爬虫知识--03

数据存mysql

import requests
from bs4 import BeautifulSoup
import pymysql

# 链接数据库pymysql
conn = pymysql.connect(
    user='root',
    password="JIAJIA",
    host='127.0.0.1',
    database='cnblogs',
    port=3306,
)
cursor = conn.cursor()
cursor = conn.cursor()

# 爬数据
res = requests.get('https://www.autohome.com.cn/news/1/#liststart')
# print(res.text)

# 引入解析库
soup = BeautifulSoup(res.text, 'html.parser')
ul_list = soup.find_all(name='ul', class_='article')
for ul in ul_list:
    li_list = ul.find_all(name='li')
    for li in li_list:
        h3 = li.find(name='h3')
        if h3:
            title = h3.text
            content = li.find('p').text
            url = 'https:' + li.find(name='a').attrs['href']
            img = li.find('img')['src']
            que = 'https:'
            if que not in img:
                img = que + img
            print('''
            文章标题：%s
            文章摘要：%s
            文章链接：%s
            文章图片：%s
            ''' % (title, content, url, img))

            # 把图片下载到本地
            img_res = requests.get(img)
            with open('./img/' + img.split('/')[-1], 'wb') as f:
                for line in img_res.iter_content():
                    f.write(line)

            # 存入数据库
            cursor.execute("INSERT INTO news (title,content,img,url) VALUES (%s,%s,%s,%s)",
                           args=[title, content, img, url])
            conn.commit()

上传cnblogs图片

import requests

header = {
    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/121.0.0.0 Safari/537.36 Edg/121.0.0.0',
    'Referer': 'https://i.cnblogs.com/ngsw-worker.js',
    'Cookie': '_ga_K9ZHTBGT3Z=GS1.2.1702215735.1.0.1702215735.0.0.0; __gads=ID=5a6b96e7dce231d8:T=1700637208:RT=1706056779:S=ALNI_MYRXZOHhBLLyQzpN1We4NUdfMGYQQ; __gpi=UID=00000c93f4b05ac4:T=1700637208:RT=1706056779:S=ALNI_MatObgQktmbem_0SXLJiayLziRINg; cto_bundle=m0Ggk19mR1BnZTB1dExyaHNvT01CN1dQM0RISU9WUUtIZXVsRW43OXpxJTJCM2ZRMEpncWxtRUpqUiUyRnJ4TkdJQTFlWWdJcWUlMkZub2I4ZWQ1M0w3bk9YRjFmSVozNFBZaFp1MkJEeDhMMjZCc1R4YTFLanNXVVhyNGZIOHJpUGpEREZUJTJGMG9pTmI3YURQWWQ3enlTVVVaY3RoNXNVQSUzRCUzRA; _ga_7DSFGJNPL4=GS1.1.1708417388.1.0.1708417388.0.0.0; Hm_lvt_866c9be12d4a814454792b1fd0fed295=1706056778,1708328454,1708521052; _ga=GA1.1.1528785063.1700619377; .AspNetCore.Antiforgery.b8-pDmTq1XM=CfDJ8DZoAyJmInJHoSwqM1IbzdQjW57WAADzO55EUBOrBFTDFAwXvmddFdsOwcd91uR9v1A23sMz6-hAGYwVrJHEH5c21hYY0xIfJ0-q3IBRlH4fIIoIFR2YMrT0AFNY6FN6LMlo2UEaQP9bnNUglQExnvY; .Cnblogs.AspNetCore.Cookies=CfDJ8DZoAyJmInJHoSwqM1IbzdTqIZnOOklAz_Oo1-ESSjFQVR0Sus_rZ18eCIcP-akb1YjYH-Uic23MF7u9RZHFjwmkhNbiUWC7xbFsYnzTcgu-nLsl5S6ZxcwEU3JGwjspeBFIqJ6kAA52NE1KnJKSkjvgKyfQL4FTAWpP6W7djbWxgFRsMY_eHqjRRS6L0-6dGePy8BFcwCNKm8yfVDN_wgKlsW6AztrWX9UB00sjnrBp0-QmK13o4qf8uLCw4eznsmlPWv_X1F6cNkp1Y6KODCznqgBcsOybycWBKS9vV2uIQ6-36t6HHxqNxL8ChBv73C3rTKfUtSMtLVYtBBE-goLnnqoVTfHqPAdhtj7USyVivYiEcBWAZrEScJ0kJy_9_yyMKiSIgjpOD7aQzzhD1U13tgnOO9ukBjSL-DH8BMODa7dcjfVlkm_8_osfLiwl7Y7oUTlxkIfIhTex2gyUsTIeWuvxRjOG-1kunQwCKpf6r2pZ_zmchTeWtg0hmR8FrHSH0JiT0GiQJZx7RBUZj3RmFfffkHanyu2KGXzMF_ekixBdCpzYAie5TkbQk1uHKoKrAMgdrGBWvRetLEcXfBoDtYgAmw3yDzLavMmVM1Ue; .CNBlogsCookie=4CC9FDFAB5122BA83534E90C476C1A6247EAF22D393555C1F9B6AA6247A9015208A126D736C7FD60A945836247262F06D289A6B0BF8B8B0E2981F8291311DD502E97F3DBD91ECD511D3EAF92566F14674A86FB9B; _ga_3Q0DVSGN10=GS1.1.1708521056.3.1.1708521352.28.0.0; Hm_lpvt_866c9be12d4a814454792b1fd0fed295=1708521353; _ga_M95P3TTWJZ=GS1.1.1708521052.31.1.1708521356.0.0.0; XSRF-TOKEN=CfDJ8DZoAyJmInJHoSwqM1IbzdRGjy4Kotcb_jOERQehTlGIpqZdhn-Zue6X3BLKyWSLj8cJCfqtCnXEFdWPWy1OLD64D1M7CWuHNCnJDZYsyExMi6EjFarOowjQTTuyzlGLPgG-6TnhkfDNGW1xsjqEuzDuINZ_HRV_fC0k_-nPEvGEU_bOuPqJXS3X6F7tGa0DwA; _ga_C2LFP3RFGH=GS1.1.1708521243.1.1.1708521421.0.0.0'
}

