bdlat_nullablevaluefunctions Component

Provide a namespace defining nullable value functions. More...

Namespaces
namespace	bdlat_NullableValueFunctions

Detailed Description

Outline

Purpose
Classes
Description
- Usage
- Example 1: Defining a "Nullable" Type

Purpose:: Provide a namespace defining nullable value functions.

Classes:

bdlat_NullableValueFunctions

namespace for "nullable" value functions

See also:: Component bdlb_nullablevalue

Description:: The bdlat_NullableValueFunctions namespace provided in this component defines parameterized functions that expose "nullable" behavior for "nullable" types. See the bdlat package-level documentation for a full description of "nullable" types.

The functions in this namespace allow users to:

make the nullable object contain a value (makeValue),
manipulate the value contained in a nullable object using a parameterized manipulator functor (manipulateValue),
access the value contained in a nullable object using a parameterized accessor functor (accessValue), and
check whether the nullable object is null or not (isNull).

: A type becomes part of the bdlat "nullable" framework by creating, in the namespace where the type is defined, specializations of the following four (free) function templates:

A type becomes part of the bdlat "nullable" framework by creating, in the namespace where the type is defined, overloads of the following two (free) functions and two (free) function templates. Note that the placeholder YOUR_TYPE is not a template argument and should be replaced with the name of the type being plugged into the framework.

  // MANIPULATORS
  void bdlat_nullableValueMakeValue(YOUR_TYPE *object);
      // Assign to the specified "nullable" 'object' the default value for
      // the contained type (i.e., 'ValueType()').

  template <class MANIPULATOR>
  int bdlat_nullableValueManipulateValue(YOUR_TYPE    *object,
                                         MANIPULATOR&  manipulator);
      // Invoke the specified 'manipulator' on the address of the value
      // stored in the specified "nullable" 'object'.  Return the value from
      // the invocation of 'manipulator'.  The behavior is undefined if
      // 'object' contains a null value.

  // ACCESSORS
  template <class ACCESSOR>
  int bdlat_nullableValueAccessValue(const YOUR_TYPE& object,
                                     ACCESSOR&        accessor);
      // Invoke the specified 'accessor' on a 'const'-reference to the value
      // stored in the specified "nullable" 'object'.  Return the value from
      // the invocation of 'accessor'.  The behavior is undefined if 'object'
      // contains a null value.

  bool bdlat_nullableValueIsNull(const YOUR_TYPE& object);
      // Return 'true' if the specified "nullable" 'object' contains a null
      // value, and 'false' otherwise.

The "nullable" type must also define two meta-functions in the bdlat_NullableValueFunctions namespace:

the meta-function IsNullableValue contains a compile-time constant VALUE that is non-zero if the parameterized TYPE exposes "nullable" behavior, and
the ValueType meta-function contains a typedef Type that specifies the type of the value that can be stored in the parameterized "nullable" type.

: Note that bdlb::NullableValue<TYPE> is already part of the bldat infrastructure for "nullable" types because this component also provides overloads of the required functions and meta-function specializations.

Usage:: ------ The following code illustrate the usage of this component.

Example 1: Defining a "Nullable" Type:

- - - - - - - - - - - - - - - - - - Suppose you had a type whose value could be in a "null" state.

  namespace BloombergLP {
  namespace mine {

  struct MyNullableValue {

      // DATA
      bool d_isNull;
      int  d_value;

      // CREATORS
      MyNullableValue()
      {
          d_isNull = true;
      }
  };

  }  // close namespace mine
  }  // close enterprise namespace

We can now make mine::MyNullableValue expose "nullable" behavior by implementing the necessary bdlta_NullableValueFunctions for MyNullableValue inside the mine namespace and defining the required meta-functions withing the bdlat_NullableValueFunctions namespace.

First, we should forward declare all the functions that we will implement inside the mine namespace:

  namespace BloombergLP {
  namespace mine {

  // MANIPULATORS
  void bdlat_nullableValueMakeValue(MyNullableValue *object);
      // Assign to the specified "nullable" 'object' the default value for
      // the contained type (i.e., 'ValueType()').

  template <class MANIPULATOR>
  int bdlat_nullableValueManipulateValue(MyNullableValue *object,
                                         MANIPULATOR&     manipulator);
      // Invoke the specified 'manipulator' on the address of the value
      // stored in the specified "nullable" 'object'.  Return the value from
      // the invocation of 'manipulator'.  The behavior is undefined if
      // 'object' contains a null value.

