I've been thinking about a process for getting press mentions from Google Alerts into the wiki. Here are my initial thoughts on a script/process: I'm thinking we can do this with a script that takes the Alerts RSS feed for a particular language (I set up a bunch at http://venus1-jeb.rhcloud.com/), parses out date, link, title and author (in the goog alerts rss feed, author = publication), and sends out this info to the marketing list, formatted for cutting and pasting into the wiki. A member of the marketing team would then check out the links to see that they are in fact press mentions, take note of the Author of the piece (since the google alert doesn't include this), and paste them into the wiki page. It could happen once per week, and would arrive at the list looking something like this: Hi, for the week of $WEEK, here are the English press mentions from Google Alerts. Please have a look at these links, and paste them into the page at $WHERE if they look good. Take note of the Author of the item, and include that information when you paste it into the wiki. When you're done, please reply to this message on the list so we know it's already handled. |- | Date (YYYY-MM-DD) | [http://www.url-of-article Title of Article] | Publication name | AUTHOR | ENGLISH | Notes |- | Date (YYYY-MM-DD) | [http://www.url-of-article Title of Article] | Publication name | AUTHOR | ENGLISH | Notes *** Other thoughts: 1. I played around a bit w/ feedparser (http://packages.python.org/feedparser/introduction.html), which should do for writing this script. 2. It may be good to record the items harvested in a mysql db, and dedupe them based on URL. 3. The URLs in google alerts come wrapped in google tracking stuff, and that'll have to be removed by the script. 4. The same script could format things a bit differently and send the mentions to the FWN list as a proto-In the News blurb, to help automate that, as well. 5. I figure this can be hosted in openshift. 6. If we wanted, we could populate the "Notes" field with the blurb from the rss entry. 7. Might be cool for the script to go look at the URL and try to figure out the author, and addtl info, perhaps as a future rev enhancement. I'll get going on this, and as soon as I have something to look at, I'll put it in github and send out the link. Regards, Jason -- marketing mailing list marketing@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/marketing
