Automatic tag generation for UltimateTagWarrior with Python
Saturday, November 25th, 2006After importing categories as tags, I realized that the tag cloud at the bottom of the site was pretty lonely. Whatever is a programmer to do who has access to the underlying database for the posts and the tags? Hmmmmmm. Last night, I started putting together a python script to parse all of the posts of the site and generate a word-frequency list. Essentially, it finds all of the words of the site that meet the following requirements:
- A frequency of 10
- >3 letters. We don’t care about short words.
- Aren’t in a banned list of words that occur in regular speech like “it, was, green, dog, blue, friday, one”. The list is pretty big (and configurable).
From that exported word-frequency list, I selected 80 tags for my site which I preconfigured as the official tag list. From the official tag list, the appropriate wp_tags and wp_tag2post entries are created for tags that occur in each post, and in a matter of milliseconds, python has created them in the database.
Check out the tag cloud at the bottom. Pretty, eh?
This was all made possible because of mysqldb, a python implementation of the mysql libraries in python’s standard database API. In English: every database type in python inherits from a common API. That means that the code I wrote in python would be the same if it were for sqlite or posgresql.
Let me know if you want the python script. It worked wonders for me 
November 25th, 2006 at 4:21 pm
Nice post and do send me a copy of the script as it has given my name a official tag on your site.
I don’t think I will use it but reviewing and learning from it will be more funny anyway.
- Michiel
December 8th, 2006 at 3:20 pm
I was happy for your marriage (even though that is small) that Bethany figures so prominently. And, it was good to see a BIG Joomla! and a smaller Drupal. Your code is socially acceptable.
December 29th, 2006 at 4:23 am
Please send me a copy.
March 29th, 2007 at 6:21 pm
Hi Brent,
I don’t know Python, but I’m wondering if you could nonetheless send me the script, as I am trying to get something similar going for my blog, albeit with “simple tagging”, PHP, and mySQL.
Thanks a lot,
Tobias
May 26th, 2008 at 2:46 am
Hey I definitely can use a script like this for my site. Though I’m not that well versed in python I definitely can find a way to make it work for me!!