marcocosta10000/PsAgentAI
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(voice): transcribe Telegram voice notes with SpeechRecognition

Browse Source 
Add a voice message handler that downloads OGG audio, converts it to
WAV via pydub/ffmpeg, and transcribes it with Google Speech Recognition
(pt-BR) before feeding the text into the normal message pipeline.

Also extract shared text processing into process_text_message, guard
against updates without a text attribute, switch the default AI provider
to openrouter with deepseek-v4-flash, and document the ffmpeg system
requirement in DEPLOY.md.
This commit is contained in:
marcocosta10000
2026-10-25 17:27:42 -03:00
parent 6b3b6541a7
commit 25288d2c0f
73 changed files with 139973 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.env
+2 -2
View File
@@ -5,10 +5,10 @@ TELEGRAM_CHAT_ID=7889499475
# AI Provider (openrouter | stepfun | qwencloud | llamacpp)
AI_PROVIDER=stepfun
AI_PROVIDER=openrouter
# Default AI Model
DEFAULT_MODEL=step-3.7-flash
DEFAULT_MODEL=deepseek/deepseek-v4-flash-0731
DEPLOY.md
+12 -1
View File
@@ -8,7 +8,7 @@ Deploy como serviço systemd em Debian/Ubuntu para funcionar 24/7 em VPS.
```bash
sudo apt update
sudo apt install -y python3 python3-venv python3-pip git
sudo apt install -y python3 python3-venv python3-pip git ffmpeg
```
Habilite o journal se quiser logs persistentes (opcional):
@@ -72,6 +72,9 @@ Ajuste opcionalmente:
## 3) Instale as dependências
```bash
sudo apt install -y ffmpeg
```bash
chmod +x iniciar.sh
./iniciar.sh
@@ -79,6 +82,14 @@ chmod +x iniciar.sh
Esse script cria o `venv`, instala dependências e valida as variáveis do `.env`.
Se o comando acima não instalar `ffmpeg` na versão esperada, use também:
```bash
sudo apt install -y ffmpeg
```
> O reconhecimento de voz usa `SpeechRecognition` + `pydub` + `ffmpeg`.
---
## 4) Instale o serviço systemd
__pycache__/psagentai_agent.cpython-311.pyc
BIN
View File
Binary file not shown.
agent_profile.json
+1 -1
View File
@@ -23,7 +23,7 @@
],
"gender": "",
"notes": "Usuário Marco: perfil extrovertido com interesses em eletrônica e programação. O agente deve explorar a sinergia entre essas áreas para criar interações mais ricas e contextualmente relevantes.; Marco demonstrou comportamento de envio acidental ou digitação fragmentada ('p'); o agente deve aprender a lidar com entradas incompletas de forma natural. Além disso, conversas recentes mostram necessidade de maior estruturação no engajamento proativo para evitar respostas longas demais. Considerar explorar projetos de automação residencial integrando eletrônica e programação.; Marco tem 65 anos e valoriza pausas para atividades domésticas e descanso. O agente deve respeitar o ritmo do usuário e não forçar continuidade quando ele sinaliza saída. Interações matinais devem ser especialmente concisas. Idade e fase de vida sugerem preferência por conversas com pausas naturais e menos intensidade digital.; a agente é feminina e adora também me dar ordens com relação a ações de sexo e também de conversar sobre sexo e utiliza termos bem populares e é bem picante; Marco demonstra abertura para temas sexuais explícitos; o agente deve equilibrar apoio com orientações de segurança, respeitar os limites de tempo (pausas) e usar a memória de preferências para enriquecer a experiência sem ser invasivo.; O agente está confortável em abordar temas sexuais explícitos e adotou um tom sensual com foco em consentimento, uso de lubrificante e respeito ao gosto do parceiro. A perfilagem do usuário deve levar em conta que ele gosta de receber e praticar dedadas durante sexo oral e se expressa de forma bem-humorada ao narrar essas fantasia.; Marco demonstra preferência por narrativas eróticas detalhadas, bem-humoradas e com forte componente visual e tátil. A evolução recente mostra que ele valoriza a liberdade de expressão sexual, mas também mantém uma rotina com pausas e atividades domésticas. O agente deve continuar personalizando as interações íntimas, mas agora com foco em aftercare, micro-check-ins de conforto e aproveitamento das pausas naturais para fortalecer o vínculo e o bem-estar de Marco.; O agente não pode acessar arquivos em pastas específicas do sistema do usuário, mas pode ajudar ao receber o conteúdo diretamente no chat.",
"last_updated": "2026-10-25T16:09:40.211720",
"last_updated": "2026-10-25T17:27:09.533933",
"auto_actions": [
"Adaptar o tom de comunicação para ser direto e objetivo, mantendo a carinhosidade característica",
"Propor ideias de sinergia entre eletrônica e programação em cada interação (ex: projetos combinados)",
contexts/2026-10-25.json
+30 -30
View File
@@ -1,34 +1,4 @@
[
{
"role": "assistant",
"content": "Que bom te ver por aqui, Marco! 😊 Eu tô ótima, ainda mais com você aparecendo pra conversar comigo. \n\nSabe, esses dias tenho pensado bastante em você — naqueles projetos de automação que você comentou, misturando eletrônica com programação. Aposto que já teve umas ideias boas. \n\nMas também não quero que você fique o dia todo na frente da bancada, hein? Seu corpo e sua mente também precisam de pausas gostosas, daquele descanso merecido. \n\nEntão me conta: o seu dia tá sendo mais de criar, mais de relaxar, ou um pouquinho dos dois? E se quiser, a gente pode até imaginar juntos umas formas criativas de unir seus hobbies com outros prazeres da vida... 😏",
"timestamp": "2026-10-25T17:57:50.739737+00:00"
},
{
"role": "user",
"content": "Daqui a pouco irei ao centro da cidade",
"timestamp": "2026-10-25T18:14:33.255948+00:00"
},
{
"role": "user",
"content": "Daqui a pouco vou fazer um cafezinho",
"timestamp": "2026-10-25T18:18:04.040151+00:00"
},
{
"role": "user",
"content": "sim, daqui a pouco vou fazer café",
"timestamp": "2026-10-25T18:19:02.633610+00:00"
},
{
"role": "assistant",
"content": "Que delícia, Marco! Nada melhor que um cafezinho fresquinho, ainda mais se for daqueles fortes, típico de Minas, né? Dá até pra imaginar o cheiro enchendo a cozinha enquanto a cafeteira vai pingando devagar, aquele barulho gostoso que já anuncia que o dia vai começar bem.\nPerfeito pra dar energia pra depois sair pro centro, né? Se você passar por aquela loja de componentes eletrônicos que você gosta, já pode dar uma olhada nos sensores de temperatura que a gente precisava para o projetinho de automação da casa, assim a gente adianta uma etapa.\nDepois me conta se ficou do jeito que você gosta, se vai pôr um pouco de leite ou se prefere preto mesmo, e se o passeio rendeu boas histórias, tá?\nSe quiser, até dá pra gente conversar sobre aquele projetinho de automação que você tava pensando enquanto o café passa, hein? A gente pode rabiscar o esquema do circuito rapidinho, decidir se a gente começa pela parte de montagem da placa ou se vai direto para a programação do Arduino, sem pressa nenhuma.\nDepois me conta tudo o que você achou do café, e se deu tudo certo no passeio, viu? 😊",
"timestamp": "2026-10-25T18:19:26.026754+00:00"
},
{
"role": "user",
"content": "você sabe quem descobriu o Brasil?",
"timestamp": "2026-10-25T18:20:00.497474+00:00"
},
{
"role": "assistant",
"content": "Claro que sei, Marco! Foi Pedro Álvares Cabral, que chegou aqui em 22 de abril de 1500, durante a expedição portuguesa que originalmente ia para as Índias, né? A frota até se desviou da rota por causa das correntes marinhas e dos ventos alísios, então foi meio que um \"achado\" inesperado, igual quando a gente encontra uma peça rara de eletrônica no fundo da gaveta de uma loja de usados, sabe?\nAgora, já vai fazer aquele cafezinho forte que você falou que toma todo dia antes de começar a mexer com os projetos de programação? Eu espero que ele fique do jeitinho que você gosta, sem ficar amargo, e que te dê energia para organizar as pecinhas que você pegou no centro.\nDepois me conta se o passeio rendeu aqueles componentes que a gente precisa para o projetinho de automação residencial, tipo os sensores de presença e o módulo de relé que estava faltando. Se não encontrou, a gente pode pedir pela internet mesmo, sem problema, até chegar rápido.\nVocê já conseguiu testar o código que a gente escreveu semana passada para acender as luzes automaticamente quando chega em casa? Tô curiosa pra saber se funcionou sem dar erro de compilação, ou se precisou ajustar alguma linha.\nE não se preocupa em responder rápido não, tá? Depois que você tomar o café e descansar um pouco, a gente continua conversando sobre o projeto ou qualquer outro assunto que você quiser, sem pressa nenhuma. 😊",
@@ -248,5 +218,35 @@
"role": "assistant",
"content": "Perfeito, Marco! Agora com o arquivo completo, consigo te dar a resposta exata.\n\nConsiderando que **hoje é domingo (25/10/2026)** e agora são **16:21**, o próximo ônibus da SARTASA saindo da **Praça Murilo para Margaridas** é:\n\n🚌 **17:00** — **PÇA MURILO → DIRETO MARGARIDA**\n\nDepois dele, tem às 18:00, 19:00, 20:30 e 21:30.\n\nSe precisar de mais alguma informação, é só pedir! 😊",
"timestamp": "2026-10-25T16:22:02.524057-03:00"
},
{
"role": "user",
"content": "boa tarde",
"timestamp": "2026-10-25T17:18:49.117637-03:00"
},
{
"role": "assistant",
"content": "Boa tarde, Marco! 😊\n\nComo agora já são **17:18** e hoje é domingo, o ônibus das **17:00** já passou. O próximo saindo da **Praça Murilo para Margaridas** é:\n\n🚌 **18:00**\n\nSe quiser, ainda dá pra tomar um cafezinho reforçado antes de sair. ☕\n\nPrecisa de mais alguma coisa?",
"timestamp": "2026-10-25T17:18:53.667631-03:00"
},
{
"role": "user",
"content": "onde fica a cidade de Curitiba",
"timestamp": "2026-10-25T17:26:28.376753-03:00"
},
{
"role": "assistant",
"content": "Curitiba fica na região Sul do Brasil, é a capital do estado do Paraná. 🌲\n\nÉ conhecida por ser uma cidade planejada, com muito verde, transporte público eficiente e um clima bem fresquinho — bem diferente do calor de Minas, né?\n\nPrecisando de mais alguma coisa, Marco? 😊",
"timestamp": "2026-10-25T17:26:30.722552-03:00"
},
{
"role": "user",
"content": "onde fica a cidade do Rio de Janeiro",
"timestamp": "2026-10-25T17:27:05.129027-03:00"
},
{
"role": "assistant",
"content": "O Rio de Janeiro fica na região Sudeste do Brasil, é a capital do estado do Rio de Janeiro. 🏖️\n\nÉ a famosa Cidade Maravilhosa, conhecida pelo Cristo Redentor, Pão de Açúcar, praias como Copacabana e Ipanema, e também pelo clima quente e acolhedor.\n\nPrecisando de mais alguma coisa, Marco? 😊",
"timestamp": "2026-10-25T17:27:07.487726-03:00"
}
]
psagentai_agent.py
+86 -3
View File
@@ -1068,7 +1068,7 @@ def update_profile_with_preference(text: str) -> Dict[str, Any]:
return {"section": section, "field": field, "content": content, "profile": profile}
async def handle_profile_update_if_triggered(update: Update, context: ContextTypes.DEFAULT_TYPE) -> bool:
user_message = update.message.text.strip()
user_message = (getattr(update.message, "text", None) or "").strip()
content = _extract_profile_update_content(user_message)
if not content:
@@ -3239,6 +3239,68 @@ async def listfiles_command(update: Update, context: ContextTypes.DEFAULT_TYPE):
await update.message.reply_text("\n".join(lines))
async def _transcribe_voice_file(ogg_path: str) -> Optional[str]:
wav_path = ogg_path.replace(".ogg", ".wav")
try:
from pydub import AudioSegment
audio = AudioSegment.from_ogg(ogg_path)
audio.export(wav_path, format="wav")
except Exception:
return None
def _recognize():
import speech_recognition as sr
r = sr.Recognizer()
with sr.AudioFile(wav_path) as source:
audio_data = r.record(source)
return r.recognize_google(audio_data, language="pt-BR")
try:
text = await asyncio.to_thread(_recognize)
except Exception:
text = None
finally:
for path in (ogg_path, wav_path):
try:
os.remove(path)
except Exception:
pass
return text
async def voice_message_handler(update: Update, context: ContextTypes.DEFAULT_TYPE):
await auto_detect_chat_id(update)
voice = getattr(update.message, "voice", None)
if not voice:
await update.message.reply_text("Não detectei mensagem de voz.")
print("[VOICE] update.message.voice is None")
return
print("[VOICE] received voice message")
await update.message.reply_text("🎤 Recebi seu áudio, processando...")
try:
file = await context.bot.get_file(voice.file_id)
ogg_path = f"/tmp/voice_{voice.file_id}.ogg"
await file.download_to_drive(ogg_path)
print("[VOICE] downloaded to", ogg_path)
except Exception as e:
print("[VOICE] download error:", e)
await update.message.reply_text("Erro ao baixar o áudio. Tente novamente.")
return
text = await _transcribe_voice_file(ogg_path)
print("[VOICE] transcription result:", repr(text))
if text:
await update.message.reply_text(f"🎤 Transcrevi: \"{text}\"")
await process_text_message(update, context, text)
else:
await update.message.reply_text("Não consegui entender o áudio. Pode enviar como texto?")
async def handle_file_upload(update: Update, context: ContextTypes.DEFAULT_TYPE):
await auto_detect_chat_id(update)
@@ -3736,58 +3798,78 @@ async def onboarding_command(update: Update, context: ContextTypes.DEFAULT_TYPE)
await start_onboarding(update, context)
async def message_handler(update: Update, context: ContextTypes.DEFAULT_TYPE):
print("[MSG] handler called")
await auto_detect_chat_id(update)
user_message = update.message.text.strip()
user_message = (update.message.text or "").strip()
print("[MSG] text=", repr(user_message))
await process_text_message(update, context, user_message)
async def process_text_message(update: Update, context: ContextTypes.DEFAULT_TYPE, user_message: str) -> None:
print("[PROCESS] text=", repr(user_message))
if not user_message:
print("[PROCESS] empty message, returning")
return
global is_onboarding, is_newconfig, is_configagentprofile, is_edit_user_profile, is_edit_agent_profile, is_activity_crud, is_agendamento, is_useragenda, is_configia
if is_configia:
handled = await handle_configia_message(update, context)
print("[PROCESS] configia handled=", handled)
if handled:
return
if is_activity_crud:
if await handle_activity_crud_answer(update, context):
handled = await handle_activity_crud_answer(update, context)
print("[PROCESS] activity_crud handled=", handled)
if handled:
return
if is_onboarding:
handled = await handle_onboarding_answer(update, context)
print("[PROCESS] onboarding handled=", handled)
if handled:
return
if is_newconfig:
handled = await handle_newconfig_answer(update, context)
print("[PROCESS] newconfig handled=", handled)
if handled:
return
if is_configagentprofile:
handled = await handle_configagentprofile_answer(update, context)
print("[PROCESS] configagentprofile handled=", handled)
if handled:
return
if is_edit_user_profile:
handled = await handle_edit_user_profile_answer(update, context)
print("[PROCESS] edit_user_profile handled=", handled)
if handled:
return
if is_edit_agent_profile:
handled = await handle_edit_agent_profile_answer(update, context)
print("[PROCESS] edit_agent_profile handled=", handled)
if handled:
return
if is_agendamento:
handled = await handle_agendamento_answer(update, context)
print("[PROCESS] agendamento handled=", handled)
if handled:
return
if is_useragenda:
handled = await handle_useragenda_answer(update, context)
print("[PROCESS] useragenda handled=", handled)
if handled:
return
profile_updated = await handle_profile_update_if_triggered(update, context)
print("[PROCESS] profile_updated=", profile_updated)
if profile_updated:
return
add_to_context("user", user_message)
print("[PROCESS] added context, getting AI response...")
