strange length check error

strange length check error

[Mehdi Saada]

This : Put(Ch_Bounded1 & ' ' );
Calls on that:
function "&"
     (Left  : in     Bounded_String;
      Right : in     Character)
      return Bounded_String
   is
   begin
      if LENGTH(LEFT) + 1 > MAX_LENGTH then
         raise LENGTH_ERROR with "ERROR '&': LEFT.LENGTH := " & NATURAL'IMAGE(LENGTH(LEFT)) & " + 1 > MAX_LENGTH";
      else
 @@@     return (LENGTH(LEFT) + 1, LEFT.DATA(1..LENGTH(LEFT)) & RIGHT);
      end if;
   end "&";

The line with @@@ raises this:
raised CONSTRAINT_ERROR : p_bounded_g-p_new_bounded_g.adb:34 length check failed
The value of CH_BOUNDED1 at that moment is "pour tester le debut", its length is 21, and MAX_LENGTH est 80.

Re: strange length check error

[Anh Vo]

On Tuesday, January 16, 2018 at 12:48:45 PM UTC-8, Mehdi Saada wrote:
> This : Put(Ch_Bounded1 & ' ' );
> Calls on that:
> function "&"
>      (Left  : in     Bounded_String;
>       Right : in     Character)
>       return Bounded_String
>    is
>    begin
>       if LENGTH(LEFT) + 1 > MAX_LENGTH then
>          raise LENGTH_ERROR with "ERROR '&': LEFT.LENGTH := " & NATURAL'IMAGE(LENGTH(LEFT)) & " + 1 > MAX_LENGTH";
>       else
>  @@@     return (LENGTH(LEFT) + 1, LEFT.DATA(1..LENGTH(LEFT)) & RIGHT);
>       end if;
>    end "&";
> 
> The line with @@@ raises this:
> raised CONSTRAINT_ERROR : p_bounded_g-p_new_bounded_g.adb:34 length check failed
> The value of CH_BOUNDED1 at that moment is "pour tester le debut", its length is 21, and MAX_LENGTH est 80.

What is Bounded_String?

Re: strange length check error

[Simon Wright]

Anh Vo <anhvofrcaus@gmail.com> writes:

> On Tuesday, January 16, 2018 at 12:48:45 PM UTC-8, Mehdi Saada wrote:
>> This : Put(Ch_Bounded1 & ' ' );
>> Calls on that:
>> function "&"
>>      (Left  : in     Bounded_String;
>>       Right : in     Character)
>>       return Bounded_String
>>    is
>>    begin
>>       if LENGTH(LEFT) + 1 > MAX_LENGTH then
>>          raise LENGTH_ERROR with "ERROR '&': LEFT.LENGTH := " & NATURAL'IMAGE(LENGTH(LEFT)) & " + 1 > MAX_LENGTH";
>>       else
>>  @@@     return (LENGTH(LEFT) + 1, LEFT.DATA(1..LENGTH(LEFT)) & RIGHT);
>>       end if;
>>    end "&";
>> 
>> The line with @@@ raises this:
>> raised CONSTRAINT_ERROR : p_bounded_g-p_new_bounded_g.adb:34 length check failed
>> The value of CH_BOUNDED1 at that moment is "pour tester le debut", its length is 21, and MAX_LENGTH est 80.
>
> What is Bounded_String?

I agree. It's not the one from the standard library ((a) it's different,
(b) the standard library doesn't use this ugly, hard to read,
identifiers-in-upper-case style).

This shows why claims that you never need to use the debugger when you
write in Ada are wrong. "What, never?" "Well,hardly ever."

Re: strange length check error

[Mehdi Saada]

I get it, it's ugly, inefficient, wrong, I KNOW, but it's not mine, and I can't affor rewriting *every* piece of code that come along the exercices.
Well, I could replace the actual
type BOUNDED_STRING is record
    LENGTH: NATURAL := 0;
    DATA: STRING;
end record;

by: type Bounded_String (Length: Natural := 0) is record
    Data: String (1..Length);
end record;

I suppose it wouldn't impact the specification a lot. But I guess it wouldn't be enough to please you, Dmitry ? I swear I'll try your strings later, when the *teacher*'s exemples will work as they should.

