• soarowl (13/13) Sep 02 2010 private import
• bearophile (14/16) Sep 02 2010 You are right, there's a bug in the AA when not int keys are used:
• Steven Schveighoffer (3/17) Sep 02 2010 http://d.puremagic.com/issues/show_bug.cgi?id=4201

soarowl <soarowl yeah.net> writes:

private import
    std.stdio;

static ubyte[] data = [1, 3, 5, 7, 9, 7, 5];

void main()
{
    int[ubyte] set;
    foreach(e; data) set[e]++;
    //foreach(k, v; set) std.stdio.writef("%d: %d ", k, v);
    foreach(e; set.keys.sort) std.stdio.writef("%d ", e);
    //foreach(e; set.values) std.stdio.writef("%d ", e);
}



I expect result: 1 3 5 7 9
But result will be: 0 0 0 1 5 or something else, which not equals expaect
result.

bearophile <bearophileHUGS lycos.com> writes:

soarowl:

 I expect result: 1 3 5 7 9
 But result will be: 0 0 0 1 5 or something else, which not equals expaect
result.

You are right, there's a bug in the AA when not int keys are used:

import std.stdio: writeln;
ubyte[] data = [1, 3, 5, 7, 9, 7, 5];
void main() {
    int[ubyte] set;
    foreach (x; data)
        set[x]++;
    writeln(set.keys);
}

I think there is already a bug report about this. It's an important bug, makes
AAs partially useless.

For your next bug reports in "freeform" like this one it is better to use
D.learn newsgroup.

Bye,
bearophile

"Steven Schveighoffer" <schveiguy yahoo.com> writes:

On Thu, 02 Sep 2010 06:28:14 -0400, soarowl <soarowl yeah.net> wrote:

 private import
     std.stdio;

 static ubyte[] data = [1, 3, 5, 7, 9, 7, 5];

 void main()
 {
     int[ubyte] set;
     foreach(e; data) set[e]++;
     //foreach(k, v; set) std.stdio.writef("%d: %d ", k, v);
     foreach(e; set.keys.sort) std.stdio.writef("%d ", e);
     //foreach(e; set.values) std.stdio.writef("%d ", e);
 }



 I expect result: 1 3 5 7 9
 But result will be: 0 0 0 1 5 or something else, which not equals  
 expaect result.

http://d.puremagic.com/issues/show_bug.cgi?id=4201

-Steve