爬虫的简单爬取（京东）

python爬取京东

任务步骤

1. 爬取网站信息
2. 筛选出我们要的信息

环境库准备：python基本库、beautifulsoup库、selenium库、pyecharts库

工具：chrome、jupyter notebook

代码实现

任务一：简单爬取京东基本信息

import requests
from bs4 import BeautifulSoup
def getDATA(url):
    headers={'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/93.0.4577.82 Safari/537.36'}
   #User-Agent，浏览器的用户代理字符串，标明身份的第一层标识
   #**如果仅一个User-Agent，也很容易被网站通过相同浏览器频繁访问而识别为爬虫程序，所以一般通过使用多个User-Agent随机调用的方式，避免一个请求头长时间访问。
    source_file=r"C://**"
    try:
        parmas={
            'keyword':'pro'#搜索
        }
        r=requests.get(url=url,headers=headers,params=parmas)
 # url（请求的url地址，必需 ）,headers参数（请求头，可选）, params参数 （请求参数，可选）,proxies参数 （代理IP，可选）,verify参数（ssl证书验证，可选）        
        r.raise_for_status()#查看网络连接情况
        r.encoding='utf-8'#字符编码
        # print(r.text)
        with open (np1,'w',encoding='utf-8') as f1:
            f1.write(r.text)#写入数据
    except Exception as np:
        print(np)
#这个代码中我们用r.raise_for_status()方法，它就可以有效的判断网络连接的状态。如果网连接出现错误，那么它就会用try-except来获取一个异常，若异常出现则输出np
def extractDATA():
    source_file=r"C://**/*"
    compile_file=r"C://***/**"
    soup= BeautifulSoup(open(np1,'r',encoding='utf-8'),'lxml')
    # print(soup.title)
    test=soup.select("ul[class='gl-warp clearfix'] li")
    # print(test)
    with open(compile_file, 'w', encoding='utf-8') as f2:
        for x in test:
            w=x.select("div[class='p-price']")[0].get_text().strip()
            a=x.select("div[class='p-name p-name-type-2']")[0].get_text().strip()
            y = x.select("div[class='p-shop']")[0].get_text().strip()
            k = x.a.attrs['href'].strip()
            s='名称：'+a +'\n价格：'+w+'\n店铺：'+y+'\n链接：'+'http:'+k+'\n----------------------------'
            print(s,'\n')
            f2.write(s)
def main():
    url = ''
    getDATA(url)#调用函数
    extractDATA()
if __name__ == '__main__':
    main()

css修饰标（与上一样）

import re


def getData():
    pass

def extractData(html_path,result_list):
    with open(html_path,'r',encoding='UTF-8') as f:
        html = f.read()

    #print(html)
    r_price = r'<div class="p-price">.*?<i.*?>(.*?)</i>.*?</div>'
    price_list = re.findall(r_price, html, re.S)
    print(price_list)

    r_url = r'<div class="p-name.*?"><a.*?href="(.*?)">.*?</div>'
    url_list = re.findall(r_url, html, re.S)
    url_list = ["https:" + url for url in url_list]
    print(url_list)
    r_desc = r'<div class="p-name.*?">.*?<em>(.*?)</em>.*?</div>'
    desc_list = re.findall(r_desc, html, re.S)
    r_sub = r'<.*?>'
    desc_list = [re.sub(r_sub, '', desc, re.S) for desc in desc_list]
    print(desc_list)
    dengxian = open("C://**", 'w', encoding="utf-8")
    for item in zip(price_list, url_list, desc_list):
        print(item)
        print(item, file=dengxian)
    dengxian.close()

def saveResults():
    pass

def main():
    html_path = r'source_file'
    r_list = list()
    #getData
    extractData(html_path,r_list)

if __name__ == '__main__':
    main()

任务2：自动化爬取京东评论

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

def scrollDown(driver, step=10):
    total_height = int(driver.execute_script("return document.body.scrollHeight"))
    #自动化
    for i in range(1, total_height-1200, step):
        driver.execute_script("window.scrollTo(0, {});".format(i))
        time.sleep(0.005)
def getComments(url):

    driver = webdriver.Chrome()#使用chrmoe
    driver.get(url)#请求url
    driver.maximize_window()#全屏显示
    time.sleep(3)#相应时间

    comment_btn = driver.find_element(By.XPATH,'//li[@data-anchor="#comment"]')
    #点击所需标签（评论）
    comment_btn.click()#跳转
    time.sleep(3)

    scrollDown(driver)
    with open(r'./***','a+',encoding='utf-8')as f:
        for i in range(1,5):#跳转页数
            comments = driver.find_elements(By.CLASS_NAME,'comment-item')
            for comment in comments:
                user_info = comment.find_element_by_class_name('user-info').text
                text_info = comment.find_element(By.CLASS_NAME,'comment-con').text

