1CHARSETS(7) Linux Programmer's Manual CHARSETS(7)
2
6 charsets - character set standards and internationalization
7
9 This manual page gives an overview on different character set standards
10 and how they were used on Linux before Unicode became ubiquitous. Some
11 of this information is still helpful for people working with legacy
12 systems and documents.
13 Standards discussed include such as ASCII, GB 2312, ISO 8859, JIS,
15 KOI8-R, KS, and Unicode.
16 The primary emphasis is on character sets that were actually used by
18 locale character sets, not the myriad others that could be found in
19 data from other systems.
20 ASCII
22 ASCII (American Standard Code For Information Interchange) is the orig‐
23 inal 7-bit character set, originally designed for American English.
24 Also known as US-ASCII. It is currently described by the ISO 646:1991
25 IRV (International Reference Version) standard.
26 Various ASCII variants replacing the dollar sign with other currency
28 symbols and replacing punctuation with non-English alphabetic charac‐
29 ters to cover German, French, Spanish, and others in 7 bits emerged.
30 All are deprecated; glibc does not support locales whose character sets
31 are not true supersets of ASCII.
32 As Unicode, when using UTF-8, is ASCII-compatible, plain ASCII text
34 still renders properly on modern UTF-8 using systems.
35 ISO 8859
37 ISO 8859 is a series of 15 8-bit character sets, all of which have
38 ASCII in their low (7-bit) half, invisible control characters in posi‐
39 tions 128 to 159, and 96 fixed-width graphics in positions 160–255.
40 Of these, the most important is ISO 8859-1 ("Latin Alphabet No .1" /
42 Latin-1). It was widely adopted and supported by different systems,
43 and is gradually being replaced with Unicode. The ISO 8859-1 charac‐
44 ters are also the first 256 characters of Unicode.
45 Console support for the other 8859 character sets is available under
47 Linux through user-mode utilities (such as setfont(8)) that modify key‐
48 board bindings and the EGA graphics table and employ the "user mapping"
49 font table in the console driver.
50 Here are brief descriptions of each set:
52 8859-1 (Latin-1)
54 Latin-1 covers many West European languages such as Albanian,
55 Basque, Danish, English, Faroese, Galician, Icelandic, Irish,
56 Italian, Norwegian, Portuguese, Spanish, and Swedish. The lack
57 of the ligatures Dutch IJ/ij, French œ, and old-style „German“
58 quotation marks was considered tolerable.
59 8859-2 (Latin-2)
61 Latin-2 supports many Latin-written Central and East European
62 languages such as Bosnian, Croatian, Czech, German, Hungarian,
63 Polish, Slovak, and Slovene. Replacing Romanian ș/ț with ş/ţ
64 was considered tolerable.
65 8859-3 (Latin-3)
67 Latin-3 was designed to cover of Esperanto, Maltese, and Turk‐
68 ish, but 8859-9 later superseded it for Turkish.
69 8859-4 (Latin-4)
71 Latin-4 introduced letters for North European languages such as
72 Estonian, Latvian, and Lithuanian, but was superseded by 8859-10
73 and 8859-13.
74 8859-5 Cyrillic letters supporting Bulgarian, Byelorussian, Macedonian,
76 Russian, Serbian, and (almost completely) Ukrainian. It was
77 never widely used, see the discussion of KOI8-R/KOI8-U below.
78 8859-6 Was created for Arabic. The 8859-6 glyph table is a fixed font
80 of separate letter forms, but a proper display engine should
81 combine these using the proper initial, medial, and final forms.
82 8859-7 Was created for Modern Greek in 1987, updated in 2003.
84 8859-8 Supports Modern Hebrew without niqud (punctuation signs). Niqud
86 and full-fledged Biblical Hebrew were outside the scope of this
87 character set.
88 8859-9 (Latin-5)
90 This is a variant of Latin-1 that replaces Icelandic letters
91 with Turkish ones.
92 8859-10 (Latin-6)
94 Latin-6 added the Inuit (Greenlandic) and Sami (Lappish) letters
95 that were missing in Latin-4 to cover the entire Nordic area.
96 8859-11
98 Supports the Thai alphabet and is nearly identical to the
99 TIS-620 standard.
100 8859-12
102 This set does not exist.
103 8859-13 (Latin-7)
105 Supports the Baltic Rim languages; in particular, it includes
106 Latvian characters not found in Latin-4.
107 8859-14 (Latin-8)
109 This is the Celtic character set, covering Old Irish, Manx,
110 Gaelic, Welsh, Cornish, and Breton.
111 8859-15 (Latin-9)
113 Latin-9 is similar to the widely used Latin-1 but replaces some
114 less common symbols with the Euro sign and French and Finnish
115 letters that were missing in Latin-1.
