Watchers are not only good for logging and counting, but they can also be useful for debugging. If an agent or reference is not getting updated the way we expect, we can temporarily add a watcher to track what's happening, so we can see what's going on.
For this recipe, we'll continue the example we've been working with for the last few recipes. This time, instead of counting the data, the watch function will print the change of state to the console.
We'll use the dependencies and requirements that we did in the Managing program complexity with STM recipe. We'll also add a new import:
(import '[java.lang Thread])
Also, we'll use the data file from the Managing program complexity with STM recipe, and the lazy-read-csv and with-header functions from that recipe.
From the Maintaining data consistency with validators recipe, we'll use the functions int-rows and try-read-string.
And from the Tracking processing with watchers recipe, we'll use...