Re: [RFC] Implicit isset() in Shorthand Ternary Operator

On 18 Sep 2014, at 12:32, Pete Boere <pete@the-echoplex.net> wrote: > I'm seeing '??' as analogous to the way JS developers use '||', and I > use > that all the time when writing JS. PHP already has a direct equivalent to ||, namely the ?: operator. However, unfortunately PHP always spits out a notice on non-existent indices, unlike JS which just results in undefined, so we need to add a special operator to silence it. > Personally I wouldn't be interested in a function version because the > short-circuiting of '??' is an important distinction; not something you can > replicate with a function. Therefore having both would be confusing IMO. We could use a function *syntax*, though, but I don’t like this idea. coalesce() and ifsetor() are both ugly to me, and it’s not super obvious what they do: var_dump(coalesce($_GET[‘foobar’], 3)); vs. var_dump($_GET[‘foobar’] ?? 3); ?? also has the advantage of being shorter. > Also, not much sure about a '??=', perhaps it should be a followup RFC > should '??' be accepted. It’d make sense to do it within this RFC if possible. If it seems too controversial, it could always have a separate vote within the RFC. -- Andrea Faulds http://ajf.me/