ExternalVendorCode/ecoute
1
0
Fork 0
mirror of https://github.com/SevaSk/ecoute.git synced 2024-12-20 05:07:53 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

fixed potential temp disk memory leak

Browse Source
This commit is contained in:
SevaSk 2023-05-30 19:04:28 -04:00
parent 66da757f5f
commit 8abdb197a7
2 changed files with 13 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
AudioTranscriber.py
View File

@ -45,8 +45,14 @@ class AudioTranscriber:
who_spoke, data, time_spoken = audio_queue.get() who_spoke, data, time_spoken = audio_queue.get()
self.update_last_sample_and_phrase_status(who_spoke, data, time_spoken) self.update_last_sample_and_phrase_status(who_spoke, data, time_spoken)
source_info = self.audio_sources[who_spoke] source_info = self.audio_sources[who_spoke]
temp_file = source_info["process_data_func"](source_info["last_sample"])
text = self.audio_model.get_transcription(temp_file) text = ''
temp_file = NamedTemporaryFile(delete=False, suffix=".wav")
source_info["process_data_func"](source_info["last_sample"], temp_file.name)
text = self.audio_model.get_transcription(temp_file.name)
temp_file.close()
os.unlink(temp_file.name)
if text != '' and text.lower() != 'you': if text != '' and text.lower() != 'you':
self.update_transcript(who_spoke, text, time_spoken) self.update_transcript(who_spoke, text, time_spoken)
@ -63,23 +69,19 @@ class AudioTranscriber:
source_info["last_sample"] += data source_info["last_sample"] += data
source_info["last_spoken"] = time_spoken source_info["last_spoken"] = time_spoken
def process_mic_data(self, data): def process_mic_data(self, data, temp_file_name):
temp_file = NamedTemporaryFile().name
audio_data = sr.AudioData(data, self.audio_sources["You"]["sample_rate"], self.audio_sources["You"]["sample_width"]) audio_data = sr.AudioData(data, self.audio_sources["You"]["sample_rate"], self.audio_sources["You"]["sample_width"])
wav_data = io.BytesIO(audio_data.get_wav_data()) wav_data = io.BytesIO(audio_data.get_wav_data())
with open(temp_file, 'w+b') as f: with open(temp_file_name, 'w+b') as f:
f.write(wav_data.read()) f.write(wav_data.read())
return temp_file
def process_speaker_data(self, data): def process_speaker_data(self, data, temp_file_name):
temp_file = NamedTemporaryFile().name with wave.open(temp_file_name, 'wb') as wf:
with wave.open(temp_file, 'wb') as wf:
wf.setnchannels(self.audio_sources["Speaker"]["channels"]) wf.setnchannels(self.audio_sources["Speaker"]["channels"])
p = pyaudio.PyAudio() p = pyaudio.PyAudio()
wf.setsampwidth(p.get_sample_size(pyaudio.paInt16)) wf.setsampwidth(p.get_sample_size(pyaudio.paInt16))
wf.setframerate(self.audio_sources["Speaker"]["sample_rate"]) wf.setframerate(self.audio_sources["Speaker"]["sample_rate"])
wf.writeframes(data) wf.writeframes(data)
return temp_file
def update_transcript(self, who_spoke, text, time_spoken): def update_transcript(self, who_spoke, text, time_spoken):
source_info = self.audio_sources[who_spoke] source_info = self.audio_sources[who_spoke]

4
TranscriberModels.py
View File

@ -24,9 +24,7 @@ class WhisperTranscriber:
class APIWhisperTranscriber: class APIWhisperTranscriber:
def get_transcription(self, wav_file_path): def get_transcription(self, wav_file_path):
new_file_path = wav_file_path + '.wav' audio_file= open(wav_file_path, "rb")
os.rename(wav_file_path, new_file_path)
audio_file= open(new_file_path, "rb")
try: try:
result = openai.Audio.translate("whisper-1", audio_file) result = openai.Audio.translate("whisper-1", audio_file)
except Exception as e: except Exception as e: