| 1 | /* nls.c -- skeletal internationalization code. */ |
| 2 | |
| 3 | /* Copyright (C) 1996-2009 Free Software Foundation, Inc. |
| 4 | |
| 5 | This file is part of the GNU Readline Library (Readline), a library |
| 6 | for reading lines of text with interactive input and history editing. |
| 7 | |
| 8 | Readline is free software: you can redistribute it and/or modify |
| 9 | it under the terms of the GNU General Public License as published by |
| 10 | the Free Software Foundation, either version 3 of the License, or |
| 11 | (at your option) any later version. |
| 12 | |
| 13 | Readline is distributed in the hope that it will be useful, |
| 14 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 16 | GNU General Public License for more details. |
| 17 | |
| 18 | You should have received a copy of the GNU General Public License |
| 19 | along with Readline. If not, see <http://www.gnu.org/licenses/>. |
| 20 | */ |
| 21 | |
| 22 | #define READLINE_LIBRARY |
| 23 | |
| 24 | #if defined (HAVE_CONFIG_H) |
| 25 | # include <config.h> |
| 26 | #endif |
| 27 | |
| 28 | #include <sys/types.h> |
| 29 | |
| 30 | #include <stdio.h> |
| 31 | |
| 32 | #if defined (HAVE_UNISTD_H) |
| 33 | # include <unistd.h> |
| 34 | #endif /* HAVE_UNISTD_H */ |
| 35 | |
| 36 | #if defined (HAVE_STDLIB_H) |
| 37 | # include <stdlib.h> |
| 38 | #else |
| 39 | # include "ansi_stdlib.h" |
| 40 | #endif /* HAVE_STDLIB_H */ |
| 41 | |
| 42 | #if defined (HAVE_LOCALE_H) |
| 43 | # include <locale.h> |
| 44 | #endif |
| 45 | |
| 46 | #if defined (HAVE_LANGINFO_CODESET) |
| 47 | # include <langinfo.h> |
| 48 | #endif |
| 49 | |
| 50 | #include <ctype.h> |
| 51 | |
| 52 | #include "rldefs.h" |
| 53 | #include "readline.h" |
| 54 | #include "rlshell.h" |
| 55 | #include "rlprivate.h" |
| 56 | |
| 57 | static int utf8locale PARAMS((char *)); |
| 58 | |
| 59 | #if !defined (HAVE_SETLOCALE) |
| 60 | /* A list of legal values for the LANG or LC_CTYPE environment variables. |
| 61 | If a locale name in this list is the value for the LC_ALL, LC_CTYPE, |
| 62 | or LANG environment variable (using the first of those with a value), |
| 63 | readline eight-bit mode is enabled. */ |
| 64 | static char *legal_lang_values[] = |
| 65 | { |
| 66 | "iso88591", |
| 67 | "iso88592", |
| 68 | "iso88593", |
| 69 | "iso88594", |
| 70 | "iso88595", |
| 71 | "iso88596", |
| 72 | "iso88597", |
| 73 | "iso88598", |
| 74 | "iso88599", |
| 75 | "iso885910", |
| 76 | "koi8r", |
| 77 | 0 |
| 78 | }; |
| 79 | |
| 80 | static char *normalize_codeset PARAMS((char *)); |
| 81 | #endif /* !HAVE_SETLOCALE */ |
| 82 | |
| 83 | static char *find_codeset PARAMS((char *, size_t *)); |
| 84 | |
| 85 | static char *_rl_get_locale_var PARAMS((const char *)); |
| 86 | |
| 87 | static char * |
| 88 | _rl_get_locale_var (v) |
| 89 | const char *v; |
| 90 | { |
| 91 | char *lspec; |
| 92 | |
| 93 | lspec = sh_get_env_value ("LC_ALL"); |
