Ada202X : alternate syntax for ranged scalars

Martin

unread,

Apr 26, 2013, 6:10:01 AM4/26/13

to

The default scalar declarations are 'closed ranges', i.e. include both upper and lower in the valid range.

Sometimes we would like to specify a 'half open range', e.g. values of degree from 0.0 up to BUT NOT INCLUDING 360.0.

I'm floating (pun intended) an alternative syntax for specify such ranges (or 'intervals')...

type Degree is digits 15 range [0.0, 360.0); -- includes 0 but not 360

The current 'closed ranges' would be:

type Degree is digits 15 range [0.0, 360.0]; -- includes 0 and 360

Fully open range, i.e. not include 0.0 or 360.0:

type Degree is digits 15 range (0.0, 360.0); -- includes neither 0 nor 360

A 'reverse' half open range, i.e. not include 0.0 but including 360.0:

type Degree is digits 15 range (0.0, 360.0]; -- excludes 0 but includes 360

Usual rules would apply to the simple expression between the brackets.

Any thoughts?
-- Martin

Peter C. Chapin

unread,

Apr 26, 2013, 6:35:26 AM4/26/13

to

On 04/26/2013 06:10 AM, Martin wrote:

> The default scalar declarations are 'closed ranges', i.e. include both upper and lower in the valid range.
>
> Sometimes we would like to specify a 'half open range', e.g. values of degree from 0.0 up to BUT NOT INCLUDING 360.0.
>
> I'm floating (pun intended) an alternative syntax for specify such ranges (or 'intervals')...
>
> type Degree is digits 15 range [0.0, 360.0); -- includes 0 but not 360

I'm not sure how important this is, but your proposed syntax will cause
fits for editors that do generic brace matching and/or auto-indentation.
I understand that an Ada-aware editor could be expected to do the right
thing, but it's nice when simple algorithms allow things to 'just work.'

That said, Ada already causes simple syntax highlighting editors
problems with the use of ' for both character constants and as an
attribute delimiter. That doesn't seem like a good reason to make
matters worse, however.

Peter

Eryndlia Mavourneen

unread,

Apr 26, 2013, 8:26:49 AM4/26/13

to

On Friday, April 26, 2013 5:10:01 AM UTC-5, Martin wrote:
> The default scalar declarations are 'closed ranges', i.e. include both upper and lower in the valid range.
>
> Sometimes we would like to specify a 'half open range', e.g. values of degree from 0.0 up to BUT NOT INCLUDING 360.0.
>
> I'm floating (pun intended) an alternative syntax for specify such ranges (or 'intervals')...
>
> type Degree is digits 15 range [0.0, 360.0); -- includes 0 but not 360
>
> The current 'closed ranges' would be:
> type Degree is digits 15 range [0.0, 360.0]; -- includes 0 and 360
>
> Fully open range, i.e. not include 0.0 or 360.0:
>
> type Degree is digits 15 range (0.0, 360.0); -- includes neither 0 nor 360
>
> A 'reverse' half open range, i.e. not include 0.0 but including 360.0:
>

> Usual rules would apply to the simple expression between the brackets.
>
> Any thoughts?
> -- Martin

I agree with Peter. A syntax more closely allied with current syntax might be:

type Degree is digits 15 range 0.0 .. not 360.0; -- includes 0 but not 360

The current 'closed ranges' would be the same as now:
type Degree is digits 15 range 0.0 .. 360.0; -- includes 0 and 360

Fully open range, i.e. not include 0.0 or 360.0:

type Degree is digits 15 range not 0.0 .. not 360.0; -- includes neither 0 nor 360

That said, we must assume that floating point values get translated *exactly*; that is, 360.0 does not get translated to a machine type of 359.999999... or 360.0111111...

-- Eryndlia F. Mavourneen

AdaMagica

unread,

Apr 26, 2013, 8:53:26 AM4/26/13

to

On Friday, April 26, 2013 12:10:01 PM UTC+2, Martin wrote:

> Sometimes we would like to specify a 'half open range', e.g. values of degree from 0.0 up to BUT NOT INCLUDING 360.0.

