def dhvanify(content):
translator = Translator(os.environ['BING_SPEECH_ACCESS_KEY'])
audioOutputs = []
finalText = []
for idx, val in enumerate(content.newsItems):
val.body = ("Story #%s: " %(idx)) + val.body + " "
finalText.append(val.body)
for idx, text in enumerate(finalText):
output = translator.speak(text, "en-US", "Female", "riff-16khz-16bit-mono-pcm")
time.sleep(1)
with open("tmp/file%s.wav"%(idx), "wb") as f:
f.write(output)
audioOutputs.append(AudioSegment.from_wav("tmp/file%s.wav"%(idx)))
for idx, snippet in enumerate(audioOutputs):
if(idx == 0):
finalAudio = snippet
else:
finalAudio = finalAudio + snippet
finalAudio.export("tmp/final.wav", format="wav")
return "tmp/final.wav"
def tts(text="Hi there"):
translator = Translator(os.getenv('BING_KEY', ''))
output = translator.speak(text, "en-US", "JessaRUS", "riff-16khz-16bit-mono-pcm")
with open('output.wav', 'wb') as f:
f.write(output)
CHUNK = 1024
wf = wave.open('output.wav', 'rb')
# instantiate PyAudio (1)
p = pyaudio.PyAudio()
# open stream (2)
stream = p.open(format=p.get_format_from_width(wf.getsampwidth()),
channels=wf.getnchannels(),
rate=wf.getframerate(),
output=True)
# read data
data = wf.readframes(CHUNK)
# play stream (3)
while len(data) > 0:
stream.write(data)
data = wf.readframes(CHUNK)
# stop stream (4)
stream.stop_stream()
stream.close()
# close PyAudio (5)
p.terminate()
def __init__(self, lang, config):
super(BingTTS, self).__init__(lang, config, BingTTSValidator(self))
self.type = 'wav'
from bingtts import Translator
api = self.config.get("api_key")
self.bing = Translator(api)
self.gender = self.config.get("gender", "Male")
self.format = self.config.get("format", "riff-16khz-16bit-mono-pcm")
def __init__(self, lang, voice):
super(BingTTS, self).__init__(lang, voice, BingTTSValidator(self))
self.type = 'wav'
from bingtts import Translator
self.config = Configuration.get().get("tts", {}).get("bing", {})
api = self.config.get("api_key")
self.bing = Translator(api)
self.gender = self.config.get("gender", "Male")
self.format = self.config.get("format", "riff-16khz-16bit-mono-pcm")
def renderResult(detected):
objectDetected = detected['description']['captions'][0]['text']
print(objectDetected)
translator = Translator('[tts key]')
output = translator.speak(objectDetected, "en-US", "Female",
"riff-16khz-16bit-mono-pcm")
with open("file.wav", "w") as f:
f.write(output)
i = 15
time.sleep(0.3)
def main():