| 94 | if (lspec == 0 || *lspec == 0) |
| 95 | lspec = sh_get_env_value (v); |
| 96 | if (lspec == 0 || *lspec == 0) |
| 97 | lspec = sh_get_env_value ("LANG"); |
| 98 | |
| 99 | return lspec; |
| 100 | } |
| 101 | |
| 102 | static int |
| 103 | utf8locale (lspec) |
| 104 | char *lspec; |
| 105 | { |
| 106 | char *cp; |
| 107 | size_t len; |
| 108 | |
| 109 | #if HAVE_LANGINFO_CODESET |
| 110 | cp = nl_langinfo (CODESET); |
| 111 | return (STREQ (cp, "UTF-8") || STREQ (cp, "utf8")); |
| 112 | #else |
| 113 | cp = find_codeset (lspec, &len); |
| 114 | |
| 115 | if (cp == 0 || len < 4 || len > 5) |
| 116 | return 0; |
| 117 | return ((len == 5) ? strncmp (cp, "UTF-8", len) == 0 : strncmp (cp, "utf8", 4) == 0); |
| 118 | #endif |
| 119 | } |
| 120 | |
| 121 | /* Check for LC_ALL, LC_CTYPE, and LANG and use the first with a value |
| 122 | to decide the defaults for 8-bit character input and output. Returns |
| 123 | 1 if we set eight-bit mode. */ |
| 124 | int |
| 125 | _rl_init_eightbit () |
| 126 | { |
| 127 | /* If we have setlocale(3), just check the current LC_CTYPE category |
| 128 | value, and go into eight-bit mode if it's not C or POSIX. */ |
| 129 | #if defined (HAVE_SETLOCALE) |
| 130 | char *lspec, *t; |
| 131 | |
| 132 | /* Set the LC_CTYPE locale category from environment variables. */ |
| 133 | lspec = _rl_get_locale_var ("LC_CTYPE"); |
| 134 | /* Since _rl_get_locale_var queries the right environment variables, |
| 135 | we query the current locale settings with setlocale(), and, if |
| 136 | that doesn't return anything, we set lspec to the empty string to |
| 137 | force the subsequent call to setlocale() to define the `native' |
| 138 | environment. */ |
| 139 | if (lspec == 0 || *lspec == 0) |
| 140 | lspec = setlocale (LC_CTYPE, (char *)NULL); |
| 141 | if (lspec == 0) |
| 142 | lspec = ""; |
| 143 | t = setlocale (LC_CTYPE, lspec); |
| 144 | |
| 145 | if (t && *t) |
| 146 | _rl_utf8locale = utf8locale (t); |
| 147 | |
| 148 | if (t && *t && (t[0] != 'C' || t[1]) && (STREQ (t, "POSIX") == 0)) |
| 149 | { |
| 150 | _rl_meta_flag = 1; |
| 151 | _rl_convert_meta_chars_to_ascii = 0; |
| 152 | _rl_output_meta_chars = 1; |
| 153 | return (1); |
| 154 | } |
| 155 | else |
| 156 | return (0); |
| 157 | |
| 158 | #else /* !HAVE_SETLOCALE */ |
| 159 | char *lspec, *t; |
| 160 | int i; |
| 161 | |
| 162 | /* We don't have setlocale. Finesse it. Check the environment for the |
| 163 | appropriate variables and set eight-bit mode if they have the right |
| 164 | values. */ |
| 165 | lspec = _rl_get_locale_var ("LC_CTYPE"); |
| 166 | |
| 167 | if (lspec == 0 || (t = normalize_codeset (lspec)) == 0) |
| 168 | return (0); |
| 169 | for (i = 0; t && legal_lang_values[i]; i++) |
| 170 | if (STREQ (t, legal_lang_values[i])) |
| 171 | { |
| 172 | _rl_meta_flag = 1; |
| 173 | _rl_convert_meta_chars_to_ascii = 0; |