  // ACCESSORS
  template <class ACCESSOR>
  int bdlat_nullableValueAccessValue(const MyNullableValue& object,
                                     ACCESSOR&              accessor);
      // Invoke the specified 'accessor' on a 'const'-reference to the value
      // stored in the specified "nullable" 'object'.  Return the value from
      // the invocation of 'accessor'.  The behavior is undefined if 'object'
      // contains a null value.

  bool bdlat_nullableValueIsNull(const MyNullableValue& object);
      // Return 'true' if the specified "nullable" 'object' contains a null
      // value, and 'false' otherwise.

  }  // close namespace mine
  }  // close enterprise namespace

Then, we will implement these functions. Recall that the two (non-template) functions should be defined in some .cpp file, unless you choose to make them inline functions.

  namespace BloombergLP {

  // MANIPULATORS
  void mine::bdlat_nullableValueMakeValue(MyNullableValue *object)
  {
      assert(object);

      object->d_isNull = false;
      object->d_value  = 0;
  }

  template <class MANIPULATOR>
  int mine::bdlat_nullableValueManipulateValue(MyNullableValue *object,
                                               MANIPULATOR&     manipulator)
  {
      assert(object);
      assert(!object->d_isNull);

      return manipulator(&object->d_value);
  }

  // ACCESSORS
  template <class ACCESSOR>
  int mine::bdlat_nullableValueAccessValue(const MyNullableValue& object,
                                           ACCESSOR&              accessor)
  {
      assert(!object.d_isNull);

      return accessor(object.d_value);
  }

  bool mine::bdlat_nullableValueIsNull(const MyNullableValue& object)
  {
      return object.d_isNull;
  }

  }  // close enterprise namespace

Finally, we specialize the IsNullableValue and ValueType meta-functions in the bdlat_NullableValueFunctions namespace for the mine::MyNullableValue type:

  namespace BloombergLP {
  namespace bdlat_NullableValueFunctions {

  // TRAITS
  template <>
  struct IsNullableValue<mine::MyNullableValue> {
      enum { VALUE = 1 };
  };

  template <>
  struct ValueType<mine::MyNullableValue> {
      typedef int Type;
  };

  }  // close namespace bdlat_NullableValueFunctions
  }  // close enterprise namespace

This completes the bdlat infrastructure for mine::MyNullableValue and allows the generic software to recognize the type as a nullable abstraction.

Example 2: Using the Infrastructure Via General Methods:: The bdlat "nullable" framework provides a set of fundamental operations common to any "nullable" type. We can build upon these operations to make our own utilities, or use them on our own types that are plugged into the framework, like mine::MyNullableValue, which we created in Example 1. For example, we can use the (fundamental) operations in the bdlat_NullableValueFunctions namespace to operate on mine::NullableValue, even though they have no knowledge of that type in particular:

Two of those operations are rather basic. One simply informs whether or not an object is in the null state (the isNull method). Another sets an object to a default, non-null state (the makeValue method).

  void usageMakeObject()
  {
      BSLMF_ASSERT(bdlat_NullableValueFunctions::
                   IsNullableValue<mine::MyNullableValue>::VALUE);

      mine::MyNullableValue object;
      assert( bdlat_NullableValueFunctions::isNull(object));

      bdlat_NullableValueFunctions::makeValue(&object);
      assert(!bdlat_NullableValueFunctions::isNull(object));
  }

The other two generic methods accomplish their actions via user-supplied functors.

Let us define a generic functor that gives us access to the underlying value of the "nullable" type, if it's not null:

  template <class VALUE_TYPE>
  class GetValueAccessor {

      // DATA
      VALUE_TYPE *d_value_p;

    public:
      // CREATORS
      explicit GetValueAccessor(VALUE_TYPE *value)
      : d_value_p(value)
      {
      }

      // MANIPULATORS
      int operator()(const VALUE_TYPE& containedValue)
          // Assign the value of the specified 'containedValue' to the object
          // addressed by 'd_value_p'.
      {
          *d_value_p = containedValue;
          return 0;
      }
  };

Notice that the above class makes no assumptions about the value being accessed other than it can be copied (in the constructor) and assigned (in the operator).

