UTF-8 — an ASCII compatible multi-byte Unicode encoding
The Unicode 3.0
character set occupies a 16-bit code space. The most obvious
Unicode encoding (known as UCS-2
) consists of a sequence
of 16-bit words. Such strings can contain as parts of many
16-bit characters bytes like '\0' or '/' which have a special
meaning in filenames and other C library function parameters.
In addition, the majority of UNIX tools expects ASCII files
and can't read 16-bit words as characters without major
modifications. For these reasons, UCS-2
is not a suitable
external encoding of Unicode
in filenames, text
files, environment variables, etc. The ISO 10646 Universal Character Set
(UCS), a superset of Unicode, occupies even a
31-bit code space and the obvious UCS-4
encoding for it (a
sequence of 32-bit words) has the same problems.
The UTF-8
encoding of Unicode
and UCS
does not have these
problems and is the common way in which Unicode
is used on Unix-style
operating systems.
The UTF-8
encoding has the following nice properties:
UCS
characters
0x00000000 to 0x0000007f (the classic US-ASCII
characters)
are encoded simply as bytes 0x00 to 0x7f (ASCII
compatibility). This means that files and strings
which contain only 7-bit ASCII characters have the
same encoding under both ASCII
and UTF-8
.
All UCS
characters
> 0x7f are encoded as a multi-byte sequence
consisting only of bytes in the range 0x80 to 0xfd,
so no ASCII byte can appear as part of another
character and there are no problems with, for
example, '\0' or '/'.
The lexicographic sorting order of UCS-4
strings is
preserved.
All possible 2^31 UCS codes can be encoded using
UTF-8
.
The bytes 0xfe and 0xff are never used in the
UTF-8
encoding.
The first byte of a multi-byte sequence which
represents a single non-ASCII UCS
character is always in the
range 0xc0 to 0xfd and indicates how long this
multi-byte sequence is. All further bytes in a
multi-byte sequence are in the range 0x80 to 0xbf.
This allows easy resynchronization and makes the
encoding stateless and robust against missing
bytes.
UTF-8
encoded UCS
characters
may be up to six bytes long, however the Unicode
standard
specifies no characters above 0x10ffff, so Unicode
characters can only be up to four bytes long in
UTF-8
.
The following byte sequences are used to represent a character. The sequence to be used depends on the UCS code number of the character:
0xxxxxxx
110xxxxx
10xxxxxx
1110xxxx
10xxxxxx
10xxxxxx
11110xxx
10xxxxxx
10xxxxxx
10xxxxxx
111110xx
10xxxxxx
10xxxxxx
10xxxxxx
10xxxxxx
1111110x
10xxxxxx
10xxxxxx
10xxxxxx
10xxxxxx
10xxxxxx
The xxx
bit
positions are filled with the bits of the character code
number in binary representation. Only the shortest possible
multi-byte sequence which can represent the code number of
the character can be used.
The UCS
code values
0xd800–0xdfff (UTF-16 surrogates) as well as 0xfffe
and 0xffff (UCS non-characters) should not appear in
conforming UTF-8
streams.
The Unicode
character 0xa9 = 1010 1001 (the copyright sign) is encoded
in UTF-8 as
11000010 10101001 = 0xc2 0xa9
and character 0x2260 = 0010 0010 0110 0000 (the "not equal" symbol) is encoded as:
11100010 10001001 10100000 = 0xe2 0x89 0xa0
Users have to select a UTF-8
locale, for example
with
export LANG=en_GB.UTF-8
in order to activate the UTF-8
support in
applications.
Application software that has to be aware of the used character encoding should always set the locale with for example
setlocale(LC_CTYPE, "")
and programmers can then test the expression
strcmp(nl_langinfo(CODESET), "UTF-8") == 0
to determine whether a UTF-8
locale has been
selected and whether therefore all plaintext standard input
and output, terminal communication, plaintext file content,
filenames and environment variables are encoded in
UTF-8
.
Programmers accustomed to single-byte encodings such as
US-ASCII
or
ISO 8859 have to be
aware that two assumptions made so far are no longer valid
in UTF-8
locales.
Firstly, a single byte does not necessarily correspond any
more to a single character. Secondly, since modern terminal
emulators in UTF-8
mode also support
Chinese, Japanese, and Korean double-width characters as well
as non-spacing combining
characters, outputting a single character does
not necessarily advance the cursor by one position as it
did in ASCII
. Library
functions such as mbsrtowcs(3) and
wcswidth(3) should be
used today to count characters and cursor positions.
The official ESC sequence to switch from an ISO 2022 encoding scheme (as
used for instance by VT100 terminals) to UTF-8
is ESC % G
("\x1b%G"). The corresponding return sequence from
UTF-8
to ISO 2022
is ESC % @ ("\x1b%@"). Other ISO 2022 sequences (such as
for switching the G0 and G1 sets) are not applicable in
UTF-8 mode.
It can be hoped that in the foreseeable future,
UTF-8
will
replace ASCII
and
ISO 8859 at all
levels as the common character encoding on POSIX systems,
leading to a significantly richer environment for handling
plain text.
The Unicode
and UCS
standards require
that producers of UTF-8
shall use the
shortest form possible, for example, producing a two-byte
sequence with first byte 0xc0 is non-conforming.
Unicode 3.1 has
added the requirement that conforming programs must not
accept non-shortest forms in their input. This is for
security reasons: if user input is checked for possible
security violations, a program might check only for the
ASCII
version of "/../" or
";" or NUL and overlook that there are many
non-ASCII
ways to represent
these things in a non-shortest UTF-8
encoding.
This page is part of release 2.79 of the Linux man-pages
project. A
description of the project, and information about reporting
bugs, can be found at
http://www.kernel.org/doc/man-pages/.
Copyright (C) Markus Kuhn, 1996, 2001 This is free documentation; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. The GNU General Public License's references to "object code" and "executables" are to be interpreted as the output of any document formatting or typesetting system, including intermediate and printed output. This manual is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this manual; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA. 1995-11-26 Markus Kuhn <mskuhncip.informatik.uni-erlangen.de> First version written 2001-05-11 Markus Kuhn <mgk25cl.cam.ac.uk> Update |