1 /* vi:set ts=8 sts=4 sw=4 noet: 2 * 3 * VIM - Vi IMproved by Bram Moolenaar 4 * Multibyte extensions partly by Sung-Hoon Baek 5 * 6 * Do ":help uganda" in Vim to read copying and usage conditions. 7 * Do ":help credits" in Vim to see a list of people who contributed. 8 * See README.txt for an overview of the Vim source code. 9 */ 10 /* 11 * mbyte.c: Code specifically for handling multi-byte characters. 12 * 13 * The encoding used in the core is set with 'encoding'. When 'encoding' is 14 * changed, the following four variables are set (for speed). 15 * Currently these types of character encodings are supported: 16 * 17 * "enc_dbcs" When non-zero it tells the type of double byte character 18 * encoding (Chinese, Korean, Japanese, etc.). 19 * The cell width on the display is equal to the number of 20 * bytes. (exception: DBCS_JPNU with first byte 0x8e) 21 * Recognizing the first or second byte is difficult, it 22 * requires checking a byte sequence from the start. 23 * "enc_utf8" When TRUE use Unicode characters in UTF-8 encoding. 24 * The cell width on the display needs to be determined from 25 * the character value. 26 * Recognizing bytes is easy: 0xxx.xxxx is a single-byte 27 * char, 10xx.xxxx is a trailing byte, 11xx.xxxx is a leading 28 * byte of a multi-byte character. 29 * To make things complicated, up to six composing characters 30 * are allowed. These are drawn on top of the first char. 31 * For most editing the sequence of bytes with composing 32 * characters included is considered to be one character. 33 * "enc_unicode" When 2 use 16-bit Unicode characters (or UTF-16). 34 * When 4 use 32-but Unicode characters. 35 * Internally characters are stored in UTF-8 encoding to 36 * avoid NUL bytes. Conversion happens when doing I/O. 37 * "enc_utf8" will also be TRUE. 38 * 39 * "has_mbyte" is set when "enc_dbcs" or "enc_utf8" is non-zero. 40 * 41 * If none of these is TRUE, 8-bit bytes are used for a character. The 42 * encoding isn't currently specified (TODO). 43 * 44 * 'encoding' specifies the encoding used in the core. This is in registers, 45 * text manipulation, buffers, etc. Conversion has to be done when characters 46 * in another encoding are received or send: 47 * 48 * clipboard 49 * ^ 50 * | (2) 51 * V 52 * +---------------+ 53 * (1) | | (3) 54 * keyboard ----->| core |-----> display 55 * | | 56 * +---------------+ 57 * ^ 58 * | (4) 59 * V 60 * file 61 * 62 * (1) Typed characters arrive in the current locale. Conversion is to be 63 * done when 'encoding' is different from 'termencoding'. 64 * (2) Text will be made available with the encoding specified with 65 * 'encoding'. If this is not sufficient, system-specific conversion 66 * might be required. 67 * (3) For the GUI the correct font must be selected, no conversion done. 68 * Otherwise, conversion is to be done when 'encoding' differs from 69 * 'termencoding'. (Different in the GTK+ 2 port -- 'termencoding' 70 * is always used for both input and output and must always be set to 71 * "utf-8". gui_mch_init() does this automatically.) 72 * (4) The encoding of the file is specified with 'fileencoding'. Conversion 73 * is to be done when it's different from 'encoding'. 74 * 75 * The viminfo file is a special case: Only text is converted, not file names. 76 * Vim scripts may contain an ":encoding" command. This has an effect for 77 * some commands, like ":menutrans" 78 */ 79 80 #include "vim.h" 81 82 #ifdef WIN32UNIX 83 # ifndef WIN32_LEAN_AND_MEAN 84 # define WIN32_LEAN_AND_MEAN 85 # endif 86 # if defined(FEAT_GUI) || defined(FEAT_XCLIPBOARD) 87 # include <X11/Xwindows.h> 88 # define WINBYTE wBYTE 89 # else 90 # include <windows.h> 91 # define WINBYTE BYTE 92 # endif 93 # ifdef WIN32 94 # undef WIN32 /* Some windows.h define WIN32, we don't want that here. */ 95 # endif 96 #else 97 # define WINBYTE BYTE 98 #endif 99 100 #if (defined(MSWIN) || defined(WIN32UNIX)) && !defined(__MINGW32__) 101 # include <winnls.h> 102 #endif 103 104 #ifdef FEAT_GUI_X11 105 # include <X11/Intrinsic.h> 106 #endif 107 #ifdef X_LOCALE 108 # include <X11/Xlocale.h> 109 # if !defined(HAVE_MBLEN) && !defined(mblen) 110 # define mblen _Xmblen 111 # endif 112 #endif 113 114 #if defined(FEAT_GUI_GTK) && defined(FEAT_XIM) 115 # if GTK_CHECK_VERSION(3,0,0) 116 # include <gdk/gdkkeysyms-compat.h> 117 # else 118 # include <gdk/gdkkeysyms.h> 119 # endif 120 # ifdef MSWIN 121 # include <gdk/gdkwin32.h> 122 # else 123 # include <gdk/gdkx.h> 124 # endif 125 #endif 126 127 #ifdef HAVE_WCHAR_H 128 # include <wchar.h> 129 #endif 130 131 #if 0 132 /* This has been disabled, because several people reported problems with the 133 * wcwidth() and iswprint() library functions, esp. for Hebrew. */ 134 # ifdef __STDC_ISO_10646__ 135 # define USE_WCHAR_FUNCTIONS 136 # endif 137 #endif 138 139 static int dbcs_char2len(int c); 140 static int dbcs_char2bytes(int c, char_u *buf); 141 static int dbcs_ptr2len(char_u *p); 142 static int dbcs_ptr2len_len(char_u *p, int size); 143 static int utf_ptr2cells_len(char_u *p, int size); 144 static int dbcs_char2cells(int c); 145 static int dbcs_ptr2cells_len(char_u *p, int size); 146 static int dbcs_ptr2char(char_u *p); 147 148 /* 149 * Lookup table to quickly get the length in bytes of a UTF-8 character from 150 * the first byte of a UTF-8 string. 151 * Bytes which are illegal when used as the first byte have a 1. 152 * The NUL byte has length 1. 153 */ 154 static char utf8len_tab[256] = 155 { 156 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 157 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 158 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 159 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 160 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 161 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 162 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 163 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,5,5,5,5,6,6,1,1, 164 }; 165 166 /* 167 * Like utf8len_tab above, but using a zero for illegal lead bytes. 168 */ 169 static char utf8len_tab_zero[256] = 170 { 171 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 172 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 173 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 174 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 175 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 176 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 177 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 178 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,5,5,5,5,6,6,0,0, 179 }; 180 181 /* 182 * XIM often causes trouble. Define XIM_DEBUG to get a log of XIM callbacks 183 * in the "xim.log" file. 184 */ 185 /* #define XIM_DEBUG */ 186 #ifdef XIM_DEBUG 187 static void 188 xim_log(char *s, ...) 189 { 190 va_list arglist; 191 static FILE *fd = NULL; 192 193 if (fd == (FILE *)-1) 194 return; 195 if (fd == NULL) 196 { 197 fd = mch_fopen("xim.log", "w"); 198 if (fd == NULL) 199 { 200 emsg("Cannot open xim.log"); 201 fd = (FILE *)-1; 202 return; 203 } 204 } 205 206 va_start(arglist, s); 207 vfprintf(fd, s, arglist); 208 va_end(arglist); 209 } 210 #endif 211 212 213 /* 214 * Canonical encoding names and their properties. 215 * "iso-8859-n" is handled by enc_canonize() directly. 216 */ 217 static struct 218 { char *name; int prop; int codepage;} 219 enc_canon_table[] = 220 { 221 #define IDX_LATIN_1 0 222 {"latin1", ENC_8BIT + ENC_LATIN1, 1252}, 223 #define IDX_ISO_2 1 224 {"iso-8859-2", ENC_8BIT, 0}, 225 #define IDX_ISO_3 2 226 {"iso-8859-3", ENC_8BIT, 0}, 227 #define IDX_ISO_4 3 228 {"iso-8859-4", ENC_8BIT, 0}, 229 #define IDX_ISO_5 4 230 {"iso-8859-5", ENC_8BIT, 0}, 231 #define IDX_ISO_6 5 232 {"iso-8859-6", ENC_8BIT, 0}, 233 #define IDX_ISO_7 6 234 {"iso-8859-7", ENC_8BIT, 0}, 235 #define IDX_ISO_8 7 236 {"iso-8859-8", ENC_8BIT, 0}, 237 #define IDX_ISO_9 8 238 {"iso-8859-9", ENC_8BIT, 0}, 239 #define IDX_ISO_10 9 240 {"iso-8859-10", ENC_8BIT, 0}, 241 #define IDX_ISO_11 10 242 {"iso-8859-11", ENC_8BIT, 0}, 243 #define IDX_ISO_13 11 244 {"iso-8859-13", ENC_8BIT, 0}, 245 #define IDX_ISO_14 12 246 {"iso-8859-14", ENC_8BIT, 0}, 247 #define IDX_ISO_15 13 248 {"iso-8859-15", ENC_8BIT + ENC_LATIN9, 0}, 249 #define IDX_KOI8_R 14 250 {"koi8-r", ENC_8BIT, 0}, 251 #define IDX_KOI8_U 15 252 {"koi8-u", ENC_8BIT, 0}, 253 #define IDX_UTF8 16 254 {"utf-8", ENC_UNICODE, 0}, 255 #define IDX_UCS2 17 256 {"ucs-2", ENC_UNICODE + ENC_ENDIAN_B + ENC_2BYTE, 0}, 257 #define IDX_UCS2LE 18 258 {"ucs-2le", ENC_UNICODE + ENC_ENDIAN_L + ENC_2BYTE, 0}, 259 #define IDX_UTF16 19 260 {"utf-16", ENC_UNICODE + ENC_ENDIAN_B + ENC_2WORD, 0}, 261 #define IDX_UTF16LE 20 262 {"utf-16le", ENC_UNICODE + ENC_ENDIAN_L + ENC_2WORD, 0}, 263 #define IDX_UCS4 21 264 {"ucs-4", ENC_UNICODE + ENC_ENDIAN_B + ENC_4BYTE, 0}, 265 #define IDX_UCS4LE 22 266 {"ucs-4le", ENC_UNICODE + ENC_ENDIAN_L + ENC_4BYTE, 0}, 267 268 /* For debugging DBCS encoding on Unix. */ 269 #define IDX_DEBUG 23 270 {"debug", ENC_DBCS, DBCS_DEBUG}, 271 #define IDX_EUC_JP 24 272 {"euc-jp", ENC_DBCS, DBCS_JPNU}, 273 #define IDX_SJIS 25 274 {"sjis", ENC_DBCS, DBCS_JPN}, 275 #define IDX_EUC_KR 26 276 {"euc-kr", ENC_DBCS, DBCS_KORU}, 277 #define IDX_EUC_CN 27 278 {"euc-cn", ENC_DBCS, DBCS_CHSU}, 279 #define IDX_EUC_TW 28 280 {"euc-tw", ENC_DBCS, DBCS_CHTU}, 281 #define IDX_BIG5 29 282 {"big5", ENC_DBCS, DBCS_CHT}, 283 284 /* MS-DOS and MS-Windows codepages are included here, so that they can be 285 * used on Unix too. Most of them are similar to ISO-8859 encodings, but 286 * not exactly the same. */ 287 #define IDX_CP437 30 288 {"cp437", ENC_8BIT, 437}, /* like iso-8859-1 */ 289 #define IDX_CP737 31 290 {"cp737", ENC_8BIT, 737}, /* like iso-8859-7 */ 291 #define IDX_CP775 32 292 {"cp775", ENC_8BIT, 775}, /* Baltic */ 293 #define IDX_CP850 33 294 {"cp850", ENC_8BIT, 850}, /* like iso-8859-4 */ 295 #define IDX_CP852 34 296 {"cp852", ENC_8BIT, 852}, /* like iso-8859-1 */ 297 #define IDX_CP855 35 298 {"cp855", ENC_8BIT, 855}, /* like iso-8859-2 */ 299 #define IDX_CP857 36 300 {"cp857", ENC_8BIT, 857}, /* like iso-8859-5 */ 301 #define IDX_CP860 37 302 {"cp860", ENC_8BIT, 860}, /* like iso-8859-9 */ 303 #define IDX_CP861 38 304 {"cp861", ENC_8BIT, 861}, /* like iso-8859-1 */ 305 #define IDX_CP862 39 306 {"cp862", ENC_8BIT, 862}, /* like iso-8859-1 */ 307 #define IDX_CP863 40 308 {"cp863", ENC_8BIT, 863}, /* like iso-8859-8 */ 309 #define IDX_CP865 41 310 {"cp865", ENC_8BIT, 865}, /* like iso-8859-1 */ 311 #define IDX_CP866 42 312 {"cp866", ENC_8BIT, 866}, /* like iso-8859-5 */ 313 #define IDX_CP869 43 314 {"cp869", ENC_8BIT, 869}, /* like iso-8859-7 */ 315 #define IDX_CP874 44 316 {"cp874", ENC_8BIT, 874}, /* Thai */ 317 #define IDX_CP932 45 318 {"cp932", ENC_DBCS, DBCS_JPN}, 319 #define IDX_CP936 46 320 {"cp936", ENC_DBCS, DBCS_CHS}, 321 #define IDX_CP949 47 322 {"cp949", ENC_DBCS, DBCS_KOR}, 323 #define IDX_CP950 48 324 {"cp950", ENC_DBCS, DBCS_CHT}, 325 #define IDX_CP1250 49 326 {"cp1250", ENC_8BIT, 1250}, /* Czech, Polish, etc. */ 327 #define IDX_CP1251 50 328 {"cp1251", ENC_8BIT, 1251}, /* Cyrillic */ 329 /* cp1252 is considered to be equal to latin1 */ 330 #define IDX_CP1253 51 331 {"cp1253", ENC_8BIT, 1253}, /* Greek */ 332 #define IDX_CP1254 52 333 {"cp1254", ENC_8BIT, 1254}, /* Turkish */ 334 #define IDX_CP1255 53 335 {"cp1255", ENC_8BIT, 1255}, /* Hebrew */ 336 #define IDX_CP1256 54 337 {"cp1256", ENC_8BIT, 1256}, /* Arabic */ 338 #define IDX_CP1257 55 339 {"cp1257", ENC_8BIT, 1257}, /* Baltic */ 340 #define IDX_CP1258 56 341 {"cp1258", ENC_8BIT, 1258}, /* Vietnamese */ 342 343 #define IDX_MACROMAN 57 344 {"macroman", ENC_8BIT + ENC_MACROMAN, 0}, /* Mac OS */ 345 #define IDX_DECMCS 58 346 {"dec-mcs", ENC_8BIT, 0}, /* DEC MCS */ 347 #define IDX_HPROMAN8 59 348 {"hp-roman8", ENC_8BIT, 0}, /* HP Roman8 */ 349 #define IDX_COUNT 60 350 }; 351 352 /* 353 * Aliases for encoding names. 354 */ 355 static struct 356 { char *name; int canon;} 357 enc_alias_table[] = 358 { 359 {"ansi", IDX_LATIN_1}, 360 {"iso-8859-1", IDX_LATIN_1}, 361 {"latin2", IDX_ISO_2}, 362 {"latin3", IDX_ISO_3}, 363 {"latin4", IDX_ISO_4}, 364 {"cyrillic", IDX_ISO_5}, 365 {"arabic", IDX_ISO_6}, 366 {"greek", IDX_ISO_7}, 367 #ifdef MSWIN 368 {"hebrew", IDX_CP1255}, 369 #else 370 {"hebrew", IDX_ISO_8}, 371 #endif 372 {"latin5", IDX_ISO_9}, 373 {"turkish", IDX_ISO_9}, /* ? */ 374 {"latin6", IDX_ISO_10}, 375 {"nordic", IDX_ISO_10}, /* ? */ 376 {"thai", IDX_ISO_11}, /* ? */ 377 {"latin7", IDX_ISO_13}, 378 {"latin8", IDX_ISO_14}, 379 {"latin9", IDX_ISO_15}, 380 {"utf8", IDX_UTF8}, 381 {"unicode", IDX_UCS2}, 382 {"ucs2", IDX_UCS2}, 383 {"ucs2be", IDX_UCS2}, 384 {"ucs-2be", IDX_UCS2}, 385 {"ucs2le", IDX_UCS2LE}, 386 {"utf16", IDX_UTF16}, 387 {"utf16be", IDX_UTF16}, 388 {"utf-16be", IDX_UTF16}, 389 {"utf16le", IDX_UTF16LE}, 390 {"ucs4", IDX_UCS4}, 391 {"ucs4be", IDX_UCS4}, 392 {"ucs-4be", IDX_UCS4}, 393 {"ucs4le", IDX_UCS4LE}, 394 {"utf32", IDX_UCS4}, 395 {"utf-32", IDX_UCS4}, 396 {"utf32be", IDX_UCS4}, 397 {"utf-32be", IDX_UCS4}, 398 {"utf32le", IDX_UCS4LE}, 399 {"utf-32le", IDX_UCS4LE}, 400 {"932", IDX_CP932}, 401 {"949", IDX_CP949}, 402 {"936", IDX_CP936}, 403 {"gbk", IDX_CP936}, 404 {"950", IDX_CP950}, 405 {"eucjp", IDX_EUC_JP}, 406 {"unix-jis", IDX_EUC_JP}, 407 {"ujis", IDX_EUC_JP}, 408 {"shift-jis", IDX_SJIS}, 409 {"pck", IDX_SJIS}, /* Sun: PCK */ 410 {"euckr", IDX_EUC_KR}, 411 {"5601", IDX_EUC_KR}, /* Sun: KS C 5601 */ 412 {"euccn", IDX_EUC_CN}, 413 {"gb2312", IDX_EUC_CN}, 414 {"euctw", IDX_EUC_TW}, 415 #if defined(MSWIN) || defined(WIN32UNIX) || defined(MACOS_X) 416 {"japan", IDX_CP932}, 417 {"korea", IDX_CP949}, 418 {"prc", IDX_CP936}, 419 {"chinese", IDX_CP936}, 420 {"taiwan", IDX_CP950}, 421 {"big5", IDX_CP950}, 422 #else 423 {"japan", IDX_EUC_JP}, 424 {"korea", IDX_EUC_KR}, 425 {"prc", IDX_EUC_CN}, 426 {"chinese", IDX_EUC_CN}, 427 {"taiwan", IDX_EUC_TW}, 428 {"cp950", IDX_BIG5}, 429 {"950", IDX_BIG5}, 430 #endif 431 {"mac", IDX_MACROMAN}, 432 {"mac-roman", IDX_MACROMAN}, 433 {NULL, 0} 434 }; 435 436 #ifndef CP_UTF8 437 # define CP_UTF8 65001 /* magic number from winnls.h */ 438 #endif 439 440 /* 441 * Find encoding "name" in the list of canonical encoding names. 