Path: csiph.com!newsfeed.hal-mli.net!feeder3.hal-mli.net!newsfeed.hal-mli.net!feeder1.hal-mli.net!newsfeed.xs4all.nl!newsfeed3a.news.xs4all.nl!xs4all!newsgate.cistron.nl!newsgate.news.xs4all.nl!post.news.xs4all.nl!not-for-mail
MIME-Version: 1.0
In-Reply-To: <53060a97-44fb-4e53-a7a7-d5eeed416f62@googlegroups.com>
References: <9daf0806-02de-4447-964c-c8f8953c23e5@googlegroups.com> <10101874-2995-4acd-9851-989603f052e3@googlegroups.com> <bd158c7b-5511-472c-bce3-cc7b6b7747f6@googlegroups.com> <e2b2b9d5-56bd-4455-b8c3-205000d9acb3@googlegroups.com> <mailman.8369.1395437459.18130.python-list@python.org> <a0403642-c384-4623-bbfd-77c48285ec96@googlegroups.com> <mailman.8377.1395456097.18130.python-list@python.org> <dcd72881-7ee2-4bc3-90ab-7912b4bea05b@googlegroups.com> <lgj4sp$7ed$1@speranza.aioe.org> <532d5bd9$0$29994$c3e8da3$5496439d@news.astraweb.com> <lgoh7j$otg$1@speranza.aioe.org> <mailman.8440.1395651801.18130.python-list@python.org> <lgq8th$nj4$1@speranza.aioe.org> <mailman.8472.1395705541.18130.python-list@python.org> <0b78649a-16b3-4410-8258-e859578d62be@googlegroups.com> <mailman.8483.1395717465.18130.python-list@python.org> <53060a97-44fb-4e53-a7a7-d5eeed416f62@googlegroups.com>
Date: Tue, 25 Mar 2014 15:14:30 +1100
Subject: Re: Time we switched to unicode? (was Explanation of this Python language feature?)
From: Chris Angelico <rosuav@gmail.com>
Cc: "python-list@python.org" <python-list@python.org>
Content-Type: text/plain; charset=UTF-8
Precedence: list
Newsgroups: comp.lang.python
Message-ID: <mailman.8487.1395720873.18130.python-list@python.org>
Lines: 75
NNTP-Posting-Host: 2001:888:2000:d::a6
Xref: csiph.com comp.lang.python:68950

On Tue, Mar 25, 2014 at 2:56 PM, Rustom Mody <rustompmody@gmail.com
>> I don't know about the difference between {} in set theory and Python,
>> but the multiple uses of () actually boil down to two:
>
> In set theory {} makes sets
> In python {} makes dictionaries

That's a backward-compatibility issue. Braces in Python meant a
dictionary before it acquired a set type (at least, so I learned in
history class - I wasn't using Python back then), so empty braces have
to continue to mean empty dictionary. I sympathize with the confusion,
but short of confusing everyone terribly by changing dicts to use
something other than braces (maybe borrow Pike's mapping notation, ([
])??), I don't really see a solution.

>> 1) Grouping, which includes tuples; there's a special case whereby
>> grouping nothing makes a zero-item tuple, but everything else is just
>> the comma
>
>> 2) Functions (both definition and call)
>
>> Disambiguating them might be of some small value, but since they're
>> the same in pretty much every language under the sun, it would feel
>
> What 'they'?? I dont get: If you are talking of disambiguating function definition and call -- yeah thats overkill

No no. Disambiguating grouping and fuction definition/call. There's no
reason to have definition and call of functions differ.

> If you are talking of overlap between tuples parentheses and function (call)
> well consider
> f(x,y) vs f((x,y)) vs (x,y) vs ((x,y))
> Paren vs tuples: why do we need to write (x,) not (x)
>
> All this is because () is doing triple-duty

Tuples don't use parentheses. You only confuse yourself when you
insist on that. The only case with parens that actually makes a tuple
is the special empty tuple; imagine if that were given a name instead,
like "Empty". That would solve that confusion.

There's another minorly special case, and that's that the trailing
comma is mandatory on a one-item tuple. In all others, it's optional,
but the one-item tuple would be ambiguous otherwise.

>>> len((1,2,3,))
3
>>> len((1,2,))
2
>>> len((1,))
1

(By the way, bringing in another discussion: The comma isn't part of
the element, nor is it exactly a separator, nor is it exactly a
terminator. Just like a newline in a text file.)

So the only ambiguity is between function calls and grouping. Tuples
are just part of grouping, in that you need to disambiguate a
one-tuple-arg function call from a multiple-arg function call - as in
the above len() calls. And that's where I think it would be highly
confusing to mandate something different. Suppose we used $( )$ for
function calls, and ^( )^ for grouping:

x = ^(1 + 2)^ * 3
y = range$(x)$

What have we gained by distinguishing them? Not a lot. If an open
parenthesis immediately follows another token, it's calling that
previous token; if it follows an operator, it's grouping. Seems pretty
simple to me.

(Cue the spate of emails pointing out something I've missed that
breaks that rule, in which case call it a rule of thumb.)

ChrisA