This Ingredient for tasty testing framework allows to filter a test tree
depending on an outcome of the previous run. This may be useful in many
scenarios, especially when a test suite grows large.

For example, tasty-rerun allows:

* Rerun only tests, which failed during the last run (--rerun). Combined
  with live reloading (e. g., using ghcid or stack test --file-watch), it
  gives an ultimate power to focus on broken parts and put them back in
  shape, enjoying a tight feedback loop.

* Rerun only tests, which have beed added since the last saved test
  run. This comes handy when writing a new module, which does not affect
  other parts of the system, or adding new test cases.

* Rerun only tests, which passed during the last saved test run. Sometimes
  a part of the test suite is consistently failing (e. g., an external
  service is temporarily down), but you want be sure that you are not
  breaking anything else in course of your work.