基于python tkinter实现的电子发票识别助手
电子发票识别助手
# -*- coding: utf-8 -*-
# Author:黄成
# Date:2023/10/11
# Function: 一键式实现电子发票识别
# 注意事项: 被识别发票文件需要解密
import pdfplumber
import re
import os
import xlwt
import sys
import datetime
from collections import defaultdict
# 异常开票纳税识别号
invalid_company_ids = [
'91440300359382172R',
'91440300MA5F9JC74N',
'92440300MA5G9MUC9M',
'92440300MA5G9MNE0F',
'92440300MA5G9MMT7U',
'92440300MA5G9QF73F',
'92440300MA5G9TLJ85',
'91440300MA5FKFMBX3',
'91440300MA5FT2LM7U',
'91440300568505124A',
'91440300MA5G2W834G',
'91440300MA5G0D0G7X',
'91440300MA5GAJ7W37',
'91440300MA5FARFG4B',
'91440300596771171E',
'92440300MA5G9YJ387',
'91440300MA5DA9CY1N',
'92440300MA5F40T737',
'91440300MA5H1J5C9T',
'91110114MA04FJXF7A',
'91440300MA5CV8XN57',
'91310110MA1G99QA4Q',
'91440300MA5GUT6524',
'91440300MA5CTUAT1V',
'91440300MA5CTXCM05',
'91440300MA5EF JUP8L',
'91440300MA5F0KF938',
'91510104MA681BEA19'
]
# 创建工作簿
wb = xlwt.Workbook()
# 创建表单
sh = wb.add_sheet('发票信息')
info_titles = ['发票名称','发票代码', '发票号码','开票日期','卖方公司','卖方公司纳税人识别号','收款人','复核','开票人','购方公司','购方公司纳税人识别号','金额','检验结果']
# 写表头
for i in range(len(info_titles)):
sh.write(0,i,info_titles[i])
def re_text(bt, text):
m1 = re.search(bt, text)
if m1 is not None:
return re_block(m1[0])
def re_block(text):
return text.replace(' ', '').replace(' ', '').replace('）', '').replace(')', '').replace('：', ':')
# 购方纳税人识别号检查
def verify_buycompany_ids(inovice_filename, buytax_num):
if buytax_num[7:] != '9144030031977063XH':
print(f"{inovice_filename}公司纳税人识别号错误")
return False
return True
# 卖方纳税人识别号检查
def verify_sellcompany_ids(inovice_filename, selltax_num):
if selltax_num[7:] in invalid_company_ids:
print(f"{inovice_filename}开票方纳税识别号不合规")
return False
return True
# 购方公司名称检查
def verify_company_name(inovice_filename, buy_company_name):
if buy_company_name[3:] != 'XXX公司':
print(f"{inovice_filename}公司名称错误")
return False
return True
# 收集发票错误信息
chk_errors = defaultdict(list)
# 开票人必须填写，复核人和收款人可不填写，复核人不可以与开票人为一人
def verify_people_info(inovice_filename, drawer, reviewer, payee):
if not drawer or not re.findall(re.compile(r'[\u4e00-\u9fa5]+'),drawer):
# 开票人必须填写
chk_errors[inovice_filename].append(" |开票人信息有误| ")
# 复核人和开票人为同一人
if reviewer == drawer:
chk_errors[inovice_filename].append(" |开票人和复核人不可以为同一人| ")
if reviewer and not re.findall(re.compile(r'[\u4e00-\u9fa5]+'),reviewer):
chk_errors[inovice_filename].append(" |复核人名称非法| ")
if payee and not re.findall(re.compile(r'[\u4e00-\u9fa5]+'),payee):
chk_errors[inovice_filename].append(" |收款人名称非法| ")
else:
return True
def verify_expire(inovice_filename, invoicing_date):
# 获取当前日期
now = datetime.datetime.now()
invoicing_date_str = re.sub(r"[\u4e00-\u9fa5]+", r'/', invoicing_date,2).replace('日','')
# 将日期字符串格式化为YYYY/MM/DD格式
date = datetime.datetime.strptime(invoicing_date_str, "%Y/%m/%d")
# 计算差额月数
months_diff = (now.year - date.year) * 12 + (now.month - date.month)
