Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:125469
Feedback-ID: ifab94697:Fastmail
Precedence: bulk
MIME-Version: 1.0
Date: Sat, 07 Sep 2024 15:28:22 +0200
To: "Mike Schinkel" <mike@newclarity.net>, "Davey Shafik" <davey@php.net>,
 "Larry Garfield" <larry@garfieldtech.com>
Cc: "PHP internals" <internals@lists.php.net>
Message-ID: <2bb2b656-aace-462d-b8e3-69d23593c6fd@app.fastmail.com>
In-Reply-To: <71DEAAB9-D788-4EB9-A07F-E47211F92F78@newclarity.net>
References: <0fa39535-f22d-4eba-b4df-90abe39e683a@app.fastmail.com>
 <e73be2b0-faa9-42eb-a9f6-8308cf4f8c94@app.fastmail.com>
 <ed3847bf-af2e-4e33-ac09-66e44696fdeb@app.fastmail.com>
 <79e58673-50ec-461e-a998-736b020e4287@app.fastmail.com>
 <a3cd6d90-cfac-40d5-a59a-c9f6d9dbe5f0@app.fastmail.com>
 <928A2984-6035-4DA6-9EA7-12E85237C270@php.net>
 <71DEAAB9-D788-4EB9-A07F-E47211F92F78@newclarity.net>
Subject: Re: [PHP-DEV] bikeshed: Typed Aliases
Content-Type: multipart/alternative;
 boundary=3932e853f2484be3956cac3d6b98de5e
From: rob@bottled.codes ("Rob Landers")

--3932e853f2484be3956cac3d6b98de5e
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

On Sat, Sep 7, 2024, at 14:42, Mike Schinkel wrote:
> > On Sep 6, 2024, at 4:45 PM, Larry Garfield <larry@garfieldtech.com> =
wrote:
> > Aliases can then be used only in parameter, return, property, and in=
stanceof types.  Extends and implements are out of scope entirely.
>=20
> Is there a strong technical reason why extends and implements should b=
e out of scope?=20
>=20
> There is definite utility for this, to create a local alias in a names=
pace that can be used throughout the namespace rather than having to ref=
er to the external namespace in many different places.
>=20
> > On Sep 6, 2024, at 8:46 PM, Davey Shafik <davey@php.net> wrote:
> > I would very much prefer to either go all in with an Enum-like (whic=
h means that we can hang methods on to the value) or we need to distingu=
ish between type hints for class-likes and type hints for not-class-like=
s (*Bar anyone?).
>=20
> Allowing methods also have definite value as most use-cases I have see=
n in other languages alias in order to add methods, especially for enabl=
ing support of interfaces.
>=20
> Which, however, brings up an important distinction that other language=
s have made and which I think PHP would benefit from addressing:
>=20
> 1. Type Alias =3D> Different Name for Same Type
> 2. Type Def =3D> New Type which has all the same properties and method=
s of other type
>=20
> e.g. (being hypothetical with the syntax; bikeshed away):
>=20
> typealias  LocalWidget: Widget
>=20
> typedef  MyWidget: Widget {
>    function foo() {...}
> }
>=20
> function doSomething(Widget $widget) {...}
>=20
> $w =3D new LocalWidget;
> doSomething($w);           // This works, no problem as LocalWidget =3D=
=3D=3D Widget
>=20
> $w =3D new MyWidget;
> doSomething($w);           // This throws an error as MyWidget !=3D=3D=
 Widget
>=20
> -Mike
>=20
>=20

Hey Mike,

Keep in mind there is already single-class aliasing with well-known beha=
vior for both local and project-wide aliases. Local aliases are done thr=
ough 'use' statements, and project-wide aliases can be created by using =
the `class_alias()` function.

I feel like any aliasing for primitive/intersection/union types should f=
eel like an extension of that for local aliases. For 'project-wide' alia=
ses, I'm open to much more different syntax, like typealias or even 'ali=
as'.

