refactor: 修改项目结构

This commit is contained in:
2026-03-21 21:25:59 +08:00
parent 5b47ab1dea
commit 909f1b400c
32 changed files with 535 additions and 703 deletions
+21
View File
@@ -0,0 +1,21 @@
import re
from typing import Dict
class TextNormalizer:
"""文本规范化"""
DEFAULT_REPLACEMENTS = {
"\\[": "$$", "\\]": "$$", "\\(": "$", "\\)": "$",
'\u2018': "'", '\u2019': "'", '\u0060': "'",
'\u201C': '"', '\u201D': '"',
'\u2013': '-', '\u2014': '--', '\u2212': '-',
'\u00A0': ' ', '\u2026': '...'
}
def __init__(self, custom_rules: Dict[str, str] = None):
self.replacements = {**self.DEFAULT_REPLACEMENTS, **(custom_rules or {})}
self._pattern = re.compile('|'.join(re.escape(k) for k in self.replacements))
def normalize(self, text: str) -> str:
return self._pattern.sub(lambda m: self.replacements[m.group()], text)