Add missing test coverage for constructor kinds, generic function typedefs and generic mixin applic…

Add missing test coverage for constructor kinds, generic function typedefs, generic mixin applications, delegate, type arguments are type variables.

BUG=
R=gbracha@google.com

Committed: https://code.google.com/p/dart/source/detail?r=28270

[rmacnak, 2013-10-02 20:21:49]

Copy over delegate test from abandoned CL.

[gbracha, 2013-10-02 20:57:14]

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
File tests/lib/mirrors/generic_function_typedef_test.dart (right):

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_function_typedef_test.dart:63:
Expect.isTrue(predicateOfNum.referent.isOriginalDeclaration);
Subtle, but correct. It would be worth adding a comment explaining why this is
so.

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_function_typedef_test.dart:73:
predicateOfNum.referent.parameters[0].type);
Er, why isn't this num?

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_function_typedef_test.dart:75:
transformOfString.referent.parameters[0].type);
and this String?

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_function_typedef_test.dart:86:
transformOfString.referent.returnType);
Should be String, no?

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
File tests/lib/mirrors/generic_mixin_applications_test.dart (right):

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_mixin_applications_test.dart:40: // The extact name of
the type variable is not specified.
extact -> exact

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_mixin_applications_test.dart:42: Expect.equals(2,
reflectClass(NonGenericClass1).superclass.typeVariables.length);
reflectClass(NonGenericClass1).superclass is a mirror on Super with Mixin, aka
Super<dynamic> with Mixin<dynamic>, and should be the same as 

class SMdyn = Super<dynamic> with Mixin<dynamic;

why does this have any type variables?

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_mixin_applications_test.dart:43: Expect.equals(2,
reflectClass(NonGenericClass2).superclass.typeVariables.length);
Again, why type variables?

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_mixin_applications_test.dart:44: Expect.equals(2,
reflectClass(GenericClass1).superclass.typeVariables.length);
Ok. Here the superclass is Super<C> with Mixin<C> which could be expressed as

class SMCC<C1, C2> = Super<C1> with Mixin<C2>;

so would write

class GenericClass1<C> extends SMCC<C, C> ....

and yet, we could also have

class SMC<C1> = Super<C1> with Mixin<C1>;

so would write

class GenericClass1<C> extends SMC<C> ....

https://codereview.chromium.org/25609003/diff/5001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_mixin_applications_test.dart:70: // The extact name of
the type variable is not specified.
extact -> exact

[gbracha, 2013-10-02 21:20:39]

https://codereview.chromium.org/25609003/diff/8001/tests/lib/mirrors/generic_...
File tests/lib/mirrors/generic_mixin_applications_test.dart (right):

https://codereview.chromium.org/25609003/diff/8001/tests/lib/mirrors/generic_...
tests/lib/mirrors/generic_mixin_applications_test.dart:41: // Why 2?
So the 2 is the VM's implementation leaking information. We think we can give
the correct view via mirrors and hide these details.  An anonymous mixin
application never has type variables or type arguments, and it is its own
original declaration.

[rmacnak, 2013-10-02 22:20:33]

Address comments.
Add test for type arguments being type variables.

[rmacnak, 2013-10-02 22:21:31]

+cc Sarah

[gbracha, 2013-10-02 22:36:06]

lgtm

[rmacnak, 2013-10-03 17:55:26]

Added issue numbers

[rmacnak, 2013-10-04 17:32:29]

Committed patchset #5 manually as r28270 (presubmit successful).