Re: [DISCUSSION] Make empty() a Variadic

On 13/02/15 12:31, Andrea Faulds wrote: > >> On 13 Feb 2015, at 11:16, Andrea Faulds <ajf@ajf.me> wrote: >> >> Hey, >> >>> On 13 Feb 2015, at 07:28, Michael Wallner <mike@php.net> wrote: >>> >>> On 12/02/15 19:55, Thomas Punt wrote: >>> >>>> I'd like to propose to make empty() a variadic, where if any >>>> arguments passed in are considered empty, then false is >>>> returned >>> >>> Should that read "if any arguments passed in are considered *NOT* >>> empty, then false is returned”? >> >> No, I think it’s correct, if confusingly phrased. I believe Thomas >> is proposing variadic empty() where TRUE is returned if any of its >> arguments are empty, otherwise FALSE. So, empty($a, $b, $c) would >> be equivalent to empty($a) || empty($b) || empty($c), much like >> isset($a, $b, $c) is equivalent to (and implemented as) isset($a) >> && isset($b) && isset($c). > > Wait, I think I made a mistake. > > * Thomas proposed "if any arguments passed in are considered empty, > then false is returned”, i.e. !(empty($a) || empty($b) || empty($c)) > if his words are taken literally. This doesn’t make much sense, I > think it was a mistake. > * You suggested he may have meant "if any arguments passed in are > considered *NOT* empty, then false is returned”, i.e. (empty($a) && > empty($b) && empty($c)) > * I assume Thomas actually meant “where if any arguments passed in > are considered empty, then *true* is returned”, i.e. (empty($a) || > empty($b) || empty($c)) > > Sorry for the confusion. > > I think the || behaviour is the most useful, as it’s the analogue of > isset’s. So !empty($a, $b, $c) would work similarly to isset($a, $b, > $c), and similarly, !isset($a, $b, $c) would work similarly to > empty($a, $b, $c). > > But that’s just my opinion. :) Okay, while I think it has a tiny WTF attached, because isset() has ALL semantics and empty() would have ANY semantics, it's probably useful only in that way. -- Regards, Mike