public final class ParallelScope
extends ObjectConvenience API for running work within a pool-isolated AsyncScope.
ParallelScope combines Pool and AsyncScope
in a single call: the pool is created (or provided), used as the
executor for the scope, and shut down when the scope exits.
This is the Groovy successor to GPars' GParsPool.withPool
and PGroup scoping pattern, modernised for structured
concurrency and virtual threads.
// Groovy:
ParallelScope.withPool(4) { scope ->
def a = scope.async { cpuWork1()
def b = scope.async { cpuWork2() }
[await(a), await(b)]
}
// With a pre-configured pool:
def pool = Pool.cpu()
ParallelScope.withPool(pool) { scope ->
scope.async { work() }
null
}
pool.close()
}| Type Params | Return Type | Name and description |
|---|---|---|
<T> |
public static T |
withPool(int threads, Function<AsyncScope, T> body)Creates a fixed-size pool, executes the body within a scoped context, and shuts down the pool on exit. |
<T> |
public static T |
withPool(Pool pool, Function<AsyncScope, T> body)Executes the body within a scope backed by the given pool. |
Creates a fixed-size pool, executes the body within a scoped context, and shuts down the pool on exit.
threads - the number of threads in the poolbody - the function receiving the scopeT - the result typeExecutes the body within a scope backed by the given pool.
The pool is not shut down on exit — the caller owns the pool's lifecycle. Both AsyncScope.current and Pool.current are bound for the duration.
pool - the pool to use as executorbody - the function receiving the scopeT - the result type