______________________________________________________________________

  5   Expressions                                       [expr]

  ______________________________________________________________________

1 This clause defines the syntax, order of evaluation,  and  meaning  of
  expressions.   An  expression  is a sequence of operators and operands
  that specifies a computation.  An expression can result in a value and
  can cause side effects.

2 Operators  can  be  overloaded, that is, given meaning when applied to
  expressions of class type (_class_).  Uses of overloaded operators are
  transformed  into  function  calls as described in _over.oper_.  Over­
  loaded operators obey the rules for syntax specified in  this  clause,
  but the requirements of operand type, lvalue, and evaluation order are
  replaced by the rules for function call.  Relations between operators,
  such  as ++a meaning a+=1, are not guaranteed for overloaded operators
  (_over.oper_).1)

3 This clause defines the operators when applied to types for which they
  have not been overloaded.  Operator overloading shall not  modify  the
  rules  for  the  built-in operators, that is, for operators applied to
  types for which they are defined by  the  language  itself.   However,
  these  built-in  operators  participate  in  overload  resolution; see
  _over.match.oper_.

4 Operators can be regrouped according to the usual  mathematical  rules
  only where the operators really are associative or commutative.  Over­
  loaded operators are never assumed to be associative  or  commutative.
  Except  where noted, the order of evaluation of operands of individual
  operators and subexpressions of individual expressions is unspecified.
  In  particular,  if  a  value  is modified twice in an expression, the
  result of the expression is unspecified except where  an  ordering  is
  guaranteed by the operators involved.  For example,
          i = v[i++];      // the value of `i' is undefined
          i=7,i++,i++;     // `i' becomes 9

5 The  handling  of overflow and divide by zero in expression evaluation
  is implementation dependent.  Most  existing  implementations  of  C++
  ignore  integer  overflows.   Treatment  of  division  by zero and all
  floating  point  exceptions  vary  among  machines,  and  is   usually
  adjustable by a library function.

  _________________________
  1) Nor is it guaranteed for type bool; the left operand  of  +=  shall
  not have type bool.

6 Except  where  noted,  operands  of  types  const T,  volatile T,  T&,
  const T&, and volatile T& can be used as if they  were  of  the  plain
  type  T.  Similarly, except where noted, operands of type T* const and
  T* volatile can be used as if they were of the plain type  T*.   Simi­
  larly,  a  plain  T  can  be  used  where a volatile T or a const T is
  required.  These rules apply in  combination  so  that,  except  where
  noted,  a  T* const volatile can be used where a T* is required.  Such
  uses do not count as standard conversions when considering overloading
  resolution (_over.match_).

7 If  an  expression  initially  has the type reference to T (_dcl.ref_,
  _dcl.init.ref_), the type is adjusted to T prior to any further analy­
  sis,  the  expression designates the object or function denoted by the
  reference, and the expression  is  an  lvalue.   A  reference  can  be
  thought of as a name of an object.

8 An expression designating an object is called an object-expression.

9 User-defined  conversions of class or enum types to and from fundamen­
  tal types, pointers, and so on, can  be  defined  (_class.conv_).   If
  unambiguous  (_over.match_),  such  conversions will be applied by the
  compiler wherever a class object appears as an operand of an  operator
  or as a function argument (_expr.call_).

10Whenever  an  lvalue  expression  appears as an operand of an operator
  that  expects  an  rvalue  for  that  operand,  the   lvalue-to-rvalue
  (_conv.lval_), array-to-pointer (_conv.array_), or function-to-pointer
  (_conv.func_) standard conversion  will  be  applied  to  convert  the
  expression to an rvalue.

11Many  binary  operators  that expect operands of arithmetic type cause
  conversions and yield result types in a similar way.  The  purpose  is
  to  yield  a  common type, which is also the type of the result.  This
  pattern is called the usual arithmetic conversions.

  +-------                 BEGIN BOX 1                -------+
  Enumerations are handled correctly by  the  usual  arithmetic  conver­
  sions,  and  for  any  operator  that invokes the integral promotions.
  However, there may be other places in this Clause that fail  to  treat
  enumerations appropriately.
  +-------                  END BOX 1                 -------+

12
  --If  either operand is of type long double, the other is converted to
    long double.

  --Otherwise, if either operand is double, the other  is  converted  to
    double.

  --Otherwise,  if  either  operand  is float, the other is converted to
    float.

  --Otherwise, the integral promotions (_conv.prom_)  are  performed  on

    both operands.2)

  --Then, if either operand is unsigned long the other is  converted  to
    unsigned long.

  --Otherwise,  if one operand is a long int and the other unsigned int,
    then if a long int can represent all the values of an unsigned  int,
    the unsigned int is converted to a long int; otherwise both operands
    are converted to unsigned long int.

  --Otherwise, if either operand is long,  the  other  is  converted  to
    long.

  --Otherwise,  if either operand is unsigned, the other is converted to
    unsigned.

  --Otherwise, both operands are int.

13If the program attempts to  access  the  stored  value  of  an  object
  through an lvalue of other than one of the following types:

  --the dynamic type of the object,

  --a qualified version of the declared type of the object,

  --a  type  that  is  the  signed or unsigned type corresponding to the
    declared type of the object,

  --a type that is the signed or unsigned type corresponding to a quali­
    fied version of the declared type of the object,

  --an  aggregate  or union type that includes one of the aforementioned
    types among its members (including, recursively, a member of a  sub­
    aggregate or contained union),

  --a  type  that  is  a  (possibly  qualified)  base  class type of the
    declared type of the object,

  --a character type.3) the result is undefined.

  5.1  Primary expressions                                   [expr.prim]

1 Primary  expressions  are  literals, names, and names qualified by the
  scope resolution operator ::.

  _________________________
  2)  As a consequence, operands of type bool, wchar_t, or an enumerated
  type are converted to some integral type.
  3) The intent of this list is to specify those circumstances in  which
  an object may or may not be aliased.

          primary-expression:
                  literal
                  this
                  :: identifier
                  :: operator-function-id
                  :: qualified-id
                  ( expression )
                  id-expression

2 A literal is a primary expression.   Its  type  depends  on  its  form
  (_lex.literal_).

3 In the body of a nonstatic member function (_class.mfct_), the keyword
  this names a pointer to the object for which the function was invoked.
  The  keyword  this  shall  not be used outside a class member function
  body.

  +-------                 BEGIN BOX 2                -------+
  In a  constructor  it  is  common  practice  to  allow  this  in  mem-
  initializers.
  +-------                  END BOX 2                 -------+

4 The operator :: followed by an identifier, a qualified-id, or an oper­
  ator-function-id is a primary expression.  Its type  is  specified  by
  the declaration of the identifier, name, or operator-function-id.  The
  result is the identifier, name, or operator-function-id.   The  result
  is  an  lvalue  if  the  identifier  is.   The identifier or operator-
  function-id shall be of namespace scope.  Use of :: allows a type,  an
  object,  a  function,  or  an enumerator to be referred to even if its
  identifier has been hidden (_basic.scope_).

5 A parenthesized expression is a  primary  expression  whose  type  and
  value  are  identical to those of the unadorned expression.  The pres­
  ence of parentheses does not  affect  whether  the  expression  is  an
  lvalue.

6 A  id-expression is a restricted form of a primary-expression that can
  appear after . and -> (_expr.ref_):
          id-expression:
                  unqualified-id
                  qualified-id

          unqualified-id:
                  identifier
                  operator-function-id
                  conversion-function-id
                  ~ class-name

  +-------                 BEGIN BOX 3                -------+
  Issue: now it's allowed to  invoke  ~int(),  but  ~class-name  doesn't
  allow for that.
  +-------                  END BOX 3                 -------+

7 An  identifier  is  an  id-expression  provided  it  has been suitably
  declared (_dcl.dcl_).   For  operator-function-ids,  see  _over.oper_.
  For  conversion-function-ids, see _class.conv.fct_.  A class-name pre­
  fixed by ~ denotes a destructor; see _class.dtor_.
          qualified-id:
                  nested-name-specifier unqualified-id

8 A nested-name-specifier that names a class (_dcl.type_) followed by ::
  and the name of a member of that class (_class.mem_), or a member of a
  base of that class (_class.derived_), is a qualified-id; its  type  is
  the  data  member  type  or  function member type; it is not an object
  type.  The result is the member.  The result is an lvalue if the  mem­
  ber  is.   The  class-name might be hidden by a nontype name, in which
  case the class-name is still found  and  used.   Where  class-name  ::
  class-name  is  used, and the two class-names refer to the same class,
  this notation names the constructor (_class.ctor_).  Where  class-name
  :: ~  class-name  is used, the two class-names shall refer to the same
  class; this notation names the  destructor  (_class.dtor_).   Multiply
  qualified names, such as N1::N2::N3::n, can be used to refer to nested
  types (_class.nest_).

9 In a qualified-id, if the id-expression is  a  conversion-function-id,
  its  conversion-type-id shall denote the same type in both the context
  in which the entire qualified-id occurs and  in  the  context  of  the
  class  denoted  by the nested-name-specifier.  For the purpose of this
  evaluation, the name, if any, of  each  class  is  also  considered  a
  nested class member of that class.

