Question

I’m trying to solve exercises 8 and 9 from B. Stoustrup’s book’s, “Programming — Principles and Practice Using C++”, eighth chapter.

Exercise 8 involves writing a function named randint() that generates a random integer in the range [0;MAXINT]. What is MAXINT supposed to be? Is it the greater number possible to store in an int? Is it an argument to the function? The author doesn’t state it. The only hint given is: “The Art of Programming, Knuth, vol. 2″. That book describes an algorithm called the linear congruential method, which involves 4 parameters (see linked Wikipedia article).

Exercise 9 involves using the randint(int a, int b) function from exercise 8 to generate a random integer in the range [a;b].

Now I could write a single function implementing the linear congruential method, with 4 arguments, but that’s not what the author wants. I’m trying to understand WHAT he wants, and wondering if someone solved these exercises. Unfortunately there’s no solution to these exercises on the author’s website.

Thanks for your suggestions.

Answer 1

There is nothing in standard C++ called MAXINT.

There is a macro INT_MAX, defined in <climits> (<limits.h> in C); it expands to a constant expression whose value is the maximum value of type int. If int is 32 bits, for example, INT_MAX is probably 2147483647.

Either MAXINT is described elsewhere in the book (check the index and/or a few pages preceding the exercise), or it’s an error and Stroustrup meant to refer to INT_MAX.

If you can’t find out where this MAXINT is defined, you can probably just substitute INT_MAX.