Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:98096
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain hotmail.com designates 65.55.111.84 as permitted sender)
Message-ID: <BLU436-SMTP190C455C4DE10B3343B44E1E24A0@phx.gbl>
Content-Type: multipart/alternative;
	boundary="Apple-Mail=_FCC3A826-507A-4E19-B9BD-DBF550CE0B73"
MIME-Version: 1.0 (Mac OS X Mail 10.1 \(3251\))
Date: Tue, 31 Jan 2017 22:44:05 +0100
In-Reply-To: <a5867a6c-1427-f374-05aa-4e3c0a3876f3@gmx.de>
CC: PHP internals <internals@lists.php.net>
To: "Christoph M. Becker" <cmbecker69@gmx.de>,
 Michael Morris <tendoaki@gmail.com>,
 Larry Garfield <larry@garfieldtech.com>
References: <CAFMT4Nrt7XPMnELoAGLu=sSRCTCtzPCTaSE8HJjKfHjDXOXU_A@mail.gmail.com>
 <E6.ED.51557.B64EF885@pb1.pair.com>
 <3F428CA4-8211-44E6-9B60-62ADB47934B3@koalephant.com>
 <642a72cd-e322-0b22-452f-dfbd521aee02@php.net>
 <BLU436-SMTP4506E841E18C53E37A3CF4E24A0@phx.gbl>
 <37679ae0-eb90-2367-7fc0-8f0adf97a965@garfieldtech.com>
 <CAEUnE0cO04dQ6boek3GrykxXONH3cVuyUGSXUB2z5W0TK34_Lw@mail.gmail.com>
 <a5867a6c-1427-f374-05aa-4e3c0a3876f3@gmx.de>
Subject: Re: [PHP-DEV] [RFC][Discuss] Arrow Functions
From: bobwei9@hotmail.com (Bob Weinand)

--Apple-Mail=_FCC3A826-507A-4E19-B9BD-DBF550CE0B73
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="us-ascii"


> Am 31.01.2017 um 19:41 schrieb Christoph M. Becker =
<cmbecker69@gmx.de>:
>=20
> On 31.01.2017 at 19:16, Michael Morris wrote:
>=20
>> On Tue, Jan 31, 2017 at 12:29 PM, Larry Garfield =
<larry@garfieldtech.com>
>> wrote:
>>=20
>>> My question is why there's no mention of HHVM short closures, or the
>>> previous RFC to take that approach.  See:
>>>=20
>>> https://docs.hhvm.com/hack/lambdas/introduction
>>=20
>> For what it's worth I'd rather look at
>>=20
>> array_map( $x =3D=3D> $x + 1);
>>=20
>> than
>>=20
>> array_map( fn($x) =3D> $x + 1 )
>>=20
>> Not to mention the former isn't a bc break.
>=20
> Neither is supposed to work, though. :)
>=20
> Anyhow, I'm not sure whether I'd prefer
>=20
>  array_map($a, $b, $c =3D=3D> $a + $b + $c, $a1, $a2, $a3)
>=20
> over
>=20
>  array_map(fn($a, $b, $c) =3D> $a + $b + $c, $a1, $a2, $a3)
>=20
> or maybe
>=20
>  array_map(fn $a, $b, $c =3D> $a + $b + $c, $a1, $a2, $a3)
>=20
> Somehow, I tend to like a trailing keyword for better left-to-right
> readability.
>=20
> --=20
> Christoph M. Becker

The RFC states that we're needing some form of a prefix in order to =
avoid implementing gigantic lexer hacks.
HHVM internally implements it by parsing some part of the syntax INSIDE =
the lexer and then emitting a token before the first parameter to tell =
the parser where a short Closure begins.
Having a normal terminal symbol easily allows the parser to detect short =
Closures, without any hacks.

It is just not a practical option from a technical point of view, and =
thus not worth even considering.

Bob=

--Apple-Mail=_FCC3A826-507A-4E19-B9BD-DBF550CE0B73--