/
pinyin_timestamping.py
executable file
·143 lines (112 loc) · 4.07 KB
/
pinyin_timestamping.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
#!/usr/bin/python3
from vosk import Model, KaldiRecognizer
import sys
import os
import wave
import json
import cn2an
from pypinyin import pinyin, lazy_pinyin, Style
#input
import re
import string
from zhon.hanzi import punctuation
# define punctuation
punctuations = '''!()-[]{};:'"\,<>./?@#$%^&*_~'''
input = sys.argv[1]
stripped_input = re.sub(r'[%s]+' %punctuation, '', input)
file_name = stripped_input[:10]
person = sys.argv[2]
fps = 30.0
# wf = wave.open(sys.argv[1], "rb")
#input
print('stripped_chn_punct', stripped_input)
stripped_eng_punct = ""
for char in stripped_input:
if char not in punctuations:
stripped_eng_punct = stripped_eng_punct + char
print('stripped_eng_punct', stripped_eng_punct)
# stripped = re.sub(r'[%s]+' %punctuation, '', stripped_eng_punct)
# print('stripped_chn_punct', stripped)
len_stripped=len(stripped_eng_punct)
print('len_stripped', len_stripped)
py_input=lazy_pinyin(stripped_eng_punct)
# print('input_pinyin', py_input)
input_length=len(py_input)
text_file = open('./input_timestamp/{person}/{file_name}.txt'.format(person=person, file_name=file_name), "w")
wf = wave.open('./input_audio/{person}/{file_name}.wav'.format(person=person, file_name=file_name), "rb")
def digi_to_py(word):
digidict = {
'1': u'yi', '2': u'er', '3': u'san', '4': u'si', '5': u'wu',
'6': u'liu', '7': u'qi', '8': u'ba', '9': u'jiu', '0': u'ling'
}
patten = '([0-9]{1})'
for i in word:
if re.search(patten, i):
py = digidict[i]
word = word.replace(i, (py))
return word
if wf.getnchannels() != 1 or wf.getsampwidth() != 2 or wf.getcomptype() != "NONE":
print ("Audio file must be WAV format mono PCM.")
exit (1)
#https://github.com/alphacep/vosk-api/blob/37fbe1a52b274e980fba992af73991529e524568/python/example/test_text.py
if not os.path.exists("model"):
print ("Please download the model from https://github.com/alphacep/vosk-api/blob/master/doc/models.md and unpack as 'model' in the current folder.")
exit (1)
model = Model("model")
rec = KaldiRecognizer(model, wf.getframerate())
while True:
data = wf.readframes(1000000)
if len(data) == 0:
break
if rec.AcceptWaveform(data):
res =json.loads(rec.Result())
# print('res', res)
print('AcceptWaveform rec.Result.text', res['text'])
else:
print('rec.PartialResult', rec.PartialResult())
res = json.loads(rec.FinalResult())
print('rec.FinalResult', res['text'])
i=0
j=0
for item in res['result']:
print('item', item)
#print(item['word'])
py = lazy_pinyin(item['word'])
st = item['start']
et = item['end']
nc = len(py)
# print('nc', nc)
step = (et-st)/(nc+1)
for idx in range(nc):
tstamp=st + step * (idx + 1)
# print(tstamp, py[idx])
frame_idx = int(tstamp * fps + 0.5)
#no number, only chinese input
# n = text_file.write(str(frame_idx) + ' ' + py_input[i] + '\n')
# print(str(frame_idx), py_input[i], 'py_input[i]')
# i=i+1
if i > len_stripped-1:
break
if py_input[i].isnumeric():
# output = cn2an.an2cn(py_input[i], "low")
# print('to_chn', output)
len_digit= len(py_input[i])
# print('isdigit', py_input[i], 'len', len_digit, 'j', j)
if j<len_digit:
# print('digit', py_input[i][0])
# print(list(py_input[i]))
# digit_cn= cn2an.an2cn(py_input[i][j], "low")
# print(str(frame_idx), py_input[i][j], 'py_input[i][j]', i)
py_digit_cn = digi_to_py(py_input[i][j])
n = text_file.write(str(frame_idx) + ' ' + py_digit_cn + '\n')
print(str(frame_idx), py_digit_cn, 'py_digit_cn', i)
digit = int(py_input[i])
j=j+1
else:
j=0
i=i+1
else:
n = text_file.write(str(frame_idx) + ' ' + py_input[i] + '\n')
print(str(frame_idx), py_input[i], 'input', i)
i=i+1
text_file.close()