Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:129766
Feedback-ID: i8414410d:Fastmail
Precedence: list
MIME-Version: 1.0
Date: Fri, 16 Jan 2026 11:49:05 -0600
To: "php internals" <internals@lists.php.net>
Message-ID: <fafd2dea-63d4-471f-8f18-0b7b65e8dfd2@app.fastmail.com>
In-Reply-To: <352B0914-68EE-480B-895F-D7CB767A15B3@gmail.com>
References: <e906b61e-b951-4eb4-aa0f-700e76dae6a9@app.fastmail.com>
 <c8877bbc-b82b-4514-9651-b7dfb83b9a35@app.fastmail.com>
 <352B0914-68EE-480B-895F-D7CB767A15B3@gmail.com>
Subject: Re: [PHP-DEV] [RFC] isReadable/isWriteable property reflection
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From: larry@garfieldtech.com ("Larry Garfield")

On Wed, Jan 14, 2026, at 2:00 PM, Claude Pache wrote:
>> Le 13 janv. 2026 =C3=A0 22:30, Larry Garfield <larry@garfieldtech.com=
> a =C3=A9crit :
>>=20
>> On Wed, Nov 5, 2025, at 5:24 PM, Larry Garfield wrote:
>>> In other news, Ilija and I said a year ago that we'd take a swing at=20
>>> adding isReadable/isWriteable methods to ReflectionProperty.  Took a=20
>>> while, but here we are.  A strangely small RFC from us:
>>>=20
>>> https://wiki.php.net/rfc/isreadable-iswriteable
>>=20
>> Hi folks.
>>=20
>> The holiday blackout period is over, and there's nothing else really =
to discuss on this PR, so consider this an Intent to Vote notice for the=
 isReadable RFC.
>>=20
>> I'll call the vote later this week, baring any new serious constructi=
ve feedback.
>>=20
>> --Larry Garfield
>
> Hi,
>
> The RFC says:=20
>
>> To use =E2=80=9Cmy current scope,=E2=80=9D the `static::class` constr=
uct is an easy way to specify =E2=80=9Cwhatever class this code is runni=
ng in.=E2=80=9D
>
> The code is running in `self::class` or `__CLASS__`, not in `static::c=
lass`.

Hm, good point.  I've changed to to `self::class`, which I believe shoul=
d only be an Editorial change.  (There's no change to the design, just t=
o the explanation, and it's arguably a typo.)  I'll give it another day =
before opening the vote just in case.

> As noted in the RFC, `isReadable()` will give unavoidable=20
> false-positives (returning `true` although attempt to read the propert=
y=20
> will error). But there is also a false-negative:
>
> ```php
> class C {
>
>     function __construct(
>         readonly private mixed $foo
>     ) { }
>   =20
>     function __get($x) {
>         if ($x =3D=3D=3D 'foo') {
>             return $this->foo;
>         }
>         throw new Error('Undefined property '.$x);
>     }
>
>     function __isset($x) {
>         if ($x =3D=3D=3D 'foo')
>             return isset($this->foo); // not `true`, otherwise=20
> `isset(new C(null)->foo)` will return an incorrect result.
>         }
>         return false;
>     }
> }
>
> $c =3D new C(null);
>
> // will return false, although it is readable
> var_dump(new ReflectionProperty($c, 'foo')->isReadable(null, $c));
> ```
>
> One could skip the `__isset()` check in order to avoid the=20
> false-negative, at the cost of more false-positives. I don=E2=80=99t k=
now what=20
> is best, because I tend to avoid `__get()` like the plague anyway.

As discussed earlier in the thread and noted in the RFC, as long as hook=
s and __get support arbitrary code, we can never guarantee no false posi=
tives or negatives.  Based on the earlier discussion, the current setup =
seems like the best balance for the expected typical cases.  So we'd rat=
her stick with how it is now.

--Larry Garfield