python_day7爬虫之selenium

1.破解极验滑动验证

from selenium import webdriver
from selenium.webdriver import ActionChains
from PIL import Image
import time
import random

# 截取图片函数
def cut_image(driver):
    driver.save_screenshot('image.png')

    img = driver.find_element_by_class_name('geetest_canvas_img')

    left = img.location['x']
    upper = img.location['y']

    right = left + img.size['width']
    lower = upper + img.size['height']

    image_obj = Image.open('image.png')
    img_cut = image_obj.crop((left, upper, right, lower))
    # img_cut.show()
    return img_cut


# 得到完整图片
def get_image1(driver):

    js_code = '''
        var x = document.getElementsByClassName('geetest_canvas_fullbg')[0].style.display="block";
        console.log(x)
        '''

    time.sleep(0.2)
    driver.execute_script(js_code)

    time.sleep(0.2)
    img1 = cut_image(driver)

    return img1

# 得到缺口图片
def get_image2(driver):
    js_code = '''
        var x = document.getElementsByClassName('geetest_canvas_fullbg')[0].style.display="none";
        console.log(x)
        '''

    time.sleep(0.2)
    driver.execute_script(js_code)

    time.sleep(0.2)
    img2 = cut_image(driver)

    return img2

# 对比图片，得到滑动距离
def get_distance(img1, img2):

    # 小滑块右侧位置
    start_location = 60

    # 像素差
    threshold = 60

    for i in range(start_location, img1.size[0]):
        for j in range(img1.size[1]):

            # 获得点(i,j)处的RGB像素值
            rgb1 = img1.load()[i, j]
            rgb2 = img2.load()[i, j]
            r = abs(rgb1[0] - rgb2[0])
            g = abs(rgb1[1] - rgb2[1])
            b = abs(rgb1[2] - rgb2[2])

            if not (r < threshold and g < threshold and b < threshold):
                # 有误差，减去7
                return i-7

def get_tracks(distance):
    distance += 20
    v0 = 0
    a_list = [3, 4, 5]
    t = 0.2
    s = 0

    # 向前滑动轨迹
    forward_tracks = []
    mid = distance * 3 / 5
    while s < distance:
        if s < mid:
            a = a_list[random.randint(0, 2)]
        else:
            a = -a_list[random.randint(0, 2)]

        v = v0
        stack = v * t + 0.5 * a * (t ** 2)

        # 每次拿到的位移
        stack = round(stack)
        s += stack
        v0 = v + a * t

        forward_tracks.append(stack)

    back_tracks = [-1, -1, -2, -3, -2, -3, -2, -2, -3, -1]

    return {'forward_tracks': forward_tracks, 'back_tracks': back_tracks}


def main(user, pwd):

    driver = webdriver.Chrome()
    driver.maximize_window()

    try:
        driver.implicitly_wait(10)
        driver.get('https://account.cnblogs.com/signin?returnUrl=https%3A%2F%2Fwww.cnblogs.com%2F')

        username = driver.find_element_by_id('LoginName')
        username.send_keys(user)
        time.sleep(1)

        password = driver.find_element_by_id('Password')
        password.send_keys(pwd)
        time.sleep(1)

        submit = driver.find_element_by_id('submitBtn')
        submit.click()
        time.sleep(1)

        # 获取完整图片
        img1 = get_image1(driver)

        # 获取缺口图片
        img2 = get_image2(driver)

        # 获取移动距离
        distance = get_distance(img1, img2)

        # 获取滑动轨迹，模拟人的滑动轨迹
        tracks = get_tracks(distance)

        button = driver.find_element_by_class_name('geetest_slider_button')
        ActionChains(driver).click_and_hold(button).perform()

        for forward_track in tracks['forward_tracks']:
            ActionChains(driver).move_by_offset(xoffset=forward_track, yoffset=0).perform()
        time.sleep(0.2)

        for back_track in tracks['back_tracks']:
            ActionChains(driver).move_by_offset(xoffset=back_track, yoffset=0).perform()

        ActionChains(driver).move_by_offset(xoffset=-3, yoffset=0).perform()
        ActionChains(driver).move_by_offset(xoffset=3, yoffset=0).perform()
        time.sleep(0.1)

        ActionChains(driver).release().perform()
        time.sleep(3)
    finally:
        driver.close()

if __name__ == '__main__':

    user = '********'
    pwd = '****************'
    main(user, pwd)