ASCII

ASCII (American Standard Code for Information Interchange, generally pronounced /"{ski/ (X-SAMPA)) is a character set and a character encoding based on the Roman alphabet as used in modern English and other Western European languages. It is most commonly used by computers and other communication equipment to represent text and by control devices that work with text.

Overview

Like other character representation computer codes, ASCII specifies a correspondence between digital bit patterns and the symbols/glyphs of a written language, thus allowing digital devices to communicate with each other and to process, store, and communicate character-oriented information. The ASCII character encoding, or a compatible extension (see below), is used on nearly all common computers, especially personal computers and workstations. The preferred MIME name for this encoding is "US-ASCII".

ASCII is, strictly, a seven-bit code, meaning that it uses the bit patterns representable with seven binary digits (a range of 0 to 127 decimal) to represent character information. At the time ASCII was introduced, many computers dealt with eight-bit groups (bytes or, more specifically, octets) as the smallest unit of information; the eighth bit was commonly used as a parity bit for error checking on communication lines or other device-specific functions. Machines which did not use parity typically set the eighth bit to zero, though some systems such as PRIME machines running PRIMOS set the eighth bit of ASCII characters to one.

ASCII does not specify any way to include information about the conceptual structure or appearance of a piece of text. That requires other standards, such as those specifying markup languages. Conceptual structure can be included using XML and appearance can be specified by using HTML for relatively simple things, SGML for more complex things, or PostScript, Display PostScript, or TeX for advanced layout and font control.

ASCII was first published as a standard in 1963 by the American Standards Association (ASA), which later became ANSI. There are many variations of ASCII, but its present, most widely-used form is ANSI X3.4-1986, also standardized as ECMA-6, ISO/IEC 646:1991 International Reference Version, ITU-T Recommendation T.50 (09/92), and RFC 20. It is embedded in its probable replacement, Unicode, as the 'lowest' 128 characters. ASCII is considered by some the most successful software standard ever promulgated.

Historically, ASCII developed from telegraphic codes and its first commercial use was as a 7-bit teleprinter code promoted by Bell data services. The Bell System had been planning to use a 6-bit code derived from Fieldata that added punctuation and lower-case letter to the earlier 5-bit Baudot teleprinter code but was persuaded to instead join the ASA subcommittee that was developing ASCII. Baudot helped in the automation of sending and receiving of telegraphic messages, and took many features from Morse code; it was however, a constant length code unlike Morse code. Compared to earlier telegraph codes, the proposed Bell code and ASCII were both reordered for more convenient sorting (ie, alphabetization) of lists, and added features for devices other than teleprinters. Some ASCII features, including the 'ESCape sequence', were due to Bob Bemer.

ASCII control characters

The first thirty-two codes (numbers 0-31 decimal) in ASCII are reserved for control characters: codes that were not originally intended to carry information, but rather to control devices (such as printers) that make use of ASCII. For example, character 10 represents the "line feed" function (which causes a printer to advance its paper), and character 27 represents the "escape" key found on the top left of common keyboards.

Code 127 (all seven bits on) is another special character known as "delete" or "rubout". Though its function is similar to that of other control characters, this pattern was used so that it could be used to "erase" a section of paper tape, a popular storage medium until the 80's, by punching all possible holes at a particular character position.

Many of the ASCII control codes are to mark data packets, or to control a data transmission protocol (i.e., ENQuiry (effectively, "any stations out there?"), ACKnowledge, Negative AcKnowledge, Start Of Header, Start Of Text, End Of Text, etc). ESCape and SUBstitute permit a communications protocol to, for instance, mark binary data so that if it contains codes with the same pattern as a protocol character, the code will be processed as data.

The separator characters (Record Separator, etc.) were intended for use with magnetic tape systems.

XON and XOFF are common interpretations of two of the Device Control characters and are generally used to throttle data flow to a slow device, such as a printer, from a fast device, such as a computer so data does not overrun and be lost.

