1.识别简历信息
This commit is contained in:
parent
1d518a83c5
commit
e934b45e77
@ -79,3 +79,16 @@ async def interview_insert(
|
||||
return schemas.Msg(code=200, msg='ok', data=data)
|
||||
|
||||
|
||||
@router.post("/interview_insert")
|
||||
async def interview_insert(
|
||||
request: Request,
|
||||
data_in: schemas.Interview,
|
||||
ckdb: CKDrive = Depends(get_ck_db),
|
||||
) -> schemas.Msg:
|
||||
""" 面试情况 """
|
||||
await interview.init()
|
||||
res = interview.insert_interview_sql()
|
||||
sql = res['sql']
|
||||
insert_data = res['insert_data']
|
||||
data = await db.execute_dict(sql, insert_data)
|
||||
return schemas.Msg(code=200, msg='ok', data=data)
|
||||
@ -26,3 +26,4 @@ from .user_url import *
|
||||
from .api_module import *
|
||||
from .event_list import *
|
||||
from .interview import *
|
||||
from .interview_plan import *
|
||||
16
schemas/interview_plan.py
Normal file
16
schemas/interview_plan.py
Normal file
@ -0,0 +1,16 @@
|
||||
from typing import List, Union, Dict
|
||||
|
||||
from pydantic import BaseModel
|
||||
from typing import Optional
|
||||
|
||||
|
||||
class Interview(BaseModel):
|
||||
job_name: str =None # 应聘职位
|
||||
hr_name: str # 面试负责人
|
||||
interview_name: str # 面试官
|
||||
interview_type: str # 面试类型
|
||||
interview_sign: int # 面试签到
|
||||
feedback: int # 面试反馈
|
||||
interview_round: int # 面试轮次
|
||||
pages: int = 1 # 分页的当前页
|
||||
time_type: str # 要查询的时间范围类型
|
||||
@ -150,3 +150,13 @@ def create_neidf(resp,columnName):
|
||||
columns.insert(0, columnName)
|
||||
df = pd.DataFrame(data=date, columns=columns)
|
||||
return df
|
||||
|
||||
def random_hex():
|
||||
"""
|
||||
生成16位随机数
|
||||
:return: 随机数
|
||||
"""
|
||||
result = hex(random.randint(0,16**16)).replace('0x','').upper()
|
||||
if(len(result)<16):
|
||||
result = '0'*(16-len(result))+result
|
||||
return result
|
||||
434
utils/jianli.py
Normal file
434
utils/jianli.py
Normal file
@ -0,0 +1,434 @@
|
||||
import re
|
||||
import docx
|
||||
import os
|
||||
import copy
|
||||
from pprint import pprint
|
||||
from paddlenlp import Taskflow
|
||||
import pdfplumber
|
||||
from win32com import client as wc
|
||||
from pdf2docx import Converter
|
||||
|
||||
# 文件路径
|
||||
PATH_DATA = os.path.abspath("C:/Users/Administrator/Desktop/面试简历")
|
||||
schema = ['姓名', '所在地', '户口所在地', '籍贯', '婚姻状况', '民族', '身高', '电话', 'tel', '应聘职位', '到岗时间', '学历', '毕业学校', '专业',
|
||||
'期望薪资', '在校时间', '电子邮箱', '工作经验', 'Email', '性别', '年龄'
|
||||
]
|
||||
schema_dict = {'姓名': 'name', '所在地': 'location', '户口所在地': 'account', '婚姻状况': 'gam', '民族': 'nation', '身高': 'height',
|
||||
'电话': 'phone', '应聘职位': 'job', '到岗时间': 'come_time', '学历': 'education', '毕业学校': 'school', '专业': 'career',
|
||||
'期望薪资': 'money', '在校时间': 'at_school', '电子邮箱': 'mail', '工作经验': 'work_exp', 'Email': 'mails',
|
||||
