An experiment is a succession of processes. More exactly a directed (potentially cyclic) graph of process. A process has two possible outcomes: either success or failure. The experiment section describes the next process depending on the process outcome.

Example:

[experiment]
  entry = ["reset", ""] #special proc name: say where to start experiment
  reset = ["key_init", ""]
  key_init = ["dut", ""]
  dut = ["", "reset"]

In this experiment, the processes reset, key_init and dut are used. The special process name "entry" is used to point to the starting process (reset in the example).

Then for each process, we have

process_name = ["next process if success", "next process if failure"]

All process names must be valid, i.e. corresponding to processes defined in the "processes" section. Two special names are used:

• "entry" is the process launched at start
• "" (empty string) will interrupt the experiment if called.