116 8859-16 (Latin-10)
118 This set covers many Southeast European languages, and most im‐
119 portantly supports Romanian more completely than Latin-2.
120 KOI8-R / KOI8-U
122 KOI8-R is a non-ISO character set popular in Russia before Unicode.
123 The lower half is ASCII; the upper is a Cyrillic character set somewhat
124 better designed than ISO 8859-5. KOI8-U, based on KOI8-R, has better
125 support for Ukrainian. Neither of these sets are ISO-2022 compatible,
126 unlike the ISO 8859 series.
127 Console support for KOI8-R is available under Linux through user-mode
129 utilities that modify keyboard bindings and the EGA graphics table, and
130 employ the "user mapping" font table in the console driver.
131 GB 2312
133 GB 2312 is a mainland Chinese national standard character set used to
134 express simplified Chinese. Just like JIS X 0208, characters are
135 mapped into a 94x94 two-byte matrix used to construct EUC-CN. EUC-CN
136 is the most important encoding for Linux and includes ASCII and GB
137 2312. Note that EUC-CN is often called as GB, GB 2312, or CN-GB.
138 Big5
140 Big5 was a popular character set in Taiwan to express traditional Chi‐
141 nese. (Big5 is both a character set and an encoding.) It is a super‐
142 set of ASCII. Non-ASCII characters are expressed in two bytes. Bytes
143 0xa1–0xfe are used as leading bytes for two-byte characters. Big5 and
144 its extension were widely used in Taiwan and Hong Kong. It is not ISO
145 2022 compliant.
146 JIS X 0208
148 JIS X 0208 is a Japanese national standard character set. Though there
149 are some more Japanese national standard character sets (like JIS X
150 0201, JIS X 0212, and JIS X 0213), this is the most important one.
151 Characters are mapped into a 94x94 two-byte matrix, whose each byte is
152 in the range 0x21–0x7e. Note that JIS X 0208 is a character set, not
153 an encoding. This means that JIS X 0208 itself is not used for ex‐
154 pressing text data. JIS X 0208 is used as a component to construct en‐
155 codings such as EUC-JP, Shift_JIS, and ISO-2022-JP. EUC-JP is the most
156 important encoding for Linux and includes ASCII and JIS X 0208. In
157 EUC-JP, JIS X 0208 characters are expressed in two bytes, each of which
158 is the JIS X 0208 code plus 0x80.
159 KS X 1001
161 KS X 1001 is a Korean national standard character set. Just as JIS X
162 0208, characters are mapped into a 94x94 two-byte matrix. KS X 1001 is
163 used like JIS X 0208, as a component to construct encodings such as
164 EUC-KR, Johab, and ISO-2022-KR. EUC-KR is the most important encoding
165 for Linux and includes ASCII and KS X 1001. KS C 5601 is an older name
166 for KS X 1001.
167 ISO 2022 and ISO 4873
169 The ISO 2022 and 4873 standards describe a font-control model based on
170 VT100 practice. This model is (partially) supported by the Linux ker‐
171 nel and by xterm(1). Several ISO 2022-based character encodings have
172 been defined, especially for Japanese.
173 There are 4 graphic character sets, called G0, G1, G2, and G3, and one
175 of them is the current character set for codes with high bit zero (ini‐
176 tially G0), and one of them is the current character set for codes with
177 high bit one (initially G1). Each graphic character set has 94 or 96
178 characters, and is essentially a 7-bit character set. It uses codes
179 either 040–0177 (041–0176) or 0240–0377 (0241–0376). G0 always has
180 size 94 and uses codes 041–0176.
181 Switching between character sets is done using the shift functions ^N
183 (SO or LS1), ^O (SI or LS0), ESC n (LS2), ESC o (LS3), ESC N (SS2), ESC
184 O (SS3), ESC ~ (LS1R), ESC } (LS2R), ESC | (LS3R). The function LSn
185 makes character set Gn the current one for codes with high bit zero.
186 The function LSnR makes character set Gn the current one for codes with
187 high bit one. The function SSn makes character set Gn (n=2 or 3) the
188 current one for the next character only (regardless of the value of its
189 high order bit).
190 A 94-character set is designated as Gn character set by an escape se‐
192 quence ESC ( xx (for G0), ESC ) xx (for G1), ESC * xx (for G2), ESC +
193 xx (for G3), where xx is a symbol or a pair of symbols found in the ISO
194 2375 International Register of Coded Character Sets. For example, ESC
195 ( @ selects the ISO 646 character set as G0, ESC ( A selects the UK
196 standard character set (with pound instead of number sign), ESC ( B se‐
197 lects ASCII (with dollar instead of currency sign), ESC ( M selects a
198 character set for African languages, ESC ( ! A selects the Cuban char‐
199 acter set, and so on.
