wangweiye
/
mathdeptv2
Archived
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

修复RefineMathpix函数的两个小bug

This commit is contained in:
weiye.wang 2024-01-31 23:14:18 +08:00
parent 0470fc6e9b
commit f4f8dfbe82
1 changed files with 4 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
工具v2/database_tools.py
View File

@ -1203,7 +1203,7 @@ def RefineMathpix(raw_string): # 进一步修改mathpix得到的字符串
replacestrings = {r"\\overparen": r"\\overset\\frown", "eqslant": "eq", r"\\vec": r"\\overrightarrow ", r"\\bar": r"\\overline", r"\\lim": r"\\displaystyle\\lim", r"\\sum":r"\\displaystyle\\sum", r"\\prod":r"\\displaystyle\\prod", r"\\mid":"|", r"\^\{\\prime\}":"'",r"e\^":r"\\mathrm{e}^",r"/\s*/":r"\\parallel "}
wrongrecog = {"雉":"锥","[粗秿]圆":"椭圆","投郑":"投掷","抛郑":"抛掷","范目":"范围","揷":"插","末见":"未见","末成":"未成","针角":"钝角","幕函数":"幂函数","末知":"未知","阀值":"阈值","祖[桓晅]":"祖暅","图象":"图像","末使用":"未使用"}
string = raw_string
string = re.sub(r"\\left(?:\.?)|\\right(?:\.?)","",string) #删去括号前的\left与\right标记
string = re.sub(r"\\left(?:[\.\(\[\{|])|\\right(?:[\.\(\[\{|])","",string) #删去括号前的\left与\right标记
for s in puctuationsfulltosemi:
string = re.sub(s,puctuationsfulltosemi[s],string) #将部分全角标记替换为半角
for s in replacestrings:
@ -1236,7 +1236,9 @@ def RefineMathpix(raw_string): # 进一步修改mathpix得到的字符串
string = re.sub(r"[\\{]*\\(begin|end)\{array\}(?:\{[rcl]*\}){0,1}",lambda matchobj: "\\" + matchobj.group(1) + "{cases}",string) #将分段函数的array环境改为cases环境
string = re.sub(r"([\u4e00-\u9fa5\$])[\s]*\n\\item",lambda matchobj: matchobj.group(1)+"\\blank{50}.\n\\item",string) #给中文或公式结尾的题目最后一行加上填空的空格.
string = re.sub(r"(是|为|(?:=\$))\s*([,.;\n])",lambda matchobj: matchobj.group(1) + "\\blank{50}" + ("." if matchobj.group(2) == "\n" else "") + matchobj.group(2),string) #给行中的题目需要的地方加上空格
string = re.sub(r"([\u4e00-\u9fa5\$])(?:\\bracket\{20\})*[\.]*[\s]*\n\\fourch",lambda matchobj: matchobj.group(1)+"\\bracket{20}.\n\\fourch",string) #给中文或公式结尾的题目最后一行加上选择题的括号.
if not "\\bracket" in string:
string = re.sub(r"([\u4e00-\u9fa5\$])(?:\\bracket\{20\})*[\.]*[\s]*\n\\fourch",lambda matchobj: matchobj.group(1)+"\\bracket{20}.\n\\fourch",string) #给中文或公式结尾的题目最后一行加上选择题的括号.
string = re.sub(r"\\bracket\{(\d*)\}\$",lambda matchobj: "$"+"\\bracket{"+matchobj.group(1)+"}",string)#交换出现在$之前的括号与$的位置
string = re.sub(r"(%[^\n]*)\\blank\{50\}\.",lambda matchobj:matchobj.group(1),string) #注释行不加\blank{50}
string = re.sub(r"\\blank\{50\}\.\s*\n\\fourch",r"\\bracket{20}."+"\n"+r"\\fourch",string) #选择题前的空格改为括号
string = re.sub(r"[\\\\]*\n(\(\d{1,2}\))(?:(?!\n)\s)*",lambda matchobj: "\\\\\n"+matchobj.group(1)+" ",string) #新一行的小题号回车前加上换行符