  5.2  Postfix expressions                                   [expr.post]

1 Postfix expressions group left-to-right.
          postfix-expression:
                  primary-expression
                  postfix-expression [ expression ]
                  postfix-expression ( expression-listopt )
                  simple-type-specifier ( expression-listopt )
                  postfix-expression . id-expression
                  postfix-expression -> id-expression
                  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 )
          expression-list:
                  assignment-expression
                  expression-list , assignment-expression

  5.2.1  Subscripting                                         [expr.sub]

1 A postfix expression followed by an expression in square brackets is a
  postfix expression.  The intuitive meaning is  that  of  a  subscript.
  One  of the expressions shall have the type pointer to T and the other
  shall be of enumeration or integral type.  The result is an lvalue  of
  type T.  The type T shall be complete.  The expression E1[E2] is iden­
  tical  (by  definition)  to  *((E1)+(E2)).    See   _expr.unary_   and
  _expr.add_  for  details  of  *  and  + and _dcl.array_ for details of
  arrays.

  5.2.2  Function call                                       [expr.call]

1 There are two kinds of function call: ordinary function call and  mem­
  ber function4) (_class.mfct_) call.  A  function  call  is  a  postfix
  expression followed by parentheses containing a possibly empty, comma-
  separated list of expressions which constitute the  arguments  to  the
  function.  For ordinary function call, the postfix expression shall be
  a function name, or a pointer or reference to  function.   For  member
  function   call,   the   postfix   expression  shall  be  an  implicit
  (_class.mfct_) or explicit class member access (_expr.ref_) whose  id-
  expression  is  a function member name, or a pointer-to-member expres­
  sion  (_expr.mptr.oper_)  selecting  a  function  member.   The  first
  expression in the postfix expression is then called the object expres­
  sion, and the call is as a member of the object pointed to or referred
  to.   In  the  case  of  an  implicit class member access, the implied
  object is the one pointed to by this.  That is, a member function call
  of  the form f() is interpreted as this->f() (see _class.this_).  If a
  function or member function name is used, the name can  be  overloaded
  (_over_),  in  which  case  the  appropriate function will be selected
  according to the rules in _over.match_.  The function called in a mem­
  ber function call is normally selected according to the static type of
  the object expression (see _class.derived_), but if that  function  is
  virtual  the  function  actually  called  will  be the final overrider
  (_class.virtual_) of the selected function in the dynamic type of  the
  object expression (i.e., the type of the object pointed or referred to
  by  the  current  value  of  the  object  expression).   _class.cdtor_
  describes  the  behavior  of  virtual  function calls when the object-
  expression refers to an object under construction or destruction.

2 The type of the function call expression is the  return  type  of  the
  statically  chosen function (i.e., ignoring the virtual keyword), even
  if the type of the function actually called is different.   This  type
  shall be complete or the type void.

3 When  a  function is called, each parameter (_dcl.fct_) is initialized
  (_dcl.init.ref_, _class.copy_, _class.ctor_)  with  its  corresponding
  argument.   Standard  (_conv_) and user-defined (_class.conv_) conver­
  sions are performed.  The value  of  a  function  call  is  the  value
  returned  by  the called function except in a virtual function call if
  the return type of the final overrider is different  from  the  return
  _________________________
  4)  A static member function (_class.static_) is an ordinary function.

  type  of  the  statically chosen function, the value returned from the
  final overrider is converted to the return type of the statically cho­
  sen  function.   A  function  can change the values of its nonconstant
  parameters, but these changes cannot affect the values  of  the  argu­
  ments  except  where  a  parameter  is  of  a non-const reference type
  (_dcl.ref_).  Where a parameter is of reference type a temporary vari­
  able is introduced if needed (_dcl.type_, _lex.literal_, _lex.string_,
  _dcl.array_, _class.temporary_).  In addition, it is possible to  mod­
  ify the values of nonconstant objects through pointer parameters.

4 A  function  can  be  declared to accept fewer arguments (by declaring
  default arguments (_dcl.fct.default_)) or more arguments (by using the
  ellipsis,  ...   _dcl.fct_) than the number of parameters in the func­
  tion definition (_dcl.fct.def_).

5 If no declaration of the called function is accessible from the  scope
  of  the  call  the  program  is ill-formed.  This implies that, except
  where the ellipsis (...)  is used, a parameter is available  for  each
  argument.

6 Any  argument  of  type  float for which there is no parameter is con­
  verted to double before the call; any of char, short, or  a  bit-field
  type  for which there is no parameter are converted to int or unsigned
  by integral promotion (_conv.prom_).  Any argument of enumeration type
  is converted to int, unsigned, long, or unsigned long by integral pro­
  motion.  An object of a class for which no parameter  is  declared  is
  passed as a data structure.

  +-------                 BEGIN BOX 4                -------+
  To  ``pass  a parameter as a data structure'' means, roughly, that the
  parameter must be a PODS, and that otherwise  the  behavior  is  unde­
  fined.  This must be made more precise.
  +-------                  END BOX 4                 -------+

7 An  object  of  a class for which a parameter is declared is passed by
  initializing the parameter with the argument  by  a  constructor  call
  before the function is entered (_class.temporary_, _class.copy_).

8 The  order  of  evaluation of arguments is unspecified; take note that
  compilers differ.  All  side  effects  of  argument  expressions  take
  effect before the function is entered.  The order of evaluation of the
  postfix expression and the argument expression list is unspecified.

9 The function-to-pointer  standard  conversion  (_conv.func_)  is  sup­
  pressed on the postfix expression of a function call.

10Recursive calls are permitted.

11A  function call is an lvalue if and only if the result type is a ref­
  erence.

  5.2.3  Explicit type conversion (functional           [expr.type.conv]
       notation)

1 A  simple-type-specifier  (_dcl.type_)  followed  by  a  parenthesized
  expression-list constructs a value of the  specified  type  given  the
  expression list.  If the expression list specifies a single value, the
  expression is equivalent (in definedness, and if defined  in  meaning)
  to the corresponding cast expression (_expr.cast_).  If the expression
  list specifies more than a single value, the type  shall  be  a  class
  with  a  suitably declared constructor (_dcl.init_, _class.ctor_), and
  the expression T(x1, x2, ...)  is equivalent in effect to the declara­
  tion  T t(x1,  x2,  ...); for some invented temporary variable t, with
  the result being the value of t as an rvalue.

2 A simple-type-specifier (_dcl.type_) followed by  a  (empty)  pair  of
  parentheses  constructs a value of the specified type.  If the type is
  a class with a default constructor  (_class.ctor_),  that  constructor
  will  be  called; otherwise the result is the default value given to a
  static object of the specified type.  See also (_expr.cast_).

  5.2.4  Class member access                                  [expr.ref]

1 A postfix expression followed by a dot .)  or an arrow ->) followed by
  an  id-expression  is  a  postfix  expression.  The postfix expression
  before the dot or arrow is evaluated;5) the result of that evaluation,
  together  with  the  id-expression, determine the result of the entire
  postfix expression.

2 For the first option (dot) the  type  of  the  first  expression  (the
  object  expression)  shall  be class object (of a complete type).  For
  the second option (arrow)  the  type  of  the  first  expression  (the
  pointer  expression)  shall  be pointer to class object (of a complete
  type).  The id-expression shall name a member of  that  class,  except
  that  an  imputed  destructor can be explicitly invoked for a built-in
  type (_class.dtor_).  Therefore, if E1 has the type pointer  to  class
  X,  then  the  expression  E1->E2  is converted to the equivalent form
  (*(E1)).E2; the remainder of this  subclause  will  address  only  the
  first option (dot)6).

3 If the id-expression is a qualified-id, the  nested-name-specifier  of
  the qualified-id can specify a namespace name or a class name.  If the
  nested-name-specifier of the qualified-id specifies a namespace  name,
  the  name  is  looked  up  in the context in which the entire postfix-
  expression occurs.  If nested-name-specifier of the qualified-id spec­
  ifies  a class name, the class name is looked up as a type both in the
  class of the object expression (or the class pointed to by the pointer
  _________________________
  5) This evaluation happens even if the result is unnecessary to deter­
  mine  the  value  of the entire postfix expression, for example if the
  id-expression denotes a static member.
  6) Note that if E1 has the type pointer to class X, then (*(E1)) is an
  lvalue.

  expression)  and  the  context  in which the entire postfix-expression
  occurs.  For the purpose of this type lookup, the  name,  if  any,  of
  each  class  is  also  considered a nested class member of that class.
  These searches shall yield a single  type  which  might  be  found  in
  either  or  both  contexts.   If  the nested-name-specifier contains a
  class template-id (_temp.names_), its template-arguments are evaluated
  in the context in which the entire postfix-expression occurs.

4 Similarly,  if the id-expression is a conversion-function-id, its con­
  version-type-id shall denote the same type  in  both  the  context  in
  which  the  entire postfix-expression occurs and in the context of the
  class of the object expression (or the class pointed to by the pointer
  expression).  For the purpose of this evaluation, the name, if any, of
  each class is also considered a nested class member of that class.

5 Abbreviating object-expression.id-expression as E1.E2, then  the  type
  and  lvalue  properties  of this expression are determined as follows.
  In the remainder of this subclause, cq represents either const or  the
  absence  of  const;  vq  represents  either volatile or the absence of
  volatile.  cv represents an arbitrary set of cv-qualifiers, as defined
  in _basic.type.qualifier_.

6 If  E2  is  declared  to  have  type  reference to T, then E1.E2 is an
  lvalue; the type of E1.E2 is T.  Otherwise, one of the following rules
  applies.

