D issues are now tracked on GitHub. This Bugzilla instance remains as a read-only archive.
Issue 9152 - Regression in type inference of array of delegates
Summary: Regression in type inference of array of delegates
Status: RESOLVED INVALID
Alias: None
Product: D
Classification: Unclassified
Component: dmd (show other issues)
Version: D2
Hardware: All All
: P2 regression
Assignee: No Owner
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-12-13 09:52 UTC by hsteoh
Modified: 2012-12-13 10:00 UTC (History)
0 users

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this issue.
Description hsteoh 2012-12-13 09:52:16 UTC
This code used to compile in 2.058:

        alias void delegate(int) DgType;

        void dispatch(DgType[] funcTable, int idx, int arg) {
                funcTable[idx](arg);
        }
       
        void main() {
                dispatch([
                        (int x) { return x+x; },
                        (int x) { return x*x; },
                        (int x) { return x^^2; }
                ], 0, 1);
        }

Since 2.059, dmd is unable to correctly infer the type of the delegate array literal:

        test.d(9): Error: function test.dispatch (void delegate(int)[] funcTable, int idx, int arg) is not callable using argument types (int function(int x) pure nothrow @safe[],int,int)
        test.d(9): Error: cannot implicitly convert expression ([delegate pure nothrow @safe int(int x)
        {
        return x + x;
        }
        , delegate pure nothrow @safe int(int x)
        {
        return x * x;
        }
        , delegate pure nothrow @safe int(int x)
        {
        return int __powtmp8 = x;
         , __powtmp8 * __powtmp8;
        }
        ]) of type int function(int x) pure nothrow @safe[] to void delegate(int)[]


In this case, one can add @safe to the alias as a workaround, but it doesn't work if the same function needs to be called with another array that contains a @system delegate.
Comment 1 hsteoh 2012-12-13 10:00:13 UTC
Oops, the code snippet is wrong. I'll have to reduce the actual failing code again and file another bug.