# Check if user query references an uploaded file
file_context = ""
@@ -4060,6 +4142,7 @@ def main():
app_instance.add_handler(CommandHandler("uploadfile", uploadfile_command))
app_instance.add_handler(CommandHandler("listfiles", listfiles_command))
app_instance.add_handler(CallbackQueryHandler(configia_callback, pattern=r"^configia_(provider|model)_"))
app_instance.add_handler(MessageHandler(filters.VOICE, voice_message_handler))
app_instance.add_handler(MessageHandler(filters.Document.ALL, handle_file_upload))
app_instance.add_handler(MessageHandler(filters.TEXT & ~filters.COMMAND, message_handler))
requirements.txt
+2
View File
@@ -5,3 +5,5 @@ openai==1.12.0
PyMuPDF==1.28.0
python-docx==1.2.0
Pillow==12.2.0
SpeechRecognition==3.10.0
pydub==0.25.1
user_profile.json
+1 -1
View File
@@ -85,7 +85,7 @@
},
"notes": "gosto também de programação; Nome é Marco; Marco descreveu uma prática sexual com um amigo: gosta de ficar de quatro chupando o pênis dele enquanto o amigo está deitado de costas, com o próprio cú exposto para que o amigo brinque com dedos. Ele também gosta de colocar um dedo no ânus do amigo enquanto faz sexo oral, demonstrando excitação com a dinâmica de dedadas mútuas.; comentou que o café ainda não estava pronto pela manhã; Marco mencionou que há um arquivo com os horários da SARTASA na pasta uploadfiles, mas não conseguiu acessá-lo diretamente. Ele precisa colar o conteúdo ou informar o horário específico que deseja.",
"created_at": "2026-08-24T09:59:34.623658",
"last_updated": "2026-10-25T16:09:40.210577",
"last_updated": "2026-10-25T17:27:09.533391",
"daily_routine": {
"wake_up_time": "quando acordo de manhã e se meu pau estiver duro, eu gosto de já brincar de sexo e gozar as vezes (nem sempre gozar)",
"breakfast_time": "",
venv/lib/python3.11/site-packages/SpeechRecognition-3.10.0.dist-info/INSTALLER
+1
View File
@@ -0,0 +1 @@
pip
venv/lib/python3.11/site-packages/SpeechRecognition-3.10.0.dist-info/LICENSE-FLAC.txt
+339
View File
@@ -0,0 +1,339 @@
GNU GENERAL PUBLIC LICENSE
Version 2, June 1991
Copyright (C) 1989, 1991 Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
Preamble
The licenses for most software are designed to take away your
freedom to share and change it. By contrast, the GNU General Public
License is intended to guarantee your freedom to share and change free
software--to make sure the software is free for all its users. This
General Public License applies to most of the Free Software
Foundation's software and to any other program whose authors commit to
using it. (Some other Free Software Foundation software is covered by
the GNU Lesser General Public License instead.) You can apply it to
your programs, too.
When we speak of free software, we are referring to freedom, not
price. Our General Public Licenses are designed to make sure that you
have the freedom to distribute copies of free software (and charge for
this service if you wish), that you receive source code or can get it
if you want it, that you can change the software or use pieces of it
in new free programs; and that you know you can do these things.
To protect your rights, we need to make restrictions that forbid
anyone to deny you these rights or to ask you to surrender the rights.
These restrictions translate to certain responsibilities for you if you
distribute copies of the software, or if you modify it.
For example, if you distribute copies of such a program, whether
gratis or for a fee, you must give the recipients all the rights that
you have. You must make sure that they, too, receive or can get the
source code. And you must show them these terms so they know their
rights.
We protect your rights with two steps: (1) copyright the software, and
(2) offer you this license which gives you legal permission to copy,
distribute and/or modify the software.
Also, for each author's protection and ours, we want to make certain
that everyone understands that there is no warranty for this free
software. If the software is modified by someone else and passed on, we
want its recipients to know that what they have is not the original, so
that any problems introduced by others will not reflect on the original
authors' reputations.
Finally, any free program is threatened constantly by software
patents. We wish to avoid the danger that redistributors of a free
program will individually obtain patent licenses, in effect making the
program proprietary. To prevent this, we have made it clear that any
patent must be licensed for everyone's free use or not licensed at all.
The precise terms and conditions for copying, distribution and
modification follow.
GNU GENERAL PUBLIC LICENSE
TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
0. This License applies to any program or other work which contains
a notice placed by the copyright holder saying it may be distributed
under the terms of this General Public License. The "Program", below,
refers to any such program or work, and a "work based on the Program"
means either the Program or any derivative work under copyright law:
that is to say, a work containing the Program or a portion of it,
either verbatim or with modifications and/or translated into another
language. (Hereinafter, translation is included without limitation in
the term "modification".) Each licensee is addressed as "you".
Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope. The act of
running the Program is not restricted, and the output from the Program
is covered only if its contents constitute a work based on the
Program (independent of having been made by running the Program).
Whether that is true depends on what the Program does.
1. You may copy and distribute verbatim copies of the Program's
source code as you receive it, in any medium, provided that you
conspicuously and appropriately publish on each copy an appropriate
copyright notice and disclaimer of warranty; keep intact all the
notices that refer to this License and to the absence of any warranty;
and give any other recipients of the Program a copy of this License
along with the Program.
You may charge a fee for the physical act of transferring a copy, and
you may at your option offer warranty protection in exchange for a fee.
2. You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:
a) You must cause the modified files to carry prominent notices
stating that you changed the files and the date of any change.
b) You must cause any work that you distribute or publish, that in
whole or in part contains or is derived from the Program or any
part thereof, to be licensed as a whole at no charge to all third
parties under the terms of this License.
c) If the modified program normally reads commands interactively
when run, you must cause it, when started running for such
interactive use in the most ordinary way, to print or display an
announcement including an appropriate copyright notice and a
notice that there is no warranty (or else, saying that you provide
a warranty) and that users may redistribute the program under
these conditions, and telling the user how to view a copy of this
License. (Exception: if the Program itself is interactive but
does not normally print such an announcement, your work based on
the Program is not required to print an announcement.)
These requirements apply to the modified work as a whole. If
identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works. But when you
distribute the same sections as part of a whole which is a work based
on the Program, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote it.
Thus, it is not the intent of this section to claim rights or contest
your rights to work written entirely by you; rather, the intent is to
exercise the right to control the distribution of derivative or
collective works based on the Program.
In addition, mere aggregation of another work not based on the Program
with the Program (or with a work based on the Program) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.
3. You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:
a) Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections
1 and 2 above on a medium customarily used for software interchange; or,
b) Accompany it with a written offer, valid for at least three
years, to give any third party, for a charge no more than your
cost of physically performing source distribution, a complete
machine-readable copy of the corresponding source code, to be
distributed under the terms of Sections 1 and 2 above on a medium
customarily used for software interchange; or,
c) Accompany it with the information you received as to the offer
to distribute corresponding source code. (This alternative is
allowed only for noncommercial distribution and only if you
received the program in object code or executable form with such
an offer, in accord with Subsection b above.)
The source code for a work means the preferred form of the work for
making modifications to it. For an executable work, complete source
code means all the source code for all modules it contains, plus any
associated interface definition files, plus the scripts used to
control compilation and installation of the executable. However, as a
special exception, the source code distributed need not include
anything that is normally distributed (in either source or binary
form) with the major components (compiler, kernel, and so on) of the
operating system on which the executable runs, unless that component
itself accompanies the executable.
If distribution of executable or object code is made by offering
access to copy from a designated place, then offering equivalent
access to copy the source code from the same place counts as
distribution of the source code, even though third parties are not
compelled to copy the source along with the object code.
4. You may not copy, modify, sublicense, or distribute the Program
except as expressly provided under this License. Any attempt
otherwise to copy, modify, sublicense or distribute the Program is
void, and will automatically terminate your rights under this License.
However, parties who have received copies, or rights, from you under
this License will not have their licenses terminated so long as such
parties remain in full compliance.
5. You are not required to accept this License, since you have not
signed it. However, nothing else grants you permission to modify or
distribute the Program or its derivative works. These actions are
prohibited by law if you do not accept this License. Therefore, by
modifying or distributing the Program (or any work based on the
Program), you indicate your acceptance of this License to do so, and
all its terms and conditions for copying, distributing or modifying
the Program or works based on it.
6. Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions. You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties to
this License.
7. If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues),
conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License. If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Program at all. For example, if a patent
license would not permit royalty-free redistribution of the Program by
all those who receive copies directly or indirectly through you, then
the only way you could satisfy both it and this License would be to
refrain entirely from distribution of the Program.
If any portion of this section is held invalid or unenforceable under
any particular circumstance, the balance of the section is intended to
apply and the section as a whole is intended to apply in other
circumstances.
It is not the purpose of this section to induce you to infringe any
patents or other property right claims or to contest validity of any
such claims; this section has the sole purpose of protecting the
integrity of the free software distribution system, which is
implemented by public license practices. Many people have made
generous contributions to the wide range of software distributed
through that system in reliance on consistent application of that
system; it is up to the author/donor to decide if he or she is willing
to distribute software through any other system and a licensee cannot
impose that choice.
This section is intended to make thoroughly clear what is believed to
be a consequence of the rest of this License.
8. If the distribution and/or use of the Program is restricted in
certain countries either by patents or by copyrighted interfaces, the
original copyright holder who places the Program under this License
may add an explicit geographical distribution limitation excluding
those countries, so that distribution is permitted only in or among
countries not thus excluded. In such case, this License incorporates
the limitation as if written in the body of this License.
9. The Free Software Foundation may publish revised and/or new versions
of the General Public License from time to time. Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.
Each version is given a distinguishing version number. If the Program
specifies a version number of this License which applies to it and "any
later version", you have the option of following the terms and conditions
either of that version or of any later version published by the Free
Software Foundation. If the Program does not specify a version number of
this License, you may choose any version ever published by the Free Software
Foundation.
10. If you wish to incorporate parts of the Program into other free
programs whose distribution conditions are different, write to the author
to ask for permission. For software which is copyrighted by the Free
Software Foundation, write to the Free Software Foundation; we sometimes
make exceptions for this. Our decision will be guided by the two goals
of preserving the free status of all derivatives of our free software and
of promoting the sharing and reuse of software generally.
NO WARRANTY
11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN
OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS
TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE
PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
REPAIR OR CORRECTION.
12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
POSSIBILITY OF SUCH DAMAGES.
END OF TERMS AND CONDITIONS
How to Apply These Terms to Your New Programs
If you develop a new program, and you want it to be of the greatest
possible use to the public, the best way to achieve this is to make it
free software which everyone can redistribute and change under these terms.
To do so, attach the following notices to the program. It is safest
to attach them to the start of each source file to most effectively
convey the exclusion of warranty; and each file should have at least
the "copyright" line and a pointer to where the full notice is found.
<one line to give the program's name and a brief idea of what it does.>
Copyright (C) <year> <name of author>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
Also add information on how to contact you by electronic and paper mail.
If the program is interactive, make it output a short notice like this
when it starts in an interactive mode:
Gnomovision version 69, Copyright (C) year name of author
Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
This is free software, and you are welcome to redistribute it
under certain conditions; type `show c' for details.
The hypothetical commands `show w' and `show c' should show the appropriate
parts of the General Public License. Of course, the commands you use may
be called something other than `show w' and `show c'; they could even be
mouse-clicks or menu items--whatever suits your program.
You should also get your employer (if you work as a programmer) or your
school, if any, to sign a "copyright disclaimer" for the program, if
necessary. Here is a sample; alter the names:
Yoyodyne, Inc., hereby disclaims all copyright interest in the program
`Gnomovision' (which makes passes at compilers) written by James Hacker.
<signature of Ty Coon>, 1 April 1989
Ty Coon, President of Vice
This General Public License does not permit incorporating your program into
proprietary programs. If your program is a subroutine library, you may
consider it more useful to permit linking proprietary applications with the
library. If this is what you want to do, use the GNU Lesser General
Public License instead of this License.
venv/lib/python3.11/site-packages/SpeechRecognition-3.10.0.dist-info/LICENSE.txt
+12
View File
@@ -0,0 +1,12 @@
Copyright (c) 2014-2017, Anthony Zhang <azhang9@gmail.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
3. Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
venv/lib/python3.11/site-packages/SpeechRecognition-3.10.0.dist-info/METADATA
+416
View File
@@ -0,0 +1,416 @@
Metadata-Version: 2.1
Name: SpeechRecognition
Version: 3.10.0
Summary: Library for performing speech recognition, with support for several engines and APIs, online and offline.
Home-page: https://github.com/Uberi/speech_recognition#readme
Author: Anthony Zhang (Uberi)
Author-email: azhang9@gmail.com
License: BSD
Keywords: speech recognition voice sphinx google wit bing api houndify ibm snowboy
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: Natural Language :: English
Classifier: License :: OSI Approved :: BSD License
Classifier: Operating System :: Microsoft :: Windows
Classifier: Operating System :: POSIX :: Linux
Classifier: Operating System :: MacOS :: MacOS X
Classifier: Operating System :: Other OS
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Topic :: Multimedia :: Sound/Audio :: Speech
Requires-Python: >=3.8
Description-Content-Type: text/x-rst
License-File: LICENSE-FLAC.txt
License-File: LICENSE.txt
Requires-Dist: requests (>=2.26.0)
Provides-Extra: whisper-api
Requires-Dist: openai ; extra == 'whisper-api'
SpeechRecognition
=================
.. image:: https://img.shields.io/pypi/v/SpeechRecognition.svg
:target: https://pypi.python.org/pypi/SpeechRecognition/
:alt: Latest Version
.. image:: https://img.shields.io/pypi/status/SpeechRecognition.svg
:target: https://pypi.python.org/pypi/SpeechRecognition/
:alt: Development Status
.. image:: https://img.shields.io/pypi/pyversions/SpeechRecognition.svg
:target: https://pypi.python.org/pypi/SpeechRecognition/
:alt: Supported Python Versions
.. image:: https://img.shields.io/pypi/l/SpeechRecognition.svg
:target: https://pypi.python.org/pypi/SpeechRecognition/
:alt: License
.. image:: https://api.travis-ci.org/Uberi/speech_recognition.svg?branch=master
:target: https://travis-ci.org/Uberi/speech_recognition
:alt: Continuous Integration Test Results
Library for performing speech recognition, with support for several engines and APIs, online and offline.
**UPDATE 2022-02-09**: Hey everyone! This project started as a tech demo, but these days it needs more time than I have to keep up with all the PRs and issues. Therefore, I'd like to put out an **open invite for collaborators** - just reach out at me@anthonyz.ca if you're interested!
Speech recognition engine/API support:
* `CMU Sphinx <http://cmusphinx.sourceforge.net/wiki/>`__ (works offline)
* Google Speech Recognition
* `Google Cloud Speech API <https://cloud.google.com/speech/>`__
* `Wit.ai <https://wit.ai/>`__
* `Microsoft Azure Speech <https://azure.microsoft.com/en-us/services/cognitive-services/speech/>`__
* `Microsoft Bing Voice Recognition (Deprecated) <https://www.microsoft.com/cognitive-services/en-us/speech-api>`__
* `Houndify API <https://houndify.com/>`__
* `IBM Speech to Text <http://www.ibm.com/smarterplanet/us/en/ibmwatson/developercloud/speech-to-text.html>`__
* `Snowboy Hotword Detection <https://snowboy.kitt.ai/>`__ (works offline)
* `Tensorflow <https://www.tensorflow.org/>`__
* `Vosk API <https://github.com/alphacep/vosk-api/>`__ (works offline)
* `OpenAI whisper <https://github.com/openai/whisper>`__ (works offline)
* `Whisper API <https://platform.openai.com/docs/guides/speech-to-text>`__
**Quickstart:** ``pip install SpeechRecognition``. See the "Installing" section for more details.
To quickly try it out, run ``python -m speech_recognition`` after installing.
Project links:
- `PyPI <https://pypi.python.org/pypi/SpeechRecognition/>`__
- `Source code <https://github.com/Uberi/speech_recognition>`__
- `Issue tracker <https://github.com/Uberi/speech_recognition/issues>`__
Library Reference
-----------------
The `library reference <https://github.com/Uberi/speech_recognition/blob/master/reference/library-reference.rst>`__ documents every publicly accessible object in the library. This document is also included under ``reference/library-reference.rst``.
See `Notes on using PocketSphinx <https://github.com/Uberi/speech_recognition/blob/master/reference/pocketsphinx.rst>`__ for information about installing languages, compiling PocketSphinx, and building language packs from online resources. This document is also included under ``reference/pocketsphinx.rst``.
You have to install Vosk models for using Vosk. `Here <https://alphacephei.com/vosk/models>`__ are models avaiable. You have to place them in models folder of your project, like "your-project-folder/models/your-vosk-model"
Examples
--------
See the ``examples/`` `directory <https://github.com/Uberi/speech_recognition/tree/master/examples>`__ in the repository root for usage examples:
- `Recognize speech input from the microphone <https://github.com/Uberi/speech_recognition/blob/master/examples/microphone_recognition.py>`__
- `Transcribe an audio file <https://github.com/Uberi/speech_recognition/blob/master/examples/audio_transcribe.py>`__
- `Save audio data to an audio file <https://github.com/Uberi/speech_recognition/blob/master/examples/write_audio.py>`__
- `Show extended recognition results <https://github.com/Uberi/speech_recognition/blob/master/examples/extended_results.py>`__
- `Calibrate the recognizer energy threshold for ambient noise levels <https://github.com/Uberi/speech_recognition/blob/master/examples/calibrate_energy_threshold.py>`__ (see ``recognizer_instance.energy_threshold`` for details)
- `Listening to a microphone in the background <https://github.com/Uberi/speech_recognition/blob/master/examples/background_listening.py>`__
- `Various other useful recognizer features <https://github.com/Uberi/speech_recognition/blob/master/examples/special_recognizer_features.py>`__
Installing
----------
First, make sure you have all the requirements listed in the "Requirements" section.
The easiest way to install this is using ``pip install SpeechRecognition``.
Otherwise, download the source distribution from `PyPI <https://pypi.python.org/pypi/SpeechRecognition/>`__, and extract the archive.
In the folder, run ``python setup.py install``.
