Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:54354
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 209.85.218.42 as permitted sender)
MIME-Version: 1.0
In-Reply-To: <1312457975.1541.7.camel@guybrush>
References: <CAMyk_AZQ8cJ+iVZ=9x76kYCi4n5_FYyAg9JQQnL_BJhBUX6cCg@mail.gmail.com>
 <4E3A48F2.4000503@lerdorf.com> <CAMyk_AYxuK=o1s-HD7MGjV8u8KLksxgUfrvsx0qZdTO1MEuqJw@mail.gmail.com>
 <4E3A8014.1030503@rotorised.com> <1312457975.1541.7.camel@guybrush>
Date: Thu, 4 Aug 2011 14:39:42 +0200
Message-ID: <CAMyk_AZ+Ym2V+JZxGV3Fxv1sNhqGpk4KXtJBDga1BuXdjA0vmg@mail.gmail.com>
To: =?UTF-8?Q?Johannes_Schl=C3=BCter?= <johannes@schlueters.de>
Cc: Ryan McCue <lists@rotorised.com>, Rasmus Lerdorf <rasmus@lerdorf.com>, internals@lists.php.net
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Subject: Re: [PHP-DEV] An implementation of a short syntax for closures
From: linepogl@gmail.com (Lazare Inepologlou)

> ... ( $x ) =3D> $x + 1 for example would be ambiguous if used in an array=
 definition, but is otherwise the best in terms of readability.

> ... people wanted an easy way to grep for function declarations


A new and unique operator (like the |=3D> I have proposed) is a solution
that works because:
1. A lambda expression can be part of other expressions, like arrays.
Therefore, we cannot reuse =3D>, or ->.
2. Multiple lambda expressions can be chained. The operator must have
very low precedence and it be right associative.
3. Lambdas should be grep-able.


For those who would like to try other syntaxes that could work in PHP,
I suggest a simple and creative way to approach this. All we need is 3
operators:

  OP_A args OP_B lexical_vars OP_C expr

Operator OP_C ought to have the above properties (unique, low
precedence, right associative). On the other hand there are no
restrictions about OP_A and OP_B and we can reuse existing ones as
long as they don't conflict and they look nice.

:-)

Lazare INEPOLOGLOU
Ing=C3=A9nieur Logiciel



2011/8/4 Johannes Schl=C3=BCter <johannes@schlueters.de>:
> On Thu, 2011-08-04 at 21:18 +1000, Ryan McCue wrote:
>> Lazare Inepologlou wrote:
>> > Thank you for your interest. This is just a proposal that I have teste=
d and
>> > works. Of course, the final syntax can be different. Syntax is always =
a
>> > matter of taste :-)
>>
>> As much as I love the idea, I have to agree that using | doesn't really
>> make sense here and actually makes the readability worse, IMO. However,
>> I can't really think of a better operator. ( $x ) =3D> $x + 1 for exampl=
e
>> would be ambiguous if used in an array definition, but is otherwise the
>> best in terms of readability.
>
> If you go there you can also allow more complex statements and re-use
> symbols used to group statements, using a syntax like
>
> =C2=A0 =C2=A0 ( $x ) { $x + 1; }
>
> Oh wait - now we've reached a point which was mentioned in this thread
> already "allow to drop the function keyword".
>
> Back when I proposed to drop "function" a key argument was that people
> wanted an easy way to grep for function declarations. I accepted the
> validity of that concern and think this might be a case here, too. The
> difference of these few characters is not that much but the keyword
> gives a clear guidance what's going on, else you might end up with a
> long list of random characters next to each other ...
>
> johannes
>
>
>