Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:130502
Feedback-ID: id5114917:Fastmail
Date: Mon, 30 Mar 2026 20:36:40 +0100
To: internals@lists.php.net
Subject: Re: [PHP-DEV] [RFC] Context Managers
User-Agent: K-9 Mail for Android
In-Reply-To: <8bf1b0a7-7771-4f81-b92c-49bb019f7f9d@app.fastmail.com>
References: <a087d9fa-112a-4631-9c84-3efae641e0db@app.fastmail.com> <cf17cf19-f41e-4378-9a82-cb7f1a9d7b70@app.fastmail.com> <5d96fca3ca5418e6e9e5d8871b26477f@bastelstu.be> <8bf1b0a7-7771-4f81-b92c-49bb019f7f9d@app.fastmail.com>
Message-ID: <C50C744D-7E21-4E6F-AFC1-93113C8B0F55@rwec.co.uk>
Precedence: list
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable
From: imsop.php@rwec.co.uk ("Rowan Tommins [IMSoP]")

On 30 March 2026 19:48:05 BST, Larry Garfield <larry@garfieldtech=2Ecom> wr=
ote:

> =2E=2E=2E break is the natural keyword for that, as it already means "st=
op this control structure and go to the end of it=2E"
>
>continue means "stop this iteration of a control structure and go to the =
next one=2E"  But in this case, there is no next one=2E  switch makes it an=
 alias for break, for whatever reason lost to history, but given that it no=
w throws a warning that seems to now be considered a mistake, so we don't s=
ee a reason to propagate that mistake=2E


I agree with Tim that making break and continue define targets differently=
 is a really bad idea=2E

It's fine for a single "break;" or "continue;", but with PHP's count-based=
 targeting, it would lead to cases like this:

if ( definitely_right($loop_item) ) {
    $found_item =3D $loop_item;
    break 4;
}
elseif ( definitely_wrong($loop_item) ) {
    // targeting the same loop, but there's a couple of "using" or "switch=
" blocks in between
    continue 2;
}


If "break 2" terminates a "using" block, there are only two sane behaviour=
s for "continue 2":

- terminate that same block, as though it was a single-iteration loop
- throw an Error, because the operation is not meaningful


Looking back at the discussion of "continue targeting switch", I see I mad=
e the same point back then: https://externals=2Eio/message/102393#102500

The consensus in that discussion was to *only* add a Warning, with no plan=
 for further changes=2E It's not a deprecation, or a workaround for hard to=
 change legacy behaviour; it's just a helpful hint to the developer that th=
eir code might have a mistake in it=2E


Regards,

Rowan Tommins
[IMSoP]