if months_diff > 3:
chk_errors[inovice_filename].append(' |开票日期超过3个月| ')
return False
else:
return True
def invoice_has_noerror(inovice_filename,invoicing_date, buytax_num, selltax_num, buy_company_name,drawer,reviewer,payee):
has_noerror = True
if not verify_people_info(inovice_filename, drawer, reviewer, payee):
has_noerror = False
if not verify_expire(inovice_filename, invoicing_date):
has_noerror = False
if not verify_buycompany_ids(inovice_filename, buytax_num):
has_noerror = False
chk_errors[inovice_filename].append(' |公司纳税人识别号错误| ')
if not verify_sellcompany_ids(inovice_filename, selltax_num):
has_noerror = False
chk_errors[inovice_filename].append(' |开票方纳税人识别号不合规| ')
if not verify_company_name(inovice_filename, buy_company_name):
has_noerror = False
chk_errors[inovice_filename].append('|公司名称错误| ')
return has_noerror
else:
return has_noerror
def print_perinvoice_chkres():
if chk_errors:
for pdffile in chk_errors:
for error in chk_errors[pdffile]:
print(f'!!!Warning :{pdffile} {error}')
def check_invoice_fee(oil_fee, catering_fee, comnicat_fee):
if oil_fee < 500:
print(f'!!!Warning : 交通总费用为{oil_fee},不满足月度报销500额度')
if catering_fee < 680:
print(f'!!!Warning : 餐饮总费用为{catering_fee},不满足月度报销680额度')
if comnicat_fee < 300:
print(f'!!!Warning : 通信总费用为{comnicat_fee},不满足月度报销300额度')
# GUI 界面
import tkinter as tk
from tkinter import filedialog
from tkinter import ttk
from tkinter import scrolledtext
from tkinter import messagebox
def get_pdf(dir_path):
# 定义文件名全局变量
global files
pdf_file = []
for root, dirs, files in os.walk(dir_path):
for name in files:
if name.endswith('.pdf'):
filepath = os.path.join(root, name)
pdf_file.append(filepath)
return pdf_file, files
def choose_folder():
# 创建一个文件夹选择器
global folder_path, pdfiles
folder_path = filedialog.askdirectory()
entry_folder_path.delete(0, tk.END)
entry_folder_path.insert(0,folder_path)
#var_folder_path.set(folder_path)
pdfiles, totalfiles = get_pdf(folder_path)
entry_pdf_info.delete(0, tk.END)
var_pdf_info.set(f'一共发现{len(totalfiles)}个文件, 其中{len(pdfiles)}个PDF文件')
class myStdout(): # 重定向类
def __init__(self):
# 将其备份
self.stdoutbak = sys.stdout
self.stderrbak = sys.stderr
# 重定向
sys.stdout = self
sys.stderr = self
def write(self, info):
# info信息即标准输出sys.stdout和sys.stderr接收到的输出信息
# 在多行文本控件最后一行插入print信息
scrolltext.insert('end', info)
# 更新显示的文本，不加这句插入的信息无法显示
scrolltext.update()
# 始终显示最后一行，不加这句，当文本溢出控件最后一行时，不会自动显示最后一行
scrolltext.see(tk.END)
def restoreStd(self):
# 恢复标准输出
sys.stdout = self.stdoutbak
sys.stderr = self.stderrbak
def read_pdf(folder_path):
# 修改为自己的文件目录
row = 1
coin_sum = 0
oil_cost_sum = 0
catering_cost_sum = 0
comucate_cost_sum = 0
# 进度条初始值
progress_bar['value'] = 0
# 进度条最大值
progress_bar['maximum'] = len(pdfiles)
for pdffile in pdfiles:
# 进度条更新
progress_bar['value'] += 1
# 画面更新
root.update()
print(pdffile)