'性别': 'gender', '年龄': 'age', '籍贯': 'accounts', 'tel': 'tels'}
|
||||
|
||||
|
||||
def chkworlkandtime(listdata):
|
||||
"""
|
||||
获取工作经历中任职的公司名称和对应的在岗时间
|
||||
:param dictdata:
|
||||
:return:返回列表格式
|
||||
"""
|
||||
res = {}
|
||||
for i in listdata:
|
||||
for key, datalist in i.items():
|
||||
trueDict = {}
|
||||
for data in datalist:
|
||||
if data['text'] in trueDict:
|
||||
if data['probability'] <= trueDict[data['text']]['probability']:
|
||||
continue
|
||||
trueDict.update({
|
||||
data['text']: {
|
||||
'end': data['end'],
|
||||
'probability': data['probability'],
|
||||
'start': data['start'],
|
||||
}
|
||||
})
|
||||
trueList = []
|
||||
for key1, value1 in trueDict.items():
|
||||
value1.update({
|
||||
'text': key1
|
||||
})
|
||||
trueDict1 = copy.deepcopy(value1)
|
||||
trueList.append(trueDict1)
|
||||
trueList.sort(key=lambda item: item['start'])
|
||||
res.update({key: trueList})
|
||||
ress = []
|
||||
if res != {}:
|
||||
for i in range(len(res['公司名'])):
|
||||
date = {
|
||||
'name': res['公司名'][i]['text'],
|
||||
'time': res['时间'][i]['text']
|
||||
}
|
||||
ress.append(date)
|
||||
return ress
|
||||
|
||||
|
||||
def getText_docx(filename): # docx 转text
|
||||
"""将docx读成text"""
|
||||
doc = docx.Document(filename)
|
||||
fullText = []
|
||||
for i in doc.paragraphs: # 迭代docx文档里面的每一个段落
|
||||
fullText.append(i.text) # 保存每一个段落的文本
|
||||
numTables = doc.tables # 如果有表格的内容存放在这
|
||||
if len(numTables) > 0:
|
||||
for table in numTables:
|
||||
row_count = len(table.rows)
|
||||
col_count = len(table.columns)
|
||||
for i in range(row_count):
|
||||
for j in range(col_count):
|
||||
fullText.append(table.cell(i, j).text)
|
||||
return '\n'.join(fullText)
|
||||
|
||||
|
||||
def pdf_docx(url, filename):
|
||||
"""
|
||||
将pdf文件转为docx文件
|
||||
:param url:
|
||||
:param filename:
|
||||
:return:
|
||||
"""
|
||||
# 获取文件名称
|
||||
file_name = filename.split('.')[0]
|
||||
# pdf文件名称
|
||||
pdf_name = url + f"/{filename}"
|
||||
# docx文件名称
|
||||
docx_name = url + f"/{file_name}.docx"
|
||||
# 加载pdf文档
|
||||
cv = Converter(pdf_name)
|
||||
cv.convert(docx_name, start=0, end=12)
|
||||
cv.close()
|
||||
|
||||
|
||||
def getText_pdf(filename):
|
||||
"""将pdf读成text"""
|
||||
with pdfplumber.open(filename) as pdf_file:
|
||||
content = ''
|
||||
for i in range(len(pdf_file.pages)):
|
||||
page_text = pdf_file.pages[i]
|
||||
page_content = page_text.extract_text()
|
||||
if page_content:
|
||||
content = content + page_content + "\n"
|
||||
return content
|
||||
|
||||
|
||||
def doc_docx(url, filename):
|
||||
"""
|
||||
将doc文件转为docx文件
|
||||
:param filename:
|
||||
:return:
|
||||
"""
|
||||
word = wc.Dispatch("Word.Application")
|
||||
doc = word.Documents.Open(url + f"/{filename}")
|
||||
name = filename.split('.')[0]
|
||||
doc.SaveAs(url + f'/{name}.docx', 12) # 12为docx
|
||||
