Web automation processing "sliding verification code"

Even if the dark clouds dye the sky , The rising sun is still shining somewhere

Verification Code There are many kinds of , We often have SMS verification code 、 Voice verification code 、 Character verification code 、 Image verification code 、 Slide verification code etc. . Today, I mainly share it with my friends in web automated testing in , How to handle sliding verification code ？
First , There are several ways to deal with sliding verification codes ：

• The image processing
• adopt cookie Bypass
• Interface bypass

Today's main sharing is through The image processing Method to handle sliding verification code .
With Douban official website For example , After entering the official website , Click password to log in , After entering the account number and password , Click login , You can jump out of the sliding verification code , Then the manual operation can be successful . Now we use an automated way to operate this series of steps .

The whole process can be divided into the following steps

1. Show captcha
2. Get the picture address of slider and skateboard
3. Download sliders and skateboard pictures
4. Artificial intelligence Match the slider verification code distance (opencv)
5. Scaling and calibration slider offset
6. ActionChains Sliding unlocking
7. Add retry mechanism after sliding failure

Environmental preparation ：

pip install opencv-python

The complete code is as follows

import cv2
from selenium import webdriver
import requests
from time import sleep
from selenium.webdriver import ActionChains
from selenium.webdriver.common.by import By

driver = webdriver.Chrome()
while True:
    #1.  Display sliding verification code 
    driver.get('https://www.douban.com/')
    driver.implicitly_wait(10)
    sleep(3)
    driver.switch_to.frame(driver.find_element_by_xpath('//*[@id="anony-reg-new"]/div/div[1]/iframe'))
    driver.find_element_by_xpath('/html/body/div[1]/div[1]/ul[1]/li[2]').click()
    driver.find_element(By.XPATH,"//*[@id='username']").send_keys("18888888888")
    driver.find_element(By.XPATH,'//*[@id="password"]').send_keys("123456")
    driver.find_element(By.XPATH,'/html/body/div[1]/div[2]/div[1]/div[5]/a').click()
    sleep(2)
    #2.  Get the slider and skateboard to the picture address 
    driver.switch_to.frame(driver.find_element(By.XPATH,'//*[@id="tcaptcha_iframe"]'))
    ##2.1  slider 
    slideBlock = driver.find_element(By.XPATH,'//*[@id="slideBlock"]')
    slideBlock_url = slideBlock.get_attribute("src")
    ##2.2  Skate 
    slideBg = driver.find_element(By.XPATH,'//*[@id="slideBg"]')
    slideBg_url = slideBg.get_attribute("src")
    #print(slideBg_url)
    #print(slideBlock_url)

    #3.  Download sliders and skateboards 
    with open('images/big_images.jpg','wb') as f:
        f.write(requests.get(slideBg_url).content)
        f.close()

    with open('images/small_images.jpg','wb') as f:
        f.write(requests.get(slideBlock_url).content)
        f.close()

    #4.  After the small picture of intelligent matching slider and skateboard overlaps x Distance of 
    big_grag = cv2.imread("images/big_images.jpg",0) #  Load image in grayscale mode 
    small_grag = cv2.imread("images/small_images.jpg",0) #  Load image in grayscale mode )
    res = cv2.matchTemplate(big_grag,small_grag,cv2.TM_CCORR_NORMED) # A match object 
    cv2.imread
    value = cv2.minMaxLoc(res)
    print("value Value ",value)
    x = value[2][0]
    print(x)

    #5.  Scaling and calibration slider offset , The original picture is 680*390, the truth is that 282*162
    x = int(x*282/680)
    py = 31 - int(20*282/680)
    x = x-py
    print(x)

    #6.  adopt ActionChains Sliding unlocking 
    hk_ele = driver.find_element(By.XPATH,'//*[@id="slideBlock"]')
    action = ActionChains(driver) #  Initialize a mouse object 
    action.click_and_hold(hk_ele).perform()  # Hold down the mouse 
    action.drag_and_drop_by_offset(hk_ele,x,0).perform() # Move the slide to the specified coordinates ,0 Represents the ordinate 
    sleep(3)

    ##  If the element does not exist , That means success 
    #7.  Add retry mechanism after sliding failure 
    try:
        driver.find_element(By.XPATH,'//*[@id="captcha_close"]/div')
    except Exception as e:
        break

If you feel good , Give me a compliment , This summer , Let's make progress together .