Spaces:
Running
Running
File size: 5,193 Bytes
e82212c |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 |
import sys
# sys.path.append("/data/docker/liujing04/gpt-vits/mq-vits-s1bert_no_bert-kv_cache-new_text-emo")
sys.path.append("/data/docker/liujing04/gpt-vits/mq-vits-s1bert_no_bert-kv_cache-new_text-emo2v2")
import re
import cn2an
from pyjyutping import jyutping
from text.symbols import punctuation
normalizer = lambda x: cn2an.transform(x, "an2cn")
INITIALS = [
"aa",
"aai",
"aak",
"aap",
"aat",
"aau",
"ai",
"au",
"ap",
"at",
"ak",
"a",
"p",
"b",
"e",
"ts",
"t",
"dz",
"d",
"kw",
"k",
"gw",
"g",
"f",
"h",
"l",
"m",
"ng",
"n",
"s",
"y",
"w",
"c",
"z",
"j",
"ong",
"on",
"ou",
"oi",
"ok",
"o",
"uk",
"ung",
]
INITIALS += ["sp", "spl", "spn", "sil"]
rep_map = {
"：": ",",
"；": ",",
"，": ",",
"。": ".",
"！": "!",
"？": "?",
"\n": ".",
"·": ",",
"、": ",",
"...": "…",
"$": ".",
"“": "'",
"”": "'",
'"': "'",
"‘": "'",
"’": "'",
"（": "'",
"）": "'",
"(": "'",
")": "'",
"《": "'",
"》": "'",
"【": "'",
"】": "'",
"[": "'",
"]": "'",
"—": "-",
"～": "-",
"~": "-",
"「": "'",
"」": "'",
}
def replace_punctuation(text):
# text = text.replace("嗯", "恩").replace("呣", "母")
pattern = re.compile("|".join(re.escape(p) for p in rep_map.keys()))
replaced_text = pattern.sub(lambda x: rep_map[x.group()], text)
replaced_text = re.sub(
r"[^\u4e00-\u9fa5" + "".join(punctuation) + r"]+", "", replaced_text
)
return replaced_text
def text_normalize(text):
text = normalizer(text)
text = replace_punctuation(text)
return text
punctuation_set=set(punctuation)
def jyuping_to_initials_finals_tones(jyuping_syllables):
initials_finals = []
tones = []
word2ph = []
for syllable in jyuping_syllables:
if syllable in punctuation:
initials_finals.append(syllable)
tones.append(0)
word2ph.append(1) # Add 1 for punctuation
elif syllable == "_":
initials_finals.append(syllable)
tones.append(0)
word2ph.append(1) # Add 1 for underscore
else:
try:
tone = int(syllable[-1])
syllable_without_tone = syllable[:-1]
except ValueError:
tone = 0
syllable_without_tone = syllable
for initial in INITIALS:
if syllable_without_tone.startswith(initial):
if syllable_without_tone.startswith("nga"):
initials_finals.extend(
[
syllable_without_tone[:2],
syllable_without_tone[2:] or syllable_without_tone[-1],
]
)
# tones.extend([tone, tone])
tones.extend([-1, tone])
word2ph.append(2)
else:
final = syllable_without_tone[len(initial) :] or initial[-1]
initials_finals.extend([initial, final])
# tones.extend([tone, tone])
tones.extend([-1, tone])
word2ph.append(2)
break
assert len(initials_finals) == len(tones)
###魔改为辅音+带音调的元音
phones=[]
for a,b in zip(initials_finals,tones):
if(b not in [-1,0]):###防止粤语和普通话重合开头加Y，如果是标点，不加。
todo="%s%s"%(a,b)
else:todo=a
if(todo not in punctuation_set):todo="Y%s"%todo
phones.append(todo)
# return initials_finals, tones, word2ph
return phones, word2ph
def get_jyutping(text):
jp = jyutping.convert(text)
# print(1111111,jp)
for symbol in punctuation:
jp = jp.replace(symbol, " " + symbol + " ")
jp_array = jp.split()
return jp_array
def get_bert_feature(text, word2ph):
from text import chinese_bert
return chinese_bert.get_bert_feature(text, word2ph)
def g2p(text):
# word2ph = []
jyuping = get_jyutping(text)
# print(jyuping)
# phones, tones, word2ph = jyuping_to_initials_finals_tones(jyuping)
phones, word2ph = jyuping_to_initials_finals_tones(jyuping)
# phones = ["_"] + phones + ["_"]
# tones = [0] + tones + [0]
# word2ph = [1] + word2ph + [1]
return phones, word2ph
if __name__ == "__main__":
# text = "啊！但是《原神》是由,米哈\游自主， [研发]的一款全.新开放世界.冒险游戏"
text = "佢個鋤頭太短啦。"
text = text_normalize(text)
# phones, tones, word2ph = g2p(text)
phones, word2ph = g2p(text)
# print(phones, tones, word2ph)
print(phones, word2ph) |