On Tue, 5 Feb 2019 at 14:29, Nikita Popov <nikita....@gmail.com> wrote:
> On Tue, Feb 5, 2019 at 1:23 PM Girgias <george.bany...@gmail.com> wrote: > >> On Tue, 5 Feb 2019 at 12:22, Nikita Popov <nikita....@gmail.com> wrote: >> >>> [. . .] >>> >> >> I'm all for it but what is the scope of the RFC? >> Is it all core functions, bundled extension functions, or all extension >> functions? >> > > It affects all internal functions using the zpp APIs, which covers pretty > much all core functions, bundled extension functions and third-party > extension functions. > Thanks for the clarification, that's great Also does this means that there will be argument type hinting in core >> functions that >> could be found out via reflection? >> > > Not as a direct result of the proposal, but the RFC does remove the big > blocker for it. Once it lands we need one more change (don't actually > verify arginfo types for internal functions to avoid double type checking), > and then we can start adding the necessary type information. That will also > take some work, as we have many functions, but it's something everyone can > help with. > > Nikita > I suppose if this RFC gets voted (and implemented) there should be enough (maybe I'm wrong) time to also implement this. Really like this RFC even with the potential BC break as there are Static analysis tools (such as PHPStan, Psalm and forgot the third major one) which can help point out potential type errors. Best regards George P. Banyard
