Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:107593
Message-ID: <9A432898-EB7A-45C9-9C6F-AE6392B9B5C3@newclarity.net>
Content-Type: multipart/alternative;
	boundary="Apple-Mail=_14712301-679D-489B-8C31-862ECFF2DFF7"
Mime-Version: 1.0 (Mac OS X Mail 12.4 \(3445.104.11\))
Date: Sun, 20 Oct 2019 21:14:01 -0400
In-Reply-To: <CAP=w73U-KF2WF6oaev2tKy6oP+2xJ+JQox8zHdLOeZufj9ZVeQ@mail.gmail.com>
Cc: PHP internals <internals@lists.php.net>
To: Kosit Supanyo <webdevxp.com@gmail.com>
References: <CAP=w73VJ6pxstVzyfFkOp=esX7ne-LosWP1Eb8Ac0u-JeROPRA@mail.gmail.com>
 <C86ED976-5E52-428E-A824-0429601AB3FE@newclarity.net>
 <CAP=w73VNggn6yXnueYwfUS0EbCB6eFMGTy72DRiiTTEWdOVhMQ@mail.gmail.com>
 <3D85D31D-5DDD-42AA-B465-6ADBE72D952A@newclarity.net>
 <CAP=w73U-KF2WF6oaev2tKy6oP+2xJ+JQox8zHdLOeZufj9ZVeQ@mail.gmail.com>
Subject: Re: [PHP-DEV] 'switch-expression' and the 'type guard' unary operator
 demo
From: mike@newclarity.net (Mike Schinkel)

--Apple-Mail=_14712301-679D-489B-8C31-862ECFF2DFF7
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=utf-8

> On Oct 20, 2019, at 7:08 PM, Kosit Supanyo <webdevxp.com@gmail.com> =
wrote:
>=20
> Yes because they are statements, that's the point which I've stated =
many times.

And I am suggesting that a block-enclose expression should not be =
limited to expressions only, but instead be able to use statements in =
the future. Using commas ensures we can never go that route.

> And if your argument is that there's no place in PHP that uses commas =
within curly braces, you're wrong. PHP already has one.
>=20
> use X\Y\Z\{
>     A,
>     B,
>     C,
> };

You have a point.=20

But that use-case is not a block, it is an obscure short-hand syntax for =
which nothing else in PHP is similar. I would not expect we should use =
it as a model unless it were for a very similiar thing.

> if you did do some researches

Seriously?=20

> you might not think they are inconsistent at all. In fact I'd already =
gave you some examples in previous reply (C#, Rust).

I did not state it before because I did not want to spend time on things =
I felt were irrelevant but the fact another language uses syntax does =
not make it appropriate for PHP. What makes it appropriate for PHP is =
"feel", consistency, and how sigils are already being used, and how they =
might be used in the future.

But I am not saying you do not look to other language for inspiration. =
What I am saying is you cannot look to other languages for validation, =
at least not in a vacuum.=20

> No one is limiting anything. ... If you think that anything inside =
curly braces must be separated by semicolon,

Ok, so let's look at a scenario. Will we be able to do this in the =
future, if we decide to allow multiple expressions to result in a final =
value?
$y =3D switch ($x) {
   case 1  =3D> $arr =3D get_array(), array_pop($arr),
   case -1 =3D> $arr =3D get_array(), array_unshift($arr),
   default =3D> null,
};
What if in the future PHP implements variadic syntax for assigning to =
arrays, e.g. $var =3D 1,2,3 where $var could satisfy a $variadic =
parameter set of parameters w/o requiring `...`?   If PHP embraces that =
=E2=80=94 note GoLang has something similar =E2=80=94 then will comma =
syntax allow for it with hacky use of parentheses?
$y =3D  switch ($x) {
   case 1  =3D> $var =3D 1,2,3, $var[1],
   case -1  =3D> $var =3D 9,8,7, $var[1],
   default =3D> null,
};

> That's because `throw` do different thing at the engine level not just =
jumping things like `break`, `continue` or `goto`.

Are you saying it is literally impossible given the current engine, or =
just that it has not been done with `break`, `continue` or `goto`?

> If `break` is allowed in switch-expressions and its behavior is to =
break outer loop, don't you think it could be an another inconsistency =
or confusion? Because I'm 100% sure when people see `break` within =
`switch` they'll all think it should break `switch` not something else.

No.  I would assume the other, so I have to assume I would not be the =
only one.

> you are the one that limit yourself.

Your response makes me think you are responding in anger? I am not =
arguing in anger myself, so I hope that I am not causing you to reply in =
anger? I am merely asking posing technical challenges in hopes to ensure =
that PHP is the best it can be.  I assume we all want that?


-MIke
P.S. And I do not mean my input is needed for PHP to be the best it can =
be but only that having different opinions debated in a reasoned manner =
tends to make for a better outcome.


--Apple-Mail=_14712301-679D-489B-8C31-862ECFF2DFF7--