Chapter 6: Templates, Iterators, STL

1. What is the primary purpose of template functions?

   • A. To allow a single function to be used with varying types of arguments

   • B. To hide the name of the function from the linker (preventing duplicate symbols)

   • C. To implement container classes

   • D. To permit the use of the debugger without the -gstabs flag

A. To allow a single function to be used with varying types of arguments

1. Consider this prototype for a template function:

   
       template <class Item>
       void foo(Item x);
   

   Which is the right way to call the foo function with an integer argument i?

   • A. foo( i );

   • B. foo<int>( i );

   • C. foo<Item>( i );

   • D. foo(<int> i );

   • E. foo(<Item> i );

Answer: A. foo(i);

1. Consider the following definition:

   
       template <class Item>
       Item maximal (Item a, Item b)
       {
           if (a > b)
               return a;
           else
               return b;
       }
   

   What restrictions are placed on the Item data type for a program that uses the maximal function?

   • A. The Item data type must be either int, double, or float.

   • B. The Item data type must be one of the built-in C++ data types.

   • C. The Item data type must have a copy constructor and a > operator defined.

   • D. None of the above restrictions apply.

C. The Item data type must have a copy constructor and a > operator defined.

1. When should a function be implemented as a template function?

   • A. When the data types of the parameters all have copy constructors.

   • B. When the function depends on an underlying data type.

   • C. When the function is relatively short (usually just one line).

   • D. When the function only takes one argument.

B. When the function depends on an underlying data type.

1. What is a major difference between the header file for a toolkit of template functions and the header file for a toolkit of ordinary functions?

   • A. The ordinary function toolkit header file must have a #include for the implementation file, but the template version does not have this #include.

   • B. The template function toolkit header file must have a #include for the implementation file, but the ordinary version does not have this #include.

   • C. The ordinary function toolkit header file must have a macro guard, but the template version does not have this macro guard.

   • D. The template function toolkit header file must have a macro guard, but the ordinary version does not have this macro guard.

B. The template function toolkit header file must have a #include for the implementation file, but the ordinary version does not have this #include.

Why is it a bad idea to have a size_t parameter for a template function?

• A. Because it would require stdlib.h to be separately compiled.

• B. Because most compilers would not permit the actual argument to be 42 (an int).

• C. Because size_t values cannot be negative.

• D. Because the compiler can tell the size of something without having a size_t parameter.

C. Because size_t values cannot be negative.

1. Our original bag classes in Chapters 3-5 used a typedef to define the Item data type. What problem is solved by using a template bag class instead of these original typedef versions?

   • A. None of the typedef versions permit a program to use a bag of Strings.

   • B. With all of the typedef versions, it is difficult for a program to use several bags with different Item types.

   • C. With all of the typedef versions, the CAPACITY of the bag is fixed during compilation. A program cannot dynamically allocate a bag.

   • D. With all of the typedef versions, the Item data type must be one of the built-in C++ data types (char, int, etc.)

B. With all the typedef versions, it is difficult for a program to use several bags with different Item types.

1. Suppose bag is a template class, what is the syntax for declaring a bag b of integers?

   • A. bag b;

   • B. bag<int> b;

   • C. bag of int b;

   • D. int bag b;

B. bag<int> b;

1. Why is it recommended to first implement a class using a typedef for the Item type, before implementing the template version?

   • A. It is easier to debug the typedef version.

   • B. The typedef version requires less disk space during the design stage.

   • C. The typedef version requires less memory during execution.

   • D. The template version will not compile unless you implement the typedef version first.

A. It is easier to debug the typedef version.

When you write a template class, where does the template prefix occur?

• A. Before the template class definition

• B. Before each member function implementation.

• C. Before any other template functions that manipulate the template class.

• D. TWO of the above answers are correct.

• E. All of the (A), (B), and (C) are correct.

• D. TWO of the above answers are correct.

1. Suppose that a program wants to use both a bag of doubles and a bag of strings. Which version of the bag could you use?

   • A. You can use the array version of the non-template bag

   • B. You can use the linked list version of the non-template bag

   • C. You can use the array version of the template bag

   • D. Two of the above answers are right

   • E. Answers A, B, and C are all right

C. You can use the array version of the template bag.

1. What technique is used to provide the capability to step through items of a container class?

   • A. A copy constructor.

   • B. A default constructor.

   • C. A destructor.

   • D. An iterator.

   • E. An overloaded assignment operator.

D. An iterator.

1. Why does the new node template class require two versions of the data member function, but the node in Chapter 5 needed only one?

   • A. The Chapter 5 node was singly linked, but the node template class is doubly linked.

   • B. The Chapter 5 data function returned a copy of the data, but the data function for the node template class returns a reference to the data.

   • C. The Chapter 5 node had no iterator.

   • D. All of the above.

B. The Chapter 5 data function returned a copy of the data, but the data function for the node template class returns a reference to the data.