  --If  E2  is a static data member, and the type of E2 is T, then E1.E2
    is an lvalue; the expression designates  the  named  member  of  the
    class.  The type of E1.E2 is T.

  --If  E2  is  a  (possibly overloaded) static member function, and the
    type of E2 is cv function of (parameter type list) returning T, then
    E1.E2  is  an  lvalue;  the  expression designates the static member
    function.  The type of E1.E2 is the same type as that of E2,  namely
    cv function of (parameter type list) returning T.

  --If  E2 is a non-static data member, and the type of E1 is cq1 vq1 X,
    and the type of E2 is cq2 vq2 T, the expression designates the named
    member  of  the object designated by the first expression.  If E1 is
    an lvalue, then E1.E2 is an lvalue.  Let the notation vq12 stand for
    the  union of vq1 and vq2 ; that is, if vq1 or vq2 is volatile, then
    vq12 is volatile.  Similarly, let the notation cq12  stand  for  the
    union  of cq1 and cq2; that is, if cq1 or cq2 is const, then cq12 is
    const.  If E2 is declared to be a mutable member, then the  type  of
    E1.E2 is vq12 T.  If E2 is not declared to be a mutable member, then
    the type of E1.E2 is cq12 vq12 T.

  --If E2 is a (possibly overloaded) non-static member function, and the
    type of E2 is cv function of (parameter type list) returning T, then
    E1.E2 is not an lvalue.  The expression designates a member function
    (of some class X).  The expression can be used only as the left-hand
    operand of a member function call (_class.mfct_).  The member  func­
    tion  shall  be  at  least as cv-qualified as the left-hand operand.

    The type of E1.E2 is class X's cv member function of (parameter type
    list) returning T.

  --If E2 is a nested type, the expression E1.E2 is ill-formed.

  --If  E2 is a member constant, and the type of E2 is T, the expression
    E1.E2 is not an lvalue.  The type of E1.E2 is T.

7 Note that class objects can be  structures  (_class.mem_)  and  unions
  (_class.union_).  Classes are discussed in _class_.

  5.2.5  Increment and decrement                        [expr.post.incr]

1 The  value  obtained by applying a postfix ++ is (a copy of) the value
  that the operand had before applying the operator.  The operand  shall
  be  a  modifiable  lvalue.  The type of the operand shall be an arith­
  metic type or a pointer to object type.  After the  result  is  noted,
  the  value  of  the  object  is modified by adding 1 to it, unless the
  object is of type bool, in which case it is set to true (this  use  is
  deprecated).   The  type  of the result is the same as the type of the
  operand, but it is not an lvalue.  See also _expr.add_ and _expr.ass_.

2 The operand of postfix -- is decremented analogously to the postfix ++
  operator, except that the operand shall not be of type bool.

  5.2.6  Dynamic cast                                [expr.dynamic.cast]

1 The result of the expression dynamic_cast<T>(v) is the result of  con­
  verting the expression v to type T.  T shall be a pointer or reference
  to a complete class type, or pointer to cv void.  Types shall  not  be
  defined  in  a dynamic_cast.  The dynamic_cast operator shall not cast
  away constness (_expr.const.cast_).

2 If T is a pointer type, v shall be an rvalue of a pointer to  complete
  class  type,  and the result is an rvalue of type T.  If T is a refer­
  ence type, v shall be an lvalue of a  complete  class  type,  and  the
  result is an lvalue of the type referred to by T.

3 If  the  type of v is the same as the required result type (which, for
  convenience, will be called R in this description), or it can be  con­
  verted  to  R  via  a  qualification  conversion  (_conv.qual_) in the
  pointer case, the result is v (converted if necessary).

4 If the value of v is a null pointer value in  the  pointer  case,  the
  result is the null pointer value of type R.

5 If  T  is pointer to cv1 B and v has type pointer to cv2 D such that B
  is a base class of D, the result is a pointer to  the  unique  B  sub-
  object  of the D object pointed to by v.  Similarly, if T is reference
  to cv1 B and v has type cv2 D such that B is a base class  of  D,  the
  result  is  an  lvalue  for  the unique7) B sub-object of the D object
  referred  to by v.  In both the pointer and reference cases, cv1 shall
  _________________________
  7) The complete object pointed or referred to by v can contain other B

  be the same cv-qualification as,  or  greater  cv-qualification  than,
  cv2,  and  B shall be an accessible nonambiguous base class of D.  For
  example,
          struct B {};
          struct D : B {};
          void foo(D* dp)
          {
              B*  bp = dynamic_cast<B*>(dp);  // equivalent to B* bp = dp;
          }

6 Otherwise, v shall be a pointer to or an lvalue of a polymorphic  type
  (_class.virtual_).

7 If  T  is pointer to cv void, then the result is a pointer to the com­
  plete object (_class.base.init_) pointed to by v.  Otherwise,  a  run-
  time check is applied to see if the object pointed or referred to by v
  can be converted to the type pointed or referred to by T.

8 The run-time check logically executes like this: If, in  the  complete
  object pointed (referred) to by v, v points (refers) to an unambiguous
  base class sub-object of a T object,  the  result  is  a  pointer  (an
  lvalue  referring)  to  that  T object.  Otherwise, if the type of the
  complete object has an unambiguous public base class of  type  T,  the
  result  is  a  pointer (reference) to the T sub-object of the complete
  object.  Otherwise, the run-time check fails.

  +-------                 BEGIN BOX 5                -------+
  Comment from Bill Gibbons: the  original  papers  allowed  all  strict
  downcasts  from  accessible  bases.  This wording does not.  The para­
  graph can be fixed by changing the first instance  of  ``an  unambigu­
  ous'' to ``a public.''
  +-------                  END BOX 5                 -------+

9 The  value  of a failed cast to pointer type is the null pointer value
  of the required result type.  A failed cast to reference  type  throws
  bad_cast (_lib.bad.cast_).  For example,
          class A { virtual void f(); };
          class B { virtual void g(); };
          class D : public virtual A, private B {};
          void g()
          {
              D   d;
              B*  bp = (B*)&d;  // cast needed to break protection
              A*  ap = &d;      // public derivation, no cast needed
              D&  dr = dynamic_cast<D&>(*bp);  // succeeds
              ap = dynamic_cast<A*>(bp);       // succeeds
              bp = dynamic_cast<B*>(ap);       // fails
              ap = dynamic_cast<A*>(&dr);      // succeeds
              bp = dynamic_cast<B*>(&dr);      // fails
          }
  _________________________
  objects as base classes, but these are ignored.

          class E : public D , public B {};
          class F : public E, public D {}
          void h()
          {
              F   f;
              A*  ap = &f;  // okay: finds unique A
              D*  dp = dynamic_cast<D*>(ap);  // fails: ambiguous
              E*  ep = (E*)ap;  // error: cast from virtual base
              E*  ep = dynamic_cast<E*>(ap);  // succeeds
          }
  _class.cdtor_  describes  the behavior of a dynamic_cast applied to an
  object under construction or destruction.

  5.2.7  Type identification                               [expr.typeid]

1 The result of a typeid expression is of  type  const type_info&.   The
  value is a reference to a type_info object (_lib.type.info_) that rep­
  resents the type-id or the type of the expression respectively.

2 If  the  expression   is   a   reference   to   a   polymorphic   type
  (_class.virtual_),    the    type_info   for   the   complete   object
  (_class.base.init_) referred to is the result.

3 If the expression is the result of applying unary * to a pointer to  a
  polymorphic type,8) then the pointer shall either be zero or point  to
  a  valid object.  If the pointer is zero, the typeid expression throws
  the bad_typeid exception (_lib.bad.typeid_).  Otherwise, the result of
  the  typeid  expression  is  the value that represents the type of the
  complete object to which the pointer points.

4 If the  expression  is  the  result  of  subscripting  (_expr.sub_)  a
  pointer,  say  p, that points to a polymorphic type,9) then the result
  of the typeid expression is that of typeid(*p).  The subscript is  not
  evaluated.

5 If  the  expression is neither a pointer nor a reference to a polymor­
  phic type, the result is the type_info representing the (static)  type
  of the expression.  The expression is not evaluated.

6 In  all  cases  typeid  ignores  the  top-level  cv-qualifiers  of its
  operand's type. For example:
          class D { ... };
          D d1;
          const D d2;
          typeid(d1) == typeid(d2);      // yields true
          typeid(D)  == typeid(const D); // yields true
          typeid(D)  == typeid(d2);      // yields true
  _________________________
  8)  If  p is a pointer, then *p, (*p), ((*p)), and so on all meet this
  requirement.
  9)  If p is a pointer to a polymorphic type and i has integral or enu­
  merated type,  then  p[i],  (p[i]),  (p)[i],  ((((p))[((i))])),  i[p],
  (i[p]), and so on all meet this requirement.

  _class.cdtor_ describes the behavior of typeid applied  to  an  object
  under construction or destrcution.

  5.2.8  Static cast                                  [expr.static.cast]

1 The  result  of the expression static_cast<T>(v) is the result of con­
  verting the expression v to type T.  If T is  a  reference  type,  the
  result  is an lvalue; otherwise, the result is an rvalue.  Types shall
  not be defined in a static_cast.  The static_cast operator  shall  not
  cast away constness.  See _expr.const.cast_.

