Using the Transform() Algorithm to Change a String's Case : Page 3

Design Improvements
There are ways to further benefit from using transform(). Suppose you need to transform a string to lowercase rather than uppercase. You change the transform() call to:

std::transform(s.begin(), 
               s.end(), 
               std::back_inserter(s1), 
               std::tolower);

It's not much of an improvement compared to the original for loop, is it? To avoid intrusive code changes such as this, use an additional level of indirection. Instead of passing a function's name as an argument, use a pointer to a function. This way, you can decouple the transform() call from the customers' requirements. Furthermore, the use of a pointer enables you to postpone the function binding to runtime:

int (*pf)(int)=tolower; 
transform(s.begin(), s.end(), s.begin(), pf); 
//lowercase

Notice that you don't need to change the transform() call now if you wish to apply yet another transformation:

pf=tocyrillic; // just an example
transform(s.begin(), s.end(), s.begin(), pf); 
// Cyrillic

Conclusions
If using transform() to convert a string to uppercase seems like overkill to you, you're probably right. The string transformation was a red herring, though. The point was to show how to use transform() to manipulate sequences in a generic fashion. By using transform(), transforming a sequence of integers to their negative or square root values is a cinch:

template <class T> class negate
{
public:
 T operator()(T t) { return -t;}
};

int arr[]={1, 2, 3};
transform(arr, 
          arr+sizeof(arr)/sizeof(int), 
          arr, 
          negate<int>()); // arr = {-1, -2, -3}

Previous Page