python爬取校园集市帖子并生成词云图

注：本篇需要python基础，json基础

前言：上篇我们学习了怎么用python获取百度热搜，在这篇中，我们将进一步学习，利用python爬取校园集市帖子并生成词云图

[toc]
灵感背景：经常在群里看见机器人转发的校园集市帖子，于是想要爬取下来分析一下

第一步，分析请求

点击链接进入浏览器页面

按下F12打开浏览器开发者界面

点击网络，按下F5刷新

刷新后如图所示

点击预览，寻找目标请求

最终找到请求

点击标头，获取url

第二步，编写代码

import requests

#发送http请求
url = 'https://u.xiaouni.com/user-api/content/article/info?id=17020'
response = requests.get(url=url)
#打印结果
print(response.text)

结果如下

{"code":200,"msg":"\u8bf7\u6c42\u6210\u529f","data":{"id":17020,"title":"\u7f8e\u56fd\u8001\u5e74\u5b66\u672f\u4f1a\u8baeGSA","content":"\u6709\u65e0\u5144\u5f1f\u59d0\u59b9\u4e4b\u524d\u53bb\u8fc7\u7f8e\u56fd\u5b66\u672f\u5e74\u4f1aGSA\uff0c\u6c42\u52a9","user_id":16872,"reading":243,"contact_id":16736,"school_id":1,"classify_id":2,"created_at":1718436786,"images":[],"art_like":0,"is_end":0,"is_top":0,"is_rec":0,"status":1,"account":"\u82f9\u679c\u53d1\u7b8d","avatar":"https:\/\/os.xiaouni.com\/uploadsimage\/2023\/09-01\/c3e65afd79583b01.jpg","zone_id":4,"is_like":0,"is_collect":0,"collect_count":0,"comment_count":0,"user":{"id":16872,"is_authority":0,"nickname":"\u5c0f\u7ea2","portrait":"https:\/\/thirdwx.qlogo.cn\/mmopen\/vi_32\/ECkvJVPXNstK1G0syFKuAvko2dF3qibGKLich2mscnYPfiaDylZrLICCG9phvUKbBaDAuk8JgbVibYyyd1AWgs9Qqg\/132","leaver_color":"#FADC3A","leaver_name":"\u5927\u4e00"},"contact":{"id":16736,"money":"0","contact_way_id":2,"contact_info":"3076689048","contact_name":"","contact_way":{"id":2,"name":"QQ"}},"school":{"id":1,"name":"\u534e\u897f"},"classify":{"id":2,"name":"\u4e07\u80fd\u6c42\u52a9","is_anonymous":0}}}

我们发现这个是json格式的内容，而且是以unicode进行编码的结果
我们进一步修改代码，获取帖子标题和内容

import json
import requests

# 发送http请求
url = 'https://u.xiaouni.com/user-api/content/article/info?id=17020'
response = requests.get(url=url)
# json格式化
json_data = json.loads(response.text)
# 获取标题和内容
title = json_data['data']['title']
content = json_data['data']['content']
# 打印标题和内容
print('title:\t'+title,'\ncontent:'+content)

结果如下

1 2	title: 美国老年学术会议GSA content:有无兄弟姐妹之前去过美国学术年会GSA，求助

成功获取集市帖子

第三步，批量获取帖子

我们改进代码，获取最近的300条帖子，写到一个txt文件中方便后续使用

import json

import requests


class getContent:
    def __init__(self, id: int):
        self.url = 'https://u.xiaouni.com/user-api/content/article/info?id=' + str(id)

    def run(self):
        # 发送http请求
        response = requests.get(url=self.url)
        # json格式化
        json_data = json.loads(response.text)
        # 获取状态码，标题和内容
        status_code = json_data['code']
        if status_code == 200:
            title = json_data['data']['title']
            content = json_data['data']['content']
            # 返回标题和内容
            return '['+title+']:'+content+'\n'
        else:
            return 'none'


for i in range(1, 301):
    id = 17020 - i
    print(i)
    text = getContent(id).run()
    if text != 'none':
        with open('content.txt', 'a', encoding='utf-8') as f:
            f.write(text)

结果如下

第四步，绘制词云图

编写一个新的python脚本，将上面获取到的帖子内容绘制成词云图
这步需要安装jieba库（注：这里只是jieba库的简单运用，并未排除一些无意义的词，若要更加出色的表现，可进行jieba库的学习）和wordcloud库

1 2	pip install jieba pip install wordcloud

接下来编写代码

import jieba # 用于文章分词
import wordcloud # 用于将词语生成词云图

# 读取文本
with open("content.txt", encoding="utf-8") as f:
    s = f.read()

# 生成分词列表
ls = jieba.lcut(s)

# 从后向前遍历列表,删除长度小于2的元素
for i in range(len(ls) - 1, -1, -1):
    if len(ls[i]) < 2:
        del ls[i]

text = ' '.join(ls)  # 连接成字符串
stopwords = ["可以", "一个", "有没有"]  # 去掉不需要显示的词
wc = wordcloud.WordCloud(font_path="msyh.ttc", width=2560, height=1600, background_color='white', max_words=300,
                         stopwords=stopwords)  # msyh.ttc电脑本地字体，防止中文乱码
wc.generate(text)  # 生成词云图
wc.to_file("1.png")  # 保存词云图片

打开1.png结果如下 (词语出现频率越高，在词云图中的字就越大)

成功实现词云图绘制