2 Any  implicit  conversion (including standard conversions and/or user-
  defined conversions; see _conv_ and _over.best.ics_) can be  performed
  explicitly  using  static_cast.  More precisely, if T t(v); is a well-
  formed declaration, for some invented temporary variable t,  then  the
  result  of  static_cast<T>(v) is defined to be the temporary t, and is
  an lvalue if T is a reference type,  and  an  rvalue  otherwise.   The
  expression v shall be an lvalue if the equivalent declaration requires
  an lvalue for v.

3 If the static_cast does not correspond to an  implicit  conversion  by
  the  above  definition, it shall perform one of the conversions listed
  below.  No other  conversion  can  be  performed  explicitly  using  a
  static_cast.

4 Any  expression  can  be  explicitly  converted  to type cv void.  The
  expression value is discarded.

5 An lvalue expression of type T1 can be cast to the type  reference  to
  T2  if an expression of type pointer to T1 can be explicitly converted
  to the type pointer to T2 using a static_cast.  That is,  a  reference
  cast   static_cast<T&>x   has   the  same  effect  as  the  conversion
  *static_cast<T*>&x with the built-in & and * operators.  The result is
  an   lvalue.   This  interpretation  is  used  only  if  the  original
  static_cast is not well-formed as an  implicit  conversion  under  the
  rules given above.  This form of reference cast creates an lvalue that
  refers to the same object as the source lvalue, but with  a  different
  type.   Consequently,  it  does  not  create  a  temporary or copy the
  object,  and  constructors  (_class.ctor_)  or  conversion   functions
  (_class.conv_) are not called.  For example,
          struct B {};
          struct D : public B {};
          D d;
          // creating a temporary for the B sub-object not allowed
          ... (const B&) d ...

6 The  inverse  of  any  standard  conversion  (_conv_) can be performed
  explicitly using static_cast  subject  to  the  restriction  that  the
  explicit  conversion does not cast away constness (_expr.const.cast_),
  and the following additional rules for specific cases:

7 A value of integral type can be explicitly converted to an enumeration
  type.   The  value  is  unchanged  if the integral value is within the

  range of the enumeration values (_dcl.enum_). Otherwise, the resulting
  enumeration value is unspecified.

8 An  rvalue  of  type pointer to cv1 B, where B is a class type, can be
  converted to an rvalue of type pointer to cv2 D, where D  is  a  class
  derived  (_class.derived_) from B, if a valid standard conversion from
  pointer to cv2 D to pointer to cv2 B exists (_conv.ptr_), cv2  is  the
  same cv-qualification as, or greater cv-qualification than, cv1, and B
  is not a virtual base class of D.  The null pointer value (_conv.ptr_)
  is  converted  to  the null pointer value of the destination type.  If
  the rvalue of type pointer to cv1 B points to a B that is  actually  a
  sub-object of an object of type D, the resulting pointer points to the
  enclosing object of type D.  Otherwise, the  result  of  the  cast  is
  undefined.

9 An  rvalue  of  type  pointer to member of D of type cv1 T can be con­
  verted to an rvalue of type pointer to member of  B  of  type  cv2  T,
  where  B  is  a base class (_class.derived_) of D, if a valid standard
  conversion from pointer to member of B of type cv2  T  to  pointer  to
  member of D of type cv2 T exists (_conv.mem_), and cv2 is the same cv-
  qualification as, or greater cv-qualification  than,  cv1.   The  null
  member  pointer  value  (_conv.mem_)  is  converted to the null member
  pointer value of the destination type.  If class B contains or  inher­
  its the original member, the resulting pointer to member points to the
  member in class B.  Otherwise, the result of the cast is undefined.

  5.2.9  Reinterpret cast                        [expr.reinterpret.cast]

1 The result of the expression reinterpret_cast<T>(v) is the  result  of
  converting  the expression v to type T.  If T is a reference type, the
  result is an lvalue; otherwise, the result is an rvalue.  Types  shall
  not  be  defined  in a reinterpret_cast.  Conversions that can be per­
  formed explicitly using reinterpret_cast are listed below.   No  other
  conversion can be performed explicitly using reinterpret_cast.

2 The  reinterpret_cast  operator  shall  not  cast  away constness; see
  _expr.const.cast_.

3 The mapping performed by reinterpret_cast  is  implementation-defined;
  it  might,  or  might not, produce a representation different from the
  original value.

4 A pointer can be explicitly  converted  to  any  integral  type  large
  enough  to  hold  it.  The mapping function is implementation-defined,
  but is intended to be unsurprising to those who  know  the  addressing
  structure of the underlying machine.

5 A  value of integral type can be explicitly converted to a pointer.  A
  pointer converted to an integer of sufficient size (if any such exists
  on the implementation) and back to the same pointer type will have its
  original value; mappings between pointers and integers  are  otherwise
  implementation-defined.

6 The  operand  of  a  pointer  cast can be an rvalue of type pointer to
  incomplete class type.  The destination type of a pointer cast can  be
  pointer  to  incomplete  class  type.   In such cases, if there is any
  inheritance relationship between the source and  destination  classes,
  the behavior is undefined.

7 A  pointer to a function can be explicitly converted to a pointer to a
  function of a different  type.   The  effect  of  calling  a  function
  through  a  pointer to a function type that differs from the type used
  in the definition of the function is undefined.  See also  _conv.ptr_.

8 A  pointer to an object can be explicitly converted to a pointer to an
  object of different type.  In general, the results of this are unspec­
  ified;  except  that converting an rvalue of type pointer to T1 to the
  type pointer to T2 (where T1 and T2 are object  types  and  where  the
  alignment  requirements  of  T2  are no stricter than those of T1) and
  back to its original type yields the original pointer value.

  +-------                 BEGIN BOX 6                -------+
  This does not allow conversion of function pointers to other  function
  pointer types and back.  Should it?
  +-------                  END BOX 6                 -------+

9 The  null  pointer value (_conv.ptr_) is converted to the null pointer
  value of the destination type.

10An rvalue of type pointer to member of X of type T1, can be explicitly
  converted  to  an rvalue of type pointer to member of Y of type T2, if
  T1 and T2 are both member function types or both  data  member  types.
  The  null  member  pointer value (_conv.mem_) is converted to the null
  member pointer value of the destination type.  In general, the  result
  of this conversion is unspecified, except that:

  --converting an rvalue of type pointer to member function to a differ­
    ent pointer to member function type and back to  its  original  type
    yields the original pointer to member value.

  --converting  an rvalue of type pointer to data member of X of type T1
    to the type pointer to data member of Y of type T2 (where the align­
    ment  requirements  of T2 are no stricter than those of T1) and back
    to its original type yields the original pointer to member value.

11Calling a member function through a pointer to member that  represents
  a  function  type that differs from the function type specified on the
  member function declaration results in undefined behavior.

12An lvalue expression of type T1 can be cast to the type  reference  to
  T2  if an expression of type pointer to T1 can be explicitly converted
  to the type pointer to T2 using a reinterpret_cast.  That is, a refer­
  ence  cast reinterpret_cast<T&>x has the same effect as the conversion
  *reinterpret_cast<T*>&x with the built-in  &  and  *  operators.   The

  result  is  an  lvalue  that  refers  to the same object as the source
  lvalue, but with a different type.  No temporary is created,  no  copy
  is  made,  and  constructors  (_class.ctor_)  or  conversion functions
  (_class.conv_) are not called.

  5.2.10  Const cast                                   [expr.const.cast]

1
  +-------                 BEGIN BOX 7                -------+
  Editorial change  from  previous  edition:  it  is  permitted  to  use
  const_cast as a no-op.
  +-------                  END BOX 7                 -------+

  The  result  of  the  expression const_cast<T>(v) is of type T.  Types
  shall not be defined in a const_cast.  Conversions that  can  be  per­
  formed explicitly using const_cast are listed below.  No other conver­
  sion shall be performed explicitly using const_cast.

2 An rvalue of type pointer to cv1 T can be explicitly converted to  the
  type  pointer  to  cv2 T, where T is any object type and where cv1 and
  cv2 are cv-qualifications , using the  cast  const_cast<cv2  T*>.   An
  lvalue  of type cv1 T can be explicitly converted to an lvalue of type
  cv2 T, where T is any object type  and  where  cv1  and  cv2  are  cv-
  qualifications,  using  the  cast const_cast<cv2 T&>.  The result of a
  pointer or reference const_cast refers to the original object.

3 An rvalue of type pointer to member of X of type cv1 T can be  explic­
  itly converted to the type pointer to member of X of type cv2 T, where
  T is a data member type and where cv1 and cv2 are cv-qualifiers, using
  the  cast  const_cast<cv2 T X::*>.   The result of a pointer to member
  const_cast will refer to the same  member  as  the  original  (uncast)
  pointer to data member.

4 The  following rules define casting away constness.  In these rules Tn
  and Xn represent types.  For two pointer types:
            Kismin(N,M)
  casting from X1 to X2 casts away constness if, for a non-pointer  type
  T (e.g., int), there does not exist an implicit conversion from:

            Tcv1,(N-K+1)*cv1,(N-K+2)*...cv1,N*
  to

            Tcv2,(N-K+1)*cv2,(M-K+2)*...cv2,M*

5 Casting from an lvalue of type T1 to an lvalue of type T2 using a ref­
  erence cast casts away constness if a cast  from  an  rvalue  of  type
  pointer to T1 to the type pointer to T2 casts away constness.

6 Casting from an rvalue of type "pointer to data member of X of type T1
  to the type pointer to data member of Y of type T2 casts  away  const­
  ness  if  a  cast  from  an  rvalue  of type pointer to T1 to the type
  pointer to T2 casts away constness.