442 * Returns -1 if not found. 443 */ 444 static int 445 enc_canon_search(char_u *name) 446 { 447 int i; 448 449 for (i = 0; i < IDX_COUNT; ++i) 450 if (STRCMP(name, enc_canon_table[i].name) == 0) 451 return i; 452 return -1; 453 } 454 455 456 /* 457 * Find canonical encoding "name" in the list and return its properties. 458 * Returns 0 if not found. 459 */ 460 int 461 enc_canon_props(char_u *name) 462 { 463 int i; 464 465 i = enc_canon_search(name); 466 if (i >= 0) 467 return enc_canon_table[i].prop; 468 #ifdef MSWIN 469 if (name[0] == 'c' && name[1] == 'p' && VIM_ISDIGIT(name[2])) 470 { 471 CPINFO cpinfo; 472 473 /* Get info on this codepage to find out what it is. */ 474 if (GetCPInfo(atoi((char *)name + 2), &cpinfo) != 0) 475 { 476 if (cpinfo.MaxCharSize == 1) /* some single-byte encoding */ 477 return ENC_8BIT; 478 if (cpinfo.MaxCharSize == 2 479 && (cpinfo.LeadByte[0] != 0 || cpinfo.LeadByte[1] != 0)) 480 /* must be a DBCS encoding */ 481 return ENC_DBCS; 482 } 483 return 0; 484 } 485 #endif 486 if (STRNCMP(name, "2byte-", 6) == 0) 487 return ENC_DBCS; 488 if (STRNCMP(name, "8bit-", 5) == 0 || STRNCMP(name, "iso-8859-", 9) == 0) 489 return ENC_8BIT; 490 return 0; 491 } 492 493 /* 494 * Set up for using multi-byte characters. 495 * Called in three cases: 496 * - by main() to initialize (p_enc == NULL) 497 * - by set_init_1() after 'encoding' was set to its default. 498 * - by do_set() when 'encoding' has been set. 499 * p_enc must have been passed through enc_canonize() already. 500 * Sets the "enc_unicode", "enc_utf8", "enc_dbcs" and "has_mbyte" flags. 501 * Fills mb_bytelen_tab[] and returns NULL when there are no problems. 502 * When there is something wrong: Returns an error message and doesn't change 503 * anything. 504 */ 505 char * 506 mb_init(void) 507 { 508 int i; 509 int idx; 510 int n; 511 int enc_dbcs_new = 0; 512 #if defined(USE_ICONV) && !defined(MSWIN) && !defined(WIN32UNIX) \ 513 && !defined(MACOS_CONVERT) 514 # define LEN_FROM_CONV 515 vimconv_T vimconv; 516 char_u *p; 517 #endif 518 519 if (p_enc == NULL) 520 { 521 /* Just starting up: set the whole table to one's. */ 522 for (i = 0; i < 256; ++i) 523 mb_bytelen_tab[i] = 1; 524 input_conv.vc_type = CONV_NONE; 525 input_conv.vc_factor = 1; 526 output_conv.vc_type = CONV_NONE; 527 return NULL; 528 } 529 530 #ifdef MSWIN 531 if (p_enc[0] == 'c' && p_enc[1] == 'p' && VIM_ISDIGIT(p_enc[2])) 532 { 533 CPINFO cpinfo; 534 535 /* Get info on this codepage to find out what it is. */ 536 if (GetCPInfo(atoi((char *)p_enc + 2), &cpinfo) != 0) 537 { 538 if (cpinfo.MaxCharSize == 1) 539 { 540 /* some single-byte encoding */ 541 enc_unicode = 0; 542 enc_utf8 = FALSE; 543 } 544 else if (cpinfo.MaxCharSize == 2 545 && (cpinfo.LeadByte[0] != 0 || cpinfo.LeadByte[1] != 0)) 546 { 547 /* must be a DBCS encoding, check below */ 548 enc_dbcs_new = atoi((char *)p_enc + 2); 549 } 550 else 551 goto codepage_invalid; 552 } 553 else if (GetLastError() == ERROR_INVALID_PARAMETER) 554 { 555 codepage_invalid: 556 return N_("E543: Not a valid codepage"); 557 } 558 } 559 #endif 560 else if (STRNCMP(p_enc, "8bit-", 5) == 0 561 || STRNCMP(p_enc, "iso-8859-", 9) == 0) 562 { 563 /* Accept any "8bit-" or "iso-8859-" name. */ 564 enc_unicode = 0; 565 enc_utf8 = FALSE; 566 } 567 else if (STRNCMP(p_enc, "2byte-", 6) == 0) 568 { 569 #ifdef MSWIN 570 /* Windows: accept only valid codepage numbers, check below. */ 571 if (p_enc[6] != 'c' || p_enc[7] != 'p' 572 || (enc_dbcs_new = atoi((char *)p_enc + 8)) == 0) 573 return e_invarg; 574 #else 575 /* Unix: accept any "2byte-" name, assume current locale. */ 576 enc_dbcs_new = DBCS_2BYTE; 577 #endif 578 } 579 else if ((idx = enc_canon_search(p_enc)) >= 0) 580 { 581 i = enc_canon_table[idx].prop; 582 if (i & ENC_UNICODE) 583 { 584 /* Unicode */ 585 enc_utf8 = TRUE; 586 if (i & (ENC_2BYTE | ENC_2WORD)) 587 enc_unicode = 2; 588 else if (i & ENC_4BYTE) 589 enc_unicode = 4; 590 else 591 enc_unicode = 0; 592 } 593 else if (i & ENC_DBCS) 594 { 595 /* 2byte, handle below */ 596 enc_dbcs_new = enc_canon_table[idx].codepage; 597 } 598 else 599 { 600 /* Must be 8-bit. */ 601 enc_unicode = 0; 602 enc_utf8 = FALSE; 603 } 604 } 605 else /* Don't know what encoding this is, reject it. */ 606 return e_invarg; 607 608 if (enc_dbcs_new != 0) 609 { 610 #ifdef MSWIN 611 /* Check if the DBCS code page is OK. */ 612 if (!IsValidCodePage(enc_dbcs_new)) 613 goto codepage_invalid; 614 #endif 615 enc_unicode = 0; 616 enc_utf8 = FALSE; 617 } 618 enc_dbcs = enc_dbcs_new; 619 has_mbyte = (enc_dbcs != 0 || enc_utf8); 620 621 #if defined(MSWIN) || defined(FEAT_CYGWIN_WIN32_CLIPBOARD) 622 enc_codepage = encname2codepage(p_enc); 623 enc_latin9 = (STRCMP(p_enc, "iso-8859-15") == 0); 624 #endif 625 626 /* Detect an encoding that uses latin1 characters. */ 627 enc_latin1like = (enc_utf8 || STRCMP(p_enc, "latin1") == 0 628 || STRCMP(p_enc, "iso-8859-15") == 0); 629 630 /* 631 * Set the function pointers. 632 */ 633 if (enc_utf8) 634 { 635 mb_ptr2len = utfc_ptr2len; 636 mb_ptr2len_len = utfc_ptr2len_len; 637 mb_char2len = utf_char2len; 638 mb_char2bytes = utf_char2bytes; 639 mb_ptr2cells = utf_ptr2cells; 640 mb_ptr2cells_len = utf_ptr2cells_len; 641 mb_char2cells = utf_char2cells; 642 mb_off2cells = utf_off2cells; 643 mb_ptr2char = utf_ptr2char; 644 mb_head_off = utf_head_off; 645 } 646 else if (enc_dbcs != 0) 647 { 648 mb_ptr2len = dbcs_ptr2len; 649 mb_ptr2len_len = dbcs_ptr2len_len; 650 mb_char2len = dbcs_char2len; 651 mb_char2bytes = dbcs_char2bytes; 652 mb_ptr2cells = dbcs_ptr2cells; 653 mb_ptr2cells_len = dbcs_ptr2cells_len; 654 mb_char2cells = dbcs_char2cells; 655 mb_off2cells = dbcs_off2cells; 656 mb_ptr2char = dbcs_ptr2char; 657 mb_head_off = dbcs_head_off; 658 } 659 else 660 { 661 mb_ptr2len = latin_ptr2len; 662 mb_ptr2len_len = latin_ptr2len_len; 663 mb_char2len = latin_char2len; 664 mb_char2bytes = latin_char2bytes; 665 mb_ptr2cells = latin_ptr2cells; 666 mb_ptr2cells_len = latin_ptr2cells_len; 667 mb_char2cells = latin_char2cells; 668 mb_off2cells = latin_off2cells; 669 mb_ptr2char = latin_ptr2char; 670 mb_head_off = latin_head_off; 671 } 672 673 /* 674 * Fill the mb_bytelen_tab[] for MB_BYTE2LEN(). 675 */ 676 #ifdef LEN_FROM_CONV 677 /* When 'encoding' is different from the current locale mblen() won't 678 * work. Use conversion to "utf-8" instead. */ 679 vimconv.vc_type = CONV_NONE; 680 if (enc_dbcs) 681 { 682 p = enc_locale(); 683 if (p == NULL || STRCMP(p, p_enc) != 0) 684 { 685 convert_setup(&vimconv, p_enc, (char_u *)"utf-8"); 686 vimconv.vc_fail = TRUE; 687 } 688 vim_free(p); 689 } 690 #endif 691 692 for (i = 0; i < 256; ++i) 693 { 694 /* Our own function to reliably check the length of UTF-8 characters, 695 * independent of mblen(). */ 696 if (enc_utf8) 697 n = utf8len_tab[i]; 698 else if (enc_dbcs == 0) 699 n = 1; 700 else 701 { 702 #if defined(MSWIN) || defined(WIN32UNIX) 703 /* enc_dbcs is set by setting 'fileencoding'. It becomes a Windows 704 * CodePage identifier, which we can pass directly in to Windows 705 * API */ 706 n = IsDBCSLeadByteEx(enc_dbcs, (WINBYTE)i) ? 2 : 1; 707 #else 708 # if defined(__amigaos4__) || defined(__ANDROID__) || \ 709 !(defined(HAVE_MBLEN) || defined(X_LOCALE)) 710 /* 711 * if mblen() is not available, character which MSB is turned on 712 * are treated as leading byte character. (note : This assumption 713 * is not always true.) 714 */ 715 n = (i & 0x80) ? 2 : 1; 716 # else 717 char buf[MB_MAXBYTES + 1]; 718 719 if (i == NUL) /* just in case mblen() can't handle "" */ 720 n = 1; 721 else 722 { 723 buf[0] = i; 724 buf[1] = 0; 725 # ifdef LEN_FROM_CONV 726 if (vimconv.vc_type != CONV_NONE) 727 { 728 /* 729 * string_convert() should fail when converting the first 730 * byte of a double-byte character. 731 */ 732 p = string_convert(&vimconv, (char_u *)buf, NULL); 733 if (p != NULL) 734 { 735 vim_free(p); 736 n = 1; 737 } 738 else 739 n = 2; 740 } 741 else 742 # endif 743 { 744 /* 745 * mblen() should return -1 for invalid (means the leading 746 * multibyte) character. However there are some platforms 747 * where mblen() returns 0 for invalid character. 748 * Therefore, following condition includes 0. 749 */ 750 vim_ignored = mblen(NULL, 0); // First reset the state. 751 if (mblen(buf, (size_t)1) <= 0) 752 n = 2; 753 else 754 n = 1; 755 } 756 } 757 # endif 758 #endif 759 } 760 761 mb_bytelen_tab[i] = n; 762 } 763 764 #ifdef LEN_FROM_CONV 765 convert_setup(&vimconv, NULL, NULL); 766 #endif 767 768 /* The cell width depends on the type of multi-byte characters. */ 769 (void)init_chartab(); 770 771 /* When enc_utf8 is set or reset, (de)allocate ScreenLinesUC[] */ 772 screenalloc(FALSE); 773 774 /* When using Unicode, set default for 'fileencodings'. */ 775 if (enc_utf8 && !option_was_set((char_u *)"fencs")) 776 set_string_option_direct((char_u *)"fencs", -1, 777 (char_u *)"ucs-bom,utf-8,default,latin1", OPT_FREE, 0); 778 779 #if defined(HAVE_BIND_TEXTDOMAIN_CODESET) && defined(FEAT_GETTEXT) 780 /* GNU gettext 0.10.37 supports this feature: set the codeset used for 781 * translated messages independently from the current locale. */ 782 (void)bind_textdomain_codeset(VIMPACKAGE, 783 enc_utf8 ? "utf-8" : (char *)p_enc); 784 #endif 785 786 #ifdef MSWIN 787 /* When changing 'encoding' while starting up, then convert the command 788 * line arguments from the active codepage to 'encoding'. */ 789 if (starting != 0) 790 fix_arg_enc(); 791 #endif 792 793 /* Fire an autocommand to let people do custom font setup. This must be 794 * after Vim has been setup for the new encoding. */ 795 apply_autocmds(EVENT_ENCODINGCHANGED, NULL, (char_u *)"", FALSE, curbuf); 796 797 #ifdef FEAT_SPELL 798 /* Need to reload spell dictionaries */ 799 spell_reload(); 800 #endif 801 802 return NULL; 803 } 804 805 /* 806 * Return the size of the BOM for the current buffer: 807 * 0 - no BOM 808 * 2 - UCS-2 or UTF-16 BOM 809 * 4 - UCS-4 BOM 810 * 3 - UTF-8 BOM 811 */ 812 int 813 bomb_size(void) 814 { 815 int n = 0; 816 817 if (curbuf->b_p_bomb && !curbuf->b_p_bin) 818 { 819 if (*curbuf->b_p_fenc == NUL) 820 { 821 if (enc_utf8) 822 { 823 if (enc_unicode != 0) 824 n = enc_unicode; 825 else 826 n = 3; 827 } 828 } 829 else if (STRCMP(curbuf->b_p_fenc, "utf-8") == 0) 830 n = 3; 831 else if (STRNCMP(curbuf->b_p_fenc, "ucs-2", 5) == 0 832 || STRNCMP(curbuf->b_p_fenc, "utf-16", 6) == 0) 833 n = 2; 834 else if (STRNCMP(curbuf->b_p_fenc, "ucs-4", 5) == 0) 835 n = 4; 836 } 837 return n; 838 } 839 840 #if defined(FEAT_QUICKFIX) || defined(PROTO) 841 /* 842 * Remove all BOM from "s" by moving remaining text. 843 */ 844 void 845 remove_bom(char_u *s) 846 { 847 if (enc_utf8) 848 { 849 char_u *p = s; 850 851 while ((p = vim_strbyte(p, 0xef)) != NULL) 852 { 853 if (p[1] == 0xbb && p[2] == 0xbf) 854 STRMOVE(p, p + 3); 855 else 856 ++p; 857 } 858 } 859 } 860 #endif 861 862 /* 863 * Get class of pointer: 864 * 0 for blank or NUL 865 * 1 for punctuation 866 * 2 for an (ASCII) word character 867 * >2 for other word characters 868 */ 869 int 870 mb_get_class(char_u *p) 871 { 872 return mb_get_class_buf(p, curbuf); 873 } 874 875 int 876 mb_get_class_buf(char_u *p, buf_T *buf) 877 { 878 if (MB_BYTE2LEN(p[0]) == 1) 879 { 880 if (p[0] == NUL || VIM_ISWHITE(p[0])) 881 return 0; 882 if (vim_iswordc_buf(p[0], buf)) 883 return 2; 884 return 1; 885 } 886 if (enc_dbcs != 0 && p[0] != NUL && p[1] != NUL) 887 return dbcs_class(p[0], p[1]); 888 if (enc_utf8) 889 return utf_class_buf(utf_ptr2char(p), buf); 890 return 0; 891 } 892 893 /* 894 * Get class of a double-byte character. This always returns 3 or bigger. 895 * TODO: Should return 1 for punctuation. 896 */ 897 int 898 dbcs_class(unsigned lead, unsigned trail) 899 { 900 switch (enc_dbcs) 901 { 902 /* please add classify routine for your language in here */ 903 904 case DBCS_JPNU: /* ? */ 905 case DBCS_JPN: 906 { 907 /* JIS code classification */ 908 unsigned char lb = lead; 909 unsigned char tb = trail; 910 911 /* convert process code to JIS */ 912 # if defined(MSWIN) || defined(WIN32UNIX) || defined(MACOS_X) 913 /* process code is SJIS */ 914 if (lb <= 0x9f) 915 lb = (lb - 0x81) * 2 + 0x21; 916 else 917 lb = (lb - 0xc1) * 2 + 0x21; 918 if (tb <= 0x7e) 919 tb -= 0x1f; 920 else if (tb <= 0x9e) 921 tb -= 0x20; 922 else 923 { 924 tb -= 0x7e; 925 lb += 1; 926 } 927 # else 928 /* 929 * XXX: Code page identification can not use with all 930 * system! So, some other encoding information 931 * will be needed. 932 * In japanese: SJIS,EUC,UNICODE,(JIS) 933 * Note that JIS-code system don't use as 934 * process code in most system because it uses 935 * escape sequences(JIS is context depend encoding). 