Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:108024
MIME-Version: 1.0
References: <CAF+90c-9ACg_v=B94Ok1tM5-xE9kO4sGQ3eL_j_qjaMWNK+QHw@mail.gmail.com>
In-Reply-To: <CAF+90c-9ACg_v=B94Ok1tM5-xE9kO4sGQ3eL_j_qjaMWNK+QHw@mail.gmail.com>
Date: Tue, 7 Jan 2020 10:42:26 +0000
Message-ID: <CALKiJKpD0uN8s_91AsSgH8Mk4w9f5JXvkghwCHbp87vZNB5cvA@mail.gmail.com>
To: PHP internals <internals@lists.php.net>
Content-Type: multipart/alternative; boundary="00000000000038565a059b8a6ff6"
Subject: Re: [PHP-DEV] [RFC] Variable syntax tweaks
From: rowan.collins@gmail.com (Rowan Tommins)

--00000000000038565a059b8a6ff6
Content-Type: text/plain; charset="UTF-8"

On Tue, 7 Jan 2020 at 10:23, Nikita Popov <nikita.ppv@gmail.com> wrote:

> I'd like to propose a small RFC, which addresses a few minor issues that
> have not been handled by the original "uniform variable syntax" RFC:
>
> https://wiki.php.net/rfc/variable_syntax_tweaks




Hi Nikita,

Thanks for this, I'm definitely a fan of this kind of consistency, even if
it rarely matters in practice. :)

Regarding instanceof, and specifically your last alternative:

> The second possibility is to relax the restrictions around the
right-hand-side of instanceof entirely. This would involve treating it as a
normal expression, and then reinterpreting plain constant accesses as class
name references instead.

Am I right in thinking that another disadvantage of this is that you
couldn't use a bare constant to define a class name, thus introducing a new
inconsistency? That is, using a bracketed form, it should be possible to
write:

const CLASS_NAME='MyClass';
var_dump( $string instanceof {CLASS_NAME} );

or:

const CLASS_NAME='MyClass';
var_dump( $string instanceof (CLASS_NAME) );

but using the "reinterpreted expression" option, you'd need to do something
hacky to force the interpretation as a constant, like:

const CLASS_NAME='MyClass';
var_dump( $string instanceof ('' . CLASS_NAME) );

Another edge case of an edge case, of course, but a mark against that
option in my view. :)


Regards,
-- 
Rowan Tommins
[IMSoP]

--00000000000038565a059b8a6ff6--