Last friday was hackday at Liip. The Migipedia-Team came together to hack around our Drupal-Setup. This post will explain what we wanted to achive, how we did it and what we have learned.
This is how we process the dataflow from the Migros-backend into the Migipedia frontent just now:
As you can see, we store the data in a “regular” relational database first and then, in a second step, store it to Drupal as an own nodetype defined with CCK. In order to reduce complexity, we now wanted to write a proof of concept to load the main data from a backend using HTTP, allowing us to trim down the Content-Type and to increase the speed of the exporter. Also it has the nice effect, to rip functionallity apart into smaller, easier testable pieces. So this would look something like this:
Bla