doc.Close()
|
||||
word.Quit()
|
||||
|
||||
|
||||
def clash(date, retain, pop):
|
||||
"""
|
||||
解决词性搜索时，最后的结果只取有值的一个
|
||||
例如'户口所在地','籍贯'只取默认的籍贯返回，（户口所在地有值把值给籍贯）
|
||||
:param date: 原数据
|
||||
:param retain: 要固定返回给前端的数据 户口所在地
|
||||
:param pop: 要删除的那个字段 籍贯
|
||||
:return:
|
||||
"""
|
||||
if date[retain] != '':
|
||||
date.pop(pop)
|
||||
else:
|
||||
date[retain] = date[pop]
|
||||
date.pop(pop)
|
||||
|
||||
|
||||
def get_date(schema, dates, schema_dict):
|
||||
"""
|
||||
把第三方获取的数据筛选出想要的基本信息
|
||||
:param schema:中文的词性标注
|
||||
:param dates:原数据
|
||||
:param schema_dict:对应中文的英文
|
||||
:return: 返回取出概率最大的基本信息数据
|
||||
"""
|
||||
date = {}
|
||||
for i in schema:
|
||||
text = dates[0].get(i, '')
|
||||
# 如果数据中没有搜到对应的键，返回空字符串
|
||||
if text == '':
|
||||
date[schema_dict[i]] = text
|
||||
else:
|
||||
if len(text) == 1:
|
||||
date[schema_dict[i]] = text[0]['text']
|
||||
else:
|
||||
aa = {}
|
||||
num = []
|
||||
for dic in text:
|
||||
aa[dic['probability']] = dic['text']
|
||||
num.append(dic['probability'])
|
||||
# 取出概率最大的值
|
||||
date[schema_dict[i]] = aa[max(num)]
|
||||
# 解决邮箱冲突的问题
|
||||
clash(date, 'mail', 'mails')
|
||||
# 解决户口所在地冲突的问题
|
||||
clash(date, 'account', 'accounts')
|
||||
# 解决电话冲突的问题
|
||||
clash(date, 'phone', 'tels')
|
||||
return date
|
||||
|
||||
|
||||
def fmtTxt(txt, istable=0):
|
||||
# 所有关键字
|
||||
chkStr = ['自我评价', '自我描述', '个人优势', '项目经历', '项目经验', '项目描述', '教育经历', '学习经历', '工作经历', '工作经验', '实习经历',
|
||||
'技能特长', '技能', '特长', '专长', '技能专长', '专业技能', '职业技能', '个人评价']
|
||||
# 自我描述
|
||||
chkList1 = ['自我评价', '自我描述', '个人优势', '个人评价']
|
||||
# 项目经验
|
||||
chkList2 = ['项目经历', '项目经验', '项目描述']
|
||||
# 教育背景
|
||||
chkList3 = ['教育经历', '学习经历']
|
||||
# 工作经历
|
||||
chkList4 = ['工作经历', '工作经验', '实习经历']
|
||||
# 个人技能
|
||||
chkList5 = ['技能特长', '技能', '特长', '专长', '技能专长', '专业技能', '职业技能']
|
||||
fmtList = [] # 返回拼接好的字符串列表
|
||||
trueIndex = 0
|
||||
fmtStr = ''
|
||||
nowChkList = []
|
||||
# 判断while循环是否需要停止
|
||||
stop_int = 0
|
||||
for index, i in enumerate(txt):
|
||||
if istable:
|
||||
text = i
|
||||
else:
|
||||
text = i.text
|
||||
# text = re.sub('\s+', '', text).lstrip() # 字符串去除空格和换行符
|
||||
|
||||
# 没有检测出关键字
|
||||
if not fmtStr:
|
||||
# 自我描述
|
||||
for i in chkList1:
|
||||
# 判断是不是以关键字开头
|
||||
if not text.startswith(i, 0):
|
||||
continue
|
||||
else:
|
||||
if i in text:
|
||||
fmtStr = text
|
||||
nowChkList = [chk for chk in chkStr if chk not in chkList1]
|
||||
# 检测出关键字证明需要继续循环
|
||||
stop_int = 1
|
||||
break
|
||||
if fmtStr:
|
||||
continue
|
||||
# 项目经验
|
||||
for i in chkList2:
|
||||
if i in text:
|
||||
fmtStr = text
|
||||
nowChkList = [chk for chk in chkStr if chk not in chkList2]
|
||||
stop_int = 1
|
||||
break
|
||||
if fmtStr:
|
||||
continue
|
||||
# 教育背景
|
||||
for i in chkList3:
|
||||
if i in text:
|
||||
fmtStr = text