Early users of ASCII adopted some of the control codes to represent "meta-information" such as end-of-line, start/end of a data element, and so on. These assignments often conflict, so part of the effort in converting data from one format to another is making the correct meta-information transformations. For example, the character(s) representing end-of-line ("new line") in text data files/streams vary from operating system to operating system. When moving files from one system to another, these characters must be recognized as end-of-line markers and converted appropriately.

Binary	Dec	Hex	Abbrev	Rep	Access	Name
0000 0000	0	00	NUL	?	^@	Null character
0000 0001	1	01	SOH	?	^A	Start of Header
0000 0010	2	02	STX	?	^B	Start of Text
0000 0011	3	03	ETX	?	^C	End of Text
0000 0100	4	04	EOT	?	^D	End of Trans
0000 0101	5	05	ENQ	?	^E	Enquiry
0000 0110	6	06	ACK	?	^F	Acknowledgement
0000 0111	7	07	BEL	?	^G	Bell
0000 1000	8	08	BS	?	^H	Backspace
0000 1001	9	09	HT	?	^I	Horizontal Tab
0000 1010	10	0A	LF	?	^J	Line feed
0000 1011	11	0B	VT	?	^K	Vertical Tab
0000 1100	12	0C	FF	?	^L	Form Feed
0000 1101	13	0D	CR	?	^M	Carriage return
0000 1110	14	0E	SO	?	^N	Shift Out
0000 1111	15	0F	SI	?	^O	Shift In
0001 0000	16	10	DLE	?	^P	Data Link Escape
0001 0001	17	11	DC1	?	^Q	Device Control 1 oft. XON
0001 0010	18	12	DC2	?	^R	Device Control 2
0001 0011	19	13	DC3	?	^S	Device Control 3 oft. XOFF
0001 0100	20	14	DC4	?	^T	Device Control 4
0001 0101	21	15	NAK	?	^U	Negative Acknowledge
0001 0110	22	16	SYN	?	^V	Synchronous Idle
0001 0111	23	17	ETB	?	^W	End of Trans. Block
0001 1000	24	18	CAN	?	^X	Cancel
0001 1001	25	19	EM	?	^Y	End of Medium
0001 1010	26	1A	SUB	?	^Z	Substitute
0001 1011	27	1B	ESC	?	^[ or ESC	Escape
0001 1100	28	1C	FS	?	^\	File Separator
0001 1101	29	1D	GS	?	^[	Group Separator
0001 1110	30	1E	RS	?	^^	Record Separator
0001 1111	31	1F	US	?	^_	Unit Separator
0111 1111	127	7F	DEL	?	DEL or Bks	Delete

In the table above, the fifth column contains glyphs reserved for representing control codes in a data stream, ie, when they must be printed or displayed rather than (or in addition to) causing action; your browser, (i.e., your HTML user agent) may require the installation of additional fonts in order to display them.

The sixth column shows the key combinations traditionally used to input control characters from a keyboard. In this column, a caret ("^") represents the "Control"/"Ctrl" key, which must be held down while pressing the next key, e.g. "^Z" means to hold down Ctrl while pressing the Z key. This notation is also sometimes used by software as a printable representation of control characters that could not be processed.

In some systems on the Internet, there is a history of the DEL control code being converted to BS in transit to a remote server. If the code was received in a text editor that did not know what do with it, the result was often "^H" appearing where the user intended to delete previous characters. "^H" persists in messages today as a deliberate humorous device, e.g. "there's a sucker^H^H^H^H^H^H^H potential customer born every minute".

ASCII printable characters

Code 32 is the "space" character, denoting the space between words, which is produced by the large space bar of a keyboard. Codes 33 to 126 are called the printable characters, which represent letters, digits, punctuation marks, and a few miscellaneous symbols.

Seven bit ASCII provided seven "national" characters and, if the combined hardware and software permit, can use overstrikes to simulate some additional international characters: a BackSpace can be followed with the grave accent (which the American and British standards, but only the American and British standards, also call "opening single quotation mark"), a tilde, or a breath mark (inverted vel).