936 */ 937 /* assume process code is JAPANESE-EUC */ 938 lb &= 0x7f; 939 tb &= 0x7f; 940 # endif 941 /* exceptions */ 942 switch (lb << 8 | tb) 943 { 944 case 0x2121: /* ZENKAKU space */ 945 return 0; 946 case 0x2122: /* TOU-TEN (Japanese comma) */ 947 case 0x2123: /* KU-TEN (Japanese period) */ 948 case 0x2124: /* ZENKAKU comma */ 949 case 0x2125: /* ZENKAKU period */ 950 return 1; 951 case 0x213c: /* prolongedsound handled as KATAKANA */ 952 return 13; 953 } 954 /* sieved by KU code */ 955 switch (lb) 956 { 957 case 0x21: 958 case 0x22: 959 /* special symbols */ 960 return 10; 961 case 0x23: 962 /* alpha-numeric */ 963 return 11; 964 case 0x24: 965 /* hiragana */ 966 return 12; 967 case 0x25: 968 /* katakana */ 969 return 13; 970 case 0x26: 971 /* greek */ 972 return 14; 973 case 0x27: 974 /* russian */ 975 return 15; 976 case 0x28: 977 /* lines */ 978 return 16; 979 default: 980 /* kanji */ 981 return 17; 982 } 983 } 984 985 case DBCS_KORU: /* ? */ 986 case DBCS_KOR: 987 { 988 /* KS code classification */ 989 unsigned char c1 = lead; 990 unsigned char c2 = trail; 991 992 /* 993 * 20 : Hangul 994 * 21 : Hanja 995 * 22 : Symbols 996 * 23 : Alpha-numeric/Roman Letter (Full width) 997 * 24 : Hangul Letter(Alphabet) 998 * 25 : Roman Numeral/Greek Letter 999 * 26 : Box Drawings 1000 * 27 : Unit Symbols 1001 * 28 : Circled/Parenthesized Letter 1002 * 29 : Hiragana/Katakana 1003 * 30 : Cyrillic Letter 1004 */ 1005 1006 if (c1 >= 0xB0 && c1 <= 0xC8) 1007 /* Hangul */ 1008 return 20; 1009 #if defined(MSWIN) || defined(WIN32UNIX) 1010 else if (c1 <= 0xA0 || c2 <= 0xA0) 1011 /* Extended Hangul Region : MS UHC(Unified Hangul Code) */ 1012 /* c1: 0x81-0xA0 with c2: 0x41-0x5A, 0x61-0x7A, 0x81-0xFE 1013 * c1: 0xA1-0xC6 with c2: 0x41-0x5A, 0x61-0x7A, 0x81-0xA0 1014 */ 1015 return 20; 1016 #endif 1017 1018 else if (c1 >= 0xCA && c1 <= 0xFD) 1019 /* Hanja */ 1020 return 21; 1021 else switch (c1) 1022 { 1023 case 0xA1: 1024 case 0xA2: 1025 /* Symbols */ 1026 return 22; 1027 case 0xA3: 1028 /* Alpha-numeric */ 1029 return 23; 1030 case 0xA4: 1031 /* Hangul Letter(Alphabet) */ 1032 return 24; 1033 case 0xA5: 1034 /* Roman Numeral/Greek Letter */ 1035 return 25; 1036 case 0xA6: 1037 /* Box Drawings */ 1038 return 26; 1039 case 0xA7: 1040 /* Unit Symbols */ 1041 return 27; 1042 case 0xA8: 1043 case 0xA9: 1044 if (c2 <= 0xAF) 1045 return 25; /* Roman Letter */ 1046 else if (c2 >= 0xF6) 1047 return 22; /* Symbols */ 1048 else 1049 /* Circled/Parenthesized Letter */ 1050 return 28; 1051 case 0xAA: 1052 case 0xAB: 1053 /* Hiragana/Katakana */ 1054 return 29; 1055 case 0xAC: 1056 /* Cyrillic Letter */ 1057 return 30; 1058 } 1059 } 1060 default: 1061 break; 1062 } 1063 return 3; 1064 } 1065 1066 /* 1067 * mb_char2len() function pointer. 1068 * Return length in bytes of character "c". 1069 * Returns 1 for a single-byte character. 1070 */ 1071 int 1072 latin_char2len(int c UNUSED) 1073 { 1074 return 1; 1075 } 1076 1077 static int 1078 dbcs_char2len( 1079 int c) 1080 { 1081 if (c >= 0x100) 1082 return 2; 1083 return 1; 1084 } 1085 1086 /* 1087 * mb_char2bytes() function pointer. 1088 * Convert a character to its bytes. 1089 * Returns the length in bytes. 1090 */ 1091 int 1092 latin_char2bytes(int c, char_u *buf) 1093 { 1094 buf[0] = c; 1095 return 1; 1096 } 1097 1098 static int 1099 dbcs_char2bytes(int c, char_u *buf) 1100 { 1101 if (c >= 0x100) 1102 { 1103 buf[0] = (unsigned)c >> 8; 1104 buf[1] = c; 1105 /* Never use a NUL byte, it causes lots of trouble. It's an invalid 1106 * character anyway. */ 1107 if (buf[1] == NUL) 1108 buf[1] = '\n'; 1109 return 2; 1110 } 1111 buf[0] = c; 1112 return 1; 1113 } 1114 1115 /* 1116 * mb_ptr2len() function pointer. 1117 * Get byte length of character at "*p" but stop at a NUL. 1118 * For UTF-8 this includes following composing characters. 1119 * Returns 0 when *p is NUL. 1120 */ 1121 int 1122 latin_ptr2len(char_u *p) 1123 { 1124 return MB_BYTE2LEN(*p); 1125 } 1126 1127 static int 1128 dbcs_ptr2len( 1129 char_u *p) 1130 { 1131 int len; 1132 1133 /* Check if second byte is not missing. */ 1134 len = MB_BYTE2LEN(*p); 1135 if (len == 2 && p[1] == NUL) 1136 len = 1; 1137 return len; 1138 } 1139 1140 /* 1141 * mb_ptr2len_len() function pointer. 1142 * Like mb_ptr2len(), but limit to read "size" bytes. 1143 * Returns 0 for an empty string. 1144 * Returns 1 for an illegal char or an incomplete byte sequence. 1145 */ 1146 int 1147 latin_ptr2len_len(char_u *p, int size) 1148 { 1149 if (size < 1 || *p == NUL) 1150 return 0; 1151 return 1; 1152 } 1153 1154 static int 1155 dbcs_ptr2len_len(char_u *p, int size) 1156 { 1157 int len; 1158 1159 if (size < 1 || *p == NUL) 1160 return 0; 1161 if (size == 1) 1162 return 1; 1163 /* Check that second byte is not missing. */ 1164 len = MB_BYTE2LEN(*p); 1165 if (len == 2 && p[1] == NUL) 1166 len = 1; 1167 return len; 1168 } 1169 1170 struct interval 1171 { 1172 long first; 1173 long last; 1174 }; 1175 1176 /* 1177 * Return TRUE if "c" is in "table[size / sizeof(struct interval)]". 1178 */ 1179 static int 1180 intable(struct interval *table, size_t size, int c) 1181 { 1182 int mid, bot, top; 1183 1184 /* first quick check for Latin1 etc. characters */ 1185 if (c < table[0].first) 1186 return FALSE; 1187 1188 /* binary search in table */ 1189 bot = 0; 1190 top = (int)(size / sizeof(struct interval) - 1); 1191 while (top >= bot) 1192 { 1193 mid = (bot + top) / 2; 1194 if (table[mid].last < c) 1195 bot = mid + 1; 1196 else if (table[mid].first > c) 1197 top = mid - 1; 1198 else 1199 return TRUE; 1200 } 1201 return FALSE; 1202 } 1203 1204 /* Sorted list of non-overlapping intervals of East Asian Ambiguous 1205 * characters, generated with ../runtime/tools/unicode.vim. */ 1206 static struct interval ambiguous[] = 1207 { 1208 {0x00a1, 0x00a1}, 1209 {0x00a4, 0x00a4}, 1210 {0x00a7, 0x00a8}, 1211 {0x00aa, 0x00aa}, 1212 {0x00ad, 0x00ae}, 1213 {0x00b0, 0x00b4}, 1214 {0x00b6, 0x00ba}, 1215 {0x00bc, 0x00bf}, 1216 {0x00c6, 0x00c6}, 1217 {0x00d0, 0x00d0}, 1218 {0x00d7, 0x00d8}, 1219 {0x00de, 0x00e1}, 1220 {0x00e6, 0x00e6}, 1221 {0x00e8, 0x00ea}, 1222 {0x00ec, 0x00ed}, 1223 {0x00f0, 0x00f0}, 1224 {0x00f2, 0x00f3}, 1225 {0x00f7, 0x00fa}, 1226 {0x00fc, 0x00fc}, 1227 {0x00fe, 0x00fe}, 1228 {0x0101, 0x0101}, 1229 {0x0111, 0x0111}, 1230 {0x0113, 0x0113}, 1231 {0x011b, 0x011b}, 1232 {0x0126, 0x0127}, 1233 {0x012b, 0x012b}, 1234 {0x0131, 0x0133}, 1235 {0x0138, 0x0138}, 1236 {0x013f, 0x0142}, 1237 {0x0144, 0x0144}, 1238 {0x0148, 0x014b}, 1239 {0x014d, 0x014d}, 1240 {0x0152, 0x0153}, 1241 {0x0166, 0x0167}, 1242 {0x016b, 0x016b}, 1243 {0x01ce, 0x01ce}, 1244 {0x01d0, 0x01d0}, 1245 {0x01d2, 0x01d2}, 1246 {0x01d4, 0x01d4}, 1247 {0x01d6, 0x01d6}, 1248 {0x01d8, 0x01d8}, 1249 {0x01da, 0x01da}, 1250 {0x01dc, 0x01dc}, 1251 {0x0251, 0x0251}, 1252 {0x0261, 0x0261}, 1253 {0x02c4, 0x02c4}, 1254 {0x02c7, 0x02c7}, 1255 {0x02c9, 0x02cb}, 1256 {0x02cd, 0x02cd}, 1257 {0x02d0, 0x02d0}, 1258 {0x02d8, 0x02db}, 1259 {0x02dd, 0x02dd}, 1260 {0x02df, 0x02df}, 1261 {0x0300, 0x036f}, 1262 {0x0391, 0x03a1}, 1263 {0x03a3, 0x03a9}, 1264 {0x03b1, 0x03c1}, 1265 {0x03c3, 0x03c9}, 1266 {0x0401, 0x0401}, 1267 {0x0410, 0x044f}, 1268 {0x0451, 0x0451}, 1269 {0x2010, 0x2010}, 1270 {0x2013, 0x2016}, 1271 {0x2018, 0x2019}, 1272 {0x201c, 0x201d}, 1273 {0x2020, 0x2022}, 1274 {0x2024, 0x2027}, 1275 {0x2030, 0x2030}, 1276 {0x2032, 0x2033}, 1277 {0x2035, 0x2035}, 1278 {0x203b, 0x203b}, 1279 {0x203e, 0x203e}, 1280 {0x2074, 0x2074}, 1281 {0x207f, 0x207f}, 1282 {0x2081, 0x2084}, 1283 {0x20ac, 0x20ac}, 1284 {0x2103, 0x2103}, 1285 {0x2105, 0x2105}, 1286 {0x2109, 0x2109}, 1287 {0x2113, 0x2113}, 1288 {0x2116, 0x2116}, 1289 {0x2121, 0x2122}, 1290 {0x2126, 0x2126}, 1291 {0x212b, 0x212b}, 1292 {0x2153, 0x2154}, 1293 {0x215b, 0x215e}, 1294 {0x2160, 0x216b}, 1295 {0x2170, 0x2179}, 1296 {0x2189, 0x2189}, 1297 {0x2190, 0x2199}, 1298 {0x21b8, 0x21b9}, 1299 {0x21d2, 0x21d2}, 1300 {0x21d4, 0x21d4}, 1301 {0x21e7, 0x21e7}, 1302 {0x2200, 0x2200}, 1303 {0x2202, 0x2203}, 1304 {0x2207, 0x2208}, 1305 {0x220b, 0x220b}, 1306 {0x220f, 0x220f}, 1307 {0x2211, 0x2211}, 1308 {0x2215, 0x2215}, 1309 {0x221a, 0x221a}, 1310 {0x221d, 0x2220}, 1311 {0x2223, 0x2223}, 1312 {0x2225, 0x2225}, 1313 {0x2227, 0x222c}, 1314 {0x222e, 0x222e}, 1315 {0x2234, 0x2237}, 1316 {0x223c, 0x223d}, 1317 {0x2248, 0x2248}, 1318 {0x224c, 0x224c}, 1319 {0x2252, 0x2252}, 1320 {0x2260, 0x2261}, 1321 {0x2264, 0x2267}, 1322 {0x226a, 0x226b}, 1323 {0x226e, 0x226f}, 1324 {0x2282, 0x2283}, 1325 {0x2286, 0x2287}, 1326 {0x2295, 0x2295}, 1327 {0x2299, 0x2299}, 1328 {0x22a5, 0x22a5}, 1329 {0x22bf, 0x22bf}, 1330 {0x2312, 0x2312}, 1331 {0x2460, 0x24e9}, 1332 {0x24eb, 0x254b}, 1333 {0x2550, 0x2573}, 1334 {0x2580, 0x258f}, 1335 {0x2592, 0x2595}, 1336 {0x25a0, 0x25a1}, 1337 {0x25a3, 0x25a9}, 1338 {0x25b2, 0x25b3}, 1339 {0x25b6, 0x25b7}, 1340 {0x25bc, 0x25bd}, 1341 {0x25c0, 0x25c1}, 1342 {0x25c6, 0x25c8}, 1343 {0x25cb, 0x25cb}, 1344 {0x25ce, 0x25d1}, 1345 {0x25e2, 0x25e5}, 1346 {0x25ef, 0x25ef}, 1347 {0x2605, 0x2606}, 1348 {0x2609, 0x2609}, 1349 {0x260e, 0x260f}, 1350 {0x261c, 0x261c}, 1351 {0x261e, 0x261e}, 1352 {0x2640, 0x2640}, 1353 {0x2642, 0x2642}, 1354 {0x2660, 0x2661}, 1355 {0x2663, 0x2665}, 1356 {0x2667, 0x266a}, 1357 {0x266c, 0x266d}, 1358 {0x266f, 0x266f}, 1359 {0x269e, 0x269f}, 1360 {0x26bf, 0x26bf}, 1361 {0x26c6, 0x26cd}, 1362 {0x26cf, 0x26d3}, 1363 {0x26d5, 0x26e1}, 1364 {0x26e3, 0x26e3}, 1365 {0x26e8, 0x26e9}, 1366 {0x26eb, 0x26f1}, 1367 {0x26f4, 0x26f4}, 1368 {0x26f6, 0x26f9}, 1369 {0x26fb, 0x26fc}, 1370 {0x26fe, 0x26ff}, 1371 {0x273d, 0x273d}, 1372 {0x2776, 0x277f}, 1373 {0x2b56, 0x2b59}, 1374 {0x3248, 0x324f}, 1375 {0xe000, 0xf8ff}, 1376 {0xfe00, 0xfe0f}, 1377 {0xfffd, 0xfffd}, 1378 {0x1f100, 0x1f10a}, 1379 {0x1f110, 0x1f12d}, 1380 {0x1f130, 0x1f169}, 1381 {0x1f170, 0x1f18d}, 1382 {0x1f18f, 0x1f190}, 1383 {0x1f19b, 0x1f1ac}, 1384 {0xe0100, 0xe01ef}, 1385 {0xf0000, 0xffffd}, 1386 {0x100000, 0x10fffd} 1387 }; 1388 1389 #if defined(FEAT_TERMINAL) || defined(PROTO) 1390 /* 1391 * utf_char2cells() with different argument type for libvterm. 1392 */ 1393 int 1394 utf_uint2cells(UINT32_T c) 1395 { 1396 if (c >= 0x100 && utf_iscomposing((int)c)) 1397 return 0; 1398 return utf_char2cells((int)c); 1399 } 1400 #endif 1401 1402 /* 1403 * For UTF-8 character "c" return 2 for a double-width character, 1 for others. 1404 * Returns 4 or 6 for an unprintable character. 1405 * Is only correct for characters >= 0x80. 1406 * When p_ambw is "double", return 2 for a character with East Asian Width 1407 * class 'A'(mbiguous). 1408 */ 1409 int 1410 utf_char2cells(int c) 1411 { 1412 /* Sorted list of non-overlapping intervals of East Asian double width 1413 * characters, generated with ../runtime/tools/unicode.vim. */ 1414 static struct interval doublewidth[] = 1415 { 1416 {0x1100, 0x115f}, 1417 {0x231a, 0x231b}, 1418 {0x2329, 0x232a}, 1419 {0x23e9, 0x23ec}, 1420 {0x23f0, 0x23f0}, 1421 {0x23f3, 0x23f3}, 1422 {0x25fd, 0x25fe}, 1423 {0x2614, 0x2615}, 1424 {0x2648, 0x2653}, 1425 {0x267f, 0x267f}, 1426 {0x2693, 0x2693}, 1427 {0x26a1, 0x26a1}, 1428 {0x26aa, 0x26ab}, 1429 {0x26bd, 0x26be}, 1430 {0x26c4, 0x26c5}, 1431 {0x26ce, 0x26ce}, 1432 {0x26d4, 0x26d4}, 1433 {0x26ea, 0x26ea}, 1434 {0x26f2, 0x26f3}, 1435 {0x26f5, 0x26f5}, 1436 {0x26fa, 0x26fa}, 1437 {0x26fd, 0x26fd}, 1438 {0x2705, 0x2705}, 1439 {0x270a, 0x270b}, 1440 {0x2728, 0x2728}, 1441 {0x274c, 0x274c}, 1442 {0x274e, 0x274e}, 1443 {0x2753, 0x2755}, 1444 {0x2757, 0x2757}, 1445 {0x2795, 0x2797}, 1446 {0x27b0, 0x27b0}, 1447 {0x27bf, 0x27bf}, 1448 {0x2b1b, 0x2b1c}, 1449 {0x2b50, 0x2b50}, 1450 {0x2b55, 0x2b55}, 1451 {0x2e80, 0x2e99}, 1452 {0x2e9b, 0x2ef3}, 1453 {0x2f00, 0x2fd5}, 1454 {0x2ff0, 0x2ffb}, 1455 {0x3000, 0x303e}, 1456 {0x3041, 0x3096}, 1457 {0x3099, 0x30ff}, 1458 {0x3105, 0x312f}, 1459 {0x3131, 0x318e}, 1460 {0x3190, 0x31ba}, 1461 {0x31c0, 0x31e3}, 1462 {0x31f0, 0x321e}, 1463 {0x3220, 0x3247}, 1464 {0x3250, 0x32fe}, 1465 {0x3300, 0x4dbf}, 1466 {0x4e00, 0xa48c}, 1467 {0xa490, 0xa4c6}, 1468 {0xa960, 0xa97c}, 1469 {0xac00, 0xd7a3}, 1470 {0xf900, 0xfaff}, 1471 {0xfe10, 0xfe19}, 1472 {0xfe30, 0xfe52}, 1473 {0xfe54, 0xfe66}, 1474 {0xfe68, 0xfe6b}, 1475 {0xff01, 0xff60}, 1476 {0xffe0, 0xffe6}, 1477 {0x16fe0, 0x16fe1}, 1478 {0x17000, 0x187f1}, 1479 {0x18800, 0x18af2}, 1480 {0x1b000, 0x1b11e}, 1481 {0x1b170, 0x1b2fb}, 1482 {0x1f004, 0x1f004}, 1483 {0x1f0cf, 0x1f0cf}, 1484 {0x1f18e, 0x1f18e}, 1485 {0x1f191, 0x1f19a}, 1486 {0x1f200, 0x1f202}, 1487 {0x1f210, 0x1f23b}, 1488 {0x1f240, 0x1f248}, 1489 {0x1f250, 0x1f251}, 1490 {0x1f260, 0x1f265}, 1491 {0x1f300, 0x1f320}, 1492 {0x1f32d, 0x1f335}, 1493 {0x1f337, 0x1f37c}, 1494 {0x1f37e, 0x1f393}, 1495 {0x1f3a0, 0x1f3ca}, 1496 {0x1f3cf, 0x1f3d3}, 1497 {0x1f3e0, 0x1f3f0}, 1498 {0x1f3f4, 0x1f3f4}, 1499 {0x1f3f8, 0x1f43e}, 1500 {0x1f440, 0x1f440}, 1501 {0x1f442, 0x1f4fc}, 1502 {0x1f4ff, 0x1f53d}, 1503 {0x1f54b, 0x1f54e}, 1504 {0x1f550, 0x1f567}, 1505 {0x1f57a, 0x1f57a}, 1506 {0x1f595, 0x1f596}, 1507 {0x1f5a4, 0x1f5a4}, 1508 {0x1f5fb, 0x1f64f}, 1509 {0x1f680, 0x1f6c5}, 1510 {0x1f6cc, 0x1f6cc}, 1511 {0x1f6d0, 0x1f6d2}, 1512 {0x1f6eb, 0x1f6ec}, 1513 {0x1f6f4, 0x1f6f9}, 1514 {0x1f910, 0x1f93e}, 1515 {0x1f940, 0x1f970}, 1516 {0x1f973, 0x1f976}, 1517 {0x1f97a, 0x1f97a}, 1518 {0x1f97c, 0x1f9a2}, 1519 {0x1f9b0, 0x1f9b9}, 1520 {0x1f9c0, 0x1f9c2}, 1521 {0x1f9d0, 0x1f9ff}, 1522 {0x20000, 0x2fffd}, 1523 {0x30000, 0x3fffd} 1524 }; 1525 1526 /* Sorted list of non-overlapping intervals of Emoji characters that don't 1527 * have ambiguous or double width, 1528 * based on http://unicode.org/emoji/charts/emoji-list.html */ 1529 static struct interval emoji_width[] = 1530 { 1531 {0x1f1e6, 0x1f1ff}, 1532 {0x1f321, 0x1f321}, 1533 {0x1f324, 0x1f32c}, 1534 {0x1f336, 0x1f336}, 1535 {0x1f37d, 0x1f37d}, 1536 {0x1f396, 0x1f397}, 1537 {0x1f399, 0x1f39b}, 1538 {0x1f39e, 0x1f39f}, 1539 {0x1f3cb, 0x1f3ce}, 1540 {0x1f3d4, 0x1f3df}, 1541 {0x1f3f3, 0x1f3f5}, 1542 {0x1f3f7, 0x1f3f7}, 1543 {0x1f43f, 0x1f43f}, 1544 {0x1f441, 0x1f441}, 1545 {0x1f4fd, 0x1f4fd}, 1546 {0x1f549, 0x1f54a}, 1547 {0x1f56f, 0x1f570}, 1548 {0x1f573, 0x1f579}, 1549 {0x1f587, 0x1f587}, 1550 {0x1f58a, 0x1f58d}, 1551 {0x1f590, 0x1f590}, 1552 {0x1f5a5, 0x1f5a5}, 1553 {0x1f5a8, 0x1f5a8}, 1554 {0x1f5b1, 0x1f5b2}, 1555 {0x1f5bc, 0x1f5bc}, 1556 {0x1f5c2, 0x1f5c4}, 1557 {0x1f5d1, 0x1f5d3}, 1558 {0x1f5dc, 0x1f5de}, 1559 {0x1f5e1, 0x1f5e1}, 1560 {0x1f5e3, 0x1f5e3}, 1561 {0x1f5e8, 0x1f5e8}, 1562 {0x1f5ef, 0x1f5ef}, 1563 {0x1f5f3, 0x1f5f3}, 1564 {0x1f5fa, 0x1f5fa}, 1565 {0x1f6cb, 0x1f6cf}, 1566 {0x1f6e0, 0x1f6e5}, 1567 {0x1f6e9, 0x1f6e9}, 1568 {0x1f6f0, 0x1f6f0}, 1569 {0x1f6f3, 0x1f6f3} 1570 }; 1571 1572 if (c >= 0x100) 1573 { 1574 #ifdef USE_WCHAR_FUNCTIONS 1575 /* 1576 * Assume the library function wcwidth() works better than our own 1577 * stuff. It should return 1 for ambiguous width chars! 