爬虫大作业（虎扑足球新闻）

Posted 2020-11-01 l_orz

import requests
from bs4 import BeautifulSoup
import jieba
from PIL import Image,ImageSequence
import numpy as np
import matplotlib.pyplot as plt
from wordcloud import WordCloud,ImageColorGenerator
def changeTitleToDict():
    f = open(\'yingchao.txt\', \'r\',encoding=\'utf-8\')
    str = f.read()
    stringList = list(jieba.cut(str))
    symbol = {"/", "（", "）" , " ", "；", "！", "、" ,  "："}
    stringSet = set(stringList) - symbol
    title_dict = {}
    for i in stringSet:
        title_dict[i] = stringList.count(i)
    print(title_dict)
    return title_dict
for i in range(1,10):
        page = i;
        hupu = \'https://voice.hupu.com/soccer/tag/496-%s.html\' % (page)
        reslist = requests.get(hupu)
        reslist.encoding = \'utf-8\'
        soup_list = BeautifulSoup(reslist.text, \'html.parser\')
        for news in soup_list.find_all(\'span\',class_=\'n1\'):
            print(news.text)
            f = open(\'yingchao.txt\', \'a\', encoding=\'utf-8\')
            f.write(news.text)
            f.close()

title_dict = changeTitleToDict()
font = r\'C:\\Windows\\Fonts\\simhei.ttf\'
content = \' \'.join(title_dict.keys())
# 根据图片生成词云
image = np.array(Image.open(\'1.jpg\'))
wordcloud = WordCloud(background_color=\'white\', font_path=font, mask=image, width=1000, height=860, margin=2).generate(content)
#字体颜色
image2 = np.array(Image.open(\'2.jpg\'))
iamge_colors = ImageColorGenerator(image2)
wordcloud.recolor(color_func=iamge_colors)

# 显示生成的词云
plt.imshow(wordcloud)
plt.axis("off")
plt.show()
wordcloud.to_file(\'3.jpg\')

背景图

字体颜色图

 

词云图