Binary Dec Hex *G 0010 0000 32 20 (?) 0010 0001 33 21 ! 0010 0010 34 22 " 0010 0011 35 23 # 0010 0100 36 24 $ 0010 0101 37 25 % 0010 0110 38 26 & 0010 0111 39 27 ' 0010 1000 40 28 ( 0010 1001 41 29 ) 0010 1010 42 2A * 0010 1011 43 2B + 0010 1100 44 2C , 0010 1101 45 2D - 0010 1110 46 2E . 0010 1111 47 2F / 0011 0000 48 30 0 0011 0001 49 31 1 0011 0010 50 32 2 0011 0011 51 33 3 0011 0100 52 34 4 0011 0101 53 35 5 0011 0110 54 36 6 0011 0111 55 37 7 0011 1000 56 38 8 0011 1001 57 39 9 0011 1010 58 3A : 0011 1011 59 3B ; 0011 1100 60 3C < 0011 1101 61 3D = 0011 1110 62 3E > 0011 1111 63 3F ?

Binary Dec Hex *G 0100 0000 64 40 @ 0100 0001 65 41 A 0100 0010 66 42 B 0100 0011 67 43 C 0100 0100 68 44 D 0100 0101 69 45 E 0100 0110 70 46 F 0100 0111 71 47 G 0100 1000 72 48 H 0100 1001 73 49 I 0100 1010 74 4A J 0100 1011 75 4B K 0100 1100 76 4C L 0100 1101 77 4D M 0100 1110 78 4E N 0100 1111 79 4F O 0101 0000 80 50 P 0101 0001 81 51 Q 0101 0010 82 52 R 0101 0011 83 53 S 0101 0100 84 54 T 0101 0101 85 55 U 0101 0110 86 56 V 0101 0111 87 57 W 0101 1000 88 58 X 0101 1001 89 59 Y 0101 1010 90 5A Z 0101 1011 91 5B [ 0101 1100 92 5C \ 0101 1101 93 5D ] 0101 1110 94 5E ^ 0101 1111 95 5F _

Binary Dec Hex *G 0110 0000 96 60 ` 0110 0001 97 61 a 0110 0010 98 62 b 0110 0011 99 63 c 0110 0100 100 64 d 0110 0101 101 65 e 0110 0110 102 66 f 0110 0111 103 67 g 0110 1000 104 68 h 0110 1001 105 69 i 0110 1010 106 6A j 0110 1011 107 6B k 0110 1100 108 6C l 0110 1101 109 6D m 0110 1110 110 6E n 0110 1111 111 6F o 0111 0000 112 70 p 0111 0001 113 71 q 0111 0010 114 72 r 0111 0011 115 73 s 0111 0100 116 74 t 0111 0101 117 75 u 0111 0110 118 76 v 0111 0111 119 77 w 0111 1000 120 78 x 0111 1001 121 79 y 0111 1010 122 7A z 0111 1011 123 7B { 0111 1100 124 7C | 0111 1101 125 7D } 0111 1110 126 7E ~

*G=Graphic
Note how uppercase characters can be converted to lowercase by adding 32 to their ASCII value; in binary, this can be accomplished simply by setting the sixth-least significant bit to 1.

Aliases for ASCII

RFC 1345, published in June 1992, and the IANA registry of character sets (http://www.iana.org/assignments/character-sets), ongoing, recognize the following case-insensitive aliases for ASCII as being suitable for use on the Internet:

• ANSI_X3.4-1968 (canonical name)
• ANSI_X3.4-1986
• ASCII
• US-ASCII (preferred MIME name)
• us
• ISO646-US
• ISO_646.irv:1991
• iso-ir-6
• IBM367
• cp367
• csASCII

Of these, only the aliases US-ASCII and ASCII are widely used. They are often found in the optional "charset" parameter in the Content-Type header of some MIME messages, in the equivalent "meta" element of some HTML documents, and in the encoding declaration part of the prolog of some XML documents.