files = {
    'imageFile': ("55.png", open('55.png', 'rb'), "image/jpeg"),
    'host': 'www.cnblogs.com',
    'uploadType': 'Paste'
}

res = requests.post('https://upload.cnblogs.com/imageuploader/CorsUpload', headers=header,  files=files)
print(res.text)

selenium介绍

# 原由：由于requests不能执行js

requests会逐个分析 ajax请求，模拟发送获取数据
使用requests爬取的数据很大概率跟在浏览器中看到的不一样
# 介绍：selenium最初是一个自动化测试工具,而爬虫中使用它主要是为了解决requests无法直接执行JavaScript代码的问题

# 本质：selenium是通过驱动浏览器，完全模拟浏览器的操作，比如跳转、输入、点击、下拉等，来拿到网页渲染之后的结果（跟咱们在浏览器看到的完全一致），支持多种浏览器

# 使用：
1、确认好，我们要驱动什么浏览器，我们用谷歌
2、下载一个浏览器驱动：驱动版本跟谷歌浏览器要对应
122.0.6261.58 ----https://googlechromelabs.github.io/chrome-for-testing/
win平台是：chromedriver.exe
3、把驱动放在环境变量下，项目根路径下

4、安装selenium 模块 pip install selenium
5、写代码
from selenium import webdriver
import time

bro = webdriver.Chrome()     # 手动打开了浏览器
bro.get('https://www.baidu.com')    # 在浏览器中输入要访问的网址,并访问
time.sleep(5)
bro.close()    # 关闭浏览器
# 如果是：122.0.6261.57这个版本，安装包如下：
https://storage.googleapis.com/chrome-for-testing-public/122.0.6261.57/win64/chromedriver-win64.zip

selenium模拟登录

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

bro = webdriver.Chrome()   # 手动打开了浏览器

bro.get('https://www.baidu.com')
# 设置等待 10s--->找标签，如果找不到--会最多等待10s钟
bro.implicitly_wait(10)

# 屏幕最大
bro.maximize_window()

# 选择器---》找标签
# submit_btn=bro.find_element(by=By.ID,value='s-top-loginbtn') # 通过id找的
submit_btn = bro.find_element(by=By.LINK_TEXT, value='登录')  # 通过a标签的文字找
submit_btn.click()   # 点击标签

