class poll_process_engine :?period:float -> pid:int -> Unixqueue.event_system ->[Unix.process_status]engine
Shell_uq instead.
This engine waits until the process with the ID pid terminates.
When this happens, the state of the engine changes to
`Done, and the argument of `Done is the process status.
The engine does not catch stopped processes.
The engine checks the process status every period seconds, and
whenever there is a Signal event on the queue. The idea of the
latter is that the user of this engine can increase the responsiveness
by defining a signal handler for SIGCHLD signals (the handler need
not to perform any special action, it must just be defined). When
the sub process terminates, a SIGCHLD signal is sent to the current
process. If the event loop happens to wait for new conditions (which
is usually very likely), a Signal event will be generated, and
the engine will check the process status very soon. Note that it is
not guaranteed that a terminating process triggers a Signal event,
although it is very likely.
You can define an empty SIGCHLD handler with:
Sys.set_signal Sys.sigchld (Sys.Signal_handle (fun _ -> ()))
period : Every period seconds the process status is checked.
Defaults to 0.1 seconds.