Benchmark Pux Router VS Mezon Router

2 min readDec 17, 2020

Hello all! The time I shall compare Pux Route with Mezon Router.

As usual we have two cases:

http server accepts request, launches php script, which handles this request, and then all script data uploads from memory. All following requests are processed in the same way. In this case very critical to launch script as soon as possible and we do not have time for long pre-compilations and preparations. Because all of it will be lost after the script will finish working;
php script is launching, initiating all internal components (and router is one of them) and then starting processing requests. This case can be organized via for example react-php. It differs from the previous case because we can spend reasonable time to pre-compile routes for faster processing.

First case

// static routes
for ($i = 0; $i < \Mezon\Benchmark\Base::$iterationsAmount; $i ++) {
    $router = \Mezon\Benchmark\RouteGenerator::generatePuxStaticRoutes(1000);    $result = $router->dispatch('/static/' . rand(0, 1000 - 1));
    $result[2]();
}// non-static routes
for ($i = 0; $i < \Mezon\Benchmark\Base::$iterationsAmount; $i ++) {
    $router = \Mezon\Benchmark\RouteGenerator::generatePuxNonStaticRoutes(1000);    $result = $router->dispatch('/param/' . rand(0, 1000 - 1) . '/1');
    $result[2]();
}

Second case

// static routes
$router = \Mezon\Benchmark\RouteGenerator::generatePuxStaticRoutes(1000);$startTime = microtime(true);
for ($i = 0; $i < \Mezon\Benchmark\Base::$iterationsAmount; $i ++) {
    $result = $router->dispatch('/static/' . rand(0, 1000 - 1));
    $result[2]();
}
return microtime(true) - $startTime;// non-static routes
$router = \Mezon\Benchmark\RouteGenerator::generatePuxNonStaticRoutes(1000);$startTime = microtime(true);
for ($i = 0; $i < \Mezon\Benchmark\Base::$iterationsAmount; $i ++) {
    $result = $router->dispatch('/param/' . rand(0, 1000 - 1) . '/1');
    $result[2]();
}
return microtime(true) - $startTime;