Recapping namespace basics

The namespace utility was added to C++ to add a scoping level beyond global scope to applications. This feature can be used to allow two or more libraries to be utilized without concern that they may contain duplicative data types, functions, or identifiers. The programmer needs to activate the desired namespace in each relevant portion of their application with the keyword using. Programmers can also create their own namespaces (usually for creating reusable library code) and activate each namespace as applicable. In the previous examples, we’ve seen the simple use of the std namespace to include cin and cout, which are instances of istream and ostream (whose definitions are found in <iostream>). Let’s review how we can create namespaces ourselves:

https://github.com/PacktPublishing/Deciphering-Object-Oriented-Programming-with-CPP/blob/main/Chapter01/Chp1-Ex11.cpp

#include <iostream>
// using namespace std; // Do not open entire std namespace
using std::cout;   // Instead, activate individual elements
using std::endl;   // within the namespace as needed
namespace DataTypes
{
    int total;
    class LinkList
    {  // full class definition … 
    };
    class Stack
    {  // full class definition …
    };
};
namespace AbstractDataTypes
{
    class Stack
    {  // full class definition …
    };
    class Queue
    {  // full class description …
    };
};
// Add entries to the AbstractDataTypes namespace
namespace AbstractDataTypes   
{
    int total;
    class Tree
    {  // full class definition …
    };
};
int main()
{
    using namespace AbstractDataTypes; //activate namespace
    using DataTypes::LinkList;    // activate only LinkList 
    LinkList list1;     // LinkList is found in DataTypes
    Stack stack1;    // Stack is found in AbstractDataTypes
    total = 5;       // total from active AbstractDataTypes
    DataTypes::total = 85;// specify non-active mbr., total
    cout << "total " << total << "\n";
    cout << "DataTypes::total " << DataTypes::total;
    cout << endl;
    return 0;        
}

In the second line of the preceding code (which is commented out), we notice the keyword using applied to indicate that we’d like to use or activate the entire std namespace. Preferably, on the following two lines of code, we can instead activate only the elements in the standard namespace that we will be needing, such as std::cout or std::endl. We can utilize using to open existing libraries (or individual elements within those libraries) that may contain useful classes; the keyword using activates the namespace to which a given library may belong. Next in the code, a user specified namespace is created called DataTypes, using the namespace keyword. Within this namespace exists a variable total, and two class definitions: LinkList and Stack. Following this namespace, a second namespace, AbstractDataTypes, is created and includes two class definitions: Stack and Queue. Additionally, the namespace AbstractDataTypes is augmented by a second occurrence of the namespace definition in which a variable total and a class definition for Tree are added.

In the main() function, first, the AbstractDataTypes namespace is opened with the using keyword. This activates all names in this namespace. Next, the keyword using is combined with the scope resolution operator (::) to only activate the LinkList class definition from the DataTypes namespace. Had there also been a LinkList class within the AbstractDataType namespace, the initial visible LinkList would now be hidden by the activation of DataTypes::LinkList.

Next, a variable of type LinkList is declared, whose definition comes from the DataTypes namespace. A variable of type Stack is next declared; though both namespaces have a Stack class definition, there is no ambiguity since only one Stack has been activated. Next, we use cin to read into total, which is active from the AbstractDataTypes namespace. Lastly, we use the scope resolution operator to explicitly read into DataTypes::total, a variable that would otherwise be hidden. One caveat to note: should two or more namespaces contain the same identifier, the one last opened will preside, hiding all previous occurrences.

It is considered good practice to activate only the elements of a namespace we wish to utilize. From the aforementioned example, we can see potential ambiguity that can otherwise arise.