Re: RFC proposal: Spread Operator for Array Destructuring Assignment

On Wed, Jun 3, 2020 at 12:02 PM Côme Chilliet < come.chilliet@fusiondirectory.org> wrote: > Le Wed, 3 Jun 2020 11:46:00 +0200, > Nikita Popov <nikita.ppv@gmail.com> a écrit : > > To give a tl;dr of existing discussions on the topic: > > > > $foobar = [2 => 2, 1 => 1, 0 => 0]; > > [$foo, ...$bar] = $foobar; > > // What is the result? > > > > Introducing this feature is primarily a matter of coming up with a > > satisfactory answer to the above question. (There is an existing > > constraint that $foo must be equal to 0 in this case, not 2.) > > Well, I guess it should behave exactly the same as: > > $foobar = [2 => 2, 1 => 1, 0 => 0]; > [$foo, $foo2, $foo3] = $foobar; > $bar = [$foo2, $foo3]; > > for consistency, no? > To be clear, the result of your semantics would be: $foo = 0; $bar = [1, 2]; Which naturally leads to the question: What happens with string keys? I guess those would just be carried over as-is? (What would the order be though? Are integer keys reindexed before string keys?) The other possible result would be: $foo = 0; $bar = [2 => 2, 1 => 1]; This leaves alone the parts of the array that were not explicitly destructured. Regards, Nikita