Why does the behavior of conversion character specifiers for FPRINTF differ from what is observed in C?

When I use the FPRINTF function with the conversion character specifiers %x, %c, %u, %d, %i or %o for an invalid input argument, MATLAB returns the same input argument expressed in exponential form.

For example, when I execute the following code:

fprintf('%x\n',-1.5)

fprintf('%c\n',-1234.5)

fprintf('%u\n',-1234.5)

fprintf('%d\n',-1.5)

fprintf('%i\n',-21.5)

fprintf('%o\n',21.5)

MATLAB returns:

 -1.500000e+000

 -1.234500e+003

 -1.234500e+003

 -1.500000e+000

 -2.150000e+001

 2.150000e+001

Rather than returning my input in exponential form, I would expect MATLAB to return warnings or errors to reflect the presence of invalid input arguments.

Answer 1

[MathWorks Support Team]

Note that although MATLAB's low-level file I/O routines are similar to those in C, they do not behave exactly the same way. This is expected behavior for MATLAB.