wangweiye
/
mathdeptv2
Archived
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
This repository has been archived on 2024-06-23. You can view files and clone it, but cannot push or open issues or pull requests.
mathdeptv2/工具/批量添加题库字段数据.ipynb

172 lines
9.0 KiB
Plaintext
Raw Blame History

{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"题号: 030169 , 字段: ans 中已修改数据: $2$\n",
"题号: 030273 , 字段: ans 中已修改数据: $5$\n",
"题号: 001677 , 字段: ans 中已修改数据: $\\dfrac\\pi 3$\n",
"题号: 003531 , 字段: ans 中已修改数据: 以$(1,0)$为圆心, $3$为半径的圆\n",
"题号: 003533 , 字段: ans 中已修改数据: $\\{4\\mathrm{i}\\}$\n",
"题号: 003455 , 字段: ans 中已修改数据: $\\dfrac\\pi 3$\n",
"题号: 004092 , 字段: ans 中已修改数据: B\n",
"题号: 003891 , 字段: ans 中已修改数据: D\n",
"题号: 001643 , 字段: ans 中已修改数据: B\n",
"题号: 000182 , 字段: ans 中已修改数据: $\\dfrac{\\sqrt{6}}6$\n",
"题号: 000187 , 字段: ans 中已修改数据: (1) 证明略; (2) 证明略\n",
"题号: 000298 , 字段: ans 中已修改数据: (1) $\\dfrac\\pi 3$; (2) $\\dfrac 56$\n",
"题号: 003495 , 字段: ans 中已修改数据: (1) $\\arctan\\sqrt{2}$; (2) $\\dfrac\\pi 4$; (3) $\\sqrt{2}$\n",
"题号: 004180 , 字段: ans 中已修改数据: (1) $\\dfrac\\pi 3$; (2) $\\dfrac{\\sqrt{2}}2$\n",
"题号: 003500 , 字段: ans 中已修改数据: (1) $2$; (2) $\\arccos \\dfrac{\\sqrt{2}}4$\n",
"题号: 003462 , 字段: ans 中已修改数据: (1) 证明略; (2) 当且仅当$\\theta=\\arcsin\\dfrac{\\sqrt{3}}3$时, 三角形$AEF$的面积最大, 最大值为$\\dfrac 12$\n",
"题号: 001506 , 字段: ans 中已修改数据: $-1$\n",
"题号: 004125 , 字段: ans 中已修改数据: $(0,1)$\n",
"题号: 002027 , 字段: ans 中已修改数据: $\\dfrac 8{25}$\n",
"题号: 030152 , 字段: ans 中已修改数据: $\\dfrac 12\\pm \\dfrac 12\\mathrm{i}$\n",
"题号: 004414 , 字段: ans 中已修改数据: $(3,+\\infty)$\n",
"题号: 001013 , 字段: ans 中已修改数据: $\\{(\\dfrac{3+\\sqrt{5}}2,\\dfrac{5+\\sqrt{5}}2),(\\dfrac{3-\\sqrt{5}}2,\\dfrac{5-\\sqrt{5}}2)\\}$\n",
"题号: 003798 , 字段: ans 中已修改数据: $\\dfrac\\pi 3$\n",
"题号: 001253 , 字段: ans 中已修改数据: $[0,\\dfrac 43]$\n",
"题号: 001510 , 字段: ans 中已修改数据: $4\\pi$\n",
"题号: 001515 , 字段: ans 中已修改数据: $-4$或$2$\n",
"题号: 004111 , 字段: ans 中已修改数据: $[\\dfrac 73,\\dfrac{13}3)$\n",
"题号: 030153 , 字段: ans 中已修改数据: $3.985\\overrightarrow{OA}+1.690\\overrightarrow{OB}$\n",
"题号: 001993 , 字段: ans 中已修改数据: C\n",
"题号: 004240 , 字段: ans 中已修改数据: D\n",
"题号: 003645 , 字段: ans 中已修改数据: C\n",
"题号: 004116 , 字段: ans 中已修改数据: A\n",
"题号: 004636 , 字段: ans 中已修改数据: (1) 当$a>2$时, 解集为$(-\\infty,-2)\\cup (0,+\\infty)$; 当$a=2$时, 解集为$\\varnothing$; 当$a<2$时, 解集为$(-2,0)$; (2) $(-\\infty,\\dfrac {34}{15})$\n",
"题号: 001494 , 字段: ans 中已修改数据: (1) 证明略; (2) $f(x)=98-x$\n",
"题号: 004098 , 字段: ans 中已修改数据: (1) $AC=\\sqrt{7}\\text{km}$, $S_{ABCD}=2\\sqrt{3}\\text{km}^2$; (2) 当且仅当$P$位于弧$\\overset\\frown{ABC}$的中点时, 改造后的新建筑用地面积最大, 最大面积为$\\dfrac{9\\sqrt{3}}4\\text{km}^2$\n",
"题号: 004424 , 字段: ans 中已修改数据: (1) $(3,4]$; (2) $(0,1]$上的值域为$\\{1\\}$, $(1,2]$上的值域为$\\{3,4\\}$, $(2,3]$上的值域为$\\{7,8,9\\}$, $(0,n]$上的值域中含有的元素个数为$\\dfrac{n(n+1)}2$; (3) $(3,+\\infty)$\n",
"题号: 004509 , 字段: ans 中已修改数据: (1) $[\\dfrac 12,+\\infty)$; (2) 不是, 证明略; (3) 证明略\n"
]
}
],
"source": [
"import os,re,json\n",
"\n",
"\"\"\"---明确数据文件位置---\"\"\"\n",
"datafile = \"文本文件/metadata.txt\"\n",
"# 双回车分隔，记录内单回车分隔列表，首行为字段名\n",
"\"\"\"---文件位置结束---\"\"\"\n",
"\n",
"def trim(string):\n",
" string = re.sub(r\"^[ \\t\\n]*\",\"\",string)\n",
" string = re.sub(r\"[ \\t\\n]*$\",\"\",string)\n",
" return string\n",
"def FloatToInt(string):\n",
" f = float(string)\n",
" if abs(f-round(f))<0.01:\n",
" f = round(f)\n",
" return f\n",
