hinet_rss.py

from BeautifulSoup import BeautifulSoup as Soup
from mechanize import Browser
from soupselect import select
from datetime import datetime

import sys
reload(sys)
sys.setdefaultencoding("utf-8")


base_url = "http://www.hinet.net/pu"
url = base_url + "/notify.htm"

browser = Browser()
resp = browser.open(url)
content = resp.read()
soup = Soup(content)

DATE_FORMAT = "%a, %d %b %Y %H:%M:%S %Z"

rss_header = r"""
<?xml version="1.0"?>
<rss version="2.0">
<channel>
<title>Hinet unofficial notification rss</title>
<link>http://www.hinet.net/pu/notify.htm</link>
<description>hinet rss, it just works</description>
<language>zh-tw</language>
<pubDate>%(RSS_DATE)s</pubDate>
<lastBuildDate>%(RSS_DATE)s</lastBuildDate>
<docs>http://foo.bar.com</docs>
<generator>Weblog Editor 2.0</generator>
"""

rss_footer = r"""
</channel>
</rss>
"""

rss_item = r"""
<item>
<title>%(TITLE)s</title>
<link>%(LINK)s</link>
<pubDate>%(DATE)s</pubDate>
<description><![CDATA[%(DESCRIPTION)s]]></description>
</item>
"""

built_date = datetime.strftime(datetime.now(), DATE_FORMAT)

print (rss_header % {'RSS_DATE':built_date}).strip()

for link in select(soup, "a"):
    href = link['href']
    if href and href.startswith('notify/'):
        date_str = href.split("_")[1]

        item_title = link.text
        item_link = base_url + "/" + href
        item_pubdate = datetime.strftime(datetime.strptime(date_str, "%Y%m%d"), DATE_FORMAT)

        item_description = str( select(Soup(browser.open(item_link).read()), ".content")[0])
        
        print rss_item % {'TITLE':item_title, 'LINK':item_link, 'DATE':item_pubdate, 'DESCRIPTION':item_description}

print rss_footer