with pdfplumber.open(pdffile) as pdf:
first_page = pdf.pages[0]
pdf_text = first_page.extract_text()
if '发票' not in pdf_text:
continue
print('--------------------------------------------------------')
fapiaodaima = re_text(re.compile(r'发票代码(.*\d+)'), pdf_text)
fapiaohaoma = re_text(re.compile(r'发票号码(.*\d+)'), pdf_text)
kaipiaoriqi = re_text(re.compile(r'开票日期(.*)'), pdf_text)
nashuishibie = re_text(re.compile(r'纳税人识别号\s*[:：]\s*([a-zA-Z0-9])\s*([a-zA-Z0-9]+)'), pdf_text)
jiaoyan = re_text(re.compile(r'校\s*验\s*码\s*:([a-zA-Z0-9 ]+)'), pdf_text)
fee = re.sub(r'小写.*[¥￥]','',re_text(re.compile(r'小写.*(.*[0-9.]+)'), pdf_text))
shoukuanren = re_text(re.compile(r'收\s*款\s*人[:：]\s*[a-zA-Z0-9\u4e00-\u9fa5]+'), pdf_text)
fuhe = re_text(re.compile(r'复\s*核[:：]\s*[a-zA-Z0-9\u4e00-\u9fa5]+'), pdf_text)
kaipiaoren = re_text(re.compile(r'开\s*票\s*人[:：]\s*[a-zA-Z0-9\u4e00-\u9fa5]+'), pdf_text)
buy_gongsi = re_text(re.compile(r'名\s*称\s*[:：]\s*([\u4e00-\u9fa5]+)'), pdf_text)
invoice_code = fapiaodaima[5:] if fapiaodaima is not None else ""
invoice_number = fapiaohaoma[5:] if fapiaohaoma is not None else ""
invoicing_date = kaipiaoriqi[5:] if kaipiaoriqi is not None else ""
purchasing_company = buy_gongsi[3:] if buy_gongsi is not None else ""
payee = shoukuanren[4:] if shoukuanren is not None else ""
reviewer = fuhe[3:] if fuhe is not None else ""
drawer = kaipiaoren[4:] if kaipiaoren is not None else ""
buyer_company_ids = nashuishibie[7:] if nashuishibie is not None else ""
print(fapiaodaima)
print(fapiaohaoma)
print(kaipiaoriqi)
print(buy_gongsi)
print(nashuishibie)
print(shoukuanren)
print(fuhe)
print(kaipiaoren)
print(f'金额:{fee}')
company = re.findall(re.compile(r'名.*称\s*[:：]\s*([\u4e00-\u9fa5]+)'), pdf_text)
tax_num = re.findall(re.compile(r'纳税人识别号\s*[:：]\s*([a-zA-Z0-9])\s*([a-zA-Z0-9]+)'), pdf_text)
tax_num = list(map(lambda eles: ''.join([ ele for ele in eles]), tax_num))
is_oil = re.findall(re.compile(r'汽.*油'), pdf_text)
is_catering = re.findall(re.compile(r'餐.*饮'), pdf_text)
is_comucate = re.findall(re.compile(r'通.*信'), pdf_text)
# 统计交通费
if is_oil:
oil_cost_sum += float(fee)
# 统计餐饮费
if is_catering:
catering_cost_sum += float(fee)
# 统计通信费
if is_comucate:
comucate_cost_sum += float(fee)
if company:
sell_gongsi = re_block(company[len(company) - 1])
print(f'卖方公司：{sell_gongsi}')
if tax_num:
sell_taxnum = re_block(tax_num[len(tax_num) - 1])
print(f'卖方公司纳税人识别号：{sell_taxnum}')
checkinfo = '发票信息正确'
if not invoice_has_noerror(pdffile, invoicing_date, nashuishibie, sell_taxnum, buy_gongsi,drawer,reviewer,payee):
checkinfo = chk_errors[pdffile]
try:
lst = [pdffile, invoice_code, invoice_number, invoicing_date, sell_gongsi, sell_taxnum, payee, reviewer, drawer,purchasing_company,buyer_company_ids, fee, checkinfo]
except TypeError:
print(">>>>>!!!info : 发票部分信息为空<<<<<")
finally:
# 填写信息入表
for i in range(len(lst)):
