def api_phonemes():
profile = request_to_profile(request, profiles_dirs)
tts_config = profile.text_to_speech
examples_path = profile.read_path(tts_config['phoneme_examples'])
# phoneme -> { word, phonemes }
logging.debug('Loading phoneme examples from %s' % examples_path)
examples_dict = utils.load_phoneme_examples(examples_path)
return jsonify(examples_dict)
def api_start_recording():
global recorder
device_index = int(request.args.get('device', -1))
if device_index < 0:
device_index = None # default device
profile = request_to_profile(request, profiles_dirs)
recorder = get_audio_recorder(profile)
recorder.start_recording(device_index)
return 'OK'
def api_train():
profile = request_to_profile(request, profiles_dirs)
start_time = time.time()
logging.info('Starting training')
train(profile)
end_time = time.time()
reload(profile)
return 'Training completed in %0.2f second(s)' % (end_time - start_time)
def api_pronounce():
profile = request_to_profile(request, profiles_dirs)
espeak_config = profile.text_to_speech['espeak']
download = request.args.get('download', 'false').lower() == 'true'
speed = int(request.args.get('speed', 80))
voice = request.args.get('voice', None)
pronounce_str = request.data.decode('utf-8').strip()
assert len(pronounce_str) > 0, 'No string to pronounce'
pronounce_type = request.args.get('type', 'phonemes')
if pronounce_type == 'phonemes':
# Load map from Sphinx to eSpeak phonemes
map_path = profile.read_path(espeak_config['phoneme_map'])
phoneme_map = utils.load_phoneme_map(map_path)
# Convert from Sphinx to espeak phonemes
espeak_str = "[['%s]]" % ''.join(
phoneme_map.get(p, p) for p in pronounce_str.split())
else:
# Speak word directly
espeak_str = pronounce_str
# Generate WAV data
espeak_command = ['espeak', '-s', str(speed)]
if voice is None:
if 'voice' in espeak_config:
# Use profile voice
voice = espeak_config['voice']
elif 'language' in profile.json:
# Use language default voice
voice = profile.json['language']
if voice is not None:
espeak_command.extend(['-v', str(voice)])
espeak_command.append(espeak_str)
with tempfile.NamedTemporaryFile(suffix='.wav', mode='wb+') as wav_file:
espeak_command.extend(['-w', wav_file.name])
logging.debug(espeak_command)
# Generate WAV data
subprocess.check_call(espeak_command)
wav_file.seek(0)
if download:
return Response(wav_file.read(), mimetype='audio/wav')
else:
subprocess.check_call(['aplay', '-t', 'wav', wav_file.name])
return espeak_str
def api_profile():
layers = request.args.get('layers', 'all')
if request.method == 'POST':
if layers == 'default':
for profiles_dir in profiles_dirs:
profile_path = os.path.join(profiles_dir, 'defaults.json')
if os.path.exists(profile_path):
with open(profile_path, 'wb') as profile_file:
profile_file.write(request.data)
else:
profile = request_to_profile(request, profiles_dirs)
profile_path = profile.write_path('profile.json')
with open(profile_path, 'wb') as profile_file:
profile_file.write(request.data)
return 'Wrote %d byte(s) to %s' % (len(request.data), profile_path)
profile = request_to_profile(request, profiles_dirs, layers)
return jsonify(profile.json)
def api_speech_to_text():
global decoders
profile = request_to_profile(request, profiles_dirs)
wav_data = request.data
decoder = transcribe_wav(profile, wav_data, decoders.get(profile.name))
decoders[profile.name] = decoder
if decoder.hyp() is not None:
return decoder.hyp().hypstr
return ''
def api_unknown_words():
profile = request_to_profile(request, profiles_dirs)
ps_config = profile.speech_to_text['pocketsphinx']
unknown_words = {}
unknown_path = profile.read_path(ps_config['unknown_words'])
if os.path.exists(unknown_path):
for line in open(unknown_path, 'r'):
line = line.strip()
if len(line) > 0:
word, pronunciation = re.split(r'\s+', line, maxsplit=1)
unknown_words[word] = pronunciation
return jsonify(unknown_words)
def api_text_to_intent():
profile = request_to_profile(request, profiles_dirs)
text = request.data.decode()
no_hass = request.args.get('nohass', 'false').lower() == 'true'
start_time = time.time()
intent = get_intent(profile, text)
intent_sec = time.time() - start_time
intent['time_sec'] = intent_sec
if not no_hass:
# Send intent to Home Assistant
utils.send_intent(profile.home_assistant, intent)
return jsonify(intent)
def api_custom_words():
profile = request_to_profile(request, profiles_dirs)
stt_config = profile.speech_to_text