7 Note that these rules are not intended to  protect  constness  in  all
  cases.   For  instance,  conversions between pointers to functions are
  not covered because such conversions lead to values whose  use  causes
  undefined  behavior.  For the same reasons, conversions between point­
  ers to member functions, and in  particular,  the  conversion  from  a
  pointer  to a const member function to a pointer to a non-const member
  function, are not covered.  For multi-level pointers to data  members,
  or  multi-level mixed object and member pointers, the same rules apply
  as for multi-level object pointers.  That is, the member of  attribute
  is  ignored  for  purposes  of determining whether const has been cast
  away.

8 Depending on the type of the object, a  write  operation  through  the
  pointer,  lvalue or pointer to data member resulting from a const_cast
  that   casts   away   constness   may   produce   undefined   behavior
  (_dcl.type.cv_).

  +-------                      BEGIN BOX 8                     -------+
  This  will  need to be reworked once the memory model and object model
  are ironed out.
  +-------                       END BOX 8                      -------+

9 A null pointer value (_conv.ptr_) is converted  to  the  null  pointer
  value  of  the  destination  type.   The  null  member  pointer  value
  (_conv.mem_) is converted to the null member pointer value of the des­
  tination type.

  5.3  Unary expressions                                    [expr.unary]

1 Expressions with unary operators group right-to-left.
          unary-expression:
                  postfix-expression
                  ++  unary-expression
                  --  unary-expression
                  unary-operator cast-expression
                  sizeof unary-expression
                  sizeof ( type-id )
                  new-expression
                  delete-expression
          unary-operator: one of
                  *  &  +  -  !  ~

  5.3.1  Unary operators                                 [expr.unary.op]

1 The  unary  *  operator  means  indirection: the expression shall be a
  pointer, and the result is an lvalue referring to the object to  which
  the expression points.  If the type of the expression is pointer to T,
  the type of the result is T.

2 The result of the unary & operator is a pointer to its  operand.   The
  operand  shall  be an lvalue or a qualified-id.  In the first case, if
  the type of the expression is T, the type of the result is pointer  to
  T.  In particular, the address of an object of type cv T is pointer to

  cv T, with the same cv-qualifiers.  For example,  the  address  of  an
  object  of type const int has type pointer to const int.  For a quali­
  fied-id, if the member is a nonstatic member of class C of type T, the
  type  of  the  result  is pointer to member of class C of type T.  For
  example:
          struct A { int i; };
          struct B : A { };
          ... &B::i ... // has type "int A::*"
  For a static member of type T, the type is plain pointer to  T.   Note
  that a pointer to member is only formed when an explicit & is used and
  its operand is a qualified-id not enclosed in parentheses.  For  exam­
  ple,   the  expression  &(qualified-id),  where  the  qualified-id  is
  enclosed in parentheses, does not form an expression of  type  pointer
  to  member.   Neither does qualified-id, and there is no implicit con­
  version from the type nonstatic member function to the type pointer to
  member  function,  as  there is from an lvalue of function type to the
  type pointer to function  (_conv.func_).   Nor  is  &unqualified-id  a
  pointer to member, even within the scope of unqualified-id's class.

  +-------                      BEGIN BOX 9                     -------+
  This  section  probably needs to take into account const and its rela­
  tionship to mutable.
  +-------                       END BOX 9                      -------+

3 The address of an object of incomplete type can be taken, but only  if
  the complete type of that object does not have the address-of operator
  (operator&()) overloaded; no diagnostic is required.

4 The address of an overloaded function (_over_) can be taken only in  a
  context that uniquely determines which version of the overloaded func­
  tion is referred to (see _over.over_).  Note that  since  the  context
  might  determine  whether  the operand is a static or nonstatic member
  function, the context can also affect whether the expression has  type
  pointer to function or pointer to member function.

5 The operand of the unary + operator shal have arithmetic, enumeration,
  or pointer type and the result is the value of the argument.  Integral
  promotion  is performed on integral or enumeration operands.  The type
  of the result is the type of the promoted operand.

6 The operand of the unary - operator shall have arithmetic or  enumera­
  tion  type  and  the  result is the negation of its operand.  Integral
  promotion is performed on integral or enumeration operands.  The nega­
  tive of an unsigned quantity is computed by subtracting its value from
  2n, where n is the number of bits in the promoted operand.   The  type
  of the result is the type of the promoted operand.

7 The  operand  of the logical negation operator !  is converted to bool
  (_conv.bool_); its value is true if the converted operand is false and
  false otherwise.  The type of the result is bool.

8 The  operand  of ~ shall have integral or enumeration type; the result
  is the one's complement of its operand.  Integral promotions are  per­
  formed.  The type of the result is the type of the promoted operand.

  5.3.2  Increment and decrement                         [expr.pre.incr]

1 The operand of prefix ++ is modified by adding 1, or set to true if it
  is bool (this use is deprecated).  The operand shall be  a  modifiable
  lvalue.   The  type  of  the  operand shall be an arithmetic type or a
  pointer to a completely-defined object type.  The  value  is  the  new
  value  of the operand; it is an lvalue.  If x is not of type bool, the
  expression ++x is equivalent to x+=1.  See the discussions of addition
  (_expr.add_)  and assignment operators (_expr.ass_) for information on
  conversions.

2 The operand of prefix -- is decremented analogously to the  prefix  ++
  operator, except that the operand shall not be of type bool.

  5.3.3  Sizeof                                            [expr.sizeof]

1 The  sizeof  operator  yields the size, in bytes, of its operand.  The
  operand is either an expression, which is not evaluated, or  a  paren­
  thesized  type-id.   The  sizeof  operator  shall not be applied to an
  expression that has function or incomplete type, or to  the  parenthe­
  sized  name  of  such  a  type, or to an lvalue that designates a bit-
  field.  A byte is unspecified by the language except in terms  of  the
  value   of   sizeof;   sizeof(char)   is   1,   but  sizeof(bool)  and
  sizeof(wchar_t) are implementation-defined.  10)

2 When applied to a reference, the result is the size of the  referenced
  object.  When applied to a class, the result is the number of bytes in
  an object of that class including any  padding  required  for  placing
  such  objects  in  an array.  The size of any class or class object is
  greater than zero.  When applied to an array, the result is the  total
  number  of bytes in the array.  This implies that the size of an array
  of n elements is n times the size of an element.

3 The sizeof operator can be applied to a pointer  to  a  function,  but
  shall not be applied directly to a function.

4 The  lvalue-to-rvalue  (_conv.lval_), array-to-pointer (_conv.array_),
  and function-to-pointer (_conv.func_) standard  conversions  are  sup­
  pressed on the operand of sizeof.

5 Types shall not be defined in a sizeof expression.

6 The  result  is a constant of type size_t, an implementation-dependent
  unsigned  integral  type  defined  in  the  standard   header   <cstd­
  def>(_lib.support.types_).

  _________________________
  10) sizeof(bool) is not required to be 1.

  5.3.4  New                                                  [expr.new]

1 The  new-expression  attempts  to  create  an  object  of  the type-id
  (_dcl.name_) to which it is applied.  This type shall  be  a  complete
  object or array type (_intro.memory_, _basic.types_).
          new-expression:
                  ::opt new new-placementopt new-type-id new-initializeropt
                  ::opt new new-placementopt ( type-id ) new-initializeropt
          new-placement:
                  ( expression-list )
          new-type-id:
                  type-specifier-seq new-declaratoropt
          new-declarator:
                  * cv-qualifier-seqopt new-declaratoropt
                   ::opt nested-name-specifier * cv-qualifier-seqopt new-declaratoropt
                  direct-new-declarator
          direct-new-declarator:
                  [ expression ]
                  direct-new-declarator [ constant-expression ]
          new-initializer:
                  ( expression-listopt )
  Entities  created  by  a  new-expression have dynamic storage duration
  (_basic.stc.dynamic_).  That is, the lifetime of such an entity is not
  restricted  to  the scope in which it is created.  If the entity is an
  object, the new-expression returns a pointer to  the  object  created.
  If it is an array, the new-expression returns a pointer to the initial
  element of the array.

2 The new-type in a new-expression is the longest possible  sequence  of
  new-declarators.   This prevents ambiguities between declarator opera­
  tors &, *, [], and their expression counterparts.  For example,
          new int*i;     // syntax error: parsed as `(new int*) i'
                         //               not as `(new int)*i'
  The * is the pointer declarator and not the multiplication operator.

3 Parentheses shall not appear in a new-type-id used as the operand  for
  new.  For example,

4         new int(*[10])();       // error
  is ill-formed because the binding is
          (new int) (*[10])();    // error
  The  explicitly  parenthesized version of the new operator can be used
  to create objects of compound types (_basic.compound_).  For example,
          new (int (*[10])());
  allocates an array of 10 pointers to functions (taking no argument and
  returning int).

5 The  type-specifier-seq  shall not contain class declarations, or enu­
  meration declarations.

6 When the allocated object  is  an  array  (that  is,  the  direct-new-
  declarator  syntax  is  used  or the new-type-id or type-id denotes an
  array type), the new-expression yields a pointer to the  initial  ele­
  ment (if any) of the array.  Thus, both new int and new int[10] return

  an int* and the type of new int[i][10] is int (*)[10].

