www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.learn - Cannot initialize associative array.

reply dcoder <dcoder devnull.dev> writes:
Hello.  I have the following d code which fails to compile.  All 3
initializations of mywords fail.  What am I doing wrong?

thanks.


$ dmd --help
Digital Mars D Compiler v2.042
Copyright (c) 1999-2010 by Digital Mars written by Walter Bright
Documentation: http://www.digitalmars.com/d/2.0/index.html
Usage:


import std.algorithm, std.stdio, std.string;

//uint[string] mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
//auto mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
const uint[const string] mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1,
"Dog" : 1 ];

void main() {
  foreach( line; stdin.byLine()) {
    foreach( word; split(strip(line))) {
      string s = word.idup;
      if( s in mywords)
	writefln( "%s is a word I know.", s);
      else
	writefln( "%s is not a word I know.", s);
    }
  }

  return;
}
Jun 22 2010
parent reply dcoder <dcoder devenull.dev> writes:
Sorry, I forgot to put some compiler output:

For the declaration: uint[string] mywords = [ "Hello" : 1, "World" : 1, "Cat" :
1,
"Dog" : 1 ];


I get:

$ dmd test_01.d
test_01.d(3): Error: non-constant expression
["Hello":1u,"World":1u,"Cat":1u,"Dog":1u]
Jun 22 2010
parent reply Bernard Helyer <b.helyer gmail.com> writes:
On Tue, 22 Jun 2010 21:32:48 +0000, dcoder wrote:


 Sorry, I forgot to put some compiler output:
 
 For the declaration: uint[string] mywords = [ "Hello" : 1, "World" : 1,
 "Cat" : 1, "Dog" : 1 ];
 
 
 I get:
 
 $ dmd test_01.d
 test_01.d(3): Error: non-constant expression
 ["Hello":1u,"World":1u,"Cat":1u,"Dog":1u]


AAs can't be assigned to at compile time (:[). You'll have to use a 
static constructor or an initiliasation function to assign to one at 
global scope.
Jun 22 2010
next sibling parent bearophile <bearophileHUGS lycos.com> writes:
Bernard Helyer:

 AAs can't be assigned to at compile time (:[).


You can define enum ones, this works:

import std.stdio;
enum int[string] aa = ["foo": 10];
void main() {
    writeln(cast(bool)("foo" in aa));
    writeln(aa["foo"]);
    writeln(cast(bool)("hello" in aa));
}


But this code:

import std.stdio;
immutable int[string] aa = ["foo": 10];
void main() {
    writeln(cast(bool)("foo" in aa));
    writeln(aa["foo"]);
    writeln(cast(bool)("hello" in aa));
}

Raises the compilation error:
test.d(2): Error: non-constant expression ["foo":10]
In theory aa here is a constant expression :-)

Bye,
bearophile
Jun 22 2010
prev sibling parent reply dcoder <dcoder devnull.dev> writes:
 For the declaration:
 uint[string] mywords = [ "Hello" : 1, "World" : 1,
 "Cat" : 1, "Dog" : 1 ];


 I get:

 $ dmd test_01.d
 test_01.d(3): Error: non-constant expression
 ["Hello":1u,"World":1u,"Cat":1u,"Dog":1u]





 AAs can't be assigned to at compile time (:[). You'll have to use a
 static constructor or an initiliasation function to assign to one at
 global scope.


thanks Bernard.

I see, so for global associative arrays, I can't do it.  That's good to know,
and
I guess that will force me into better coding habits.  Anyways, I just wish the
compiler could give me a better message than what I got.  :(

So, I moved the initialization to inside the main function, and now it works.
Great.  I think we need to put this question in the FAQ.

Many thanks.
Jun 22 2010
parent reply =?UTF-8?B?QWxpIMOHZWhyZWxp?= <acehreli yahoo.com> writes:
dcoder wrote:


 So, I moved the initialization to inside the main function, and now 


it works.

 Great.  I think we need to put this question in the FAQ.


For future reference, if it really needs to be global:

uint[string] mywords;

static this()
{
     mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
}

Ali
Jun 22 2010
next sibling parent reply =?UTF-8?B?QWxpIMOHZWhyZWxp?= <acehreli yahoo.com> writes:
Ali Çehreli wrote:

 dcoder wrote:
 
  > So, I moved the initialization to inside the main function, and now 
 it works.
  > Great.  I think we need to put this question in the FAQ.
 
 For future reference, if it really needs to be global:
 
 uint[string] mywords;
 
 static this()
 {
     mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
 }


Could someone please verify whether the above is really necessary? Is it 
actually a dmd bug that we need to use 'static this()' to initialize an 
associative array?

Ali
Jun 23 2010
next sibling parent Pelle <pelle.mansson gmail.com> writes:
On 06/23/2010 09:41 AM, Ali Çehreli wrote:

 Ali Çehreli wrote:

 dcoder wrote:


 So, I moved the initialization to inside the main function, and now


 it works.

 Great. I think we need to put this question in the FAQ.


