语音识别数据列表和字典处理脚本

			import os
	source_file = 'D:\PythonTest\ASR_work\data_thchs30\data'
	def source_get(source_file):
	    train_file = source_file
	    label_lst = []
	    wav_lst = []
	    for root, dirs, files in os.walk(train_file):
	        for file in files:
	            if file.endswith('.wav') or file.endswith('.WAV'):
	                wav_file = os.sep.join([root, file])
	                label_file = wav_file + '.trn'
	                wav_lst.append(wav_file)
	                label_lst.append(label_file)
	            
	    return label_lst, wav_lst
	
	label_lst, wav_lst = source_get(source_file)
	# with open('s.txt','w+',encoding="utf-8") as f:
	#     for i in wav_lst:
	#         i = i.split("\\")[-1]
	#         i = i.split('.')[0]
	#         f.write(i+'\n')
	
	def read_label(label_file):
	    with open(label_file, 'r', encoding='utf8') as f:
	        data = f.readlines()
	        return data[0]
	
	# print(read_label(label_lst[0]))
	
	def gen_label_data(label_lst):
	    label_data = []
	    for label_file in label_lst:
	        pny = read_label(label_file)
	        label_data.append(pny.strip('\n'))
	    return label_data
	
	label_data = gen_label_data(label_lst)
	print(label_data[0:2])
	print(wav_lst[0:2])
	with open('ss.txt','w+',encoding="utf-8") as f:
	    for i in range(len(wav_lst)):
	        ii = wav_lst[i].split("\\")[-1]
	        ii = ii.split('.')[0]
	        line = label_data[i].split()
	        l = ''.join(line)
	        l = ' '.join(l)
	        l = ii+' '+l
	        f.write(l+'\n')


	
	with open('./aishell_transcript_v0.8.txt','r',encoding='utf-8') as f:
	    lines = f.readlines()
	    vocab = []
	    for line in lines:
	        line = line.split()
	        liness = line[1:]
	        l = ''.join(liness)
	        l = ' '.join(l)
	        l = l.split(' ')
	        for pny in l:
	            if pny not in vocab:
	                vocab.append(pny)
	    vocab.append('_')
	
	with open('./sss.txt','w',encoding='utf-8') as fr:
	    for i in vocab:
	        fr.write(i)
	        fr.write('\n')