:- use_module(library(spawn)).
retirement(EmployeeId,Name,HireDate) :-
% start long-running computations in the background
spawn(query_company_ldap(EmployeeId,Employee)),
spawn(query_hr_database(EmployeeId,Job)),
do_something_productive_while_waiting,
% extract desired details from Employee and Job
% (blocks until background computations are done)
Employee = employee(Name,_,_),
Job = job(_,HireDate,_).
some_numbers(PiN,Rand) :-
% start long-running computations in the background
async(calculate_nth_digit_of_pi(21_734,PiN),PiToken),
async(ask_randomorg_for_a_number(Rand), RandToken),
% explicitly block until they're done
await(PiToken),
writeln(PiN), % now it has a value
await(RandToken),
writeln(Rand). % now it has a value too
SWI Prolog has great concurrency primitives. For some common operations, those primitives are too low level. You end up reading through a lot of message queue and threading boilerplate code. This library provides high level concurrency predicates which handle all the messy details.
Our aim is to make this library as easy to use as call/1.
Using SWI-Prolog 6.3 or later:
?- pack_install(spawn).
This module uses semantic versioning.
Source code available and pull requests accepted at http://github.com/mndrix/spawn
| spawn.pl | ||
|---|---|---|
| async/2 | Like async/3 with default options. | |
| async/3 | Seek solutions to Goal in a background thread. | |
| await/1 | Wait for solutions from an async/3 call. | |
| lazy/1 | Postpone execution of goal until needed. | |
| spawn/1 | Like spawn/2 with default options. | |
| spawn/2 | Seek solutions to Goal in a background thread. | |