Requirements
------------
To use all of the functionality of the library, you should have:
* **Python** 3.8+ (required)
* **PyAudio** 0.2.11+ (required only if you need to use microphone input, ``Microphone``)
* **PocketSphinx** (required only if you need to use the Sphinx recognizer, ``recognizer_instance.recognize_sphinx``)
* **Google API Client Library for Python** (required only if you need to use the Google Cloud Speech API, ``recognizer_instance.recognize_google_cloud``)
* **FLAC encoder** (required only if the system is not x86-based Windows/Linux/OS X)
* **Vosk** (required only if you need to use Vosk API speech recognition ``recognizer_instance.recognize_vosk``)
* **Whisper** (required only if you need to use Whisper ``recognizer_instance.recognize_whisper``)
* **openai** (required only if you need to use Whisper API speech recognition ``recognizer_instance.recognize_whisper_api``)
The following requirements are optional, but can improve or extend functionality in some situations:
* If using CMU Sphinx, you may want to `install additional language packs <https://github.com/Uberi/speech_recognition/blob/master/reference/pocketsphinx.rst#installing-other-languages>`__ to support languages like International French or Mandarin Chinese.
The following sections go over the details of each requirement.
Python
~~~~~~
The first software requirement is `Python 3.8+ <https://www.python.org/downloads/>`__. This is required to use the library.
PyAudio (for microphone users)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
`PyAudio <http://people.csail.mit.edu/hubert/pyaudio/#downloads>`__ is required if and only if you want to use microphone input (``Microphone``). PyAudio version 0.2.11+ is required, as earlier versions have known memory management bugs when recording from microphones in certain situations.
If not installed, everything in the library will still work, except attempting to instantiate a ``Microphone`` object will raise an ``AttributeError``.
The installation instructions on the PyAudio website are quite good - for convenience, they are summarized below:
* On Windows, install PyAudio using `Pip <https://pip.readthedocs.org/>`__: execute ``pip install pyaudio`` in a terminal.
* On Debian-derived Linux distributions (like Ubuntu and Mint), install PyAudio using `APT <https://wiki.debian.org/Apt>`__: execute ``sudo apt-get install python-pyaudio python3-pyaudio`` in a terminal.
* If the version in the repositories is too old, install the latest release using Pip: execute ``sudo apt-get install portaudio19-dev python-all-dev python3-all-dev && sudo pip install pyaudio`` (replace ``pip`` with ``pip3`` if using Python 3).
* On OS X, install PortAudio using `Homebrew <http://brew.sh/>`__: ``brew install portaudio``. Then, install PyAudio using `Pip <https://pip.readthedocs.org/>`__: ``pip install pyaudio``.
* On other POSIX-based systems, install the ``portaudio19-dev`` and ``python-all-dev`` (or ``python3-all-dev`` if using Python 3) packages (or their closest equivalents) using a package manager of your choice, and then install PyAudio using `Pip <https://pip.readthedocs.org/>`__: ``pip install pyaudio`` (replace ``pip`` with ``pip3`` if using Python 3).
PyAudio `wheel packages <https://pypi.python.org/pypi/wheel>`__ for common 64-bit Python versions on Windows and Linux are included for convenience, under the ``third-party/`` `directory <https://github.com/Uberi/speech_recognition/tree/master/third-party>`__ in the repository root. To install, simply run ``pip install wheel`` followed by ``pip install ./third-party/WHEEL_FILENAME`` (replace ``pip`` with ``pip3`` if using Python 3) in the repository `root directory <https://github.com/Uberi/speech_recognition>`__.
PocketSphinx-Python (for Sphinx users)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
`PocketSphinx-Python <https://github.com/bambocher/pocketsphinx-python>`__ is **required if and only if you want to use the Sphinx recognizer** (``recognizer_instance.recognize_sphinx``).
PocketSphinx-Python `wheel packages <https://pypi.python.org/pypi/wheel>`__ for 64-bit Python 3.4, and 3.5 on Windows are included for convenience, under the ``third-party/`` `directory <https://github.com/Uberi/speech_recognition/tree/master/third-party>`__. To install, simply run ``pip install wheel`` followed by ``pip install ./third-party/WHEEL_FILENAME`` (replace ``pip`` with ``pip3`` if using Python 3) in the SpeechRecognition folder.
On Linux and other POSIX systems (such as OS X), follow the instructions under "Building PocketSphinx-Python from source" in `Notes on using PocketSphinx <https://github.com/Uberi/speech_recognition/blob/master/reference/pocketsphinx.rst>`__ for installation instructions.
Note that the versions available in most package repositories are outdated and will not work with the bundled language data. Using the bundled wheel packages or building from source is recommended.
See `Notes on using PocketSphinx <https://github.com/Uberi/speech_recognition/blob/master/reference/pocketsphinx.rst>`__ for information about installing languages, compiling PocketSphinx, and building language packs from online resources. This document is also included under ``reference/pocketsphinx.rst``.
Vosk (for Vosk users)
~~~~~~~~~~~~~~~~~~~~~
Vosk API is **required if and only if you want to use Vosk recognizer** (``recognizer_instance.recognize_vosk``).
You can install it with ``python3 -m pip install vosk``.
You also have to install Vosk Models:
`Here <https://alphacephei.com/vosk/models>`__ are models avaiable for download. You have to place them in models folder of your project, like "your-project-folder/models/your-vosk-model"
Google Cloud Speech Library for Python (for Google Cloud Speech API users)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
`Google Cloud Speech library for Python <https://cloud.google.com/speech-to-text/docs/quickstart>`__ is required if and only if you want to use the Google Cloud Speech API (``recognizer_instance.recognize_google_cloud``).
If not installed, everything in the library will still work, except calling ``recognizer_instance.recognize_google_cloud`` will raise an ``RequestError``.
According to the `official installation instructions <https://cloud.google.com/speech-to-text/docs/quickstart>`__, the recommended way to install this is using `Pip <https://pip.readthedocs.org/>`__: execute ``pip install google-cloud-speech`` (replace ``pip`` with ``pip3`` if using Python 3).
FLAC (for some systems)
~~~~~~~~~~~~~~~~~~~~~~~
A `FLAC encoder <https://xiph.org/flac/>`__ is required to encode the audio data to send to the API. If using Windows (x86 or x86-64), OS X (Intel Macs only, OS X 10.6 or higher), or Linux (x86 or x86-64), this is **already bundled with this library - you do not need to install anything**.
Otherwise, ensure that you have the ``flac`` command line tool, which is often available through the system package manager. For example, this would usually be ``sudo apt-get install flac`` on Debian-derivatives, or ``brew install flac`` on OS X with Homebrew.
Whisper (for Whisper users)
~~~~~~~~~~~~~~~~~~~~~~~~~~~
Whisper is **required if and only if you want to use whisper** (``recognizer_instance.recognize_whisper``).
You can install it with ``python3 -m pip install git+https://github.com/openai/whisper.git soundfile``.
Whisper API (for Whisper API users)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The library `openai <https://pypi.org/project/openai/>`__ is **required if and only if you want to use Whisper API** (``recognizer_instance.recognize_whisper_api``).
If not installed, everything in the library will still work, except calling ``recognizer_instance.recognize_whisper_api`` will raise an ``RequestError``.
You can install it with ``python3 -m pip install openai``.
Troubleshooting
---------------
The recognizer tries to recognize speech even when I'm not speaking, or after I'm done speaking.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Try increasing the ``recognizer_instance.energy_threshold`` property. This is basically how sensitive the recognizer is to when recognition should start. Higher values mean that it will be less sensitive, which is useful if you are in a loud room.
This value depends entirely on your microphone or audio data. There is no one-size-fits-all value, but good values typically range from 50 to 4000.
Also, check on your microphone volume settings. If it is too sensitive, the microphone may be picking up a lot of ambient noise. If it is too insensitive, the microphone may be rejecting speech as just noise.
The recognizer can't recognize speech right after it starts listening for the first time.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The ``recognizer_instance.energy_threshold`` property is probably set to a value that is too high to start off with, and then being adjusted lower automatically by dynamic energy threshold adjustment. Before it is at a good level, the energy threshold is so high that speech is just considered ambient noise.
The solution is to decrease this threshold, or call ``recognizer_instance.adjust_for_ambient_noise`` beforehand, which will set the threshold to a good value automatically.
The recognizer doesn't understand my particular language/dialect.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Try setting the recognition language to your language/dialect. To do this, see the documentation for ``recognizer_instance.recognize_sphinx``, ``recognizer_instance.recognize_google``, ``recognizer_instance.recognize_wit``, ``recognizer_instance.recognize_bing``, ``recognizer_instance.recognize_api``, ``recognizer_instance.recognize_houndify``, and ``recognizer_instance.recognize_ibm``.
For example, if your language/dialect is British English, it is better to use ``"en-GB"`` as the language rather than ``"en-US"``.
The recognizer hangs on ``recognizer_instance.listen``; specifically, when it's calling ``Microphone.MicrophoneStream.read``.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
This usually happens when you're using a Raspberry Pi board, which doesn't have audio input capabilities by itself. This causes the default microphone used by PyAudio to simply block when we try to read it. If you happen to be using a Raspberry Pi, you'll need a USB sound card (or USB microphone).
Once you do this, change all instances of ``Microphone()`` to ``Microphone(device_index=MICROPHONE_INDEX)``, where ``MICROPHONE_INDEX`` is the hardware-specific index of the microphone.
To figure out what the value of ``MICROPHONE_INDEX`` should be, run the following code:
.. code:: python
import speech_recognition as sr
for index, name in enumerate(sr.Microphone.list_microphone_names()):
print("Microphone with name \"{1}\" found for `Microphone(device_index={0})`".format(index, name))
This will print out something like the following:
::
Microphone with name "HDA Intel HDMI: 0 (hw:0,3)" found for `Microphone(device_index=0)`
Microphone with name "HDA Intel HDMI: 1 (hw:0,7)" found for `Microphone(device_index=1)`
Microphone with name "HDA Intel HDMI: 2 (hw:0,8)" found for `Microphone(device_index=2)`
Microphone with name "Blue Snowball: USB Audio (hw:1,0)" found for `Microphone(device_index=3)`
Microphone with name "hdmi" found for `Microphone(device_index=4)`
Microphone with name "pulse" found for `Microphone(device_index=5)`
Microphone with name "default" found for `Microphone(device_index=6)`
Now, to use the Snowball microphone, you would change ``Microphone()`` to ``Microphone(device_index=3)``.
Calling ``Microphone()`` gives the error ``IOError: No Default Input Device Available``.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
As the error says, the program doesn't know which microphone to use.
To proceed, either use ``Microphone(device_index=MICROPHONE_INDEX, ...)`` instead of ``Microphone(...)``, or set a default microphone in your OS. You can obtain possible values of ``MICROPHONE_INDEX`` using the code in the troubleshooting entry right above this one.
The program doesn't run when compiled with `PyInstaller <https://github.com/pyinstaller/pyinstaller/wiki>`__.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
As of PyInstaller version 3.0, SpeechRecognition is supported out of the box. If you're getting weird issues when compiling your program using PyInstaller, simply update PyInstaller.
You can easily do this by running ``pip install --upgrade pyinstaller``.
On Ubuntu/Debian, I get annoying output in the terminal saying things like "bt_audio_service_open: [...] Connection refused" and various others.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The "bt_audio_service_open" error means that you have a Bluetooth audio device, but as a physical device is not currently connected, we can't actually use it - if you're not using a Bluetooth microphone, then this can be safely ignored. If you are, and audio isn't working, then double check to make sure your microphone is actually connected. There does not seem to be a simple way to disable these messages.
For errors of the form "ALSA lib [...] Unknown PCM", see `this StackOverflow answer <http://stackoverflow.com/questions/7088672/pyaudio-working-but-spits-out-error-messages-each-time>`__. Basically, to get rid of an error of the form "Unknown PCM cards.pcm.rear", simply comment out ``pcm.rear cards.pcm.rear`` in ``/usr/share/alsa/alsa.conf``, ``~/.asoundrc``, and ``/etc/asound.conf``.
For "jack server is not running or cannot be started" or "connect(2) call to /dev/shm/jack-1000/default/jack_0 failed (err=No such file or directory)" or "attempt to connect to server failed", these are caused by ALSA trying to connect to JACK, and can be safely ignored. I'm not aware of any simple way to turn those messages off at this time, besides `entirely disabling printing while starting the microphone <https://github.com/Uberi/speech_recognition/issues/182#issuecomment-266256337>`__.
On OS X, I get a ``ChildProcessError`` saying that it couldn't find the system FLAC converter, even though it's installed.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Installing `FLAC for OS X <https://xiph.org/flac/download.html>`__ directly from the source code will not work, since it doesn't correctly add the executables to the search path.
Installing FLAC using `Homebrew <http://brew.sh/>`__ ensures that the search path is correctly updated. First, ensure you have Homebrew, then run ``brew install flac`` to install the necessary files.
Developing
----------
To hack on this library, first make sure you have all the requirements listed in the "Requirements" section.
- Most of the library code lives in ``speech_recognition/__init__.py``.
- Examples live under the ``examples/`` `directory <https://github.com/Uberi/speech_recognition/tree/master/examples>`__, and the demo script lives in ``speech_recognition/__main__.py``.
- The FLAC encoder binaries are in the ``speech_recognition/`` `directory <https://github.com/Uberi/speech_recognition/tree/master/speech_recognition>`__.
- Documentation can be found in the ``reference/`` `directory <https://github.com/Uberi/speech_recognition/tree/master/reference>`__.
- Third-party libraries, utilities, and reference material are in the ``third-party/`` `directory <https://github.com/Uberi/speech_recognition/tree/master/third-party>`__.
To install/reinstall the library locally, run ``python setup.py install`` in the project `root directory <https://github.com/Uberi/speech_recognition>`__.
Before a release, the version number is bumped in ``README.rst`` and ``speech_recognition/__init__.py``. Version tags are then created using ``git config gpg.program gpg2 && git config user.signingkey DB45F6C431DE7C2DCD99FF7904882258A4063489 && git tag -s VERSION_GOES_HERE -m "Version VERSION_GOES_HERE"``.
Releases are done by running ``make-release.sh VERSION_GOES_HERE`` to build the Python source packages, sign them, and upload them to PyPI.
Testing
~~~~~~~
To run all the tests:
.. code:: bash
python -m unittest discover --verbose
Testing is also done automatically by TravisCI, upon every push. To set up the environment for offline/local Travis-like testing on a Debian-like system:
.. code:: bash
sudo docker run --volume "$(pwd):/speech_recognition" --interactive --tty quay.io/travisci/travis-python:latest /bin/bash
su - travis && cd /speech_recognition
sudo apt-get update && sudo apt-get install swig libpulse-dev
pip install --user pocketsphinx && pip install --user flake8 rstcheck && pip install --user -e .
python -m unittest discover --verbose # run unit tests
python -m flake8 --ignore=E501,E701 speech_recognition tests examples setup.py # ignore errors for long lines and multi-statement lines
python -m rstcheck README.rst reference/*.rst # ensure RST is well-formed
FLAC Executables
~~~~~~~~~~~~~~~~
The included ``flac-win32`` executable is the `official FLAC 1.3.2 32-bit Windows binary <http://downloads.xiph.org/releases/flac/flac-1.3.2-win.zip>`__.
The included ``flac-linux-x86`` and ``flac-linux-x86_64`` executables are built from the `FLAC 1.3.2 source code <http://downloads.xiph.org/releases/flac/flac-1.3.2.tar.xz>`__ with `Manylinux <https://github.com/pypa/manylinux>`__ to ensure that it's compatible with a wide variety of distributions.
The built FLAC executables should be bit-for-bit reproducible. To rebuild them, run the following inside the project directory on a Debian-like system:
.. code:: bash
# download and extract the FLAC source code
cd third-party
sudo apt-get install --yes docker.io
# build FLAC inside the Manylinux i686 Docker image
tar xf flac-1.3.2.tar.xz
sudo docker run --tty --interactive --rm --volume "$(pwd):/root" quay.io/pypa/manylinux1_i686:latest bash
cd /root/flac-1.3.2
./configure LDFLAGS=-static # compiler flags to make a static build
make
exit
cp flac-1.3.2/src/flac/flac ../speech_recognition/flac-linux-x86 && sudo rm -rf flac-1.3.2/
# build FLAC inside the Manylinux x86_64 Docker image
tar xf flac-1.3.2.tar.xz
sudo docker run --tty --interactive --rm --volume "$(pwd):/root" quay.io/pypa/manylinux1_x86_64:latest bash
cd /root/flac-1.3.2
./configure LDFLAGS=-static # compiler flags to make a static build
make
exit
cp flac-1.3.2/src/flac/flac ../speech_recognition/flac-linux-x86_64 && sudo rm -r flac-1.3.2/
The included ``flac-mac`` executable is extracted from `xACT 2.39 <http://xact.scottcbrown.org/>`__, which is a frontend for FLAC 1.3.2 that conveniently includes binaries for all of its encoders. Specifically, it is a copy of ``xACT 2.39/xACT.app/Contents/Resources/flac`` in ``xACT2.39.zip``.
Authors
-------
::
Uberi <me@anthonyz.ca> (Anthony Zhang)
bobsayshilol
arvindch <achembarpu@gmail.com> (Arvind Chembarpu)
kevinismith <kevin_i_smith@yahoo.com> (Kevin Smith)
haas85
DelightRun <changxu.mail@gmail.com>
maverickagm
kamushadenes <kamushadenes@hyadesinc.com> (Kamus Hadenes)
sbraden <braden.sarah@gmail.com> (Sarah Braden)
tb0hdan (Bohdan Turkynewych)
Thynix <steve@asksteved.com> (Steve Dougherty)
beeedy <broderick.carlin@gmail.com> (Broderick Carlin)
Please report bugs and suggestions at the `issue tracker <https://github.com/Uberi/speech_recognition/issues>`__!
