Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:129484
Feedback-ID: i8414410d:Fastmail
Precedence: list
MIME-Version: 1.0
Date: Mon, 01 Dec 2025 14:06:27 -0600
To: "php internals" <internals@lists.php.net>
Message-ID: <dc487d89-ab46-4e82-8aa6-a54838029fcb@app.fastmail.com>
In-Reply-To: <ce70cf53-5503-4be7-b47c-9dda877fed7e@rwec.co.uk>
References: <a087d9fa-112a-4631-9c84-3efae641e0db@app.fastmail.com>
 <AM8P250MB0170975953345DCE66B6B9B6E2C5A@AM8P250MB0170.EURP250.PROD.OUTLOOK.COM>
 <d282c072-0403-4e33-b17b-bfcdd105e9c2@rwec.co.uk>
 <AM8P250MB017018D517EABEA87D64378BE2C5A@AM8P250MB0170.EURP250.PROD.OUTLOOK.COM>
 <74abada6-cb34-490e-abeb-517e39b3ffa0@app.fastmail.com>
 <d250c9413e1d561033a1bd735dd1d3cd@bastelstu.be>
 <ce70cf53-5503-4be7-b47c-9dda877fed7e@rwec.co.uk>
Subject: Re: [PHP-DEV] [RFC] Context Managers
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From: larry@garfieldtech.com ("Larry Garfield")

On Mon, Dec 1, 2025, at 9:38 AM, Rowan Tommins [IMSoP] wrote:
> On 01/12/2025 14:26, Tim D=C3=BCsterhus wrote:
>> You are wrong here. Manually interacting with an Iterator is not just=20
>> safe, it is also necessary for some use cases. The simplest example=20
>> would be iterating two Iterators in lock-step (i.e. doing a `zip()`=20
>> operation / array_combine()).=20
>
>
> It is certainly *possible* to use the Iterator methods safely without =
a=20
> foreach() construct or built-in aggregate; but it's also easy to miss=20
> the implications, and cause very confusing behaviour. For instance, yo=
u=20
> could call next() in two different pieces of code, and each would miss=20
> half the items; or you could call rewind() on an iterator that was=20
> already in use elsewhere.
>
> The same is true of the ContextManager interface: you could certainly=20
> call enterContext() and exitContext() manually with no problems at all=
.=20
> It might even be necessary, for much the same reasons as iterators: yo=
u=20
> might want to write code that aggregates two context managers in a=20
> specific arrangement.
>
>
> As I understand it, your concern is that someone will call=20
> enterContext() without exitContext(). You could do exactly the same=20
> thing if you manually handle an Iterator.
>
> Consider this method:
>
> function getDataIterator(): Generator {
>  =C2=A0 =C2=A0 $this->acquireLock();
>  =C2=A0 =C2=A0 while ( $this->hasMoreResults() ) {
>  =C2=A0 =C2=A0 =C2=A0 =C2=A0 yield $this->getNextResult();
>  =C2=A0 =C2=A0 }
>  =C2=A0 =C2=A0 $this->releaseLock();
> }
>
> If the user calls $it =3D $foo->getDataIterator(), they can call=20
> $it->current() and get a result, but never call $it->next(). They coul=
d=20
> even use it in a foreach() loop but `break` out before consuming all t=
he=20
> results. In such cases, the lock would never be released, unless there=
's=20
> an extra safety check in the destructor.

Excellent example, thank you.

> I think the currently proposed names actually make that *less* likely=20
> for a ContextManager: if you enter something, you probably want to exi=
t=20
> it later.
>
> We could reinforce that further by also prefixing the method names wit=
h=20
> "__", but we don't have any precedent for that, and misuse would still=20
> be possible.

Python uses magic methods here, and we considered it, but given that you=
 will basically always want both methods we felt an interface was easier=
 and provided better type checks.  If there were multiple callbacks (say=
, a different one for a success exit and a fail exit), then magic method=
s might make more sense.

If you're suggesting having interface methods that begin with __, I agre=
e that would be novel for no particular value.  You can't screw up Conte=
xtManagers through misuse any more than you can screw up Iterator or Arr=
ayAccess, and we've survived just fine with those for literally decades.

--Larry Garfield