CSC 331

Programming Languages

Data Types

Lecture 7 – Data Types CSC 331 1

 Data Types
• What is a data type?
A name with certain attributes:
– The values that can be stored, the internal representation,
the operations, …
• A data type is a set of values
• e.g., int in Java:
int x;
xIntegers= [-2147483648, 2147483647]
• A data type is also a set of operations on the values

Lecture 7 – Data Types 2

Why are data types important?
• Example: z = x / y; (Java)
– int x, y; x=5; y=2;
• Integer division, x/y results in 2.
• int z: z = 2;
• double z: z=2.0;
– double x, y; x=5; y=2;
• floating-point division, x/y results in 2.5
• int z: wrong!
• double z: z=2.5;

Lecture 7 – Data Types CSC 331 3

 Java Types

Primitive Reference

boolean Numeric Array class interface

Type structure
Integral Floating point of Java
char float
byte double
short
int
long
Lecture 7 – Data Types CSC 331 4
 C Types

Basic Derived

void Numeric Pointer Array Function struct union

Integral Floating

float
Type structure
double
(signed) enum long double of C
(unsigned)
char
int
short int
long int

Lecture 7 – Data Types CSC 331 5

 Simple Data Types
• No internal structure:
e.g., integer, double, character, and boolean.
• Often directly supported in hardware.
– machine dependency
• Most predefined types are simple types.
– Exceptions: String in Java.
• Some simple types are not predefined
• Enumerated types
• Subrange types

Lecture 7 – Data Types CSC 331 6

 Enumerated Types
Ordered set, whose elements are named and listed explicitly.

• Examples:
enum Color_Type {Red, Green, Blue}; ( C )
type Color_Type is (Red, Green, Blue); ( Ada )
datatype Color_Type = Red | Green | Blue; ( ML )

• Operations: ?
Successor and predecessor

Lecture 7 – Data Types CSC 331 7

 Ada Example
type Color_Type is (Red, Green, Blue);

x : Color_Type := Green;
x : Color_Type’Succ(x);
x : Color_Type’Pred(x);
put(x); -- prints GREEN

• No assumptions about the internal representation of

values
• Print the value name itself
8
 Pascal Example
type
cardsuit = (club, diamond, heart, spade);
card = record
suit: cardsuit;
value: 1 .. 13;
end;
var
hand: array [ 1 .. 13 ] of card;

• Succ(diamond) = heart; Pred(spade) = heart;

• club < heart; is true.
• for acard := club to heart do
9
 C Example
#include <stdio.h>
enum Color {Red, Green, Blue};
enum Courses {CSE1111=1, CSE3302=3, CSE3310=3, CSE5555=4};
main() {
enum Color x = Green;
enum Courses c = CSE3302;
x++;
printf("%d\n“,x);
printf("%d\n",Blue+1);
printf("%d\n",c);
return 0;
}
• Enum in C is simply int
• Can customize the values
10
 Java Example
public enum Planet { MERCURY (2.4397e6), EARTH (6.37814e6);
private final double radius; // in meters
Planet(double radius) {this.radius = radius; }
private double radius() { return radius; }

public static void main(String[] args) {

for (Planet p : Planet.values())
System.out.printf(“The radius of %s is %f\n", p, p.radius());
}
}

java.util.Enumeration has different meaning

for (Enumeration e = v.elements() ; e.hasMoreElements() ;)
System.out.println(e.nextElement());
Lecture 7 – Data Types 11
Evaluation of Enumeration Types
• Efficiency – e.g., compiler can select and use a compact
efficient representation (e.g., small integers)

• Readability -- e.g. no need to code a color as a number

• Maintainability – e.g., adding a new color doesn’t require

updating hard-coded constants.

• Reliability -- e.g. compiler can check operations and ranges of

value.