# 点击短信登录---》睡3s钟
sms_submit = bro.find_element(By.ID, 'TANGRAM__PSP_11__changeSmsCodeItem')
sms_submit.click()
time.sleep(3)

# 点击账号登录
username_submit = bro.find_element(By.ID, 'TANGRAM__PSP_11__changePwdCodeItem')
username_submit.click()
time.sleep(3)

# 找到账号和密码---》输入
username = bro.find_element(By.ID, 'TANGRAM__PSP_11__userName')
username.send_keys('19856014525')  # 向输入框中写内容
time.sleep(1)

password = bro.find_element(By.ID, 'TANGRAM__PSP_11__password')
password.send_keys('19856014525')  # 向输入框中写内容
time.sleep(1)

# 点击接受---》点击登录
accept = bro.find_element(By.ID, 'TANGRAM__PSP_11__isAgree')
accept.click()
time.sleep(1)

submit = bro.find_element(By.ID, 'TANGRAM__PSP_11__submit')
submit.click()
time.sleep(5)

bro.close()

无头浏览器

# 概念：无头浏览器是不加载页面也可以执行浏览器中的操作

# 过程：会去环境变量找驱动，找的过程比较慢，但是找到过一次，再运行就快了
# 这个是谷歌浏览器可执行文件的位置(可以不配, 环境变量中有)

指定驱动的位置，具体需要参照文档
# options.binary_location = r"C:\Users\oldboy\AppData\Local\Google\Chrome\Application\chrome.exe"
# options.xx='D:\Python27\crawl_day03\chromedriver.exe'
from selenium import webdriver
from selenium.webdriver.chrome.options import Options

options = Options()
options.add_argument('blink-settings=imagesEnabled=false')  # 不加载图片, 提升速度
options.add_argument('--headless')  # 浏览器不提供可视化页面. linux下如果系统不支持可视化不加这条会启动失败



bro = webdriver.Chrome(options=options)
bro.get('https://www.cnblogs.com/liuqingzheng/p/16005896.html')

print('破解知乎登陆' in bro.page_source)
print(bro.page_source)

搜索标签

# 搜索标签：

By.ID    # 根据id号查找标签
By.NAME    # 根据name属性查找标签
By.TAG_NAME    # 根据标签查找标签
By.CLASS_NAME # 按类名找
By.LINK_TEXT   # a标签文字
By.PARTIAL_LINK_TEXT  # a标签文字，模糊匹配
By.CSS_SELECTOR # 按css选择器找
By.XPATH    #按xpath找

# 标签属性，位置，大小，文本：

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

bro = webdriver.Chrome()
bro.get('https://www.cnblogs.com/liuqingzheng/p/16005896.html')
bro.implicitly_wait(10)
bro.maximize_window()
# bro.find_element() # 找一个
# bro.find_elements() # 找所有

# 1 按id找---找到点赞---》点击它--->使用id找
number=bro.find_element(By.ID,'digg_count')
number.click()

#2 按标签名找   找出页面中所有a标签  按标签名找
a_list=bro.find_elements(By.TAG_NAME,'a')
print(len(a_list))

# 3 按类名找
dig=bro.find_element(By.CLASS_NAME,'diggit')
dig.click()

# 4 按 a 标签 文字找 By.LINK_TEXT
res=bro.find_element(By.LINK_TEXT,'分布式爬虫')
print(res.text)
# print(res.get_attribute('href'))
# res.click()


# 5 a标签文字，模糊匹配  By.PARTIAL_LINK_TEXT
res=bro.find_element(By.PARTIAL_LINK_TEXT,'分布式')
print(res.text)
# print(res.get_attribute('href'))
# res.click()

# 6 css 解析
res=bro.find_element(By.CSS_SELECTOR,'a#cb_post_title_url>span')
res=bro.find_element(By.CSS_SELECTOR,'#cb_post_title_url > span')
print(res.get_attribute('role'))
print(res.text)

# 7 xpath解析--->不会xpath语法
res=bro.find_element(By.XPATH,'//*[@id="cb_post_title_url"]/span')
print(res.get_attribute('role'))
print(res.text)

time.sleep(5)
bro.close()