"\n",
"with open(datafile,\"r\",encoding=\"utf8\") as f:\n",
" data = f.read()\n",
"pos = data.index(\"\\n\")\n",
"field = data[:pos].strip()\n",
"appending_data = data[pos:]\n",
"\n",
"with open(r\"../题库0.3/Problems.json\",\"r\",encoding = \"utf8\") as f:\n",
" database = f.read()\n",
"pro_dict = json.loads(database)\n",
"with open(r\"../题库0.3/LessonObj.json\",\"r\",encoding = \"utf8\") as f:\n",
" database = f.read()\n",
"obj_dict = json.loads(database)\n",
"\n",
"#该字段列表可能需要更新\n",
"fields = [\"content\",\"objs\",\"tags\",\"genre\",\"ans\",\"solution\",\"duration\",\"usages\",\"origin\",\"edit\",\"same\",\"related\",\"remark\",\"space\"]\n",
"\n",
"if field in fields:\n",
" field_type = type(pro_dict[\"000001\"][field])\n",
" datalist = [record.strip() for record in appending_data.split(\"\\n\\n\") if len(trim(record)) > 0]\n",
" for record in datalist:\n",
" id = re.findall(r\"^[\\d]{1,}\",record)[0]\n",
" data = record[len(id):].strip()\n",
" id = id.zfill(6)\n",
" if not id in pro_dict:\n",
" print(\"题号:\",id,\"不在数据库中.\")\n",
" break\n",
" \n",
" #字符串类型字段添加数据\n",
" elif field_type == str and data in pro_dict[id][field]:\n",
" print(\"题号:\",id,\", 字段:\",field,\"中已有该数据:\",data)\n",
" elif field_type == str and not data in pro_dict[id][field] and not field == \"ans\" and not field == \"space\":\n",
" origin_data = pro_dict[id][field]\n",
" new_data = trim(origin_data + \"\\n\" + data)\n",
" pro_dict[id][field] = new_data\n",
" print(\"题号:\",id,\", 字段:\",field,\"中已添加数据:\",data)\n",
" elif field_type == str and not data in pro_dict[id][field] and field == \"ans\" or field == \"space\":\n",
" pro_dict[id][field] = data\n",
" print(\"题号:\",id,\", 字段:\",field,\"中已修改数据:\",data)\n",
" \n",
" #数值类型字段添加数据\n",
" elif (field_type == int or field_type == float) and abs(float(data) - pro_dict[id][field])<0.01:\n",
" print(\"题号:\",id,\", 字段:\",field,\"中已有该数据:\",FloatToInt(data))\n",
" elif (field_type == int or field_type == float) and abs(float(data) - pro_dict[id][field])>=0.01:\n",
" pro_dict[id][field] = FloatToInt(data)\n",
" print(\"题号:\",id,\", 字段:\",field,\"中已修改数据:\",FloatToInt(data))\n",
" \n",
" #列表类型字段添加数据\n",
" elif field_type == list:\n",
" cell_data_list = [d.strip() for d in data.split(\"\\n\")]\n",
" for cell_data in cell_data_list:\n",
" if cell_data in pro_dict[id][field]:\n",
" print(\"题号:\",id,\", 字段:\",field,\"中已有该数据:\",cell_data)\n",
" elif not field == \"objs\":\n",
" pro_dict[id][field].append(cell_data)\n",
" print(\"题号:\",id,\", 字段:\",field,\"中已添加数据:\",cell_data)\n",
" else:\n",
" if not cell_data in obj_dict and not cell_data.upper() == \"KNONE\":\n",
" print(\"题号:\",id,\", 字段:\",field,\"目标编号有误:\",cell_data)\n",
" else:\n",
" pro_dict[id][field].append(cell_data.upper())\n",
" print(\"题号:\",id,\", 字段:\",field,\"中已添加数据:\",cell_data.upper())\n",
"\n",
"with open(r\"../题库0.3/Problems.json\",\"w\",encoding = \"utf8\") as f:\n",
" f.write(json.dumps(pro_dict,indent=4,ensure_ascii=False))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3.8.8 ('base')",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.8"
},
"orig_nbformat": 4,
"vscode": {
"interpreter": {
"hash": "d311ffef239beb3b8f3764271728f3972d7b090c974f8e972fcdeedf230299ac"
}
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Reference in New Issue View Git Blame Copy Permalink