Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:78882
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 209.85.212.176 as permitted sender)
Message-ID: <54638904.20605@gmail.com>
Date: Wed, 12 Nov 2014 16:21:24 +0000
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:31.0) Gecko/20100101 Thunderbird/31.2.0
MIME-Version: 1.0
To: internals@lists.php.net
References: <CALW=WECCELQOZpqB3foM6T+Qz+0v3hWT-mdTgCmGApEG4vxNcg@mail.gmail.com> <CAEZPtU44YsM1s_zJqPhAEu6cbqxh4k=A35qOd_z=XWErNf5wPA@mail.gmail.com> <5457AF2F.90808@php.net> <5457BDB7.8070701@garfieldtech.com> <CAFMT4NrgLzzpMt5Vm_Q84hXJ65U2jjqtLFm=ALYqwtiyxp0LjA@mail.gmail.com> <CAEZPtU4urLHc9FM0GohQJwkZD-TYKTVuPXcUMXsG5cAScHz3mw@mail.gmail.com> <54589A8D.3020607@sugarcrm.com> <CAEZPtU5qfcjCrcWRKW+ETF4uBp46rwVbv9nCdJoSUF+M+wKQdA@mail.gmail.com> <CAEPJdumXG9TC7xG8-YmGbQiwPMzwYvE_jXWvOQuNmz+U_=+ZaQ@mail.gmail.com> <1C3F4FA3-ABD5-4F6F-A898-F63AC1C723D5@ajf.me> <54591A76.8070302@sugarcrm.com> <967E30E5-71CB-40F8-9AE2-733D327DE197@ajf.me> <CAFp73XtxhW6t0k5TMa5n1ErC_7vur55Sa9oLtUM1P+X1Wugemg@mail.gmail.com> <545945A5.2090204@sugarcrm.com> <C12D441D-81CB-4270-8CE3-5A8BF2B27238@ajf.me> <CAFp73Xu++tT7fXE878i8CsibwNtiVJFcHQtzXi-iZA6=FXqeFA@mail.gmail.com> <CAFp73XuSaj4s3aqTfoOMCM6mAUM0BBncbRTyMW6uaMRgHhTphQ@mail.gmail.com> <CADyq6sLcYXxpyDzxrLVxS+5wgYYKh_TewgEcEAvg63p7LaNZqQ@mail.gmail.com>
In-Reply-To: <CADyq6sLcYXxpyDzxrLVxS+5wgYYKh_TewgEcEAvg63p7LaNZqQ@mail.gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [PHP-DEV] Annotation PHP 7
From: rowan.collins@gmail.com (Rowan Collins)

Marco Pivetta wrote on 05/11/2014 14:02:
> For example, this alternative approach perfectly fits the current
> doctrine/annotations use-case:
>
> use Doctrine\ORM\Mapping\Entity;
> use Doctrine\ORM\Mapping\Table;
> use Doctrine\ORM\Mapping\Id;
> use Doctrine\ORM\Mapping\GeneratedValue;
> use Doctrine\ORM\Mapping\Column;
>
> [Entity::class => []]
> [Table::class => ['name' => 'foo_table']]
> class Foo
> {
>      [Id::class => []]
>      [GeneratedValue::class => [GeneratedValue::UUID]]
>      [Column::class => ['name' => 'bar_column', 'type' => 'string']]
>      private $bar;
> }
>
> I did nothing special here, just using plain old associative arrays and
> built-in `::class` pseudo-constants: this leaves space for libraries that
> do want to use annotations, but not the way doctrine/annotations does it
> (calling the constructor of an annotation), and the annotations themselves
> can still decide what to do with nested values.
>
> No autoloading is going on, no code execution, nothing: plain and simple
> arrays which may as well be used for phpdocumentor as well, if needed.

One problem with using *just* array syntax, without any new keyword or 
symbols, is that it is extremely close to being existing valid syntax. 
This may well cause problems in the parser, and would certainly be 
confusing to users.

Specifically, a class annotation might look like this:

['Blah\\Annotation\\Table' => ['name' => 'foo_table']]
class Foo {
}

But add a single semi-colon, and you have valid PHP code 
[http://3v4l.org/inkul]:

['Blah\\Annotation\\Table' => ['name' => 'foo_table']];
class Foo {
}

This works because an expression, by itself, is a valid statement, as 
long as it's terminated with a semi-colon, and before the "class" 
keyword, you are just in global execution scope (a problem many other 
languages don't have to worry about, because statements can't exist 
outside blocks).

The same reasoning rules out simple uses of @ as the prefix, because 
this is also valid code [http://3v4l.org/TCBbC]:

@Blah\Annotation\Table( ['name' => 'foo_table'] );
class Foo {
}

Amusingly, the @ suppresses the non-existent function error, making this 
look almost like it "worked".

Having a single semi-colon change an annotation into a silent statement 
seems like an extremely bad idea. The "inside" of an annotation can look 
like an array or constructor call, but we need something distinct to 
introduce (or wrap) it, assuming we want it positioned before the 
"class" keyword, as in other languages.

Regards,
-- 
Rowan Collins
[IMSoP]