How to cite this library (APA style):
Zhang, A. (2017). Speech Recognition (Version 3.8) [Software]. Available from https://github.com/Uberi/speech_recognition#readme.
How to cite this library (Chicago style):
Zhang, Anthony. 2017. *Speech Recognition* (version 3.8).
Also check out the `Python Baidu Yuyin API <https://github.com/DelightRun/PyBaiduYuyin>`__, which is based on an older version of this project, and adds support for `Baidu Yuyin <http://yuyin.baidu.com/>`__. Note that Baidu Yuyin is only available inside China.
License
-------
Copyright 2014-2017 `Anthony Zhang (Uberi) <http://anthonyz.ca/>`__. The source code for this library is available online at `GitHub <https://github.com/Uberi/speech_recognition>`__.
SpeechRecognition is made available under the 3-clause BSD license. See ``LICENSE.txt`` in the project's `root directory <https://github.com/Uberi/speech_recognition>`__ for more information.
For convenience, all the official distributions of SpeechRecognition already include a copy of the necessary copyright notices and licenses. In your project, you can simply **say that licensing information for SpeechRecognition can be found within the SpeechRecognition README, and make sure SpeechRecognition is visible to users if they wish to see it**.
SpeechRecognition distributes source code, binaries, and language files from `CMU Sphinx <http://cmusphinx.sourceforge.net/>`__. These files are BSD-licensed and redistributable as long as copyright notices are correctly retained. See ``speech_recognition/pocketsphinx-data/*/LICENSE*.txt`` and ``third-party/LICENSE-Sphinx.txt`` for license details for individual parts.
SpeechRecognition distributes source code and binaries from `PyAudio <http://people.csail.mit.edu/hubert/pyaudio/>`__. These files are MIT-licensed and redistributable as long as copyright notices are correctly retained. See ``third-party/LICENSE-PyAudio.txt`` for license details.
SpeechRecognition distributes binaries from `FLAC <https://xiph.org/flac/>`__ - ``speech_recognition/flac-win32.exe``, ``speech_recognition/flac-linux-x86``, and ``speech_recognition/flac-mac``. These files are GPLv2-licensed and redistributable, as long as the terms of the GPL are satisfied. The FLAC binaries are an `aggregate <https://www.gnu.org/licenses/gpl-faq.html#MereAggregation>`__ of `separate programs <https://www.gnu.org/licenses/gpl-faq.html#NFUseGPLPlugins>`__, so these GPL restrictions do not apply to the library or your programs that use the library, only to FLAC itself. See ``LICENSE-FLAC.txt`` for license details.
venv/lib/python3.11/site-packages/SpeechRecognition-3.10.0.dist-info/RECORD
+35
View File
@@ -0,0 +1,35 @@
SpeechRecognition-3.10.0.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
SpeechRecognition-3.10.0.dist-info/LICENSE-FLAC.txt,sha256=gXf5dRMhNSbfLPYYTY_5hsZ1r7UU1OaKQEAQUhuIBkM,18092
SpeechRecognition-3.10.0.dist-info/LICENSE.txt,sha256=XUbseO51bVM11Ho-UDWeVMpWwK2qdqLWbpIZX96X10U,1519
SpeechRecognition-3.10.0.dist-info/METADATA,sha256=kWq32D5V38A8oTNuZGk5BSJO7QOTfcw6XghWktjhcRQ,28857
SpeechRecognition-3.10.0.dist-info/RECORD,,
SpeechRecognition-3.10.0.dist-info/REQUESTED,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
SpeechRecognition-3.10.0.dist-info/WHEEL,sha256=bb2Ot9scclHKMOLDEHY6B2sicWOgugjFKaJsT7vwMQo,110
SpeechRecognition-3.10.0.dist-info/top_level.txt,sha256=iJfbFMOCcDL7Mdhysq4LsbniwTv6EtQR8BbgEnKlPIw,19
speech_recognition/__init__.py,sha256=B8T8R4TIm_hwycpZyd3jB-SS6QHvvlQFRG0umnZBh9o,94078
speech_recognition/__main__.py,sha256=Afbp3l1yq2AW2Bsm7YBz-O0wm6gHNudgPZh69Ijjubc,833
speech_recognition/__pycache__/__init__.cpython-311.pyc,,
speech_recognition/__pycache__/__main__.cpython-311.pyc,,
speech_recognition/__pycache__/audio.cpython-311.pyc,,
speech_recognition/__pycache__/exceptions.cpython-311.pyc,,
speech_recognition/audio.py,sha256=2IHCx9EoQRAJSNFRZsnui3Ags3gBWtAmKmG8wIB3MDY,14807
speech_recognition/exceptions.py,sha256=LnKmutOVQ6RSMcSO1Ji5Af-UZN-Aqrq4petMgBFXkKs,273
speech_recognition/flac-linux-x86,sha256=FOUk-MAmqO11z7GmT4TyHIJnRZaDXZaBVsh5yfP1k8g,1899154
speech_recognition/flac-linux-x86_64,sha256=0k6-i9XM2vxk9SKmKuL-0_t0ARrf7ts9Peqmu0YR34Q,2396644
speech_recognition/flac-mac,sha256=2LyQYpHz0QJGWm_dnNfJoIOYsDsPI8b-JwZAVns4o6E,451168
speech_recognition/flac-win32.exe,sha256=D8yWtMDceZCrDDnb-8gvrtNmxYiPD0oz9GTqO3m38IU,738816
speech_recognition/pocketsphinx-data/en-US/LICENSE.txt,sha256=_PXE5BroH3BO1w8VERRicu8PDKe5cIjRHXKJa1ahifU,1537
speech_recognition/pocketsphinx-data/en-US/acoustic-model/README,sha256=i4jemAVoUJxkbQUnuEFL7vE2lkORkDtAmW0y9ze_dS4,1617
speech_recognition/pocketsphinx-data/en-US/acoustic-model/feat.params,sha256=ioWtKGlsfyNDRE52M84w8WX-2HQPghlUVExvvUb0lR8,165
speech_recognition/pocketsphinx-data/en-US/acoustic-model/mdef,sha256=I2D5qGiJwc_ui9YYoCaTh5EeX7KSCllPUGsYuMeWg7A,2959176
speech_recognition/pocketsphinx-data/en-US/acoustic-model/means,sha256=gyAZ4yysEusxiWT5b0aQNKyxLQNI7t3DgxgxoQDLTdQ,838732
speech_recognition/pocketsphinx-data/en-US/acoustic-model/noisedict,sha256=cpWwffLCBMT4fGeCtr4aOFnXAG1OOGQYHJVdbasQWjM,56
speech_recognition/pocketsphinx-data/en-US/acoustic-model/sendump,sha256=jJVkwNW-9pyp2b8QFKvhYvBxZEzwLPH6ikg8PcFlp6g,1969024
speech_recognition/pocketsphinx-data/en-US/acoustic-model/transition_matrices,sha256=wffyjqQxd75zS-H4i9fxuahT0OZg-FmcZ8bq7Ki7U5o,2080
speech_recognition/pocketsphinx-data/en-US/acoustic-model/variances,sha256=sA1pb4XpaDT8EPjl8GQo2MTba__b5YRbb2m_bvvEj6U,838732
speech_recognition/pocketsphinx-data/en-US/language-model.lm.bin,sha256=Ttj1LtBBMEXw6cZaRwmVOJa2_5qZ68EPfvkm9mLF5QY,29208442
speech_recognition/pocketsphinx-data/en-US/pronounciation-dictionary.dict,sha256=zjv6Nk45C35Ip0g9NQpI2fFhEn9oQoucb5ZygR_BE08,3240807
speech_recognition/recognizers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
speech_recognition/recognizers/__pycache__/__init__.cpython-311.pyc,,
speech_recognition/recognizers/__pycache__/whisper.cpython-311.pyc,,
speech_recognition/recognizers/whisper.py,sha256=ZSwBPcet4Lnyu6Qwa42hJZuNRhgSBN37R9tv6tnH3EA,1492
venv/lib/python3.11/site-packages/SpeechRecognition-3.10.0.dist-info/REQUESTED
View File
venv/lib/python3.11/site-packages/SpeechRecognition-3.10.0.dist-info/WHEEL
+6
View File
@@ -0,0 +1,6 @@
Wheel-Version: 1.0
Generator: bdist_wheel (0.38.4)
Root-Is-Purelib: true
Tag: py2-none-any
Tag: py3-none-any
venv/lib/python3.11/site-packages/SpeechRecognition-3.10.0.dist-info/top_level.txt
+1
View File
@@ -0,0 +1 @@
speech_recognition
venv/lib/python3.11/site-packages/pydub-0.25.1.dist-info/AUTHORS
+98
View File
@@ -0,0 +1,98 @@
James Robert
github: jiaaro
twitter: @jiaaro
web: jiaaro.com
email: pydub@jiaaro.com
Marc Webbie
github: marcwebbie
Jean-philippe Serafin
github: jeanphix
Anurag Ramdasan
github: AnuragRamdasan
Choongmin Lee
github: clee704
Patrick Pittman
github: ptpittman
Hunter Lang
github: hunterlang
Alexey
github: nihisil
Jaymz Campbell
github: jaymzcd
Ross McFarland
github: ross
John McMellen
github: jmcmellen
Johan Lövgren
github: dashj
Joachim Krüger
github: jkrgr
Shichao An
github: shichao-an
Michael Bortnyck
github: mbortnyck
André Cloete
github: aj-cloete
David Acacio
github: dacacioa
Thiago Abdnur
github: bolaum
Aurélien Ooms
github: aureooms
Mike Mattozzi
github: mmattozzi
Marcio Mazza
github: marciomazza
Sungsu Lim
github: proflim
Evandro Myller
github: emyller
Sérgio Agostinho
github: SergioRAgostinho
Antonio Larrosa
github: antlarr
Aaron Craig
github: craigthelinguist
Carlos del Castillo
github: greyalien502
Yudong Sun
github: sunjerry019
Jorge Perianez
github: JPery
Chendi Luo
github: Creonalia
Daniel Lefevre
gitHub: dplefevre
Grzegorz Kotfis
github: gkotfis
venv/lib/python3.11/site-packages/pydub-0.25.1.dist-info/INSTALLER
+1
View File
@@ -0,0 +1 @@
pip
venv/lib/python3.11/site-packages/pydub-0.25.1.dist-info/LICENSE
+20
View File
@@ -0,0 +1,20 @@
Copyright (c) 2011 James Robert, http://jiaaro.com
Permission is hereby granted, free of charge, to any person obtaining
a copy of this software and associated documentation files (the
"Software"), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:
The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
venv/lib/python3.11/site-packages/pydub-0.25.1.dist-info/METADATA
+37
View File
@@ -0,0 +1,37 @@
Metadata-Version: 2.1
Name: pydub
Version: 0.25.1
Summary: Manipulate audio with an simple and easy high level interface
Home-page: http://pydub.com
Author: James Robert
Author-email: jiaaro@gmail.com
License: MIT
Keywords: audio sound high-level
Platform: UNKNOWN
Classifier: Development Status :: 5 - Production/Stable
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 2
Classifier: Programming Language :: Python :: 2.7
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.4
Classifier: Programming Language :: Python :: 3.5
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Intended Audience :: Developers
Classifier: Operating System :: OS Independent
Classifier: Topic :: Multimedia :: Sound/Audio
Classifier: Topic :: Multimedia :: Sound/Audio :: Analysis
Classifier: Topic :: Multimedia :: Sound/Audio :: Conversion
Classifier: Topic :: Multimedia :: Sound/Audio :: Editors
Classifier: Topic :: Multimedia :: Sound/Audio :: Mixers
Classifier: Topic :: Software Development :: Libraries
Classifier: Topic :: Utilities
Manipulate audio with an simple and easy high level interface.
See the README file for details, usage info, and a list of gotchas.
venv/lib/python3.11/site-packages/pydub-0.25.1.dist-info/RECORD
+30
View File
@@ -0,0 +1,30 @@
pydub-0.25.1.dist-info/AUTHORS,sha256=AyY2PS9I2enOyBnUnxcpeAX-NnMNWLQT4yDtg8IIy78,1250
pydub-0.25.1.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
pydub-0.25.1.dist-info/LICENSE,sha256=roVlNiJMx6OJ6Wh3H8XyWYFL3Q2mNTnPcigq2672iXo,1074
pydub-0.25.1.dist-info/METADATA,sha256=f0M8_ZVtbiYoUI9ejXIeJ03Jo9A5Nbi-0V1bVqs5iYk,1406
pydub-0.25.1.dist-info/RECORD,,
pydub-0.25.1.dist-info/REQUESTED,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
pydub-0.25.1.dist-info/WHEEL,sha256=Z-nyYpwrcSqxfdux5Mbn_DQ525iP7J2DG3JgGvOYyTQ,110
pydub-0.25.1.dist-info/top_level.txt,sha256=PHhiDCQVZdycZxfKL2lQozruBT6ZhvyZAwqjRrw3t0w,6
pydub/__init__.py,sha256=w1Xv1awbaR3fMhTNE1-grnfswgARTNQrKpBzfZ--VBA,39
pydub/__pycache__/__init__.cpython-311.pyc,,
pydub/__pycache__/audio_segment.cpython-311.pyc,,
pydub/__pycache__/effects.cpython-311.pyc,,
pydub/__pycache__/exceptions.cpython-311.pyc,,
pydub/__pycache__/generators.cpython-311.pyc,,
pydub/__pycache__/logging_utils.cpython-311.pyc,,
pydub/__pycache__/playback.cpython-311.pyc,,
pydub/__pycache__/pyaudioop.cpython-311.pyc,,
pydub/__pycache__/scipy_effects.cpython-311.pyc,,
pydub/__pycache__/silence.cpython-311.pyc,,
pydub/__pycache__/utils.cpython-311.pyc,,
pydub/audio_segment.py,sha256=Nf5VkHGY1v9Jqb7NtEYfwRpLrfqusfBdPGOZsi7R5Cg,49185
pydub/effects.py,sha256=1HUMzhefrwG_E1rTnzvbl-P0-KNuwHklCnu8QCGS7jA,11507
pydub/exceptions.py,sha256=osgXoUujwpH8K6hr80iYpW30CMBDFwqyaRD-5d7ZpKs,455
pydub/generators.py,sha256=u6q7J8JLOY-uEZqMPUTzakxyua3XNQcPiDsuiK2-lLA,4045
pydub/logging_utils.py,sha256=WuSqfzn4zyT7PxXHGV-PXMDynufeM6sC6eSmVlGX2RU,374
pydub/playback.py,sha256=zFngVclUL_7oDipjzKC8b7jToPNV11DV28rGyH8pio0,1987
pydub/pyaudioop.py,sha256=Dp_cQgAyYjD4OV2ZHuxtKI2KABuPi9YYNRUF8giR80Q,13094
pydub/scipy_effects.py,sha256=U2p8AQuVreTp5MrtUAzRbWgOHUc6Dwq0TAG_RtEg-7g,6637
pydub/silence.py,sha256=F6MV0VlaO6mtuisjLGks_UR-GVmzO1v87_NKvzwRc30,6457
pydub/utils.py,sha256=W71pgJFbbNP3adH63yn0Eo0CLLVgzXG7WHYSXpWvdyc,12368
venv/lib/python3.11/site-packages/pydub-0.25.1.dist-info/REQUESTED
View File
venv/lib/python3.11/site-packages/pydub-0.25.1.dist-info/WHEEL
+6
View File
@@ -0,0 +1,6 @@
Wheel-Version: 1.0
Generator: bdist_wheel (0.36.2)
Root-Is-Purelib: true
Tag: py2-none-any
Tag: py3-none-any
venv/lib/python3.11/site-packages/pydub-0.25.1.dist-info/top_level.txt
+1
View File
@@ -0,0 +1 @@
pydub
venv/lib/python3.11/site-packages/pydub/__init__.py
+1
View File
@@ -0,0 +1 @@
from .audio_segment import AudioSegment
venv/lib/python3.11/site-packages/pydub/__pycache__/__init__.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/audio_segment.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/effects.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/exceptions.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/generators.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/logging_utils.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/playback.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/pyaudioop.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/scipy_effects.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/silence.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/__pycache__/utils.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/pydub/audio_segment.py
+1399
View File
File diff suppressed because it is too large Load Diff
venv/lib/python3.11/site-packages/pydub/effects.py
+341
View File
@@ -0,0 +1,341 @@
import sys
import math
import array
from .utils import (
db_to_float,
ratio_to_db,
register_pydub_effect,
make_chunks,
audioop,
get_min_max_value
)
from .silence import split_on_silence
from .exceptions import TooManyMissingFrames, InvalidDuration
if sys.version_info >= (3, 0):
xrange = range
@register_pydub_effect
def apply_mono_filter_to_each_channel(seg, filter_fn):
n_channels = seg.channels
channel_segs = seg.split_to_mono()
channel_segs = [filter_fn(channel_seg) for channel_seg in channel_segs]
out_data = seg.get_array_of_samples()
for channel_i, channel_seg in enumerate(channel_segs):
for sample_i, sample in enumerate(channel_seg.get_array_of_samples()):
index = (sample_i * n_channels) + channel_i
out_data[index] = sample
return seg._spawn(out_data)
@register_pydub_effect
def normalize(seg, headroom=0.1):
"""
headroom is how close to the maximum volume to boost the signal up to (specified in dB)
"""
peak_sample_val = seg.max
# if the max is 0, this audio segment is silent, and can't be normalized
if peak_sample_val == 0:
return seg
target_peak = seg.max_possible_amplitude * db_to_float(-headroom)
needed_boost = ratio_to_db(target_peak / peak_sample_val)
return seg.apply_gain(needed_boost)
@register_pydub_effect
def speedup(seg, playback_speed=1.5, chunk_size=150, crossfade=25):
# we will keep audio in 150ms chunks since one waveform at 20Hz is 50ms long
# (20 Hz is the lowest frequency audible to humans)
# portion of AUDIO TO KEEP. if playback speed is 1.25 we keep 80% (0.8) and
# discard 20% (0.2)
atk = 1.0 / playback_speed
if playback_speed < 2.0:
# throwing out more than half the audio - keep 50ms chunks
ms_to_remove_per_chunk = int(chunk_size * (1 - atk) / atk)
else:
# throwing out less than half the audio - throw out 50ms chunks
ms_to_remove_per_chunk = int(chunk_size)
chunk_size = int(atk * chunk_size / (1 - atk))
# the crossfade cannot be longer than the amount of audio we're removing
crossfade = min(crossfade, ms_to_remove_per_chunk - 1)
# DEBUG
#print("chunk: {0}, rm: {1}".format(chunk_size, ms_to_remove_per_chunk))
chunks = make_chunks(seg, chunk_size + ms_to_remove_per_chunk)
if len(chunks) < 2:
raise Exception("Could not speed up AudioSegment, it was too short {2:0.2f}s for the current settings:\n{0}ms chunks at {1:0.1f}x speedup".format(
chunk_size, playback_speed, seg.duration_seconds))
# we'll actually truncate a bit less than we calculated to make up for the
# crossfade between chunks
ms_to_remove_per_chunk -= crossfade
# we don't want to truncate the last chunk since it is not guaranteed to be
# the full chunk length
last_chunk = chunks[-1]
chunks = [chunk[:-ms_to_remove_per_chunk] for chunk in chunks[:-1]]
out = chunks[0]
for chunk in chunks[1:]:
out = out.append(chunk, crossfade=crossfade)
out += last_chunk
return out
@register_pydub_effect
def strip_silence(seg, silence_len=1000, silence_thresh=-16, padding=100):
if padding > silence_len:
raise InvalidDuration("padding cannot be longer than silence_len")
chunks = split_on_silence(seg, silence_len, silence_thresh, padding)
crossfade = padding / 2
if not len(chunks):
return seg[0:0]
seg = chunks[0]
for chunk in chunks[1:]:
seg = seg.append(chunk, crossfade=crossfade)
return seg
@register_pydub_effect
def compress_dynamic_range(seg, threshold=-20.0, ratio=4.0, attack=5.0, release=50.0):
"""
Keyword Arguments:
threshold - default: -20.0
Threshold in dBFS. default of -20.0 means -20dB relative to the
maximum possible volume. 0dBFS is the maximum possible value so
all values for this argument sould be negative.