| 174 | _rl_output_meta_chars = 1; |
| 175 | break; |
| 176 | } |
| 177 | xfree (t); |
| 178 | return (legal_lang_values[i] ? 1 : 0); |
| 179 | |
| 180 | #endif /* !HAVE_SETLOCALE */ |
| 181 | } |
| 182 | |
| 183 | #if !defined (HAVE_SETLOCALE) |
| 184 | static char * |
| 185 | normalize_codeset (codeset) |
| 186 | char *codeset; |
| 187 | { |
| 188 | size_t namelen, i; |
| 189 | int len, all_digits; |
| 190 | char *wp, *retval; |
| 191 | |
| 192 | codeset = find_codeset (codeset, &namelen); |
| 193 | |
| 194 | if (codeset == 0) |
| 195 | return (codeset); |
| 196 | |
| 197 | all_digits = 1; |
| 198 | for (len = 0, i = 0; i < namelen; i++) |
| 199 | { |
| 200 | if (ISALNUM ((unsigned char)codeset[i])) |
| 201 | { |
| 202 | len++; |
| 203 | all_digits &= _rl_digit_p (codeset[i]); |
| 204 | } |
| 205 | } |
| 206 | |
| 207 | retval = (char *)malloc ((all_digits ? 3 : 0) + len + 1); |
| 208 | if (retval == 0) |
| 209 | return ((char *)0); |
| 210 | |
| 211 | wp = retval; |
| 212 | /* Add `iso' to beginning of an all-digit codeset */ |
| 213 | if (all_digits) |
| 214 | { |
| 215 | *wp++ = 'i'; |
| 216 | *wp++ = 's'; |
| 217 | *wp++ = 'o'; |
| 218 | } |
| 219 | |
| 220 | for (i = 0; i < namelen; i++) |
| 221 | if (ISALPHA ((unsigned char)codeset[i])) |
| 222 | *wp++ = _rl_to_lower (codeset[i]); |
| 223 | else if (_rl_digit_p (codeset[i])) |
| 224 | *wp++ = codeset[i]; |
| 225 | *wp = '\0'; |
| 226 | |
| 227 | return retval; |
| 228 | } |
| 229 | #endif /* !HAVE_SETLOCALE */ |
| 230 | |
| 231 | /* Isolate codeset portion of locale specification. */ |
| 232 | static char * |
| 233 | find_codeset (name, lenp) |
| 234 | char *name; |
| 235 | size_t *lenp; |
| 236 | { |
| 237 | char *cp, *language, *result; |
| 238 | |
| 239 | cp = language = name; |
| 240 | result = (char *)0; |
| 241 | |
| 242 | while (*cp && *cp != '_' && *cp != '@' && *cp != '+' && *cp != ',') |
| 243 | cp++; |
| 244 | |
| 245 | /* This does not make sense: language has to be specified. As |
| 246 | an exception we allow the variable to contain only the codeset |
| 247 | name. Perhaps there are funny codeset names. */ |
| 248 | if (language == cp) |
| 249 | { |
| 250 | *lenp = strlen (language); |
| 251 | result = language; |
| 252 | } |
| 253 | else |
| 254 | { |
| 255 | /* Next is the territory. */ |
| 256 | if (*cp == '_') |
| 257 | do |
| 258 | ++cp; |
| 259 | while (*cp && *cp != '.' && *cp != '@' && *cp != '+' && *cp != ',' && *cp != '_'); |
| 260 | |
| 261 | /* Now, finally, is the codeset. */ |
| 262 | result = cp; |
| 263 | if (*cp == '.') |
| 264 | do |
| 265 | ++cp; |
| 266 | while (*cp && *cp != '@'); |
| 267 | |
| 268 | if (cp - result > 2) |
| 269 | { |
| 270 | result++; |
| 271 | *lenp = cp - result; |
| 272 | } |
| 273 | else |
| 274 | { |
| 275 | *lenp = strlen (language); |
| 276 | result = language; |
| 277 | } |
| 278 | } |
| 279 | |
| 280 | return result; |
| 281 | } |