问题
I have a huge list of URLs and my task is to feed them to a python script which should spit out the feed urls if there are any. Is there an API library or code out there that can help?
回答1:
I second waffle paradox in recommending Beautiful Soup for parsing the HTML and then getting the <link rel="alternate"> tags, where the feeds are referenced. The code I usually use:
from BeautifulSoup import BeautifulSoup as parser
def detect_feeds_in_HTML(input_stream):
""" examines an open text stream with HTML for referenced feeds.
This is achieved by detecting all ``link`` tags that reference a feed in HTML.
:param input_stream: an arbitrary opened input stream that has a :func:`read` method.
:type input_stream: an input stream (e.g. open file or URL)
:return: a list of tuples ``(url, feed_type)``
:rtype: ``list(tuple(str, str))``
"""
# check if really an input stream
if not hasattr(input_stream, "read"):
raise TypeError("An opened input *stream* should be given, was %s instead!" % type(input_stream))
result = []
# get the textual data (the HTML) from the input stream
html = parser(input_stream.read())
# find all links that have an "alternate" attribute
feed_urls = html.findAll("link", rel="alternate")
# extract URL and type
for feed_link in feed_urls:
url = feed_link.get("href", None)
# if a valid URL is there
if url:
result.append(url)
return result
回答2:
There's feedfinder:
>>> import feedfinder
>>>
>>> feedfinder.feed('scripting.com')
'http://scripting.com/rss.xml'
>>>
>>> feedfinder.feeds('scripting.com')
['http://delong.typepad.com/sdj/atom.xml',
'http://delong.typepad.com/sdj/index.rdf',
'http://delong.typepad.com/sdj/rss.xml']
>>>
回答3:
I don't know any existing library, but Atom or RSS feeds are usually indicated with a <link> tag in the <head> section as such:
<link rel="alternative" type="application/rss+xml" href="http://link.to/feed">
<link rel="alternative" type="application/atom+xml" href="http://link.to/feed">
Straightforward way would be downloading and parsing these URL's with an HTML parser like lxml.html and getting the href attribute of relevant <link> tags.
回答4:
Depending on how well-formed the information in these feeds are (e.g., Are all the links in the form of http://.../? Do you know if they will all be in href or link tags? Are all the links in the feeds going to be to other feeds? etc.), I'd recommend anything from a simple regex to a straight-up parsing module to extract links from the feeds.
As far as parsing modules go, I can only recommend beautiful soup. Though even the best parser will only go so far--esp in the case I mentioned above, if you can't guarantee all links in the data are going to be links to other feeds; then you have to do some additional crawling and probing on your own.
来源：https://stackoverflow.com/questions/7883581/automatically-extracting-feed-links-atom-rss-etc-from-webpages