1578 */ 1579 int n = wcwidth(c); 1580 1581 if (n < 0) 1582 return 6; /* unprintable, displays <xxxx> */ 1583 if (n > 1) 1584 return n; 1585 #else 1586 if (!utf_printable(c)) 1587 return 6; /* unprintable, displays <xxxx> */ 1588 if (intable(doublewidth, sizeof(doublewidth), c)) 1589 return 2; 1590 #endif 1591 if (p_emoji && intable(emoji_width, sizeof(emoji_width), c)) 1592 return 2; 1593 } 1594 1595 /* Characters below 0x100 are influenced by 'isprint' option */ 1596 else if (c >= 0x80 && !vim_isprintc(c)) 1597 return 4; /* unprintable, displays <xx> */ 1598 1599 if (c >= 0x80 && *p_ambw == 'd' && intable(ambiguous, sizeof(ambiguous), c)) 1600 return 2; 1601 1602 return 1; 1603 } 1604 1605 /* 1606 * mb_ptr2cells() function pointer. 1607 * Return the number of display cells character at "*p" occupies. 1608 * This doesn't take care of unprintable characters, use ptr2cells() for that. 1609 */ 1610 int 1611 latin_ptr2cells(char_u *p UNUSED) 1612 { 1613 return 1; 1614 } 1615 1616 int 1617 utf_ptr2cells( 1618 char_u *p) 1619 { 1620 int c; 1621 1622 /* Need to convert to a wide character. */ 1623 if (*p >= 0x80) 1624 { 1625 c = utf_ptr2char(p); 1626 /* An illegal byte is displayed as <xx>. */ 1627 if (utf_ptr2len(p) == 1 || c == NUL) 1628 return 4; 1629 /* If the char is ASCII it must be an overlong sequence. */ 1630 if (c < 0x80) 1631 return char2cells(c); 1632 return utf_char2cells(c); 1633 } 1634 return 1; 1635 } 1636 1637 int 1638 dbcs_ptr2cells(char_u *p) 1639 { 1640 /* Number of cells is equal to number of bytes, except for euc-jp when 1641 * the first byte is 0x8e. */ 1642 if (enc_dbcs == DBCS_JPNU && *p == 0x8e) 1643 return 1; 1644 return MB_BYTE2LEN(*p); 1645 } 1646 1647 /* 1648 * mb_ptr2cells_len() function pointer. 1649 * Like mb_ptr2cells(), but limit string length to "size". 1650 * For an empty string or truncated character returns 1. 1651 */ 1652 int 1653 latin_ptr2cells_len(char_u *p UNUSED, int size UNUSED) 1654 { 1655 return 1; 1656 } 1657 1658 static int 1659 utf_ptr2cells_len(char_u *p, int size) 1660 { 1661 int c; 1662 1663 /* Need to convert to a wide character. */ 1664 if (size > 0 && *p >= 0x80) 1665 { 1666 if (utf_ptr2len_len(p, size) < utf8len_tab[*p]) 1667 return 1; /* truncated */ 1668 c = utf_ptr2char(p); 1669 /* An illegal byte is displayed as <xx>. */ 1670 if (utf_ptr2len(p) == 1 || c == NUL) 1671 return 4; 1672 /* If the char is ASCII it must be an overlong sequence. */ 1673 if (c < 0x80) 1674 return char2cells(c); 1675 return utf_char2cells(c); 1676 } 1677 return 1; 1678 } 1679 1680 static int 1681 dbcs_ptr2cells_len(char_u *p, int size) 1682 { 1683 /* Number of cells is equal to number of bytes, except for euc-jp when 1684 * the first byte is 0x8e. */ 1685 if (size <= 1 || (enc_dbcs == DBCS_JPNU && *p == 0x8e)) 1686 return 1; 1687 return MB_BYTE2LEN(*p); 1688 } 1689 1690 /* 1691 * mb_char2cells() function pointer. 1692 * Return the number of display cells character "c" occupies. 1693 * Only takes care of multi-byte chars, not "^C" and such. 1694 */ 1695 int 1696 latin_char2cells(int c UNUSED) 1697 { 1698 return 1; 1699 } 1700 1701 static int 1702 dbcs_char2cells(int c) 1703 { 1704 /* Number of cells is equal to number of bytes, except for euc-jp when 1705 * the first byte is 0x8e. */ 1706 if (enc_dbcs == DBCS_JPNU && ((unsigned)c >> 8) == 0x8e) 1707 return 1; 1708 /* use the first byte */ 1709 return MB_BYTE2LEN((unsigned)c >> 8); 1710 } 1711 1712 /* 1713 * Return the number of cells occupied by string "p". 1714 * Stop at a NUL character. When "len" >= 0 stop at character "p[len]". 1715 */ 1716 int 1717 mb_string2cells(char_u *p, int len) 1718 { 1719 int i; 1720 int clen = 0; 1721 1722 for (i = 0; (len < 0 || i < len) && p[i] != NUL; i += (*mb_ptr2len)(p + i)) 1723 clen += (*mb_ptr2cells)(p + i); 1724 return clen; 1725 } 1726 1727 /* 1728 * mb_off2cells() function pointer. 1729 * Return number of display cells for char at ScreenLines[off]. 1730 * We make sure that the offset used is less than "max_off". 1731 */ 1732 int 1733 latin_off2cells(unsigned off UNUSED, unsigned max_off UNUSED) 1734 { 1735 return 1; 1736 } 1737 1738 int 1739 dbcs_off2cells(unsigned off, unsigned max_off) 1740 { 1741 /* never check beyond end of the line */ 1742 if (off >= max_off) 1743 return 1; 1744 1745 /* Number of cells is equal to number of bytes, except for euc-jp when 1746 * the first byte is 0x8e. */ 1747 if (enc_dbcs == DBCS_JPNU && ScreenLines[off] == 0x8e) 1748 return 1; 1749 return MB_BYTE2LEN(ScreenLines[off]); 1750 } 1751 1752 int 1753 utf_off2cells(unsigned off, unsigned max_off) 1754 { 1755 return (off + 1 < max_off && ScreenLines[off + 1] == 0) ? 2 : 1; 1756 } 1757 1758 /* 1759 * mb_ptr2char() function pointer. 1760 * Convert a byte sequence into a character. 1761 */ 1762 int 1763 latin_ptr2char(char_u *p) 1764 { 1765 return *p; 1766 } 1767 1768 static int 1769 dbcs_ptr2char(char_u *p) 1770 { 1771 if (MB_BYTE2LEN(*p) > 1 && p[1] != NUL) 1772 return (p[0] << 8) + p[1]; 1773 return *p; 1774 } 1775 1776 /* 1777 * Convert a UTF-8 byte sequence to a wide character. 1778 * If the sequence is illegal or truncated by a NUL the first byte is 1779 * returned. 1780 * For an overlong sequence this may return zero. 1781 * Does not include composing characters, of course. 1782 */ 1783 int 1784 utf_ptr2char(char_u *p) 1785 { 1786 int len; 1787 1788 if (p[0] < 0x80) /* be quick for ASCII */ 1789 return p[0]; 1790 1791 len = utf8len_tab_zero[p[0]]; 1792 if (len > 1 && (p[1] & 0xc0) == 0x80) 1793 { 1794 if (len == 2) 1795 return ((p[0] & 0x1f) << 6) + (p[1] & 0x3f); 1796 if ((p[2] & 0xc0) == 0x80) 1797 { 1798 if (len == 3) 1799 return ((p[0] & 0x0f) << 12) + ((p[1] & 0x3f) << 6) 1800 + (p[2] & 0x3f); 1801 if ((p[3] & 0xc0) == 0x80) 1802 { 1803 if (len == 4) 1804 return ((p[0] & 0x07) << 18) + ((p[1] & 0x3f) << 12) 1805 + ((p[2] & 0x3f) << 6) + (p[3] & 0x3f); 1806 if ((p[4] & 0xc0) == 0x80) 1807 { 1808 if (len == 5) 1809 return ((p[0] & 0x03) << 24) + ((p[1] & 0x3f) << 18) 1810 + ((p[2] & 0x3f) << 12) + ((p[3] & 0x3f) << 6) 1811 + (p[4] & 0x3f); 1812 if ((p[5] & 0xc0) == 0x80 && len == 6) 1813 return ((p[0] & 0x01) << 30) + ((p[1] & 0x3f) << 24) 1814 + ((p[2] & 0x3f) << 18) + ((p[3] & 0x3f) << 12) 1815 + ((p[4] & 0x3f) << 6) + (p[5] & 0x3f); 1816 } 1817 } 1818 } 1819 } 1820 /* Illegal value, just return the first byte */ 1821 return p[0]; 1822 } 1823 1824 /* 1825 * Convert a UTF-8 byte sequence to a wide character. 1826 * String is assumed to be terminated by NUL or after "n" bytes, whichever 1827 * comes first. 1828 * The function is safe in the sense that it never accesses memory beyond the 1829 * first "n" bytes of "s". 1830 * 1831 * On success, returns decoded codepoint, advances "s" to the beginning of 1832 * next character and decreases "n" accordingly. 1833 * 1834 * If end of string was reached, returns 0 and, if "n" > 0, advances "s" past 1835 * NUL byte. 1836 * 1837 * If byte sequence is illegal or incomplete, returns -1 and does not advance 1838 * "s". 1839 */ 1840 static int 1841 utf_safe_read_char_adv(char_u **s, size_t *n) 1842 { 1843 int c, k; 1844 1845 if (*n == 0) /* end of buffer */ 1846 return 0; 1847 1848 k = utf8len_tab_zero[**s]; 1849 1850 if (k == 1) 1851 { 1852 /* ASCII character or NUL */ 1853 (*n)--; 1854 return *(*s)++; 1855 } 1856 1857 if ((size_t)k <= *n) 1858 { 1859 /* We have a multibyte sequence and it isn't truncated by buffer 1860 * limits so utf_ptr2char() is safe to use. Or the first byte is 1861 * illegal (k=0), and it's also safe to use utf_ptr2char(). */ 1862 c = utf_ptr2char(*s); 1863 1864 /* On failure, utf_ptr2char() returns the first byte, so here we 1865 * check equality with the first byte. The only non-ASCII character 1866 * which equals the first byte of its own UTF-8 representation is 1867 * U+00C3 (UTF-8: 0xC3 0x83), so need to check that special case too. 1868 * It's safe even if n=1, else we would have k=2 > n. */ 1869 if (c != (int)(**s) || (c == 0xC3 && (*s)[1] == 0x83)) 1870 { 1871 /* byte sequence was successfully decoded */ 1872 *s += k; 1873 *n -= k; 1874 return c; 1875 } 1876 } 1877 1878 /* byte sequence is incomplete or illegal */ 1879 return -1; 1880 } 1881 1882 /* 1883 * Get character at **pp and advance *pp to the next character. 1884 * Note: composing characters are skipped! 1885 */ 1886 int 1887 mb_ptr2char_adv(char_u **pp) 1888 { 1889 int c; 1890 1891 c = (*mb_ptr2char)(*pp); 1892 *pp += (*mb_ptr2len)(*pp); 1893 return c; 1894 } 1895 1896 /* 1897 * Get character at **pp and advance *pp to the next character. 1898 * Note: composing characters are returned as separate characters. 1899 */ 1900 int 1901 mb_cptr2char_adv(char_u **pp) 1902 { 1903 int c; 1904 1905 c = (*mb_ptr2char)(*pp); 1906 if (enc_utf8) 1907 *pp += utf_ptr2len(*pp); 1908 else 1909 *pp += (*mb_ptr2len)(*pp); 1910 return c; 1911 } 1912 1913 #if defined(FEAT_ARABIC) || defined(PROTO) 1914 /* 1915 * Check if the character pointed to by "p2" is a composing character when it 1916 * comes after "p1". For Arabic sometimes "ab" is replaced with "c", which 1917 * behaves like a composing character. 1918 */ 1919 int 1920 utf_composinglike(char_u *p1, char_u *p2) 1921 { 1922 int c2; 1923 1924 c2 = utf_ptr2char(p2); 1925 if (utf_iscomposing(c2)) 1926 return TRUE; 1927 if (!arabic_maycombine(c2)) 1928 return FALSE; 1929 return arabic_combine(utf_ptr2char(p1), c2); 1930 } 1931 #endif 1932 1933 /* 1934 * Convert a UTF-8 byte string to a wide character. Also get up to MAX_MCO 1935 * composing characters. 1936 */ 1937 int 1938 utfc_ptr2char( 1939 char_u *p, 1940 int *pcc) /* return: composing chars, last one is 0 */ 1941 { 1942 int len; 1943 int c; 1944 int cc; 1945 int i = 0; 1946 1947 c = utf_ptr2char(p); 1948 len = utf_ptr2len(p); 1949 1950 /* Only accept a composing char when the first char isn't illegal. */ 1951 if ((len > 1 || *p < 0x80) 1952 && p[len] >= 0x80 1953 && UTF_COMPOSINGLIKE(p, p + len)) 1954 { 1955 cc = utf_ptr2char(p + len); 1956 for (;;) 1957 { 1958 pcc[i++] = cc; 1959 if (i == MAX_MCO) 1960 break; 1961 len += utf_ptr2len(p + len); 1962 if (p[len] < 0x80 || !utf_iscomposing(cc = utf_ptr2char(p + len))) 1963 break; 1964 } 1965 } 1966 1967 if (i < MAX_MCO) /* last composing char must be 0 */ 1968 pcc[i] = 0; 1969 1970 return c; 1971 } 1972 1973 /* 1974 * Convert a UTF-8 byte string to a wide character. Also get up to MAX_MCO 1975 * composing characters. Use no more than p[maxlen]. 1976 */ 1977 int 1978 utfc_ptr2char_len( 1979 char_u *p, 1980 int *pcc, /* return: composing chars, last one is 0 */ 1981 int maxlen) 1982 { 1983 int len; 1984 int c; 1985 int cc; 1986 int i = 0; 1987 1988 c = utf_ptr2char(p); 1989 len = utf_ptr2len_len(p, maxlen); 1990 /* Only accept a composing char when the first char isn't illegal. */ 1991 if ((len > 1 || *p < 0x80) 1992 && len < maxlen 1993 && p[len] >= 0x80 1994 && UTF_COMPOSINGLIKE(p, p + len)) 1995 { 1996 cc = utf_ptr2char(p + len); 1997 for (;;) 1998 { 1999 pcc[i++] = cc; 2000 if (i == MAX_MCO) 2001 break; 2002 len += utf_ptr2len_len(p + len, maxlen - len); 2003 if (len >= maxlen 2004 || p[len] < 0x80 2005 || !utf_iscomposing(cc = utf_ptr2char(p + len))) 2006 break; 2007 } 2008 } 2009 2010 if (i < MAX_MCO) /* last composing char must be 0 */ 2011 pcc[i] = 0; 2012 2013 return c; 2014 } 2015 2016 /* 2017 * Convert the character at screen position "off" to a sequence of bytes. 2018 * Includes the composing characters. 2019 * "buf" must at least have the length MB_MAXBYTES + 1. 2020 * Only to be used when ScreenLinesUC[off] != 0. 2021 * Returns the produced number of bytes. 2022 */ 2023 int 2024 utfc_char2bytes(int off, char_u *buf) 2025 { 2026 int len; 2027 int i; 2028 2029 len = utf_char2bytes(ScreenLinesUC[off], buf); 2030 for (i = 0; i < Screen_mco; ++i) 2031 { 2032 if (ScreenLinesC[i][off] == 0) 2033 break; 2034 len += utf_char2bytes(ScreenLinesC[i][off], buf + len); 2035 } 2036 return len; 2037 } 2038 2039 /* 2040 * Get the length of a UTF-8 byte sequence, not including any following 2041 * composing characters. 2042 * Returns 0 for "". 2043 * Returns 1 for an illegal byte sequence. 2044 */ 2045 int 2046 utf_ptr2len(char_u *p) 2047 { 2048 int len; 2049 int i; 2050 2051 if (*p == NUL) 2052 return 0; 2053 len = utf8len_tab[*p]; 2054 for (i = 1; i < len; ++i) 2055 if ((p[i] & 0xc0) != 0x80) 2056 return 1; 2057 return len; 2058 } 2059 2060 /* 2061 * Return length of UTF-8 character, obtained from the first byte. 2062 * "b" must be between 0 and 255! 2063 * Returns 1 for an invalid first byte value. 2064 */ 2065 int 2066 utf_byte2len(int b) 2067 { 2068 return utf8len_tab[b]; 2069 } 2070 2071 /* 2072 * Get the length of UTF-8 byte sequence "p[size]". Does not include any 2073 * following composing characters. 2074 * Returns 1 for "". 2075 * Returns 1 for an illegal byte sequence (also in incomplete byte seq.). 2076 * Returns number > "size" for an incomplete byte sequence. 2077 * Never returns zero. 2078 */ 2079 int 2080 utf_ptr2len_len(char_u *p, int size) 2081 { 2082 int len; 2083 int i; 2084 int m; 2085 2086 len = utf8len_tab[*p]; 2087 if (len == 1) 2088 return 1; /* NUL, ascii or illegal lead byte */ 2089 if (len > size) 2090 m = size; /* incomplete byte sequence. */ 2091 else 2092 m = len; 2093 for (i = 1; i < m; ++i) 2094 if ((p[i] & 0xc0) != 0x80) 2095 return 1; 2096 return len; 2097 } 2098 2099 /* 2100 * Return the number of bytes the UTF-8 encoding of the character at "p" takes. 2101 * This includes following composing characters. 