Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:109034
To: Rowan Tommins <rowan.collins@gmail.com>, "internals@lists.php.net"
	<internals@lists.php.net>
Thread-Topic: [PHP-DEV] RFC idea: Block scoped variables with "let $x = expr"
Thread-Index: AQHV+ug0CGaQ9aFqj0aNq48f2k+kqahJ7oQAgAACegI=
Date: Sun, 15 Mar 2020 18:17:42 +0000
Message-ID:
 <DM5PR07MB30671D2DEA2C7D2139584593F9F80@DM5PR07MB3067.namprd07.prod.outlook.com>
References:
 <DM5PR07MB3067D2DEAD4B3C85FBDFD559F9F80@DM5PR07MB3067.namprd07.prod.outlook.com>,<77731b0f-70d9-788c-f6bf-7663542aaa7b@gmail.com>
In-Reply-To: <77731b0f-70d9-788c-f6bf-7663542aaa7b@gmail.com>
Accept-Language: en-CA, en-US
Content-Language: en-CA
x-ms-exchange-messagesentrepresentingtype: 1
x-tmn: [Q/pr5f5kmcmAZxBM8sPFHCFMTh5Xl2m30Yp171lLc3R8L3xJ/jmfnTVqkYIk7Yfi]
x-ms-publictraffictype: Email
x-incomingheadercount: 45
x-eopattributedmessage: 0
x-ms-office365-filtering-correlation-id: 9edf67fb-d5ff-479f-9641-08d7c90d2725
x-ms-traffictypediagnostic: BN7NAM10HT094:
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info:
 Y7nb7RzPr9mh/JsWuW62ASMb2hxLcs2YhoNbkvyZHK7L0lFOsu1yr4LIwYfYxodLNiMxkcBeepIhOKW0ufyTdQa6D6ePjESVKMMi2fU0X94Pm00ZwwdvEBUdjt/W8I2Z639iBVvFXkYdrizZW0n2Y2Oer+XDxq5pV0U4bwx/apRSFkj1fFqJS+6V6XkVd/k0oSRDQFPgNRMwR5/WLtaOTg3QfZgwutFba/hG1qOfXl8=
x-ms-exchange-antispam-messagedata:
 4Oy1uV1QAsXs7Z+L3K6LtCrY3929tf+0O7XykhFLnG+pKcwLCBVcIlkokkpdR4Epu6a9Vl5Tex/G/ho82fd9ElYq17Q9oYQgC/ZXUknI8TI5R92C1PIdynAB2hG0sMwJ723nNIoH/I8NN/tgPVUGXLdut3YuyLM18qSpgGT8H9dU0kAkXS/eo8mqKFejsGQ1H8WsYTgSz7bAe9vj/QFsRQ==
x-ms-exchange-transport-forked: True
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-OriginatorOrg: hotmail.com
X-MS-Exchange-CrossTenant-RMS-PersistedConsumerOrg: 00000000-0000-0000-0000-000000000000
X-MS-Exchange-CrossTenant-Network-Message-Id: 9edf67fb-d5ff-479f-9641-08d7c90d2725
X-MS-Exchange-CrossTenant-rms-persistedconsumerorg: 00000000-0000-0000-0000-000000000000
X-MS-Exchange-CrossTenant-originalarrivaltime: 15 Mar 2020 18:17:42.3507
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Internet
X-MS-Exchange-CrossTenant-id: 84df9e7f-e9f6-40af-b435-aaaaaaaaaaaa
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BN7NAM10HT094
Subject: Re: [PHP-DEV] RFC idea: Block scoped variables with "let $x = expr"
From: tysonandre775@hotmail.com (tyson andre)

> Intuitively, I would expect this:=0A=
>=0A=
> {=0A=
> =A0 =A0 =A0let $x =3D new Foo;=0A=
> =A0 =A0 =A0$x->bar();=0A=
> }=0A=
> somethingElse();=0A=
>=0A=
> to be equivalent to this:=0A=
>=0A=
> let $x =3D new Foo;=0A=
> $x->bar();=0A=
> unset($x);=0A=
> somethingElse();=0A=
=0A=
If this feature freed variables when going out of scope,=0A=
it would be compiled more like the following (if there were multiple lets):=
=0A=
=0A=
JavaScript doesn't have destructors, but PHP does, which makes the implemen=
tation a tiny bit more complex.=0A=
=0A=
```=0A=
try {=0A=
=A0 =A0 $x =3D new Foo();=0A=
=A0 =A0 $otherLet =3D new Other();=0A=
=A0 =A0 $x->bar();=0A=
} finally {=0A=
=A0 =A0 // try/finally is useful if unset($otherLet) could throw from __des=
truct=0A=
=A0 =A0 try {=0A=
=A0 =A0 =A0 =A0 unset($otherLet);=0A=
=A0 =A0 } finally {=0A=
=A0 =A0 =A0 =A0 unset($x);=0A=
=A0 =A0 }=0A=
}=0A=
somethingElse();=0A=
```=0A=
=0A=
Actually, thinking about this again, performance is hopefully less of a con=
cern.=0A=
I could safely avoid using `try/finally` if this was outside of a `try` blo=
ck,=0A=
as long as this was in a function scope instead of a global scope.=0A=
(All variable go out of scope for uncaught exceptions)=0A=
This idea is still a work in progress.=0A=
=0A=
It would definitely be useful to free immediately for objects (or arrays wi=
th objects) which call `__destruct`=0A=
when the reference count goes to 0, such as RAII patterns.=0A=
=0A=
My concern about adding an extra `unset()` is that it would add some perfor=
mance overhead,=0A=
discouraging using this feature, but opcache should be able to optimize the=
 unset out when unnecessary.=0A=
=0A=
I touched on the reasons for avoiding `try` blocks earlier, but hopefully t=
hey're not needed.=0A=
The global scope is less likely to be vital to performance in most cases.=
=0A=
=0A=
> - Freeing it immediately would likely require the equivalent of a try{} f=
inally{} to handle exceptions.=0A=
> =A0 try statements prevent opcache from optimizing a function, the last t=
ime I checked.=0A=
=0A=
See https://github.com/php/php-src/blob/43443857b74503246ee4ca25859b302ed0e=
bc078/ext/opcache/Optimizer/dfa_pass.c#L42-L49=