|
||||
nowChkList = [chk for chk in chkStr if chk not in chkList3]
|
||||
stop_int = 1
|
||||
break
|
||||
if fmtStr:
|
||||
continue
|
||||
# 工作经历
|
||||
for i in chkList4:
|
||||
# 判断是不是以关键字开头
|
||||
if not text.startswith(i, 0):
|
||||
continue
|
||||
else:
|
||||
if i in text:
|
||||
fmtStr = text
|
||||
nowChkList = [chk for chk in chkStr if chk not in chkList4]
|
||||
stop_int = 1
|
||||
break
|
||||
if fmtStr:
|
||||
continue
|
||||
# 个人技能
|
||||
for i in chkList5:
|
||||
if i in text:
|
||||
fmtStr = text
|
||||
nowChkList = [chk for chk in chkStr if chk not in chkList5]
|
||||
stop_int = 1
|
||||
break
|
||||
continue
|
||||
else:
|
||||
isTure = 1
|
||||
for i in nowChkList:
|
||||
if i in text:
|
||||
isTure = 0
|
||||
break
|
||||
if isTure:
|
||||
fmtStr += text
|
||||
continue
|
||||
else:
|
||||
fmtStrTrue = fmtStr
|
||||
fmtList.append(fmtStrTrue)
|
||||
trueIndex = index
|
||||
# fmtStr = ''
|
||||
# nowChkList = []
|
||||
# 剩余没有检索的部分
|
||||
txt1 = txt[trueIndex:]
|
||||
return fmtList, txt1, stop_int
|
||||
|
||||
# 当列表全部检索完毕需要停止循环
|
||||
if fmtStr:
|
||||
fmtStrTrue = fmtStr
|
||||
fmtList.append(fmtStrTrue)
|
||||
stop_int = 0
|
||||
txt1 = txt[trueIndex:]
|
||||
return fmtList, txt1, stop_int
|
||||
|
||||
|
||||
def fmtList(txtlist, dates):
|
||||
chkList1 = ['自我评价', '自我描述', '个人优势']
|
||||
chkList2 = ['项目经历', '项目经验', '项目描述']
|
||||
chkList3 = ['教育经历', '学习经历']
|
||||
chkList4 = ['工作经历', '工作经验', '实习经历']
|
||||
chkList5 = ['技能特长', '技能', '特长', '专长', '技能专长', '专业技能', '职业技能']
|
||||
# 自我评价
|
||||
review = []
|
||||
# 项目经验
|
||||
project = []
|
||||
# 工作经验
|
||||
work = []
|
||||
# 教育经验
|
||||
upgrade = []
|
||||
# 技能特长
|
||||
specialty = []
|
||||
for text in txtlist:
|
||||
ischk = 0
|
||||
# 自我评价
|
||||
for i in chkList1:
|
||||
if i in text:
|
||||
review.append(text)
|
||||
ischk = 1
|
||||
break
|
||||
if ischk:
|
||||
continue
|
||||
# 项目经验
|
||||
for i in chkList2:
|
||||
if i in text:
|
||||
project.append(text)
|
||||
ischk = 1
|
||||
break
|
||||
if ischk:
|
||||
continue
|
||||
# 工作经验
|
||||
for i in chkList4:
|
||||
if i in text:
|
||||
work.append(text)
|
||||
ischk = 1
|
||||
break
|
||||
if ischk:
|
||||
continue
|
||||
# 教育经历
|
||||
for i in chkList3:
|
||||
if i in text:
|
||||
upgrade.append(text)
|
||||
ischk = 1
|
||||
break
|
||||
if ischk:
|
||||
continue
|
||||
# 自我评价
|
||||
for i in chkList5:
|
||||
if i in text:
|
||||
specialty.append(text)
|
||||
ischk = 1
|
||||
break
|
||||
if ischk:
|
||||
continue
|
||||
# 取出工作经验里面的公司名和时间
|
||||
work_list = []
|
||||
if len(work) > 0:
|
||||
works = ''
|
||||
for i in work:
|
||||
works += i
|
||||
schema = ['公司名', '时间']
|
||||
ie = Taskflow('information_extraction', schema=schema)
|
||||
text_lists = ie(works)
|
||||
work_list = chkworlkandtime(text_lists)
|
||||
