Back to DFS's Pascal Page

Data Types in Turbo Pascal

(Preliminary)

Follow these fast links:

• Basic Data Types
  • Char
    • ASCII
    • EBCDIC
  • Boolean
  • Integer
  • Longint
  • Real
  • String
• Programmer-Defined Data Types
  • Enumerated Type
  • Record
• Binary and Other Numeration Systems

Back to Top

Char

The Char data type is used for individual 8-bit codes, many of which are used to represent printable characters. Each Char requires one byte of memory and can have a value from 0 to 255 (or 0 to 11111111 in binary). This data type, along with Integer and Boolean, belongs to the class of data types known as ordinals. An ordinal is one of a set of values, each of which has both a predecessor and a successor except for the first and the last. In Pascal, a Char literal is delimited by single quotes, e.g., 'C'.

A String can be viewed as an array of Chars.

Many different encoding schemes have been used to represent characters on computers. Below are two of the most widely used, ASCII and EBCDIC.

Back to Top

ASCII

The original American Standard Code for Information Interchange (ASCII) used only 7 of the bits, with values from 0₁₀ to 127₁₀ or 0₁₆ to 7F₁₆. The first 32 codes and the last one are for controling communication and peripherals. The other 95 codes are for printable characters. They were intended to handle English only and they do not even do that well. Note that the hexadecimal value for a character is composed of the column value as the high-order digit and the row value as the low-order digit, e.g., the value for lowercase 'a' is 61₁₆. The space character is 20₁₆. Click here to see what the control code mnemonics for the codes 00₁₆ to 1F₁₆ stand for.

Back to Top

EBCDIC

An earlier, but now rarely used, encoding system was the Extended Binary Coded Decimal Interchange Code (EBCDIC), which was used on IBM mainframes. It reflected the system of holes used on punch cards. It has fallen into disfavor partially because of the use of nonsequential codes for the alphabetic characters. Note that the hexadecimal value for a character is composed of the row value as the high-order digit and the column value as the low-order digit, e.g., the value for lowercase 'a' is 81₁₆. The space character is 40₁₆. The nonindented sections of the table indicate unused codes. Click here to see what the control code mnemonics for the codes 00₁₆ to 3F₁₆ stand for.

Back to Top

Boolean

A variable of the Boolean type can have only two values, TRUE and FALSE. A good compiler, therefore, can use a single bit for a boolean variable, packing up to eight such variables into a single byte of memory. Zero (0) is used to represent FALSE and one (1) is used for TRUE.

A boolean is one of the ordinal types. Internally, FALSE is treated like 0 and TRUE like 1. So TRUE is the successor of FALSE and FALSE is the predecessor of TRUE.

The result of an evaluation of a Boolean expression can be assigned to a Boolean variable, e.g.,

boo : Boolean;

boo := 5 < 4;
writeln(boo);

will result in the word FALSE being printed.

Booleans are typically used as flags to store what a current condition is. This can be seen in a reworking of the program on the User-Controlled Repetition with the Repeat-Until Loop page.

QuitProgram : boolean; {Loop control variable}

Repeat
   .
   .
   .
   {-------------------User Control------------------}
   write('Do you want to do another? (Y/N) ');
   readln(response);
   if (upcase(response) = 'N') then
      QuitProgram := TRUE
Until QuitProgram

Back to Top

Integer

An Integer is a two-byte ordinal data type. The range possible is from -32768 to 32767. The largest value is assigned to the Pascal constant MaxInt. In the binary representation, the highest-order bit of the 16-bit value is known as the sign bit -- if it is a one (1), the value is negative.

A error which frequently occurs in programming is the choice of a data type which is too small to hold a particular value. Consider the following code, for example.

x : Integer;

x := MaxInt;
x := x + 1;
writeln(x)

This code will print the value -32768. This is known as overflow. Examine the following table.

The following binary addition demonstrates how the value stored in x became negative.

 0111 1111 1111 1111
                  +1
--------------------
 1000 0000 0000 0000

The result of the addition is negative, as can be seen by the highest-order bit being a 1. The resultant value is the binary representation of the smallest value that can be stored in an Integer variable.

Back to Top

Longint

An LongInt is a four-byte ordinal data type, making it twice as long as an Integer in terms of bits. The range possible is from -2147483648 to 2147483647. The largest value is assigned to the Pascal constant MaxLongInt. In the binary representation, the highest-order bit of the 32-bit value is known as the sign bit -- if it is a one (1), the value is negative.

x : LongInt;

x := MaxLongInt;
x := x + 1;
writeln(x)

This code will print the value -2147483648. This is known as overflow. The following table shows the smallest and largest values in decimal, binary and hexadecimal that can be stored in a LongInt.

