def transfer(NAPs, wavs, wav_audio_ids):
audio_memory = AudioMemory()
for audio_id, nappers in enumerate(NAPs):
for i in range(len(nappers)):
NAP = nappers[i]
if len(NAP):
clean_key, overlap_key = audio_memory._digitize(NAP)
crude_hash = utils.d_hash(NAP, hash_size=8)
fine_hash = utils.d_hash(NAP, hash_size=16)
wav_file = wavs[audio_id][i]
try:
audio_segment = AudioSegment(
audio_id, crude_hash, fine_hash, wav_file,
wav_audio_ids[(wav_file, audio_id)])
except:
print '{} {} barfed'.format(wav_file, audio_id)
continue
audio_memory._insert(audio_memory.NAP_intervals, clean_key,
audio_segment)
audio_memory._insert(audio_memory.NAP_intervals, overlap_key,
audio_segment)
audio_memory._insert(audio_memory.audio_ids, audio_id,
audio_segment)
return audio_memory
def transfer(NAPs, wavs, wav_audio_ids):
audio_memory = AudioMemory()
for audio_id, nappers in enumerate(NAPs):
for i in range(len(nappers)):
NAP = nappers[i]
if len(NAP):
clean_key, overlap_key = audio_memory._digitize(NAP)
crude_hash = utils.d_hash(NAP, hash_size=8)
fine_hash = utils.d_hash(NAP, hash_size=16)
wav_file = wavs[audio_id][i]
try:
audio_segment = AudioSegment(audio_id, crude_hash, fine_hash, wav_file, wav_audio_ids[(wav_file, audio_id)])
except:
print '{} {} barfed'.format(wav_file, audio_id)
continue
audio_memory._insert(audio_memory.NAP_intervals, clean_key, audio_segment)
audio_memory._insert(audio_memory.NAP_intervals, overlap_key, audio_segment)
audio_memory._insert(audio_memory.audio_ids, audio_id, audio_segment)
return audio_memory
def find(self, NAP):
crude_hash = utils.d_hash(NAP, hash_size=8)
fine_hash = utils.d_hash(NAP, hash_size=16)
clean_key, overlap_key = self._digitize(NAP)
unsorted = self.NAP_intervals[clean_key] if clean_key in self.NAP_intervals else [] + self.NAP_intervals[overlap_key] if overlap_key in self.NAP_intervals else []
return sorted(unsorted, key = lambda x: utils.hamming_distance(fine_hash, x.fine_hash))[0] if len(unsorted) else [], crude_hash, fine_hash, clean_key, overlap_key