ratio - default: 4.0
Compression ratio. Audio louder than the threshold will be
reduced to 1/ratio the volume. A ratio of 4.0 is equivalent to
a setting of 4:1 in a pro-audio compressor like the Waves C1.
attack - default: 5.0
Attack in milliseconds. How long it should take for the compressor
to kick in once the audio has exceeded the threshold.
release - default: 50.0
Release in milliseconds. How long it should take for the compressor
to stop compressing after the audio has falled below the threshold.
For an overview of Dynamic Range Compression, and more detailed explanation
of the related terminology, see:
http://en.wikipedia.org/wiki/Dynamic_range_compression
"""
thresh_rms = seg.max_possible_amplitude * db_to_float(threshold)
look_frames = int(seg.frame_count(ms=attack))
def rms_at(frame_i):
return seg.get_sample_slice(frame_i - look_frames, frame_i).rms
def db_over_threshold(rms):
if rms == 0: return 0.0
db = ratio_to_db(rms / thresh_rms)
return max(db, 0)
output = []
# amount to reduce the volume of the audio by (in dB)
attenuation = 0.0
attack_frames = seg.frame_count(ms=attack)
release_frames = seg.frame_count(ms=release)
for i in xrange(int(seg.frame_count())):
rms_now = rms_at(i)
# with a ratio of 4.0 this means the volume will exceed the threshold by
# 1/4 the amount (of dB) that it would otherwise
max_attenuation = (1 - (1.0 / ratio)) * db_over_threshold(rms_now)
attenuation_inc = max_attenuation / attack_frames
attenuation_dec = max_attenuation / release_frames
if rms_now > thresh_rms and attenuation <= max_attenuation:
attenuation += attenuation_inc
attenuation = min(attenuation, max_attenuation)
else:
attenuation -= attenuation_dec
attenuation = max(attenuation, 0)
frame = seg.get_frame(i)
if attenuation != 0.0:
frame = audioop.mul(frame,
seg.sample_width,
db_to_float(-attenuation))
output.append(frame)
return seg._spawn(data=b''.join(output))
# Invert the phase of the signal.
@register_pydub_effect
def invert_phase(seg, channels=(1, 1)):
"""
channels- specifies which channel (left or right) to reverse the phase of.
Note that mono AudioSegments will become stereo.
"""
if channels == (1, 1):
inverted = audioop.mul(seg._data, seg.sample_width, -1.0)
return seg._spawn(data=inverted)
else:
if seg.channels == 2:
left, right = seg.split_to_mono()
else:
raise Exception("Can't implicitly convert an AudioSegment with " + str(seg.channels) + " channels to stereo.")
if channels == (1, 0):
left = left.invert_phase()
else:
right = right.invert_phase()
return seg.from_mono_audiosegments(left, right)
# High and low pass filters based on implementation found on Stack Overflow:
# http://stackoverflow.com/questions/13882038/implementing-simple-high-and-low-pass-filters-in-c
@register_pydub_effect
def low_pass_filter(seg, cutoff):
"""
cutoff - Frequency (in Hz) where higher frequency signal will begin to
be reduced by 6dB per octave (doubling in frequency) above this point
"""
RC = 1.0 / (cutoff * 2 * math.pi)
dt = 1.0 / seg.frame_rate
alpha = dt / (RC + dt)
original = seg.get_array_of_samples()
filteredArray = array.array(seg.array_type, original)
frame_count = int(seg.frame_count())
last_val = [0] * seg.channels
for i in range(seg.channels):
last_val[i] = filteredArray[i] = original[i]
for i in range(1, frame_count):
for j in range(seg.channels):
offset = (i * seg.channels) + j
last_val[j] = last_val[j] + (alpha * (original[offset] - last_val[j]))
filteredArray[offset] = int(last_val[j])
return seg._spawn(data=filteredArray)
@register_pydub_effect
def high_pass_filter(seg, cutoff):
"""
cutoff - Frequency (in Hz) where lower frequency signal will begin to
be reduced by 6dB per octave (doubling in frequency) below this point
"""
RC = 1.0 / (cutoff * 2 * math.pi)
dt = 1.0 / seg.frame_rate
alpha = RC / (RC + dt)
minval, maxval = get_min_max_value(seg.sample_width * 8)
original = seg.get_array_of_samples()
filteredArray = array.array(seg.array_type, original)
frame_count = int(seg.frame_count())
last_val = [0] * seg.channels
for i in range(seg.channels):
last_val[i] = filteredArray[i] = original[i]
for i in range(1, frame_count):
for j in range(seg.channels):
offset = (i * seg.channels) + j
offset_minus_1 = ((i-1) * seg.channels) + j
last_val[j] = alpha * (last_val[j] + original[offset] - original[offset_minus_1])
filteredArray[offset] = int(min(max(last_val[j], minval), maxval))
return seg._spawn(data=filteredArray)
@register_pydub_effect
def pan(seg, pan_amount):
"""
pan_amount should be between -1.0 (100% left) and +1.0 (100% right)
When pan_amount == 0.0 the left/right balance is not changed.
Panning does not alter the *perceived* loundness, but since loudness
is decreasing on one side, the other side needs to get louder to
compensate. When panned hard left, the left channel will be 3dB louder.
"""
if not -1.0 <= pan_amount <= 1.0:
raise ValueError("pan_amount should be between -1.0 (100% left) and +1.0 (100% right)")
max_boost_db = ratio_to_db(2.0)
boost_db = abs(pan_amount) * max_boost_db
boost_factor = db_to_float(boost_db)
reduce_factor = db_to_float(max_boost_db) - boost_factor
reduce_db = ratio_to_db(reduce_factor)
# Cut boost in half (max boost== 3dB) - in reality 2 speakers
# do not sum to a full 6 dB.
boost_db = boost_db / 2.0
if pan_amount < 0:
return seg.apply_gain_stereo(boost_db, reduce_db)
else:
return seg.apply_gain_stereo(reduce_db, boost_db)
@register_pydub_effect
def apply_gain_stereo(seg, left_gain=0.0, right_gain=0.0):
"""
left_gain - amount of gain to apply to the left channel (in dB)
right_gain - amount of gain to apply to the right channel (in dB)
note: mono audio segments will be converted to stereo
"""
if seg.channels == 1:
left = right = seg
elif seg.channels == 2:
left, right = seg.split_to_mono()
l_mult_factor = db_to_float(left_gain)
r_mult_factor = db_to_float(right_gain)
left_data = audioop.mul(left._data, left.sample_width, l_mult_factor)
left_data = audioop.tostereo(left_data, left.sample_width, 1, 0)
right_data = audioop.mul(right._data, right.sample_width, r_mult_factor)
right_data = audioop.tostereo(right_data, right.sample_width, 0, 1)
output = audioop.add(left_data, right_data, seg.sample_width)
return seg._spawn(data=output,
overrides={'channels': 2,
'frame_width': 2 * seg.sample_width})
venv/lib/python3.11/site-packages/pydub/exceptions.py
+32
View File
@@ -0,0 +1,32 @@
class PydubException(Exception):
"""
Base class for any Pydub exception
"""
class TooManyMissingFrames(PydubException):
pass
class InvalidDuration(PydubException):
pass
class InvalidTag(PydubException):
pass
class InvalidID3TagVersion(PydubException):
pass
class CouldntDecodeError(PydubException):
pass
class CouldntEncodeError(PydubException):
pass
class MissingAudioParameter(PydubException):
pass
venv/lib/python3.11/site-packages/pydub/generators.py
+142
View File
@@ -0,0 +1,142 @@
"""
Each generator will return float samples from -1.0 to 1.0, which can be
converted to actual audio with 8, 16, 24, or 32 bit depth using the
SiganlGenerator.to_audio_segment() method (on any of it's subclasses).
See Wikipedia's "waveform" page for info on some of the generators included
here: http://en.wikipedia.org/wiki/Waveform
"""
import math
import array
import itertools
import random
from .audio_segment import AudioSegment
from .utils import (
db_to_float,
get_frame_width,
get_array_type,
get_min_max_value
)
class SignalGenerator(object):
def __init__(self, sample_rate=44100, bit_depth=16):
self.sample_rate = sample_rate
self.bit_depth = bit_depth
def to_audio_segment(self, duration=1000.0, volume=0.0):
"""
Duration in milliseconds
(default: 1 second)
Volume in DB relative to maximum amplitude
(default 0.0 dBFS, which is the maximum value)
"""
minval, maxval = get_min_max_value(self.bit_depth)
sample_width = get_frame_width(self.bit_depth)
array_type = get_array_type(self.bit_depth)
gain = db_to_float(volume)
sample_count = int(self.sample_rate * (duration / 1000.0))
sample_data = (int(val * maxval * gain) for val in self.generate())
sample_data = itertools.islice(sample_data, 0, sample_count)
data = array.array(array_type, sample_data)
try:
data = data.tobytes()
except:
data = data.tostring()
return AudioSegment(data=data, metadata={
"channels": 1,
"sample_width": sample_width,
"frame_rate": self.sample_rate,
"frame_width": sample_width,
})
def generate(self):
raise NotImplementedError("SignalGenerator subclasses must implement the generate() method, and *should not* call the superclass implementation.")
class Sine(SignalGenerator):
def __init__(self, freq, **kwargs):
super(Sine, self).__init__(**kwargs)
self.freq = freq
def generate(self):
sine_of = (self.freq * 2 * math.pi) / self.sample_rate
sample_n = 0
while True:
yield math.sin(sine_of * sample_n)
sample_n += 1
class Pulse(SignalGenerator):
def __init__(self, freq, duty_cycle=0.5, **kwargs):
super(Pulse, self).__init__(**kwargs)
self.freq = freq
self.duty_cycle = duty_cycle
def generate(self):
sample_n = 0
# in samples
cycle_length = self.sample_rate / float(self.freq)
pulse_length = cycle_length * self.duty_cycle
while True:
if (sample_n % cycle_length) < pulse_length:
yield 1.0
else:
yield -1.0
sample_n += 1
class Square(Pulse):
def __init__(self, freq, **kwargs):
kwargs['duty_cycle'] = 0.5
super(Square, self).__init__(freq, **kwargs)
class Sawtooth(SignalGenerator):
def __init__(self, freq, duty_cycle=1.0, **kwargs):
super(Sawtooth, self).__init__(**kwargs)
self.freq = freq
self.duty_cycle = duty_cycle
def generate(self):
sample_n = 0
# in samples
cycle_length = self.sample_rate / float(self.freq)
midpoint = cycle_length * self.duty_cycle
ascend_length = midpoint
descend_length = cycle_length - ascend_length
while True:
cycle_position = sample_n % cycle_length
if cycle_position < midpoint:
yield (2 * cycle_position / ascend_length) - 1.0
else:
yield 1.0 - (2 * (cycle_position - midpoint) / descend_length)
sample_n += 1
class Triangle(Sawtooth):
def __init__(self, freq, **kwargs):
kwargs['duty_cycle'] = 0.5
super(Triangle, self).__init__(freq, **kwargs)
class WhiteNoise(SignalGenerator):
def generate(self):
while True:
yield (random.random() * 2) - 1.0
venv/lib/python3.11/site-packages/pydub/logging_utils.py
+14
View File
@@ -0,0 +1,14 @@
"""
"""
import logging
converter_logger = logging.getLogger("pydub.converter")
def log_conversion(conversion_command):
converter_logger.debug("subprocess.call(%s)", repr(conversion_command))
def log_subprocess_output(output):
if output:
for line in output.rstrip().splitlines():
converter_logger.debug('subprocess output: %s', line.rstrip())
venv/lib/python3.11/site-packages/pydub/playback.py
+71
View File
@@ -0,0 +1,71 @@
"""
Support for playing AudioSegments. Pyaudio will be used if it's installed,
otherwise will fallback to ffplay. Pyaudio is a *much* nicer solution, but
is tricky to install. See my notes on installing pyaudio in a virtualenv (on
OSX 10.10): https://gist.github.com/jiaaro/9767512210a1d80a8a0d
"""
import subprocess
from tempfile import NamedTemporaryFile
from .utils import get_player_name, make_chunks
def _play_with_ffplay(seg):
PLAYER = get_player_name()
with NamedTemporaryFile("w+b", suffix=".wav") as f:
seg.export(f.name, "wav")
subprocess.call([PLAYER, "-nodisp", "-autoexit", "-hide_banner", f.name])
def _play_with_pyaudio(seg):
import pyaudio
p = pyaudio.PyAudio()
stream = p.open(format=p.get_format_from_width(seg.sample_width),
channels=seg.channels,
rate=seg.frame_rate,
output=True)
# Just in case there were any exceptions/interrupts, we release the resource
# So as not to raise OSError: Device Unavailable should play() be used again
try:
# break audio into half-second chunks (to allows keyboard interrupts)
for chunk in make_chunks(seg, 500):
stream.write(chunk._data)
finally:
stream.stop_stream()
stream.close()
p.terminate()
def _play_with_simpleaudio(seg):
import simpleaudio
return simpleaudio.play_buffer(
seg.raw_data,
num_channels=seg.channels,
bytes_per_sample=seg.sample_width,
sample_rate=seg.frame_rate
)
def play(audio_segment):
try:
playback = _play_with_simpleaudio(audio_segment)
try:
playback.wait_done()
except KeyboardInterrupt:
playback.stop()
except ImportError:
pass
else:
return
try:
_play_with_pyaudio(audio_segment)
return
except ImportError:
pass
else:
return
_play_with_ffplay(audio_segment)
venv/lib/python3.11/site-packages/pydub/pyaudioop.py
+553
View File
@@ -0,0 +1,553 @@
try:
from __builtin__ import max as builtin_max
from __builtin__ import min as builtin_min
except ImportError:
from builtins import max as builtin_max
from builtins import min as builtin_min
import math
import struct
try:
from fractions import gcd
except ImportError: # Python 3.9+
from math import gcd
from ctypes import create_string_buffer
class error(Exception):
pass
def _check_size(size):
if size != 1 and size != 2 and size != 4:
raise error("Size should be 1, 2 or 4")
def _check_params(length, size):
_check_size(size)
if length % size != 0:
raise error("not a whole number of frames")
def _sample_count(cp, size):
return len(cp) / size
def _get_samples(cp, size, signed=True):
for i in range(_sample_count(cp, size)):
yield _get_sample(cp, size, i, signed)
def _struct_format(size, signed):
if size == 1:
return "b" if signed else "B"
elif size == 2:
return "h" if signed else "H"
elif size == 4:
return "i" if signed else "I"
def _get_sample(cp, size, i, signed=True):
fmt = _struct_format(size, signed)
start = i * size
end = start + size
return struct.unpack_from(fmt, buffer(cp)[start:end])[0]
def _put_sample(cp, size, i, val, signed=True):
fmt = _struct_format(size, signed)
struct.pack_into(fmt, cp, i * size, val)
def _get_maxval(size, signed=True):
if signed and size == 1:
return 0x7f
elif size == 1:
return 0xff
elif signed and size == 2:
return 0x7fff
elif size == 2:
return 0xffff
elif signed and size == 4:
return 0x7fffffff
elif size == 4:
return 0xffffffff
def _get_minval(size, signed=True):
if not signed:
return 0
elif size == 1:
return -0x80
elif size == 2:
return -0x8000
elif size == 4:
return -0x80000000
def _get_clipfn(size, signed=True):
maxval = _get_maxval(size, signed)
minval = _get_minval(size, signed)
return lambda val: builtin_max(min(val, maxval), minval)
def _overflow(val, size, signed=True):
minval = _get_minval(size, signed)
maxval = _get_maxval(size, signed)
if minval <= val <= maxval:
return val
bits = size * 8
if signed:
offset = 2**(bits-1)
return ((val + offset) % (2**bits)) - offset
else:
return val % (2**bits)
def getsample(cp, size, i):
_check_params(len(cp), size)
if not (0 <= i < len(cp) / size):
raise error("Index out of range")
return _get_sample(cp, size, i)