Back to Top

Real

A Real is a six-byte non-ordinal data type. This means that it cannot be used as the loop control variable in a for-to-do loop.

The range of a positive Real variable is from 2.9 x 10^-39 to 1.7 x 10³⁸; approximately the same holds for negative values. A Real maintains about 11 significant digits.

Consider the following code fragment:

x : Real;

x := 13.2;
writeln(x);

This will result in 1.3200000000E+01 being printed. This is know as E-notation and is equivalent to 1.32 x 10¹ in scientific notation.

Since normal people find reading such notation bothersome, Pascal allows you to specify the number of decimal places to be used in the output. For example, the following

writeln(x:1:2)

would print 13.20 as output. The value after the first colon specifies the field width and the second specifies the number of decimal places to be used. If the first value is less than the total length of the string being printed, the string is printed quad left (left justified); if the first value is more than the total length, the string is printed quad right (right justified) within the field. Thus, decimal values such as those used for money can be made to line up:

x : Real;
y : Real;

x := 13.2;
y := 11;
writeln(x:7:2);
writeln(y:7:2);

would result in the following output, where the carets are used to indicate wordspaces.

^^13.20
^^11.00

Back to Top

Back to ASCII

ASCII CONTROL CODES
00	NUL	NULl	10	DLE	Data Link Escape
01	SOH	Start Of Heading	11	DC1	Device Control 1
02	STX	end of heading or Start of TeXt	12	DC2	Device Control 2
03	ETX	End of TeXt	13	DC3	Device Control 3
04	EOT	End Of Transmission	14	DC4	Device Control 4
05	ENQ	ENQuiry (to request identification)	15	NAK	Negative AcKnowledge
06	ACK	ACKnowledge	16	SYN	SYNchronouos idle
07	BEL	ring BELl	17	ETB	End of Transmission Block
08	BS	BackSpace	18	CAN	CANcel previous transmission
09	HT	Horizontal Tab	19	EM	End of Medium
0A	LF	Line Feed	1A	SUB	SUBstitute a character for another
0B	VT	Vertical Tab	1B	ESC	ESCape
0C	FF	Form Feed	1C	FS	File Separator
0D	CR	Carriage Return	1D	GS	Group Separator
0E	SO	Shift Out (begin non-ASCII bit string)	1E	RS	Record Separator
0F	SI	Shift In (end non-ASCII bit string)	1F	US	Unit Separator

Back to Top

Back to EBCDIC

EBCDIC CONTROL CODES
00	NUL	NULl	10	DLE	Data Link Escape	20	DS	Digit Select	30
01	SOH	Start Of Heading	11	DC1	Device Control 1	21	SOS	Start Of Significance	31
02	STX	end of heading or Start of TeXt	12	DC2	Device Control 2	22	FS	File Separator	32	SYN	SYNchronous idle
03	ETX	End of TeXt	13	DC3	Device Control 3	23			33
04	PF	Punch ofF	14	RES	REStore	24	BYP	BYPass	34	PN	Punch oN
05	HT	Horizontal Tab	15	NL	NewLine	25	LF	Line Feed	35	RS	Record Separator
06	LC	Lower Case	16	BS	BackSpace	26	ETB	End of Transmission Block	36	UC	Upper Case
07	DEL	DELete	17	IL	IdLe	27	ESC	ESCape	37	EOT	End Of Transmission
08			18	CAN	CANcel previous transmission	28			38
09			19	EM	End of Medium	29			39
0A	SMM	repeat	1A	CC	unit backspace	2A	SM	Start Message	3A
0B	VT	Vertical Tab	1B			2B			3B
0C	FF	Form Feed	1C	IFS	Interchange File Separator	2C			3C	DC4	Device Control 4
0D	CR	Carriage Return	1D	IGS	Interchange Group Separator	2D	ENQ	ENQuiry (to request identification)	3D	NAK	Negative AcKnowledge
0E	SO	Shift Out (begin non-EBCDIC bit string)	1E	IRS	Interchange Record Separator	2E	ACK	ACKnowledge	3E
0F	SI	Shift In (end non-EBCDIC bit string)	1F	IUS	Interchange Unit Separator	2F	BEL	ring BELl	3F	SUB	SUBstitute a character for another