a/src-python/model.py b/src-python/model.py index bbf43604..e78ed857 100644 --- a/src-python/model.py +++ b/src-python/model.py @@ -10,7 +10,7 @@ from time import sleep from queue import Queue from threading import Thread from requests import get as requests_get -from typing import Callable +from typing import Callable, Optional, cast from packaging.version import parse from flashtext import KeywordProcessor @@ -106,6 +106,9 @@ class Model: self.websocket_server_loop = False self.websocket_server_alive = False self.th_websocket_server = None + # default no-op callbacks for energy check functions + self.check_mic_energy_fnc: Callable[[float], None] = lambda v: None + self.check_speaker_energy_fnc: Callable[[float], None] = lambda v: None def checkTranslatorCTranslate2ModelWeight(self, weight_type:str): return checkCTranslate2Weight(config.PATH_LOCAL, weight_type) @@ -291,9 +294,9 @@ class Model: if self.transliterator is not None: self.transliterator = None - def convertMessageToTransliteration(self, message: str, hiragana: bool=True, romaji: bool=True) -> str: + def convertMessageToTransliteration(self, message: str, hiragana: bool=True, romaji: bool=True) -> list: if hiragana is False and romaji is False: - return message + return [] keys_to_keep = {"orig"} if hiragana: @@ -574,9 +577,10 @@ class Model: # self.mic_get_energy.stop() # self.mic_get_energy = None - def startCheckMicEnergy(self, fnc:Callable[[float], None]=None) -> None: - if isinstance(fnc, Callable): - self.check_mic_energy_fnc = fnc + def startCheckMicEnergy(self, fnc:Optional[Callable[[float], None]]=None) -> None: + # fnc may be None or a callable. Use cast after checking for None to satisfy type checker. + if fnc is not None: + self.check_mic_energy_fnc = cast(Callable[[float], None], fnc) mic_host_name = config.SELECTED_MIC_HOST mic_device_name = config.SELECTED_MIC_DEVICE @@ -596,7 +600,7 @@ class Model: errorLogging() sleep(0.01) - mic_energy_queue = Queue() + mic_energy_queue: Queue = Queue() mic_device = selected_mic_device[0] self.mic_energy_recorder = SelectedMicEnergyRecorder(mic_device) self.mic_energy_recorder.recordIntoQueue(mic_energy_queue) @@ -614,17 +618,18 @@ class Model: self.mic_energy_recorder.stop() self.mic_energy_recorder = None - def startSpeakerTranscript(self, fnc): + def startSpeakerTranscript(self, fnc:Optional[Callable[[dict], None]]=None) -> None: speaker_device_name = config.SELECTED_SPEAKER_DEVICE speaker_device_list = device_manager.getSpeakerDevices() selected_speaker_device = [device for device in speaker_device_list if device["name"] == speaker_device_name] if len(selected_speaker_device) == 0 or speaker_device_name == "NoDevice": - fnc({"text": False, "language": None}) + # fnc may be None; only call if callable + if callable(fnc): + fnc({"text": False, "language": None}) else: - speaker_audio_queue = Queue() - # speaker_energy_queue = Queue() + speaker_audio_queue: Queue = Queue() speaker_device = selected_speaker_device[0] record_timeout = config.SPEAKER_RECORD_TIMEOUT phrase_timeout = config.SPEAKER_PHRASE_TIMEOUT @@ -708,9 +713,10 @@ class Model: # self.speaker_get_energy.stop() # self.speaker_get_energy = None - def startCheckSpeakerEnergy(self, fnc:Callable[[float], None]=None) -> None: - if isinstance(fnc, Callable): - self.check_speaker_energy_fnc = fnc + def startCheckSpeakerEnergy(self, fnc:Optional[Callable[[float], None]]=None) -> None: + # Accept None as default and assign safely with cast after None-check + if fnc is not None: + self.check_speaker_energy_fnc = cast(Callable[[float], None], fnc) speaker_device_name = config.SELECTED_SPEAKER_DEVICE speaker_device_list = device_manager.getSpeakerDevices() @@ -720,7 +726,7 @@ class Model: self.check_speaker_energy_fnc(False) else: def sendSpeakerEnergy(): - if speaker_energy_queue.empty() is False: + if not speaker_energy_queue.empty(): energy = speaker_energy_queue.get() try: self.check_speaker_energy_fnc(energy) @@ -728,7 +734,7 @@ class Model: errorLogging() sleep(0.01) - speaker_energy_queue = Queue() + speaker_energy_queue: Queue = Queue() speaker_device = selected_speaker_device[0] self.speaker_energy_recorder = SelectedSpeakerEnergyRecorder(speaker_device) self.speaker_energy_recorder.recordIntoQueue(speaker_energy_queue) @@ -746,9 +752,12 @@ class Model: self.speaker_energy_recorder.stop() self.speaker_energy_recorder = None - def createOverlayImageSmallLog(self, message:str, your_language:str, translation:list, target_language:dict): - target_language = [data["language"] for data in target_language.values() if data["enable"] is True] - return self.overlay_image.createOverlayImageSmallLog(message, your_language, translation, target_language) + def createOverlayImageSmallLog(self, message:Optional[str], your_language:Optional[str], translation:list, target_language:Optional[dict]) -> object: + # target_language may be provided as dict or None + target_language_list = [] + if isinstance(target_language, dict): + target_language_list = [data["language"] for data in target_language.values() if data.get("enable") is True] + return self.overlay_image.createOverlayImageSmallLog(message, your_language, translation, target_language_list) def createOverlayImageSmallMessage(self, message): ui_language = config.UI_LANGUAGE @@ -797,9 +806,12 @@ class Model: if (self.overlay.settings[size]["ui_scaling"] != config.OVERLAY_SMALL_LOG_SETTINGS["ui_scaling"]): self.overlay.updateUiScaling(config.OVERLAY_SMALL_LOG_SETTINGS["ui_scaling"], size) - def createOverlayImageLargeLog(self, message_type:str, message:str, your_language:str, translation:list, target_language:dict): - target_language = [data["language"] for data in target_language.values() if data["enable"] is True] - return self.overlay_image.createOverlayImageLargeLog(message_type, message, your_language, translation, target_language) + def createOverlayImageLargeLog(self, message_type:str, message:Optional[str], your_language:Optional[str], translation:list, target_language:Optional[dict]=None): + # normalize target_language dict -> list of language strings + target_language_list = [] + if isinstance(target_language, dict): + target_language_list = [data["language"] for data in target_language.values() if data.get("enable") is True] + return self.overlay_image.createOverlayImageLargeLog(message_type, message, your_language, translation, target_language_list) def createOverlayImageLargeMessage(self, message): ui_language = config.UI_LANGUAGE diff --git a/src-python/models/__init__.py b/src-python/models/__init__.py new file mode 100644 index 00000000..4d9d80d5 --- /dev/null +++ b/src-python/models/__init__.py @@ -0,0 +1,5 @@ +"""models package init for static analysis and packaging.""" + +__all__ = [ + # subpackages are discovered implicitly +] diff --git a/src-python/models/overlay/__init__.py b/src-python/models/overlay/__init__.py new file mode 100644 index 00000000..33f43c7b --- /dev/null +++ b/src-python/models/overlay/__init__.py @@ -0,0 +1,5 @@ +"""models.overlay package init for static analysis.""" + +from . import overlay_utils # re-export helper for ease-of-use in tooling + +__all__ = ["overlay_utils"] diff --git a/src-python/models/transcription/transcription_recorder.py b/src-python/models/transcription/transcription_recorder.py index f30c071f..b574013c 100644 --- a/src-python/models/transcription/transcription_recorder.py +++ b/src-python/models/transcription/transcription_recorder.py @@ -1,7 +1,6 @@ from speech_recognition import Recognizer, Microphone from pyaudiowpatch import get_sample_size, paInt16 from datetime import datetime -from queue import Queue class BaseRecorder: def __init__(self, source, energy_threshold, dynamic_energy_threshold, record_timeout): diff --git a/src-python/scripts/find_doc_tokens.py b/src-python/scripts/find_doc_tokens.py index b9e1c95f..7b9fae1a 100644 --- a/src-python/scripts/find_doc_tokens.py +++ b/src-python/scripts/find_doc_tokens.py @@ -7,7 +7,6 @@ tokens = [ 'transcription_mic', 'transcription_speaker', 'selected_translation_compute_device', - '/run/selected_translation_compute_device', '/run/transcription_mic', '/run/transcription_speaker', ] diff --git a/src-python/utils.py b/src-python/utils.py index fab62d51..fe4faa90 100644 --- a/src-python/utils.py +++ b/src-python/utils.py @@ -1,5 +1,5 @@ import base64 -from typing import Any +from typing import Any, List, Dict import json import traceback import logging @@ -79,7 +79,7 @@ def isValidIpAddress(ip_address: str) -> bool: except ValueError: return False -def getComputeDeviceList() -> dict: +def getComputeDeviceList() -> List[Dict[str, Any]]: compute_types = [ { "device": "cpu", @@ -191,8 +191,8 @@ def printLog(log:str, data:Any=None) -> None: "data": str(data), } process_logger.info(response) - response = json.dumps(response) - print(response, flush=True) + serialized = json.dumps(response) + print(serialized, flush=True) def printResponse(status:int, endpoint:str, result:Any=None) -> None: global process_logger