This functor can be used to fetch the value of our nullable object:

  void usageGetValue()
  {
      mine::MyNullableValue object;

      bdlat_NullableValueFunctions::makeValue(&object);
      assert(!bdlat_NullableValueFunctions::isNull(object));

      int value;
      GetValueAccessor<int> accessor(&value);

      int rc = bdlat_NullableValueFunctions::accessValue(object, accessor);
      assert(0 == rc);
      assert(0 == value);
  }

Notice that we did not invoke accessValue until object had been set to a non-null state. Doing otherwise would have led to undefined behavior.

Finally, let's define a functor to set the state of a nullable object:

  template <class VALUE_TYPE>
  class SetValueManipulator {

      // DATA
      VALUE_TYPE d_value;

    public:
      // CREATORS
      explicit SetValueManipulator(const VALUE_TYPE& value)
      : d_value(value)
      {
      }

      // ACCESSOR
      int operator()(VALUE_TYPE *value) const
      {
          *value = d_value;
          return 0;
      }
  };

As with the previous functor, this functor has no knowledge of the nullable type to which it will be applied. The only assumption here is that the value (type) of our nullable type can be copy constructed and copy assigned.

Let us use this functor to modify one of our nullable objects:

  void usageSetValue()
  {
      mine::MyNullableValue object;

      bdlat_NullableValueFunctions::makeValue(&object);
      assert(!bdlat_NullableValueFunctions::isNull(object));

      SetValueManipulator<int> manipulator(42);
      int rcm = bdlat_NullableValueFunctions::manipulateValue(&object,
                                                              manipulator);
      assert(0 == rcm);

      // Confirm that the object was set to the expected state.

      int value;
      GetValueAccessor<int> accessor(&value);

      int rca = bdlat_NullableValueFunctions::accessValue(object, accessor);
      assert( 0 == rca);
      assert(42 == value);
  }

Example 3: Defining Utility Functions:: Creating functor objects for each operation can be tedious and error prone; consequently, those types are often executed via utility functions.

: Suppose we want to create utilities for getting and setting the value associated with an arbitrary "nullable" type.

: These functors make minimal assumptions of VALUE_TYPE, merely that it is copy constructable and copy assignable.

  struct NullableValueUtil {

      // CLASS METHODS
      template <class NULLABLE_VALUE_TYPE>
      static int getValue(
          typename bdlat_NullableValueFunctions
                              ::ValueType<NULLABLE_VALUE_TYPE>::Type *value,
          const NULLABLE_VALUE_TYPE&                                  object)
          // Load to the specified 'value' the value of the specified
          // nullable value 'object'.  This function template requires that
          // the specified 'NULLABLE_VALUE_TYPE' is a 'bdlat' "nullable"
          // type.  The behavior is undefined unless 'object' is in a
          // non-null state (i.e.,
          // 'false == bdlat_NullableValueFunctions::isNull(object))'.
      {
          BSLMF_ASSERT(bdlat_NullableValueFunctions
                              ::IsNullableValue<NULLABLE_VALUE_TYPE>::VALUE);

          BSLS_ASSERT(!bdlat_NullableValueFunctions::isNull(object));

          typedef typename bdlat_NullableValueFunctions
                            ::ValueType<NULLABLE_VALUE_TYPE>::Type ValueType;

          GetValueAccessor<ValueType> valueAccessor(value);
          return bdlat_NullableValueFunctions::accessValue(object,
                                                           valueAccessor);
      }

      template <class NULLABLE_VALUE_TYPE>
      static int setValue(NULLABLE_VALUE_TYPE                        *object,
                          const typename bdlat_NullableValueFunctions
                             ::ValueType<NULLABLE_VALUE_TYPE>::Type&  value)
          // Set the value of the specified 'object' to the specified
          // 'value'.  This function template requires that the specified
          // 'NULLABLE_VALUE_TYPE' is a 'bdlat' "nullable" type.  The
          // behavior is undefined unless 'object' is in a non-null state
          // (i.e., 'false == bdlat_NullableValueFunctions::isNull(object))'.
          // Note that a "nullable" object can be put into a non-null state
          // by the 'bdlat_NullableValueFunctions::makeValue' function
          // overload for the 'NULLABLE_VALUE_TYPE'.
      {
          BSLMF_ASSERT(bdlat_NullableValueFunctions
                              ::IsNullableValue<NULLABLE_VALUE_TYPE>::VALUE);

          BSLS_ASSERT(object);
          BSLS_ASSERT(!bdlat_NullableValueFunctions::isNull(*object));

          typedef typename bdlat_NullableValueFunctions
                            ::ValueType<NULLABLE_VALUE_TYPE>::Type ValueType;

