python鐖櫕xpath鑿滈笩鏁欑▼(Python爬虫：选项卡切换和xpath解析)

选项卡切换和xpath解析 京东商品50页数据爬虫

from selenium.webdriver import Chrome, ChromeOptions
from selenium.webdriver.common.keys import Keys
import time, csv
from bs4 import BeautifulSoup

options = ChromeOptions()
options.add_experimental_option('excludeSwitches', ['enable-automation'])
options.add_experimental_option("prefs", {"profile.managed_default_content_settings.images": 2})
# 如果浏览器对象是全局变量，操作结束后浏览器不会自动关闭；如果是局部变量，函数结束后会自动关闭浏览器
b = Chrome(options=options)
b.implicitly_wait(10)

# 创建writer
writer = csv.writer(open('files/运动鞋数据.csv', 'w', encoding='utf-8', newline=''))
writer.writerow(['商品名', '价格', '评论数', '商家', '商品链接', '图片', '商家链接'])


def analysis_data(html: str):
    soup = BeautifulSoup(html, 'lxml')
    all_goods_li = soup.select('.gl-warp>li')
    all_data = []
    for goods in all_goods_li:
        a = goods.select_one('.p-img>a')
        title = a.attrs['title']
        goods_url = 'https:' + a.attrs['href']
        img_dict = a.select_one('img').attrs
        if 'src' in img_dict:
            img_url = 'https:' + img_dict['src']
        else:
            img_url = 'https:' + img_dict['data-lazy-img']
        price = goods.select_one('.p-price i').text
        comment_count = goods.select_one('.p-commit a').text
        shop = goods.select_one('.curr-shop')
        shop_name = shop.text
        shop_url = shop.attrs['href']
        all_data.append([title, price, comment_count, shop_name, goods_url, img_url, shop_url])

    writer.writerows(all_data)
    print('====================数据写入成功====================')


def get_net_data():
    # 1. 打开京东搜索 运动鞋
    b.get('https://www.jd.com')
    search = b.find_element_by_id('key')
    search.send_keys('运动鞋')
    search.send_keys(Keys.ENTER)

    for _ in range(50):
        # 2. 滚动页面
        # height = 1000
        for _ in range(10):
            b.execute_script('window.scrollBy(0,800)')
            time.sleep(1)

        # 3. 获取网页数据，并且解析数据
        analysis_data(b.page_source)
        # print(b.page_source)

        # 4.点击下一页
        b.find_element_by_class_name('pn-next').click()


if __name__ == '__main__':
    get_net_data()

中国知网选项卡切换

from selenium.webdriver import Chrome, ChromeOptions
from selenium.webdriver.common.keys import Keys
from bs4 import BeautifulSoup
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.common.by import By
import time

# 1. 打开中国知网
options = ChromeOptions()
options.add_experimental_option('excludeSwitches', ['enable-automation'])
options.add_experimental_option("prefs", {"profile.managed_default_content_settings.images": 2})
b = Chrome(options=options)
b.implicitly_wait(5)
b.get('https://kns.cnki.net/kns8')

# 2. 搜索相关文献
search = b.find_element_by_id('txt_search')
search.send_keys('人工智能')
search.send_keys(Keys.ENTER)

# 3.获取搜索结果
wait = WebDriverWait(b, 5)
wait.until(EC.presence_of_element_located((By.CLASS_NAME, 'result-table-list')))
search_result = b.find_elements_by_css_selector('.result-table-list .name>a')

# 4. 点击每个搜索结果，获取详情页数据
for a in search_result:
    # 点击每个结果
    a.click()
    time.sleep(1)

    # 切换窗口
    # 浏览器对象.window_handles   -  获取当前浏览器中所有的窗口对象，返回一个列表
    # 切换选项卡: 浏览器对象.switch_to.window(窗口对象)
    b.switch_to.window(b.window_handles[-1])
    print(b.page_source)
    b.close()