Re: strange length check error

[AdaMagica]

Am Mittwoch, 17. Januar 2018 13:13:49 UTC+1 schrieb Mehdi Saada:
> I get it, it's ugly, inefficient, wrong, I KNOW, but it's not mine, and I can't affor rewriting *every* piece of code that come along the exercices.
> Well, I could replace the actual
> type BOUNDED_STRING is record
>     LENGTH: NATURAL := 0;
>     DATA: STRING (1 .. 80);    <---- I guess *you* forgot this or else it's illegal
> end record;

Then of course this will raise an exception:
return (LENGTH(LEFT) + 1, LEFT.DATA(1..LENGTH(LEFT)) & RIGHT);
since
LEFT.DATA(1..LENGTH(LEFT)) & RIGHT is shorter than 80.

Who is the *teacher*?

Re: strange length check error

[Simon Wright]

AdaMagica <christ-usch.grein@t-online.de> writes:

> Am Mittwoch, 17. Januar 2018 13:13:49 UTC+1 schrieb Mehdi Saada:
>> I get it, it's ugly, inefficient, wrong, I KNOW, but it's not mine,
>> and I can't affor rewriting *every* piece of code that come along
>> the exercices.
>> Well, I could replace the actual
>> type BOUNDED_STRING is record
>>     LENGTH: NATURAL := 0;
>>     DATA: STRING (1 .. 80); <---- I guess *you* forgot this or else
>> it's illegal
>> end record;
>
> Then of course this will raise an exception:
> return (LENGTH(LEFT) + 1, LEFT.DATA(1..LENGTH(LEFT)) & RIGHT);
> since
> LEFT.DATA(1..LENGTH(LEFT)) & RIGHT is shorter than 80.
>
> Who is the *teacher*?

I'm guessing that this is a generic, with parameter MAX_LENGTH, so maybe

   type BOUNDED_STRING is record
      LENGTH: NATURAL := 0;
      DATA: STRING (1 .. MAX_LENGTH);
   end record;

& then this should be OK

   return
     (LENGTH(LEFT) + 1,
      LEFT.DATA(1..LENGTH(LEFT))
        & RIGHT
        & LEFT.DATA((LENGTH(LEFT) + 1)..MAX_LENGTH));

Re: strange length check error

[Jeffrey R. Carter]

On 01/17/2018 02:41 PM, Simon Wright wrote:
> 
>     return
>       (LENGTH(LEFT) + 1,
>        LEFT.DATA(1..LENGTH(LEFT))
>          & RIGHT
>          & LEFT.DATA((LENGTH(LEFT) + 1)..MAX_LENGTH));

I think the last part should be Left.Data (Left.Length + 1 .. Max_Length - 1) or 
(Left.Length + 2 .. Max_Length). I guess it doesn't matter what you put in those 
unused positions; you could also have (Left.Length + 2 .. Max_Length => <>).

I'd probably have written

    Result : Bounded_String;
begin -- "&"
    Result.Length := Left.Length + 1;
    Result.Data (1 .. Result.Length) := Left (1 .. Left.Length) & Right;

    return Result;

-- 
Jeff Carter
"We'll make Rock Ridge think it's a chicken
that got caught in a tractor's nuts!"
Blazing Saddles
87

Re: strange length check error

[Simon Wright]

"Jeffrey R. Carter" <spam.jrcarter.not@spam.not.acm.org> writes:

> On 01/17/2018 02:41 PM, Simon Wright wrote:
>>
>>     return
>>       (LENGTH(LEFT) + 1,
>>        LEFT.DATA(1..LENGTH(LEFT))
>>          & RIGHT
>>          & LEFT.DATA((LENGTH(LEFT) + 1)..MAX_LENGTH));
>
> I think the last part should be Left.Data (Left.Length + 1
> .. Max_Length - 1) or (Left.Length + 2 .. Max_Length). I guess it
> doesn't matter what you put in those unused positions; you could also
> have (Left.Length + 2 .. Max_Length => <>).

Should have been (Left.Length + 2 .. Max_Length). That'll teach me not
to post code without testing it ..

> I'd probably have written
>
>    Result : Bounded_String;
> begin -- "&"
>    Result.Length := Left.Length + 1;
>    Result.Data (1 .. Result.Length) := Left (1 .. Left.Length) & Right;
>
>    return Result;

There was a check in the original that the input wasn't already at its
max length. That aside, agreed.

