message-bus

v2

Benchmarks are written with PHPBench and live under benchmarks/. They are not part of mago's checked [source] paths, this is harness/tooling code, not library source, so it isn't held to the same lint/analyze conventions as src/.

Running the benchmarks

composer benchmark

This runs phpbench run --report=aggregate using phpbench.json.dist.

What's measured

  • MiddlewarePipeBench: dispatches a command through MessageBus -> MiddlewarePipe with a single pass-through middleware ahead of the terminal MessageHandlerMiddleware. Handler resolution is stubbed with a static resolver wired to the default HandleStrategy, so the number isolates the cost of pipeline traversal plus the strategy seam (StrategyInterface::handlerMethod(), the is_callable() guard, and dynamic dispatch).

  • MessageBusDispatchBench: the total cost of handling a command end-to-end through a container-backed setup (Laminas ServiceManager), including real MessageHandlerResolver container lookups, at 1 and 5 pass-through middleware.

Latest results

Recorded on 2026-08-13, PHP 8.4.24 (Linux x86_64, WSL2), 1000 revolutions / 5 iterations per subject.

Benchmark Subject Set Mode (μs) RStDev

MiddlewarePipeBench

benchHandle

1 middleware

4.112266

±2.04%

MessageBusDispatchBench

benchHandle

1 middleware

6.089102

±8.65%

MessageBusDispatchBench

benchHandle

5 middleware

7.209593

±2.70%

V2 adds the strategy seam to MessageHandlerMiddleware: resolve the handler, call StrategyInterface::handlerMethod(), guard the name with is_callable(), dispatch through the dynamic method call, and coerce the result to ResultInterface. The 1-middleware dispatch figure carries a wide ±8.65% RStDev, so treat that single delta as indicative rather than precise.

Re-run composer benchmark and update this table after any change likely to affect pipeline or resolver performance.