爬虫-selenium

selenium是一个web的自动化测试工具 可直接运行在浏览器上 支持所有
主流的浏览器 可以接收指令 让浏览器自动加载页面 获取需要的数据 甚至
页面截取
chromedriver 是一个驱动Chrome浏览器的驱动程序

下载网址：
https://developer.aliyun.com/mirror/NPM?from=tnpm

选择windows 下载

根据chrome版本选择下载
解压后放到python3.6 目录下

selenium学习.py

from selenium import webdriver
driver = webdriver.Chrome()
import time

# 打开百度
driver.get('https://www.baidu.com')
# 窗口最大化
driver.maximize_window()

time.sleep(3)
# 关闭当前的窗口
driver.close()

time.sleep(3)

# 退出驱动关闭所有窗口
driver.quit()

定位元素.py

# @ Time : 2021/3/1 17:40
# @ Author : Ellen

from selenium import webdriver
from selenium.webdriver.common.by import By


driver = webdriver.Chrome()
driver.get('https://www.baidu.com')
# 1 id 通过send_keys('python')来输入内容
# driver.find_element_by_id('kw').send_keys('python')
# driver.find_element(By.ID, 'kw').send_keys('java')

# class_name
# driver.find_element_by_class_name('s_ipt').send_keys('jerry')
# driver.find_element(By.CLASS_NAME, 's_ipt').send_keys('kugu')

# name
# driver.find_element_by_name('wd').send_keys('feng')
# driver.find_element(By.NAME, 'wd').send_keys('ELLEN')

# 通过标签名字去定位 用得不多
# head = driver.find_elements_by_tag_name('head')
# print(head)

# 通过xpath来定位  //*[@id="kw"]
# driver.find_element_by_xpath('//input[@id="kw"]').send_keys('php')
# driver.find_element_by_xpath('//*[@id="kw"]').send_keys('hao')

# 通过css选择器
# driver.find_element_by_css_selector('.s_ipt').send_keys('zhang')

inputTag = driver.find_elements_by_tag_name('input')
print(len(inputTag))

操作下拉菜单.py

# @ Time : 2021/3/1 21:04
# @ Author : Ellen
from selenium import webdriver
from selenium.webdriver.common.by import By
import time
from selenium.webdriver.support.ui import Select

driver = webdriver.Chrome()
driver.get('https://www.17sucai.com/pins/demo-show?id=5926')

# 切换iframe
driver.switch_to_frame(driver.find_element_by_id('iframe'))

selectTag = Select(driver.find_element_by_class_name('nojs'))

# selectTag = driver.find_element_by_class_name('nojs')
# driver.find_elements_by_name('country-wrap')
# 选择方式
# 1 根据值来选择
# selectTag.select_by_value('JP')

# 2 根据索引值来选择
selectTag.select_by_index(3)

操作非select标签.py

# @ Time : 2021/3/1 21:37
# @ Author : Ellen

from selenium import webdriver
from selenium.webdriver.common.by import By
import time
from selenium.webdriver.support.ui import Select

driver = webdriver.Chrome()
driver.get('https://www.17sucai.com/pins/demo-show?id=5926')

# 切换iframe  id = 'dk_container_country-nofake'
# driver.switch_to_frame(driver.find_element_by_id('iframe'))
driver.switch_to.frame(driver.find_element_by_id('iframe'))

# selectTag = Select(driver.find_element_by_id('dk_container_country-nofake'))
divTag = driver.find_element_by_id('dk_container_country-nofake').click()
# selectTag.select_by_index(3)   # 报错 AttributeError: 'WebElement' object has no attribute 'select_by_index'

key = int(input('请输入:'))

if key == 1:
    driver.find_element_by_xpath('//*[@id="dk_container_country-nofake"]/div/ul/li[1]/a').click()
elif key == 2:
    driver.find_element_by_xpath('//*[@id="dk_container_country-nofake"]/div/ul/li[2]/a').click()
    

以上代码分析网址如下：
https://www.17sucai.com/boards/53562.html



模拟登录豆瓣.py

# @ Time : 2021/3/1 22:10
# @ Author : Ellen

from selenium import webdriver
import time

driver = webdriver.Chrome()
driver.get('https://www.douban.com/')

# 切换iframe
login_iframe = driver.find_element_by_xpath('//div[@class="login"]/iframe')
driver.switch_to.frame(login_iframe)

time.sleep(2)
# 切换登录方式
# driver.find_element_by_xpath('/html/body/div[1]/div[1]/ul[1]/li[2]')
driver.find_element_by_xpath('//li[@class ="account-tab-account"]').click()

# 输入账号和密码
driver.find_element_by_id('username').send_keys('xxxxx')
driver.find_element_by_id('password').send_keys('xxxxx')

# 点击按钮  不能有空格 要删除前面的btn
driver.find_element_by_class_name('btn-account').click()

行为链.py

# @ Time : 2021/3/1 22:55
# @ Author : Ellen

# https://selenium-python.readthedocs.io/api.html

from selenium import webdriver
from selenium.webdriver import ActionChains

driver = webdriver.Chrome()
driver.get('https://www.baidu.com')
# 定位输入框
inputTag = driver.find_element_by_id('kw')
# 定位百度按钮
summitBth = driver.find_element_by_id('su')
# 实例化对象
actions = ActionChains(driver)
# 把鼠标移动到输入框
actions.move_to_element(inputTag)
# 输入内容
actions.send_keys_to_element(inputTag, 'python')

# 点击按键百度一下
actions.move_to_element(summitBth)
actions.click()

# 点击右键
actions.context_click()

# 提交行为链
actions.perform()