200 A 96-character set is designated as Gn character set by an escape se‐
202 quence ESC - xx (for G1), ESC . xx (for G2) or ESC / xx (for G3). For
203 example, ESC - G selects the Hebrew alphabet as G1.
204 A multibyte character set is designated as Gn character set by an es‐
206 cape sequence ESC $ xx or ESC $ ( xx (for G0), ESC $ ) xx (for G1), ESC
207 $ * xx (for G2), ESC $ + xx (for G3). For example, ESC $ ( C selects
208 the Korean character set for G0. The Japanese character set selected
209 by ESC $ B has a more recent version selected by ESC & @ ESC $ B.
210 ISO 4873 stipulates a narrower use of character sets, where G0 is fixed
212 (always ASCII), so that G1, G2 and G3 can be invoked only for codes
213 with the high order bit set. In particular, ^N and ^O are not used
214 anymore, ESC ( xx can be used only with xx=B, and ESC ) xx, ESC * xx,
215 ESC + xx are equivalent to ESC - xx, ESC . xx, ESC / xx, respectively.
216 TIS-620
218 TIS-620 is a Thai national standard character set and a superset of
219 ASCII. In the same fashion as the ISO 8859 series, Thai characters are
220 mapped into 0xa1–0xfe.
221 Unicode
223 Unicode (ISO 10646) is a standard which aims to unambiguously represent
224 every character in every human language. Unicode's structure permits
225 20.1 bits to encode every character. Since most computers don't in‐
226 clude 20.1-bit integers, Unicode is usually encoded as 32-bit integers
227 internally and either a series of 16-bit integers (UTF-16) (needing two
228 16-bit integers only when encoding certain rare characters) or a series
229 of 8-bit bytes (UTF-8).
230 Linux represents Unicode using the 8-bit Unicode Transformation Format
232 (UTF-8). UTF-8 is a variable length encoding of Unicode. It uses 1
233 byte to code 7 bits, 2 bytes for 11 bits, 3 bytes for 16 bits, 4 bytes
234 for 21 bits, 5 bytes for 26 bits, 6 bytes for 31 bits.
235 Let 0,1,x stand for a zero, one, or arbitrary bit. A byte 0xxxxxxx
237 stands for the Unicode 00000000 0xxxxxxx which codes the same symbol as
238 the ASCII 0xxxxxxx. Thus, ASCII goes unchanged into UTF-8, and people
239 using only ASCII do not notice any change: not in code, and not in file
240 size.
241 A byte 110xxxxx is the start of a 2-byte code, and 110xxxxx 10yyyyyy is
243 assembled into 00000xxx xxyyyyyy. A byte 1110xxxx is the start of a
244 3-byte code, and 1110xxxx 10yyyyyy 10zzzzzz is assembled into xxxxyyyy
245 yyzzzzzz. (When UTF-8 is used to code the 31-bit ISO 10646 then this
246 progression continues up to 6-byte codes.)
247 For most texts in ISO 8859 character sets, this means that the charac‐
249 ters outside of ASCII are now coded with two bytes. This tends to ex‐
250 pand ordinary text files by only one or two percent. For Russian or
251 Greek texts, this expands ordinary text files by 100%, since text in
252 those languages is mostly outside of ASCII. For Japanese users this
253 means that the 16-bit codes now in common use will take three bytes.
254 While there are algorithmic conversions from some character sets (espe‐
255 cially ISO 8859-1) to Unicode, general conversion requires carrying
256 around conversion tables, which can be quite large for 16-bit codes.
257 Note that UTF-8 is self-synchronizing: 10xxxxxx is a tail, any other
259 byte is the head of a code. Note that the only way ASCII bytes occur
260 in a UTF-8 stream, is as themselves. In particular, there are no em‐
261 bedded NULs ('\0') or '/'s that form part of some larger code.
262 Since ASCII, and, in particular, NUL and '/', are unchanged, the kernel
264 does not notice that UTF-8 is being used. It does not care at all what
265 the bytes it is handling stand for.
266 Rendering of Unicode data streams is typically handled through "sub‐
268 font" tables which map a subset of Unicode to glyphs. Internally the
269 kernel uses Unicode to describe the subfont loaded in video RAM. This
270 means that in the Linux console in UTF-8 mode, one can use a character
271 set with 512 different symbols. This is not enough for Japanese, Chi‐
272 nese, and Korean, but it is enough for most other purposes.
273
275 iconv(1), ascii(7), iso_8859-1(7), unicode(7), utf-8(7)
276
278 This page is part of release 5.10 of the Linux man-pages project. A
279 description of the project, information about reporting bugs, and the
280 latest version of this page, can be found at
281 https://www.kernel.org/doc/man-pages/.
282Linux 2020-08-13 CHARSETS(7)