# recognized_text = recognize()
# print(recognized_text)
# bing = BingSpeechAPI()
# bing.text_to_speech(text='Can I have some coffee?')
from bingtts import Translator
translator = Translator(os.getenv('BING_KEY', ''))
output = translator.speak("This is a text to speech translation", "en-US",
"Female", "riff-16khz-16bit-mono-pcm")
with open("file.wav", "w") as f:
f.write(output)
def synthesize_speech(recipe_id, steps_json, ing_pos):
# TODO: Do unhardcode path, it is relative to calling PWD
wav_root = "./website/static/wav/"
if not os.path.exists(wav_root + recipe_id):
tts = Translator('50dbceaa8a424c1c89ea9c33d3a4eec0')
speech_text = [step_raw['step'] for step_raw in steps_json]
# TODO: optimize speech
speech_bin = [
tts.speak(step_raw, "en-US", "Female", "riff-16khz-16bit-mono-pcm")
for step_raw in speech_text
]
ing_bin = [
tts.speak(ing_raw, "en-US", "Female", "riff-16khz-16bit-mono-pcm")
for ing_raw in ing_pos
]
# create the recipe ID holding the recipe wave files
try:
print('creating folder ' + wav_root + recipe_id)
os.makedirs(wav_root + recipe_id)
except OSError as e:
if e.errno != errno.EEXIST:
raise
wav_list = []
for ing_num in xrange(len(ing_pos)):
fname = 'ing_%d' % ing_num
wav_list.append('%s' % fname)
with open("%s%s/%s.wav" % (wav_root, recipe_id, fname), 'w') as f:
f.write(ing_bin[ing_num])
for step in xrange(len(speech_text)):
fname = 'step_%d' % step
wav_list.append('%s' % fname)
with open("%s%s/%s.wav" % (wav_root, recipe_id, fname), 'w') as f:
f.write(speech_bin[step])
return wav_list
else:
return [
file[:-4] for file in os.listdir('%s%s' % (wav_root, recipe_id))
if file.endswith(".wav")
]
def __init__(self, lang, config):
super(BingTTS, self).__init__(lang, config, BingTTSValidator(self))
self.type = 'wav'
from bingtts import Translator
self.config = Configuration.get().get("tts", {}).get("bing", {})
api = self.config.get("api_key")
self.bing = Translator(api)
self.gender = self.config.get("gender", "Male")
self.format = self.config.get("format", "riff-16khz-16bit-mono-pcm")
def tts(text):
translator = Translator('b2efca95d4b74c3ca0d1e5b0729963fe')
output = translator.speak(text, "en-US", "Male",
"riff-16khz-16bit-mono-pcm")
# print(type(output))
with open("file.wav", "w") as f:
f.write(output)
chunk = 1024
wf = wave.open(r"file.wav", 'rb')
p = pyaudio.PyAudio()
stream = p.open(format=p.get_format_from_width(wf.getsampwidth()),
channels=wf.getnchannels(),
rate=wf.getframerate(),
output=True)
while True:
data = wf.readframes(chunk)
if data == "": break
stream.write(data)
stream.close()
p.terminate()
def synthesize_speech_one_item(input_string, input_language, voice,
output_name, output_dir):
from bingtts import Translator
#translator = Translator(BING_SUBSCRIPTION_KEY)
bing_key = BING_SUBSCRIPTION_KEY
# bing_key = keyring.get_password("bing_api", "obi")
if bing_key == None:
print("Please set BING_SUBSCRIPTION_KEY")
return
translator = Translator(bing_key)
output = translator.speak(input_string, input_language, voice,
"audio-16khz-64kbitrate-mono-mp3 ")
output_filename = output_name.strip().replace(" ", "_")
output_filename = '{}/{}_{}.mp3'.format(output_dir, output_filename,
input_language)
with open(output_filename, "wb") as f:
f.write(output)
print('Audio content written to file "{}"'.format(output_filename))
class BingTTS(TTS):
def __init__(self, lang, config):
super(BingTTS, self).__init__(lang, config, BingTTSValidator(self))
self.type = 'wav'
from bingtts import Translator
api = self.config.get("api_key")
self.bing = Translator(api)
self.gender = self.config.get("gender", "Male")
self.format = self.config.get("format", "riff-16khz-16bit-mono-pcm")
def get_tts(self, sentence, wav_file):
output = self.bing.speak(sentence, self.lang, self.gender, self.format)
with open(wav_file, "w") as f:
f.write(output)
