cajeta.concurrent.Semaphore — counting permit pool that bounds the number of concurrent operations against an unbounded set of consumers. Built on the lock

condition-variable intrinsics: acquire waits for a free permit then takes one, release returns one and wakes a waiter. If a workload is statically batchable into “K at a time”, a chunked scope { spawn K times } expresses the limit in the program’s shape; reach for Semaphore when consumers are unbounded and no scope can serve as the batching boundary.

Semaphore gate = heap Semaphore(2);
gate.acquire();
int32 avail = gate.availablePermits();   // 1
gate.release();

Methods

Signature
`Semaphore(int32 initial)` ⚑	Create a semaphore seeded with `initial` free permits — the cap on concurrent holders
`void acquire()`	Take one permit, blocking (parking the fiber) until one is free
`void release()`	Return one permit, waking a waiter so it can re-check
`void withPermit(() -> void fn)`	Run `fn` while holding a permit, releasing it even if `fn` throws
`int32 availablePermits()`	Current free-permit count (snapshot; advisory under contention)

⚑ = @EntryPoint

Semaphore

Methods

See also