2102 */ 2103 int 2104 utfc_ptr2len(char_u *p) 2105 { 2106 int len; 2107 int b0 = *p; 2108 #ifdef FEAT_ARABIC 2109 int prevlen; 2110 #endif 2111 2112 if (b0 == NUL) 2113 return 0; 2114 if (b0 < 0x80 && p[1] < 0x80) /* be quick for ASCII */ 2115 return 1; 2116 2117 /* Skip over first UTF-8 char, stopping at a NUL byte. */ 2118 len = utf_ptr2len(p); 2119 2120 /* Check for illegal byte. */ 2121 if (len == 1 && b0 >= 0x80) 2122 return 1; 2123 2124 /* 2125 * Check for composing characters. We can handle only the first six, but 2126 * skip all of them (otherwise the cursor would get stuck). 2127 */ 2128 #ifdef FEAT_ARABIC 2129 prevlen = 0; 2130 #endif 2131 for (;;) 2132 { 2133 if (p[len] < 0x80 || !UTF_COMPOSINGLIKE(p + prevlen, p + len)) 2134 return len; 2135 2136 /* Skip over composing char */ 2137 #ifdef FEAT_ARABIC 2138 prevlen = len; 2139 #endif 2140 len += utf_ptr2len(p + len); 2141 } 2142 } 2143 2144 /* 2145 * Return the number of bytes the UTF-8 encoding of the character at "p[size]" 2146 * takes. This includes following composing characters. 2147 * Returns 0 for an empty string. 2148 * Returns 1 for an illegal char or an incomplete byte sequence. 2149 */ 2150 int 2151 utfc_ptr2len_len(char_u *p, int size) 2152 { 2153 int len; 2154 #ifdef FEAT_ARABIC 2155 int prevlen; 2156 #endif 2157 2158 if (size < 1 || *p == NUL) 2159 return 0; 2160 if (p[0] < 0x80 && (size == 1 || p[1] < 0x80)) /* be quick for ASCII */ 2161 return 1; 2162 2163 /* Skip over first UTF-8 char, stopping at a NUL byte. */ 2164 len = utf_ptr2len_len(p, size); 2165 2166 /* Check for illegal byte and incomplete byte sequence. */ 2167 if ((len == 1 && p[0] >= 0x80) || len > size) 2168 return 1; 2169 2170 /* 2171 * Check for composing characters. We can handle only the first six, but 2172 * skip all of them (otherwise the cursor would get stuck). 2173 */ 2174 #ifdef FEAT_ARABIC 2175 prevlen = 0; 2176 #endif 2177 while (len < size) 2178 { 2179 int len_next_char; 2180 2181 if (p[len] < 0x80) 2182 break; 2183 2184 /* 2185 * Next character length should not go beyond size to ensure that 2186 * UTF_COMPOSINGLIKE(...) does not read beyond size. 2187 */ 2188 len_next_char = utf_ptr2len_len(p + len, size - len); 2189 if (len_next_char > size - len) 2190 break; 2191 2192 if (!UTF_COMPOSINGLIKE(p + prevlen, p + len)) 2193 break; 2194 2195 /* Skip over composing char */ 2196 #ifdef FEAT_ARABIC 2197 prevlen = len; 2198 #endif 2199 len += len_next_char; 2200 } 2201 return len; 2202 } 2203 2204 /* 2205 * Return the number of bytes the UTF-8 encoding of character "c" takes. 2206 * This does not include composing characters. 2207 */ 2208 int 2209 utf_char2len(int c) 2210 { 2211 if (c < 0x80) 2212 return 1; 2213 if (c < 0x800) 2214 return 2; 2215 if (c < 0x10000) 2216 return 3; 2217 if (c < 0x200000) 2218 return 4; 2219 if (c < 0x4000000) 2220 return 5; 2221 return 6; 2222 } 2223 2224 /* 2225 * Convert Unicode character "c" to UTF-8 string in "buf[]". 2226 * Returns the number of bytes. 2227 */ 2228 int 2229 utf_char2bytes(int c, char_u *buf) 2230 { 2231 if (c < 0x80) /* 7 bits */ 2232 { 2233 buf[0] = c; 2234 return 1; 2235 } 2236 if (c < 0x800) /* 11 bits */ 2237 { 2238 buf[0] = 0xc0 + ((unsigned)c >> 6); 2239 buf[1] = 0x80 + (c & 0x3f); 2240 return 2; 2241 } 2242 if (c < 0x10000) /* 16 bits */ 2243 { 2244 buf[0] = 0xe0 + ((unsigned)c >> 12); 2245 buf[1] = 0x80 + (((unsigned)c >> 6) & 0x3f); 2246 buf[2] = 0x80 + (c & 0x3f); 2247 return 3; 2248 } 2249 if (c < 0x200000) /* 21 bits */ 2250 { 2251 buf[0] = 0xf0 + ((unsigned)c >> 18); 2252 buf[1] = 0x80 + (((unsigned)c >> 12) & 0x3f); 2253 buf[2] = 0x80 + (((unsigned)c >> 6) & 0x3f); 2254 buf[3] = 0x80 + (c & 0x3f); 2255 return 4; 2256 } 2257 if (c < 0x4000000) /* 26 bits */ 2258 { 2259 buf[0] = 0xf8 + ((unsigned)c >> 24); 2260 buf[1] = 0x80 + (((unsigned)c >> 18) & 0x3f); 2261 buf[2] = 0x80 + (((unsigned)c >> 12) & 0x3f); 2262 buf[3] = 0x80 + (((unsigned)c >> 6) & 0x3f); 2263 buf[4] = 0x80 + (c & 0x3f); 2264 return 5; 2265 } 2266 /* 31 bits */ 2267 buf[0] = 0xfc + ((unsigned)c >> 30); 2268 buf[1] = 0x80 + (((unsigned)c >> 24) & 0x3f); 2269 buf[2] = 0x80 + (((unsigned)c >> 18) & 0x3f); 2270 buf[3] = 0x80 + (((unsigned)c >> 12) & 0x3f); 2271 buf[4] = 0x80 + (((unsigned)c >> 6) & 0x3f); 2272 buf[5] = 0x80 + (c & 0x3f); 2273 return 6; 2274 } 2275 2276 #if defined(FEAT_TERMINAL) || defined(PROTO) 2277 /* 2278 * utf_iscomposing() with different argument type for libvterm. 2279 */ 2280 int 2281 utf_iscomposing_uint(UINT32_T c) 2282 { 2283 return utf_iscomposing((int)c); 2284 } 2285 #endif 2286 2287 /* 2288 * Return TRUE if "c" is a composing UTF-8 character. This means it will be 2289 * drawn on top of the preceding character. 2290 * Based on code from Markus Kuhn. 2291 */ 2292 int 2293 utf_iscomposing(int c) 2294 { 2295 /* Sorted list of non-overlapping intervals. 2296 * Generated by ../runtime/tools/unicode.vim. */ 2297 static struct interval combining[] = 2298 { 2299 {0x0300, 0x036f}, 2300 {0x0483, 0x0489}, 2301 {0x0591, 0x05bd}, 2302 {0x05bf, 0x05bf}, 2303 {0x05c1, 0x05c2}, 2304 {0x05c4, 0x05c5}, 2305 {0x05c7, 0x05c7}, 2306 {0x0610, 0x061a}, 2307 {0x064b, 0x065f}, 2308 {0x0670, 0x0670}, 2309 {0x06d6, 0x06dc}, 2310 {0x06df, 0x06e4}, 2311 {0x06e7, 0x06e8}, 2312 {0x06ea, 0x06ed}, 2313 {0x0711, 0x0711}, 2314 {0x0730, 0x074a}, 2315 {0x07a6, 0x07b0}, 2316 {0x07eb, 0x07f3}, 2317 {0x07fd, 0x07fd}, 2318 {0x0816, 0x0819}, 2319 {0x081b, 0x0823}, 2320 {0x0825, 0x0827}, 2321 {0x0829, 0x082d}, 2322 {0x0859, 0x085b}, 2323 {0x08d3, 0x08e1}, 2324 {0x08e3, 0x0903}, 2325 {0x093a, 0x093c}, 2326 {0x093e, 0x094f}, 2327 {0x0951, 0x0957}, 2328 {0x0962, 0x0963}, 2329 {0x0981, 0x0983}, 2330 {0x09bc, 0x09bc}, 2331 {0x09be, 0x09c4}, 2332 {0x09c7, 0x09c8}, 2333 {0x09cb, 0x09cd}, 2334 {0x09d7, 0x09d7}, 2335 {0x09e2, 0x09e3}, 2336 {0x09fe, 0x09fe}, 2337 {0x0a01, 0x0a03}, 2338 {0x0a3c, 0x0a3c}, 2339 {0x0a3e, 0x0a42}, 2340 {0x0a47, 0x0a48}, 2341 {0x0a4b, 0x0a4d}, 2342 {0x0a51, 0x0a51}, 2343 {0x0a70, 0x0a71}, 2344 {0x0a75, 0x0a75}, 2345 {0x0a81, 0x0a83}, 2346 {0x0abc, 0x0abc}, 2347 {0x0abe, 0x0ac5}, 2348 {0x0ac7, 0x0ac9}, 2349 {0x0acb, 0x0acd}, 2350 {0x0ae2, 0x0ae3}, 2351 {0x0afa, 0x0aff}, 2352 {0x0b01, 0x0b03}, 2353 {0x0b3c, 0x0b3c}, 2354 {0x0b3e, 0x0b44}, 2355 {0x0b47, 0x0b48}, 2356 {0x0b4b, 0x0b4d}, 2357 {0x0b56, 0x0b57}, 2358 {0x0b62, 0x0b63}, 2359 {0x0b82, 0x0b82}, 2360 {0x0bbe, 0x0bc2}, 2361 {0x0bc6, 0x0bc8}, 2362 {0x0bca, 0x0bcd}, 2363 {0x0bd7, 0x0bd7}, 2364 {0x0c00, 0x0c04}, 2365 {0x0c3e, 0x0c44}, 2366 {0x0c46, 0x0c48}, 2367 {0x0c4a, 0x0c4d}, 2368 {0x0c55, 0x0c56}, 2369 {0x0c62, 0x0c63}, 2370 {0x0c81, 0x0c83}, 2371 {0x0cbc, 0x0cbc}, 2372 {0x0cbe, 0x0cc4}, 2373 {0x0cc6, 0x0cc8}, 2374 {0x0cca, 0x0ccd}, 2375 {0x0cd5, 0x0cd6}, 2376 {0x0ce2, 0x0ce3}, 2377 {0x0d00, 0x0d03}, 2378 {0x0d3b, 0x0d3c}, 2379 {0x0d3e, 0x0d44}, 2380 {0x0d46, 0x0d48}, 2381 {0x0d4a, 0x0d4d}, 2382 {0x0d57, 0x0d57}, 2383 {0x0d62, 0x0d63}, 2384 {0x0d82, 0x0d83}, 2385 {0x0dca, 0x0dca}, 2386 {0x0dcf, 0x0dd4}, 2387 {0x0dd6, 0x0dd6}, 2388 {0x0dd8, 0x0ddf}, 2389 {0x0df2, 0x0df3}, 2390 {0x0e31, 0x0e31}, 2391 {0x0e34, 0x0e3a}, 2392 {0x0e47, 0x0e4e}, 2393 {0x0eb1, 0x0eb1}, 2394 {0x0eb4, 0x0eb9}, 2395 {0x0ebb, 0x0ebc}, 2396 {0x0ec8, 0x0ecd}, 2397 {0x0f18, 0x0f19}, 2398 {0x0f35, 0x0f35}, 2399 {0x0f37, 0x0f37}, 2400 {0x0f39, 0x0f39}, 2401 {0x0f3e, 0x0f3f}, 2402 {0x0f71, 0x0f84}, 2403 {0x0f86, 0x0f87}, 2404 {0x0f8d, 0x0f97}, 2405 {0x0f99, 0x0fbc}, 2406 {0x0fc6, 0x0fc6}, 2407 {0x102b, 0x103e}, 2408 {0x1056, 0x1059}, 2409 {0x105e, 0x1060}, 2410 {0x1062, 0x1064}, 2411 {0x1067, 0x106d}, 2412 {0x1071, 0x1074}, 2413 {0x1082, 0x108d}, 2414 {0x108f, 0x108f}, 2415 {0x109a, 0x109d}, 2416 {0x135d, 0x135f}, 2417 {0x1712, 0x1714}, 2418 {0x1732, 0x1734}, 2419 {0x1752, 0x1753}, 2420 {0x1772, 0x1773}, 2421 {0x17b4, 0x17d3}, 2422 {0x17dd, 0x17dd}, 2423 {0x180b, 0x180d}, 2424 {0x1885, 0x1886}, 2425 {0x18a9, 0x18a9}, 2426 {0x1920, 0x192b}, 2427 {0x1930, 0x193b}, 2428 {0x1a17, 0x1a1b}, 2429 {0x1a55, 0x1a5e}, 2430 {0x1a60, 0x1a7c}, 2431 {0x1a7f, 0x1a7f}, 2432 {0x1ab0, 0x1abe}, 2433 {0x1b00, 0x1b04}, 2434 {0x1b34, 0x1b44}, 2435 {0x1b6b, 0x1b73}, 2436 {0x1b80, 0x1b82}, 2437 {0x1ba1, 0x1bad}, 2438 {0x1be6, 0x1bf3}, 2439 {0x1c24, 0x1c37}, 2440 {0x1cd0, 0x1cd2}, 2441 {0x1cd4, 0x1ce8}, 2442 {0x1ced, 0x1ced}, 2443 {0x1cf2, 0x1cf4}, 2444 {0x1cf7, 0x1cf9}, 2445 {0x1dc0, 0x1df9}, 2446 {0x1dfb, 0x1dff}, 2447 {0x20d0, 0x20f0}, 2448 {0x2cef, 0x2cf1}, 2449 {0x2d7f, 0x2d7f}, 2450 {0x2de0, 0x2dff}, 2451 {0x302a, 0x302f}, 2452 {0x3099, 0x309a}, 2453 {0xa66f, 0xa672}, 2454 {0xa674, 0xa67d}, 2455 {0xa69e, 0xa69f}, 2456 {0xa6f0, 0xa6f1}, 2457 {0xa802, 0xa802}, 2458 {0xa806, 0xa806}, 2459 {0xa80b, 0xa80b}, 2460 {0xa823, 0xa827}, 2461 {0xa880, 0xa881}, 2462 {0xa8b4, 0xa8c5}, 2463 {0xa8e0, 0xa8f1}, 2464 {0xa8ff, 0xa8ff}, 2465 {0xa926, 0xa92d}, 2466 {0xa947, 0xa953}, 2467 {0xa980, 0xa983}, 2468 {0xa9b3, 0xa9c0}, 2469 {0xa9e5, 0xa9e5}, 2470 {0xaa29, 0xaa36}, 2471 {0xaa43, 0xaa43}, 2472 {0xaa4c, 0xaa4d}, 2473 {0xaa7b, 0xaa7d}, 2474 {0xaab0, 0xaab0}, 2475 {0xaab2, 0xaab4}, 2476 {0xaab7, 0xaab8}, 2477 {0xaabe, 0xaabf}, 2478 {0xaac1, 0xaac1}, 2479 {0xaaeb, 0xaaef}, 2480 {0xaaf5, 0xaaf6}, 2481 {0xabe3, 0xabea}, 2482 {0xabec, 0xabed}, 2483 {0xfb1e, 0xfb1e}, 2484 {0xfe00, 0xfe0f}, 2485 {0xfe20, 0xfe2f}, 2486 {0x101fd, 0x101fd}, 2487 {0x102e0, 0x102e0}, 2488 {0x10376, 0x1037a}, 2489 {0x10a01, 0x10a03}, 2490 {0x10a05, 0x10a06}, 2491 {0x10a0c, 0x10a0f}, 2492 {0x10a38, 0x10a3a}, 2493 {0x10a3f, 0x10a3f}, 2494 {0x10ae5, 0x10ae6}, 2495 {0x10d24, 0x10d27}, 2496 {0x10f46, 0x10f50}, 2497 {0x11000, 0x11002}, 2498 {0x11038, 0x11046}, 2499 {0x1107f, 0x11082}, 2500 {0x110b0, 0x110ba}, 2501 {0x11100, 0x11102}, 2502 {0x11127, 0x11134}, 2503 {0x11145, 0x11146}, 2504 {0x11173, 0x11173}, 2505 {0x11180, 0x11182}, 2506 {0x111b3, 0x111c0}, 2507 {0x111c9, 0x111cc}, 2508 {0x1122c, 0x11237}, 2509 {0x1123e, 0x1123e}, 2510 {0x112df, 0x112ea}, 2511 {0x11300, 0x11303}, 2512 {0x1133b, 0x1133c}, 2513 {0x1133e, 0x11344}, 2514 {0x11347, 0x11348}, 2515 {0x1134b, 0x1134d}, 2516 {0x11357, 0x11357}, 2517 {0x11362, 0x11363}, 2518 {0x11366, 0x1136c}, 2519 {0x11370, 0x11374}, 2520 {0x11435, 0x11446}, 2521 {0x1145e, 0x1145e}, 2522 {0x114b0, 0x114c3}, 2523 {0x115af, 0x115b5}, 2524 {0x115b8, 0x115c0}, 2525 {0x115dc, 0x115dd}, 2526 {0x11630, 0x11640}, 2527 {0x116ab, 0x116b7}, 2528 {0x1171d, 0x1172b}, 2529 {0x1182c, 0x1183a}, 2530 {0x11a01, 0x11a0a}, 2531 {0x11a33, 0x11a39}, 2532 {0x11a3b, 0x11a3e}, 2533 {0x11a47, 0x11a47}, 2534 {0x11a51, 0x11a5b}, 2535 {0x11a8a, 0x11a99}, 2536 {0x11c2f, 0x11c36}, 2537 {0x11c38, 0x11c3f}, 2538 {0x11c92, 0x11ca7}, 2539 {0x11ca9, 0x11cb6}, 2540 {0x11d31, 0x11d36}, 2541 {0x11d3a, 0x11d3a}, 2542 {0x11d3c, 0x11d3d}, 2543 {0x11d3f, 0x11d45}, 2544 {0x11d47, 0x11d47}, 2545 {0x11d8a, 0x11d8e}, 2546 {0x11d90, 0x11d91}, 2547 {0x11d93, 0x11d97}, 2548 {0x11ef3, 0x11ef6}, 2549 {0x16af0, 0x16af4}, 2550 {0x16b30, 0x16b36}, 2551 {0x16f51, 0x16f7e}, 2552 {0x16f8f, 0x16f92}, 2553 {0x1bc9d, 0x1bc9e}, 2554 {0x1d165, 0x1d169}, 2555 {0x1d16d, 0x1d172}, 2556 {0x1d17b, 0x1d182}, 2557 {0x1d185, 0x1d18b}, 2558 {0x1d1aa, 0x1d1ad}, 2559 {0x1d242, 0x1d244}, 2560 {0x1da00, 0x1da36}, 2561 {0x1da3b, 0x1da6c}, 2562 {0x1da75, 0x1da75}, 2563 {0x1da84, 0x1da84}, 2564 {0x1da9b, 0x1da9f}, 2565 {0x1daa1, 0x1daaf}, 2566 {0x1e000, 0x1e006}, 2567 {0x1e008, 0x1e018}, 2568 {0x1e01b, 0x1e021}, 2569 {0x1e023, 0x1e024}, 2570 {0x1e026, 0x1e02a}, 2571 {0x1e8d0, 0x1e8d6}, 2572 {0x1e944, 0x1e94a}, 2573 {0xe0100, 0xe01ef} 2574 }; 2575 2576 return intable(combining, sizeof(combining), c); 2577 } 2578 2579 /* 2580 * Return TRUE for characters that can be displayed in a normal way. 2581 * Only for characters of 0x100 and above! 2582 */ 2583 int 2584 utf_printable(int c) 2585 { 2586 #ifdef USE_WCHAR_FUNCTIONS 2587 /* 2588 * Assume the iswprint() library function works better than our own stuff. 2589 */ 2590 return iswprint(c); 2591 #else 2592 /* Sorted list of non-overlapping intervals. 