C++ grammar refactoring

[guwirth]

Align grammar with latest C++ standard.

e.g. @Bertk By the way, the sequence of the statements within postfixExpression is not correct and the definition for postfix-expression [ braced-init-list ]is missing. New parser errors occurred after moving primary-expression to the binginning of the rule and I stopped this activity.
Excerpt from N4527 Annex A Grammar summary:

postfix-expression:
primary-expression
postfix-expression [ expression ]
postfix-expression [ braced-init-list ]
postfix-expression ( expression-listopt)
simple-type-specifier ( expression-listopt)
typename-specifier ( expression-listopt)
simple-type-specifier braced-init-list
typename-specifier braced-init-list
postfix-expression . templateopt id-expression
postfix-expression -> templateopt id-expression
postfix-expression . pseudo-destructor-name
postfix-expression -> pseudo-destructor-name
postfix-expression ++
postfix-expression --
dynamic_cast < type-id > ( expression )
static_cast < type-id > ( expression )
reinterpret_cast < type-id > ( expression )
const_cast < type-id > ( expression )
typeid ( expression )
typeid ( type-id )

[guwirth]

From @Bertk: By the way, the C++ standard (n4527) defines

class-name:
      identifier
      simple-template-id

but we use :

  b.firstOf(
    simpleTemplateId,
    b.sequence(b.optional("::"), IDENTIFIER)
    )
  );

[guwirth]

There are some lines which shall be checked as well - I feel the b.optional("::") should be removed:

Line 960:     b.rule(usingDirective).is(b.optional(attributeSpecifier), CxxKeyword.USING, CxxKeyword.NAMESPACE, b.optional("::"), b.optional(nestedNameSpecifier), namespaceName, ";");
Line 1461:             b.sequence(b.optional("::"), b.optional(nestedNameSpecifier), typeName, b.optional("^")),
Line 1462:             b.sequence(b.optional("::"), b.optional(nestedNameSpecifier), CxxKeyword.TEMPLATE, templateId, "^")
Line 1519:         b.sequence(b.optional("::"), b.optional(nestedNameSpecifier), innerSimpleTemplateId),
Line 1520:         b.sequence(b.optional("::"), b.optional(nestedNameSpecifier), CxxKeyword.TEMPLATE, innerSimpleTemplateId),

[guwirth]

Some of the issues are fixed/improved with #743. Bigger refactoring is difficult without having better parser error information #476.

[guwirth]

@Bertk there are also some issues from you solved. Maybe you can also test it?

[Bertk]

I did a small test and it is OK. Did anybody check the performance impact? I got the impression the analysis takes some minutes more but this might also related to other modifications.
I will do some more tests in the evening.

[jmecosta]

for me now its also ok, did not see any impact in this small project i run