return (wav_file, None) # No phonemes
class BingTTS(TTS):
def __init__(self, lang, config):
super(BingTTS, self).__init__(lang, config, BingTTSValidator(self))
self.type = 'wav'
from bingtts import Translator
self.config = Configuration.get().get("tts", {}).get("bing", {})
api = self.config.get("api_key")
self.bing = Translator(api)
self.gender = self.config.get("gender", "Male")
self.format = self.config.get("format", "riff-16khz-16bit-mono-pcm")
def get_tts(self, sentence, wav_file):
output = self.bing.speak(sentence, self.lang, self.gender,
self.format)
with open(wav_file, "w") as f:
f.write(output)
return (wav_file, None) # No phonemes
class TextToSpeech():
def __init__(self, bingTtsKey):
self.translator = Translator(bingTtsKey)
self.ttsAudio = {}
def playAudio(self, audio):
CHUNK = 1024
f = io.BytesIO()
f.write(audio)
f.seek(0)
wf = wave.Wave_read(f)
p = pyaudio.PyAudio()
stream = p.open(format=p.get_format_from_width(wf.getsampwidth()),
channels=wf.getnchannels(),
rate=wf.getframerate(),
output=True)
data = wf.readframes(CHUNK)
while data != b'':
stream.write(data)
data = wf.readframes(CHUNK)
stream.stop_stream()
stream.close()
p.terminate()
def Text2Speech(self, text):
text = text.lower()
audio = self.ttsAudio.get(text)
if audio == None:
print('audio not found')
audio = self.translator.speak(
text, "en-AU", "Catherine", "riff-16khz-16bit-mono-pcm")
self.ttsAudio[text] = audio
self.playAudio(audio)
def __init__(self, bingTtsKey):
self.translator = Translator(bingTtsKey)
self.ttsAudio = {}
from telebot import TeleBot
from bingtts import Translator
telegram_token = '<TELEGRAM_TOKEN>'
azure_token = '<AZURE_TOKEN>'
telebot = TeleBot(telegram_token)
translator = Translator(azure_token)
@telebot.message_handler(content_types=['text'])
def handle_speech(message):
user_id = message.chat.id
text = message.text
speech = translator.speak(text, "en-US", "BenjaminRUS",
"audio-16khz-32kbitrate-mono-mp3")
res = telebot.send_voice(user_id, speech)
if __name__ == '__main__':
telebot.polling(none_stop=True)
# create filename base on MD5 hash
filename = "{}.wav".format(m.hexdigest())
if ttsCache:
# If our file already exists, just return it so we don't have to do an API call...
if os.path.isfile(os.path.join(ttsCache, filename)):
try:
shutil.copy2(os.path.join(ttsCache, filename), destLoc)
sys.exit(0)
except (Exception) as e:
print("Could not copy cached file: {}".format(e))
print("Exiting...")
sys.exit(1)
# Wait to import this until after we check cache to keep things as speedy as possible...
from bingtts import Translator
translator = Translator(ttsApiKey)
try:
data = translator.speak(
ttsText.encode('utf-8').decode('latin-1'), ttsLang, ttsVoice,
ttsFormat)
except (Exception) as e:
print("Error retrieving speech file: {}".format(e))
sys.exit(1)
if ttsCache:
try:
with open(destLoc, 'wb') as f:
f.write(data)
except (Exception) as e:
print("Couldn't write to file {}: {}".format(destLoc, e))
sys.exit(1)
# create filename base on MD5 hash
filename = str(m.hexdigest()) + ".wav"
if ttsCache:
# If our file already exists, just return it so we don't have to do an API call...
if os.path.isfile(ttsCache + '/' + filename):
try:
shutil.copy2(ttsCache + '/' + filename, destLoc)
sys.exit(0)
except (Exception) as e:
print "Could not copy cached file:", e
print "Exiting..."
sys.exit(1)
# Wait to import this until after we check cache to keep things as speedy as possible...
from bingtts import Translator
translator = Translator(ttsApiKey)
try:
data = translator.speak(ttsText, ttsLang, ttsGender, ttsFormat)
except (Exception) as e:
print "Error retrieving speech file:", e
sys.exit(1)
if ttsCache:
try:
with open(destLoc, 'w') as f:
f.write(data)
except (Exception) as e:
print "Couldn't write to file", destLoc, ":", e
sys.exit(1)
try:
with open(ttsCache + '/' + filename, 'w') as f:
