| 1 | /* Header for multibyte character handler. |
| 2 | Copyright (C) 1995, 1997, 1998 Electrotechnical Laboratory, JAPAN. |
| 3 | Licensed to the Free Software Foundation. |
| 4 | |
| 5 | This file is part of GNU Emacs. |
| 6 | |
| 7 | GNU Emacs is free software; you can redistribute it and/or modify |
| 8 | it under the terms of the GNU General Public License as published by |
| 9 | the Free Software Foundation; either version 2, or (at your option) |
| 10 | any later version. |
| 11 | |
| 12 | GNU Emacs is distributed in the hope that it will be useful, |
| 13 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 15 | GNU General Public License for more details. |
| 16 | |
| 17 | You should have received a copy of the GNU General Public License |
| 18 | along with GNU Emacs; see the file COPYING. If not, write to |
| 19 | the Free Software Foundation, Inc., 59 Temple Place - Suite 330, |
| 20 | Boston, MA 02111-1307, USA. */ |
| 21 | |
| 22 | #ifndef EMACS_CHARSET_H |
| 23 | #define EMACS_CHARSET_H |
| 24 | |
| 25 | /* #define BYTE_COMBINING_DEBUG */ |
| 26 | |
| 27 | /*** GENERAL NOTE on CHARACTER SET (CHARSET) *** |
| 28 | |
| 29 | A character set ("charset" hereafter) is a meaningful collection |
| 30 | (i.e. language, culture, functionality, etc) of characters. Emacs |
| 31 | handles multiple charsets at once. Each charset corresponds to one |
| 32 | of ISO charsets. Emacs identifies a charset by a unique |
| 33 | identification number, whereas ISO identifies a charset by a triplet |
| 34 | of DIMENSION, CHARS and FINAL-CHAR. So, hereafter, just saying |
| 35 | "charset" means an identification number (integer value). |
| 36 | |
| 37 | The value range of charset is 0x00, 0x81..0xFE. There are four |
| 38 | kinds of charset depending on DIMENSION (1 or 2) and CHARS (94 or |
| 39 | 96). For instance, a charset of DIMENSION2_CHARS94 contains 94x94 |
| 40 | characters. |
| 41 | |
| 42 | Within Emacs Lisp, a charset is treated as a symbol which has a |
| 43 | property `charset'. The property value is a vector containing |
| 44 | various information about the charset. For readability of C codes, |
| 45 | we use the following convention for C variable names: |
| 46 | charset_symbol: Emacs Lisp symbol of a charset |
| 47 | charset_id: Emacs Lisp integer of an identification number of a charset |
| 48 | charset: C integer of an identification number of a charset |
| 49 | |
| 50 | Each charset (except for ascii) is assigned a base leading-code |
| 51 | (range 0x80..0x9E). In addition, a charset of greater than 0xA0 |
| 52 | (whose base leading-code is 0x9A..0x9D) is assigned an extended |
| 53 | leading-code (range 0xA0..0xFE). In this case, each base |
| 54 | leading-code specify the allowable range of extended leading-code as |
| 55 | shown in the table below. A leading-code is used to represent a |
| 56 | character in Emacs' buffer and string. |
| 57 | |
| 58 | We call a charset which has extended leading-code as "private |
| 59 | charset" because those are mainly for a charset which is not yet |
| 60 | registered by ISO. On the contrary, we call a charset which does |
| 61 | not have extended leading-code as "official charset". |
| 62 | |
| 63 | --------------------------------------------------------------------------- |
| 64 | charset dimension base leading-code extended leading-code |
| 65 | --------------------------------------------------------------------------- |
| 66 | 0x00 official dim1 -- none -- -- none -- |
| 67 | (ASCII) |
| 68 | 0x01..0x7F --never used-- |
| 69 | 0x80 official dim1 -- none -- -- none -- |
| 70 | (eight-bit-graphic) |
| 71 | 0x81..0x8F official dim1 same as charset -- none -- |
| 72 | 0x90..0x99 official dim2 same as charset -- none -- |
| 73 | 0x9A..0x9D --never used-- |
| 74 | 0x9E official dim1 same as charset -- none -- |
| 75 | (eight-bit-control) |
| 76 | 0x9F --never used-- |
| 77 | 0xA0..0xDF private dim1 0x9A same as charset |
| 78 | of 1-column width |
| 79 | 0xE0..0xEF private dim1 0x9B same as charset |
| 80 | of 2-column width |
| 81 | 0xF0..0xF4 private dim2 0x9C same as charset |
| 82 | of 1-column width |
| 83 | 0xF5..0xFE private dim2 0x9D same as charset |
| 84 | of 2-column width |
| 85 | 0xFF --never used-- |
| 86 | --------------------------------------------------------------------------- |
| 87 | |
| 88 | */ |
| 89 | |
| 90 | /* Definition of special leading-codes. */ |
| 91 | /* Leading-code followed by extended leading-code. */ |
| 92 | #define LEADING_CODE_PRIVATE_11 0x9A /* for private DIMENSION1 of 1-column */ |
| 93 | #define LEADING_CODE_PRIVATE_12 0x9B /* for private DIMENSION1 of 2-column */ |
| 94 | #define LEADING_CODE_PRIVATE_21 0x9C /* for private DIMENSION2 of 1-column */ |
| 95 | #define LEADING_CODE_PRIVATE_22 0x9D /* for private DIMENSION2 of 2-column */ |
| 96 | |
| 97 | #define LEADING_CODE_8_BIT_CONTROL 0x9E /* for `eight-bit-control' */ |
| 98 | |
| 99 | /* Extended leading-code. */ |
| 100 | /* Start of each extended leading-codes. */ |
| 101 | #define LEADING_CODE_EXT_11 0xA0 /* follows LEADING_CODE_PRIVATE_11 */ |
| 102 | #define LEADING_CODE_EXT_12 0xE0 /* follows LEADING_CODE_PRIVATE_12 */ |
| 103 | #define LEADING_CODE_EXT_21 0xF0 /* follows LEADING_CODE_PRIVATE_21 */ |
| 104 | #define LEADING_CODE_EXT_22 0xF5 /* follows LEADING_CODE_PRIVATE_22 */ |
| 105 | /* Maximum value of extended leading-codes. */ |
| 106 | #define LEADING_CODE_EXT_MAX 0xFE |
| 107 | |
| 108 | /* Definition of minimum/maximum charset of each DIMENSION. */ |
| 109 | #define MIN_CHARSET_OFFICIAL_DIMENSION1 0x81 |
| 110 | #define MAX_CHARSET_OFFICIAL_DIMENSION1 0x8F |
| 111 | #define MIN_CHARSET_OFFICIAL_DIMENSION2 0x90 |
| 112 | #define MAX_CHARSET_OFFICIAL_DIMENSION2 0x99 |
| 113 | #define MIN_CHARSET_PRIVATE_DIMENSION1 LEADING_CODE_EXT_11 |
| 114 | #define MIN_CHARSET_PRIVATE_DIMENSION2 LEADING_CODE_EXT_21 |
| 115 | |
| 116 | /* Maximum value of overall charset identification number. */ |
| 117 | #define MAX_CHARSET 0xFE |
| 118 | |
| 119 | /* Definition of special charsets. */ |
| 120 | #define CHARSET_ASCII 0 /* 0x00..0x7F */ |
| 121 | #define CHARSET_8_BIT_CONTROL 0x9E /* 0x80..0x9F */ |
| 122 | #define CHARSET_8_BIT_GRAPHIC 0x80 /* 0xA0..0xFF */ |
| 123 | |
| 124 | extern int charset_latin_iso8859_1; /* ISO8859-1 (Latin-1) */ |
| 125 | extern int charset_jisx0208_1978; /* JISX0208.1978 (Japanese Kanji old set) */ |
| 126 | extern int charset_jisx0208; /* JISX0208.1983 (Japanese Kanji) */ |
| 127 | extern int charset_katakana_jisx0201; /* JISX0201.Kana (Japanese Katakana) */ |
| 128 | extern int charset_latin_jisx0201; /* JISX0201.Roman (Japanese Roman) */ |
| 129 | extern int charset_big5_1; /* Big5 Level 1 (Chinese Traditional) */ |
| 130 | extern int charset_big5_2; /* Big5 Level 2 (Chinese Traditional) */ |
| 131 | |
| 132 | /* Check if CH is an ASCII character or a base leading-code. |
| 133 | Nowadays, any byte can be the first byte of a character in a |
| 134 | multibyte buffer/string. So this macro name is not appropriate. */ |
| 135 | #define CHAR_HEAD_P(ch) ((unsigned char) (ch) < 0xA0) |
| 136 | |
| 137 | /*** GENERAL NOTE on CHARACTER REPRESENTATION *** |
| 138 | |
| 139 | At first, the term "character" or "char" is used for a multilingual |
| 140 | character (of course, including ASCII character), not for a byte in |
| 141 | computer memory. We use the term "code" or "byte" for the latter |
| 142 | case. |
| 143 | |
| 144 | A character is identified by charset and one or two POSITION-CODEs. |
| 145 | POSITION-CODE is the position of the character in the charset. A |
| 146 | character of DIMENSION1 charset has one POSITION-CODE: POSITION-CODE-1. |
| 147 | A character of DIMENSION2 charset has two POSITION-CODE: |
| 148 | POSITION-CODE-1 and POSITION-CODE-2. The code range of |
| 149 | POSITION-CODE is 0x20..0x7F. |
| 150 | |
| 151 | Emacs has two kinds of representation of a character: multi-byte |
| 152 | form (for buffer and string) and single-word form (for character |
| 153 | object in Emacs Lisp). The latter is called "character code" here |
| 154 | after. Both representations encode the information of charset and |
| 155 | POSITION-CODE but in a different way (for instance, MSB of |
| 156 | POSITION-CODE is set in multi-byte form). |
| 157 | |
| 158 | For details of multi-byte form, see the section "2. Emacs internal |
| 159 | format handlers" of `coding.c'. |
| 160 | |
| 161 | Emacs uses 19 bits for a character code. The bits are divided into |
| 162 | 3 fields: FIELD1(5bits):FIELD2(7bits):FIELD3(7bits). |
| 163 | |
| 164 | A character code of DIMENSION1 character uses FIELD2 to hold charset |
| 165 | and FIELD3 to hold POSITION-CODE-1. A character code of DIMENSION2 |
| 166 | character uses FIELD1 to hold charset, FIELD2 and FIELD3 to hold |
| 167 | POSITION-CODE-1 and POSITION-CODE-2 respectively. |
| 168 | |
| 169 | More precisely... |
| 170 | |
| 171 | FIELD2 of DIMENSION1 character (except for ascii, eight-bit-control, |
| 172 | and eight-bit-graphic) is "charset - 0x70". This is to make all |
| 173 | character codes except for ASCII and 8-bit codes greater than 256. |
| 174 | So, the range of FIELD2 of DIMENSION1 character is 0, 1, or |
| 175 | 0x11..0x7F. |
| 176 | |
| 177 | FIELD1 of DIMENSION2 character is "charset - 0x8F" for official |
| 178 | charset and "charset - 0xE0" for private charset. So, the range of |
| 179 | FIELD1 of DIMENSION2 character is 0x01..0x1E. |
| 180 | |
| 181 | ----------------------------------------------------------------------------- |
| 182 | charset FIELD1 (5-bit) FIELD2 (7-bit) FIELD3 (7-bit) |
| 183 | ----------------------------------------------------------------------------- |
| 184 | ascii 0 0 0x00..0x7F |
| 185 | eight-bit-control 0 1 0x00..0x1F |
| 186 | eight-bit-graphic 0 1 0x20..0x7F |
| 187 | DIMENSION1 0 charset - 0x70 POSITION-CODE-1 |
| 188 | DIMENSION2(o) charset - 0x8F POSITION-CODE-1 POSITION-CODE-2 |
| 189 | DIMENSION2(p) charset - 0xE0 POSITION-CODE-1 POSITION-CODE-2 |
| 190 | ----------------------------------------------------------------------------- |
| 191 | "(o)": official, "(p)": private |
| 192 | ----------------------------------------------------------------------------- |
| 193 | */ |
| 194 | |
| 195 | /* Masks of each field of character code. */ |
| 196 | #define CHAR_FIELD1_MASK (0x1F << 14) |
| 197 | #define CHAR_FIELD2_MASK (0x7F << 7) |
| 198 | #define CHAR_FIELD3_MASK 0x7F |
| 199 | |
| 200 | /* Macros to access each field of character C. */ |
| 201 | #define CHAR_FIELD1(c) (((c) & CHAR_FIELD1_MASK) >> 14) |
| 202 | #define CHAR_FIELD2(c) (((c) & CHAR_FIELD2_MASK) >> 7) |
| 203 | #define CHAR_FIELD3(c) ((c) & CHAR_FIELD3_MASK) |
| 204 | |
| 205 | /* Minimum character code of character of each DIMENSION. */ |
| 206 | #define MIN_CHAR_OFFICIAL_DIMENSION1 \ |
| 207 | ((MIN_CHARSET_OFFICIAL_DIMENSION1 - 0x70) << 7) |
| 208 | #define MIN_CHAR_PRIVATE_DIMENSION1 \ |
| 209 | ((MIN_CHARSET_PRIVATE_DIMENSION1 - 0x70) << 7) |
| 210 | #define MIN_CHAR_OFFICIAL_DIMENSION2 \ |
| 211 | ((MIN_CHARSET_OFFICIAL_DIMENSION2 - 0x8F) << 14) |
| 212 | #define MIN_CHAR_PRIVATE_DIMENSION2 \ |
| 213 | ((MIN_CHARSET_PRIVATE_DIMENSION2 - 0xE0) << 14) |
| 214 | /* Maximum character code currently used plus 1. */ |
| 215 | #define MAX_CHAR (0x1F << 14) |
| 216 | |
| 217 | /* 1 if C is a single byte character, else 0. */ |
| 218 | #define SINGLE_BYTE_CHAR_P(c) ((unsigned) (c) < 0x100) |
| 219 | |
| 220 | /* 1 if BYTE is an ASCII character in itself, in multibyte mode. */ |
| 221 | #define ASCII_BYTE_P(byte) ((byte) < 0x80) |
| 222 | |
| 223 | /* A char-table containing information of each character set. |
| 224 | |
| 225 | Unlike ordinary char-tables, this doesn't contain any nested table. |
| 226 | Only the top level elements are used. Each element is a vector of |
| 227 | the following information: |
| 228 | CHARSET-ID, BYTES, DIMENSION, CHARS, WIDTH, DIRECTION, |
| 229 | LEADING-CODE-BASE, LEADING-CODE-EXT, |
| 230 | ISO-FINAL-CHAR, ISO-GRAPHIC-PLANE, |
| 231 | REVERSE-CHARSET, SHORT-NAME, LONG-NAME, DESCRIPTION, |
| 232 | PLIST. |
| 233 | |
| 234 | CHARSET-ID (integer) is the identification number of the charset. |
| 235 | |
| 236 | BYTES (integer) is the length of multi-byte form of a character in |
| 237 | the charset: one of 1, 2, 3, and 4. |
| 238 | |
| 239 | DIMENSION (integer) is the number of bytes to represent a character: 1 or 2. |
| 240 | |
| 241 | CHARS (integer) is the number of characters in a dimension: 94 or 96. |
| 242 | |
| 243 | WIDTH (integer) is the number of columns a character in the charset |
| 244 | occupies on the screen: one of 0, 1, and 2.. |
| 245 | |
| 246 | DIRECTION (integer) is the rendering direction of characters in the |
| 247 | charset when rendering. If 0, render from left to right, else |
| 248 | render from right to left. |
| 249 | |
| 250 | LEADING-CODE-BASE (integer) is the base leading-code for the |
| 251 | charset. |
| 252 | |
| 253 | LEADING-CODE-EXT (integer) is the extended leading-code for the |
| 254 | charset. All charsets of less than 0xA0 has the value 0. |
| 255 | |
| 256 | ISO-FINAL-CHAR (character) is the final character of the |
| 257 | corresponding ISO 2022 charset. It is -1 for such a character |
| 258 | that is used only internally (e.g. `eight-bit-control'). |
| 259 | |
| 260 | ISO-GRAPHIC-PLANE (integer) is the graphic plane to be invoked |
| 261 | while encoding to variants of ISO 2022 coding system, one of the |
| 262 | following: 0/graphic-plane-left(GL), 1/graphic-plane-right(GR). It |
| 263 | is -1 for such a character that is used only internally |
| 264 | (e.g. `eight-bit-control'). |
| 265 | |
| 266 | REVERSE-CHARSET (integer) is the charset which differs only in |