What about the attributes Succ and Pred RM 3.5 (22-27)?

type Angle is digits D;
subtype Normalized_Angle is Angle range -180.0 .. Angle'Pred (+180.0);

Shark8

unread,

Apr 26, 2013, 10:46:42 AM4/26/13

to

Yes, that should work much better.

Jeffrey Carter

unread,

Apr 26, 2013, 12:55:08 PM4/26/13

to

On 04/26/2013 03:10 AM, Martin wrote:
> The default scalar declarations are 'closed ranges', i.e. include both upper and lower in the valid range.
>
> Sometimes we would like to specify a 'half open range', e.g. values of degree from 0.0 up to BUT NOT INCLUDING 360.0.
>
> I'm floating (pun intended) an alternative syntax for specify such ranges (or 'intervals')...
>
> type Degree is digits 15 range [0.0, 360.0); -- includes 0 but not 360

I dislike the proposal. It's far too easy to miss the subtle distinction between
a bracket and a parenthesis.

--
Jeff Carter
"My legs are gray, my ears are gnarled, my eyes are old and bent."
Monty Python's Life of Brian
81

Stephen Leake

unread,

Apr 27, 2013, 10:11:27 PM4/27/13

to

+1

--
-- Stephe

Randy Brukardt

unread,

May 1, 2013, 10:11:50 PM5/1/13

to

"Eryndlia Mavourneen" <eryn...@gmail.com> wrote in message
news:3861a3cd-2513-4871...@googlegroups.com...
>...

> That said, we must assume that floating point values get translated
> *exactly*; that is,
>360.0 does not get translated to a machine type of 359.999999... or
>360.0111111...

Which of course makes no sense for most literals; it can only make sense for
values that already correspond to machine numbers, and those already require
exact conversion.

The problem I see with such syntax is what precisely happens when the bound
is not specified as a machine number? Say:
type My_Flt digits 5 range not 0.6 .. not 2.6;

Is this the same as the range without the "nots"? Or is it requiring
something less than some unknown value in a particular interval? I don't see
how that would be useful. Or are you going to require selecting a
*particular* machine number in this case? That would be a massive change
from the behavior of Ada currently, and probably would break some programs
(that aren't portable but do work) be changing the values of their literals.

The bounds of the range have to be machine numbers, as they don't have to be
static; as such, we don't care about values that aren't machine numbers. So
what exactly is this gaining over:

subtype My_Flt is Float range Float'Succ(0.0)..Float'Pred(360.0);

Randy.

Eryndlia Mavourneen

unread,

May 2, 2013, 9:29:32 AM5/2/13

to

On Wednesday, May 1, 2013 9:11:50 PM UTC-5, Randy Brukardt wrote:
> "Eryndlia Mavourneen" <eryn...@gmail.com> wrote in message
> news:3861a3cd-2513-4871...@googlegroups.com...
>
> >...
> > That said, we must assume that floating point values get translated
> > *exactly*; that is,
> >360.0 does not get translated to a machine type of 359.999999... or
> >360.0111111...
>

> So what exactly is this gaining over:
> subtype My_Flt is Float range Float'Succ(0.0)..Float'Pred(360.0);
>
> Randy.

Nothing that I see. I prefer the already available syntax of 'Pred and 'Succ. My note was merely offering up an alternative syntax to the Op's, if something special were required for any reason.

-- Eryndlia

Jerry

unread,

May 2, 2013, 7:05:35 PM5/2/13

to

If Ada introduces [ and ] please let them be for array indices.
Jerry

Martin

unread,

May 14, 2013, 6:30:21 PM5/14/13

to

Some good points against...

For reference, the inspiration for the syntax was ISO 31-11 (https://en.wikipedia.org/wiki/ISO_31-11)... Which I'm using to try and map between a modelling tool and a semantic analysis tool.

-- Martin