2593 * 0xd800-0xdfff is reserved for UTF-16, actually illegal. */ 2594 static struct interval nonprint[] = 2595 { 2596 {0x070f, 0x070f}, {0x180b, 0x180e}, {0x200b, 0x200f}, {0x202a, 0x202e}, 2597 {0x206a, 0x206f}, {0xd800, 0xdfff}, {0xfeff, 0xfeff}, {0xfff9, 0xfffb}, 2598 {0xfffe, 0xffff} 2599 }; 2600 2601 return !intable(nonprint, sizeof(nonprint), c); 2602 #endif 2603 } 2604 2605 /* Sorted list of non-overlapping intervals of all Emoji characters, 2606 * based on http://unicode.org/emoji/charts/emoji-list.html */ 2607 static struct interval emoji_all[] = 2608 { 2609 {0x203c, 0x203c}, 2610 {0x2049, 0x2049}, 2611 {0x2122, 0x2122}, 2612 {0x2139, 0x2139}, 2613 {0x2194, 0x2199}, 2614 {0x21a9, 0x21aa}, 2615 {0x231a, 0x231b}, 2616 {0x2328, 0x2328}, 2617 {0x23cf, 0x23cf}, 2618 {0x23e9, 0x23f3}, 2619 {0x23f8, 0x23fa}, 2620 {0x24c2, 0x24c2}, 2621 {0x25aa, 0x25ab}, 2622 {0x25b6, 0x25b6}, 2623 {0x25c0, 0x25c0}, 2624 {0x25fb, 0x25fe}, 2625 {0x2600, 0x2604}, 2626 {0x260e, 0x260e}, 2627 {0x2611, 0x2611}, 2628 {0x2614, 0x2615}, 2629 {0x2618, 0x2618}, 2630 {0x261d, 0x261d}, 2631 {0x2620, 0x2620}, 2632 {0x2622, 0x2623}, 2633 {0x2626, 0x2626}, 2634 {0x262a, 0x262a}, 2635 {0x262e, 0x262f}, 2636 {0x2638, 0x263a}, 2637 {0x2640, 0x2640}, 2638 {0x2642, 0x2642}, 2639 {0x2648, 0x2653}, 2640 {0x265f, 0x2660}, 2641 {0x2663, 0x2663}, 2642 {0x2665, 0x2666}, 2643 {0x2668, 0x2668}, 2644 {0x267b, 0x267b}, 2645 {0x267e, 0x267f}, 2646 {0x2692, 0x2697}, 2647 {0x2699, 0x2699}, 2648 {0x269b, 0x269c}, 2649 {0x26a0, 0x26a1}, 2650 {0x26aa, 0x26ab}, 2651 {0x26b0, 0x26b1}, 2652 {0x26bd, 0x26be}, 2653 {0x26c4, 0x26c5}, 2654 {0x26c8, 0x26c8}, 2655 {0x26ce, 0x26cf}, 2656 {0x26d1, 0x26d1}, 2657 {0x26d3, 0x26d4}, 2658 {0x26e9, 0x26ea}, 2659 {0x26f0, 0x26f5}, 2660 {0x26f7, 0x26fa}, 2661 {0x26fd, 0x26fd}, 2662 {0x2702, 0x2702}, 2663 {0x2705, 0x2705}, 2664 {0x2708, 0x270d}, 2665 {0x270f, 0x270f}, 2666 {0x2712, 0x2712}, 2667 {0x2714, 0x2714}, 2668 {0x2716, 0x2716}, 2669 {0x271d, 0x271d}, 2670 {0x2721, 0x2721}, 2671 {0x2728, 0x2728}, 2672 {0x2733, 0x2734}, 2673 {0x2744, 0x2744}, 2674 {0x2747, 0x2747}, 2675 {0x274c, 0x274c}, 2676 {0x274e, 0x274e}, 2677 {0x2753, 0x2755}, 2678 {0x2757, 0x2757}, 2679 {0x2763, 0x2764}, 2680 {0x2795, 0x2797}, 2681 {0x27a1, 0x27a1}, 2682 {0x27b0, 0x27b0}, 2683 {0x27bf, 0x27bf}, 2684 {0x2934, 0x2935}, 2685 {0x2b05, 0x2b07}, 2686 {0x2b1b, 0x2b1c}, 2687 {0x2b50, 0x2b50}, 2688 {0x2b55, 0x2b55}, 2689 {0x3030, 0x3030}, 2690 {0x303d, 0x303d}, 2691 {0x3297, 0x3297}, 2692 {0x3299, 0x3299}, 2693 {0x1f004, 0x1f004}, 2694 {0x1f0cf, 0x1f0cf}, 2695 {0x1f170, 0x1f171}, 2696 {0x1f17e, 0x1f17f}, 2697 {0x1f18e, 0x1f18e}, 2698 {0x1f191, 0x1f19a}, 2699 {0x1f1e6, 0x1f1ff}, 2700 {0x1f201, 0x1f202}, 2701 {0x1f21a, 0x1f21a}, 2702 {0x1f22f, 0x1f22f}, 2703 {0x1f232, 0x1f23a}, 2704 {0x1f250, 0x1f251}, 2705 {0x1f300, 0x1f321}, 2706 {0x1f324, 0x1f393}, 2707 {0x1f396, 0x1f397}, 2708 {0x1f399, 0x1f39b}, 2709 {0x1f39e, 0x1f3f0}, 2710 {0x1f3f3, 0x1f3f5}, 2711 {0x1f3f7, 0x1f4fd}, 2712 {0x1f4ff, 0x1f53d}, 2713 {0x1f549, 0x1f54e}, 2714 {0x1f550, 0x1f567}, 2715 {0x1f56f, 0x1f570}, 2716 {0x1f573, 0x1f57a}, 2717 {0x1f587, 0x1f587}, 2718 {0x1f58a, 0x1f58d}, 2719 {0x1f590, 0x1f590}, 2720 {0x1f595, 0x1f596}, 2721 {0x1f5a4, 0x1f5a5}, 2722 {0x1f5a8, 0x1f5a8}, 2723 {0x1f5b1, 0x1f5b2}, 2724 {0x1f5bc, 0x1f5bc}, 2725 {0x1f5c2, 0x1f5c4}, 2726 {0x1f5d1, 0x1f5d3}, 2727 {0x1f5dc, 0x1f5de}, 2728 {0x1f5e1, 0x1f5e1}, 2729 {0x1f5e3, 0x1f5e3}, 2730 {0x1f5e8, 0x1f5e8}, 2731 {0x1f5ef, 0x1f5ef}, 2732 {0x1f5f3, 0x1f5f3}, 2733 {0x1f5fa, 0x1f64f}, 2734 {0x1f680, 0x1f6c5}, 2735 {0x1f6cb, 0x1f6d2}, 2736 {0x1f6e0, 0x1f6e5}, 2737 {0x1f6e9, 0x1f6e9}, 2738 {0x1f6eb, 0x1f6ec}, 2739 {0x1f6f0, 0x1f6f0}, 2740 {0x1f6f3, 0x1f6f9}, 2741 {0x1f910, 0x1f93a}, 2742 {0x1f93c, 0x1f93e}, 2743 {0x1f940, 0x1f945}, 2744 {0x1f947, 0x1f970}, 2745 {0x1f973, 0x1f976}, 2746 {0x1f97a, 0x1f97a}, 2747 {0x1f97c, 0x1f9a2}, 2748 {0x1f9b0, 0x1f9b9}, 2749 {0x1f9c0, 0x1f9c2}, 2750 {0x1f9d0, 0x1f9ff} 2751 }; 2752 2753 /* 2754 * Get class of a Unicode character. 2755 * 0: white space 2756 * 1: punctuation 2757 * 2 or bigger: some class of word character. 2758 */ 2759 int 2760 utf_class(int c) 2761 { 2762 return utf_class_buf(c, curbuf); 2763 } 2764 2765 int 2766 utf_class_buf(int c, buf_T *buf) 2767 { 2768 /* sorted list of non-overlapping intervals */ 2769 static struct clinterval 2770 { 2771 unsigned int first; 2772 unsigned int last; 2773 unsigned int class; 2774 } classes[] = 2775 { 2776 {0x037e, 0x037e, 1}, /* Greek question mark */ 2777 {0x0387, 0x0387, 1}, /* Greek ano teleia */ 2778 {0x055a, 0x055f, 1}, /* Armenian punctuation */ 2779 {0x0589, 0x0589, 1}, /* Armenian full stop */ 2780 {0x05be, 0x05be, 1}, 2781 {0x05c0, 0x05c0, 1}, 2782 {0x05c3, 0x05c3, 1}, 2783 {0x05f3, 0x05f4, 1}, 2784 {0x060c, 0x060c, 1}, 2785 {0x061b, 0x061b, 1}, 2786 {0x061f, 0x061f, 1}, 2787 {0x066a, 0x066d, 1}, 2788 {0x06d4, 0x06d4, 1}, 2789 {0x0700, 0x070d, 1}, /* Syriac punctuation */ 2790 {0x0964, 0x0965, 1}, 2791 {0x0970, 0x0970, 1}, 2792 {0x0df4, 0x0df4, 1}, 2793 {0x0e4f, 0x0e4f, 1}, 2794 {0x0e5a, 0x0e5b, 1}, 2795 {0x0f04, 0x0f12, 1}, 2796 {0x0f3a, 0x0f3d, 1}, 2797 {0x0f85, 0x0f85, 1}, 2798 {0x104a, 0x104f, 1}, /* Myanmar punctuation */ 2799 {0x10fb, 0x10fb, 1}, /* Georgian punctuation */ 2800 {0x1361, 0x1368, 1}, /* Ethiopic punctuation */ 2801 {0x166d, 0x166e, 1}, /* Canadian Syl. punctuation */ 2802 {0x1680, 0x1680, 0}, 2803 {0x169b, 0x169c, 1}, 2804 {0x16eb, 0x16ed, 1}, 2805 {0x1735, 0x1736, 1}, 2806 {0x17d4, 0x17dc, 1}, /* Khmer punctuation */ 2807 {0x1800, 0x180a, 1}, /* Mongolian punctuation */ 2808 {0x2000, 0x200b, 0}, /* spaces */ 2809 {0x200c, 0x2027, 1}, /* punctuation and symbols */ 2810 {0x2028, 0x2029, 0}, 2811 {0x202a, 0x202e, 1}, /* punctuation and symbols */ 2812 {0x202f, 0x202f, 0}, 2813 {0x2030, 0x205e, 1}, /* punctuation and symbols */ 2814 {0x205f, 0x205f, 0}, 2815 {0x2060, 0x27ff, 1}, /* punctuation and symbols */ 2816 {0x2070, 0x207f, 0x2070}, /* superscript */ 2817 {0x2080, 0x2094, 0x2080}, /* subscript */ 2818 {0x20a0, 0x27ff, 1}, /* all kinds of symbols */ 2819 {0x2800, 0x28ff, 0x2800}, /* braille */ 2820 {0x2900, 0x2998, 1}, /* arrows, brackets, etc. */ 2821 {0x29d8, 0x29db, 1}, 2822 {0x29fc, 0x29fd, 1}, 2823 {0x2e00, 0x2e7f, 1}, /* supplemental punctuation */ 2824 {0x3000, 0x3000, 0}, /* ideographic space */ 2825 {0x3001, 0x3020, 1}, /* ideographic punctuation */ 2826 {0x3030, 0x3030, 1}, 2827 {0x303d, 0x303d, 1}, 2828 {0x3040, 0x309f, 0x3040}, /* Hiragana */ 2829 {0x30a0, 0x30ff, 0x30a0}, /* Katakana */ 2830 {0x3300, 0x9fff, 0x4e00}, /* CJK Ideographs */ 2831 {0xac00, 0xd7a3, 0xac00}, /* Hangul Syllables */ 2832 {0xf900, 0xfaff, 0x4e00}, /* CJK Ideographs */ 2833 {0xfd3e, 0xfd3f, 1}, 2834 {0xfe30, 0xfe6b, 1}, /* punctuation forms */ 2835 {0xff00, 0xff0f, 1}, /* half/fullwidth ASCII */ 2836 {0xff1a, 0xff20, 1}, /* half/fullwidth ASCII */ 2837 {0xff3b, 0xff40, 1}, /* half/fullwidth ASCII */ 2838 {0xff5b, 0xff65, 1}, /* half/fullwidth ASCII */ 2839 {0x20000, 0x2a6df, 0x4e00}, /* CJK Ideographs */ 2840 {0x2a700, 0x2b73f, 0x4e00}, /* CJK Ideographs */ 2841 {0x2b740, 0x2b81f, 0x4e00}, /* CJK Ideographs */ 2842 {0x2f800, 0x2fa1f, 0x4e00}, /* CJK Ideographs */ 2843 }; 2844 2845 int bot = 0; 2846 int top = sizeof(classes) / sizeof(struct clinterval) - 1; 2847 int mid; 2848 2849 /* First quick check for Latin1 characters, use 'iskeyword'. */ 2850 if (c < 0x100) 2851 { 2852 if (c == ' ' || c == '\t' || c == NUL || c == 0xa0) 2853 return 0; /* blank */ 2854 if (vim_iswordc_buf(c, buf)) 2855 return 2; /* word character */ 2856 return 1; /* punctuation */ 2857 } 2858 2859 /* binary search in table */ 2860 while (top >= bot) 2861 { 2862 mid = (bot + top) / 2; 2863 if (classes[mid].last < (unsigned int)c) 2864 bot = mid + 1; 2865 else if (classes[mid].first > (unsigned int)c) 2866 top = mid - 1; 2867 else 2868 return (int)classes[mid].class; 2869 } 2870 2871 /* emoji */ 2872 if (intable(emoji_all, sizeof(emoji_all), c)) 2873 return 3; 2874 2875 /* most other characters are "word" characters */ 2876 return 2; 2877 } 2878 2879 int 2880 utf_ambiguous_width(int c) 2881 { 2882 return c >= 0x80 && (intable(ambiguous, sizeof(ambiguous), c) 2883 || intable(emoji_all, sizeof(emoji_all), c)); 2884 } 2885 2886 /* 2887 * Code for Unicode case-dependent operations. Based on notes in 2888 * http://www.unicode.org/Public/UNIDATA/CaseFolding.txt 2889 * This code uses simple case folding, not full case folding. 2890 * Last updated for Unicode 5.2. 2891 */ 2892 2893 /* 2894 * The following tables are built by ../runtime/tools/unicode.vim. 2895 * They must be in numeric order, because we use binary search. 2896 * An entry such as {0x41,0x5a,1,32} means that Unicode characters in the 2897 * range from 0x41 to 0x5a inclusive, stepping by 1, are changed to 2898 * folded/upper/lower by adding 32. 2899 */ 2900 typedef struct 2901 { 2902 int rangeStart; 2903 int rangeEnd; 2904 int step; 2905 int offset; 2906 } convertStruct; 2907 2908 static convertStruct foldCase[] = 2909 { 2910 {0x41,0x5a,1,32}, 2911 {0xb5,0xb5,-1,775}, 2912 {0xc0,0xd6,1,32}, 2913 {0xd8,0xde,1,32}, 2914 {0x100,0x12e,2,1}, 2915 {0x132,0x136,2,1}, 2916 {0x139,0x147,2,1}, 2917 {0x14a,0x176,2,1}, 2918 {0x178,0x178,-1,-121}, 2919 {0x179,0x17d,2,1}, 2920 {0x17f,0x17f,-1,-268}, 2921 {0x181,0x181,-1,210}, 2922 {0x182,0x184,2,1}, 2923 {0x186,0x186,-1,206}, 2924 {0x187,0x187,-1,1}, 2925 {0x189,0x18a,1,205}, 2926 {0x18b,0x18b,-1,1}, 2927 {0x18e,0x18e,-1,79}, 2928 {0x18f,0x18f,-1,202}, 2929 {0x190,0x190,-1,203}, 2930 {0x191,0x191,-1,1}, 2931 {0x193,0x193,-1,205}, 2932 {0x194,0x194,-1,207}, 2933 {0x196,0x196,-1,211}, 2934 {0x197,0x197,-1,209}, 2935 {0x198,0x198,-1,1}, 2936 {0x19c,0x19c,-1,211}, 2937 {0x19d,0x19d,-1,213}, 2938 {0x19f,0x19f,-1,214}, 2939 {0x1a0,0x1a4,2,1}, 2940 {0x1a6,0x1a6,-1,218}, 2941 {0x1a7,0x1a7,-1,1}, 2942 {0x1a9,0x1a9,-1,218}, 2943 {0x1ac,0x1ac,-1,1}, 2944 {0x1ae,0x1ae,-1,218}, 2945 {0x1af,0x1af,-1,1}, 2946 {0x1b1,0x1b2,1,217}, 2947 {0x1b3,0x1b5,2,1}, 2948 {0x1b7,0x1b7,-1,219}, 2949 {0x1b8,0x1bc,4,1}, 2950 {0x1c4,0x1c4,-1,2}, 2951 {0x1c5,0x1c5,-1,1}, 2952 {0x1c7,0x1c7,-1,2}, 2953 {0x1c8,0x1c8,-1,1}, 2954 {0x1ca,0x1ca,-1,2}, 2955 {0x1cb,0x1db,2,1}, 2956 {0x1de,0x1ee,2,1}, 2957 {0x1f1,0x1f1,-1,2}, 2958 {0x1f2,0x1f4,2,1}, 2959 {0x1f6,0x1f6,-1,-97}, 2960 {0x1f7,0x1f7,-1,-56}, 2961 {0x1f8,0x21e,2,1}, 2962 {0x220,0x220,-1,-130}, 2963 {0x222,0x232,2,1}, 2964 {0x23a,0x23a,-1,10795}, 2965 {0x23b,0x23b,-1,1}, 2966 {0x23d,0x23d,-1,-163}, 2967 {0x23e,0x23e,-1,10792}, 2968 {0x241,0x241,-1,1}, 2969 {0x243,0x243,-1,-195}, 2970 {0x244,0x244,-1,69}, 2971 {0x245,0x245,-1,71}, 2972 {0x246,0x24e,2,1}, 2973 {0x345,0x345,-1,116}, 2974 {0x370,0x372,2,1}, 2975 {0x376,0x376,-1,1}, 2976 {0x37f,0x37f,-1,116}, 2977 {0x386,0x386,-1,38}, 2978 {0x388,0x38a,1,37}, 2979 {0x38c,0x38c,-1,64}, 2980 {0x38e,0x38f,1,63}, 2981 {0x391,0x3a1,1,32}, 2982 {0x3a3,0x3ab,1,32}, 2983 {0x3c2,0x3c2,-1,1}, 2984 {0x3cf,0x3cf,-1,8}, 2985 {0x3d0,0x3d0,-1,-30}, 2986 {0x3d1,0x3d1,-1,-25}, 2987 {0x3d5,0x3d5,-1,-15}, 2988 {0x3d6,0x3d6,-1,-22}, 2989 {0x3d8,0x3ee,2,1}, 2990 {0x3f0,0x3f0,-1,-54}, 2991 {0x3f1,0x3f1,-1,-48}, 2992 {0x3f4,0x3f4,-1,-60}, 2993 {0x3f5,0x3f5,-1,-64}, 2994 {0x3f7,0x3f7,-1,1}, 2995 {0x3f9,0x3f9,-1,-7}, 2996 {0x3fa,0x3fa,-1,1}, 2997 {0x3fd,0x3ff,1,-130}, 2998 {0x400,0x40f,1,80}, 2999 {0x410,0x42f,1,32}, 3000 {0x460,0x480,2,1}, 3001 {0x48a,0x4be,2,1}, 3002 {0x4c0,0x4c0,-1,15}, 3003 {0x4c1,0x4cd,2,1}, 3004 {0x4d0,0x52e,2,1}, 3005 {0x531,0x556,1,48}, 3006 {0x10a0,0x10c5,1,7264}, 3007 {0x10c7,0x10cd,6,7264}, 3008 {0x13f8,0x13fd,1,-8}, 3009 {0x1c80,0x1c80,-1,-6222}, 3010 {0x1c81,0x1c81,-1,-6221}, 3011 {0x1c82,0x1c82,-1,-6212}, 3012 {0x1c83,0x1c84,1,-6210}, 3013 {0x1c85,0x1c85,-1,-6211}, 3014 {0x1c86,0x1c86,-1,-6204}, 3015 {0x1c87,0x1c87,-1,-6180}, 3016 {0x1c88,0x1c88,-1,35267}, 3017 {0x1c90,0x1cba,1,-3008}, 3018 {0x1cbd,0x1cbf,1,-3008}, 3019 {0x1e00,0x1e94,2,1}, 3020 {0x1e9b,0x1e9b,-1,-58}, 3021 {0x1e9e,0x1e9e,-1,-7615}, 3022 {0x1ea0,0x1efe,2,1}, 3023 {0x1f08,0x1f0f,1,-8}, 3024 {0x1f18,0x1f1d,1,-8}, 3025 {0x1f28,0x1f2f,1,-8}, 3026 {0x1f38,0x1f3f,1,-8}, 3027 {0x1f48,0x1f4d,1,-8}, 3028 {0x1f59,0x1f5f,2,-8}, 3029 {0x1f68,0x1f6f,1,-8}, 3030 {0x1f88,0x1f8f,1,-8}, 3031 {0x1f98,0x1f9f,1,-8}, 3032 {0x1fa8,0x1faf,1,-8}, 3033 {0x1fb8,0x1fb9,1,-8}, 3034 {0x1fba,0x1fbb,1,-74}, 3035 {0x1fbc,0x1fbc,-1,-9}, 3036 {0x1fbe,0x1fbe,-1,-7173}, 3037 {0x1fc8,0x1fcb,1,-86}, 3038 {0x1fcc,0x1fcc,-1,-9}, 3039 {0x1fd8,0x1fd9,1,-8}, 3040 {0x1fda,0x1fdb,1,-100}, 3041 {0x1fe8,0x1fe9,1,-8}, 3042 {0x1fea,0x1feb,1,-112}, 3043 {0x1fec,0x1fec,-1,-7}, 3044 {0x1ff8,0x1ff9,1,-128}, 3045 {0x1ffa,0x1ffb,1,-126}, 3046 {0x1ffc,0x1ffc,-1,-9}, 3047 {0x2126,0x2126,-1,-7517}, 3048 {0x212a,0x212a,-1,-8383}, 3049 {0x212b,0x212b,-1,-8262}, 3050 {0x2132,0x2132,-1,28}, 3051 {0x2160,0x216f,1,16}, 3052 {0x2183,0x2183,-1,1}, 3053 {0x24b6,0x24cf,1,26}, 3054 {0x2c00,0x2c2e,1,48}, 3055 {0x2c60,0x2c60,-1,1}, 3056 {0x2c62,0x2c62,-1,-10743}, 3057 {0x2c63,0x2c63,-1,-3814}, 3058 {0x2c64,0x2c64,-1,-10727}, 3059 {0x2c67,0x2c6b,2,1}, 3060 {0x2c6d,0x2c6d,-1,-10780}, 3061 {0x2c6e,0x2c6e,-1,-10749}, 3062 {0x2c6f,0x2c6f,-1,-10783}, 3063 {0x2c70,0x2c70,-1,-10782}, 3064 {0x2c72,0x2c75,3,1}, 3065 {0x2c7e,0x2c7f,1,-10815}, 3066 {0x2c80,0x2ce2,2,1}, 3067 {0x2ceb,0x2ced,2,1}, 3068 {0x2cf2,0xa640,31054,1}, 3069 {0xa642,0xa66c,2,1}, 3070 {0xa680,0xa69a,2,1}, 3071 {0xa722,0xa72e,2,1}, 3072 {0xa732,0xa76e,2,1}, 3073 {0xa779,0xa77b,2,1}, 3074 {0xa77d,0xa77d,-1,-35332}, 3075 {0xa77e,0xa786,2,1}, 3076 {0xa78b,0xa78b,-1,1}, 3077 {0xa78d,0xa78d,-1,-42280}, 3078 {0xa790,0xa792,2,1}, 3079 {0xa796,0xa7a8,2,1}, 3080 {0xa7aa,0xa7aa,-1,-42308}, 3081 {0xa7ab,0xa7ab,-1,-42319}, 3082 {0xa7ac,0xa7ac,-1,-42315}, 3083 {0xa7ad,0xa7ad,-1,-42305}, 3084 {0xa7ae,0xa7ae,-1,-42308}, 3085 {0xa7b0,0xa7b0,-1,-42258}, 3086 {0xa7b1,0xa7b1,-1,-42282}, 3087 {0xa7b2,0xa7b2,-1,-42261}, 3088 {0xa7b3,0xa7b3,-1,928}, 3089 {0xa7b4,0xa7b8,2,1}, 3090 {0xab70,0xabbf,1,-38864}, 3091 {0xff21,0xff3a,1,32}, 3092 {0x10400,0x10427,1,40}, 3093 {0x104b0,0x104d3,1,40}, 3094 {0x10c80,0x10cb2,1,64}, 3095 {0x118a0,0x118bf,1,32}, 3096 {0x16e40,0x16e5f,1,32}, 3097 {0x1e900,0x1e921,1,34} 3098 }; 3099 3100 /* 3101 * Generic conversion function for case operations. 3102 * Return the converted equivalent of "a", which is a UCS-4 character. Use 3103 * the given conversion "table". Uses binary search on "table". 3104 */ 3105 static int 3106 utf_convert( 3107 int a, 3108 convertStruct table[], 3109 int tableSize) 3110 { 3111 int start, mid, end; /* indices into table */ 3112 int entries = tableSize / sizeof(convertStruct); 3113 3114 start = 0; 3115 end = entries; 3116 while (start < end) 3117 { 3118 /* need to search further */ 3119 mid = (end + start) / 2; 3120 if (table[mid].rangeEnd < a) 3121 start = mid + 1; 3122 else 3123 end = mid; 3124 } 3125 if (start < entries 3126 && table[start].rangeStart <= a 3127 && a <= table[start].rangeEnd 3128 && (a - table[start].rangeStart) % table[start].step == 0) 3129 return (a + table[start].offset); 3130 else 3131 return a; 3132 } 3133 3134 /* 3135 * Return the folded-case equivalent of "a", which is a UCS-4 character. Uses 3136 * simple case folding. 