| 267 | LEFT-TO-RIGHT value from the charset. If there's no such a |
| 268 | charset, the value is -1. |
| 269 | |
| 270 | SHORT-NAME (string) is the short name to refer to the charset. |
| 271 | |
| 272 | LONG-NAME (string) is the long name to refer to the charset. |
| 273 | |
| 274 | DESCRIPTION (string) is the description string of the charset. |
| 275 | |
| 276 | PLIST (property list) may contain any type of information a user |
| 277 | want to put and get by functions `put-charset-property' and |
| 278 | `get-charset-property' respectively. */ |
| 279 | extern Lisp_Object Vcharset_table; |
| 280 | |
| 281 | /* Macros to access various information of CHARSET in Vcharset_table. |
| 282 | We provide these macros for efficiency. No range check of CHARSET. */ |
| 283 | |
| 284 | /* Return entry of CHARSET (C integer) in Vcharset_table. */ |
| 285 | #define CHARSET_TABLE_ENTRY(charset) \ |
| 286 | XCHAR_TABLE (Vcharset_table)->contents[((charset) == CHARSET_ASCII \ |
| 287 | ? 0 : (charset) + 128)] |
| 288 | |
| 289 | /* Return information INFO-IDX of CHARSET. */ |
| 290 | #define CHARSET_TABLE_INFO(charset, info_idx) \ |
| 291 | XVECTOR (CHARSET_TABLE_ENTRY (charset))->contents[info_idx] |
| 292 | |
| 293 | #define CHARSET_ID_IDX (0) |
| 294 | #define CHARSET_BYTES_IDX (1) |
| 295 | #define CHARSET_DIMENSION_IDX (2) |
| 296 | #define CHARSET_CHARS_IDX (3) |
| 297 | #define CHARSET_WIDTH_IDX (4) |
| 298 | #define CHARSET_DIRECTION_IDX (5) |
| 299 | #define CHARSET_LEADING_CODE_BASE_IDX (6) |
| 300 | #define CHARSET_LEADING_CODE_EXT_IDX (7) |
| 301 | #define CHARSET_ISO_FINAL_CHAR_IDX (8) |
| 302 | #define CHARSET_ISO_GRAPHIC_PLANE_IDX (9) |
| 303 | #define CHARSET_REVERSE_CHARSET_IDX (10) |
| 304 | #define CHARSET_SHORT_NAME_IDX (11) |
| 305 | #define CHARSET_LONG_NAME_IDX (12) |
| 306 | #define CHARSET_DESCRIPTION_IDX (13) |
| 307 | #define CHARSET_PLIST_IDX (14) |
| 308 | /* Size of a vector of each entry of Vcharset_table. */ |
| 309 | #define CHARSET_MAX_IDX (15) |
| 310 | |
| 311 | /* And several more macros to be used frequently. */ |
| 312 | #define CHARSET_BYTES(charset) \ |
| 313 | XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_BYTES_IDX)) |
| 314 | #define CHARSET_DIMENSION(charset) \ |
| 315 | XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_DIMENSION_IDX)) |
| 316 | #define CHARSET_CHARS(charset) \ |
| 317 | XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_CHARS_IDX)) |
| 318 | #define CHARSET_WIDTH(charset) \ |
| 319 | XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_WIDTH_IDX)) |
| 320 | #define CHARSET_DIRECTION(charset) \ |
| 321 | XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_DIRECTION_IDX)) |
| 322 | #define CHARSET_LEADING_CODE_BASE(charset) \ |
| 323 | XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_BASE_IDX)) |
| 324 | #define CHARSET_LEADING_CODE_EXT(charset) \ |
| 325 | XFASTINT (CHARSET_TABLE_INFO (charset, CHARSET_LEADING_CODE_EXT_IDX)) |
| 326 | #define CHARSET_ISO_FINAL_CHAR(charset) \ |
| 327 | XINT (CHARSET_TABLE_INFO (charset, CHARSET_ISO_FINAL_CHAR_IDX)) |
| 328 | #define CHARSET_ISO_GRAPHIC_PLANE(charset) \ |
| 329 | XINT (CHARSET_TABLE_INFO (charset, CHARSET_ISO_GRAPHIC_PLANE_IDX)) |
| 330 | #define CHARSET_REVERSE_CHARSET(charset) \ |
| 331 | XINT (CHARSET_TABLE_INFO (charset, CHARSET_REVERSE_CHARSET_IDX)) |
| 332 | |
| 333 | /* Macros to specify direction of a charset. */ |
| 334 | #define CHARSET_DIRECTION_LEFT_TO_RIGHT 0 |
| 335 | #define CHARSET_DIRECTION_RIGHT_TO_LEFT 1 |
| 336 | |
| 337 | /* A vector of charset symbol indexed by charset-id. This is used |
| 338 | only for returning charset symbol from C functions. */ |
| 339 | extern Lisp_Object Vcharset_symbol_table; |
| 340 | |
| 341 | /* Return symbol of CHARSET. */ |
| 342 | #define CHARSET_SYMBOL(charset) \ |
| 343 | XVECTOR (Vcharset_symbol_table)->contents[charset] |
| 344 | |
| 345 | /* 1 if CHARSET is in valid value range, else 0. */ |
| 346 | #define CHARSET_VALID_P(charset) \ |
| 347 | ((charset) == 0 \ |
| 348 | || ((charset) > 0x80 && (charset) <= MAX_CHARSET_OFFICIAL_DIMENSION2) \ |
| 349 | || ((charset) >= MIN_CHARSET_PRIVATE_DIMENSION1 \ |
| 350 | && (charset) <= MAX_CHARSET) \ |
| 351 | || ((charset) == CHARSET_8_BIT_CONTROL) \ |
| 352 | || ((charset) == CHARSET_8_BIT_GRAPHIC)) |
| 353 | |
| 354 | /* 1 if CHARSET is already defined, else 0. */ |
| 355 | #define CHARSET_DEFINED_P(charset) \ |
| 356 | (((charset) >= 0) && ((charset) <= MAX_CHARSET) \ |
| 357 | && !NILP (CHARSET_TABLE_ENTRY (charset))) |
| 358 | |
| 359 | /* Since the information CHARSET-BYTES and CHARSET-WIDTH of |
| 360 | Vcharset_table can be retrieved only by the first byte of |
| 361 | multi-byte form (an ASCII code or a base leading-code), we provide |
| 362 | here tables to be used by macros BYTES_BY_CHAR_HEAD and |
| 363 | WIDTH_BY_CHAR_HEAD for faster information retrieval. */ |
| 364 | extern int bytes_by_char_head[256]; |
| 365 | extern int width_by_char_head[256]; |
| 366 | |
| 367 | #define BYTES_BY_CHAR_HEAD(char_head) \ |
| 368 | (ASCII_BYTE_P (char_head) ? 1 : bytes_by_char_head[char_head]) |
| 369 | #define WIDTH_BY_CHAR_HEAD(char_head) \ |
| 370 | (ASCII_BYTE_P (char_head) ? 1 : width_by_char_head[char_head]) |
| 371 | |
| 372 | /* Charset of the character C. */ |
| 373 | #define CHAR_CHARSET(c) \ |
| 374 | (SINGLE_BYTE_CHAR_P (c) \ |
| 375 | ? (ASCII_BYTE_P (c) \ |
| 376 | ? CHARSET_ASCII \ |
| 377 | : (c) < 0xA0 ? CHARSET_8_BIT_CONTROL : CHARSET_8_BIT_GRAPHIC) \ |
| 378 | : ((c) < MIN_CHAR_OFFICIAL_DIMENSION2 \ |
| 379 | ? CHAR_FIELD2 (c) + 0x70 \ |
| 380 | : ((c) < MIN_CHAR_PRIVATE_DIMENSION2 \ |
| 381 | ? CHAR_FIELD1 (c) + 0x8F \ |
| 382 | : CHAR_FIELD1 (c) + 0xE0))) |
| 383 | |
| 384 | /* Check if two characters C1 and C2 belong to the same charset. */ |
| 385 | #define SAME_CHARSET_P(c1, c2) \ |
| 386 | (c1 < MIN_CHAR_OFFICIAL_DIMENSION2 \ |
| 387 | ? (c1 & CHAR_FIELD2_MASK) == (c2 & CHAR_FIELD2_MASK) \ |
| 388 | : (c1 & CHAR_FIELD1_MASK) == (c2 & CHAR_FIELD1_MASK)) |
| 389 | |
| 390 | /* Return a character of which charset is CHARSET and position-codes |
| 391 | are C1 and C2. DIMENSION1 character ignores C2. */ |
| 392 | #define MAKE_CHAR(charset, c1, c2) \ |
| 393 | ((charset) == CHARSET_ASCII \ |
| 394 | ? (c1) & 0x7F \ |
| 395 | : (((charset) == CHARSET_8_BIT_CONTROL \ |
| 396 | || (charset) == CHARSET_8_BIT_GRAPHIC) \ |
| 397 | ? ((c1) & 0x7F) | 0x80 \ |
| 398 | : (! CHARSET_DEFINED_P (charset) || CHARSET_DIMENSION (charset) == 1 \ |
| 399 | ? (((charset) - 0x70) << 7) | ((c1) <= 0 ? 0 : (c1)) \ |
| 400 | : ((((charset) \ |
| 401 | - ((charset) < MIN_CHARSET_PRIVATE_DIMENSION2 ? 0x8F : 0xE0)) \ |
| 402 | << 14) \ |
| 403 | | ((c2) <= 0 ? 0 : ((c2) & 0x7F)) \ |
| 404 | | ((c1) <= 0 ? 0 : (((c1) & 0x7F) << 7)))))) |
| 405 | |
| 406 | |
| 407 | /* If GENERICP is nonzero, return nonzero iff C is a valid normal or |
| 408 | generic character. If GENERICP is zero, return nonzero iff C is a |
| 409 | valid normal character. */ |
| 410 | #define CHAR_VALID_P(c, genericp) \ |
| 411 | ((c) >= 0 \ |
| 412 | && (SINGLE_BYTE_CHAR_P (c) || char_valid_p (c, genericp))) |
| 413 | |
| 414 | /* This default value is used when nonascii-translation-table or |
| 415 | nonascii-insert-offset fail to convert unibyte character to a valid |
| 416 | multibyte character. This makes a Latin-1 character. */ |
| 417 | |
| 418 | #define DEFAULT_NONASCII_INSERT_OFFSET 0x800 |
| 419 | |
| 420 | /* Parse multibyte string STR of length LENGTH and set BYTES to the |
| 421 | byte length of a character at STR. */ |
| 422 | |
| 423 | #ifdef BYTE_COMBINING_DEBUG |
| 424 | |
| 425 | #define PARSE_MULTIBYTE_SEQ(str, length, bytes) \ |
| 426 | do { \ |
| 427 | int i = 1; \ |
| 428 | while (i < (length) && ! CHAR_HEAD_P ((str)[i])) i++; \ |
| 429 | (bytes) = BYTES_BY_CHAR_HEAD ((str)[0]); \ |
| 430 | if ((bytes) > i) \ |
| 431 | abort (); \ |
| 432 | } while (0) |
| 433 | |
| 434 | #else /* not BYTE_COMBINING_DEBUG */ |
| 435 | |
| 436 | #define PARSE_MULTIBYTE_SEQ(str, length, bytes) \ |
| 437 | (bytes) = BYTES_BY_CHAR_HEAD ((str)[0]) |
| 438 | |
| 439 | #endif /* not BYTE_COMBINING_DEBUG */ |
| 440 | |
| 441 | /* Return 1 iff the byte sequence at unibyte string STR (LENGTH bytes) |
| 442 | is valid as a multibyte form. If valid, by a side effect, BYTES is |
| 443 | set to the byte length of the multibyte form. */ |
| 444 | |
| 445 | #define UNIBYTE_STR_AS_MULTIBYTE_P(str, length, bytes) \ |
| 446 | (((str)[0] < 0x80 || (str)[0] >= 0xA0) \ |
| 447 | ? (bytes) = 1 \ |
| 448 | : (((bytes) = BYTES_BY_CHAR_HEAD ((str)[0])), \ |
| 449 | ((str)[0] != LEADING_CODE_8_BIT_CONTROL \ |
| 450 | && (bytes) <= (length) \ |
| 451 | && !CHAR_HEAD_P ((str)[1]) \ |
| 452 | && ((bytes) == 2 \ |
| 453 | || (!CHAR_HEAD_P ((str)[2]) \ |
| 454 | && ((bytes) == 3 \ |
| 455 | || !CHAR_HEAD_P ((str)[3]))))))) |
| 456 | |
| 457 | /* Return 1 iff the byte sequence at multibyte string STR is valid as |
| 458 | a unibyte form. By a side effect, BYTES is set to the byte length |
| 459 | of one character at STR. */ |
| 460 | |
| 461 | #define MULTIBYTE_STR_AS_UNIBYTE_P(str, bytes) \ |
| 462 | ((bytes) = BYTES_BY_CHAR_HEAD ((str)[0]), \ |
| 463 | (str)[0] != LEADING_CODE_8_BIT_CONTROL) |
| 464 | |
| 465 | /* The charset of character C is stored in CHARSET, and the |
| 466 | position-codes of C are stored in C1 and C2. |
| 467 | We store -1 in C2 if the dimension of the charset is 1. */ |
| 468 | |
| 469 | #define SPLIT_CHAR(c, charset, c1, c2) \ |
| 470 | (SINGLE_BYTE_CHAR_P (c) \ |
| 471 | ? ((charset = ASCII_BYTE_P (c) \ |
| 472 | ? CHARSET_ASCII \ |
| 473 | : (c) < 0xA0 ? CHARSET_8_BIT_CONTROL : CHARSET_8_BIT_GRAPHIC), \ |
| 474 | c1 = (c), c2 = -1) \ |
| 475 | : ((c) & CHAR_FIELD1_MASK \ |
| 476 | ? (charset = (CHAR_FIELD1 (c) \ |
| 477 | + ((c) < MIN_CHAR_PRIVATE_DIMENSION2 ? 0x8F : 0xE0)), \ |
| 478 | c1 = CHAR_FIELD2 (c), \ |
| 479 | c2 = CHAR_FIELD3 (c)) \ |
| 480 | : (charset = CHAR_FIELD2 (c) + 0x70, \ |
| 481 | c1 = CHAR_FIELD3 (c), \ |
| 482 | c2 = -1))) |
| 483 | |
| 484 | /* Return 1 iff character C has valid printable glyph. */ |
| 485 | #define CHAR_PRINTABLE_P(c) (ASCII_BYTE_P (c) || char_printable_p (c)) |
| 486 | |
| 487 | /* The charset of the character at STR is stored in CHARSET, and the |
| 488 | position-codes are stored in C1 and C2. |
| 489 | We store -1 in C2 if the character is just 2 bytes. */ |
| 490 | |
| 491 | #define SPLIT_STRING(str, len, charset, c1, c2) \ |
| 492 | ((BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) < 2 \ |
| 493 | || BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) > len \ |
| 494 | || split_string (str, len, &charset, &c1, &c2) < 0) \ |
| 495 | ? c1 = *(str), charset = CHARSET_ASCII \ |
| 496 | : charset) |
| 497 | |
| 498 | /* Mapping table from ISO2022's charset (specified by DIMENSION, |
| 499 | CHARS, and FINAL_CHAR) to Emacs' charset. Should be accessed by |
| 500 | macro ISO_CHARSET_TABLE (DIMENSION, CHARS, FINAL_CHAR). */ |
| 501 | extern int iso_charset_table[2][2][128]; |
| 502 | |
| 503 | #define ISO_CHARSET_TABLE(dimension, chars, final_char) \ |
| 504 | iso_charset_table[XINT (dimension) - 1][XINT (chars) > 94][XINT (final_char)] |
| 505 | |
| 506 | #define BASE_LEADING_CODE_P(c) (BYTES_BY_CHAR_HEAD ((unsigned char) (c)) > 1) |
| 507 | |
| 508 | /* Return how many bytes C will occupy in a multibyte buffer. */ |
| 509 | #define CHAR_BYTES(c) \ |
| 510 | (SINGLE_BYTE_CHAR_P (c) \ |
| 511 | ? ((ASCII_BYTE_P (c) || (c) >= 0xA0) ? 1 : 2) \ |
| 512 | : char_bytes (c)) |
| 513 | |
| 514 | /* The following two macros CHAR_STRING and STRING_CHAR are the main |
| 515 | entry points to convert between Emacs two types of character |
| 516 | representations: multi-byte form and single-word form (character |
| 517 | code). */ |
| 518 | |
| 519 | /* Store multi-byte form of the character C in STR. The caller should |
| 520 | allocate at least MAX_MULTIBYTE_LENGTH bytes area at STR in |
| 521 | advance. Returns the length of the multi-byte form. If C is an |
| 522 | invalid character code, signal an error. */ |
| 523 | |
| 524 | #define CHAR_STRING(c, str) \ |
| 525 | (ASCII_BYTE_P (c) \ |
| 526 | ? (*(str) = (unsigned char)(c), 1) \ |
| 527 | : char_to_string (c, (unsigned char *) str)) |
| 528 | |
| 529 | /* Return a character code of the character of which multi-byte form |
| 530 | is at STR and the length is LEN. If STR doesn't contain valid |
| 531 | multi-byte form, only the first byte in STR is returned. */ |
| 532 | |
| 533 | #define STRING_CHAR(str, len) \ |
| 534 | (BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) == 1 \ |
| 535 | ? (unsigned char) *(str) \ |
| 536 | : string_to_char (str, len, 0)) |
| 537 | |
| 538 | /* This is like STRING_CHAR but the third arg ACTUAL_LEN is set to the |
| 539 | length of the multi-byte form. Just to know the length, use |
| 540 | MULTIBYTE_FORM_LENGTH. */ |
| 541 | |
| 542 | #define STRING_CHAR_AND_LENGTH(str, len, actual_len) \ |
| 543 | (BYTES_BY_CHAR_HEAD ((unsigned char) *(str)) == 1 \ |
| 544 | ? ((actual_len) = 1), (unsigned char) *(str) \ |
| 545 | : string_to_char (str, len, &(actual_len))) |
| 546 | |
| 547 | /* Fetch the "next" character from Lisp string STRING at byte position |
| 548 | BYTEIDX, character position CHARIDX. Store it into OUTPUT. |
| 549 | |
| 550 | All the args must be side-effect-free. |
| 551 | BYTEIDX and CHARIDX must be lvalues; |
| 552 | we increment them past the character fetched. */ |
| 553 | |
| 554 | #define FETCH_STRING_CHAR_ADVANCE(OUTPUT, STRING, CHARIDX, BYTEIDX) \ |
| 555 | if (1) \ |
| 556 | { \ |
| 557 | CHARIDX++; \ |
| 558 | if (STRING_MULTIBYTE (STRING)) \ |
| 559 | { \ |
