发票查验GUI-基于selenium

1.项目介绍

基于Selenium的GUI发票查询小工具
该工具可以批量的将PDF或PNG和JPG格式的发票信息提取，
执行发票查验，将查询结果图保存到本地。

2.环境依赖

Python3

Chrome

ChromeDerive

3.目录结构

├── Chrome      谷歌浏览器

├── db          存放百度文字识别的token

├── pic         存放验证码图片和分离的色彩图片

├── utils       工具目录

│   ├── Automation.py   爬虫执行步骤

│   ├── BrowerDrive.py  初始化浏览器

│   ├── Dialog.py       对话框

│   ├── TaskEvent.py    任务事件

│   ├── TaskThread.py   任务线程

│   ├── VerifCodeOCR.py 验证码处理

├── v3.py       界面代码

4.项目实现

编写一个基于selenium初始化浏览器对象的一个类

# BrowserDrive.py
from selenium.webdriver import Chrome
from selenium.webdriver.chrome.options import Options


class ChromeDrive:
    browser = None

    def __init__(self, chrome_path, driver_path, stealth_path):
        """
        :param chrome_path: chrome浏览器地址
        :param driver_path: chrome浏览器驱动地址
        :param stealth_path: stealth.min.js文件地址
        """
        options = self.initOptions(chrome=chrome_path)
        self.browser = self.initBrowser(driver=driver_path, options=options, stealth=stealth_path)

    def initOptions(self, chrome):
        """
        :param chrome: 浏览器路径 ...\\chrome.exe
        :return: 浏览器配置
        """
        try:
            # 实例化浏览器选项
            options = Options()
            # 无头模式
            options.add_argument("--headless")
            # 打开页面加速
            options.page_load_strategy = 'eager'
            # 设置忽略ssl证书认证的错误，或者接收不信任的认证
            options.add_argument('ignore-certificate-errors')
            # 关闭 Chrome正受到自动测试软件的控制
            options.add_experimental_option('excludeSwitches', ['enable-automation'])
            # UA
            options.add_argument(
                'user-agent=Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) '
                'Chrome/96.0.46664.45 Safari/537.36'
            )
            # 最大化窗口启动
            options.add_argument('--start-maximized')
            # 设置分辨率
            options.add_argument('window-size=1920x1080')
            # 指定浏览器路径
            options.binary_location = chrome
            return options
        except Exception as e:
            print(e)
            return None

    def initBrowser(self, driver, options, stealth):
        """
        :param driver: 驱动地址 ...\\chromedriver.exe
        :param options: 浏览器配置
        :param stealth: stealth.min.js文件地址
        :return: 浏览器对象
        """
        try:
            # 实例化驱动器
            browser = Chrome(executable_path=driver, options=options)
            # 读取配置文件
            with open(stealth, mode="r", encoding="utf-8") as f:
                js = f.read()
            # 变更配置
            browser.execute_cdp_cmd("Page.addScriptToEvaluateOnNewDocument", {"source": js})
            return browser
        except Exception as e:
            print(e)
            return None

爬虫的步骤不编写了,照着来就行
下面介绍验证码的色彩分离，
图像识别基于的是ddddocr

# VerifyCodeOCR.py
import cv2 as cv
import ddddocr
import os


PIC_DIR = os.path.join(os.getcwd(), 'pic')

CODE_PATH = os.path.join(PIC_DIR, 'code.png')
RED_PATH = os.path.join(PIC_DIR, 'red.png')
YELLOW_PATH = os.path.join(PIC_DIR, 'yellow.png')
BLUE_PATH = os.path.join(PIC_DIR, 'blue.png')
BLACK_PATH = os.path.join(PIC_DIR, 'black.png')


# noinspection PyUnresolvedReferences
def splitColor(filename):
    """
    :param filename: 验证码图片地址
    :return: 分离验证码中的色彩: 黑色, 红色, 蓝色, 黄色
    """
    try:
        # noinspection PyUnresolvedReferences
        img = cv.imread(filename)
        height, width = img.shape[:2]
        ...

        for i in range(height):
            for j in range(width):
                '''取值范围设定'''
                bgr = img[i, j]
                ...
            
        cv.imwrite(RED_PATH, redBlank)
        cv.imwrite(YELLOW_PATH, yellowBlank)
        cv.imwrite(BLUE_PATH, blueBlank)
        cv.imwrite(BLACK_PATH, blackBlank)
    except Exception as e:
        print(e)


def getPicStr(color):
    """
    :param color: 验证码颜色
    :return: 验证码文字
    """
    try:
        ocr = ddddocr.DdddOcr()
        fileName = os.path.join(PIC_DIR, f'{color}.png')
        with open(fileName, 'rb') as fp:
            imageBytes = fp.read()
        result = ocr.classification(imageBytes)
        return result
    except Exception as e:
        print(e)

编写界面部分

# v3.py
import sys
from PyQt5.QtWidgets import QApplication, QWidget, QDesktopWidget, QHBoxLayout, QVBoxLayout, QMessageBox
from PyQt5.QtWidgets import QPushButton, QTableWidget,  QTableWidgetItem, QLabel
from PyQt5.QtGui import QCloseEvent
from utils.TaskEvent import addTaskBtnClickEvent, initTaskBtnClickEvent, execTaskBtnClickEvent, stopTaskBtnClickEvent
from utils.TaskEvent import reInitTaskBtnClickEvent, reExecTaskBtnClickEvent, logonTaskBtnClickEvent, delTaskBtnClickEvent


