Ada 202x and square brackets

Ada 202x and square brackets

[Lucretia]

Hi,

I've noticed whilst looking through the grammar of draft 17 there are a lot of new uses of '[' ']' without much explanation. Can someone elaborate?

Ta,
Luke.

Re: Ada 202x and square brackets

[Lucretia]

Ah found it.

Re: Ada 202x and square brackets

[Jere]

On Tuesday, February 19, 2019 at 5:01:05 PM UTC-5, Lucretia wrote:
> Hi,
> 
> I've noticed whilst looking through the grammar of draft 17 there are a lot of new uses of '[' ']' without much explanation. Can someone elaborate?
> 
> Ta,
> Luke.

I know you found it, but for other's benefit, which did you find.  I 
searched through and found them used in "reduction expressions" [1], 
but I don't know if that was the only use

[1]: http://www.ada-auth.org/standards/2xrm/html/RM-4-5-10.html#I3068

Re: Ada 202x and square brackets

[Jere]

On Wednesday, February 20, 2019 at 9:32:14 AM UTC-5, Jere wrote:
> On Tuesday, February 19, 2019 at 5:01:05 PM UTC-5, Lucretia wrote:
> > Hi,
> > 
> > I've noticed whilst looking through the grammar of draft 17 there are a lot of new uses of '[' ']' without much explanation. Can someone elaborate?
> > 
> > Ta,
> > Luke.
> 
> I know you found it, but for other's benefit, which did you find.  I 
> searched through and found them used in "reduction expressions" [1], 
> but I don't know if that was the only use
> 
> [1]: http://www.ada-auth.org/standards/2xrm/html/RM-4-5-10.html#I3068

Sorry, meant "value sequence" of reduction expressions

Re: Ada 202x and square brackets

[Lucretia]

On Wednesday, 20 February 2019 14:32:14 UTC, Jere  wrote:
> On Tuesday, February 19, 2019 at 5:01:05 PM UTC-5, Lucretia wrote:
> > Hi,
> > 
> > I've noticed whilst looking through the grammar of draft 17 there are a lot of new uses of '[' ']' without much explanation. Can someone elaborate?
> > 
> > Ta,
> > Luke.
> 
> I know you found it, but for other's benefit, which did you find.  I 
> searched through and found them used in "reduction expressions" [1], 
> but I don't know if that was the only use
> 
> [1]: http://www.ada-auth.org/standards/2xrm/html/RM-4-5-10.html#I3068

It was aggregates for containers, so you can do Set := [1,2,3,4] or something like that. I didn't read the details.

http://www.ada-auth.org/standards/2xaarm/html/AA-4-3-5.html

Re: Ada 202x and square brackets

[Randy Brukardt]

"Lucretia" <laguest9000@googlemail.com> wrote in message 
news:b773bbfa-9fd0-431a-8ff3-462b2c5ccd15@googlegroups.com...
> On Wednesday, 20 February 2019 14:32:14 UTC, Jere  wrote:
>> On Tuesday, February 19, 2019 at 5:01:05 PM UTC-5, Lucretia wrote:
>> > Hi,
>> >
>> > I've noticed whilst looking through the grammar of draft 17 there are a 
>> > lot of new uses of '[' ']' without much explanation. Can someone 
>> > elaborate?
>> >
>> > Ta,
>> > Luke.
>>
>> I know you found it, but for other's benefit, which did you find.  I
>> searched through and found them used in "reduction expressions" [1],
>> but I don't know if that was the only use
>>
>> [1]: http://www.ada-auth.org/standards/2xrm/html/RM-4-5-10.html#I3068
>
> It was aggregates for containers, so you can do Set := [1,2,3,4] or 
> something like that. I didn't read the details.
>
> http://www.ada-auth.org/standards/2xaarm/html/AA-4-3-5.html

You can also use them around array aggregates, so that you can write a 
single component array aggregate:

             My_Arr : constant Some_Unconstrained_Array_Type := [1];

Container aggregates are roughly user-defined array aggrgeates. 
(Alternatively, you can [maybe should] think of an array as a built-in 
container.)

                                  Randy.

Re: Ada 202x and square brackets

[Lucretia]

On Wednesday, 20 February 2019 23:47:53 UTC, Randy Brukardt  wrote:

>              My_Arr : constant Some_Unconstrained_Array_Type := [1];
> 
> Container aggregates are roughly user-defined array aggrgeates. 
> (Alternatively, you can [maybe should] think of an array as a built-in 
> container.)

Do you think the language is getting a tad too complicated now?