7 Every constant-expression in a direct-new-declarator shall be an inte­
  gral  constant  expression  (_expr.const_)  with  a  strictly positive
  value.  The expression in a direct-new-declarator shall be of integral
  type (_basic.fundamental_) with a non-negative value.  For example, if
  n is a variable of  type  int,  then  new float[n][5]  is  well-formed
  (because   n  is  the  expression  of  a  direct-new-declarator),  but
  new float[5][n]  is  ill-formed  (because  n  is   not   a   constant-
  expression).  If n is negative, the effect of new float[n][5] is unde­
  fined.

8 When the value of the expression in a direct-new-declarator  is  zero,
  an  array  with no elements is allocated.  The pointer returned by the
  new-expression will be non-null and distinct from the pointer  to  any
  other object.

9 Storage  for  the  object created by a new-expression is obtained from
  the appropriate allocation function  (_basic.stc.dynamic.allocation_).
  When  the  allocation  function  is called, the first argument will be
  amount of space requested (which might be larger than the size of  the
  object being created only if that object is an array).

10An  implementation  provides default definitions of the global alloca­
  tion functions operator new() for non-arrays (_lib.new.delete.single_)
  and  operator new[]() for arrays (_lib.new.delete.array_).  A C++ pro­
  gram  can  provide  alternative   definitions   of   these   functions
  (_lib.replacement.functions_),    and/or    class-specific    versions
  (_class.free_).

11The new-placement syntax can be used to supply additional arguments to
  an  allocation function.  Overloading resolution is done by assembling
  an argument list from the amount of space requested (the  first  argu­
  ment)  and  the  expressions  in  the  new-placement  part of the new-
  expression, if used (the second and succeeding arguments).

12For example:

  --new T results in a call of operator new(sizeof(T)),

  --new(2,f) T results in a call of operator new(sizeof(T),2,f),

  --new T[5] results in a call of operator new[](sizeof(T)*5+x), and

  --new(2,f) T[5]       results        in        a        call        of
    operator new[](sizeof(T)*5+y,2,f).   Here, x and y are non-negative,
    implementation-defined values representing  array  allocation  over­
    head.  They might vary from one use of new to another.

13The  return  value  from the allocation function, if non-null, will be
  assumed to point to a block of appropriately aligned available storage
  of  the  requested  size, and the object will be created in that block
  (but not necessarily at the beginning of the block, if the  object  is
  an array).

14A  new-expression  for  a  class  calls  one of the class constructors
  (_class.ctor_) to initialize i the object.  An object of a  class  can
  be  created  by  new  only  if suitable arguments are provided for the
  class' constructors by the new-initializer, or  if  the  class  has  a
  default constructor.11) If no user-declared constructor is used and  a
  new-initializer  is provided, the new-initializer shall be of the form
  (expression) or (); if the expression is present, it shall be of class
  type and is used to initialize the object.

15No  initializers  can be specified for arrays.  Arrays of objects of a
  class can be created by a new-expression  only  if  the  class  has  a
  default  constructor.12) In that case, the default constructor will be
  called for each element of the array, in order of increasing  address.

16Access and ambiguity control are done for both the allocation function
  and the constructor (_class.ctor_, _class.free_).

17The allocation function can indicate failure by throwing  a  bad_alloc
  exception (_except_, _lib.bad.alloc_).  In this case no initialization
  is done.

18If the constructor throws an exception and the new-expression does not
  contain    a    new-placement,    then   the   deallocation   function
  (_basic.stc.dynamic.deallocation_, _class.free_) is used to  free  the
  memory  in  which  the  object  was being constructed, after which the
  exception continues to propagate in the context of the new-expression.

19The  way the object was allocated determines how it is freed: if it is
  allocated by ::new, then it is freed by ::delete,  and  if  it  is  an
  array, it is freed by delete[] or ::delete[] as appropriate.

  +-------                     BEGIN BOX 10                     -------+
  This  is  a  correction to San Diego resolution 3.5, which on its face
  seems to require that whether  to  use  delete  or  delete[]  must  be
  decided purely on syntactic grounds.  I believe the intent of the com­
  mittee was to make the form of delete correspond to the  form  of  the
  corresponding new.
  +-------                      END BOX 10                      -------+

20Whether  the  allocation function is called before evaluating the con­
  structor arguments, after evaluating  the  constructor  arguments  but
  before  entering  the  constructor,  or  by  the constructor itself is
  unspecified.  It is also unspecified whether the arguments to  a  con­
  structor  are  evaluated  if  the allocation function returns the null
  pointer or throws an exception.

  _________________________
  11)  This  means  that  struct  s{};  s* ps = new s; is allowed on the
  grounds that class s has an implicitly-declared default constructor.
  12) PODS structs have an implicitly-declared default constructor.

  5.3.5  Delete                                            [expr.delete]

1 The   delete-expression   operator   destroys   a   complete    object
  (_intro.memory_) or array created by a new-expression.
          delete-expression:
                  ::opt delete cast-expression
                  ::opt delete [ ] cast-expression
  The  first alternative is for non-array objects, and the second is for
  arrays.  The result has type void.

2 In either alternative, if the value of the operand of  delete  is  the
  null  pointer  the  operation  has no effect.  Otherwise, in the first
  alternative (delete object), the value of the operand of delete  shall
  be a pointer to a non-array object created by a new-expression without
  a  new-placement  specification,  or  a  pointer   to   a   sub-object
  (_intro.memory_)   representing   a  base  class  of  such  an  object
  (_class.derived_).

  +-------                     BEGIN BOX 11                     -------+
  Issue: ... or a class with an unambiguous conversion to such a pointer
  type ...
  +-------                      END BOX 11                      -------+

  In  the second alternative (delete array), the value of the operand of
  delete shall be a pointer to an  array  created  by  a  new-expression
  without a new-placement specification.

3 In  the  first  alternative (delete object), if the static type of the
  operand is different from its dynamic type, the static type shall have
  a virtual destructor or the result is undefined.  In the second alter­
  native (delete array) if the dynamic type of the object to be  deleted
  is a class that has a destructor and its static type is different from
  its dynamic type, the result is undefined.

  +-------                     BEGIN BOX 12                     -------+
  This should probably be tightened  to  require  that  the  static  and
  dynamic types match, period.
  +-------                      END BOX 12                      -------+

4 The  deletion  of an object might change its value.  If the expression
  denoting the object in a delete-expression is a modifiable lvalue, any
  attempt   to   access  its  value  after  the  deletion  is  undefined
  (_basic.stc.dynamic.deallocation_).

5 If the class of the object being deleted is incomplete at the point of
  deletion and the class has a destructor or an allocation function or a
  deallocation function, the result is undefined.

6 The delete-expression will invoke the  destructor  (if  any)  for  the
  object  or the elements of the array being deleted.  In the case of an
  array, the elements will be destroyed in order of  decreasing  address
  (that is, in reverse order of construction).

7 To  free  the  storage  pointed  to, the delete-expression will call a
  deallocation function (_basic.stc.dynamic.deallocation_).

8 An implementation provides default definitions of the global dealloca­
  tion       functions       operator delete()       for      non-arrays
  (_lib.new.delete.single_)   and   operator delete[]()    for    arrays
  (_lib.new.delete.array_).  A C++ program can provide alternative defi­
  nitions  of  these  functions  (_lib.replacement.functions_),   and/or
  class-specific versions (_class.free_).

9 Access  and ambiguity control are done for both the deallocation func­
  tion and the destructor (_class.dtor_, _class.free_).

  5.4  Explicit type conversion (cast notation)              [expr.cast]

1 The result of the expression (T) cast-expression is  of  type  T.   An
  explicit  type  conversion  can be expressed using functional notation
  (_expr.type.conv_),  a   type   conversion   operator   (dynamic_cast,
  static_cast, reinterpret_cast, const_cast), or the cast notation.
          cast-expression:
                  unary-expression
                  ( type-id ) cast-expression

2 Types shall not be defined in casts.

3 Any  type conversion not mentioned below and not explicitly defined by
  the user (_class.conv_) is ill-formed.

4 The conversions performed by static_cast  (_expr.static.cast_),  rein­
  terpret_cast           (_expr.reinterpret.cast_),           const_cast
  (_expr.const.cast_), or any sequence thereof, can be  performed  using
  the  cast  notation  of  explicit  type conversion.  The same semantic
  restrictions and behaviors apply.  If a given conversion can  be  per­
  formed  using  either static_cast or reinterpret_cast, the static_cast
  interpretation is used.

5 In addition to those conversions, a pointer to an object of a  derived
  class  (_class.derived_)  can  be explicitly converted to a pointer to
  any of its  base  classes  regardless  of  accessibility  restrictions
  (_class.access.base_),   provided   the   conversion   is  unambiguous
  (_class.member.lookup_).  The resulting pointer will refer to the con­
  tained object of the base class.

  5.5  Pointer-to-member operators                      [expr.mptr.oper]

1 The pointer-to-member operators ->* and .*  group left-to-right.
          pm-expression:
                  cast-expression
                  pm-expression .* cast-expression
                  pm-expression ->* cast-expression

2 The  binary  operator  .*  binds its second operand, which shall be of
  type pointer to member of T to its first operand, which  shall  be  of
  class T or of a class of which T is an unambiguous and accessible base

  class.  The result is an object or a function of the type specified by
  the second operand.

3 The  binary  operator  ->* binds its second operand, which shall be of
  type pointer to member of T to its first operand, which  shall  be  of
  type  pointer  to T or pointer to a class of which T is an unambiguous
  and accessible base class.  The result is an object or a  function  of
  the type specified by the second operand.