ps_config = stt_config['pocketsphinx']
custom_words_path = profile.write_path(ps_config['custom_words'])
if request.method == 'POST':
# Update custom words
with open(custom_words_path, 'wb') as custom_words_file:
custom_words_file.write(request.data)
return 'Wrote %s byte(s) to %s' % (len(
request.data), custom_words_path)
# Return custom_words
if not os.path.exists(custom_words_path):
return '' # no custom_words yet
# Return file contents
return send_file(open(custom_words_path, 'rb'), mimetype='text/plain')
def api_sentences():
profile = request_to_profile(request, profiles_dirs)
stt_config = profile.speech_to_text
if request.method == 'POST':
# Update sentences
sentences_path = profile.write_path(stt_config['sentences_ini'])
with open(sentences_path, 'wb') as sentences_file:
sentences_file.write(request.data)
return 'Wrote %s byte(s) to %s' % (len(
request.data), sentences_path)
# Return sentences
sentences_path = profile.read_path(stt_config['sentences_ini'])
if not os.path.exists(sentences_path):
return '' # no sentences yet
# Return file contents
return send_file(open(sentences_path, 'rb'), mimetype='text/plain')
def api_lookup():
n = int(request.args.get('n', 5))
assert n > 0, 'No pronunciations requested'
voice = request.args.get('voice', None)
profile = request_to_profile(request, profiles_dirs)
ps_config = profile.speech_to_text['pocketsphinx']
espeak_config = profile.text_to_speech['espeak']
word = request.data.decode('utf-8').strip().lower()
assert len(word) > 0, 'No word to look up'
logging.debug('Getting pronunciations for %s' % word)
# Load base and custom dictionaries
base_dictionary_path = profile.read_path(ps_config['base_dictionary'])
custom_path = profile.read_path(ps_config['custom_words'])
word_dict = {}
for word_dict_path in [base_dictionary_path, custom_path]:
if os.path.exists(word_dict_path):
with open(word_dict_path, 'r') as dictionary_file:
utils.read_dict(dictionary_file, word_dict)
result = utils.lookup_word(word, word_dict, profile, n=n)
# Get phonemes from eSpeak
espeak_command = ['espeak', '-q', '-x']
if voice is None:
if 'voice' in espeak_config:
# Use profile voice
voice = espeak_config['voice']
elif 'language' in profile.json:
# Use language default voice
voice = profile.json['language']
espeak_command.extend(['-v', voice, word])
logging.debug(espeak_command)
result['espeak_phonemes'] = subprocess.check_output(
espeak_command).decode()
return jsonify(result)
def api_speech_to_intent():
profile = request_to_profile(request, profiles_dirs)
wav_data = request.data
no_hass = request.args.get('nohass', 'false').lower() == 'true'
start_time = time.time()
intent = speech_to_intent(profile, wav_data, no_hass)
intent_sec = time.time() - start_time
intent['time_sec'] = intent_sec
if intent is not None:
return jsonify(intent)
return jsonify({
'text': '',
'intent': '',
'entities': [],
'time_sec': intent_sec
})
def api_stop_recording():
global decoders, recorder
no_hass = request.args.get('nohass', 'false').lower() == 'true'
if recorder is not None:
record_buffer = recorder.stop_recording()
logging.debug('Stopped recording (got %s byte(s))' %
len(record_buffer))
profile = request_to_profile(request, profiles_dirs)
# Convert to WAV
with io.BytesIO() as wav_data:
with wave.open(wav_data, mode='wb') as wav_file:
wav_file.setframerate(16000)
wav_file.setsampwidth(2)
wav_file.setnchannels(1)
wav_file.writeframesraw(record_buffer)
wav_data.seek(0)
record_buffer = None
# speech to text
decoder = transcribe_wav(profile, wav_data.read(),
decoders.get(profile.name))
decoders[profile.name] = decoder
# text to intent
if decoder.hyp() is not None:
text = decoder.hyp().hypstr
logging.info(text)
intent = get_intent(profile, text)
if not no_hass:
# Send intent to Home Assistant
utils.send_intent(profile.home_assistant, intent)
return jsonify(intent)
return jsonify({'text': '', 'intent': '', 'entities': []})
def api_listen_for_wake():
profile = request_to_profile(request, profiles_dirs)
no_hass = request.args.get('nohass', 'false').lower() == 'true'
return jsonify(listen_for_wake(profile, no_hass))
def api_microphones():
profile = request_to_profile(request, profiles_dirs)
mics = get_audio_recorder(profile).get_microphones()
return jsonify(mics)
def api_reload():
profile = request_to_profile(request, profiles_dirs)
reload(profile)
return 'Reloaded profile "%s"' % profile.name