Re: [RFC][DISCUSSION] debug_backtrace alternative as an array of StackFrame objects

Hi Levi, śr., 8 lip 2020 o 22:25 Levi Morrison <levi.morrison@datadoghq.com> napisał(a): > Any opposition for a parameter to limit stack depth? It's something > I'd like on the old API too. > I've updated the RFC with the same arguments as in debug_backtrace() function. I'm also considering adding a third option for int $offset = 0 in some known cases we want by default skip more frames at the beginning when they're known this would reduce copying values from skipped frames. I've also added two additional properties and related methods: 1) property ?string $object_class = null and method `public function getObjectClass(): ?string` They expose an object class name which is useful in case we don't wanna populate frames with the objects. The additional information could fulfil the https://bugs.php.net/bug.php?id=45351 if frames as objects would be got accepted as a replacement in Exception::getTrace(). 2) property ?\Closure $closure = null and method `public function getClosure(): ?\Closure` They expose a closure which is useful in case the frame was produced by closure call. This was added to keep the function property compatibility with frames produced by debug_backtrace(). The additional information could fulfil the https://bugs.php.net/bug.php?id=62325 if frames as objects would be got accepted as a replacement in Exception::getTrace(). Cheers, -- Michał Marcin Brzuchalski