Re: Ada 202x and square brackets

[Randy Brukardt]

"Lucretia" <laguest9000@googlemail.com> wrote in message 
news:be2e3cf3-e452-4501-a960-822bf6432cd9@googlegroups.com...
> On Wednesday, 20 February 2019 23:47:53 UTC, Randy Brukardt  wrote:
>
>>              My_Arr : constant Some_Unconstrained_Array_Type := [1];
>>
>> Container aggregates are roughly user-defined array aggrgeates.
>> (Alternatively, you can [maybe should] think of an array as a built-in
>> container.)
>
> Do you think the language is getting a tad too complicated now?

How so? The idea has been that an abstract data type should be able to 
declare all of the same capabilities as a built-in type. The main remaining 
holes were literals and aggregates, and those have been addressed now.

It can be complicated to build such an ADT, but that's not something that is 
done often (much like with generics). OTOH, the user experience should be 
fairly seamless -- you can write the same thing whether or not the type is 
built-in or user-defined, and (if properly defined), they'll mean the same 
thing. Most people won't care how a user-defined aggregate is constructed 
any more than they care how a vector container supports indexing. (At least, 
that's the goal.)

                           Randy.

Re: Ada 202x and square brackets

[AdaMagica]

So we can now mix () and [] at will?

([1,2,3], (7,8,9)) for a two-dim. array.

[1,2,3] & (7,8,9) for concatenation.

At least (a,b,c] is not allowed.

I'd say a rule to not mix them in an expression is due.

Re: Ada 202x and square brackets

[J-P. Rosen]

Le 22/02/2019 à 11:34, AdaMagica a écrit :
> So we can now mix () and [] at will?
> 
> ([1,2,3], (7,8,9)) for a two-dim. array.
> 
> [1,2,3] & (7,8,9) for concatenation.
> 
> At least (a,b,c] is not allowed.
> 
> I'd say a rule to not mix them in an expression is due.
> 
There are cases where you want to mix for clarity:
([1,2], [3,4])

So what would your rule be?

-- 
J-P. Rosen
Adalog
2 rue du Docteur Lombard, 92441 Issy-les-Moulineaux CEDEX
Tel: +33 1 45 29 21 52, Fax: +33 1 45 29 25 00
http://www.adalog.fr

Re: Ada 202x and square brackets

[AdaMagica]

Am Freitag, 22. Februar 2019 12:53:00 UTC+1 schrieb J-P. Rosen:
> There are cases where you want to mix for clarity:
> ([1,2], [3,4])
I do not see that this enhances clarity, but rather raises questions about the difference.
I see new coding standards emerge saying: Use [] only for containers, not arrays. (I've seen lots of stupid rules in such coding standards.) So this would then be an array aggregate of container subaggregates.

> So what would your rule be?
Prhaps something like J.2(4)?

Re: Ada 202x and square brackets

[J-P. Rosen]

Le 22/02/2019 à 13:37, AdaMagica a écrit :
> Am Freitag, 22. Februar 2019 12:53:00 UTC+1 schrieb J-P. Rosen:
>> There are cases where you want to mix for clarity: ([1,2], [3,4])
> I do not see that this enhances clarity, but rather raises questions
> about the difference. I see new coding standards emerge saying: Use
> [] only for containers, not arrays. (I've seen lots of stupid rules
> in such coding standards.) So this would then be an array aggregate
> of container subaggregates.
A possible rule would be, when you have nested aggregates, to alternate
between parentheses and brackets. Or the rule you suggest. Or whatever
people find useful - i.e. better not put "good taste" rules, leave that
to tools.

-- 
J-P. Rosen
Adalog
2 rue du Docteur Lombard, 92441 Issy-les-Moulineaux CEDEX
Tel: +33 1 45 29 21 52, Fax: +33 1 45 29 25 00
http://www.adalog.fr

Re: Ada 202x and square brackets

[AdaMagica]

Am Freitag, 22. Februar 2019 17:56:56 UTC+1 schrieb J-P. Rosen:
> A possible rule would be, when you have nested aggregates, to alternate
> between parentheses and brackets. Or the rule you suggest. Or whatever
> people find useful - i.e. better not put "good taste" rules, leave that
> to tools.

You're right, the RM is not about taste.

Re: Ada 202x and square brackets

[Randy Brukardt]

"AdaMagica" <christ-usch.grein@t-online.de> wrote in message 
news:e5f7df5a-4e83-4c1f-82ad-ee286627ffbc@googlegroups.com...
> So we can now mix () and [] at will?
>
> ([1,2,3], (7,8,9)) for a two-dim. array.
>
> [1,2,3] & (7,8,9) for concatenation.
>
> At least (a,b,c] is not allowed.
>
> I'd say a rule to not mix them in an expression is due.

I raised that argument, and found out I was literally the only one that 
thought that was a good idea. I'm happy to hear that I'm not 100% alone, but 
it's not happening in the Standard.

Indeed, it's been suggested that:

([1,2,3], [7,8,9]) is a good idea a two-dim. array.

Supposedly, this will be the provence of style tools and rules.

                         Randy.