horiyouta's picture
202406210928
fd6a905
import re
import cn2an
from style_bert_vits2.nlp.symbols import PUNCTUATIONS
__REPLACE_MAP = {
":": ",",
"οΌ›": ",",
",": ",",
"。": ".",
"!": "!",
"?": "?",
"\n": ".",
"Β·": ",",
"、": ",",
"...": "…",
"$": ".",
"β€œ": "'",
"”": "'",
'"': "'",
"β€˜": "'",
"’": "'",
"(": "'",
"οΌ‰": "'",
"(": "'",
")": "'",
"γ€Š": "'",
"》": "'",
"【": "'",
"】": "'",
"[": "'",
"]": "'",
"β€”": "-",
"~": "-",
"~": "-",
"γ€Œ": "'",
"」": "'",
}
def normalize_text(text: str) -> str:
numbers = re.findall(r"\d+(?:\.?\d+)?", text)
for number in numbers:
text = text.replace(number, cn2an.an2cn(number), 1)
text = replace_punctuation(text)
return text
def replace_punctuation(text: str) -> str:
text = text.replace("ε—―", "恩").replace("ε‘£", "母")
pattern = re.compile("|".join(re.escape(p) for p in __REPLACE_MAP))
replaced_text = pattern.sub(lambda x: __REPLACE_MAP[x.group()], text)
replaced_text = re.sub(
r"[^\u4e00-\u9fa5" + "".join(PUNCTUATIONS) + r"]+", "", replaced_text
)
return replaced_text