Courtesy of Charles Nicholas at UMBC

Lecture 7 – Data Types 12
C Example for Maintainability
enum Color {White, Green, Blue, Black};
enum Color {White, Yellow, Green, Blue, Black};
main(){
enum Color x = Black;
int i = x;
while (i >= White){
if (i < Green)
printf(“this is a light color!\n”);
i--;
}
}
What if no enumeration?
if (i < 1) printf(“this is a light color!\n”);
Has to be changed to:
if (i < 2) printf(“this is a light color!\n”);
Lecture 7 – Data Types 13
Ada Example for Reliability
type DAY is (MON, TUE, WED, THU, FRI, SAT, SUN);
type DIRECTION is (NORTH, EAST, SOUTH, WEST);

GOAL : DIRECTION;
TODAY : DAY;
START : DAY;

TODAY := MON;
GOAL := WEST;
START := TODAY;

TODAY := WEST; -- Illegal: WEST is not a DAY value

TODAY := 5; -- Illegal: 5 is not a DAY value
TODAY := TODAY + START; -- Illegal: "+" is not defined for DAYS

Lecture 7 – Data Types 14

 Subrange Types
Contiguous subsets of simple types, with a least and greatest element.
• Example:
type Digit_Type is range 0..9; ( Ada)
• Not available in C,C++,Java. Need to use something like:
byte digit; //-128..127
...
if (digit>9 || digit <0) throw new DigitException();
• defined over ordinal types:
– ordered, every value has a next/previous element
• E.g., integer, enumerations, and subrange itself

Lecture 7 – Data Types CSC 331 15

Type constructors: Defining New Types

• Type constructors as set operations:

– Cartesian product
– Union
– Subset
– Functions (Arrays)
• Some type constructors do not correspond to set operations
(e.g., pointers)
• Some set operators don’t have corresponding type
constructors (e.g., intersection)

Lecture 7 – Data Types CSC 331 16

 Cartesian Product
• Ordered Pairs of elements from U and V
U  V = {(u, v) | uU and vV}

• Operations:
– projection
p1: U  V → U; p2 : U  V → V
p1((u,v))=u; p2((u,v))=v

Lecture 7 – Data Types CSC 331 17

 Examples
• struct in C
struct IntCharReal
{ int i;
char c;
double r;
}
int  char  double
• record in Ada
type IntCharReal is record
i: integer;
c: character;
r: float;
end record;

Lecture 7 – Data Types CSC 331 18

 The same type?
struct IntCharReal
{ int i;
char c;
double r;
}

struct IntCharReal
{ char c;
int i;
double r;
}
Lecture 7 – Data Types CSC 331 19
 The same type?
struct IntCharReal
{ int i;
char c;
double r;
}

struct IntCharReal
{ int j;
char ch;
double d;
}
Lecture 7 – Data Types CSC 331 20
 Record/structure are not exactly
Cartesian products
• Component selector: projection by component
names
struct IntCharReal x;
x.i;

• Most languages consider component names to be

part of the type.
• Thus the previous two types can be considered
different, even though they represent the same
Cartesian product.
Lecture 7 – Data Types 21
 ML: Pure Cartesian Product
type IntCharReal = int * char * real;

• (2, #"a", 3.14)

• #3(2, #"a", 3.14) = 3.14

Lecture 7 – Data Types CSC 331 22

 Union
• U  V = { x | xU or xV}
– data items with different types are stored in overlapping
region, reduce memory allocation.
– Only one type of value is valid at one time.
– E.g.,
union IntOrReal {
int i;
double r;
}

• Different from records?

Lecture 7 – Data Types CSC 331 23

 Undiscriminated Union in C
union IntOrReal {
int i;
double r;
}
union IntOrReal x;
x.i = 1;
printf(“%f\n”, x.r);

• Can be unsafe

Lecture 7 – Data Types 24

Create Discriminated Union in C++
struct IntOrReal {
bool isInt;
union {
int i;
double r;
};
};

IntOrReal x;
x.isInt = true;
x.i = 1;
...
if (x.isInt) printf(“%d\n”, x.i);
else printf(“%f\n”, x.r);