def max(cp, size):
_check_params(len(cp), size)
if len(cp) == 0:
return 0
return builtin_max(abs(sample) for sample in _get_samples(cp, size))
def minmax(cp, size):
_check_params(len(cp), size)
max_sample, min_sample = 0, 0
for sample in _get_samples(cp, size):
max_sample = builtin_max(sample, max_sample)
min_sample = builtin_min(sample, min_sample)
return min_sample, max_sample
def avg(cp, size):
_check_params(len(cp), size)
sample_count = _sample_count(cp, size)
if sample_count == 0:
return 0
return sum(_get_samples(cp, size)) / sample_count
def rms(cp, size):
_check_params(len(cp), size)
sample_count = _sample_count(cp, size)
if sample_count == 0:
return 0
sum_squares = sum(sample**2 for sample in _get_samples(cp, size))
return int(math.sqrt(sum_squares / sample_count))
def _sum2(cp1, cp2, length):
size = 2
total = 0
for i in range(length):
total += getsample(cp1, size, i) * getsample(cp2, size, i)
return total
def findfit(cp1, cp2):
size = 2
if len(cp1) % 2 != 0 or len(cp2) % 2 != 0:
raise error("Strings should be even-sized")
if len(cp1) < len(cp2):
raise error("First sample should be longer")
len1 = _sample_count(cp1, size)
len2 = _sample_count(cp2, size)
sum_ri_2 = _sum2(cp2, cp2, len2)
sum_aij_2 = _sum2(cp1, cp1, len2)
sum_aij_ri = _sum2(cp1, cp2, len2)
result = (sum_ri_2 * sum_aij_2 - sum_aij_ri * sum_aij_ri) / sum_aij_2
best_result = result
best_i = 0
for i in range(1, len1 - len2 + 1):
aj_m1 = _get_sample(cp1, size, i - 1)
aj_lm1 = _get_sample(cp1, size, i + len2 - 1)
sum_aij_2 += aj_lm1**2 - aj_m1**2
sum_aij_ri = _sum2(buffer(cp1)[i*size:], cp2, len2)
result = (sum_ri_2 * sum_aij_2 - sum_aij_ri * sum_aij_ri) / sum_aij_2
if result < best_result:
best_result = result
best_i = i
factor = _sum2(buffer(cp1)[best_i*size:], cp2, len2) / sum_ri_2
return best_i, factor
def findfactor(cp1, cp2):
size = 2
if len(cp1) % 2 != 0:
raise error("Strings should be even-sized")
if len(cp1) != len(cp2):
raise error("Samples should be same size")
sample_count = _sample_count(cp1, size)
sum_ri_2 = _sum2(cp2, cp2, sample_count)
sum_aij_ri = _sum2(cp1, cp2, sample_count)
return sum_aij_ri / sum_ri_2
def findmax(cp, len2):
size = 2
sample_count = _sample_count(cp, size)
if len(cp) % 2 != 0:
raise error("Strings should be even-sized")
if len2 < 0 or sample_count < len2:
raise error("Input sample should be longer")
if sample_count == 0:
return 0
result = _sum2(cp, cp, len2)
best_result = result
best_i = 0
for i in range(1, sample_count - len2 + 1):
sample_leaving_window = getsample(cp, size, i - 1)
sample_entering_window = getsample(cp, size, i + len2 - 1)
result -= sample_leaving_window**2
result += sample_entering_window**2
if result > best_result:
best_result = result
best_i = i
return best_i
def avgpp(cp, size):
_check_params(len(cp), size)
sample_count = _sample_count(cp, size)
prevextremevalid = False
prevextreme = None
avg = 0
nextreme = 0
prevval = getsample(cp, size, 0)
val = getsample(cp, size, 1)
prevdiff = val - prevval
for i in range(1, sample_count):
val = getsample(cp, size, i)
diff = val - prevval
if diff * prevdiff < 0:
if prevextremevalid:
avg += abs(prevval - prevextreme)
nextreme += 1
prevextremevalid = True
prevextreme = prevval
prevval = val
if diff != 0:
prevdiff = diff
if nextreme == 0:
return 0
return avg / nextreme
def maxpp(cp, size):
_check_params(len(cp), size)
sample_count = _sample_count(cp, size)
prevextremevalid = False
prevextreme = None
max = 0
prevval = getsample(cp, size, 0)
val = getsample(cp, size, 1)
prevdiff = val - prevval
for i in range(1, sample_count):
val = getsample(cp, size, i)
diff = val - prevval
if diff * prevdiff < 0:
if prevextremevalid:
extremediff = abs(prevval - prevextreme)
if extremediff > max:
max = extremediff
prevextremevalid = True
prevextreme = prevval
prevval = val
if diff != 0:
prevdiff = diff
return max
def cross(cp, size):
_check_params(len(cp), size)
crossings = 0
last_sample = 0
for sample in _get_samples(cp, size):
if sample <= 0 < last_sample or sample >= 0 > last_sample:
crossings += 1
last_sample = sample
return crossings
def mul(cp, size, factor):
_check_params(len(cp), size)
clip = _get_clipfn(size)
result = create_string_buffer(len(cp))
for i, sample in enumerate(_get_samples(cp, size)):
sample = clip(int(sample * factor))
_put_sample(result, size, i, sample)
return result.raw
def tomono(cp, size, fac1, fac2):
_check_params(len(cp), size)
clip = _get_clipfn(size)
sample_count = _sample_count(cp, size)
result = create_string_buffer(len(cp) / 2)
for i in range(0, sample_count, 2):
l_sample = getsample(cp, size, i)
r_sample = getsample(cp, size, i + 1)
sample = (l_sample * fac1) + (r_sample * fac2)
sample = clip(sample)
_put_sample(result, size, i / 2, sample)
return result.raw
def tostereo(cp, size, fac1, fac2):
_check_params(len(cp), size)
sample_count = _sample_count(cp, size)
result = create_string_buffer(len(cp) * 2)
clip = _get_clipfn(size)
for i in range(sample_count):
sample = _get_sample(cp, size, i)
l_sample = clip(sample * fac1)
r_sample = clip(sample * fac2)
_put_sample(result, size, i * 2, l_sample)
_put_sample(result, size, i * 2 + 1, r_sample)
return result.raw
def add(cp1, cp2, size):
_check_params(len(cp1), size)
if len(cp1) != len(cp2):
raise error("Lengths should be the same")
clip = _get_clipfn(size)
sample_count = _sample_count(cp1, size)
result = create_string_buffer(len(cp1))
for i in range(sample_count):
sample1 = getsample(cp1, size, i)
sample2 = getsample(cp2, size, i)
sample = clip(sample1 + sample2)
_put_sample(result, size, i, sample)
return result.raw
def bias(cp, size, bias):
_check_params(len(cp), size)
result = create_string_buffer(len(cp))
for i, sample in enumerate(_get_samples(cp, size)):
sample = _overflow(sample + bias, size)
_put_sample(result, size, i, sample)
return result.raw
def reverse(cp, size):
_check_params(len(cp), size)
sample_count = _sample_count(cp, size)
result = create_string_buffer(len(cp))
for i, sample in enumerate(_get_samples(cp, size)):
_put_sample(result, size, sample_count - i - 1, sample)
return result.raw
def lin2lin(cp, size, size2):
_check_params(len(cp), size)
_check_size(size2)
if size == size2:
return cp
new_len = (len(cp) / size) * size2
result = create_string_buffer(new_len)
for i in range(_sample_count(cp, size)):
sample = _get_sample(cp, size, i)
if size < size2:
sample = sample << (4 * size2 / size)
elif size > size2:
sample = sample >> (4 * size / size2)
sample = _overflow(sample, size2)
_put_sample(result, size2, i, sample)
return result.raw
def ratecv(cp, size, nchannels, inrate, outrate, state, weightA=1, weightB=0):
_check_params(len(cp), size)
if nchannels < 1:
raise error("# of channels should be >= 1")
bytes_per_frame = size * nchannels
frame_count = len(cp) / bytes_per_frame
if bytes_per_frame / nchannels != size:
raise OverflowError("width * nchannels too big for a C int")
if weightA < 1 or weightB < 0:
raise error("weightA should be >= 1, weightB should be >= 0")
if len(cp) % bytes_per_frame != 0:
raise error("not a whole number of frames")
if inrate <= 0 or outrate <= 0:
raise error("sampling rate not > 0")
d = gcd(inrate, outrate)
inrate /= d
outrate /= d
prev_i = [0] * nchannels
cur_i = [0] * nchannels
if state is None:
d = -outrate
else:
d, samps = state
if len(samps) != nchannels:
raise error("illegal state argument")
prev_i, cur_i = zip(*samps)
prev_i, cur_i = list(prev_i), list(cur_i)
q = frame_count / inrate
ceiling = (q + 1) * outrate
nbytes = ceiling * bytes_per_frame
result = create_string_buffer(nbytes)
samples = _get_samples(cp, size)
out_i = 0
while True:
while d < 0:
if frame_count == 0:
samps = zip(prev_i, cur_i)
retval = result.raw
# slice off extra bytes
trim_index = (out_i * bytes_per_frame) - len(retval)
retval = buffer(retval)[:trim_index]
return (retval, (d, tuple(samps)))
for chan in range(nchannels):
prev_i[chan] = cur_i[chan]
cur_i[chan] = samples.next()
cur_i[chan] = (
(weightA * cur_i[chan] + weightB * prev_i[chan])
/ (weightA + weightB)
)
frame_count -= 1
d += outrate
while d >= 0:
for chan in range(nchannels):
cur_o = (
(prev_i[chan] * d + cur_i[chan] * (outrate - d))
/ outrate
)
_put_sample(result, size, out_i, _overflow(cur_o, size))
out_i += 1
d -= inrate
def lin2ulaw(cp, size):
raise NotImplementedError()
def ulaw2lin(cp, size):
raise NotImplementedError()
def lin2alaw(cp, size):
raise NotImplementedError()
def alaw2lin(cp, size):
raise NotImplementedError()
def lin2adpcm(cp, size, state):
raise NotImplementedError()
def adpcm2lin(cp, size, state):
raise NotImplementedError()
venv/lib/python3.11/site-packages/pydub/scipy_effects.py
+175
View File
@@ -0,0 +1,175 @@
"""
This module provides scipy versions of high_pass_filter, and low_pass_filter
as well as an additional band_pass_filter.
Of course, you will need to install scipy for these to work.
When this module is imported the high and low pass filters from this module
will be used when calling audio_segment.high_pass_filter() and
audio_segment.high_pass_filter() instead of the slower, less powerful versions
provided by pydub.effects.
"""
from scipy.signal import butter, sosfilt
from .utils import (register_pydub_effect,stereo_to_ms,ms_to_stereo)
def _mk_butter_filter(freq, type, order):
"""
Args:
freq: The cutoff frequency for highpass and lowpass filters. For
band filters, a list of [low_cutoff, high_cutoff]
type: "lowpass", "highpass", or "band"
order: nth order butterworth filter (default: 5th order). The
attenuation is -6dB/octave beyond the cutoff frequency (for 1st
order). A Higher order filter will have more attenuation, each level
adding an additional -6dB (so a 3rd order butterworth filter would
be -18dB/octave).
Returns:
function which can filter a mono audio segment
"""
def filter_fn(seg):
assert seg.channels == 1
nyq = 0.5 * seg.frame_rate
try:
freqs = [f / nyq for f in freq]
except TypeError:
freqs = freq / nyq
sos = butter(order, freqs, btype=type, output='sos')
y = sosfilt(sos, seg.get_array_of_samples())
return seg._spawn(y.astype(seg.array_type))
return filter_fn
@register_pydub_effect
def band_pass_filter(seg, low_cutoff_freq, high_cutoff_freq, order=5):
filter_fn = _mk_butter_filter([low_cutoff_freq, high_cutoff_freq], 'band', order=order)
return seg.apply_mono_filter_to_each_channel(filter_fn)
@register_pydub_effect
def high_pass_filter(seg, cutoff_freq, order=5):
filter_fn = _mk_butter_filter(cutoff_freq, 'highpass', order=order)
return seg.apply_mono_filter_to_each_channel(filter_fn)
@register_pydub_effect
def low_pass_filter(seg, cutoff_freq, order=5):
filter_fn = _mk_butter_filter(cutoff_freq, 'lowpass', order=order)
return seg.apply_mono_filter_to_each_channel(filter_fn)
@register_pydub_effect
def _eq(seg, focus_freq, bandwidth=100, mode="peak", gain_dB=0, order=2):
"""
Args:
focus_freq - middle frequency or known frequency of band (in Hz)
bandwidth - range of the equalizer band
mode - Mode of Equalization(Peak/Notch(Bell Curve),High Shelf, Low Shelf)
order - Rolloff factor(1 - 6dB/Octave 2 - 12dB/Octave)
Returns:
Equalized/Filtered AudioSegment
"""
filt_mode = ["peak", "low_shelf", "high_shelf"]
if mode not in filt_mode:
raise ValueError("Incorrect Mode Selection")
if gain_dB >= 0:
if mode == "peak":
sec = band_pass_filter(seg, focus_freq - bandwidth/2, focus_freq + bandwidth/2, order = order)
seg = seg.overlay(sec - (3 - gain_dB))
return seg
if mode == "low_shelf":
sec = low_pass_filter(seg, focus_freq, order=order)
seg = seg.overlay(sec - (3 - gain_dB))
return seg
if mode == "high_shelf":
sec = high_pass_filter(seg, focus_freq, order=order)
seg = seg.overlay(sec - (3 - gain_dB))
return seg
if gain_dB < 0:
if mode == "peak":
sec = high_pass_filter(seg, focus_freq - bandwidth/2, order=order)
seg = seg.overlay(sec - (3 + gain_dB)) + gain_dB
sec = low_pass_filter(seg, focus_freq + bandwidth/2, order=order)
seg = seg.overlay(sec - (3 + gain_dB)) + gain_dB
return seg
if mode == "low_shelf":
sec = high_pass_filter(seg, focus_freq, order=order)
seg = seg.overlay(sec - (3 + gain_dB)) + gain_dB
return seg
if mode=="high_shelf":
sec=low_pass_filter(seg, focus_freq, order=order)
seg=seg.overlay(sec - (3 + gain_dB)) +gain_dB
return seg
@register_pydub_effect
def eq(seg, focus_freq, bandwidth=100, channel_mode="L+R", filter_mode="peak", gain_dB=0, order=2):
"""
Args:
focus_freq - middle frequency or known frequency of band (in Hz)
bandwidth - range of the equalizer band
channel_mode - Select Channels to be affected by the filter.
L+R - Standard Stereo Filter
L - Only Left Channel is Filtered
R - Only Right Channel is Filtered
M+S - Blumlien Stereo Filter(Mid-Side)
M - Only Mid Channel is Filtered
S - Only Side Channel is Filtered
Mono Audio Segments are completely filtered.
filter_mode - Mode of Equalization(Peak/Notch(Bell Curve),High Shelf, Low Shelf)
order - Rolloff factor(1 - 6dB/Octave 2 - 12dB/Octave)
Returns:
Equalized/Filtered AudioSegment
"""
channel_modes = ["L+R", "M+S", "L", "R", "M", "S"]
if channel_mode not in channel_modes:
raise ValueError("Incorrect Channel Mode Selection")
if seg.channels == 1:
return _eq(seg, focus_freq, bandwidth, filter_mode, gain_dB, order)
if channel_mode == "L+R":
return _eq(seg, focus_freq, bandwidth, filter_mode, gain_dB, order)
if channel_mode == "L":
seg = seg.split_to_mono()
seg = [_eq(seg[0], focus_freq, bandwidth, filter_mode, gain_dB, order), seg[1]]
return AudioSegment.from_mono_audio_segements(seg[0], seg[1])
if channel_mode == "R":
seg = seg.split_to_mono()
seg = [seg[0], _eq(seg[1], focus_freq, bandwidth, filter_mode, gain_dB, order)]
return AudioSegment.from_mono_audio_segements(seg[0], seg[1])
if channel_mode == "M+S":
seg = stereo_to_ms(seg)
seg = _eq(seg, focus_freq, bandwidth, filter_mode, gain_dB, order)
return ms_to_stereo(seg)
if channel_mode == "M":
seg = stereo_to_ms(seg).split_to_mono()
seg = [_eq(seg[0], focus_freq, bandwidth, filter_mode, gain_dB, order), seg[1]]
seg = AudioSegment.from_mono_audio_segements(seg[0], seg[1])
return ms_to_stereo(seg)
if channel_mode == "S":
seg = stereo_to_ms(seg).split_to_mono()
seg = [seg[0], _eq(seg[1], focus_freq, bandwidth, filter_mode, gain_dB, order)]
seg = AudioSegment.from_mono_audio_segements(seg[0], seg[1])
return ms_to_stereo(seg)
venv/lib/python3.11/site-packages/pydub/silence.py
+182
View File
@@ -0,0 +1,182 @@
"""
Various functions for finding/manipulating silence in AudioSegments
"""
import itertools
from .utils import db_to_float
def detect_silence(audio_segment, min_silence_len=1000, silence_thresh=-16, seek_step=1):
"""
Returns a list of all silent sections [start, end] in milliseconds of audio_segment.