sh.write(row, i, lst[i])
row += 1
# 累计金额
coin_sum += float(fee)
print('--------------------------------------------------------')
print('#----------------------检查结果--------------------------#')
print(f'交通费合计：{oil_cost_sum}')
print(f'餐饮费合计：{catering_cost_sum}')
print(f'通信费合计：{comucate_cost_sum}')
print(f'合计金额：{coin_sum}')
check_invoice_fee(oil_cost_sum, catering_cost_sum, comucate_cost_sum)
print_perinvoice_chkres()
sh.write(row, len(info_titles) - 1, f"合计：{coin_sum}")
print('#-------------------------------------------------------#')
def export_to_excel():
current_path = os.path.dirname(os.path.realpath(sys.argv[0]))
savexlsx = '发票信息.xls'
xlspath = '\\'.join((current_path, savexlsx))
print(f"xls存放路径为{xlspath}")
if os.path.exists(xlspath):
os.remove(xlspath)
tk.messagebox.showinfo(title='提示', message=f'{xlspath}导出成功！')
# 保存
wb.save('发票信息.xls')
root = tk.Tk()
# 设置文本框的标题为 "电子发票识别助手"
root.title("电子发票识别助手___Developed By ==> cheng Version:20231026")
root.geometry("1000x400")
# 创建一个标签,用于显示文件夹名称
tk.Label(root, text='选择路径: ', font= 20).place(x=50, y=25)
# 输入框显示文件夹路径
var_folder_path = tk.StringVar() #大模型生成
entry_folder_path = tk.Entry(root, textvariable=var_folder_path, fg='grey',font='15')
entry_folder_path.place(x=150, y=25, width=500, height=30 )
entry_folder_path.insert(0,'请选择文件存储路径：')
tk.Label(root, text='遍历情况: ', font = 20).place(x=50, y=75)
# 输入框显示PDF文件数量
var_pdf_info = tk.StringVar()
entry_pdf_info = tk.Entry(root, textvariable=var_pdf_info, fg='grey', font='15')
entry_pdf_info.place(x=150, y=75, width=500, height=30 )
entry_pdf_info.insert(0,'一共发现0个文件, 其中0个PDF文件')
tk.Label(root, text='进度情况: ', font = 20).place(x=50, y=125)
# 输入框显示PDF识别进度
var_pdf_progress = tk.StringVar()
# 进度条
progress_bar = ttk.Progressbar(root)
progress_bar.place(x=150, y=125, width=500, height=30)
# 创建一个按钮,用于选择文件夹
btn_selectfolder= tk.Button(root, text='浏览遍历文件夹', font = 18, command=choose_folder)
btn_selectfolder.place(x=700, y=23)
btn_selectfolder.config(width='15')
# 创建一个按钮,用于显示PDF文件信息
btn_verifypdf= tk.Button(root, text='识别PDF文件', font = 18, command=lambda:read_pdf(folder_path))
btn_verifypdf.place(x=700, y=73)
btn_verifypdf.config(width='15')
# 创建一个按钮,用于导出excel
btn_exportexcel= tk.Button(root, text='导出excel', font = 18, command=export_to_excel)
btn_exportexcel.place(x=700, y=123)
btn_exportexcel.config(width='15')
# 创建滚动文本框，来显示打印信息
scrolltext = scrolledtext.ScrolledText(root, width=100, height=15, font=('黑体', 10))
scrolltext.place(x=150,y=175)
# 创建一个 Text 对象
my_text = tk.Text(root)
# 设置文本框的文本内容
my_text.delete(1.0, tk.END)
my_text.insert(tk.END, "请确认被检测PDF文件均已被解密!!!")
# 创建一个 Text 对象并将其添加到应用程序中
my_text.place(x=150,y=380)
my_text.config(font=("Courier", 10), fg="red", bg='light grey',width=30)
# 实例化重定向类
mystd = myStdout()
root.mainloop()
# 恢复标准输出
mystd.restoreStd()
扫一扫
1675
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
