http://d.puremagic.com/issues/show_bug.cgi?id=11258

           Summary: Static field access problem
           Product: D
           Version: D2
          Platform: x86
        OS/Version: Windows
            Status: NEW
          Keywords: rejects-valid
          Severity: regression
          Priority: P2
         Component: DMD
        AssignedTo: nobody@puremagic.com
        ReportedBy: bearophile_hugs@eml.cc


--- Comment #0 from bearophile_hugs@eml.cc 2013-10-14 05:15:19 PDT ---
struct Foo {
    static int spam = 1;
}
struct Bar {
    static assert(__traits(compiles, { auto x = Foo.spam; })); // line5
    static void foo() {
        static assert(__traits(compiles, { auto x = Foo.spam; }));
    }
}
void main() {}


With dmd 2.064beta this program gives:


test.d(5): Error: static assert  (__traits(compiles, ()

{

auto x = Foo.spam;

}

)) is false


In dmd 2.063 this used to work.

(Additionally, I'd like those error messages to show less newlines).

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=11258



--- Comment #1 from Kenji Hara <k.hara.pg@gmail.com> 2013-10-14 06:45:10 PDT ---
(In reply to comment #0)
> In dmd 2.063 this used to work.

This is a regression from 2.062, and same error occurs with 2.063.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=11258


Kenji Hara <k.hara.pg@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID


--- Comment #2 from Kenji Hara <k.hara.pg@gmail.com> 2013-10-14 07:26:07 PDT ---
(In reply to comment #1)
> (In reply to comment #0)
> > In dmd 2.063 this used to work.
> 
> This is a regression from 2.062, and same error occurs with 2.063.

Sorry, this is NOT a regression. Current behavior is intended.

In class/struct DeclDefs scope, directly using function literal is not allowed.

struct Bar {
    int var;
    auto dg = { return var; };
    // Error: function literals cannot be class members
}

Against that, __traits(compiles) has *no* special handling. Therefore,

struct Bar {
    int var;
    static assert(__traits(compiles, { return var; }));
    // catch "cannot be class members" error, then returns false
}

The static assert correctly fails. Therefor current behavior is intended.

On the other hand, this error check is stopped inside typeof operator. It's also an intended behavior.

So, in this case, please use is(typeof({ ... })) instead.

====

Note: The old bug, existed in 2.062 and earlier, was fixed in here: https://github.com/D-Programming-Language/dmd/commit/0679c4c31ba78853873933324c69cd90e7714d15#diff-ffa5582af7d723c487d4a11ac6743b85R84

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=11258



--- Comment #3 from bearophile_hugs@eml.cc 2013-10-14 09:42:37 PDT ---
(In reply to comment #2)

> Sorry, this is NOT a regression. Current behavior is intended.

Thank you Kenji, this was tricky for me.


Regarding the extra newlines in the error message:


test.d(5): Error: static assert  (__traits(compiles, ()

{

auto x = Foo.spam;

}

)) is false


Do you want me to open another "minor importance" bug report on it?

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=11258



--- Comment #4 from Kenji Hara <k.hara.pg@gmail.com> 2013-10-14 10:11:39 PDT ---
(In reply to comment #3)
> Regarding the extra newlines in the error message:
> 
> 
> test.d(5): Error: static assert  (__traits(compiles, ()
> 
> {
> 
> auto x = Foo.spam;
> 
> }
> 
> )) is false
> 
> 
> Do you want me to open another "minor importance" bug report on it?

I hope you that.
And I'd ask you one question: If the lambda has quite big body, how it will be
output?

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=11258



--- Comment #5 from bearophile_hugs@eml.cc 2013-10-14 15:05:50 PDT ---
(In reply to comment #4)

> I hope you that.
> And I'd ask you one question: If the lambda has quite big body, how it will be
> output?

OK, I have opened Issue 11263 .

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------