4 If  the  result  of  .*  or ->* is a function, then that result can be
  used only as the operand for the function call operator ().  For exam­
  ple,
          (ptr_to_obj->*ptr_to_mfct)(10);
  calls  the  member  function  denoted  by  ptr_to_mfct  for the object
  pointed to by ptr_to_obj.  The result of a .*  expression is an lvalue
  only  if  its  first  operand is an lvalue and its second operand is a
  pointer to data member.  The result of an ->* expression is an  lvalue
  only if its second operand is a pointer to data member.  If the second
  operand is the null pointer to member value (_conv.mem_),  the  result
  is undefined.

  5.6  Multiplicative operators                               [expr.mul]

1 The multiplicative operators *, /, and % group left-to-right.
          multiplicative-expression:
                  pm-expression
                  multiplicative-expression * pm-expression
                  multiplicative-expression / pm-expression
                  multiplicative-expression % pm-expression

2 The  operands of * and / shall have arithmetic type; the operands of %
  shall have integral type.  The usual arithmetic conversions  are  per­
  formed on the operands and determine the type of the result.

3 The binary * operator indicates multiplication.

4 The  binary  / operator yields the quotient, and the binary % operator
  yields the remainder from the division of the first expression by  the
  second.   If  the second operand of / or % is zero the result is unde­
  fined; otherwise (a/b)*b + a%b is equal to a.  If  both  operands  are
  nonnegative then the remainder is nonnegative; if not, the sign of the
  remainder is implementation dependent.

  5.7  Additive operators                                     [expr.add]

1 The additive operators + and - group left-to-right.  The usual  arith­
  metic conversions are performed for operands of arithmetic type.
          additive-expression:
                  multiplicative-expression
                  additive-expression + multiplicative-expression
                  additive-expression - multiplicative-expression
  For  addition, either both operands shall have arithmetic type, or one
  operand shall be a pointer to a completely defined object type and the

  other shall have integral type.

2 For subtraction, one of the following shall hold:

  --both operands have arithmetic type;

  --both  operands  are pointers to qualified or unqualified versions of
    the same completely defined object type; or

  --the left operand is a pointer to a completely  defined  object  type
    and the right operand has integral type.

3 If  both  operands  have arithmetic type, the usual arithmetic conver­
  sions are performed on them.  The result of the binary +  operator  is
  the  sum  of the operands.  The result of the binary - operator is the
  difference resulting from the subtraction of the second  operand  from
  the first.

4 For  the  purposes  of these operators, a pointer to a nonarray object
  behaves the same as a pointer to the first  element  of  an  array  of
  length one with the type of the object as its element type.

5 When  an  expression  that has integral type is added to or subtracted
  from a pointer, the result has the type of the  pointer  operand.   If
  the  pointer  operand points to an element of an array object, and the
  array is large enough, the result points to an element offset from the
  original  element  such  that  the difference of the subscripts of the
  resulting and original array elements equals the integral  expression.
  In  other  words, if the expression P points to the i-th element of an
  array object, the expressions (P)+N (equivalently,  N+(P))  and  (P)-N
  (where  N has the value n) point to, respectively, the i+n-th and i-n-
  th elements of the array object, provided they  exist.   Moreover,  if
  the  expression  P  points to the last element of an array object, the
  expression (P)+1 points one past the last element of the array object,
  and  if  the expression Q points one past the last element of an array
  object, the expression (Q)-1 points to the last element of  the  array
  object.   If both the pointer operand and the result point to elements
  of the same array object, or one past the last element  of  the  array
  object,  the  evaluation shall not produce an overflow; otherwise, the
  behavior is undefined.  If the result is used as  an  operand  of  the
  unary  *  operator,  the behavior is undefined unless both the pointer
  operand and the result point to elements of the same array object,  or
  the  pointer  operand  points  one  past  the last element of an array
  object and the result points to an element of the same array object.

6 When two pointers to elements of the same array object are subtracted,
  the  result  is the difference of the subscripts of the two array ele­
  ments.  The type of the result  is  an  implementation-defined  signed
  integral  type;  this  type  shall be the same type that is defined as
  ptrdiff_t in the <cstddef> header (_lib.support.types_).  As with  any
  other  arithmetic  overflow,  if  the result does not fit in the space
  provided, the behavior is undefined.  In other words, if  the  expres­
  sions P and Q point to, respectively, the i-th and j-th elements of an

  array object, the expression (P)-(Q) has the value  i-j  provided  the
  value  fits  in an object of type ptrdiff_t.  Moreover, if the expres­
  sion P points either to an element of an array object or one past  the
  last  element  of  an array object, and the expression Q points to the
  last element of the same array object, the expression ((Q)+1)-(P)  has
  the same value as ((Q)-(P))+1 and as -((P)-((Q)+1)), and has the value
  zero if the expression P points one past the last element of the array
  object,  even though the expression (Q)+1 does not point to an element
  of the array object.  Unless both pointers point to  elements  of  the
  same  array  object, or one past the last element of the array object,
  the behavior is undefined.13)

  5.8  Shift operators                                      [expr.shift]

1 The shift operators << and >> group left-to-right.
          shift-expression:
                  additive-expression
                  shift-expression << additive-expression
                  shift-expression >> additive-expression
  The operands shall be of integral type  and  integral  promotions  are
  performed.   The  type  of  the  result  is  that of the promoted left
  operand.  The result is undefined if the right operand is negative, or
  greater  than  or  equal  to  the  length in bits of the promoted left
  operand.  The value of E1 << E2 is E1 (interpreted as a  bit  pattern)
  left-shifted  E2  bits; vacated bits are zero-filled.  The value of E1
  >> E2 is E1 right-shifted E2 bit positions.  The right shift is  guar­
  anteed  to  be logical (zero-fill) if E1 has an unsigned type or if it
  has a nonnegative value; otherwise the result is implementation depen­
  dent.

  5.9  Relational operators                                   [expr.rel]

1 The  relational  operators  group  left-to-right, but this fact is not
  very useful; a<b<c means (a<b)<c and not (a<b)&&(b<c).

  _________________________
  13) Another way to approach pointer arithmetic is first to convert the
  pointer(s)  to  character  pointer(s): In this scheme the integral ex­
  pression added to or subtracted from the converted  pointer  is  first
  multiplied  by  the  size of the object originally pointed to, and the
  resulting pointer is converted back to the original type.  For pointer
  subtraction, the result of the difference between the character point­
  ers is similarly divided by the size of the object originally  pointed
  to.

7 When viewed in this way, an implementation need only provide one extra
  byte (which might overlap another object in the  program)  just  after
  the  end  of the object in order to satisfy the one past the last ele­
  ment requirements.

          relational-expression:
                  shift-expression
                  relational-expression < shift-expression
                  relational-expression > shift-expression
                  relational-expression <= shift-expression
                  relational-expression >= shift-expression
  The operands shall have arithmetic or pointer type.  The  operators  <
  (less  than),  >  (greater  than),  <= (less than or equal to), and >=
  (greater than or equal to) all yield false or true.  The type  of  the
  result is bool.

2 The usual arithmetic conversions are performed on arithmetic operands.
  Pointer conversions are performed on pointer operands to bring them to
  the  same  type,  which shall be a qualified or unqualified version of
  the type of one of the operands.  This implies that any pointer can be
  compared to an integral constant expression evaluating to zero and any
  pointer can be compared to a pointer of qualified or unqualified  type
  void*  (in  the  latter case the pointer is first converted to void*).
  Pointers to objects or functions of the same type (after pointer  con­
  versions)  can  be  compared; the result depends on the relative posi­
  tions of the pointed-to objects or functions in the address space.

3 If two pointers of the same type point to the same object or function,
  or  both  point  one past the end of the same array, or are both null,
  they compare equal.  If two pointers of the same type point to differ­
  ent  objects  or  functions, or only one of them is null, they compare
  unequal.  If two pointers point to nonstatic data members of the  same
  object,  the pointer to the later declared member compares higher pro­
  vided the two members  not  separated  by  an  access-specifier  label
  (_class.access.spec_) and provided their class is not a union.  If two
  pointers point to nonstatic members of the same object separated by an
  access-specifier  label  (_class.access.spec_)  the result is unspeci­
  fied.  If two pointers point to data members of the same  union,  they
  compare  equal  (after  conversion  to  void*,  if necessary).  If two
  pointers point to elements of the same array or one beyond the end  of
  the  array,  the  pointer to the object with the higher subscript com­
  pares higher.  Other pointer comparisons are implementation-defined.

  5.10  Equality operators                                     [expr.eq]

1         equality-expression:
                  relational-expression
                  equality-expression == relational-expression
                  equality-expression != relational-expression
  The == (equal to) and the != (not equal to) operators  have  the  same
  semantic  restrictions, conversions, and result type as the relational
  operators except for their lower precedence  and  truth-value  result.
  (Thus  a<b  ==  c<d  is true whenever a<b and c<d have the same truth-
  value.)

2 In addition, pointers to members of the same  type  can  be  compared.
  Pointer  to  member conversions (_conv.mem_) are performed.  A pointer
  to member can be compared to  an  integral  constant  expression  that

  evaluates  to  zero.   If one operand is a pointer to a virtual member
  function and the other is not the null pointer to  member  value,  the
  result is unspecified.