                star_info = comment.find_element_by_xpath('.//div[@class="comment-column J-comment-column"]/div').get_attribute('class').strip()[-5:]
                days_info = comment.find_element_by_xpath('.//div[@class="order-info"]/span[last()]').text
                xian = comment.find_element_by_class_name('user-level').text
                if xian == "PLUS会员":
                    print('\n'"---------------"'\n名称', user_info, '\n会员状态：', xian, '\n评论', text_info, '\n星级', star_info, '\n评论时间',days_info)
                    f.writelines('\n'"---------------"'\n名称:'+ user_info+ '\n会员状态：'+ xian+'\n评论:'+text_info+'\n星级:'+star_info+'\n评论时间:'+days_info)
                else:
                    print( '\n'"---------------"'\n名称', user_info, '\n会员状态：', "普通会员", '\n评论', text_info, '\n星级', star_info, '\n评论时间',days_info)
                    f.writelines(
                        '\n'"---------------"'\n名称:' + user_info + '\n会员状态：' + "普通会员" + '\n评论:' + text_info + '\n星级:' + star_info + '\n评论时间:' + days_info)

            next_btn = driver.find_element_by_xpath('.//a[@class="ui-pager-next"]')
            #所需点击标签
            next_btn.click()
            time.sleep(3)
            scrollDown(driver)




def main():
    url=r'https://item.jd.com/100021707422.html'
    getComments(url)
if __name__ == '__main__':
    main()

任务3：生成词云

#jieba、pandas用来处理数据，数据源以xls格式存储的，这里用pandas进行处理
from pyecharts.charts import WordCloud
from wordcloud import wordcloud
from selenium import webdriver
import time, jieba, pyecharts
import pyecharts.options as opts
# from selenium.webdriver.common.by import By                                                #方法的引用

def getWord(word_list):       #创建函数
    with open(r'C://**',encoding='utf-8') as f :
        txt = f.read()                                                                                          #读取（评论）文件
        re_move = ["，", "。", " ", '\n', '\xa0',':','*','/d']                                               # 去除无效数据
        for i in re_move:
            txt = txt.replace(i, " ")
            #判断是否存在，若存在则删除
        word_lists = jieba.lcut(txt)
        # 使用精确模式对文本进行分词
        getWordFreq(word_lists, word_count={})
        #通过键值对的形式存储词语及其出现的次数


def getWordFreq(word_list, word_count):
    stopwords = [line.strip() for line in open('A:\\1\评论1.txt', encoding='utf-8').readlines()]                #遍历并返回
    #----readlines() 方法用于读取所有行(直到结束符 EOF)并返回列表。    ----strip() 方法用于移除字符串头尾指定的字符（默认为空格）或字符序列
    #sjstopwords = [line.strip() for line in open('"B:\\3206574001\FileRecv\hit_stopwords.txt"', encoding='utf-8').readlines()]
    for word in word_list: 
        #循环
        if word not in stopwords:
            # if word not in sjstopwords:
                word_count[word] = word_count.get(word, 0) + 1
                #统计词语的出现的次数
        if len(word) == 1:                                                                              # 单个词语不计算在内
            continue
            #跳出
        else:
            word_count[word] = word_count.get(word, 0) + 1
            # 遍历所有词语，每出现一次其对应的值加 1
    items = list(word_count.items())
    #统计次数
    items.sort(key=lambda x: x[1], reverse=True)
    # 根据词语出现的次数进行从大到小排序
    print(items)
    drawWordCLoud(word_count)


def drawWordCLoud(word_count):
    wCloud = WordCloud()
    #color_mask = imread("bg.png")                                                                       # bg.png背景图片
    #d = path.dirname(__file__)                                                                    # 当前文件文件夹所在目录
    # cloud = WordCloud(
    #     font_path="STXINGKA.ttf",                                                                             #华文行楷
    #     font_path=path.join(d,'simsun.ttc'),                                                           #  设置字体  宋体
    #     background_color='white',                                                                           # 设置背景色
    #     mask=color_mask,                                                                                     # 词云形状
    #     max_words=2000,                                                                                    #允许最大词汇
    #     max_font_size=400,                                                               # 最大号字体，如果不指定则为图像高度
    #     max_font_size = 150,                                                       # 画布宽度和高度，如果设置了mask则不会生效
    #     prefer_horizontal = 0.8                                           # 词语水平摆放的频率，默认为0.9.即竖直摆放的频率为0.1
    #     )
    # word_cloud = cloud.generate(cut_text)                                                                     # 产生词云
    wCloud.set_global_opts(title_opts=opts.TitleOpts(title="高频评论词"))                                             #标题

    wCloud.add(
        series_name='评论',
        data_pair=list(word_count.items()),
        shape='pentagon'
    )
    wCloud.render(r'wordCloud.html')                                                                   #构建图片生成网页文件
    print("已成功构建 wordCloud.html 文件，请访问它以查看最终效果。")


def main():
    word_file = open('C:/**', encoding="utf-8")、
    #打开文件（评论文件）
    word_list = word_file.read()
    #  read() 方法用于从文件读取指定的字节数，如果未给定或为负则读取所有
    # print(word_list)
    # word_count = {}
    getWord(word_list)
    #调用
    # getWordFreq(word_list,word_count)


if __name__ == '__main__':                                                                 #main方法
    main()

侵权请联系必删