www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - D Uniform initialization {}

reply Patric Dexheimer <patric.dexheimer gmail.com> writes:
Quite sure that this was already discussed, but.. any chance of 
this on D?
(one of the few things that i miss from c++)

There are a lot of places where it should make the code clear.
I always have to create shorter aliases for the most used 
structs. (which i think is awkward sometimes)

I know there is the case of being ambiguous with lambdas, but 
after reading this thread 
https://forum.dlang.org/thread/nud21i$o29$1 digitalmars.com
uniform initialization comes to my mind again :)
Oct 21 2016
parent reply Daniel Kozak via Digitalmars-d <digitalmars-d puremagic.com> writes:
Dne 21.10.2016 v 20:49 Patric Dexheimer via Digitalmars-d napsal(a):


 Quite sure that this was already discussed, but.. any chance of this 
 on D?


No (I hope so)

 There are a lot of places where it should make the code clear.


Can you elaborate on this?

 I always have to create shorter aliases for the most used structs. 
 (which i think is awkward sometimes)


Why? (I do not see any relation to Uniform initialization)
Oct 21 2016
parent reply Patric Dexheimer <patric.dexheimer gmail.com> writes:
On Friday, 21 October 2016 at 19:20:25 UTC, Daniel Kozak wrote:

 Dne 21.10.2016 v 20:49 Patric Dexheimer via Digitalmars-d 
 napsal(a):


 Quite sure that this was already discussed, but.. any chance 
 of this on D?


 No (I hope so)

 There are a lot of places where it should make the code clear.


 Can you elaborate on this?

 I always have to create shorter aliases for the most used 
 structs. (which i think is awkward sometimes)


 Why? (I do not see any relation to Uniform initialization)


egs:
//D
alias vec3 = Tuple!(float, "x", float, "y", float, "z");
vec3[] vectors = [
    vec3(1.0,0.0,1.0),
    vec3(2.0,1.0,1.0),
    vec3(3.0,2.0,1.0)
];
//C++ equivalent
vec3 vectors[] = {
    {1.0,0.0,1.0},
    {2.0,1.0,1.0},
    {3.0,2.0,1.0}
};

//D
auto return_value = get_struct(); //don´t need to write the 
return type
set_struct( StructName(value1, value2) );
//C++
set_struct( {value1, value2} ); //don´t need to write the 
argument type

//D in case of large struct names
alias v = VeryLargeStructName; //not cool
v[] vectors = [
    v(1.0,0.0,1.0),
    v(2.0,1.0,1.0),
    v(3.0,2.0,1.0)
];


I find myself falling with frequency on examples that will 
benefit from the c++ uniform initialization.

"No (I hope so)"

Can you explain why you think is a bad idea?
Oct 21 2016
parent reply Daniel Kozak via Digitalmars-d <digitalmars-d puremagic.com> writes:
Dne 21.10.2016 v 23:21 Patric Dexheimer via Digitalmars-d napsal(a):


 On Friday, 21 October 2016 at 19:20:25 UTC, Daniel Kozak wrote:

 Dne 21.10.2016 v 20:49 Patric Dexheimer via Digitalmars-d napsal(a):


 Quite sure that this was already discussed, but.. any chance of this 
 on D?


 No (I hope so)

 There are a lot of places where it should make the code clear.


 Can you elaborate on this?

 I always have to create shorter aliases for the most used structs. 
 (which i think is awkward sometimes)


 Why? (I do not see any relation to Uniform initialization)


 egs:
 //D
 alias vec3 = Tuple!(float, "x", float, "y", float, "z");
 vec3[] vectors = [
    vec3(1.0,0.0,1.0),
    vec3(2.0,1.0,1.0),
    vec3(3.0,2.0,1.0)
 ];
 //C++ equivalent
 vec3 vectors[] = {
    {1.0,0.0,1.0},
    {2.0,1.0,1.0},
    {3.0,2.0,1.0}
 };


this works for D too:

import std.stdio;

struct S
{
     int a;
     int b;
}
void main()
{
     S[] s = [{ 1, 2 }];
     writeln(s[0]);
}

 //D
 auto return_value = get_struct(); //don´t need to write the return type
 set_struct( StructName(value1, value2) );
 //C++
 set_struct( {value1, value2} ); //don´t need to write the argument type


OK this does not work but I do not think it is releated to Uniform 
initialization, but it is more something like cast to parametr type or 
something like that

 Can you explain why you think is a bad idea?


Because there is no need. In c++ it is disaster because there is milion 
way how to initialize something, it is really hard to understand and 
inconsistent
Oct 21 2016
parent reply Patric Dexheimer <patric.dexheimer gmail.com> writes:
S[] s = [{ 1, 2 }];

Nice, did´n knew that it worked.

On Friday, 21 October 2016 at 21:41:16 UTC, Daniel Kozak wrote:

 Because there is no need. In c++ it is disaster because there 
 is milion way how to initialize something, it is really hard to 
 understand and inconsistent


I never really felt lost about it with c++, but the argument 
holds true anyway :)
Oct 22 2016
parent reply Seb <seb wilzba.ch> writes:
On Saturday, 22 October 2016 at 21:26:53 UTC, Patric Dexheimer 
wrote:

 S[] s = [{ 1, 2 }];

 Nice, did´n knew that it worked.

 On Friday, 21 October 2016 at 21:41:16 UTC, Daniel Kozak wrote:

 Because there is no need. In c++ it is disaster because there 
 is milion way how to initialize something, it is really hard 
 to understand and inconsistent


 I never really felt lost about it with c++, but the argument 
 holds true anyway :)


There has been a abandoned proposal for struct initialization:
https://github.com/dlang/DIPs/pull/22
Oct 22 2016
parent mogu <mogucpp 163.com> writes:
On Sunday, 23 October 2016 at 01:31:47 UTC, Seb wrote:

 On Saturday, 22 October 2016 at 21:26:53 UTC, Patric Dexheimer 
 wrote:

 S[] s = [{ 1, 2 }];

 Nice, did´n knew that it worked.

 On Friday, 21 October 2016 at 21:41:16 UTC, Daniel Kozak wrote:

 Because there is no need. In c++ it is disaster because there 
 is milion way how to initialize something, it is really hard 
 to understand and inconsistent


 I never really felt lost about it with c++, but the argument 
 holds true anyway :)


 There has been a abandoned proposal for struct initialization:
 https://github.com/dlang/DIPs/pull/22


It has been closed only because of inactivity.
Oct 22 2016