Hugging Face's logo

Spaces:
paranov
/
MeloTTS
Build error

App Files Community
MeloTTS
File size: 6,038 Bytes 
4300fed
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
import re
import os
import glob
import numpy as np
import soundfile as sf
import torchaudio
from txtsplit import txtsplit
def split_sentence(text, min_len=10, language_str='EN'):
    if language_str in ['EN', 'FR', 'ES', 'SP', 'DE', 'RU']:
        sentences = split_sentences_latin(text, min_len=min_len)
    else:
        sentences = split_sentences_zh(text, min_len=min_len)
    return sentences

def split_sentences_latin(text, min_len=10):
    text = re.sub('[。！？；]', '.', text)
    text = re.sub('[，]', ',', text)
    text = re.sub('[“”]', '"', text)
    text = re.sub('[‘’]', "'", text)
    text = re.sub(r"[\<\>\(\)\[\]\"\«\»]+", "", text)
    return txtsplit(text, 512, 512)
    # 将文本中的换行符、空格和制表符替换为空格
    # text = re.sub('[\n\t ]+', ' ', text)
    # # 在标点符号后添加一个空格
    # text = re.sub('([,.!?;])', r'\1 $#!', text)
    # # 分隔句子并去除前后空格
    # sentences = [s.strip() for s in text.split('$#!')]
    # if len(sentences[-1]) == 0: del sentences[-1]

    # new_sentences = []
    # new_sent = []
    # count_len = 0
    # for ind, sent in enumerate(sentences):
    #     # print(sent)
    #     new_sent.append(sent)
    #     count_len += len(sent.split(" "))
    #     if count_len > min_len or ind == len(sentences) - 1:
    #         count_len = 0
    #         new_sentences.append(' '.join(new_sent))
    #         new_sent = []
    # return merge_short_sentences_en(new_sentences)

def split_sentences_zh(text, min_len=10):
    text = re.sub('[。！？；]', '.', text)
    text = re.sub('[，]', ',', text)
    # 将文本中的换行符、空格和制表符替换为空格
    text = re.sub('[\n\t ]+', ' ', text)
    # 在标点符号后添加一个空格
    text = re.sub('([,.!?;])', r'\1 $#!', text)
    # 分隔句子并去除前后空格
    # sentences = [s.strip() for s in re.split('(。|！|？|；)', text)]
    sentences = [s.strip() for s in text.split('$#!')]
    if len(sentences[-1]) == 0: del sentences[-1]

    new_sentences = []
    new_sent = []
    count_len = 0
    for ind, sent in enumerate(sentences):
        new_sent.append(sent)
        count_len += len(sent)
        if count_len > min_len or ind == len(sentences) - 1:
            count_len = 0
            new_sentences.append(' '.join(new_sent))
            new_sent = []
    return merge_short_sentences_zh(new_sentences)

def merge_short_sentences_en(sens):
    """Avoid short sentences by merging them with the following sentence.

    Args:
        List[str]: list of input sentences.

    Returns:
        List[str]: list of output sentences.
    """
    sens_out = []
    for s in sens:
        # If the previous sentense is too short, merge them with
        # the current sentence.
        if len(sens_out) > 0 and len(sens_out[-1].split(" ")) <= 2:
            sens_out[-1] = sens_out[-1] + " " + s
        else:
            sens_out.append(s)
    try:
        if len(sens_out[-1].split(" ")) <= 2:
            sens_out[-2] = sens_out[-2] + " " + sens_out[-1]
            sens_out.pop(-1)
    except:
        pass
    return sens_out

def merge_short_sentences_zh(sens):
    # return sens
    """Avoid short sentences by merging them with the following sentence.

    Args:
        List[str]: list of input sentences.

    Returns:
        List[str]: list of output sentences.
    """
    sens_out = []
    for s in sens:
        # If the previous sentense is too short, merge them with
        # the current sentence.
        if len(sens_out) > 0 and len(sens_out[-1]) <= 2:
            sens_out[-1] = sens_out[-1] + " " + s
        else:
            sens_out.append(s)
    try:
        if len(sens_out[-1]) <= 2:
            sens_out[-2] = sens_out[-2] + " " + sens_out[-1]
            sens_out.pop(-1)
    except:
        pass
    return sens_out


if __name__ == '__main__':
    zh_text = "好的，我来给你讲一个故事吧。从前有一个小姑娘，她叫做小红。小红非常喜欢在森林里玩耍，她经常会和她的小伙伴们一起去探险。有一天，小红和她的小伙伴们走到了森林深处，突然遇到了一只凶猛的野兽。小红的小伙伴们都吓得不敢动弹，但是小红并没有被吓倒，她勇敢地走向野兽，用她的智慧和勇气成功地制服了野兽，保护了她的小伙伴们。从那以后，小红变得更加勇敢和自信，成为了她小伙伴们心中的英雄。"
    en_text = "I didn’t know what to do. I said please kill her because it would be better than being kidnapped,” Ben, whose surname CNN is not using for security concerns, said on Wednesday. “It’s a nightmare. I said ‘please kill her, don’t take her there.’"
    sp_text = "¡Claro! ¿En qué tema te gustaría que te hable en español? Puedo proporcionarte información o conversar contigo sobre una amplia variedad de temas, desde cultura y comida hasta viajes y tecnología. ¿Tienes alguna preferencia en particular?"
    fr_text = "Bien sûr ! En quelle matière voudriez-vous que je vous parle en français ? Je peux vous fournir des informations ou discuter avec vous sur une grande variété de sujets, que ce soit la culture, la nourriture, les voyages ou la technologie. Avez-vous une préférence particulière ?"
    de_text = 'Es war das Wichtigste was wir sichern wollten da es keine Möglichkeit gab eine 20 Megatonnen- H- Bombe ab zu werfen von einem 30, C124.'
    ru_text = 'Но он был во многом, как-бы, всё равно что сын плантатора, так как являлся сыном человека, у которого было в собственности много чего.'
    print(split_sentence(zh_text, language_str='ZH'))
    print(split_sentence(en_text, language_str='EN'))
    print(split_sentence(sp_text, language_str='SP'))
    print(split_sentence(fr_text, language_str='FR'))
    print(split_sentence(de_text, language_str='DE'))
    print(split_sentence(ru_text, language_str='RU'))