Homework 2: Templates and the standard library

In the lecture, we talked about STL being a misnomer. STL was a historical predecessor of parts of the C++ standard library. Nowadays, we would refer to those parts by name. What are the four components of STL that are now part of the standard library?

Suppose we have a C++ class called Student. There are many ways one can create a collection Student objects. Among these: one can use a normal C++ array or an STL vector, and one could create a collection of objects, or of pointers to objects (presumably allocated on the heap.)

1. (2 pts) Write a line of C++ code that declares an array a that can hold 10 objects of type Student.

2. (2 pts) Write a line of C++ code that declares an array b that can hold 10 pointers to objects of type Student.

3. (2 pts) Write a line of C++ code that declares a standard library vector c that can hold 10 objects of type Student.

4. (2 pts) Write a line of C++ code that declares a standard library vector d that can hold 10 pointers to objects of type Student.

In the previous problem, you were asked to write four different declarations, 1, 2, 3, 4. Each of these has "pros" and "cons". There are also differences among them that we might describe as "neutral", for whether they are a pro/con depends on the context. For the "pros" and "cons" below, please indicate which of the options above (1, 2, 3, 4) the statement applies to. Note that in some cases, you may have to indicate more than one number.

Circle the numbers to which the statement applies. If you mess up, cross out all the letters and list the letters in the space to the right.

In your own words, what is the "problem" for which "templates for functions" is a solution? I'm looking for a brief description–a single sentence, or at most 2-3 sentences that get to the central point, a description of the problem, not a detailed description of everything you know about templates, and certainly not a sentence copied, word-for-word, from either textbook.

The C++ syntax for function templates includes template <class Item>. The name Item in the example above is preceeded by the keyword class. One might infer from that that Item should be the name of some class, e.g. class Student, or class Roster, etc. But that is not a correct assumption. Explain why briefly.

Let's say you have a function int calc(int operand1, int operand2, char op) where op is either + or -. Instead of overloading this function for each numerical datatype in C++ (e.g. also defining it for float, double, etc., you want to create a generic function that is far more flexible. In the space below, define calc as a template function. If op is neither + nor - , print an error message on std::cerr and call the system function exit(1); (you may assume that #include <cstdlib> has already been done.)Note that once we've covered exceptions, that would be a better approach than the cerr/exit technique.