Variants of ASCII

As computer technology spread throughout the world, many variations of ASCII were developed by corporations and standards bodies in order to facilitate the expression of non-English languages that still used Roman-based alphabets. Some of these variations can be considered to be ASCII extensions, although the term is sometimes misapplied to cover all variants, including those that don't preserve ASCII's character map in the 7-bit range.

ISO 646 (1972) was the first attempt to remedy the English bias, although it created compatibility problems, since it was still a seven-bit character set. No additional codes were available, so some were re-assigned in language-specific variants. It was thus impossible to know what character was represented by a code without knowing what variant was in use, and text processing systems were generally able to cope with only one variant, anyway.

Eventually, improved technology brought out-of-band means to represent the information formerly encoded in the eighth bit of each byte, freeing this bit to add another 128 additional character codes for new assignments. For example, IBM developed eight-bit code pages, such as code page 437, which replaced the control characters with graphic symbols such as smiley faces, and mapped additional graphic characters to the upper 128 bytes. These code pages were supported in hardware by IBM PC manufacturers, as well as in operating systems such as MS-DOS.

Eight-bit standards such as ISO/IEC 8859 were true extensions of ASCII, leaving the original character mapping intact and just adding additional values above the 7-bit range. This enabled a broader range of languages to be represented, but these standards were still plagued with incompatibilities and limitations. Still, ISO/IEC 8859-1 and original 7-bit ASCII are the most common character encodings in use today.

Unicode and ISO/IEC 10646: the Universal Character Set, have a much wider array of characters, and their various encoding forms are rapidly supplanting ISO/IEC 8859 and ASCII in many environments. While ASCII is defined in terms of 7-bit codes, Unicode and the UCS are defined in terms of relatively abstract "code points": non-negative integer numbers that can be mapped, using different encoding forms and schemes, to sequences of 1 or more 8-bit bytes. To permit backward compatibility, Unicode and the UCS assign the first 128 code points to the same characters as ASCII. ASCII can therefore be thought of as being a 7-bit encoding scheme for a very small subset of Unicode and the UCS. The popular UTF-8 encoding form prescribes the use of one to four 8-bit code values for each code point character, and happens to be identical to ASCII for the code values below 128. Other encoding forms such as UTF-16 resemble ASCII in how they represent the first 128 characters of Unicode, but tend to use 16 or 32 bits per character, so they are not entirely compatible without conversions.

The portmanteau word ASCIIbetical has evolved to describe the collation of data in ASCII code order rather than "standard" alphabetical order (which requires some tricky computation, and varies with language).

ASCII contains many characters which were not commonly used, or at least spoken of, outside of the computing context; the "popularization" of these characters required that names be agreed upon for them. See the pronunciation guide in the external links, below.

ASCIIZ or ASCIZ is an adjective used to refer to a null-terminated ASCII string.

External links

• Standard ECMA-6: 7-bit Coded Character Set 6th edition (December 1991) (http://www.ecma-international.org/publications/standards/Ecma-006.htm)
• IANA Registry: C0 Set of ISO 646 (http://www.itscj.ipsj.or.jp/ISO-IR/001.pdf)
• IANA Registry: ISO 646, USA Version X3.4 - 1968 (http://www.itscj.ipsj.or.jp/ISO-IR/006.pdf)
• ASCII: American Standard Code for Information Infiltration by Tom Jennings (http://www.wps.com/projects/codes/)
• Jargon File: ASCIIbetical (http://www.catb.org/~esr/jargon/html/A/ASCIIbetical-order.html)
• Bob Bemer's home page (http://www.bobbemer.com) some historical notes about ASCII from one of its designers
• A pronunciation guide (http://www.speech.cs.cmu.edu/~sburke/stuff/pronunciation-guide.txt) for ASCII characters (some are more whimsical than others; see especially the end of the list)

---

ASCII is also a name of one of the oldest and most prestigious computer magazines published in Japan. See ASCII (magazine)