.\"	$OpenBSD: utf8.7,v 1.9 2022/02/18 10:24:32 jsg Exp $
.\"
.\" Copyright (c) 2017 Ted Unangst <tedu@openbsd.org>
.\"
.\" Permission to use, copy, modify, and distribute this software for any
.\" purpose with or without fee is hereby granted, provided that the above
.\" copyright notice and this permission notice appear in all copies.
.\"
.\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
.\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
.\"
.Dd $Mdocdate: February 18 2022 $
.Dt UTF8 7
.Os
.Sh NAME
.Nm utf8
.Nd UTF-8 text encoding
.Sh DESCRIPTION
UTF-8 is a multibyte character encoding for Unicode text.
It is the preferred format for non ASCII text.
.Pp
Unicode codepoints are encoded as follows:
.Bl -tag -width Ds
.It U+0000 \(en U+007F:
One byte: 0....... (compatible with ASCII)
.It U+0080 \(en U+07FF:
Two bytes: 110..... 10......
.It U+0800 \(en U+D7FF and U+E000 \(en U+FFFF:
Three bytes: 1110.... 10...... 10......
.It U+10000 \(en U+10FFFF:
Four bytes: 11110... 10...... 10...... 10......
.El
.Pp
The bits shown as dots contain the codepoint represented as a binary
integer.
.Pp
Bytes starting with the bit pattern 11...... are called UTF-8 start
bytes, and those starting with 10...... UTF-8 continuation bytes.
The number of leading 1 bits in a start byte indicates the total
number of bytes used to encode the codepoint, including the start
byte.
.Pp
Encodings using more bytes than required are invalid.
In particular, 11000000 and 11000001 are not valid start bytes,
the byte after 11100000 must be at least 10100000,
and the byte after 11110000 must be at least 10010000.
.Pp
The ranges U+D800 to U+DFFF and U+110000 to U+1FFFFF
do not contain valid Unicode codepoints.
Consequently, the corresponding three- and four-byte UTF-8 sequences
are invalid.
The highest valid byte after 11101101 is 10011111,
the highest valid byte of the form 1111.... is 11110100,
and the highest valid byte after 11110100 is 10001111.
.Pp
To summarize, the following is a complete list of bytes
that are invalid in all contexts:
.Pp
.Bl -tag -width 5n -offset 4n -compact
.It c0\(enc1
two-byte sequence that has to be encoded as a single byte
.It f5\(enf7
four-byte sequence beyond the Unicode range
.It f8\(enff
invalid sequence of five or more bytes
.El
.Pp
The following is a complete list of invalid two-byte combinations
of the form 11...... 10...... that consist of two valid bytes:
.Pp
.Bl -tag -width 9n -offset 4n -compact
.It e080\(ene09f
three-byte sequence that has to be encoded as two bytes
.It eda0\(enedbf
start of a UTF-16 surrogate, which is not valid UTF-8
.It f080\(enf08f
four-byte sequence that has to be encoded as three bytes
.It f490\(enf4bf
four-byte sequence beyond the Unicode range
.El
.Sh SEE ALSO
.Xr locale 1 ,
.Xr ascii 7
.Sh STANDARDS
.Rs
.%A F. Yergeau
.%D November 2003
.%R RFC 3629
.%T UTF-8, a transformation format of ISO 10646
.Re
.Pp
.Lk https://www.unicode.org/versions/latest/ "The Unicode Standard"
.Pp
.Lk https://www.unicode.org/reports/tr44/ "The Unicode Character Database"