# review自我评价, project项目经验，work工作经验，work具体工作的公司和时间，upgrade教育经历，specialty技能特长
|
||||
dates.update({
|
||||
'review': review,
|
||||
'project': project,
|
||||
'work': work,
|
||||
'work_list': work_list,
|
||||
'upgrade': upgrade,
|
||||
'specialty': specialty,
|
||||
})
|
||||
|
||||
return dates
|
||||
|
||||
|
||||
def get_resume():
|
||||
for root, dirs, files in os.walk(PATH_DATA):
|
||||
for file in files: # 一个file就是一份简历
|
||||
url = PATH_DATA + f"/{file}"
|
||||
if os.path.splitext(file)[1] == '.pdf':
|
||||
pdf_docx(PATH_DATA, file) # 转为docx
|
||||
name = file.split('.')[0]
|
||||
open_txt = docx.Document(PATH_DATA + f"/{name}.docx") # 打开docx
|
||||
os.remove(PATH_DATA + f"/{name}.docx") # 删除生成的文件
|
||||
txt = getText_pdf(url) # 打开pdf格式文件转txt
|
||||
# txt = getText_docx(PATH_DATA + f"\{name}.docx")
|
||||
elif os.path.splitext(file)[1] == '.docx':
|
||||
open_txt = docx.Document(url) # 打开docx，将用来读取每一段的内容
|
||||
txt = getText_docx(url) # 打开docx格式文件转txt
|
||||
elif os.path.splitext(file)[1] == '.doc':
|
||||
doc_docx(PATH_DATA, file) # 转为docx
|
||||
name = file.split('.')[0]
|
||||
open_txt = docx.Document(PATH_DATA + f"/{name}.docx") # 打开docx
|
||||
txt = getText_docx(PATH_DATA + f"/{name}.docx") # 打开docx格式文件转txt
|
||||
os.remove(PATH_DATA + f"/{name}.docx") # 删除生成的文件
|
||||
ie = Taskflow('information_extraction', schema=schema) # 花费时间会安装文件
|
||||
# pprint(ie(txt)) # 姓名，电话，电子邮箱，民族，毕业院校，专业，工作经验，婚姻状况
|
||||
# 获取的基础数据
|
||||
text_lists = ie(txt)
|
||||
# 处理后的基本数据
|
||||
dates = get_date(schema, text_lists, schema_dict)
|
||||
# 打开docx获取的每一段数据
|
||||
txt_list = open_txt.paragraphs
|
||||
# 获取的文档内容
|
||||
txt_list1 = []
|
||||
stop_int = 1
|
||||
txt1 = txt_list
|
||||
while stop_int:
|
||||
txt_list2, txt1, stop_int = fmtTxt(txt1)
|
||||
txt_list1 += txt_list2
|
||||
# print(txt_list1)
|
||||
numTables = open_txt.tables # 获取表格里面的内容
|
||||
table_list = []
|
||||
if len(numTables) > 0:
|
||||
for table in numTables:
|
||||
row_count = len(table.rows)
|
||||
col_count = len(table.columns)
|
||||
for i in range(row_count):
|
||||
for j in range(col_count):
|
||||
texts = table.cell(i, j).text
|
||||
# texts = re.sub('\s+', '', texts).lstrip() # 字符串去除空格和换行符
|
||||
if not texts:
|
||||
continue
|
||||
if texts in table_list:
|
||||
continue
|
||||
table_list.append(texts)
|
||||
if table_list:
|
||||
stop_table = 1
|
||||
table1 = table_list
|
||||
while stop_table:
|
||||
table_list2, table1, stop_table = fmtTxt(table1, istable=1)
|
||||
txt_list1 += table_list2
|
||||
# print(txt_list1)
|
||||
# review自我评价,project项目经验，work工作经验，upgrade教育经历，specialty技能特长
|
||||
# 把两部分的数据合起来返回前端,数据都在dates中
|
||||
fmtList(txt_list1, dates)
|
||||
# pprint(dates)
|
||||
a = 1
|
||||
|
||||
return dates
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
get_resume()
|
||||
Loading…
Reference in New Issue
Block a user