          SetValueManipulator<ValueType> manipulator(value);
          return bdlat_NullableValueFunctions::manipulateValue(object,
                                                               manipulator);
      }
  };

Now, we can use these functors to write generic utility functions for getting and setting the value types of arbitrary "nullable" classes.

  void myUsageScenario()
  {
      mine::MyNullableValue object;
      assert(bdlat_NullableValueFunctions::isNull(object));

      bdlat_NullableValueFunctions::makeValue(&object);
      assert(!bdlat_NullableValueFunctions::isNull(object));

      typedef
      bdlat_NullableValueFunctions::ValueType<mine::MyNullableValue>::Type
                                                               MyValueType;

      int rcs = NullableValueUtil::setValue(&object, MyValueType(42));
      assert(0 == rcs);

      MyValueType value;
      int rcg = NullableValueUtil::getValue(&value, object);
      assert( 0 == rcg);
      assert(42 == value);
  }

Example 4: Achieving Type Independence:

Finally, suppose we have another type such as your::YourNullableType, shown below:

  namespace BloombergLP {
  namespace your {

  class YourNullableValue {

      // DATA
      bool        d_isNull;
      bsl::string d_value;

    public:
      // CREATORS
      YourNullableValue()
      : d_isNull(true)
      , d_value()
      {
      }

      // MANIPULATORS
      void makeValue()
      {
          d_isNull = false;
          d_value.clear();
      }

      void makeNull()
      {
          d_isNull = true;
          d_value.clear();
      }

      bsl::string& value()
      {
          assert(!d_isNull);

          return d_value;
      }

      // ACCESSORS
      const bsl::string& value() const
      {
          assert(!d_isNull);

          return d_value;
      }
      bool isNull() const
      {
          return d_isNull;
      }
  };

  }  // close namespace your
  }  // close enterprise namespace

Notice that while there are many similarities to mine::MyNullableValue there are clearly differences:

The value type is bsl::string, not int.
Attributes are accessed via accessor methods, not public data members.

Nevertheless, since your::YourNullableValue also provides the functions and types expected by the bdlat infrastructure (not shown) we can successfully use your::YourNullableValue value instead of mine::MyNullableValue in the previous usage scenario, with no other changes:

  void yourUsageScenario()
  {
      your::YourNullableValue object;  // YOUR NULLABLE TYPE
      assert(bdlat_NullableValueFunctions::isNull(object));

      bdlat_NullableValueFunctions::makeValue(&object);
      assert(!bdlat_NullableValueFunctions::isNull(object));

      typedef
      bdlat_NullableValueFunctions::ValueType<your::YourNullableValue>::Type
                                                               YourValueType;

      int rcs = NullableValueUtil::setValue(&object, YourValueType("NB"));
      assert(0 == rcs);

      YourValueType value;
      int rcg = NullableValueUtil::getValue(&value, object);
      assert(  0  == rcg);
      assert("NB" == value);
  }

Notice that syntax and order of bdlat_NullableValueFunction functions calls have not been changed. The only difference is that the contained type has changed from int to bsl::string.

Finally, instead of defining a new "nullable" type, we could substitute the existing type template bdlb::NullableValue. Note that this component provides specializations of the bdlat_nullableValueFunctions for that type. Since the accessor and manipulator functions we created earlier are type neutral, we can simply drop bdlb::NullableValue<float> into our familiar scenario:

  void anotherUsageScenario()
  {
      bdlb::NullableValue<float> object;  // BDE NULLABLE TYPE
      assert(bdlat_NullableValueFunctions::isNull(object));

      bdlat_NullableValueFunctions::makeValue(&object);
      assert(!bdlat_NullableValueFunctions::isNull(object));

      typedef
      bdlat_NullableValueFunctions::ValueType<bdlb::NullableValue<float> >
                                                  ::Type AnotherValueType;

      int rcs = NullableValueUtil::setValue(&object, AnotherValueType(2.0));
      assert(0 == rcs);

      AnotherValueType value;
      int rcg = NullableValueUtil::getValue(&value, object);
      assert(0   == rcg);
      assert(2.0 == value);
  }

Component bdlat_nullablevaluefunctions [Package bdlat]

Namespaces

Detailed Description

Component bdlat_nullablevaluefunctions
[Package bdlat]