def active_listen():
"""
Actively listens for speech to translate into text
:return: speech input as a text string
"""
global r
# use the default microphone as the audio source
with tts.ignore_stderr():
with speech_recognition.Microphone() as src:
# listen for 1 second to adjust energy threshold for ambient noise
# r.adjust_for_ambient_noise(src)
log.info("Active listening... ")
tts.play_mp3('double-beep.mp3') # <-- Windows Multithread error
# listen for the first phrase and extract it into audio data
audio = r.listen(src)
msg = ''
try:
msg = r.recognize_google(audio) # recognize speech using Google STT
except speech_recognition.UnknownValueError:
print("Google Speech Recognition could not understand audio")
except speech_recognition.RequestError as e:
print("Could not request results from Google STT; {0}".format(e))
except:
print("Unknown exception occurred!")
finally:
return msg
def action(self, text):
mod_name = self.module.lower().strip().replace(' ', '')
if 'disable' in self.enable.lower() or 'remove' in self.enable.lower():
log.info("Attempting to disable '"+mod_name+"'")
brain.inst.disable_mod(mod_name)
else:
log.info("Attempting to enable '"+mod_name+"'")
brain.inst.enable_mod(mod_name)
def mod_select(self, mods):
""" Prompt user to specify which module to use to respond """
print('\n~ Which module (greedy) would you like me to use to respond?')
print('~ Choices: ' + str([mod.name for mod in mods])[1:-1] + '\n')
mod_select = input('> ')
for mod in mods:
if re.search('^.*\\b' + mod.name + '\\b.*$', mod_select,
re.IGNORECASE):
return mod
log.info('No module name found.')
def listen_keyword():
""" Passively listens for the WAKE_UP_WORD string """
with tts.ignore_stderr():
global decoder, p
stream = p.open(format=pyaudio.paInt16, channels=1, rate=16000,
input=True, frames_per_buffer=1024)
stream.start_stream()
p.get_default_input_device_info()
log.info("Waiting to be woken up... ")
decoder.start_utt()
while True:
buf = stream.read(1024)
decoder.process_raw(buf, False, False)
if decoder.hyp() and decoder.hyp().hypstr == settings.WAKE_UP_WORD:
decoder.end_utt()
return
decoder.end_utt()
def open(self, url=None, new_tab=False):
if not self.driver:
try:
# print(settings.CHROME_DRIVER)
# print(os.path.isfile(settings.CHROME_DRIVER))
if not os.path.isfile(settings.CHROME_DRIVER):
raise Exception
self.driver = webdriver.Chrome(settings.CHROME_DRIVER)
except:
print(traceback.format_exc())
self.driver = webdriver.Firefox()
else:
if new_tab:
log.info('Opening new tab...')
self.driver.find_element_by_tag_name('body').send_keys(NW_TAB)
if url:
if not url[0:4] == 'http':
url = 'https://' + url.replace(' ', '')
self.driver.get(url)
def run(self):
""" Listen for input, match the modules and respond """
while True:
if self.quit_flag:
break
try:
if settings.USE_STT:
stt.listen_keyword()
text = stt.active_listen()
else:
text = input('> ')
if not text:
log.info('No text input received.')
continue
self.match_mods(text)
self.execute_mods(text)
except OSError as e:
if 'Invalid input device' in str(e):
log.error(settings.NO_MIC + '\n')
settings.USE_STT = False
continue
else:
raise Exception
except (EOFError, KeyboardInterrupt):
log.info('Shutting down...')
break
except:
log.error("(runtime error)")
self.error()
log.info('Arrivederci.')
def speak(phrase,
cache=False,
filename='default',
show_text=True,
log_text=True): # cache is normally false
"""Speaks a given text phrase
:param phrase: text string to speak
:param cache: if True, store .mp3 in 'media/responses'
:param filename: filename if cached
:param show_text: if True, store .mp3 in 'media/responses'
:param cache: if True, store .mp3 in 'media/responses'
"""
if show_text:
log.info(phrase)
if not settings.USE_TTS:
log.info('SPOKEN: ' + phrase)
return
try:
phrase = phrase[:settings.MAX_CHAR]
tts = gTTS(text=phrase, lang=settings.LANG)
if not cache:
with tempfile.NamedTemporaryFile(mode='wb',
suffix='.mp3',
delete=False) as f:
(temp_path, temp_name) = os.path.split(f.name)
tts.write_to_fp(f)
play_mp3(temp_name, temp_path)
os.remove(os.path.join(temp_path, temp_name))
else:
filename = os.path.join(settings.RESPONSES_DIR, filename + '.mp3')
tts.save(filename)
log.info('Saved to: ' + filename)
except HTTPError as e:
log.error('Google TTS might not be updated: ' + str(e))
except Exception as e:
log.error('Unknown Google TTS issue: ' + str(e))
def disable_mod(self, name):
for mod in mods.mod_lib:
if mod.name == name:
mod.enabled = False
log.info(mod.name + " disabled.")
def enable_mod(self, name):
for mod in mods.mod_lib:
if mod.name == name:
mod.enabled = True
log.info(mod.name + " enabled.")
def search(self, q):
log.info('Answering with Google...')
self.open(GOOGLE_URL + quote_plus(q), new_tab=False)
def action(self, text):
log.info('\n~ Searching Google...\n')
api_lib['voice_browse_api'].search(text)
def enable_mod(name):
global mod_lib
for mod in mod_lib:
if name in mod.name:
log.info('Enabling: ' + name + '\n')
mod.enabled = True
def disable_mod(name):
global mod_lib
for mod in mod_lib:
if name in mod.name:
log.info('Disabling: ' + name + '\n')
mod.enabled = False
def list_mods():
""" Print modules in order """
global mod_lib
log.info('Module Order: ' + str([mod.name
for mod in mod_lib])[1:-1] + '\n')