• Safe now
• or not?
Lecture 7 – Data Types 25
 Discriminated Union in Ada
• Variant record (with tag or discriminator)
type Disc is (IsInt, IsReal);
type IntOrReal (which: Disc) is
record
case which is
when IsInt => i: integer;
when IsReal => r: float;
end case;
end record;
...
x: IntOrReal := (IsReal, 2.3);
put (x.i); -- generates ERROR

• Safe: programmers won’t be able to create

inconsistent data
Lecture 7 – Data Types CSC 331 26
 Discriminated Union in Pascal
• Variant record

• Can be unsafe:
– First, the tag is optional
– Second, the tag can be set inconsistently.

Lecture 7 – Data Types 27

 Discriminated Union in ML
datatype IntOrReal =
IsInt of int | IsReal of real;

• val x = IsReal(2.3);

28
 How about Java?
• Is there record or union in java? Why?

Lecture 7 – Data Types 29

 “Union” in Java
• public abstract class A {…};
public class B extends A {…};
public class C extends A {…};

Abstract class A: union of B and C.

• Discriminated union: instanceof

Lecture 7 – Data Types 30

 Subset
• U = { v | v satisfies certain conditions and vV}

• Ada subtype

• Example 1
– type Digit_Type is range 0..9;
– subtype IntDigit_Type is integer range 0..9;

Lecture 7 – Data Types 31

 subtype in Ada
• Example 2
type Disc is (IsInt, IsReal);
type IntOrReal (which: Disc) is
record
case which is
when IsInt => i: integer;
when IsReal => r: float;
end case;
end record;

subtype IRInt is IntOrReal(IsInt);

subtype IRReal is IntOrReal(IsReal);

x: IRReal := 2.3;

Lecture 7 – Data Types 32

 Powerset
• P(U) = { U’ | U’  U }
• Example: Pascal
set of <ordinal type>

– var S: set of 1.. 10;

– var S: 1.. 10;
What’s the difference?

• The order isn’t significant though

unordered, values are distinct
Lecture 7 – Data Types 33
 set of in Pascal
var S, T: set of 1 .. 10;
S := [1, 2, 3, 5, 7];
T := [1 .. 6];

Set operations can be performed on the variables.

What are these?
• T := S*T;
• If T = [1, 3, 5] then …;
• x = 3; if x in S then …;
• if S<=T then …;

•  (*)  (+) − (-) = (=)  (<>)

•  (>)  (>=)  (<)  (<=)  (in)
Lecture 7 – Data Types 34
 Arrays and Functions
f: U → V

index type component type

• [0, …] (C/C++/Java)
• Ordinal type (Ada/Pascal)

Lecture 7 – Data Types 35

 C/C++
Allocated on stack, size statically specified.

typedef int TenIntArray [10];

typedef int IntArray [];

TenIntArray x;
int y[5];
int z[]={1,2,3,4};
IntArray w={1,2};
IntArray w; //illegal
int n = ... //from user input
int a[n]; //illegal
Lecture 7 – Data Types 36
 Java
Allocated on heap, size dynamically specified;
Size can be obtained by .length

int n = ... //from user input

int [] x = new int [n];
System.out.println(x.length);

Lecture 7 – Data Types 37

 Ada
size dynamically specified;
Set of subscripts

type IntToInt is array(integer range <>) of integer;

get(n); //from user input

x: IntToInt(1..n);
for i in x’range loop
put(x(i));
end loop;

38
 Multi-dimensional arrays
• C/C++
int x[10][20];

• Java ??
int [][] x = new int [10][20];

• Python ??

Lecture 7 – Data Types 39

 Storage
• row-major form
x[1,-10],x[1,-9],…,x[1,10],x[2,-10],…,x[2,10],x[3,-10],…

• column-major form
x[1,-10],x[2,-10],…,x[10,-10],x[1,-9],…,x[10,-9],x[1,-8],…

• C/C++
int array_max(int a[][20], int size)
• Java
int array_max(int [][][]a)

Lecture 7 – Data Types 40