| 560 | unsigned char *ptr = &XSTRING (STRING)->data[BYTEIDX]; \ |
| 561 | int space_left = XSTRING (STRING)->size_byte - BYTEIDX; \ |
| 562 | int actual_len; \ |
| 563 | \ |
| 564 | OUTPUT = STRING_CHAR_AND_LENGTH (ptr, space_left, actual_len); \ |
| 565 | BYTEIDX += actual_len; \ |
| 566 | } \ |
| 567 | else \ |
| 568 | OUTPUT = XSTRING (STRING)->data[BYTEIDX++]; \ |
| 569 | } \ |
| 570 | else |
| 571 | |
| 572 | /* Like FETCH_STRING_CHAR_ADVANCE but assume STRING is multibyte. */ |
| 573 | |
| 574 | #define FETCH_STRING_CHAR_ADVANCE_NO_CHECK(OUTPUT, STRING, CHARIDX, BYTEIDX) \ |
| 575 | if (1) \ |
| 576 | { \ |
| 577 | unsigned char *fetch_string_char_ptr = &XSTRING (STRING)->data[BYTEIDX]; \ |
| 578 | int fetch_string_char_space_left = XSTRING (STRING)->size_byte - BYTEIDX; \ |
| 579 | int actual_len; \ |
| 580 | \ |
| 581 | OUTPUT \ |
| 582 | = STRING_CHAR_AND_LENGTH (fetch_string_char_ptr, \ |
| 583 | fetch_string_char_space_left, actual_len); \ |
| 584 | \ |
| 585 | BYTEIDX += actual_len; \ |
| 586 | CHARIDX++; \ |
| 587 | } \ |
| 588 | else |
| 589 | |
| 590 | /* Like FETCH_STRING_CHAR_ADVANCE but fetch character from the current |
| 591 | buffer. */ |
| 592 | |
| 593 | #define FETCH_CHAR_ADVANCE(OUTPUT, CHARIDX, BYTEIDX) \ |
| 594 | if (1) \ |
| 595 | { \ |
| 596 | CHARIDX++; \ |
| 597 | if (!NILP (current_buffer->enable_multibyte_characters)) \ |
| 598 | { \ |
| 599 | unsigned char *ptr = BYTE_POS_ADDR (BYTEIDX); \ |
| 600 | int space_left = ((CHARIDX < GPT ? GPT_BYTE : Z_BYTE) - BYTEIDX); \ |
| 601 | int actual_len; \ |
| 602 | \ |
| 603 | OUTPUT= STRING_CHAR_AND_LENGTH (ptr, space_left, actual_len); \ |
| 604 | BYTEIDX += actual_len; \ |
| 605 | } \ |
| 606 | else \ |
| 607 | { \ |
| 608 | OUTPUT = *(BYTE_POS_ADDR (BYTEIDX)); \ |
| 609 | BYTEIDX++; \ |
| 610 | } \ |
| 611 | } \ |
| 612 | else |
| 613 | |
| 614 | /* Return the length of the multi-byte form at string STR of length LEN. */ |
| 615 | |
| 616 | #define MULTIBYTE_FORM_LENGTH(str, len) \ |
| 617 | (BYTES_BY_CHAR_HEAD (*(unsigned char *)(str)) == 1 \ |
| 618 | ? 1 \ |
| 619 | : multibyte_form_length (str, len)) |
| 620 | |
| 621 | #ifdef emacs |
| 622 | |
| 623 | /* Increase the buffer byte position POS_BYTE of the current buffer to |
| 624 | the next character boundary. This macro relies on the fact that |
| 625 | *GPT_ADDR and *Z_ADDR are always accessible and the values are |
| 626 | '\0'. No range checking of POS. */ |
| 627 | |
| 628 | #ifdef BYTE_COMBINING_DEBUG |
| 629 | |
| 630 | #define INC_POS(pos_byte) \ |
| 631 | do { \ |
| 632 | unsigned char *p = BYTE_POS_ADDR (pos_byte); \ |
| 633 | if (BASE_LEADING_CODE_P (*p)) \ |
| 634 | { \ |
| 635 | int len, bytes; \ |
| 636 | len = Z_BYTE - pos_byte; \ |
| 637 | PARSE_MULTIBYTE_SEQ (p, len, bytes); \ |
| 638 | pos_byte += bytes; \ |
| 639 | } \ |
| 640 | else \ |
| 641 | pos_byte++; \ |
| 642 | } while (0) |
| 643 | |
| 644 | #else /* not BYTE_COMBINING_DEBUG */ |
| 645 | |
| 646 | #define INC_POS(pos_byte) \ |
| 647 | do { \ |
| 648 | unsigned char *p = BYTE_POS_ADDR (pos_byte); \ |
| 649 | pos_byte += BYTES_BY_CHAR_HEAD (*p); \ |
| 650 | } while (0) |
| 651 | |
| 652 | #endif /* not BYTE_COMBINING_DEBUG */ |
| 653 | |
| 654 | /* Decrease the buffer byte position POS_BYTE of the current buffer to |
| 655 | the previous character boundary. No range checking of POS. */ |
| 656 | #define DEC_POS(pos_byte) \ |
| 657 | do { \ |
| 658 | unsigned char *p, *p_min; \ |
| 659 | \ |
| 660 | pos_byte--; \ |
| 661 | if (pos_byte < GPT_BYTE) \ |
| 662 | p = BEG_ADDR + pos_byte - 1, p_min = BEG_ADDR; \ |
| 663 | else \ |
| 664 | p = BEG_ADDR + GAP_SIZE + pos_byte - 1, p_min = GAP_END_ADDR; \ |
| 665 | if (p > p_min && !CHAR_HEAD_P (*p)) \ |
| 666 | { \ |
| 667 | unsigned char *pend = p--; \ |
| 668 | int len, bytes; \ |
| 669 | while (p > p_min && !CHAR_HEAD_P (*p)) p--; \ |
| 670 | len = pend + 1 - p; \ |
| 671 | PARSE_MULTIBYTE_SEQ (p, len, bytes); \ |
| 672 | if (bytes == len) \ |
| 673 | pos_byte -= len - 1; \ |
| 674 | } \ |
| 675 | } while (0) |
| 676 | |
| 677 | /* Increment both CHARPOS and BYTEPOS, each in the appropriate way. */ |
| 678 | |
| 679 | #define INC_BOTH(charpos, bytepos) \ |
| 680 | do \ |
| 681 | { \ |
| 682 | (charpos)++; \ |
| 683 | if (NILP (current_buffer->enable_multibyte_characters)) \ |
| 684 | (bytepos)++; \ |
| 685 | else \ |
| 686 | INC_POS ((bytepos)); \ |
| 687 | } \ |
| 688 | while (0) |
| 689 | |
| 690 | /* Decrement both CHARPOS and BYTEPOS, each in the appropriate way. */ |
| 691 | |
| 692 | #define DEC_BOTH(charpos, bytepos) \ |
| 693 | do \ |
| 694 | { \ |
