HP C/iX Reference Manual (31506-90011)

Chapter 4 49
Type Conversions
4 Type Conversions
The use of different types of data within C programs creates a need for data type
conversions. For example, some circumstances that may require a type conversion are
when a program assigns one variable to another, when it passes arguments to functions, or
when it tries to evaluate an expression containing operands of different types. C performs
data conversions in these situations.
• Assignment — Assignment operations cause some implicit type conversions. This
makes arithmetic operations easier to write. Assigning an integer type variable to a
floating type variable causes an automatic conversion from the integer type to the
floating type.
• Function call — Arguments to functions are implicitly converted following a number
of 'widening' conversions. For example, characters are automatically converted to
integers when passed as function arguments in the absence of a prototype.
• Normal conversions — In preparation for arithmetic or logical operations, the
compiler automatically converts from one type to another. Also, if two operands are not
of the same type, one or both may be converted to a common type before the operation is
performed.
• Casting — You can explicitly force a conversion from one type to another using a cast
operation.
• Returned values — Values returned from a function are automatically converted to
the function's type. For example, if a function was declared to return a double and the
return statement has an integer expression, the integer value is automatically
converted to a double.
Conversions from one type to another do not always cause an actual physical change in
representation. Converting a 16-bit short int into a 64-bit double causes a representational
change. Converting a 16-bit signed short int to a 16-bit unsigned short int does not cause a
representational change.