After the F2F meeting, we start to put everything together and work at full scale. The first thing is to define what’s the outcome of the this summer. While our goal is to build a system that can run some queries on the workflow traces or the workflow itself and we already built a RPQ engine, so we sketch the following diagram as a overall design of the system.
We sensed that the user generally won’t care about how the system works, they prefer to treat it as a black box, only if there is something wrong, they will look into the box itself. So we decide to start with a system that takes minimal number of inputs and outputs a visualized result that is easier to look into. By the end of this week, the system could take .dot file and .xml file as input, along with a query you want to ask, then highlight the result on the original graph formatted as a pdf file. In more detail, we extend the existing RPQ engine and integrate together with other formatting mappers into one bash file and call different programs. I’ll demonstrate to our group next Tuesday with the Hamming workflow trace example and show how the result looks like.
Next week, I’ll try to run more examples on the system. Also interestingly, I’ll take a look at how to define RPQ/4 and write down a set of rules in datalog for that.