Re: strange length check error

[Mehdi Saada]

Sorry. It was indeed Data: String (1..Max_Length) wher Max_Length is 80. Teacher is Daniel Feneuille. French.
But I understand now... I used to assign strings of different length. Of course it's illegal.

Re: strange length check error

[Mehdi Saada]

Sorry. It was indeed Data: String (1..Max_Length) wher Max_Length is 80. Teacher is Daniel Feneuille. French. 
But I understand now... I used to assign strings of different length. I'm rewriting things accordingly.

Re: strange length check error

[Mehdi Saada]

It's done. I ran the test program, and fortunately, it tells nothing about the core package I spent time on. At least it complies with the specification, though the behavior might be partialy wrong, I'll see that later. But it still shows error with IO child package, which I modified too but contains only:


   function Get_Line (
         File : File_Type := Standard_Input )
     return Bounded_String;

   procedure Put_Line (
         File :        File_Type := Standard_Output;
         Item : in     Bounded_String );

   procedure Put (
         File :        File_Type := Standard_Output;
         Item : in     Bounded_String );

as specs, and as bodies:

   procedure Put (
         File :        File_Type := Standard_Output;
         Item : in     Bounded_String ) is
   begin
      Put(File, Item.Data );
   end Put;

   procedure Put_Line (
         File :        File_Type := Standard_Output;
         Item : in     Bounded_String ) is
   begin
      Put_Line (File, Item.Data );
   end Put_Line;

   function Get_Line (
         File : File_Type := Standard_Input )
     return Bounded_String is
      Ch : Unbounded_String;
   begin
      Ch := Get_Line(File);
      return (Length(Ch), Slice(Ch,1,Length(Ch)));
   end Get_Line;

Originally, there was subprograms with, and without the File parameters. I didn't get why, and put instead default file parameters. I suspect there is a reason I never saw such default parameters in generic package before, but I couldn't help but give a try...

There are the errors:
crea_ora.adb:32:07: no candidate interpretations match the actuals:
crea_ora.adb:32:07: too many arguments in call to "Put_Line"
crea_ora.adb:32:17: expected private type "Ada.Text_Io.File_Type"
crea_ora.adb:32:17: found private type "Bounded_String" defined at p_bounded_g.ads:21, instance at p_bounded_80.ads:6
crea_ora.adb:32:17:   ==> in call to "Put_Line" at p_bounded_g-p_bounded_g_io.ads:17, instance at p_bounded_io_80.ads:9
crea_ora.adb:32:17:   ==> in call to "Put_Line" at a-textio.ads:259

One such piece for each call on Put, Put_Line or Get_Line.

Re: strange length check error

[Simon Wright]

Not sure what the problem is, but this looks like a case where you
should *not* 'use Ada.Text_IO;', especially in the spec.

Unrelated - do you think that in your procedure Put, for example, you
should only output the valid contents of Item.Data? (1 .. Item.Length)

Re: strange length check error

[Mehdi Saada]

Ah ? Ah how am I gonna use File_Type then ? I can't do otherwise.
There is only valid content, since now the definition is
type Bounded_String (Length: Length_Range) is
    Data: String;
end record;
I changed everything accordingly. The package merely creates bigger or smaller Strings dynamically. I don't use Text_IO outside the IO child package.
If there is a way to do IO without Ada.Text_IO, I'm open. 

Re: strange length check error

[Niklas Holsti]

On 18-01-17 21:43 , Simon Wright wrote:
 > Not sure what the problem is, but this looks like a case where you
 > should *not* 'use Ada.Text_IO;', especially in the spec.

On 18-01-17 23:37 , Mehdi Saada wrote:
> Ah ? Ah how am I gonna use File_Type then ? I can't do otherwise.

By qualifying with the name of the package in which it is defined:

    function Get_Line (
       File : Ada.Text_IO.File_Type := Ada.Text_IO.Standard_Input )
    return Bounded_String;

> If there is a way to do IO without Ada.Text_IO, I'm open.

As I understood Simon (but perhaps I should have let him answer), the 
point was to say only

    with Ada.Text_IO;

but *not* add

    use Ada.Text_IO;

because you are defining operations (Get_Line, Put_Line, ...) with the 
same names as operations in Ada.Text_IO, and you could get into problems 
with one definition hiding another.

The question "to 'use', or not to 'use'" is a big discussion in 
Ada-land, with people having different styles and preferences.

