Get a weekly digest from a Pelican blog

Posted by Bob on Sat 24 December 2016 in Tools • 2 min read

Prep work

We built this blog in Pelican, adding this in adds an RSS feed:

FEED_RSS = 'feeds/all.rss.xml'

And voila, after pushing this change we have our RSS feed.

Script (use PyPi!)

The script is on github in our new blog repo.

No need to re-invent the wheel, PyPI (Python Package Index) has so much good stuff, feedparser is just what we need. It can take both a remote as well as local xml file, so you don't even need requests.

This single line parses the feed into a comprehensive data structure:

feed = feedparser.parse(xml)

Which you can then easily consume:

for article in feed['entries']:
    # ... filtering
    yield article

The only thing I had to add was some timestamp conversations/calculations to go x days back (the returned feed data has a convenient time.struct_time field).

Mail digest as txt/html in a cronjob

I left this for sendmail which accepts a mailheader, see here. So this is my weekly cronjob:

# html email 
0 7 * * 6 cat pybites_header <(python3 /path/to/pybites_digest/ 7 1) | sendmail -t

# text version for copy+paste into social media (no need to cat header file)
10 7 * * 6 python3 /path/to/pybites_digest/ 7 | mailx -s "Weekly PyBites digest (txt ed)" <email>
  • First arg is "days back" = 7 = one week / 2nd arg = html True

  • You might need to do a export PYTHONPATH=/path/to/python3.x/site-packages if you installed Python3 in your $HOME on a shared hosting provider.

  • The '<()' syntax is a nice way in Unix to join in output from a subprocess.

Don't miss any post

If you want to receive these weekly digests please subscribe to our blog or join our FB group.

See an error in this post? Please submit a pull request on Github.

>>> next(PyBites)