Re: Array of records with default values not propagating to array

[Bojan Bozovic <bozovic.bojan@gmail.com>]

On Sunday, February 4, 2018 at 11:07:10 AM UTC+1, Niklas Holsti wrote:
> On 18-02-04 06:59 , Bojan Bozovic wrote:
> > On Saturday, February 3, 2018 at 12:06:52 PM UTC+1, Bojan Bozovic wrote:
> >> When I defined a record, and set a default value for some of its
> >> components, these don't propagate when I make array of these
> >> records. Here is code that can be compiled, but I wonder do I
> >> need to explicitly assign value for 'others' in 'Initialize'
> >> procedure.
> 
> This is just a rule in Ada: in a record aggregate, one has to specify 
> values for all the "needed components", whether or not the record type 
> declaration supplies default values for some components.
> 
> This is different from a subprogram call, in which one can omit the 
> parameter associations for parameters that have a default value.
> 
> The default values of record components are used in two cases: when one 
> creates an object of the type, without giving an initial value, such as 
> (in your code):
> 
>        Temporary_Data : Card_Info;
> 
> Here, the default values in the Card_Info declaration come into play, 
> and define the initial values of those components of Temporary_Data.
> 
> Here it means that Temporary_Data.Trump is False, and Name and 
> Divinatory_Meaning are null unbounded strings. The component 
> Temporary_Data.Trump_Value has an undefined value, as it is not given 
> any default value in the declaration of Card_Info.
> 
> However, in your code, defining explicit default values of 
> Null_Unbounded_String for the components Name and Divinatory_Meaning is 
> unnecessary, because every object of type Unbounded_String is 
> initialised, by default, to Null_Unbounded_String.
> 
> The second case where default values of record components are used is in 
> a record aggreggate where a component association says " ... => <>". 
> Here, the "<>" means "use the default value, if there is one".
> 
> So, in your code, if a record aggregate for a Card_Info record says "... 
> => <>" for the components Name and Divinatory_Meaning, it means that 
> these components will be given the value Null_Unbounded_String.
> 
> >> When I omit "others =>" in Initialize, I get compilation errors
> >> at these lines both on FSF compiler and AdaCore GPL compiler on
> >> Windows. Is this intended behavior or I am missing something?
> 
> It is intended behaviour. The relevant Reference Manual rule is 
> http://www.ada-auth.org/standards/rm12_w_tc1/html/RM-4-3-1.html#p16 .
> 
> > The compiler does compile assignment of others => <> to uninitialized
> > record components as Mr. Simon Wright noticed above,
> 
> Why do you believe that the components (Name and Divinatory_Meaning) are 
> uninitialised? In my tests of the code, they are initialised, as 
> expected, to Null_Unbounded_String.
> 
> > I have reproduced the bug both with GNAT 2017 GPL and FSF GNAT on
> > Cygwin (6.4.0).
> 
> So far, I don't think there is any bug.
> 
> -- 
> Niklas Holsti
> Tidorum Ltd
> niklas holsti tidorum fi
>        .      @       .

When in record declaration making components uninitialized like:

Name : Ada.Strings.Unbounded.Unbounded_String;

I had made a mistake to believe these had no proper initialization and shouldn't have been assigned by others => <> in assignment during Initialize, but explicitly, as Simon Wright originally said. I don't know enough about the language itself to discern if something isn't implemented correctly by the compiler, but it is, as Simon Wright rectified his stance on this, so I just sent some spam to AdaCore about that, but I don't think it matters. ;)