 For future reference, if it really needs to be global:

 uint[string] mywords;

 static this()
 {
 mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
 }


 Could someone please verify whether the above is really necessary? Is it
 actually a dmd bug that we need to use 'static this()' to initialize an
 associative array?

 Ali


I say it's a bug, a literal like that should be a constant expression. 
Someone report it!
Jun 23 2010
prev sibling next sibling parent bearophile <bearophileHUGS lycos.com> writes:
Ali Çehreli:

 Could someone please verify whether the above is really necessary?


An initialization inside some runtime function/initializator is necessary
unless the AA is an enum.



 Is it 
 actually a dmd bug that we need to use 'static this()' to initialize an 
 associative array?


According to the way D AAs are designed it's not a bug. But of course you can
think of possible enhancements or changes to the design.

Bye,
bearophile
Jun 23 2010
prev sibling parent reply Bernard Helyer <b.helyer gmail.com> writes:
On Wed, 23 Jun 2010 00:41:45 -0700, Ali Çehreli wrote:


 Ali Çehreli wrote:

 dcoder wrote:
 
  > So, I moved the initialization to inside the main function, and now
 it works.
  > Great.  I think we need to put this question in the FAQ.
 
 For future reference, if it really needs to be global:
 
 uint[string] mywords;
 
 static this()
 {
     mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
 }


 
 Could someone please verify whether the above is really necessary? Is it
 actually a dmd bug that we need to use 'static this()' to initialize an
 associative array?
 
 Ali


I can't remember where exactly I read it, but there's a line in the docs 
specifically forbidding the use of AAs in constant expressions, so it's 
by design AFAIK.
Jun 23 2010
parent reply Andrej Mitrovic <andrej.mitrovich gmail.com> writes:
Bernard Helyer Wrote:


 On Wed, 23 Jun 2010 00:41:45 -0700, Ali Çehreli wrote:
 

 Ali Çehreli wrote:

 dcoder wrote:
 
  > So, I moved the initialization to inside the main function, and now
 it works.
  > Great.  I think we need to put this question in the FAQ.
 
 For future reference, if it really needs to be global:
 
 uint[string] mywords;
 
 static this()
 {
     mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
 }


 
 Could someone please verify whether the above is really necessary? Is it
 actually a dmd bug that we need to use 'static this()' to initialize an
 associative array?
 
 Ali


 
 I can't remember where exactly I read it, but there's a line in the docs 
 specifically forbidding the use of AAs in constant expressions, so it's 
 by design AFAIK.


http://www.digitalmars.com/d/2.0/expression.html#ArrayLiteral

"Associative array literals are a comma-separated list of key:value pairs...
*An AssocArrayLiteral cannot be used to statically initialize anything. *"
Jun 24 2010
parent Andrej Mitrovic <andrej.mitrovich gmail.com> writes:
Andrej Mitrovic Wrote:


 Bernard Helyer Wrote:
 

 On Wed, 23 Jun 2010 00:41:45 -0700, Ali Çehreli wrote:
 

 Ali Çehreli wrote:

 dcoder wrote:
 
  > So, I moved the initialization to inside the main function, and now
 it works.
  > Great.  I think we need to put this question in the FAQ.
 
 For future reference, if it really needs to be global:
 
 uint[string] mywords;
 
 static this()
 {
     mywords = [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
 }


 
 Could someone please verify whether the above is really necessary? Is it
 actually a dmd bug that we need to use 'static this()' to initialize an
 associative array?
 
 Ali


 
 I can't remember where exactly I read it, but there's a line in the docs 
 specifically forbidding the use of AAs in constant expressions, so it's 
 by design AFAIK.


 
 http://www.digitalmars.com/d/2.0/expression.html#ArrayLiteral
 
 "Associative array literals are a comma-separated list of key:value pairs...
 *An AssocArrayLiteral cannot be used to statically initialize anything. *"


Woops, correct link:
http://www.digitalmars.com/d/2.0/expression.html#AssocArrayLiteral
Jun 24 2010
prev sibling parent "Rory McGuire" <rmcguire neonova.co.za> writes:
On Wed, 23 Jun 2010 00:30:40 +0200, Ali =C3=87ehreli <acehreli yahoo.com> w=
rote:


 dcoder wrote:

  > So, I moved the initialization to inside the main function, and now =


=20

 it works.
  > Great.  I think we need to put this question in the FAQ.

 For future reference, if it really needs to be global:

 uint[string] mywords;

 static this()
 {
      mywords =3D [ "Hello" : 1, "World" : 1, "Cat" : 1, "Dog" : 1 ];
 }

 Ali




 from what I have read in TDPL so far (about half way now), there is no =20
mention of this limitation.
It just says that [key:value] is how you would statically initialize a AA.

Does this mean that it is just a D implementation issue. To me its seems =
=20
like this should be a defined part
of the language. Also the compiler should really rewrite it to your above =
=20
code anyway, surely?

-Rory
Jun 22 2010