Friday, 28 August 2015

Hacking Universal Jobmatch

The Department of Work and Pensions insists that all jobseekers use their online jobsearch website, 'Universal Jobmatch'.

Unfortunately, it can be quite difficult to use, with needing to click through a load of pages filled with mostly agency adverts, which are often duplicates, and sometimes cross-posted to incorrect locations despite the website offering an option to search within a particular radius.

So I thought I'd write a Python script to convert the information to a simple HTML table that one can simply scroll down through without needing to click through 20 or so pages.

I have now tidied it up a bit at put it on Bitbucket: Universal Jobmatch Spam Soup. As I used the Beautiful Soup library and it is written in Python I thought 'Spam Soup' was an appropriate name.

It doesn't really do much to the results, in the way of filtering them beyond what the query returns, but it does make it easier to read I think.


  1. Well done. (Might want to make it a higher default number of days? Removing "&tm={d}" from line 166 removed constant "dead parrot" response for me - with any keyword.) Can see how this could make it easier to read and thus help people, unlike UJ itself.

    Grim, galling subject, so can't say it has lightened my mood though.

    Meur ras dhis.

  2. Does the 'ex-parrot' message appear below the list of jobs, or in place of it?

  3. No worries - tis only in place of the list of jobs - so i think that your program is working correctly. It was just that by my own neglecting to put in a -d parameter of more days, say "-d 5", and then also by happening to search on a Sunday for a specific keyword (when specific 'posts' are unlikely to be advertised), it happened that no adverts with that keyword had been entered in the previous day. So, your script works fine i think. Good starting point !

    PS Da via gweles post genes a-dro dhe'n "gor-loor" !