Re: [RFC] [Discussion] Shorter Attribute Syntax Change

I think that verbosity is not a problem if compared to "strange mixed symbols", mainly to new users. Google it is a bit hard "what means double at". And "using attribute" is very clear. Anyway, I think that is valid we use it for now until we have a good symbol arrangement, and on future we could have the "verbose syntax" ("using attribute()") and shortened syntax (to be decided, but like "#[...]"). Atenciosamente, David Rodrigues Em qui., 30 de jul. de 2020 às 11:09, tyson andre <tysonandre775@hotmail.com> escreveu: > Hi David, > > > I would like to suggests the syntax "using attribute(Attribute, ...)". It > > is more clear and should not create BC. > > I'd agree that it's implementable and works with the tokenizer. > > My main objection is the verbosity, which is the reason I assume many > other languages have fairly short/distinct attribute syntax. > > ``` > // Assume Unused tells IDEs/analyzers/linters not to warn about unused > parameters, > // and Override warns about a method not being an override > > using attribute(Override) > public function apiMethod( > using attribute(Unused) > array $params, > using attribute(Unused) > bool $featureEnabled, > using attribute(Unused) > array $options, > ) {...} > ``` > > - Tyson