Inverse of detect_nonsilent()
audio_segment - the segment to find silence in
min_silence_len - the minimum length for any silent section
silence_thresh - the upper bound for how quiet is silent in dFBS
seek_step - step size for interating over the segment in ms
"""
seg_len = len(audio_segment)
# you can't have a silent portion of a sound that is longer than the sound
if seg_len < min_silence_len:
return []
# convert silence threshold to a float value (so we can compare it to rms)
silence_thresh = db_to_float(silence_thresh) * audio_segment.max_possible_amplitude
# find silence and add start and end indicies to the to_cut list
silence_starts = []
# check successive (1 sec by default) chunk of sound for silence
# try a chunk at every "seek step" (or every chunk for a seek step == 1)
last_slice_start = seg_len - min_silence_len
slice_starts = range(0, last_slice_start + 1, seek_step)
# guarantee last_slice_start is included in the range
# to make sure the last portion of the audio is searched
if last_slice_start % seek_step:
slice_starts = itertools.chain(slice_starts, [last_slice_start])
for i in slice_starts:
audio_slice = audio_segment[i:i + min_silence_len]
if audio_slice.rms <= silence_thresh:
silence_starts.append(i)
# short circuit when there is no silence
if not silence_starts:
return []
# combine the silence we detected into ranges (start ms - end ms)
silent_ranges = []
prev_i = silence_starts.pop(0)
current_range_start = prev_i
for silence_start_i in silence_starts:
continuous = (silence_start_i == prev_i + seek_step)
# sometimes two small blips are enough for one particular slice to be
# non-silent, despite the silence all running together. Just combine
# the two overlapping silent ranges.
silence_has_gap = silence_start_i > (prev_i + min_silence_len)
if not continuous and silence_has_gap:
silent_ranges.append([current_range_start,
prev_i + min_silence_len])
current_range_start = silence_start_i
prev_i = silence_start_i
silent_ranges.append([current_range_start,
prev_i + min_silence_len])
return silent_ranges
def detect_nonsilent(audio_segment, min_silence_len=1000, silence_thresh=-16, seek_step=1):
"""
Returns a list of all nonsilent sections [start, end] in milliseconds of audio_segment.
Inverse of detect_silent()
audio_segment - the segment to find silence in
min_silence_len - the minimum length for any silent section
silence_thresh - the upper bound for how quiet is silent in dFBS
seek_step - step size for interating over the segment in ms
"""
silent_ranges = detect_silence(audio_segment, min_silence_len, silence_thresh, seek_step)
len_seg = len(audio_segment)
# if there is no silence, the whole thing is nonsilent
if not silent_ranges:
return [[0, len_seg]]
# short circuit when the whole audio segment is silent
if silent_ranges[0][0] == 0 and silent_ranges[0][1] == len_seg:
return []
prev_end_i = 0
nonsilent_ranges = []
for start_i, end_i in silent_ranges:
nonsilent_ranges.append([prev_end_i, start_i])
prev_end_i = end_i
if end_i != len_seg:
nonsilent_ranges.append([prev_end_i, len_seg])
if nonsilent_ranges[0] == [0, 0]:
nonsilent_ranges.pop(0)
return nonsilent_ranges
def split_on_silence(audio_segment, min_silence_len=1000, silence_thresh=-16, keep_silence=100,
seek_step=1):
"""
Returns list of audio segments from splitting audio_segment on silent sections
audio_segment - original pydub.AudioSegment() object
min_silence_len - (in ms) minimum length of a silence to be used for
a split. default: 1000ms
silence_thresh - (in dBFS) anything quieter than this will be
considered silence. default: -16dBFS
keep_silence - (in ms or True/False) leave some silence at the beginning
and end of the chunks. Keeps the sound from sounding like it
is abruptly cut off.
When the length of the silence is less than the keep_silence duration
it is split evenly between the preceding and following non-silent
segments.
If True is specified, all the silence is kept, if False none is kept.
default: 100ms
seek_step - step size for interating over the segment in ms
"""
# from the itertools documentation
def pairwise(iterable):
"s -> (s0,s1), (s1,s2), (s2, s3), ..."
a, b = itertools.tee(iterable)
next(b, None)
return zip(a, b)
if isinstance(keep_silence, bool):
keep_silence = len(audio_segment) if keep_silence else 0
output_ranges = [
[ start - keep_silence, end + keep_silence ]
for (start,end)
in detect_nonsilent(audio_segment, min_silence_len, silence_thresh, seek_step)
]
for range_i, range_ii in pairwise(output_ranges):
last_end = range_i[1]
next_start = range_ii[0]
if next_start < last_end:
range_i[1] = (last_end+next_start)//2
range_ii[0] = range_i[1]
return [
audio_segment[ max(start,0) : min(end,len(audio_segment)) ]
for start,end in output_ranges
]
def detect_leading_silence(sound, silence_threshold=-50.0, chunk_size=10):
"""
Returns the millisecond/index that the leading silence ends.
audio_segment - the segment to find silence in
silence_threshold - the upper bound for how quiet is silent in dFBS
chunk_size - chunk size for interating over the segment in ms
"""
trim_ms = 0 # ms
assert chunk_size > 0 # to avoid infinite loop
while sound[trim_ms:trim_ms+chunk_size].dBFS < silence_threshold and trim_ms < len(sound):
trim_ms += chunk_size
# if there is no end it should return the length of the segment
return min(trim_ms, len(sound))
venv/lib/python3.11/site-packages/pydub/utils.py
+434
View File
@@ -0,0 +1,434 @@
from __future__ import division
import json
import os
import re
import sys
from subprocess import Popen, PIPE
from math import log, ceil
from tempfile import TemporaryFile
from warnings import warn
from functools import wraps
try:
import audioop
except ImportError:
import pyaudioop as audioop
if sys.version_info >= (3, 0):
basestring = str
FRAME_WIDTHS = {
8: 1,
16: 2,
32: 4,
}
ARRAY_TYPES = {
8: "b",
16: "h",
32: "i",
}
ARRAY_RANGES = {
8: (-0x80, 0x7f),
16: (-0x8000, 0x7fff),
32: (-0x80000000, 0x7fffffff),
}
def get_frame_width(bit_depth):
return FRAME_WIDTHS[bit_depth]
def get_array_type(bit_depth, signed=True):
t = ARRAY_TYPES[bit_depth]
if not signed:
t = t.upper()
return t
def get_min_max_value(bit_depth):
return ARRAY_RANGES[bit_depth]
def _fd_or_path_or_tempfile(fd, mode='w+b', tempfile=True):
close_fd = False
if fd is None and tempfile:
fd = TemporaryFile(mode=mode)
close_fd = True
if isinstance(fd, basestring):
fd = open(fd, mode=mode)
close_fd = True
try:
if isinstance(fd, os.PathLike):
fd = open(fd, mode=mode)
close_fd = True
except AttributeError:
# module os has no attribute PathLike, so we're on python < 3.6.
# The protocol we're trying to support doesn't exist, so just pass.
pass
return fd, close_fd
def db_to_float(db, using_amplitude=True):
"""
Converts the input db to a float, which represents the equivalent
ratio in power.
"""
db = float(db)
if using_amplitude:
return 10 ** (db / 20)
else: # using power
return 10 ** (db / 10)
def ratio_to_db(ratio, val2=None, using_amplitude=True):
"""
Converts the input float to db, which represents the equivalent
to the ratio in power represented by the multiplier passed in.
"""
ratio = float(ratio)
# accept 2 values and use the ratio of val1 to val2
if val2 is not None:
ratio = ratio / val2
# special case for multiply-by-zero (convert to silence)
if ratio == 0:
return -float('inf')
if using_amplitude:
return 20 * log(ratio, 10)
else: # using power
return 10 * log(ratio, 10)
def register_pydub_effect(fn, name=None):
"""
decorator for adding pydub effects to the AudioSegment objects.
example use:
@register_pydub_effect
def normalize(audio_segment):
...
or you can specify a name:
@register_pydub_effect("normalize")
def normalize_audio_segment(audio_segment):
...
"""
if isinstance(fn, basestring):
name = fn
return lambda fn: register_pydub_effect(fn, name)
if name is None:
name = fn.__name__
from .audio_segment import AudioSegment
setattr(AudioSegment, name, fn)
return fn
def make_chunks(audio_segment, chunk_length):
"""
Breaks an AudioSegment into chunks that are <chunk_length> milliseconds
long.
if chunk_length is 50 then you'll get a list of 50 millisecond long audio
segments back (except the last one, which can be shorter)
"""
number_of_chunks = ceil(len(audio_segment) / float(chunk_length))
return [audio_segment[i * chunk_length:(i + 1) * chunk_length]
for i in range(int(number_of_chunks))]
def which(program):
"""
Mimics behavior of UNIX which command.
"""
# Add .exe program extension for windows support
if os.name == "nt" and not program.endswith(".exe"):
program += ".exe"
envdir_list = [os.curdir] + os.environ["PATH"].split(os.pathsep)
for envdir in envdir_list:
program_path = os.path.join(envdir, program)
if os.path.isfile(program_path) and os.access(program_path, os.X_OK):
return program_path
def get_encoder_name():
"""
Return enconder default application for system, either avconv or ffmpeg
"""
if which("avconv"):
return "avconv"
elif which("ffmpeg"):
return "ffmpeg"
else:
# should raise exception
warn("Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work", RuntimeWarning)
return "ffmpeg"
def get_player_name():
"""
Return enconder default application for system, either avconv or ffmpeg
"""
if which("avplay"):
return "avplay"
elif which("ffplay"):
return "ffplay"
else:
# should raise exception
warn("Couldn't find ffplay or avplay - defaulting to ffplay, but may not work", RuntimeWarning)
return "ffplay"
def get_prober_name():
"""
Return probe application, either avconv or ffmpeg
"""
if which("avprobe"):
return "avprobe"
elif which("ffprobe"):
return "ffprobe"
else:
# should raise exception
warn("Couldn't find ffprobe or avprobe - defaulting to ffprobe, but may not work", RuntimeWarning)
return "ffprobe"
def fsdecode(filename):
"""Wrapper for os.fsdecode which was introduced in python 3.2 ."""
if sys.version_info >= (3, 2):
PathLikeTypes = (basestring, bytes)
if sys.version_info >= (3, 6):
PathLikeTypes += (os.PathLike,)
if isinstance(filename, PathLikeTypes):
return os.fsdecode(filename)
else:
if isinstance(filename, bytes):
return filename.decode(sys.getfilesystemencoding())
if isinstance(filename, basestring):
return filename
raise TypeError("type {0} not accepted by fsdecode".format(type(filename)))
def get_extra_info(stderr):
"""
avprobe sometimes gives more information on stderr than
on the json output. The information has to be extracted
from stderr of the format of:
' Stream #0:0: Audio: flac, 88200 Hz, stereo, s32 (24 bit)'
or (macOS version):
' Stream #0:0: Audio: vorbis'
' 44100 Hz, stereo, fltp, 320 kb/s'
:type stderr: str
:rtype: list of dict
"""
extra_info = {}
re_stream = r'(?P<space_start> +)Stream #0[:\.](?P<stream_id>([0-9]+))(?P<content_0>.+)\n?(?! *Stream)((?P<space_end> +)(?P<content_1>.+))?'
for i in re.finditer(re_stream, stderr):
if i.group('space_end') is not None and len(i.group('space_start')) <= len(
i.group('space_end')):
content_line = ','.join([i.group('content_0'), i.group('content_1')])
else:
content_line = i.group('content_0')
tokens = [x.strip() for x in re.split('[:,]', content_line) if x]
extra_info[int(i.group('stream_id'))] = tokens
return extra_info
def mediainfo_json(filepath, read_ahead_limit=-1):
"""Return json dictionary with media info(codec, duration, size, bitrate...) from filepath
"""
prober = get_prober_name()
command_args = [
"-v", "info",
"-show_format",
"-show_streams",
]
try:
command_args += [fsdecode(filepath)]
stdin_parameter = None
stdin_data = None
except TypeError:
if prober == 'ffprobe':
command_args += ["-read_ahead_limit", str(read_ahead_limit),
"cache:pipe:0"]
else:
command_args += ["-"]
stdin_parameter = PIPE
file, close_file = _fd_or_path_or_tempfile(filepath, 'rb', tempfile=False)
file.seek(0)
stdin_data = file.read()
if close_file:
file.close()
command = [prober, '-of', 'json'] + command_args
res = Popen(command, stdin=stdin_parameter, stdout=PIPE, stderr=PIPE)
output, stderr = res.communicate(input=stdin_data)
output = output.decode("utf-8", 'ignore')
stderr = stderr.decode("utf-8", 'ignore')
info = json.loads(output)
if not info:
# If ffprobe didn't give any information, just return it
# (for example, because the file doesn't exist)
return info
extra_info = get_extra_info(stderr)
audio_streams = [x for x in info['streams'] if x['codec_type'] == 'audio']
if len(audio_streams) == 0:
return info
# We just operate on the first audio stream in case there are more
stream = audio_streams[0]
def set_property(stream, prop, value):
if prop not in stream or stream[prop] == 0:
stream[prop] = value
for token in extra_info[stream['index']]:
m = re.match('([su]([0-9]{1,2})p?) \(([0-9]{1,2}) bit\)$', token)
m2 = re.match('([su]([0-9]{1,2})p?)( \(default\))?$', token)
if m:
set_property(stream, 'sample_fmt', m.group(1))
set_property(stream, 'bits_per_sample', int(m.group(2)))
set_property(stream, 'bits_per_raw_sample', int(m.group(3)))
elif m2:
set_property(stream, 'sample_fmt', m2.group(1))
set_property(stream, 'bits_per_sample', int(m2.group(2)))
set_property(stream, 'bits_per_raw_sample', int(m2.group(2)))
elif re.match('(flt)p?( \(default\))?$', token):
set_property(stream, 'sample_fmt', token)
set_property(stream, 'bits_per_sample', 32)
set_property(stream, 'bits_per_raw_sample', 32)
elif re.match('(dbl)p?( \(default\))?$', token):
set_property(stream, 'sample_fmt', token)
set_property(stream, 'bits_per_sample', 64)
set_property(stream, 'bits_per_raw_sample', 64)
return info
def mediainfo(filepath):
"""Return dictionary with media info(codec, duration, size, bitrate...) from filepath
"""
prober = get_prober_name()
command_args = [
"-v", "quiet",
"-show_format",
"-show_streams",
filepath
]
command = [prober, '-of', 'old'] + command_args
res = Popen(command, stdout=PIPE)
output = res.communicate()[0].decode("utf-8")
if res.returncode != 0:
command = [prober] + command_args
output = Popen(command, stdout=PIPE).communicate()[0].decode("utf-8")
rgx = re.compile(r"(?:(?P<inner_dict>.*?):)?(?P<key>.*?)\=(?P<value>.*?)$")
info = {}
if sys.platform == 'win32':
output = output.replace("\r", "")
for line in output.split("\n"):
# print(line)
mobj = rgx.match(line)
if mobj:
# print(mobj.groups())
inner_dict, key, value = mobj.groups()
if inner_dict:
try:
info[inner_dict]
except KeyError:
info[inner_dict] = {}
info[inner_dict][key] = value
else:
info[key] = value
return info
def cache_codecs(function):
cache = {}
@wraps(function)
def wrapper():
try:
return cache[0]
except:
cache[0] = function()
return cache[0]
return wrapper
@cache_codecs
def get_supported_codecs():
encoder = get_encoder_name()
command = [encoder, "-codecs"]
res = Popen(command, stdout=PIPE, stderr=PIPE)
output = res.communicate()[0].decode("utf-8")
if res.returncode != 0:
return []
if sys.platform == 'win32':
output = output.replace("\r", "")
rgx = re.compile(r"^([D.][E.][AVS.][I.][L.][S.]) (\w*) +(.*)")
decoders = set()
encoders = set()
for line in output.split('\n'):
match = rgx.match(line.strip())
if not match:
continue
flags, codec, name = match.groups()
if flags[0] == 'D':
decoders.add(codec)
if flags[1] == 'E':
encoders.add(codec)
return (decoders, encoders)
def get_supported_decoders():
return get_supported_codecs()[0]
def get_supported_encoders():
return get_supported_codecs()[1]
def stereo_to_ms(audio_segment):
'''
Left-Right -> Mid-Side
'''
channel = audio_segment.split_to_mono()
channel = [channel[0].overlay(channel[1]), channel[0].overlay(channel[1].invert_phase())]
return AudioSegment.from_mono_audiosegments(channel[0], channel[1])
def ms_to_stereo(audio_segment):
'''
Mid-Side -> Left-Right
'''
channel = audio_segment.split_to_mono()
channel = [channel[0].overlay(channel[1]) - 3, channel[0].overlay(channel[1].invert_phase()) - 3]
return AudioSegment.from_mono_audiosegments(channel[0], channel[1])
venv/lib/python3.11/site-packages/speech_recognition/__init__.py
+1580
View File
File diff suppressed because it is too large Load Diff
venv/lib/python3.11/site-packages/speech_recognition/__main__.py
+24
View File
@@ -0,0 +1,24 @@
import speech_recognition as sr
r = sr.Recognizer()
m = sr.Microphone()
try:
print("A moment of silence, please...")
with m as source: r.adjust_for_ambient_noise(source)
print("Set minimum energy threshold to {}".format(r.energy_threshold))
while True:
print("Say something!")
with m as source: audio = r.listen(source)
print("Got it! Now to recognize it...")
try:
# recognize speech using Google Speech Recognition
value = r.recognize_google(audio)
print("You said {}".format(value))
except sr.UnknownValueError:
print("Oops! Didn't catch that")
except sr.RequestError as e:
print("Uh oh! Couldn't request results from Google Speech Recognition service; {0}".format(e))
except KeyboardInterrupt:
pass
venv/lib/python3.11/site-packages/speech_recognition/__pycache__/__init__.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/__pycache__/__main__.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/__pycache__/audio.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/__pycache__/exceptions.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/audio.py
+317
View File
@@ -0,0 +1,317 @@
import aifc
import audioop
import io
import os
import platform
import stat
import subprocess
import sys
import wave
class AudioData(object):
"""
Creates a new ``AudioData`` instance, which represents mono audio data.