3137 */ 3138 int 3139 utf_fold(int a) 3140 { 3141 if (a < 0x80) 3142 /* be fast for ASCII */ 3143 return a >= 0x41 && a <= 0x5a ? a + 32 : a; 3144 return utf_convert(a, foldCase, (int)sizeof(foldCase)); 3145 } 3146 3147 static convertStruct toLower[] = 3148 { 3149 {0x41,0x5a,1,32}, 3150 {0xc0,0xd6,1,32}, 3151 {0xd8,0xde,1,32}, 3152 {0x100,0x12e,2,1}, 3153 {0x130,0x130,-1,-199}, 3154 {0x132,0x136,2,1}, 3155 {0x139,0x147,2,1}, 3156 {0x14a,0x176,2,1}, 3157 {0x178,0x178,-1,-121}, 3158 {0x179,0x17d,2,1}, 3159 {0x181,0x181,-1,210}, 3160 {0x182,0x184,2,1}, 3161 {0x186,0x186,-1,206}, 3162 {0x187,0x187,-1,1}, 3163 {0x189,0x18a,1,205}, 3164 {0x18b,0x18b,-1,1}, 3165 {0x18e,0x18e,-1,79}, 3166 {0x18f,0x18f,-1,202}, 3167 {0x190,0x190,-1,203}, 3168 {0x191,0x191,-1,1}, 3169 {0x193,0x193,-1,205}, 3170 {0x194,0x194,-1,207}, 3171 {0x196,0x196,-1,211}, 3172 {0x197,0x197,-1,209}, 3173 {0x198,0x198,-1,1}, 3174 {0x19c,0x19c,-1,211}, 3175 {0x19d,0x19d,-1,213}, 3176 {0x19f,0x19f,-1,214}, 3177 {0x1a0,0x1a4,2,1}, 3178 {0x1a6,0x1a6,-1,218}, 3179 {0x1a7,0x1a7,-1,1}, 3180 {0x1a9,0x1a9,-1,218}, 3181 {0x1ac,0x1ac,-1,1}, 3182 {0x1ae,0x1ae,-1,218}, 3183 {0x1af,0x1af,-1,1}, 3184 {0x1b1,0x1b2,1,217}, 3185 {0x1b3,0x1b5,2,1}, 3186 {0x1b7,0x1b7,-1,219}, 3187 {0x1b8,0x1bc,4,1}, 3188 {0x1c4,0x1c4,-1,2}, 3189 {0x1c5,0x1c5,-1,1}, 3190 {0x1c7,0x1c7,-1,2}, 3191 {0x1c8,0x1c8,-1,1}, 3192 {0x1ca,0x1ca,-1,2}, 3193 {0x1cb,0x1db,2,1}, 3194 {0x1de,0x1ee,2,1}, 3195 {0x1f1,0x1f1,-1,2}, 3196 {0x1f2,0x1f4,2,1}, 3197 {0x1f6,0x1f6,-1,-97}, 3198 {0x1f7,0x1f7,-1,-56}, 3199 {0x1f8,0x21e,2,1}, 3200 {0x220,0x220,-1,-130}, 3201 {0x222,0x232,2,1}, 3202 {0x23a,0x23a,-1,10795}, 3203 {0x23b,0x23b,-1,1}, 3204 {0x23d,0x23d,-1,-163}, 3205 {0x23e,0x23e,-1,10792}, 3206 {0x241,0x241,-1,1}, 3207 {0x243,0x243,-1,-195}, 3208 {0x244,0x244,-1,69}, 3209 {0x245,0x245,-1,71}, 3210 {0x246,0x24e,2,1}, 3211 {0x370,0x372,2,1}, 3212 {0x376,0x376,-1,1}, 3213 {0x37f,0x37f,-1,116}, 3214 {0x386,0x386,-1,38}, 3215 {0x388,0x38a,1,37}, 3216 {0x38c,0x38c,-1,64}, 3217 {0x38e,0x38f,1,63}, 3218 {0x391,0x3a1,1,32}, 3219 {0x3a3,0x3ab,1,32}, 3220 {0x3cf,0x3cf,-1,8}, 3221 {0x3d8,0x3ee,2,1}, 3222 {0x3f4,0x3f4,-1,-60}, 3223 {0x3f7,0x3f7,-1,1}, 3224 {0x3f9,0x3f9,-1,-7}, 3225 {0x3fa,0x3fa,-1,1}, 3226 {0x3fd,0x3ff,1,-130}, 3227 {0x400,0x40f,1,80}, 3228 {0x410,0x42f,1,32}, 3229 {0x460,0x480,2,1}, 3230 {0x48a,0x4be,2,1}, 3231 {0x4c0,0x4c0,-1,15}, 3232 {0x4c1,0x4cd,2,1}, 3233 {0x4d0,0x52e,2,1}, 3234 {0x531,0x556,1,48}, 3235 {0x10a0,0x10c5,1,7264}, 3236 {0x10c7,0x10cd,6,7264}, 3237 {0x13a0,0x13ef,1,38864}, 3238 {0x13f0,0x13f5,1,8}, 3239 {0x1c90,0x1cba,1,-3008}, 3240 {0x1cbd,0x1cbf,1,-3008}, 3241 {0x1e00,0x1e94,2,1}, 3242 {0x1e9e,0x1e9e,-1,-7615}, 3243 {0x1ea0,0x1efe,2,1}, 3244 {0x1f08,0x1f0f,1,-8}, 3245 {0x1f18,0x1f1d,1,-8}, 3246 {0x1f28,0x1f2f,1,-8}, 3247 {0x1f38,0x1f3f,1,-8}, 3248 {0x1f48,0x1f4d,1,-8}, 3249 {0x1f59,0x1f5f,2,-8}, 3250 {0x1f68,0x1f6f,1,-8}, 3251 {0x1f88,0x1f8f,1,-8}, 3252 {0x1f98,0x1f9f,1,-8}, 3253 {0x1fa8,0x1faf,1,-8}, 3254 {0x1fb8,0x1fb9,1,-8}, 3255 {0x1fba,0x1fbb,1,-74}, 3256 {0x1fbc,0x1fbc,-1,-9}, 3257 {0x1fc8,0x1fcb,1,-86}, 3258 {0x1fcc,0x1fcc,-1,-9}, 3259 {0x1fd8,0x1fd9,1,-8}, 3260 {0x1fda,0x1fdb,1,-100}, 3261 {0x1fe8,0x1fe9,1,-8}, 3262 {0x1fea,0x1feb,1,-112}, 3263 {0x1fec,0x1fec,-1,-7}, 3264 {0x1ff8,0x1ff9,1,-128}, 3265 {0x1ffa,0x1ffb,1,-126}, 3266 {0x1ffc,0x1ffc,-1,-9}, 3267 {0x2126,0x2126,-1,-7517}, 3268 {0x212a,0x212a,-1,-8383}, 3269 {0x212b,0x212b,-1,-8262}, 3270 {0x2132,0x2132,-1,28}, 3271 {0x2160,0x216f,1,16}, 3272 {0x2183,0x2183,-1,1}, 3273 {0x24b6,0x24cf,1,26}, 3274 {0x2c00,0x2c2e,1,48}, 3275 {0x2c60,0x2c60,-1,1}, 3276 {0x2c62,0x2c62,-1,-10743}, 3277 {0x2c63,0x2c63,-1,-3814}, 3278 {0x2c64,0x2c64,-1,-10727}, 3279 {0x2c67,0x2c6b,2,1}, 3280 {0x2c6d,0x2c6d,-1,-10780}, 3281 {0x2c6e,0x2c6e,-1,-10749}, 3282 {0x2c6f,0x2c6f,-1,-10783}, 3283 {0x2c70,0x2c70,-1,-10782}, 3284 {0x2c72,0x2c75,3,1}, 3285 {0x2c7e,0x2c7f,1,-10815}, 3286 {0x2c80,0x2ce2,2,1}, 3287 {0x2ceb,0x2ced,2,1}, 3288 {0x2cf2,0xa640,31054,1}, 3289 {0xa642,0xa66c,2,1}, 3290 {0xa680,0xa69a,2,1}, 3291 {0xa722,0xa72e,2,1}, 3292 {0xa732,0xa76e,2,1}, 3293 {0xa779,0xa77b,2,1}, 3294 {0xa77d,0xa77d,-1,-35332}, 3295 {0xa77e,0xa786,2,1}, 3296 {0xa78b,0xa78b,-1,1}, 3297 {0xa78d,0xa78d,-1,-42280}, 3298 {0xa790,0xa792,2,1}, 3299 {0xa796,0xa7a8,2,1}, 3300 {0xa7aa,0xa7aa,-1,-42308}, 3301 {0xa7ab,0xa7ab,-1,-42319}, 3302 {0xa7ac,0xa7ac,-1,-42315}, 3303 {0xa7ad,0xa7ad,-1,-42305}, 3304 {0xa7ae,0xa7ae,-1,-42308}, 3305 {0xa7b0,0xa7b0,-1,-42258}, 3306 {0xa7b1,0xa7b1,-1,-42282}, 3307 {0xa7b2,0xa7b2,-1,-42261}, 3308 {0xa7b3,0xa7b3,-1,928}, 3309 {0xa7b4,0xa7b8,2,1}, 3310 {0xff21,0xff3a,1,32}, 3311 {0x10400,0x10427,1,40}, 3312 {0x104b0,0x104d3,1,40}, 3313 {0x10c80,0x10cb2,1,64}, 3314 {0x118a0,0x118bf,1,32}, 3315 {0x16e40,0x16e5f,1,32}, 3316 {0x1e900,0x1e921,1,34} 3317 }; 3318 3319 static convertStruct toUpper[] = 3320 { 3321 {0x61,0x7a,1,-32}, 3322 {0xb5,0xb5,-1,743}, 3323 {0xe0,0xf6,1,-32}, 3324 {0xf8,0xfe,1,-32}, 3325 {0xff,0xff,-1,121}, 3326 {0x101,0x12f,2,-1}, 3327 {0x131,0x131,-1,-232}, 3328 {0x133,0x137,2,-1}, 3329 {0x13a,0x148,2,-1}, 3330 {0x14b,0x177,2,-1}, 3331 {0x17a,0x17e,2,-1}, 3332 {0x17f,0x17f,-1,-300}, 3333 {0x180,0x180,-1,195}, 3334 {0x183,0x185,2,-1}, 3335 {0x188,0x18c,4,-1}, 3336 {0x192,0x192,-1,-1}, 3337 {0x195,0x195,-1,97}, 3338 {0x199,0x199,-1,-1}, 3339 {0x19a,0x19a,-1,163}, 3340 {0x19e,0x19e,-1,130}, 3341 {0x1a1,0x1a5,2,-1}, 3342 {0x1a8,0x1ad,5,-1}, 3343 {0x1b0,0x1b4,4,-1}, 3344 {0x1b6,0x1b9,3,-1}, 3345 {0x1bd,0x1bd,-1,-1}, 3346 {0x1bf,0x1bf,-1,56}, 3347 {0x1c5,0x1c5,-1,-1}, 3348 {0x1c6,0x1c6,-1,-2}, 3349 {0x1c8,0x1c8,-1,-1}, 3350 {0x1c9,0x1c9,-1,-2}, 3351 {0x1cb,0x1cb,-1,-1}, 3352 {0x1cc,0x1cc,-1,-2}, 3353 {0x1ce,0x1dc,2,-1}, 3354 {0x1dd,0x1dd,-1,-79}, 3355 {0x1df,0x1ef,2,-1}, 3356 {0x1f2,0x1f2,-1,-1}, 3357 {0x1f3,0x1f3,-1,-2}, 3358 {0x1f5,0x1f9,4,-1}, 3359 {0x1fb,0x21f,2,-1}, 3360 {0x223,0x233,2,-1}, 3361 {0x23c,0x23c,-1,-1}, 3362 {0x23f,0x240,1,10815}, 3363 {0x242,0x247,5,-1}, 3364 {0x249,0x24f,2,-1}, 3365 {0x250,0x250,-1,10783}, 3366 {0x251,0x251,-1,10780}, 3367 {0x252,0x252,-1,10782}, 3368 {0x253,0x253,-1,-210}, 3369 {0x254,0x254,-1,-206}, 3370 {0x256,0x257,1,-205}, 3371 {0x259,0x259,-1,-202}, 3372 {0x25b,0x25b,-1,-203}, 3373 {0x25c,0x25c,-1,42319}, 3374 {0x260,0x260,-1,-205}, 3375 {0x261,0x261,-1,42315}, 3376 {0x263,0x263,-1,-207}, 3377 {0x265,0x265,-1,42280}, 3378 {0x266,0x266,-1,42308}, 3379 {0x268,0x268,-1,-209}, 3380 {0x269,0x269,-1,-211}, 3381 {0x26a,0x26a,-1,42308}, 3382 {0x26b,0x26b,-1,10743}, 3383 {0x26c,0x26c,-1,42305}, 3384 {0x26f,0x26f,-1,-211}, 3385 {0x271,0x271,-1,10749}, 3386 {0x272,0x272,-1,-213}, 3387 {0x275,0x275,-1,-214}, 3388 {0x27d,0x27d,-1,10727}, 3389 {0x280,0x283,3,-218}, 3390 {0x287,0x287,-1,42282}, 3391 {0x288,0x288,-1,-218}, 3392 {0x289,0x289,-1,-69}, 3393 {0x28a,0x28b,1,-217}, 3394 {0x28c,0x28c,-1,-71}, 3395 {0x292,0x292,-1,-219}, 3396 {0x29d,0x29d,-1,42261}, 3397 {0x29e,0x29e,-1,42258}, 3398 {0x345,0x345,-1,84}, 3399 {0x371,0x373,2,-1}, 3400 {0x377,0x377,-1,-1}, 3401 {0x37b,0x37d,1,130}, 3402 {0x3ac,0x3ac,-1,-38}, 3403 {0x3ad,0x3af,1,-37}, 3404 {0x3b1,0x3c1,1,-32}, 3405 {0x3c2,0x3c2,-1,-31}, 3406 {0x3c3,0x3cb,1,-32}, 3407 {0x3cc,0x3cc,-1,-64}, 3408 {0x3cd,0x3ce,1,-63}, 3409 {0x3d0,0x3d0,-1,-62}, 3410 {0x3d1,0x3d1,-1,-57}, 3411 {0x3d5,0x3d5,-1,-47}, 3412 {0x3d6,0x3d6,-1,-54}, 3413 {0x3d7,0x3d7,-1,-8}, 3414 {0x3d9,0x3ef,2,-1}, 3415 {0x3f0,0x3f0,-1,-86}, 3416 {0x3f1,0x3f1,-1,-80}, 3417 {0x3f2,0x3f2,-1,7}, 3418 {0x3f3,0x3f3,-1,-116}, 3419 {0x3f5,0x3f5,-1,-96}, 3420 {0x3f8,0x3fb,3,-1}, 3421 {0x430,0x44f,1,-32}, 3422 {0x450,0x45f,1,-80}, 3423 {0x461,0x481,2,-1}, 3424 {0x48b,0x4bf,2,-1}, 3425 {0x4c2,0x4ce,2,-1}, 3426 {0x4cf,0x4cf,-1,-15}, 3427 {0x4d1,0x52f,2,-1}, 3428 {0x561,0x586,1,-48}, 3429 {0x10d0,0x10fa,1,3008}, 3430 {0x10fd,0x10ff,1,3008}, 3431 {0x13f8,0x13fd,1,-8}, 3432 {0x1c80,0x1c80,-1,-6254}, 3433 {0x1c81,0x1c81,-1,-6253}, 3434 {0x1c82,0x1c82,-1,-6244}, 3435 {0x1c83,0x1c84,1,-6242}, 3436 {0x1c85,0x1c85,-1,-6243}, 3437 {0x1c86,0x1c86,-1,-6236}, 3438 {0x1c87,0x1c87,-1,-6181}, 3439 {0x1c88,0x1c88,-1,35266}, 3440 {0x1d79,0x1d79,-1,35332}, 3441 {0x1d7d,0x1d7d,-1,3814}, 3442 {0x1e01,0x1e95,2,-1}, 3443 {0x1e9b,0x1e9b,-1,-59}, 3444 {0x1ea1,0x1eff,2,-1}, 3445 {0x1f00,0x1f07,1,8}, 3446 {0x1f10,0x1f15,1,8}, 3447 {0x1f20,0x1f27,1,8}, 3448 {0x1f30,0x1f37,1,8}, 3449 {0x1f40,0x1f45,1,8}, 3450 {0x1f51,0x1f57,2,8}, 3451 {0x1f60,0x1f67,1,8}, 3452 {0x1f70,0x1f71,1,74}, 3453 {0x1f72,0x1f75,1,86}, 3454 {0x1f76,0x1f77,1,100}, 3455 {0x1f78,0x1f79,1,128}, 3456 {0x1f7a,0x1f7b,1,112}, 3457 {0x1f7c,0x1f7d,1,126}, 3458 {0x1f80,0x1f87,1,8}, 3459 {0x1f90,0x1f97,1,8}, 3460 {0x1fa0,0x1fa7,1,8}, 3461 {0x1fb0,0x1fb1,1,8}, 3462 {0x1fb3,0x1fb3,-1,9}, 3463 {0x1fbe,0x1fbe,-1,-7205}, 3464 {0x1fc3,0x1fc3,-1,9}, 3465 {0x1fd0,0x1fd1,1,8}, 3466 {0x1fe0,0x1fe1,1,8}, 3467 {0x1fe5,0x1fe5,-1,7}, 3468 {0x1ff3,0x1ff3,-1,9}, 3469 {0x214e,0x214e,-1,-28}, 3470 {0x2170,0x217f,1,-16}, 3471 {0x2184,0x2184,-1,-1}, 3472 {0x24d0,0x24e9,1,-26}, 3473 {0x2c30,0x2c5e,1,-48}, 3474 {0x2c61,0x2c61,-1,-1}, 3475 {0x2c65,0x2c65,-1,-10795}, 3476 {0x2c66,0x2c66,-1,-10792}, 3477 {0x2c68,0x2c6c,2,-1}, 3478 {0x2c73,0x2c76,3,-1}, 3479 {0x2c81,0x2ce3,2,-1}, 3480 {0x2cec,0x2cee,2,-1}, 3481 {0x2cf3,0x2cf3,-1,-1}, 3482 {0x2d00,0x2d25,1,-7264}, 3483 {0x2d27,0x2d2d,6,-7264}, 3484 {0xa641,0xa66d,2,-1}, 3485 {0xa681,0xa69b,2,-1}, 3486 {0xa723,0xa72f,2,-1}, 3487 {0xa733,0xa76f,2,-1}, 3488 {0xa77a,0xa77c,2,-1}, 3489 {0xa77f,0xa787,2,-1}, 3490 {0xa78c,0xa791,5,-1}, 3491 {0xa793,0xa797,4,-1}, 3492 {0xa799,0xa7a9,2,-1}, 3493 {0xa7b5,0xa7b9,2,-1}, 3494 {0xab53,0xab53,-1,-928}, 3495 {0xab70,0xabbf,1,-38864}, 3496 {0xff41,0xff5a,1,-32}, 3497 {0x10428,0x1044f,1,-40}, 3498 {0x104d8,0x104fb,1,-40}, 3499 {0x10cc0,0x10cf2,1,-64}, 3500 {0x118c0,0x118df,1,-32}, 3501 {0x16e60,0x16e7f,1,-32}, 3502 {0x1e922,0x1e943,1,-34} 3503 }; 3504 3505 /* 3506 * Return the upper-case equivalent of "a", which is a UCS-4 character. Use 3507 * simple case folding. 3508 */ 3509 int 3510 utf_toupper(int a) 3511 { 3512 /* If 'casemap' contains "keepascii" use ASCII style toupper(). */ 3513 if (a < 128 && (cmp_flags & CMP_KEEPASCII)) 3514 return TOUPPER_ASC(a); 3515 3516 #if defined(HAVE_TOWUPPER) && defined(__STDC_ISO_10646__) 3517 /* If towupper() is available and handles Unicode, use it. */ 3518 if (!(cmp_flags & CMP_INTERNAL)) 3519 return towupper(a); 3520 #endif 3521 3522 /* For characters below 128 use locale sensitive toupper(). */ 3523 if (a < 128) 3524 return TOUPPER_LOC(a); 3525 3526 /* For any other characters use the above mapping table. */ 3527 return utf_convert(a, toUpper, (int)sizeof(toUpper)); 3528 } 3529 3530 int 3531 utf_islower(int a) 3532 { 3533 /* German sharp s is lower case but has no upper case equivalent. */ 3534 return (utf_toupper(a) != a) || a == 0xdf; 3535 } 3536 3537 /* 3538 * Return the lower-case equivalent of "a", which is a UCS-4 character. Use 3539 * simple case folding. 3540 */ 3541 int 3542 utf_tolower(int a) 3543 { 3544 /* If 'casemap' contains "keepascii" use ASCII style tolower(). */ 3545 if (a < 128 && (cmp_flags & CMP_KEEPASCII)) 3546 return TOLOWER_ASC(a); 3547 3548 #if defined(HAVE_TOWLOWER) && defined(__STDC_ISO_10646__) 3549 /* If towlower() is available and handles Unicode, use it. */ 3550 if (!(cmp_flags & CMP_INTERNAL)) 3551 return towlower(a); 3552 #endif 3553 3554 /* For characters below 128 use locale sensitive tolower(). */ 3555 if (a < 128) 3556 return TOLOWER_LOC(a); 3557 3558 /* For any other characters use the above mapping table. */ 3559 return utf_convert(a, toLower, (int)sizeof(toLower)); 3560 } 3561 3562 int 3563 utf_isupper(int a) 3564 { 3565 return (utf_tolower(a) != a); 3566 } 3567 3568 static int 3569 utf_strnicmp( 3570 char_u *s1, 3571 char_u *s2, 3572 size_t n1, 3573 size_t n2) 3574 { 3575 int c1, c2, cdiff; 3576 char_u buffer[6]; 3577 3578 for (;;) 3579 { 3580 c1 = utf_safe_read_char_adv(&s1, &n1); 3581 c2 = utf_safe_read_char_adv(&s2, &n2); 3582 3583 if (c1 <= 0 || c2 <= 0) 3584 break; 3585 3586 if (c1 == c2) 3587 continue; 3588 3589 cdiff = utf_fold(c1) - utf_fold(c2); 3590 if (cdiff != 0) 3591 return cdiff; 3592 } 3593 3594 /* some string ended or has an incomplete/illegal character sequence */ 3595 3596 if (c1 == 0 || c2 == 0) 3597 { 3598 /* some string ended. shorter string is smaller */ 3599 if (c1 == 0 && c2 == 0) 3600 return 0; 3601 return c1 == 0 ? -1 : 1; 3602 } 3603 3604 /* Continue with bytewise comparison to produce some result that 3605 * would make comparison operations involving this function transitive. 3606 * 3607 * If only one string had an error, comparison should be made with 3608 * folded version of the other string. In this case it is enough 3609 * to fold just one character to determine the result of comparison. */ 3610 3611 if (c1 != -1 && c2 == -1) 3612 { 3613 n1 = utf_char2bytes(utf_fold(c1), buffer); 3614 s1 = buffer; 3615 } 3616 else if (c2 != -1 && c1 == -1) 3617 { 3618 n2 = utf_char2bytes(utf_fold(c2), buffer); 3619 s2 = buffer; 3620 } 3621 3622 while (n1 > 0 && n2 > 0 && *s1 != NUL && *s2 != NUL) 3623 { 3624 cdiff = (int)(*s1) - (int)(*s2); 3625 if (cdiff != 0) 3626 return cdiff; 3627 3628 s1++; 3629 s2++; 3630 n1--; 3631 n2--; 3632 } 3633 3634 if (n1 > 0 && *s1 == NUL) 3635 n1 = 0; 3636 if (n2 > 0 && *s2 == NUL) 3637 n2 = 0; 3638 3639 if (n1 == 0 && n2 == 0) 3640 return 0; 3641 return n1 == 0 ? -1 : 1; 3642 } 3643 3644 /* 3645 * Version of strnicmp() that handles multi-byte characters. 3646 * Needed for Big5, Shift-JIS and UTF-8 encoding. Other DBCS encodings can 3647 * probably use strnicmp(), because there are no ASCII characters in the 3648 * second byte. 3649 * Returns zero if s1 and s2 are equal (ignoring case), the difference between 3650 * two characters otherwise. 