Conversions

Within these castle walls be forged Mavens of Computer Science ...

— Merlin, The Coder

Boeing, N780BA, B747-409(LCF) Dreamlifter

Prerequisites

• W1031 Positive Integers
• W1032 Negative Integers
• W1033 Characters

Introduction

In some cases we have a value of one type but want to use it in another context, as a different type. Depending on language, this may be referred to as "type conversion", "type casting", or "type coercion". Often, type casting refers to the mere re-interpretation of existing bits, while type conversion creates a new representation of data as a different type.

Type Casting

When type casting, we consider an existing bit pattern as other than that originally declared. For example, consider the following C program:

#include <stdio.h>
int main(int argc, char *argv[]) {
  int c = 65;
  printf("The value is: %i", c);
  printf("The value is: %c", c);

  int d = -c;
  printf("The value is: %i\n", d);
  printf("The value is: %u\n", d);
  return 0;
}

On line 3, we declare an integer variable, "c" and assign the initial value, 65. On line 4, we print the integer value. On line 5, we use the printf command to indicate that we actually have a character, rather than an integer. The existing bit pattern is then re-interpreted and the character "A" is printed. (If you don't remember why it's "A", review W1033 Character Encoding.

Type Conversion

Type conversion is a more extensive process than type casting because it involves the creation of a new value based on an existing value of a different type. As we learned previously, Doubles and Integers are stored using very different bit patterns, i.e. 1.0 (a Double) is very different from 1 (an Int). But we'll encounter many situations where we'll want to perform an operation on values of more than one type, for example, and an Int to a Double. In these cases, we'll need to form a new value of the appropriate type using an existing value. Let's look at an example in Swift:

let n = 1
let d = 2.0
let r = n + d

Line 3 will result in the error: error: binary operator '+' cannot be applied to operands of type 'Int' and 'Double'. This is because Swift is protecting us from accidentally adding operands of different types. If we choose to do so, we'll need to do so explicitly. We can accomplish this by creating a new, unnamed, temporary constant of the desired type as follows:

let n = 1
let d = 2.0
let r = Double(n) + d

Pay close attention to line 3, where we create a Double from n, and then add the result to the Double d.

We use the same methodology, i.e. creating new, temporary constants for other types of conversions, such as from Int to String:

let s = "Seventeen"
let d = 2.0
let r = s + String(d)

It's important to note that the operators will be interpreted in accordance with the operands after any conversions have taken place. As such, the operator on line 3 above specifies concatenation rather than addition.

Key Concepts

Key Concepts

• Changing the type of a value may be referred to, depending on language, as "type conversion", "type casting", or "type coercion".
• Type casting refers to the mere re-interpretation of existing bits
• Type conversion creates a new representation of data as a different type
• Operators will be interpreted in accordance with the operands after any conversions have taken place

References

• Type Conversion (Wikipedia)
• Type Casting (Swift Documentation)