from email.Utils import formatdate
import urllib
import re
import time
import datetime

TITLE = "Nemi"
SITELINK = "http://www.dn.se/serier/nemi"

data = urllib.urlopen("http://www.dn.se/serier/nemi").read()

#matchExp = re.compile(r'/polopoly_fs/(.*?)!image/(.*?).gif_gen/derivatives/teaser-small/(.*?).gif')
matchExp = re.compile(r'<a href="/serier/nemi/.*"><img.*?src="/polopoly_fs/(.*?)!image/(.*?).gif_gen/derivatives/teaser-small/(.*?).gif".*?/></a>')
matches = re.findall(matchExp, data)

urls = []
for match in matches:
	urls.append("http://www.dn.se/polopoly_fs/%s!image/%s.gif" % match[0:2])

rssTemplate = """
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<rss xmlns:dc="http://purl.org/dc/elements/1.1/" version="2.0">
<channel>
<title>{TITLE}</title>
<link>{SITELINK}</link>
<description></description>
<language>sv</language>
<pubDate>{PUBDATE}</pubDate>
<lastBuildDate>{PUBDATE}</lastBuildDate>
<docs>http://blogs.law.harvard.edu/tech/rss</docs>
{ITEMS}
</channel>
</rss>
"""

itemTemplate = """
<item>
	<title>{TITLE}</title>
	<link>{LINK}</link>
	<description>
	<![CDATA[
		<img src="{LINK}" />
	]]>
	</description>
	<dc:date>{DATE}</dc:date>
</item>
"""

timestamp = time.time()
items = ""
for url in urls:
	title = str(datetime.date.fromtimestamp(timestamp))
	date = formatdate(timestamp)

	item = itemTemplate.strip()
	item = item.replace("{LINK}", url)
	item = item.replace("{TITLE}", title)
	item = item.replace("{DATE}", date)
	items += item + "\n"

	timestamp -= 86400

output = rssTemplate.strip()
output = output.replace("{TITLE}", TITLE)
output = output.replace("{SITELINK}", SITELINK)
output = output.replace("{PUBDATE}", formatdate(time.time()))
output = output.replace("{ITEMS}", items)

print output