Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:126186
Feedback-ID: i8414410d:Fastmail
Precedence: bulk
MIME-Version: 1.0
Date: Sat, 28 Dec 2024 16:59:29 -0600
To: "php internals" <internals@lists.php.net>
Message-ID: <430e1c1c-4cde-42f2-94f1-d9153239fd70@app.fastmail.com>
In-Reply-To: <087a01db596a$e7525660$b5f70320$@glaive.pro>
References: <087a01db596a$e7525660$b5f70320$@glaive.pro>
Subject: Re: [PHP-DEV] [RFC] [Discussion] Optional interfaces
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From: larry@garfieldtech.com ("Larry Garfield")

On Sat, Dec 28, 2024, at 2:56 PM, Juris Evertovskis wrote:
> Hi,
>=20
> I would like to propose a new syntax that let=E2=80=99s you implement =
an=20
> interface only if it exists.
>=20
> `class MyClass extends ?OptionalInterface {}`
>=20
> If the `OptionalInterface` exists, the class implements it. If no such=20
> interface can be found, it is ignored.
>=20
> https://wiki.php.net/rfc/optional-interfaces=20
>=20
> The need to declare compatibility with a certain interface without=20
> requiring the interface itself comes up when you=E2=80=99re writing co=
de that=20
> should work with various versions of other libraries or PHP itself and=20
> when a library provides interoperability with some other library or a=20
> PHP extension, but does not require it.
>=20
> Although this would mainly be used by library developers, the consumer=
s=20
> of those libraries would also benefit as this syntax would clearly=20
> communicate the intention of optional interoperability. IDEs and stati=
c=20
> analysis tools could also have easier time with `?OptionalInterface`=20
> instead of class definitions inside an `if (interface_exists(=E2=80=A6=
)) { =E2=80=A6 }=20
> else {=E2=80=A6}`.
>=20
> I=E2=80=99ve not entirely sure about the naming as an =E2=80=9Coptiona=
l interface=E2=80=9D=20
> feels like an oxymoron. I=E2=80=99ve also considered names like non-re=
quired=20
> interfaces, conditional interfaces, optional/conditional implementatio=
n=20
> or even soft interfaces/soft implementation. Please let me know if the=20
> naming itself bothers you.
>=20
> Previous discussion: https://externals.io/message/125967
> WIP implementation: https://github.com/php/php-src/pull/17288
>=20
> Looking forward to your feedback and suggestions!
> Juris

Going by the design/RFC, this seems reasonable to me, and I can see the =
use for it.  My only concern would be if the implementation makes anythi=
ng more complex for the engine or compiler that could cause future optim=
ization problems, but I'm not qualified to answer that.

It may help the RFC to include some non-trivial real-world-ish examples,=
 in addition to the A and B stuff.  (You can probably steal some from th=
e projects linked to.)

--Larry Garfield