• d-bugmail puremagic.com (32/32) Sep 30 2009 http://d.puremagic.com/issues/show_bug.cgi?id=3353
• d-bugmail puremagic.com (50/50) Sep 30 2009 http://d.puremagic.com/issues/show_bug.cgi?id=3353
• d-bugmail puremagic.com (12/12) Oct 06 2009 http://d.puremagic.com/issues/show_bug.cgi?id=3353

d-bugmail puremagic.com writes:

http://d.puremagic.com/issues/show_bug.cgi?id=3353

           Summary: storage class of a member function is propagated to
                    default arguments
           Product: D
           Version: 2.032
          Platform: Other
        OS/Version: Windows
            Status: NEW
          Keywords: rejects-valid
          Severity: normal
          Priority: P2
         Component: DMD
        AssignedTo: nobody puremagic.com
        ReportedBy: r.sagitario gmx.de



PDT ---
Compiling the module test.d:

public struct foo
{
    public this(real aleft)
    {
    } 
}

class bar
{
    final void fun(foo arg = foo(0.)) { } 
}

produces:
test.d(11): Error: variable test.bar.__ctmp1 final cannot be applied to
variable

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

d-bugmail puremagic.com writes:

http://d.puremagic.com/issues/show_bug.cgi?id=3353


Rainer Schuetze <r.sagitario gmx.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |patch



PDT ---
This happens, because the default argument is evaluated in the same scope as
the function. The following patch opens a new scope for argument semantics,
clearing the storage class (maybe some flags should be kept?):

Index: mtype.c
===================================================================
--- mtype.c    (revision 196)
+++ mtype.c    (working copy)
   -4065,11 +4065,14   
     if (tf->parameters)
     {    size_t dim = Argument::dim(tf->parameters);

+    Scope* argsc = sc->push(); // arguments must be evaluated in a different
scope
+    argsc->stc = STCundefined; // as we don't want to inherit storage class
from the function
+
     for (size_t i = 0; i < dim; i++)
     {   Argument *arg = Argument::getNth(tf->parameters, i);

         tf->inuse++;
-        arg->type = arg->type->semantic(loc,sc);
+        arg->type = arg->type->semantic(loc,argsc);
         if (tf->inuse == 1) tf->inuse--;

         arg->type = arg->type->addStorageClass(arg->storageClass);
   -4094,9 +4097,9   

         if (arg->defaultArg)
         {
-        arg->defaultArg = arg->defaultArg->semantic(sc);
-        arg->defaultArg = resolveProperties(sc, arg->defaultArg);
-        arg->defaultArg = arg->defaultArg->implicitCastTo(sc, arg->type);
+        arg->defaultArg = arg->defaultArg->semantic(argsc);
+        arg->defaultArg = resolveProperties(argsc, arg->defaultArg);
+        arg->defaultArg = arg->defaultArg->implicitCastTo(argsc, arg->type);
         }

         /* If arg turns out to be a tuple, the number of parameters may
   -4107,6 +4110,7   
         i--;
         }
     }
+    argsc->pop();
     }
     if (tf->next)
     tf->deco = tf->merge()->deco;

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

d-bugmail puremagic.com writes:

http://d.puremagic.com/issues/show_bug.cgi?id=3353


Walter Bright <bugzilla digitalmars.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |bugzilla digitalmars.com
         Resolution|                            |FIXED



02:19:18 PDT ---
Fixed dmd 1.048 and 2.033

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