-- 
Niklas Holsti
Tidorum Ltd
niklas holsti tidorum fi
       .      @       .

Re: strange length check error

[Simon Wright]

Niklas Holsti <niklas.holsti@tidorum.invalid> writes:

> As I understood Simon (but perhaps I should have let him answer), the
> point was to say only
>
>    with Ada.Text_IO;
>
> but *not* add
>
>    use Ada.Text_IO;

Yes.

Re: strange length check error

[Simon Clubley]

On 2018-01-17, Simon Wright <simon@pushface.org> wrote:
>
> I agree. It's not the one from the standard library ((a) it's different,
> (b) the standard library doesn't use this ugly, hard to read,
> identifiers-in-upper-case style).
>

Someone was obviously following the Ada 83 style when they wrote that
code...

Not as annoying as some Wirth case sensitive languages however where
the keywords themselves are in uppercase. (Oberon comes to mind here.)

Simon.

-- 
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world

Re: strange length check error

[Dmitry A. Kazakov]

On 16/01/2018 21:48, Mehdi Saada wrote:
> This : Put(Ch_Bounded1 & ' ' );
> Calls on that:
> function "&"
>       (Left  : in     Bounded_String;
>        Right : in     Character)
>        return Bounded_String
>     is
>     begin
>        if LENGTH(LEFT) + 1 > MAX_LENGTH then
>           raise LENGTH_ERROR with "ERROR '&': LEFT.LENGTH := " & NATURAL'IMAGE(LENGTH(LEFT)) & " + 1 > MAX_LENGTH";
>        else
>   @@@     return (LENGTH(LEFT) + 1, LEFT.DATA(1..LENGTH(LEFT)) & RIGHT);
>        end if;
>     end "&";
> 
> The line with @@@ raises this:
> raised CONSTRAINT_ERROR : p_bounded_g-p_new_bounded_g.adb:34 length check failed
> The value of CH_BOUNDED1 at that moment is "pour tester le debut", its length is 21, and MAX_LENGTH est 80.

Why do you care? It is a wrong design anyway. The whole idea of 
bounded-length strings is avoid reallocation and copying upon string 
modification. [Which is why you should never use them]

Thus, *all* operations on bounded-length strings must be in-place, 
except constructors. If you have "&" that defeats the only purpose of 
having such strings. Use String.

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de

Re: strange length check error

[Robert Eachus]

On Tuesday, January 16, 2018 at 3:48:45 PM UTC-5, Mehdi Saada wrote:
> This : Put(Ch_Bounded1 & ' ' );
> Calls on that:
> function "&"
>      (Left  : in     Bounded_String;
>       Right : in     Character)
>       return Bounded_String
>    is
...
>  @@@     return (LENGTH(LEFT) + 1, LEFT.DATA(1..LENGTH(LEFT)) & RIGHT);
>       end if;
>    end "&";
> 
> The line with @@@ raises this:
> raised CONSTRAINT_ERROR : p_bounded_g-p_new_bounded_g.adb:34 length check failed
> The value of CH_BOUNDED1 at that moment is "pour tester le debut", its length is 21, and MAX_LENGTH est 80.

Arg!  Horrible code, I'm not surprised that others missed it.  This is a recursive call on "&" with a longer LENGTH each time.  Eventually LENGTH equals MAX_LENGTH and the error is raised.

Re: strange length check error

[Mehdi Saada]

> Arg!  Horrible code, I'm not surprised that others missed it.  This is a recursive call on "&" with a longer LENGTH each time.  Eventually LENGTH equals MAX_LENGTH and the error is raised.

No, I doubt it, the type of the expression LEFT.DATA(1..LENGTH(LEFT)) & RIGHT is String, and the "&" is the standard string concatenation operator... Or please show me the parenthese I'm eventually missing ?

Re: strange length check error

[Robert Eachus]

On Tuesday, January 23, 2018 at 11:41:58 AM UTC-5, Mehdi Saada wrote:
 
> No, I doubt it, the type of the expression LEFT.DATA(1..LENGTH(LEFT)) & RIGHT is String, and the "&" is the standard string concatenation operator... Or please show me the parenthese I'm eventually missing ?

Actually it is the double parenthesis after LENGTH(LEFT)) that I messed up.  Tried to fix/delete the post almost immediately, but Google Groups was having none of it--wouldn't even let me post a correction.