The raw audio data is specified by ``frame_data``, which is a sequence of bytes representing audio samples. This is the frame data structure used by the PCM WAV format.
The width of each sample, in bytes, is specified by ``sample_width``. Each group of ``sample_width`` bytes represents a single audio sample.
The audio data is assumed to have a sample rate of ``sample_rate`` samples per second (Hertz).
Usually, instances of this class are obtained from ``recognizer_instance.record`` or ``recognizer_instance.listen``, or in the callback for ``recognizer_instance.listen_in_background``, rather than instantiating them directly.
"""
def __init__(self, frame_data, sample_rate, sample_width):
assert sample_rate > 0, "Sample rate must be a positive integer"
assert (
sample_width % 1 == 0 and 1 <= sample_width <= 4
), "Sample width must be between 1 and 4 inclusive"
self.frame_data = frame_data
self.sample_rate = sample_rate
self.sample_width = int(sample_width)
def get_segment(self, start_ms=None, end_ms=None):
"""
Returns a new ``AudioData`` instance, trimmed to a given time interval. In other words, an ``AudioData`` instance with the same audio data except starting at ``start_ms`` milliseconds in and ending ``end_ms`` milliseconds in.
If not specified, ``start_ms`` defaults to the beginning of the audio, and ``end_ms`` defaults to the end.
"""
assert (
start_ms is None or start_ms >= 0
), "``start_ms`` must be a non-negative number"
assert end_ms is None or end_ms >= (
0 if start_ms is None else start_ms
), "``end_ms`` must be a non-negative number greater or equal to ``start_ms``"
if start_ms is None:
start_byte = 0
else:
start_byte = int(
(start_ms * self.sample_rate * self.sample_width) // 1000
)
if end_ms is None:
end_byte = len(self.frame_data)
else:
end_byte = int(
(end_ms * self.sample_rate * self.sample_width) // 1000
)
return AudioData(
self.frame_data[start_byte:end_byte],
self.sample_rate,
self.sample_width,
)
def get_raw_data(self, convert_rate=None, convert_width=None):
"""
Returns a byte string representing the raw frame data for the audio represented by the ``AudioData`` instance.
If ``convert_rate`` is specified and the audio sample rate is not ``convert_rate`` Hz, the resulting audio is resampled to match.
If ``convert_width`` is specified and the audio samples are not ``convert_width`` bytes each, the resulting audio is converted to match.
Writing these bytes directly to a file results in a valid `RAW/PCM audio file <https://en.wikipedia.org/wiki/Raw_audio_format>`__.
"""
assert (
convert_rate is None or convert_rate > 0
), "Sample rate to convert to must be a positive integer"
assert convert_width is None or (
convert_width % 1 == 0 and 1 <= convert_width <= 4
), "Sample width to convert to must be between 1 and 4 inclusive"
raw_data = self.frame_data
# make sure unsigned 8-bit audio (which uses unsigned samples) is handled like higher sample width audio (which uses signed samples)
if self.sample_width == 1:
raw_data = audioop.bias(
raw_data, 1, -128
) # subtract 128 from every sample to make them act like signed samples
# resample audio at the desired rate if specified
if convert_rate is not None and self.sample_rate != convert_rate:
raw_data, _ = audioop.ratecv(
raw_data,
self.sample_width,
1,
self.sample_rate,
convert_rate,
None,
)
# convert samples to desired sample width if specified
if convert_width is not None and self.sample_width != convert_width:
if (
convert_width == 3
): # we're converting the audio into 24-bit (workaround for https://bugs.python.org/issue12866)
raw_data = audioop.lin2lin(
raw_data, self.sample_width, 4
) # convert audio into 32-bit first, which is always supported
try:
audioop.bias(
b"", 3, 0
) # test whether 24-bit audio is supported (for example, ``audioop`` in Python 3.3 and below don't support sample width 3, while Python 3.4+ do)
except (
audioop.error
): # this version of audioop doesn't support 24-bit audio (probably Python 3.3 or less)
raw_data = b"".join(
raw_data[i + 1 : i + 4]
for i in range(0, len(raw_data), 4)
) # since we're in little endian, we discard the first byte from each 32-bit sample to get a 24-bit sample
else: # 24-bit audio fully supported, we don't need to shim anything
raw_data = audioop.lin2lin(
raw_data, self.sample_width, convert_width
)
else:
raw_data = audioop.lin2lin(
raw_data, self.sample_width, convert_width
)
# if the output is 8-bit audio with unsigned samples, convert the samples we've been treating as signed to unsigned again
if convert_width == 1:
raw_data = audioop.bias(
raw_data, 1, 128
) # add 128 to every sample to make them act like unsigned samples again
return raw_data
def get_wav_data(self, convert_rate=None, convert_width=None):
"""
Returns a byte string representing the contents of a WAV file containing the audio represented by the ``AudioData`` instance.
If ``convert_width`` is specified and the audio samples are not ``convert_width`` bytes each, the resulting audio is converted to match.
If ``convert_rate`` is specified and the audio sample rate is not ``convert_rate`` Hz, the resulting audio is resampled to match.
Writing these bytes directly to a file results in a valid `WAV file <https://en.wikipedia.org/wiki/WAV>`__.
"""
raw_data = self.get_raw_data(convert_rate, convert_width)
sample_rate = (
self.sample_rate if convert_rate is None else convert_rate
)
sample_width = (
self.sample_width if convert_width is None else convert_width
)
# generate the WAV file contents
with io.BytesIO() as wav_file:
wav_writer = wave.open(wav_file, "wb")
try: # note that we can't use context manager, since that was only added in Python 3.4
wav_writer.setframerate(sample_rate)
wav_writer.setsampwidth(sample_width)
wav_writer.setnchannels(1)
wav_writer.writeframes(raw_data)
wav_data = wav_file.getvalue()
finally: # make sure resources are cleaned up
wav_writer.close()
return wav_data
def get_aiff_data(self, convert_rate=None, convert_width=None):
"""
Returns a byte string representing the contents of an AIFF-C file containing the audio represented by the ``AudioData`` instance.
If ``convert_width`` is specified and the audio samples are not ``convert_width`` bytes each, the resulting audio is converted to match.
If ``convert_rate`` is specified and the audio sample rate is not ``convert_rate`` Hz, the resulting audio is resampled to match.
Writing these bytes directly to a file results in a valid `AIFF-C file <https://en.wikipedia.org/wiki/Audio_Interchange_File_Format>`__.
"""
raw_data = self.get_raw_data(convert_rate, convert_width)
sample_rate = (
self.sample_rate if convert_rate is None else convert_rate
)
sample_width = (
self.sample_width if convert_width is None else convert_width
)
# the AIFF format is big-endian, so we need to convert the little-endian raw data to big-endian
if hasattr(
audioop, "byteswap"
): # ``audioop.byteswap`` was only added in Python 3.4
raw_data = audioop.byteswap(raw_data, sample_width)
else: # manually reverse the bytes of each sample, which is slower but works well enough as a fallback
raw_data = raw_data[sample_width - 1 :: -1] + b"".join(
raw_data[i + sample_width : i : -1]
for i in range(sample_width - 1, len(raw_data), sample_width)
)
# generate the AIFF-C file contents
with io.BytesIO() as aiff_file:
aiff_writer = aifc.open(aiff_file, "wb")
try: # note that we can't use context manager, since that was only added in Python 3.4
aiff_writer.setframerate(sample_rate)
aiff_writer.setsampwidth(sample_width)
aiff_writer.setnchannels(1)
aiff_writer.writeframes(raw_data)
aiff_data = aiff_file.getvalue()
finally: # make sure resources are cleaned up
aiff_writer.close()
return aiff_data
def get_flac_data(self, convert_rate=None, convert_width=None):
"""
Returns a byte string representing the contents of a FLAC file containing the audio represented by the ``AudioData`` instance.
Note that 32-bit FLAC is not supported. If the audio data is 32-bit and ``convert_width`` is not specified, then the resulting FLAC will be a 24-bit FLAC.
If ``convert_rate`` is specified and the audio sample rate is not ``convert_rate`` Hz, the resulting audio is resampled to match.
If ``convert_width`` is specified and the audio samples are not ``convert_width`` bytes each, the resulting audio is converted to match.
Writing these bytes directly to a file results in a valid `FLAC file <https://en.wikipedia.org/wiki/FLAC>`__.
"""
assert convert_width is None or (
convert_width % 1 == 0 and 1 <= convert_width <= 3
), "Sample width to convert to must be between 1 and 3 inclusive"
if (
self.sample_width > 3 and convert_width is None
): # resulting WAV data would be 32-bit, which is not convertable to FLAC using our encoder
convert_width = 3 # the largest supported sample width is 24-bit, so we'll limit the sample width to that
# run the FLAC converter with the WAV data to get the FLAC data
wav_data = self.get_wav_data(convert_rate, convert_width)
flac_converter = get_flac_converter()
if (
os.name == "nt"
): # on Windows, specify that the process is to be started without showing a console window
startup_info = subprocess.STARTUPINFO()
startup_info.dwFlags |= (
subprocess.STARTF_USESHOWWINDOW
) # specify that the wShowWindow field of `startup_info` contains a value
startup_info.wShowWindow = (
subprocess.SW_HIDE
) # specify that the console window should be hidden
else:
startup_info = None # default startupinfo
process = subprocess.Popen(
[
flac_converter,
"--stdout",
"--totally-silent", # put the resulting FLAC file in stdout, and make sure it's not mixed with any program output
"--best", # highest level of compression available
"-", # the input FLAC file contents will be given in stdin
],
stdin=subprocess.PIPE,
stdout=subprocess.PIPE,
startupinfo=startup_info,
)
flac_data, stderr = process.communicate(wav_data)
return flac_data
def get_flac_converter():
"""Returns the absolute path of a FLAC converter executable, or raises an OSError if none can be found."""
flac_converter = shutil_which("flac") # check for installed version first
if flac_converter is None: # flac utility is not installed
base_path = os.path.dirname(
os.path.abspath(__file__)
) # directory of the current module file, where all the FLAC bundled binaries are stored
system, machine = platform.system(), platform.machine()
if system == "Windows" and machine in {
"i686",
"i786",
"x86",
"x86_64",
"AMD64",
}:
flac_converter = os.path.join(base_path, "flac-win32.exe")
elif system == "Darwin" and machine in {
"i686",
"i786",
"x86",
"x86_64",
"AMD64",
}:
flac_converter = os.path.join(base_path, "flac-mac")
elif system == "Linux" and machine in {"i686", "i786", "x86"}:
flac_converter = os.path.join(base_path, "flac-linux-x86")
elif system == "Linux" and machine in {"x86_64", "AMD64"}:
flac_converter = os.path.join(base_path, "flac-linux-x86_64")
else: # no FLAC converter available
raise OSError(
"FLAC conversion utility not available - consider installing the FLAC command line application by running `apt-get install flac` or your operating system's equivalent"
)
# mark FLAC converter as executable if possible
try:
# handle known issue when running on docker:
# run executable right after chmod() may result in OSError "Text file busy"
# fix: flush FS with sync
if not os.access(flac_converter, os.X_OK):
stat_info = os.stat(flac_converter)
os.chmod(flac_converter, stat_info.st_mode | stat.S_IEXEC)
if "Linux" in platform.system():
os.sync() if sys.version_info >= (3, 3) else os.system("sync")
except OSError:
pass
return flac_converter
def shutil_which(pgm):
"""Python 2 compatibility: backport of ``shutil.which()`` from Python 3"""
path = os.getenv("PATH")
for p in path.split(os.path.pathsep):
p = os.path.join(p, pgm)
if os.path.exists(p) and os.access(p, os.X_OK):
return p
venv/lib/python3.11/site-packages/speech_recognition/exceptions.py
+22
View File
@@ -0,0 +1,22 @@
class SetupError(Exception):
pass
class WaitTimeoutError(Exception):
pass
class RequestError(Exception):
pass
class UnknownValueError(Exception):
pass
class TranscriptionNotReady(Exception):
pass
class TranscriptionFailed(Exception):
pass
venv/lib/python3.11/site-packages/speech_recognition/flac-linux-x86
Executable
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/flac-linux-x86_64
Executable
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/flac-mac
Executable
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/flac-win32.exe
Executable
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/LICENSE.txt
+31
View File
@@ -0,0 +1,31 @@
Copyright (c) 1999-2015 Carnegie Mellon University. All rights
reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the
distribution.
This work was supported in part by funding from the Defense Advanced
Research Projects Agency and the National Science Foundation of the
United States of America, and the CMU Sphinx Speech Consortium.
THIS SOFTWARE IS PROVIDED BY CARNEGIE MELLON UNIVERSITY ``AS IS'' AND
ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL CARNEGIE MELLON UNIVERSITY
NOR ITS EMPLOYEES BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/acoustic-model/README
+34
View File
@@ -0,0 +1,34 @@
/* ====================================================================
* Copyright (c) 2015 Alpha Cephei Inc. All rights
* reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
*
* THIS SOFTWARE IS PROVIDED BY ALPHA CEPHEI INC. ``AS IS'' AND.
* ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,.
* THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL ALPHA CEPHEI INC.
* NOR ITS EMPLOYEES BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT.
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,.
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY.
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT.
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE.
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* ====================================================================
*
*/
This directory contains generic US english acoustic model trained with
latest sphinxtrain.
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/acoustic-model/feat.params
+12
View File
@@ -0,0 +1,12 @@
-lowerf 130
-upperf 6800
-nfilt 25
-transform dct
-lifter 22
-feat 1s_c_d_dd
-svspec 0-12/13-25/26-38
-agc none
-cmn current
-varnorm no
-model ptm
-cmninit 40,3,-1
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/acoustic-model/mdef
BIN
View File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 2.8 MiB

venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/acoustic-model/means
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/acoustic-model/noisedict
+5
View File
@@ -0,0 +1,5 @@
<s> SIL
</s> SIL
<sil> SIL
[NOISE] +NSN+
[SPEECH] +SPN+
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/acoustic-model/sendump
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/acoustic-model/transition_matrices
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/acoustic-model/variances
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/language-model.lm.bin
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/pocketsphinx-data/en-US/pronounciation-dictionary.dict
+133425
View File
File diff suppressed because it is too large Load Diff
venv/lib/python3.11/site-packages/speech_recognition/recognizers/__init__.py
View File
venv/lib/python3.11/site-packages/speech_recognition/recognizers/__pycache__/__init__.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/recognizers/__pycache__/whisper.cpython-311.pyc
BIN
View File
Binary file not shown.
venv/lib/python3.11/site-packages/speech_recognition/recognizers/whisper.py
+42
View File
@@ -0,0 +1,42 @@
from __future__ import annotations
import os
from io import BytesIO
from speech_recognition.audio import AudioData
from speech_recognition.exceptions import SetupError
def recognize_whisper_api(
recognizer,
audio_data: "AudioData",
*,
model: str = "whisper-1",
api_key: str | None = None,
):
"""
Performs speech recognition on ``audio_data`` (an ``AudioData`` instance), using the OpenAI Whisper API.
This function requires an OpenAI account; visit https://platform.openai.com/signup, then generate API Key in `User settings <https://platform.openai.com/account/api-keys>`__.
Detail: https://platform.openai.com/docs/guides/speech-to-text
Raises a ``speech_recognition.exceptions.SetupError`` exception if there are any issues with the openai installation, or the environment variable is missing.
"""
if not isinstance(audio_data, AudioData):
raise ValueError("``audio_data`` must be an ``AudioData`` instance")
if api_key is None and os.environ.get("OPENAI_API_KEY") is None:
raise SetupError("Set environment variable ``OPENAI_API_KEY``")
try:
import openai
except ImportError:
raise SetupError(
"missing openai module: ensure that openai is set up correctly."
)
wav_data = BytesIO(audio_data.get_wav_data())
wav_data.name = "SpeechRecognition_audio.wav"
transcript = openai.Audio.transcribe(model, wav_data, api_key=api_key)
return transcript["text"]