# 状态变更: 待初始化, 初始化中, 待执行/初始化异常, 执行中, 执行完成/执行异常


class MainWindow(QWidget):

    def __init__(self):
        super().__init__()

        self.tableWidget = None
        self.initUI()

    def closeEvent(self, event: QCloseEvent) -> None:
        reply = QMessageBox.question(
            self,
            "提示",
            "是否要退出？",
            QMessageBox.Yes | QMessageBox.No,
            QMessageBox.No
        )
        if reply == QMessageBox.Yes:
            event.accept()
        else:
            event.ignore()

    def initUI(self):
        # 窗体标题和尺寸
        self.setWindowTitle('增值税发票查询工具')
        # 窗体尺寸
        self.resize(1200, 720)
        # 窗体位置
        qr = self.frameGeometry()
        cp = QDesktopWidget().availableGeometry().center()
        qr.moveCenter(cp)

        # 界面父布局
        mainLayout = QVBoxLayout()

        # 添加顶部菜单栏布局
        mainLayout.addLayout(self.initHeaderLayout())
        # 添加表格布局
        mainLayout.addLayout(self.initTableLayout())
        # 添加底部菜单栏布局
        mainLayout.addLayout(self.initFooterLayout())
        # mainLayout.addStretch()  # 添加弹簧, 用于控件置顶
        # 给窗体设置元素排列方式
        self.setLayout(mainLayout)

    # noinspection PyUnresolvedReferences
    def initHeaderLayout(self):
        # 创建顶部菜单栏布局
        headerLayout = QHBoxLayout()
        # 创建开始按钮和导出按钮
        execBtn = QPushButton("执行")
        initBtn = QPushButton("初始化")
        addBtn = QPushButton("添加")
        stopBtn = QPushButton("终止")
        # 添加点击事件
        addBtn.clicked.connect(lambda: addTaskBtnClickEvent(window=self))
        initBtn.clicked.connect(lambda: initTaskBtnClickEvent(window=self))
        execBtn.clicked.connect(lambda: execTaskBtnClickEvent(window=self))
        stopBtn.clicked.connect(lambda: stopTaskBtnClickEvent(window=self))
        # 将按钮添加到顶部菜单栏布局中
        headerLayout.addWidget(execBtn)
        headerLayout.addWidget(initBtn)
        headerLayout.addWidget(addBtn)
        headerLayout.addWidget(stopBtn)
        headerLayout.addStretch()  # 弹簧
        # 返回布局
        return headerLayout

    
    def initTableLayout(self):
        # 初始化表格布局
        tableLayout = QHBoxLayout()
        # 构建表头参数
        tableHeader = [
            {"field": "fileName", "text": "文件名称", "width": 200},
            {"field": "filePath", "text": "文件路径", "width": 300},
            {"field": "status", "text": "状态", "width": 100},
            {"field": "invoiceCode", "text": "发票代码", "width": 100},
            {"field": "invoiceNum", "text": "发票号码", "width": 100},
            {"field": "invoiceDate", "text": "发票日期", "width": 100},
            {"field": "totalAmount", "text": "不含税金额|校验码", "width": 200},
            {"field": "errorMsg", "text": "异常消息", "width": 200},
        ]
        # 添加表格控件
        tableWidget = QTableWidget(0, len(tableHeader))
        self.tableWidget = tableWidget
        # 循环添加表头
        for idx, info in enumerate(tableHeader):
            item = QTableWidgetItem()
            item.setText(info['text'])

            self.tableWidget.setHorizontalHeaderItem(idx, item)
            self.tableWidget.setColumnWidth(idx, info['width'])

        # 表格添加到表格布局中
        tableLayout.addWidget(tableWidget)
        # 返回表格布局
        return tableLayout

 
    def initFooterLayout(self):
        # 创建底部菜单栏布局
        footerLayout = QHBoxLayout()
        # 创建文本框显示框和按钮
        versionView = QLabel("version: v3", self)
        reInitBtn = QPushButton("重新初始化")
        reExecBtn = QPushButton("重新执行")
        delBtn = QPushButton("删除校验项")
        logonBtn = QPushButton("API配置")
        # 点击事件
        # noinspection PyUnresolvedReferences
        reInitBtn.clicked.connect(lambda: reInitTaskBtnClickEvent(window=self))
        # noinspection PyUnresolvedReferences
        reExecBtn.clicked.connect(lambda: reExecTaskBtnClickEvent(window=self))
        # noinspection PyUnresolvedReferences
        delBtn.clicked.connect(lambda: delTaskBtnClickEvent(window=self))
        # noinspection PyUnresolvedReferences
        logonBtn.clicked.connect(lambda: logonTaskBtnClickEvent(window=self))
        # 将文本框和按钮加载到布局中
        footerLayout.addWidget(versionView)
        footerLayout.addStretch()
        footerLayout.addWidget(reInitBtn)
        footerLayout.addWidget(reExecBtn)
        footerLayout.addWidget(delBtn)
        footerLayout.addWidget(logonBtn)

        # 返回底部菜单栏布局
        return footerLayout


if __name__ == '__main__':
    app = QApplication(sys.argv)
    window = MainWindow()
    window.show()
    sys.exit(app.exec_())

5.参考文章

防止 Selenium 被检测
参考文章: https://blog.csdn.net/cqcre/article/details/110944075

色彩分离
参考文章: https://blog.csdn.net/m0_50616665/article/details/124810344

ddddocr食用方法
参考文章: https://blog.csdn.net/fun_sn/article/details/125421983

使用PyQt5编写GUI
参考文章: https://www.bilibili.com/video/BV1EY4y187yD

在打包过程中，出现了以下2点问题
1. PyQt5打包后运行selenium有黑框
参考文章: https://www.jianshu.com/p/3fadf8f7c203
2. ddddocr打包
参考文章: https://blog.csdn.net/weixin_46010646/article/details/124926207
评论区qq8533的评论