    # 切换回原窗口
    b.switch_to.window(b.window_handles[0])

xpath数据解析

python基于xpath做数据解析需要使用的库是lxml，使用时需要导入

from lxml import etree

1.lxml第三方库

作用：lxml既可以解析xml内容也可以解析html内容

xpath相关术语

术语	解释
树	整个网页或者xml文件对应的结构
节点	标签、元素
根节点	最外层的标签
绝对路径	从根节点开始写的路径
相对路径	用.表示当前节点
节点内容(内容)	双标签的标签内容
属性	标签属性

2.xml数据结构

xml数据和json数据一样，是一种通用的数据格式

#json数据示例
supermarket = {
    "name": "永辉超市",
    "address": "肖家河大厦",
    "staff": [
        {"name":},
        {}
        {},
    ]
}


#xml数据示例

    肖家河大厦
    永辉超市
    
        张三
        小明
        小红
        小花
    
    
        
            泡面
            3.5
            120
        
        
            火腿肠
            1.5
            305
        
        
            矿泉水
            1.5
            1200
        
        
            巧克力
            11.5
            50
        
    

3.xpath语法 - 获取标签 1.创建树并且获取根节点

etree.XML(xml数据)    -   创建xml树结构并且返回根节点
etree.HTML(html数据)  -   创建html树结构并且返回根节点

root = etree.XML(open('超市.xml', encoding='utf-8').read())

2.通过路径获取标签

节点对象.xpath(路径)   -  返回值是列表，列表中的元素是通过指定路径找到的所有标签对应的节点对象

三种路径	表示方法
绝对路径	/绝对路径
相对路径	用 . 表示当前节点 (xpath前面是哪个节点，当前节点就是谁)
	用 … 表示当前节点的父节点
任意路径	//路径

# ==================1.绝对路径====================
# 绝对路径一定从根节点开始往下写，和xpath前面是哪个节点无关
result = root.xpath('/supermarket/all_goods/goods/name/text()')
print(result)

all_goods_node = root.xpath('/supermarket/all_goods')[0]
result = all_goods_node.xpath('/supermarket/all_goods/goods/name/text()')
print(result)

# ==================2.相对路径====================
result = root.xpath('./all_goods/goods/name/text()')
print(result)

result = all_goods_node.xpath('./goods/name/text()')
print(result)

result = all_goods_node.xpath('../all_goods/goods/name/text()')
print(result)

goods_list = all_goods_node.xpath('./goods')
for goods in goods_list:
    print(goods.xpath('./name/text()')[0], end=' ')
    print(goods.xpath('./pirce/text()')[0], end=' ')
    print(goods.xpath('./count/text()')[0])

# ==================3.任意路径====================
result = root.xpath('//name/text()')
print(result)       # ['永辉超市', '泡面', '火腿肠', '矿泉水', '巧克力']

result = root.xpath('//goods/name/text()')
print(result)       # ['泡面', '火腿肠', '矿泉水', '巧克力']

3.获取标签内容

获取标签的路径/text()

4.获取标签属性值

获取标签的路径/@属性名

result = root.xpath('//staff/@position')
print(result)       # ['收营员', '售货员', '售货员', '促销员']

staff = root.xpath('//staff')[0]
print(staff.xpath('./@position')[0], staff.xpath('./@class')[0])

5.谓语 - 条件 位置相关的条件

操作	含义
标签名[N]	第N个指定标签
标签名[last()]	最后一个指定标签
标签名[last()-N]	倒数第(N+1)个标签
标签名[position()	获取所有位置值(从1开始)小于N的标签
标签名[position()>N]	获取所有位置值(从1开始)大于N的标签
标签名[position()<=N]	获取所有位置值(从1开始)小于等于N的标签
标签名[position()>=N]	获取所有位置值(从1开始)大于等于N的标签