| 695 | (charpos)--; \ |
| 696 | if (NILP (current_buffer->enable_multibyte_characters)) \ |
| 697 | (bytepos)--; \ |
| 698 | else \ |
| 699 | DEC_POS ((bytepos)); \ |
| 700 | } \ |
| 701 | while (0) |
| 702 | |
| 703 | /* Increase the buffer byte position POS_BYTE of the current buffer to |
| 704 | the next character boundary. This macro relies on the fact that |
| 705 | *GPT_ADDR and *Z_ADDR are always accessible and the values are |
| 706 | '\0'. No range checking of POS_BYTE. */ |
| 707 | |
| 708 | #ifdef BYTE_COMBINING_DEBUG |
| 709 | |
| 710 | #define BUF_INC_POS(buf, pos_byte) \ |
| 711 | do { \ |
| 712 | unsigned char *p = BUF_BYTE_ADDRESS (buf, pos_byte); \ |
| 713 | if (BASE_LEADING_CODE_P (*p)) \ |
| 714 | { \ |
| 715 | int len, bytes; \ |
| 716 | len = BUF_Z_BYTE (buf) - pos_byte; \ |
| 717 | PARSE_MULTIBYTE_SEQ (p, len, bytes); \ |
| 718 | pos_byte += bytes; \ |
| 719 | } \ |
| 720 | else \ |
| 721 | pos_byte++; \ |
| 722 | } while (0) |
| 723 | |
| 724 | #else /* not BYTE_COMBINING_DEBUG */ |
| 725 | |
| 726 | #define BUF_INC_POS(buf, pos_byte) \ |
| 727 | do { \ |
| 728 | unsigned char *p = BUF_BYTE_ADDRESS (buf, pos_byte); \ |
| 729 | pos_byte += BYTES_BY_CHAR_HEAD (*p); \ |
| 730 | } while (0) |
| 731 | |
| 732 | #endif /* not BYTE_COMBINING_DEBUG */ |
| 733 | |
| 734 | /* Decrease the buffer byte position POS_BYTE of the current buffer to |
| 735 | the previous character boundary. No range checking of POS_BYTE. */ |
| 736 | #define BUF_DEC_POS(buf, pos_byte) \ |
| 737 | do { \ |
| 738 | unsigned char *p, *p_min; \ |
| 739 | pos_byte--; \ |
| 740 | if (pos_byte < BUF_GPT_BYTE (buf)) \ |
| 741 | { \ |
| 742 | p = BUF_BEG_ADDR (buf) + pos_byte - 1; \ |
| 743 | p_min = BUF_BEG_ADDR (buf); \ |
| 744 | } \ |
| 745 | else \ |
| 746 | { \ |
| 747 | p = BUF_BEG_ADDR (buf) + BUF_GAP_SIZE (buf) + pos_byte - 1; \ |
| 748 | p_min = BUF_GAP_END_ADDR (buf); \ |
| 749 | } \ |
| 750 | if (p > p_min && !CHAR_HEAD_P (*p)) \ |
| 751 | { \ |
| 752 | unsigned char *pend = p--; \ |
| 753 | int len, bytes; \ |
| 754 | while (p > p_min && !CHAR_HEAD_P (*p)) p--; \ |
| 755 | len = pend + 1 - p; \ |
| 756 | PARSE_MULTIBYTE_SEQ (p, len, bytes); \ |
| 757 | if (bytes == len) \ |
| 758 | pos_byte -= len - 1; \ |
| 759 | } \ |
| 760 | } while (0) |
| 761 | |
| 762 | #endif /* emacs */ |
| 763 | |
| 764 | /* This is the maximum byte length of multi-byte sequence. */ |
| 765 | #define MAX_MULTIBYTE_LENGTH 4 |
| 766 | |
| 767 | extern void invalid_character P_ ((int)); |
| 768 | |
| 769 | extern int translate_char P_ ((Lisp_Object, int, int, int, int)); |
| 770 | extern int split_string P_ ((const unsigned char *, int, int *, |
| 771 | unsigned char *, unsigned char *)); |
| 772 | extern int char_to_string P_ ((int, unsigned char *)); |
| 773 | extern int string_to_char P_ ((const unsigned char *, int, int *)); |
| 774 | extern int char_printable_p P_ ((int c)); |
| 775 | extern int multibyte_form_length P_ ((const unsigned char *, int)); |
| 776 | extern void parse_str_as_multibyte P_ ((unsigned char *, int, int *, int *)); |
| 777 | extern int str_as_multibyte P_ ((unsigned char *, int, int, int *)); |
| 778 | extern int str_to_multibyte P_ ((unsigned char *, int, int)); |
| 779 | extern int str_as_unibyte P_ ((unsigned char *, int)); |
| 780 | extern int get_charset_id P_ ((Lisp_Object)); |
| 781 | extern int find_charset_in_text P_ ((unsigned char *, int, int, int *, |
| 782 | Lisp_Object)); |
| 783 | extern int strwidth P_ ((unsigned char *, int)); |
| 784 | extern int char_bytes P_ ((int)); |
| 785 | extern int char_valid_p P_ ((int, int)); |
| 786 | |
| 787 | extern Lisp_Object Vtranslation_table_vector; |
| 788 | |
| 789 | /* Return a translation table of id number ID. */ |
| 790 | #define GET_TRANSLATION_TABLE(id) \ |
| 791 | (XCDR(XVECTOR(Vtranslation_table_vector)->contents[(id)])) |
| 792 | |
| 793 | /* A char-table for characters which may invoke auto-filling. */ |
| 794 | extern Lisp_Object Vauto_fill_chars; |
| 795 | |
| 796 | /* Copy LEN bytes from FROM to TO. This macro should be used only |
| 797 | when a caller knows that LEN is short and the obvious copy loop is |
| 798 | faster than calling bcopy which has some overhead. Copying a |
| 799 | multibyte sequence of a multibyte character is the typical case. */ |
| 800 | |
| 801 | #define BCOPY_SHORT(from, to, len) \ |
| 802 | do { \ |
| 803 | int i = len; \ |
| 804 | unsigned char *from_p = from, *to_p = to; \ |
| 805 | while (i--) *to_p++ = *from_p++; \ |
| 806 | } while (0) |
| 807 | |
| 808 | #endif /* EMACS_CHARSET_H */ |