Path: csiph.com!x330-a1.tempe.blueboxinc.net!usenet.pasdenom.info!aioe.org!news.stack.nl!newsfeed.xs4all.nl!newsfeed6.news.xs4all.nl!xs4all!post.news.xs4all.nl!not-for-mail
To: python-list@python.org
From: Robert Kern <robert.kern@gmail.com>
Subject: Re: Grammar for classes
Date: Tue, 20 Dec 2011 17:20:12 +0000
References: <CAN1F8qWV1j-2VDvMyh47dxfj4G_yXSP8rcUdmpBY-HLrD5gaUQ@mail.gmail.com> <jcppjc$va7$1@dough.gmane.org> <CAN1F8qXjE0zPrUjC-Kih=gCm=pv4mwGPCvT+Y5EE3XVwVw3WwA@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.7; rv:8.0) Gecko/20111105 Thunderbird/8.0
In-Reply-To: <CAN1F8qXjE0zPrUjC-Kih=gCm=pv4mwGPCvT+Y5EE3XVwVw3WwA@mail.gmail.com>
Precedence: list
Newsgroups: comp.lang.python
Message-ID: <mailman.3871.1324401626.27778.python-list@python.org>
Lines: 56
NNTP-Posting-Host: 2001:888:2000:d::a6
Xref: x330-a1.tempe.blueboxinc.net comp.lang.python:17591

On 12/20/11 5:05 PM, Joshua Landau wrote:
> On 20 December 2011 10:55, Robert Kern <robert.kern@gmail.com
> <mailto:robert.kern@gmail.com>> wrote:
>
>     On 12/20/11 1:34 AM, Joshua Landau wrote:
>
>         In reading thorough the syntax defined in the reference
>         <http://docs.python.org/py3k/__reference/compound_stmts.html#__class-definitions
>         <http://docs.python.org/py3k/reference/compound_stmts.html#class-definitions>>,
>
>         the class statement has surprised me.
>
>         It says that the inheritance part of the class can accept
>         comprehensions. What
>         does this mean?
>         I've tried:
>         "class A(x for x in ()): pass"
>         but this doesn't need the extra clause as "x for x in ()" is an
>         expression, and
>         thus this evaluates:
>         "class A(x for x in (),): pass"
>         although again it won't be a valid class anytime soon.
>
>         So what is this clause for?
>
>
>     I suspect that it's harder to make a grammar rule that allows every kind of
>     expression except for generator expressions than it is just to reuse the
>     "testlist" rule and let the runtime reject the generator object when it goes
>     to construct the class.
>
>
> The thing is, as far as I can tell, the grammar would be the same without the
> added clause. That is because it's a valid grammar anyway. The only exception I
> can think of is if generators were treated exceptionally. Thus, while I agree it
> makes complete sense to make it a runtime error, the clause in the specification
> seems to be fluff.
>
> Also - what is this "testlist" rule thee speaks of?

Sorry, I was looking at the actual 2.7 Grammar file. "testlist" has been renamed 
to "arglist" in the 3.2 version of the Grammar, but the substantive point I made 
remains (to the extent that it is relevant at all).

That said, the documentation for both versions differs from a simple restatement 
of the Grammar rules, so I'm not sure exactly why the documentation is written 
the way it is.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
  that is made terrible by our own mad attempt to interpret it as though it had
  an underlying truth."
   -- Umberto Eco