From e420742a4147813c95ac940b59af761769777d14 Mon Sep 17 00:00:00 2001 From: Denis Lehmann Date: Thu, 27 Feb 2020 13:14:28 +0100 Subject: [PATCH] add main.py --- main.py | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) create mode 100644 main.py diff --git a/main.py b/main.py new file mode 100644 index 0000000..b9149e1 --- /dev/null +++ b/main.py @@ -0,0 +1,28 @@ +import bs4 +import feedparser +import urllib.request + +# Get content of a webpage +def get_content(url): + page = urllib.request.Request(url, headers = {'User-Agent': 'Mozilla/5.0'}) + infile = urllib.request.urlopen(page).read() + data = infile.decode('ISO-8859-1') + soup = bs4.BeautifulSoup(data,features = 'html.parser') + return soup + +# Get entries of a RSS feed +def get_entries(url): + feed = feedparser.parse(url) + return feed.entries + +def main(): + entries = get_entries("https://nixos.org/blogs.xml") + for e in entries: + print(e.title) + print(get_content(entries[0].link)) + + + +if __name__ == '__main__': + main() +