用Python抓取RSS feed

时间:2012-02-12 21:05:17

标签: python regex rss screen-scraping

我是Python和编程的新手,所以如果问题非常愚蠢,请原谅。

我一直在关注RSS抓取的this教程,但是当我试图收集相应链接到正在收集的文章的标题时,我得到Python的“列表索引超出范围”错误

这是我的代码:

from urllib import urlopen
from BeautifulSoup import BeautifulSoup
import re

source  = urlopen('http://feeds.huffingtonpost.com/huffingtonpost/raw_feed').read()

title = re.compile('<title>(.*)</title>')
link = re.compile('<link>(.*)</link>')

find_title = re.findall(title, source)
find_link = re.findall(link, source)

literate = []
literate[:] = range(1, 16)

for i in literate:
    print find_title[i]
    print find_link[i]

当我只告诉它检索标题时它执行正常,但是当我想要检索标题它们相应的链接时立即抛出索引错误。

非常感谢任何帮助。

2 个答案:

答案 0 :(得分:6)

您可以使用feedparser module to parse an RSS feed from a given url

#!/usr/bin/env python
import feedparser # pip install feedparser

d = feedparser.parse('http://feeds.huffingtonpost.com/huffingtonpost/latestnews')
# .. skipped handling http errors, cacheing ..

for e in d.entries:
    print(e.title)
    print(e.link)
    print(e.description)
    print("\n") # 2 newlines

输出

Even Critics Of Safety Net Increasingly Depend On It
http://www.huffingtonpost.com/2012/02/12/safety-net-benefits_n_1271867.html
<p>Ki Gulbranson owns a logo apparel shop, deals in 
<!-- ... snip ... -->

Christopher Cain, Atlanta Anti-Gay Attack Suspect, Arrested And
Charged With Aggravated Assault And Robbery
http://www.huffingtonpost.com/2012/02/12/atlanta-anti-gay-suspect-christopher-cain-arrested_n_1271811.html
<p>ATLANTA -- Atlanta police have arrested a suspect 
<!-- ... snip ... -->

使用regular expressions to parse rss(xml)可能不是一个好主意。

答案 1 :(得分:1)

我认为您使用错误的正则表达式从页面中提取链接。

>>> link = re.compile('<link rel="alternate" type="text/html" href=(.*)')
>>> find_link = re.findall(link, source)
>>> find_link[1].strip()
'"http://www.huffingtonpost.com/andrew-brandt/the-peyton-predicament-pa_b_1271834.html" />'
>>> len(find_link)
15
>>>

查看您网页的html source,您会发现这些链接未包含在内 <link></link>模式。

实际上模式是<link rel="alternate" type="text/html" href= links here

这就是你的正则表达式无效的原因。