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 multi-byte Unicode encoding
36 character set occupies a 16-bit code space. The most obvious
37 Unicode encoding (known as
39 consists of a sequence of 16-bit words. Such strings can contain as
40 parts of many 16-bit characters bytes like '\\0' or '/' which have a
41 special meaning in filenames and other C library function parameters.
42 In addition, the majority of UNIX tools expects ASCII files and can't
43 read 16-bit words as characters without major modifications. For these
46 is not a suitable external encoding of
48 in filenames, text files, environment variables, etc. The
49 .BR "ISO 10646 Universal Character Set (UCS)" ,
50 a superset of Unicode, occupies even a 31-bit code space and the obvious
52 encoding for it (a sequence of 32-bit words) has the same problems.
60 does not have these problems and is the common way in which
62 is used on Unix-style operating systems.
66 encoding has the following nice properties:
70 characters 0x00000000 to 0x0000007f (the classic
72 characters) are encoded simply as bytes 0x00 to 0x7f (ASCII
73 compatibility). This means that files and strings which contain only
74 7-bit ASCII characters have the same encoding under both
82 characters > 0x7f are encoded as a multi-byte sequence
83 consisting only of bytes in the range 0x80 to 0xfd, so no ASCII
84 byte can appear as part of another character and there are no
85 problems with e.g. '\\0' or '/'.
88 The lexicographic sorting order of
93 All possible 2^31 UCS codes can be encoded using
97 The bytes 0xfe and 0xff are never used in the
102 The first byte of a multi-byte sequence which represents a single non-ASCII
104 character is always in the range 0xc0 to 0xfd and indicates how long
105 this multi-byte sequence is. All further bytes in a multi-byte sequence
106 are in the range 0x80 to 0xbf. This allows easy resynchronization and
107 makes the encoding stateless and robust against missing bytes.
113 characters may be up to six bytes long, however the
115 standard specifies no characters above 0x10ffff, so Unicode characters
116 can only be up to four bytes long in
119 The following byte sequences are used to represent a character. The
120 sequence to be used depends on the UCS code number of the character:
122 0x00000000 \- 0x0000007F:
125 0x00000080 \- 0x000007FF:
129 0x00000800 \- 0x0000FFFF:
134 0x00010000 \- 0x001FFFFF:
140 0x00200000 \- 0x03FFFFFF:
147 0x04000000 \- 0x7FFFFFFF:
157 bit positions are filled with the bits of the character code number in
158 binary representation. Only the shortest possible multi-byte sequence
159 which can represent the code number of the character can be used.
163 code values 0xd800\(en0xdfff (UTF-16 surrogates) as well as 0xfffe and
164 0xffff (UCS non-characters) should not appear in conforming
170 character 0xa9 = 1010 1001 (the copyright sign) is encoded
174 11000010 10101001 = 0xc2 0xa9
177 and character 0x2260 = 0010 0010 0110 0000 (the "not equal" symbol) is
181 11100010 10001001 10100000 = 0xe2 0x89 0xa0
183 .SH "APPLICATION NOTES"
184 Users have to select a
186 locale, for example with
189 export LANG=en_GB.UTF-8
192 in order to activate the
194 support in applications.
196 Application software that has to be aware of the used character
197 encoding should always set the locale with for example
200 setlocale(LC_CTYPE, "")
203 and programmers can then test the expression
206 strcmp(nl_langinfo(CODESET), "UTF-8") == 0
209 to determine whether a
211 locale has been selected and whether
212 therefore all plaintext standard input and output, terminal
213 communication, plaintext file content, filenames and environment
214 variables are encoded in
217 Programmers accustomed to single-byte encodings such as
221 have to be aware that two assumptions made so far are no longer valid
224 locales. Firstly, a single byte does not necessarily correspond any
225 more to a single character. Secondly, since modern terminal emulators
228 mode also support Chinese, Japanese, and Korean
229 .B double-width characters
230 as well as non-spacing
231 .BR "combining characters" ,
232 outputting a single character does not necessarily advance the cursor
233 by one position as it did in
235 Library functions such as
239 should be used today to count characters and cursor positions.
241 The official ESC sequence to switch from an
243 encoding scheme (as used for instance by VT100 terminals) to
246 ("\\x1b%G"). The corresponding return sequence from
248 to ISO 2022 is ESC % @ ("\\x1b%@"). Other ISO 2022 sequences (such as
249 for switching the G0 and G1 sets) are not applicable in UTF-8 mode.
251 It can be hoped that in the foreseeable future,
257 at all levels as the common character encoding on POSIX systems,
258 leading to a significantly richer environment for handling plain text.
261 .BR Unicode " and " UCS
262 standards require that producers of
264 shall use the shortest form possible, e.g., producing a two-byte
265 sequence with first byte 0xc0 is non-conforming.
267 has added the requirement that conforming programs must not accept
268 non-shortest forms in their input. This is for security reasons: if
269 user input is checked for possible security violations, a program
270 might check only for the
272 version of "/../" or ";" or NUL and overlook that there are many
274 ways to represent these things in a non-shortest
278 ISO/IEC 10646-1:2000, Unicode 3.1, RFC\ 2279, Plan 9.
280 Markus Kuhn <mgk25@cl.cam.ac.uk>