# 第2个
result = root.xpath('//all_goods/goods[2]/name/text()')[0]
print(result)       # 火腿肠

# 最后一个
result = root.xpath('//all_goods/goods[last()]/name/text()')[0]
print(result)       # 巧克力

# 倒数第2个
result = root.xpath('//all_goods/goods[last()-1]/name/text()')[0]
print(result)       # 矿泉水


result = root.xpath('//all_goods/goods[position()<=3]/name/text()')
print(result)

和属性相关的条件

操作	含义
标签名[@属性名]	获取拥有指定属性的指定标签
标签名[@属性名=值]	获取指定属性为指定的指定标签

result = root.xpath('//goods[@class]/name/text()')
print(result)

result = root.xpath('//goods[@id]/name/text()')
print(result)

# //div[@]   == div.c2
result = root.xpath('//goods[@]/name/text()')
print(result)

和标签内容相关的条件

用子标签的标签内容对父标签进行筛选

操作
标签对象[子标签名>数据]
标签对象[子标签名<数据]
标签对象[子标签名<=数据]
标签对象[子标签名>=数据]
标签对象[子标签名=数据]

# 获取价格大于3的商品的名称
result = root.xpath('//goods[pirce>3]/name/text()')
print(result)       # ['泡面', '巧克力']

# 获取商品名为"巧克力"的商品的名字
result = root.xpath('//goods[name="巧克力"]/name/text()')
print(result)       # ['巧克力']

result = root.xpath('//goods[name="巧克力"]/count/text()')
print(result)

6. 通配符 - *

用*代替任何标签或者任何属性

result = root.xpath('/text()')
print(result)

xpath解析豆瓣电影

import requests
from lxml import etree


def get_html(url):
    headers = {
        'user-agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/99.0.4844.51 Safari/537.36'
    }
    response = requests.get(url, headers=headers)

    # 解析数据
    html = etree.HTML(response.text)
    # 方法1：
    # all_li = html.xpath('//ol[@]/li/div')
    # for div in all_li:
    #     name = div.xpath('./div[2]/div[1]/a/span[1]/text()')[0]
    #     score = div.xpath('./div[2]/div[2]/div/span[2]/text()')[0]
    #     print(name, score)

    # 方法2：
    names = html.xpath('//div[@]/a/span[1]/text()')
    scores = html.xpath('//span[@]/text()')
    # [[name, score], [name2, score2]]
    all_data = list(map(lambda name, score: [name, score], names, scores))
    print(all_data)


if __name__ == '__main__':
    get_html('https://movie.douban.com/top250')

练习：爬取英雄联盟所有英雄及皮肤并装入文件

import requests, json, os, csv


def get_all_skins(url: str):
    headers = {
        'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/99.0.4844.51 Safari/537.36'
    }
    response = requests.get(url, headers=headers)
    all_skins = json.loads(response.text)['skins']
    all_hero_skin = []
    for skin in all_skins:
        skin_name = skin['name']
        all_hero_skin.append([skin_name])

    writer = csv.writer(open(f'{file_path}/皮肤.csv', 'a', encoding='utf-8', newline=''))
    writer.writerows(all_hero_skin)


def get_all_hero():
    url = 'https://game.gtimg.cn/images/lol/act/img/js/heroList/hero_list.js'
    headers = {
        'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/99.0.4844.51 Safari/537.36'
    }
    response = requests.get(url, headers=headers)

    all_hero = json.loads(response.text)['hero']
    for hero in all_hero:
        number = hero['heroId']
        name = hero['title']
        title = hero['alias']
        img_url = f'https://game.gtimg.cn/images/lol/act/img/champion/{title}.png'
        game_skin_url = f'https://game.gtimg.cn/images/lol/act/img/js/hero/{number}.js'
        global file_path
        file_path = os.path.join(r'files', name)
        os.mkdir(file_path)
        get_all_skins(game_skin_url)


if __name__ == '__main__':
    get_all_hero()