等待元素

# 1、隐士等待：等待标签加载好

# 原由：在find_element 找标签时，代码执行很快，标签可能还没加载出来，取不到会报错

# 设置隐士等待：   bro.implicitly_wait(10)

        加了这一句，当取标签时，如果标签没加载好，等待最多10s

        等标签加载出来后，找到了继续往后走

# 2、显示等待(不好用)
   每找一个标签，都要给它设置一次等待
# 以后，都在访问到某个地址后，加入这句话即可        bro.implicitly_wait(10)

执行js

bro.execute_script('js代码')

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

bro = webdriver.Chrome()
bro.get('https://www.pearvideo.com/category_1')
bro.implicitly_wait(10)
bro.maximize_window()
# 1 基本使用
bro.execute_script('alert("美女")')

# 2 打印出一些变量
bro.execute_script('console.log(urlMap)')
bro.execute_script('alert(JSON.stringify(urlMap))')


# 3 新建选项卡
bro.execute_script('open()')

# 4 滑动屏幕
bro.execute_script('scrollTo(0,document.documentElement.scrollHeight)')

# 5 获取当前访问地址
bro.execute_script('alert(location)')
bro.execute_script('location="http://www.baidu.com"')

# 6 打印cookie
bro.execute_script('alert(document.cookie)')


time.sleep(10)
bro.close()

选项卡

打开选项卡(通过执行js)：bro.execute_script('open()')
切换到某个选项卡：bro.switch_to.window(bro.window_handles[1])
关闭选项卡：bro.close()
关闭浏览器：bro.quit()

from selenium import webdriver
import time
bro = webdriver.Chrome()
bro.get('https://www.pearvideo.com/')
bro.implicitly_wait(10)
print(bro.window_handles)
# 开启选项卡
bro.execute_script('window.open()')
# 获取出所有选项卡

bro.switch_to.window(bro.window_handles[1]) # 切换到第一个个选项卡
bro.get('http://www.taobao.com')

time.sleep(2)
bro.switch_to.window(bro.window_handles[0]) # 切换到某个选项卡
bro.get('http://www.baidu.com')

time.sleep(2)
bro.execute_script('window.open()')
bro.execute_script('window.open()')
bro.close() # 关闭选项卡

time.sleep(2)
bro.quit()  # 关闭页面

模拟浏览器前进后退

前进：bro.forward()

后退：bro.back()

from selenium import webdriver
import time
bro = webdriver.Chrome()
bro.get('https://www.pearvideo.com/')
bro.implicitly_wait(10)

# 获取出所有选项卡
time.sleep(2)
bro.get('http://www.taobao.com')

time.sleep(2)

bro.get('http://www.baidu.com')
time.sleep(2)
bro.back()
time.sleep(2)
bro.back() # 前进
time.sleep(2)
bro.forward()  # 后退
bro.quit()  # 关闭页面

自动登录cnblogs

import time
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
import json
from selenium.webdriver.common.by import By

# 去掉自动化软件控制的检测
options = Options()
options.add_argument("--disable-blink-features=AutomationControlled")  # 去掉自动化控制
bro = webdriver.Chrome(options=options)


bro.get('https://www.cnblogs.com/')
bro.implicitly_wait(10)    # 等待十秒
bro.maximize_window()  # 屏幕最大化
login_btn = bro.find_element(By.LINK_TEXT, '登录')
login_btn.click()     # selenium自动点击登录按钮

time.sleep(2)

# 找到用户名和密码输入框
username = bro.find_element(By.CSS_SELECTOR, '#mat-input-0')
password = bro.find_element(By.ID, 'mat-input-1')

submit_btn = bro.find_element(By.CSS_SELECTOR,
                              'body > app-root > app-sign-in-layout > div > div > app-sign-in > app-content-container > div > div > div > form > div > button')
# 验证码
code=bro.find_element(By.ID,'rectMask')
time.sleep(1)


username.send_keys('糖果爱上我125')
time.sleep(1)
password.send_keys('20020308Zjq.')
time.sleep(1)
submit_btn.click()  # 一种情况直接登录成功   一种情况会弹出验证码
code.click()
time.sleep(10)

# 让程序先停在这---》手动操作浏览器---》把验证码搞好---》程序再继续往下走
# 到现在，是登录成功的状态
# 取出cookie存起来
cookies = bro.get_cookies()
with open('cnblogs.json', 'w', encoding='utf-8') as f:
    json.dump(cookies, f)

time.sleep(2)
bro.close()