  5.11  Bitwise AND operator                              [expr.bit.and]

1         and-expression:
                  equality-expression
                  and-expression & equality-expression
  The usual arithmetic conversions are performed; the result is the bit­
  wise function of the operands.  The operator applies only to  integral
  operands.

  5.12  Bitwise exclusive OR operator                         [expr.xor]

1         exclusive-or-expression:
                  and-expression
                  exclusive-or-expression ^ and-expression
  The usual arithmetic conversions are performed; the result is the bit­
  wise exclusive function of the operands.  The operator applies only to
  integral operands.

  5.13  Bitwise inclusive OR operator                          [expr.or]

1         inclusive-or-expression:
                  exclusive-or-expression
                  inclusive-or-expression | exclusive-or-expression
  The usual arithmetic conversions are performed; the result is the bit­
  wise inclusive function of its operands.  The operator applies only to
  integral operands.

  5.14  Logical AND operator                              [expr.log.and]

1         logical-and-expression:
                  inclusive-or-expression
                  logical-and-expression && inclusive-or-expression
  The && operator groups left-to-right.  The operands are both converted
  to type bool (_conv.bool_).  The result is true if both  operands  are
  true and false otherwise.  Unlike &, && guarantees left-to-right eval­
  uation: the second operand is not evaluated if the  first  operand  is
  false.

2 The result is a bool.  All side effects of the first expression except
  for destruction of temporaries (_class.temporary_) happen  before  the
  second expression is evaluated.

  5.15  Logical OR operator                                [expr.log.or]

1         logical-or-expression:
                  logical-and-expression
                  logical-or-expression || logical-and-expression
  The || operator groups left-to-right.  The operands are both converted
  to bool (_conv.bool_).  It returns true if either of its  operands  is

  true,  and  false  otherwise.   Unlike  |, || guarantees left-to-right
  evaluation; moreover, the second operand is not evaluated if the first
  operand evaluates to true.

2 The result is a bool.  All side effects of the first expression except
  for destruction of temporaries (_class.temporary_) happen  before  the
  second expression is evaluated.

  5.16  Conditional operator                                 [expr.cond]

1         conditional-expression:
                  logical-or-expression
                  logical-or-expression ? expression : assignment-expression
  Conditional  expressions group right-to-left.  The first expression is
  converted to bool (_conv.bool_).  It is evaluated and if it  is  true,
  the  result  of  the conditional expression is the value of the second
  expression, otherwise that of the third expression.  All side  effects
  of   the  first  expression  except  for  destruction  of  temporaries
  (_class.temporary_) happen before the second or  third  expression  is
  evaluated.

2 If  either  the  second  or  third  expression  is  a throw-expression
  (_except.throw_), the result is of the type of the other.

3 If both the second and the third expressions are of  arithmetic  type,
  then  if  they are of the same type the result is of that type; other­
  wise the usual arithmetic conversions are performed to bring them to a
  common  type.  Otherwise, if both the second and the third expressions
  are either a pointer or an integral constant expression that evaluates
  to  zero, pointer conversions (_conv.ptr_) are performed to bring them
  to a common type, which shall be a qualified or unqualified version of
  the  type of either the second or the third expression.  Otherwise, if
  both the second and the third expressions are either a pointer to mem­
  ber or an integral constant expression that evaluates to zero, pointer
  to member conversions (_conv.mem_) are performed to bring  them  to  a
  common type14) which shall be a qualified or  unqualified  version  of
  the  type of either the second or the third expression.  Otherwise, if
  both the second and the third expressions are lvalues of related class
  types, they are converted to a common type as if by a cast to a refer­
  ence to the common type (_expr.static.cast_).  Otherwise, if both  the
  second  and  the third expressions are of the same class T, the common
  type is T.  Otherwise, if both the second and  the  third  expressions
  have  type cv void, the common type is cv void.  Otherwise the expres­
  sion is ill formed.  The result has the common type; only one  of  the
  second and third expressions is evaluated.  The result is an lvalue if
  the second and the third operands are of the same type  and  both  are
  lvalues.

  _________________________
  14)  This is one instance in which the composite type, as described in
  the C Standard, is still employed in C++.

  5.17  Assignment operators                                  [expr.ass]

1 There  are  several assignment operators, all of which group right-to-
  left.  All require a modifiable lvalue as their left operand, and  the
  type  of  an  assignment  expression is that of its left operand.  The
  result of the assignment operation is the value  stored  in  the  left
  operand after the assignment has taken place; the result is an lvalue.
          assignment-expression:
                  conditional-expression
                  unary-expression assignment-operator assignment-expression
                  throw-expression
          assignment-operator: one of
                  =  *=  /=  %=   +=  -=  >>=  <<=  &=  ^=  |=

2 In simple assignment (=), the value of the expression replaces that of
  the object referred to by the left operand.

3 If  the left operand is not of class type, the expression is converted
  to the unqualified type of the left operand using standard conversions
  (_conv_) and/or user-defined conversions (_class.conv_), as necessary.

4 Assignment to objects of a class (_class_) X is defined by  the  func­
  tion   X::operator=()   (_over.ass_).   Unless  the  user  defines  an
  X::operator=(),  the  default   version   is   used   for   assignment
  (_class.copy_).  This implies that an object of a class derived from X
  (directly   or   indirectly)   by   unambiguous   public    derivation
  (_class.derived_) can be assigned to an X.

5 For  class  objects, assignment is not in general the same as initial­
  ization (_dcl.init_, _class.ctor_, _class.init_, _class.copy_).

6 When the left operand of an assignment operator denotes a reference to
  T, the operation assigns to the object of type T denoted by the refer­
  ence.

7 The behavior of an expression of the form E1 op= E2 is  equivalent  to
  E1=E1 op  E2 except that E1 is evaluated only once.  E1 shall not have
  bool type.  In += and -=, E1 can be a pointer to a  possibly-qualified
  completely  defined  object type, in which case E2 shall have integral
  type and is converted as explained in _expr.add_; In all other  cases,
  E1 and E2 shall have arithmetic type.

8 See _except.throw_ for throw expressions.

  5.18  Comma operator                                      [expr.comma]

1 The comma operator groups left-to-right.
          expression:
                  assignment-expression
                  expression , assignment-expression
  A  pair of expressions separated by a comma is evaluated left-to-right
  and the value of the left expression is discarded.  All  side  effects
  of  the  left  expression  are  performed before the evaluation of the

  right expression.  The type and value of the result are the  type  and
  value  of  the  right  operand;  the  result is an lvalue if its right
  operand is.

2 In contexts where comma is given a special meaning,  for  example,  in
  lists  of arguments to functions (_expr.call_) and lists of initializ­
  ers (_dcl.init_), the comma operator as described in this  clause  can
  appear only in parentheses; for example,
          f(a, (t=3, t+2), c);
  has three arguments, the second of which has the value 5.

  5.19  Constant expressions                                [expr.const]

1 In  several places, C++ requires expressions that evaluate to an inte­
  gral constant: as array  bounds  (_dcl.array_),  as  case  expressions
  (_stmt.switch_), as bit-field lengths (_class.bit_), and as enumerator
  initializers (_dcl.enum_).
          constant-expression:
                  conditional-expression
  An   integral   constant-expression   can   involve   only    literals
  (_lex.literal_),  enumerators, const values of integral types initial­
  ized with constant expressions (_dcl.init_), and  sizeof  expressions.
  Floating  constants  (_lex.fcon_)  can appear only if they are cast to
  integral types.  Only type conversions to integral types can be  used.
  In particular, except in sizeof expressions, functions, class objects,
  pointers, or references shall not be used, and assignment,  increment,
  decrement, function-call, or comma operators shall not be used.

2 Other  expressions  are  considered  constant-expressions only for the
  purpose     of     non-local     static     object      initialization
  (_basic.start.init_).  Such constant expressions shall evaluate to one
  of the following:

  --a null pointer constant (_conv.ptr_),

  --a null member pointer value (_conv.mem_),

  --an arithmetic constant expression,

  --an address constant,

  --an address constant for an object type plus  or  minus  an  integral
    constant expression, or

  --a pointer to member constant expression.

3 An arithmetic constant expression shall have arithmetic type and shall
  only have operands that are integer constants  (_lex.icon_),  floating
  constants  (_lex.fcon_), enumerators, character constants (_lex.ccon_)
  and sizeof expressions (_expr.sizeof_).  Casts operators in an  arith­
  metic  constant  expression  shall  only  convert  arithmetic types to
  arithmetic types, except as part of an operand to the sizeof operator.

4 An address constant is a pointer to an lvalue designating an object of
  static storage duration or a function.  The pointer shall  be  created
  explicitly, using the unary & operator, or implicitly using an expres­
  sion of array (_conv.array_) or function (_conv.func_) type.  The sub­
  scripting operator [] and the class member access .  and -> operators,
  the & and * unary operators, and pointer casts (except  dynamic_casts,
  _expr.dynamic.cast_)  can  be  used in the creation of an address con­
  stant, but the value of an object shall not be accessed by the use  of
  these  operators.  An expression that designates the address of a mem­
  ber or base class of a non-POD class  object  (_class_)  is  never  an
  address constant expression (_class.cdtor_).  Function calls shall not
  be used in an address constant expression, even  if  the  function  is
  inline and has a reference return type.

5 A  pointer  to  member  constant expression shall be created using the
  unary & operator applied to a qualified-id operand  (_expr.unary.op_).