dasemu/Transcriptarr
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Default LanguageCode inputs to from_string and fix detect_langauge

Browse Source
This commit is contained in:
McCloudS
2024-12-03 11:35:40 -07:00
committed by GitHub
parent 84243bb34e
commit bf9d29475f
1 changed files with 17 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
subgen.py
View File

@@ -68,21 +68,21 @@ detect_language_length = os.getenv('DETECT_LANGUAGE_LENGTH', 30)
detect_language_start_offset = os.getenv('DETECT_LANGUAGE_START_OFFSET', int(0)) detect_language_start_offset = os.getenv('DETECT_LANGUAGE_START_OFFSET', int(0))
skipifexternalsub = convert_to_bool(os.getenv('SKIPIFEXTERNALSUB', False)) skipifexternalsub = convert_to_bool(os.getenv('SKIPIFEXTERNALSUB', False))
skip_if_to_transcribe_sub_already_exist = convert_to_bool(os.getenv('SKIP_IF_TO_TRANSCRIBE_SUB_ALREADY_EXIST', True)) skip_if_to_transcribe_sub_already_exist = convert_to_bool(os.getenv('SKIP_IF_TO_TRANSCRIBE_SUB_ALREADY_EXIST', True))
skipifinternalsublang = LanguageCode.from_iso_639_2(os.getenv('SKIPIFINTERNALSUBLANG', '')) skipifinternalsublang = LanguageCode.from_string(os.getenv('SKIPIFINTERNALSUBLANG', ''))
skip_lang_codes_list = ( skip_lang_codes_list = (
[LanguageCode.from_iso_639_2(code) for code in os.getenv("SKIP_LANG_CODES", "").split("|")] [LanguageCode.from_string(code) for code in os.getenv("SKIP_LANG_CODES", "").split("|")]
if os.getenv('SKIP_LANG_CODES') if os.getenv('SKIP_LANG_CODES')
else [] else []
) )
force_detected_language_to = LanguageCode.from_iso_639_2(os.getenv('FORCE_DETECTED_LANGUAGE_TO', '')) force_detected_language_to = LanguageCode.from_string(os.getenv('FORCE_DETECTED_LANGUAGE_TO', ''))
preferred_audio_languages = ( preferred_audio_languages = (
[LanguageCode.from_iso_639_2(code) for code in os.getenv('PREFERRED_AUDIO_LANGUAGES', 'eng').split("|")] [LanguageCode.from_string(code) for code in os.getenv('PREFERRED_AUDIO_LANGUAGES', 'eng').split("|")]
if os.getenv('PREFERRED_AUDIO_LANGUAGES') if os.getenv('PREFERRED_AUDIO_LANGUAGES')
else [] else []
) # in order of preferrence ) # in order of preferrence
limit_to_preferred_audio_languages = convert_to_bool(os.getenv('LIMIT_TO_PREFERRED_AUDIO_LANGUAGE', False)) #TODO: add support for this limit_to_preferred_audio_languages = convert_to_bool(os.getenv('LIMIT_TO_PREFERRED_AUDIO_LANGUAGE', False)) #TODO: add support for this
skip_if_audio_track_is_in_list = ( skip_if_audio_track_is_in_list = (
[LanguageCode.from_iso_639_2(code) for code in os.getenv('SKIP_IF_AUDIO_TRACK_IS', '').split("|")] [LanguageCode.from_string(code) for code in os.getenv('SKIP_IF_AUDIO_TRACK_IS', '').split("|")]
if os.getenv('SKIP_IF_AUDIO_TRACK_IS') if os.getenv('SKIP_IF_AUDIO_TRACK_IS')
else [] else []
) )
@@ -129,13 +129,12 @@ def transcription_worker():
while True: while True:
task = task_queue.get() task = task_queue.get()
logger.info(f"Task {task['path']} is being handled by Subgen.")
if 'Bazarr-' in task['path']:
logging.info(f"Task {task['path']} is being handled by ASR.")
if "type" in task and task["type"] == "detect_language": if "type" in task and task["type"] == "detect_language":
detect_language_task(task['path']) detect_language_task(task['path'])
if 'Bazarr-' in task['path']:
logging.info(f"Task {task['path']} is being handled by ASR.")
else: else:
logging.info(f"Task {task['path']} is being handled by Subgen.")
gen_subtitles(task['path'], task['transcribe_or_translate'], task['force_language']) gen_subtitles(task['path'], task['transcribe_or_translate'], task['force_language'])
task_queue.task_done() task_queue.task_done()
# show queue # show queue
@@ -370,7 +369,7 @@ async def asr(
random_name = ''.join(random.choices("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ1234567890", k=6)) random_name = ''.join(random.choices("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ1234567890", k=6))
if force_detected_language_to: if force_detected_language_to:
language = force_detected_language_to.from_iso_639_1() language = force_detected_language_to.to_iso_639_1()
logging.info(f"ENV FORCE_DETECTED_LANGUAGE_TO is set: Forcing detected language to {force_detected_language_to}") logging.info(f"ENV FORCE_DETECTED_LANGUAGE_TO is set: Forcing detected language to {force_detected_language_to}")
start_time = time.time() start_time = time.time()
@@ -382,11 +381,14 @@ async def asr(
args = {} args = {}
args['progress_callback'] = progress args['progress_callback'] = progress
if not encode: file_content = audio_file.file.read()
args['audio'] = np.frombuffer(audio_file.file.read(), np.int16).flatten().astype(np.float32) / 32768.0
args['input_sr'] = 16000 if encode:
args['audio'] = file_content()
else: else:
args['audio'] = audio_file.file.read() args['audio'] = np.frombuffer(file_content, np.int16).flatten().astype(np.float32) / 32768.0
args['input_sr'] = 16000
if custom_regroup: if custom_regroup:
args['regroup'] = custom_regroup args['regroup'] = custom_regroup