Using the C++ Math Library

Specifying Library File I/O Functions

Because the MATLAB C++ Math Library file I/O functions have the same name as their C++ counterparts and because the types of their arguments are so similar, you must be careful to make sure you're calling the correct one.

This is particularly important with fprintf(). The type of the first argument to fprintf() is all important: if it is an array, the system calls the MATLAB C++ Math Library function; if it is an integer, the system calls the standard C++ function. Consider this example:

mwArray file("foo.txt"), data=rand(4);
int fd = fopen(file);
fprintf(fd, "%f", data);

The system calls the standard C++ fprintf() function because the first argument passed to fprintf() is an integer. But this is almost certainly not what the author intended; the standard C++ fprintf() uses the format string to determine how many arguments it has. In this case, it will think there is a single argument and the program will crash because the standard fprintf() function does not understand mwArray objects.

The MATLAB C++ Math Library version of sprintf() requires that you pass an mwArray as its second argument. The other arguments may be passed as character strings.

Using File I/O Functions

Example - Using File I/O Functions (ex6.cpp)