1 .\" Hey Emacs! This file is -*- nroff -*- source.
3 .\" Copyright (C) Markus Kuhn, 1996, 2001
5 .\" This is free documentation; you can redistribute it and/or
6 .\" modify it under the terms of the GNU General Public License as
7 .\" published by the Free Software Foundation; either version 2 of
8 .\" the License, or (at your option) any later version.
10 .\" The GNU General Public License's references to "object code"
11 .\" and "executables" are to be interpreted as the output of any
12 .\" document formatting or typesetting system, including
13 .\" intermediate and printed output.
15 .\" This manual is distributed in the hope that it will be useful,
16 .\" but WITHOUT ANY WARRANTY; without even the implied warranty of
17 .\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 .\" GNU General Public License for more details.
20 .\" You should have received a copy of the GNU General Public
21 .\" License along with this manual; if not, write to the Free
22 .\" Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111,
25 .\" 1995-11-26 Markus Kuhn <mskuhn@cip.informatik.uni-erlangen.de>
26 .\" First version written
27 .\" 2001-05-11 Markus Kuhn <mgk25@cl.cam.ac.uk>
30 .TH UTF-8 7 2001-05-11 "GNU" "Linux Programmer's Manual"
32 UTF-8 \- an ASCII compatible multibyte Unicode encoding
36 character set occupies a 16-bit code space.
38 Unicode encoding (known as
40 consists of a sequence of 16-bit words.
41 Such strings can contain as
42 parts of many 16-bit characters bytes
43 like \(aq\\0\(aq or \(aq/\(aq which have a
44 special meaning in filenames and other C library function arguments.
45 In addition, the majority of Unix tools expects ASCII files and can't
46 read 16-bit words as characters without major modifications.
49 is not a suitable external encoding of
51 in filenames, text files, environment variables, etc.
53 .BR "ISO 10646 Universal Character Set (UCS)" ,
54 a superset of Unicode, occupies even a 31-bit code space and the obvious
56 encoding for it (a sequence of 32-bit words) has the same problems.
64 does not have these problems and is the common way in which
66 is used on Unix-style operating systems.
70 encoding has the following nice properties:
74 characters 0x00000000 to 0x0000007f (the classic
76 characters) are encoded simply as bytes 0x00 to 0x7f (ASCII
78 This means that files and strings which contain only
79 7-bit ASCII characters have the same encoding under both
87 characters greater than 0x7f are encoded as a multibyte sequence
88 consisting only of bytes in the range 0x80 to 0xfd, so no ASCII
89 byte can appear as part of another character and there are no
90 problems with, for example, \(aq\\0\(aq or \(aq/\(aq.
93 The lexicographic sorting order of
98 All possible 2^31 UCS codes can be encoded using
102 The bytes 0xfe and 0xff are never used in the
107 The first byte of a multibyte sequence which represents a single non-ASCII
109 character is always in the range 0xc0 to 0xfd and indicates how long
110 this multibyte sequence is.
111 All further bytes in a multibyte sequence
112 are in the range 0x80 to 0xbf.
113 This allows easy resynchronization and
114 makes the encoding stateless and robust against missing bytes.
120 characters may be up to six bytes long, however the
122 standard specifies no characters above 0x10ffff, so Unicode characters
123 can only be up to four bytes long in
126 The following byte sequences are used to represent a character.
127 The sequence to be used depends on the UCS code number of the character:
129 0x00000000 \- 0x0000007F:
132 0x00000080 \- 0x000007FF:
136 0x00000800 \- 0x0000FFFF:
141 0x00010000 \- 0x001FFFFF:
147 0x00200000 \- 0x03FFFFFF:
154 0x04000000 \- 0x7FFFFFFF:
164 bit positions are filled with the bits of the character code number in
165 binary representation.
166 Only the shortest possible multibyte sequence
167 which can represent the code number of the character can be used.
171 code values 0xd800\(en0xdfff (UTF-16 surrogates) as well as 0xfffe and
172 0xffff (UCS noncharacters) should not appear in conforming
178 character 0xa9 = 1010 1001 (the copyright sign) is encoded
182 11000010 10101001 = 0xc2 0xa9
185 and character 0x2260 = 0010 0010 0110 0000 (the "not equal" symbol) is
189 11100010 10001001 10100000 = 0xe2 0x89 0xa0
191 .SS "Application Notes"
192 Users have to select a
194 locale, for example with
197 export LANG=en_GB.UTF-8
200 in order to activate the
202 support in applications.
204 Application software that has to be aware of the used character
205 encoding should always set the locale with for example
208 setlocale(LC_CTYPE, "")
211 and programmers can then test the expression
214 strcmp(nl_langinfo(CODESET), "UTF-8") == 0
217 to determine whether a
219 locale has been selected and whether
220 therefore all plaintext standard input and output, terminal
221 communication, plaintext file content, filenames and environment
222 variables are encoded in
225 Programmers accustomed to single-byte encodings such as
229 have to be aware that two assumptions made so far are no longer valid
233 Firstly, a single byte does not necessarily correspond any
234 more to a single character.
235 Secondly, since modern terminal emulators
238 mode also support Chinese, Japanese, and Korean
239 .B double-width characters
240 as well as nonspacing
241 .BR "combining characters" ,
242 outputting a single character does not necessarily advance the cursor
243 by one position as it did in
245 Library functions such as
249 should be used today to count characters and cursor positions.
251 The official ESC sequence to switch from an
253 encoding scheme (as used for instance by VT100 terminals) to
257 The corresponding return sequence from
259 to ISO 2022 is ESC % @ ("\\x1b%@").
260 Other ISO 2022 sequences (such as
261 for switching the G0 and G1 sets) are not applicable in UTF-8 mode.
263 It can be hoped that in the foreseeable future,
269 at all levels as the common character encoding on POSIX systems,
270 leading to a significantly richer environment for handling plain text.
273 .BR Unicode " and " UCS
274 standards require that producers of
276 shall use the shortest form possible, for example, producing a two-byte
277 sequence with first byte 0xc0 is nonconforming.
279 has added the requirement that conforming programs must not accept
280 non-shortest forms in their input.
281 This is for security reasons: if
282 user input is checked for possible security violations, a program
283 might check only for the
285 version of "/../" or ";" or NUL and overlook that there are many
287 ways to represent these things in a non-shortest
291 ISO/IEC 10646-1:2000, Unicode 3.1, RFC\ 2279, Plan 9.
293 .\" Markus Kuhn <mgk25@cl.cam.ac.uk>