We export Expect_test_config so that the %expect syntax uses Async, to prevent a confusing situation in which one is using Expect_test_helpers functions, which expect Async to be running, but Async isn't running. Also, we override Expect_test_config.run f so that, if f () raises, run prints the exception rather than raising. Printing works better with the expect-test workflow than an unhandled exception, because there is a .corrected file that one can accept and inspect.
module IO : sig ... endIO monad
val flush : unit -> unit IO.tFlush whatever need to be to get pending output out on file descriptor 0.
val run : (unit -> unit IO.t) -> unitRun an IO operation until completion
Synchronous check that there is no pending output on file description 0. With async, there is no guarantee that on the rhs of a IO.bind (flush ()) ... the output is completely flushed, that's why we need this.
val upon_unreleasable_issue : Expect_test_config.Upon_unreleasable_issue.tupon_unreleasable_issue specifies how to deal with output that should not be released even if it is accepted (e.g. backtraces). The default is `CR.