As far as extends/implements go, I plan to keep it the same for simple c=
lass aliases as there is utility there and the RFC already covers this.

=E2=80=94 Rob
--3932e853f2484be3956cac3d6b98de5e
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title><style type=3D"text/css">p.Mso=
Normal,p.MsoNoSpacing{margin:0}</style></head><body><div>On Sat, Sep 7, =
2024, at 14:42, Mike Schinkel wrote:<br></div><blockquote type=3D"cite" =
id=3D"qt" style=3D""><div>&gt; On Sep 6, 2024, at 4:45 PM, Larry Garfiel=
d &lt;<a href=3D"mailto:larry@garfieldtech.com">larry@garfieldtech.com</=
a>&gt; wrote:<br></div><div>&gt; Aliases can then be used only in parame=
ter, return, property, and instanceof types.&nbsp; Extends and implement=
s are out of scope entirely.<br></div><div><br></div><div>Is there a str=
ong technical reason why extends and implements should be out of scope?&=
nbsp;<br></div><div><br></div><div>There is definite utility for this, t=
o create a local alias in a namespace that can be used throughout the na=
mespace rather than having to refer to the external namespace in many di=
fferent places.<br></div><div><br></div><div>&gt; On Sep 6, 2024, at 8:4=
6 PM, Davey Shafik &lt;<a href=3D"mailto:davey@php.net">davey@php.net</a=
>&gt; wrote:<br></div><div>&gt; I would very much prefer to either go al=
l in with an Enum-like (which means that we can hang methods on to the v=
alue) or we need to distinguish between type hints for class-likes and t=
ype hints for not-class-likes (*Bar anyone?).<br></div><div><br></div><d=
iv>Allowing methods also have definite value as most use-cases I have se=
en in other languages alias in order to add methods, especially for enab=
ling support of interfaces.<br></div><div><br></div><div>Which, however,=
 brings up an important distinction that other languages have made and w=
hich I think PHP would benefit from addressing:<br></div><div><br></div>=
<div>1. Type Alias =3D&gt; Different Name for Same Type<br></div><div>2.=
 Type Def =3D&gt; New Type which has all the same properties and methods=
 of other type<br></div><div><br></div><div>e.g. (being hypothetical wit=
h the syntax; bikeshed away):<br></div><div><br></div><div>typealias&nbs=
p; LocalWidget: Widget<br></div><div><br></div><div>typedef&nbsp; MyWidg=
et: Widget {<br></div><div>&nbsp;&nbsp; function foo() {...}<br></div><d=
iv>}<br></div><div><br></div><div>function doSomething(Widget $widget) {=
...}<br></div><div><br></div><div>$w =3D new LocalWidget;<br></div><div>=
doSomething($w);&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n=
bsp; // This works, no problem as LocalWidget =3D=3D=3D Widget<br></div>=
<div><br></div><div>$w =3D new MyWidget;<br></div><div>doSomething($w);&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; // This thro=
ws an error as MyWidget !=3D=3D Widget<br></div><div><br></div><div>-Mik=
e<br></div><div><br></div><div><br></div></blockquote><div><br></div><di=
v>Hey Mike,<br></div><div><br></div><div>Keep in mind there is already s=
ingle-class aliasing with well-known behavior for both local and project=
-wide aliases. Local aliases are done through 'use' statements, and proj=
ect-wide aliases can be created by using the `class_alias()` function.<b=
r></div><div><br></div><div>I feel like any aliasing for primitive/inter=
section/union types should feel like an extension of that for local alia=
ses. For 'project-wide' aliases, I'm open to much more different syntax,=
 like typealias or even 'alias'.<br></div><div><br></div><div>As far as =
extends/implements go, I plan to keep it the same for simple class alias=
es as there is utility there and the RFC already covers this.<br></div><=
div><br></div><div id=3D"sig121229152">=E2=80=94 Rob<br></div></body></h=
tml>
--3932e853f2484be3956cac3d6b98de5e--