3651 */ 3652 int 3653 mb_strnicmp(char_u *s1, char_u *s2, size_t nn) 3654 { 3655 int i, l; 3656 int cdiff; 3657 int n = (int)nn; 3658 3659 if (enc_utf8) 3660 { 3661 return utf_strnicmp(s1, s2, nn, nn); 3662 } 3663 else 3664 { 3665 for (i = 0; i < n; i += l) 3666 { 3667 if (s1[i] == NUL && s2[i] == NUL) /* both strings end */ 3668 return 0; 3669 3670 l = (*mb_ptr2len)(s1 + i); 3671 if (l <= 1) 3672 { 3673 /* Single byte: first check normally, then with ignore case. */ 3674 if (s1[i] != s2[i]) 3675 { 3676 cdiff = MB_TOLOWER(s1[i]) - MB_TOLOWER(s2[i]); 3677 if (cdiff != 0) 3678 return cdiff; 3679 } 3680 } 3681 else 3682 { 3683 /* For non-Unicode multi-byte don't ignore case. */ 3684 if (l > n - i) 3685 l = n - i; 3686 cdiff = STRNCMP(s1 + i, s2 + i, l); 3687 if (cdiff != 0) 3688 return cdiff; 3689 } 3690 } 3691 } 3692 return 0; 3693 } 3694 3695 /* 3696 * "g8": show bytes of the UTF-8 char under the cursor. Doesn't matter what 3697 * 'encoding' has been set to. 3698 */ 3699 void 3700 show_utf8(void) 3701 { 3702 int len; 3703 int rlen = 0; 3704 char_u *line; 3705 int clen; 3706 int i; 3707 3708 /* Get the byte length of the char under the cursor, including composing 3709 * characters. */ 3710 line = ml_get_cursor(); 3711 len = utfc_ptr2len(line); 3712 if (len == 0) 3713 { 3714 msg("NUL"); 3715 return; 3716 } 3717 3718 clen = 0; 3719 for (i = 0; i < len; ++i) 3720 { 3721 if (clen == 0) 3722 { 3723 /* start of (composing) character, get its length */ 3724 if (i > 0) 3725 { 3726 STRCPY(IObuff + rlen, "+ "); 3727 rlen += 2; 3728 } 3729 clen = utf_ptr2len(line + i); 3730 } 3731 sprintf((char *)IObuff + rlen, "%02x ", 3732 (line[i] == NL) ? NUL : line[i]); /* NUL is stored as NL */ 3733 --clen; 3734 rlen += (int)STRLEN(IObuff + rlen); 3735 if (rlen > IOSIZE - 20) 3736 break; 3737 } 3738 3739 msg((char *)IObuff); 3740 } 3741 3742 /* 3743 * mb_head_off() function pointer. 3744 * Return offset from "p" to the first byte of the character it points into. 3745 * If "p" points to the NUL at the end of the string return 0. 3746 * Returns 0 when already at the first byte of a character. 3747 */ 3748 int 3749 latin_head_off(char_u *base UNUSED, char_u *p UNUSED) 3750 { 3751 return 0; 3752 } 3753 3754 int 3755 dbcs_head_off(char_u *base, char_u *p) 3756 { 3757 char_u *q; 3758 3759 /* It can't be a trailing byte when not using DBCS, at the start of the 3760 * string or the previous byte can't start a double-byte. */ 3761 if (p <= base || MB_BYTE2LEN(p[-1]) == 1 || *p == NUL) 3762 return 0; 3763 3764 /* This is slow: need to start at the base and go forward until the 3765 * byte we are looking for. Return 1 when we went past it, 0 otherwise. */ 3766 q = base; 3767 while (q < p) 3768 q += dbcs_ptr2len(q); 3769 return (q == p) ? 0 : 1; 3770 } 3771 3772 /* 3773 * Special version of dbcs_head_off() that works for ScreenLines[], where 3774 * single-width DBCS_JPNU characters are stored separately. 3775 */ 3776 int 3777 dbcs_screen_head_off(char_u *base, char_u *p) 3778 { 3779 char_u *q; 3780 3781 /* It can't be a trailing byte when not using DBCS, at the start of the 3782 * string or the previous byte can't start a double-byte. 3783 * For euc-jp an 0x8e byte in the previous cell always means we have a 3784 * lead byte in the current cell. */ 3785 if (p <= base 3786 || (enc_dbcs == DBCS_JPNU && p[-1] == 0x8e) 3787 || MB_BYTE2LEN(p[-1]) == 1 3788 || *p == NUL) 3789 return 0; 3790 3791 /* This is slow: need to start at the base and go forward until the 3792 * byte we are looking for. Return 1 when we went past it, 0 otherwise. 3793 * For DBCS_JPNU look out for 0x8e, which means the second byte is not 3794 * stored as the next byte. */ 3795 q = base; 3796 while (q < p) 3797 { 3798 if (enc_dbcs == DBCS_JPNU && *q == 0x8e) 3799 ++q; 3800 else 3801 q += dbcs_ptr2len(q); 3802 } 3803 return (q == p) ? 0 : 1; 3804 } 3805 3806 int 3807 utf_head_off(char_u *base, char_u *p) 3808 { 3809 char_u *q; 3810 char_u *s; 3811 int c; 3812 int len; 3813 #ifdef FEAT_ARABIC 3814 char_u *j; 3815 #endif 3816 3817 if (*p < 0x80) /* be quick for ASCII */ 3818 return 0; 3819 3820 /* Skip backwards over trailing bytes: 10xx.xxxx 3821 * Skip backwards again if on a composing char. */ 3822 for (q = p; ; --q) 3823 { 3824 /* Move s to the last byte of this char. */ 3825 for (s = q; (s[1] & 0xc0) == 0x80; ++s) 3826 ; 3827 /* Move q to the first byte of this char. */ 3828 while (q > base && (*q & 0xc0) == 0x80) 3829 --q; 3830 /* Check for illegal sequence. Do allow an illegal byte after where we 3831 * started. */ 3832 len = utf8len_tab[*q]; 3833 if (len != (int)(s - q + 1) && len != (int)(p - q + 1)) 3834 return 0; 3835 3836 if (q <= base) 3837 break; 3838 3839 c = utf_ptr2char(q); 3840 if (utf_iscomposing(c)) 3841 continue; 3842 3843 #ifdef FEAT_ARABIC 3844 if (arabic_maycombine(c)) 3845 { 3846 /* Advance to get a sneak-peak at the next char */ 3847 j = q; 3848 --j; 3849 /* Move j to the first byte of this char. */ 3850 while (j > base && (*j & 0xc0) == 0x80) 3851 --j; 3852 if (arabic_combine(utf_ptr2char(j), c)) 3853 continue; 3854 } 3855 #endif 3856 break; 3857 } 3858 3859 return (int)(p - q); 3860 } 3861 3862 /* 3863 * Copy a character from "*fp" to "*tp" and advance the pointers. 3864 */ 3865 void 3866 mb_copy_char(char_u **fp, char_u **tp) 3867 { 3868 int l = (*mb_ptr2len)(*fp); 3869 3870 mch_memmove(*tp, *fp, (size_t)l); 3871 *tp += l; 3872 *fp += l; 3873 } 3874 3875 /* 3876 * Return the offset from "p" to the first byte of a character. When "p" is 3877 * at the start of a character 0 is returned, otherwise the offset to the next 3878 * character. Can start anywhere in a stream of bytes. 3879 */ 3880 int 3881 mb_off_next(char_u *base, char_u *p) 3882 { 3883 int i; 3884 int j; 3885 3886 if (enc_utf8) 3887 { 3888 if (*p < 0x80) /* be quick for ASCII */ 3889 return 0; 3890 3891 /* Find the next character that isn't 10xx.xxxx */ 3892 for (i = 0; (p[i] & 0xc0) == 0x80; ++i) 3893 ; 3894 if (i > 0) 3895 { 3896 /* Check for illegal sequence. */ 3897 for (j = 0; p - j > base; ++j) 3898 if ((p[-j] & 0xc0) != 0x80) 3899 break; 3900 if (utf8len_tab[p[-j]] != i + j) 3901 return 0; 3902 } 3903 return i; 3904 } 3905 3906 /* Only need to check if we're on a trail byte, it doesn't matter if we 3907 * want the offset to the next or current character. */ 3908 return (*mb_head_off)(base, p); 3909 } 3910 3911 /* 3912 * Return the offset from "p" to the last byte of the character it points 3913 * into. Can start anywhere in a stream of bytes. 3914 */ 3915 int 3916 mb_tail_off(char_u *base, char_u *p) 3917 { 3918 int i; 3919 int j; 3920 3921 if (*p == NUL) 3922 return 0; 3923 3924 if (enc_utf8) 3925 { 3926 /* Find the last character that is 10xx.xxxx */ 3927 for (i = 0; (p[i + 1] & 0xc0) == 0x80; ++i) 3928 ; 3929 /* Check for illegal sequence. */ 3930 for (j = 0; p - j > base; ++j) 3931 if ((p[-j] & 0xc0) != 0x80) 3932 break; 3933 if (utf8len_tab[p[-j]] != i + j + 1) 3934 return 0; 3935 return i; 3936 } 3937 3938 /* It can't be the first byte if a double-byte when not using DBCS, at the 3939 * end of the string or the byte can't start a double-byte. */ 3940 if (enc_dbcs == 0 || p[1] == NUL || MB_BYTE2LEN(*p) == 1) 3941 return 0; 3942 3943 /* Return 1 when on the lead byte, 0 when on the tail byte. */ 3944 return 1 - dbcs_head_off(base, p); 3945 } 3946 3947 /* 3948 * Find the next illegal byte sequence. 3949 */ 3950 void 3951 utf_find_illegal(void) 3952 { 3953 pos_T pos = curwin->w_cursor; 3954 char_u *p; 3955 int len; 3956 vimconv_T vimconv; 3957 char_u *tofree = NULL; 3958 3959 vimconv.vc_type = CONV_NONE; 3960 if (enc_utf8 && (enc_canon_props(curbuf->b_p_fenc) & ENC_8BIT)) 3961 { 3962 /* 'encoding' is "utf-8" but we are editing a 8-bit encoded file, 3963 * possibly a utf-8 file with illegal bytes. Setup for conversion 3964 * from utf-8 to 'fileencoding'. */ 3965 convert_setup(&vimconv, p_enc, curbuf->b_p_fenc); 3966 } 3967 3968 curwin->w_cursor.coladd = 0; 3969 for (;;) 3970 { 3971 p = ml_get_cursor(); 3972 if (vimconv.vc_type != CONV_NONE) 3973 { 3974 vim_free(tofree); 3975 tofree = string_convert(&vimconv, p, NULL); 3976 if (tofree == NULL) 3977 break; 3978 p = tofree; 3979 } 3980 3981 while (*p != NUL) 3982 { 3983 /* Illegal means that there are not enough trail bytes (checked by 3984 * utf_ptr2len()) or too many of them (overlong sequence). */ 3985 len = utf_ptr2len(p); 3986 if (*p >= 0x80 && (len == 1 3987 || utf_char2len(utf_ptr2char(p)) != len)) 3988 { 3989 if (vimconv.vc_type == CONV_NONE) 3990 curwin->w_cursor.col += (colnr_T)(p - ml_get_cursor()); 3991 else 3992 { 3993 int l; 3994 3995 len = (int)(p - tofree); 3996 for (p = ml_get_cursor(); *p != NUL && len-- > 0; p += l) 3997 { 3998 l = utf_ptr2len(p); 3999 curwin->w_cursor.col += l; 4000 } 4001 } 4002 goto theend; 4003 } 4004 p += len; 4005 } 4006 if (curwin->w_cursor.lnum == curbuf->b_ml.ml_line_count) 4007 break; 4008 ++curwin->w_cursor.lnum; 4009 curwin->w_cursor.col = 0; 4010 } 4011 4012 /* didn't find it: don't move and beep */ 4013 curwin->w_cursor = pos; 4014 beep_flush(); 4015 4016 theend: 4017 vim_free(tofree); 4018 convert_setup(&vimconv, NULL, NULL); 4019 } 4020 4021 #if defined(FEAT_GUI_GTK) || defined(PROTO) 4022 /* 4023 * Return TRUE if string "s" is a valid utf-8 string. 4024 * When "end" is NULL stop at the first NUL. 4025 * When "end" is positive stop there. 4026 */ 4027 int 4028 utf_valid_string(char_u *s, char_u *end) 4029 { 4030 int l; 4031 char_u *p = s; 4032 4033 while (end == NULL ? *p != NUL : p < end) 4034 { 4035 l = utf8len_tab_zero[*p]; 4036 if (l == 0) 4037 return FALSE; /* invalid lead byte */ 4038 if (end != NULL && p + l > end) 4039 return FALSE; /* incomplete byte sequence */ 4040 ++p; 4041 while (--l > 0) 4042 if ((*p++ & 0xc0) != 0x80) 4043 return FALSE; /* invalid trail byte */ 4044 } 4045 return TRUE; 4046 } 4047 #endif 4048 4049 #if defined(FEAT_GUI) || defined(PROTO) 4050 /* 4051 * Special version of mb_tail_off() for use in ScreenLines[]. 4052 */ 4053 int 4054 dbcs_screen_tail_off(char_u *base, char_u *p) 4055 { 4056 /* It can't be the first byte if a double-byte when not using DBCS, at the 4057 * end of the string or the byte can't start a double-byte. 4058 * For euc-jp an 0x8e byte always means we have a lead byte in the current 4059 * cell. */ 4060 if (*p == NUL || p[1] == NUL 4061 || (enc_dbcs == DBCS_JPNU && *p == 0x8e) 4062 || MB_BYTE2LEN(*p) == 1) 4063 return 0; 4064 4065 /* Return 1 when on the lead byte, 0 when on the tail byte. */ 4066 return 1 - dbcs_screen_head_off(base, p); 4067 } 4068 #endif 4069 4070 /* 4071 * If the cursor moves on an trail byte, set the cursor on the lead byte. 4072 * Thus it moves left if necessary. 4073 * Return TRUE when the cursor was adjusted. 4074 */ 4075 void 4076 mb_adjust_cursor(void) 4077 { 4078 mb_adjustpos(curbuf, &curwin->w_cursor); 4079 } 4080 4081 /* 4082 * Adjust position "*lp" to point to the first byte of a multi-byte character. 4083 * If it points to a tail byte it's moved backwards to the head byte. 4084 */ 4085 void 4086 mb_adjustpos(buf_T *buf, pos_T *lp) 4087 { 4088 char_u *p; 4089 4090 if (lp->col > 0 || lp->coladd > 1) 4091 { 4092 p = ml_get_buf(buf, lp->lnum, FALSE); 4093 if (*p == NUL || (int)STRLEN(p) < lp->col) 4094 lp->col = 0; 4095 else 4096 lp->col -= (*mb_head_off)(p, p + lp->col); 4097 /* Reset "coladd" when the cursor would be on the right half of a 4098 * double-wide character. */ 4099 if (lp->coladd == 1 4100 && p[lp->col] != TAB 4101 && vim_isprintc((*mb_ptr2char)(p + lp->col)) 4102 && ptr2cells(p + lp->col) > 1) 4103 lp->coladd = 0; 4104 } 4105 } 4106 4107 /* 4108 * Return a pointer to the character before "*p", if there is one. 4109 */ 4110 char_u * 4111 mb_prevptr( 4112 char_u *line, /* start of the string */ 4113 char_u *p) 4114 { 4115 if (p > line) 4116 MB_PTR_BACK(line, p); 4117 return p; 4118 } 4119 4120 /* 4121 * Return the character length of "str". Each multi-byte character (with 4122 * following composing characters) counts as one. 4123 */ 4124 int 4125 mb_charlen(char_u *str) 4126 { 4127 char_u *p = str; 4128 int count; 4129 4130 if (p == NULL) 4131 return 0; 4132 4133 for (count = 0; *p != NUL; count++) 4134 p += (*mb_ptr2len)(p); 4135 4136 return count; 4137 } 4138 4139 #if defined(FEAT_SPELL) || defined(PROTO) 4140 /* 4141 * Like mb_charlen() but for a string with specified length. 4142 */ 4143 int 4144 mb_charlen_len(char_u *str, int len) 4145 { 4146 char_u *p = str; 4147 int count; 4148 4149 for (count = 0; *p != NUL && p < str + len; count++) 4150 p += (*mb_ptr2len)(p); 4151 4152 return count; 4153 } 4154 #endif 4155 4156 /* 4157 * Try to un-escape a multi-byte character. 4158 * Used for the "to" and "from" part of a mapping. 4159 * Return the un-escaped string if it is a multi-byte character, and advance 4160 * "pp" to just after the bytes that formed it. 4161 * Return NULL if no multi-byte char was found. 4162 */ 4163 char_u * 4164 mb_unescape(char_u **pp) 4165 { 4166 static char_u buf[6]; 4167 int n; 4168 int m = 0; 4169 char_u *str = *pp; 4170 4171 /* Must translate K_SPECIAL KS_SPECIAL KE_FILLER to K_SPECIAL and CSI 4172 * KS_EXTRA KE_CSI to CSI. 4173 * Maximum length of a utf-8 character is 4 bytes. */ 4174 for (n = 0; str[n] != NUL && m < 4; ++n) 4175 { 4176 if (str[n] == K_SPECIAL 4177 && str[n + 1] == KS_SPECIAL 4178 && str[n + 2] == KE_FILLER) 4179 { 4180 buf[m++] = K_SPECIAL; 4181 n += 2; 4182 } 4183 else if ((str[n] == K_SPECIAL 4184 # ifdef FEAT_GUI 4185 || str[n] == CSI 4186 # endif 4187 ) 4188 && str[n + 1] == KS_EXTRA 4189 && str[n + 2] == (int)KE_CSI) 4190 { 4191 buf[m++] = CSI; 4192 n += 2; 4193 } 4194 else if (str[n] == K_SPECIAL 4195 # ifdef FEAT_GUI 4196 || str[n] == CSI 4197 # endif 4198 ) 4199 break; /* a special key can't be a multibyte char */ 4200 else 4201 buf[m++] = str[n]; 4202 buf[m] = NUL; 4203 4204 /* Return a multi-byte character if it's found. An illegal sequence 4205 * will result in a 1 here. */ 4206 if ((*mb_ptr2len)(buf) > 1) 4207 { 4208 *pp = str + n + 1; 4209 return buf; 4210 } 4211 4212 /* Bail out quickly for ASCII. */ 4213 if (buf[0] < 128) 4214 break; 4215 } 4216 return NULL; 4217 } 4218 4219 /* 4220 * Return TRUE if the character at "row"/"col" on the screen is the left side 4221 * of a double-width character. 4222 * Caller must make sure "row" and "col" are not invalid! 4223 */ 4224 int 4225 mb_lefthalve(int row, int col) 4226 { 4227 #ifdef FEAT_HANGULIN 4228 if (composing_hangul) 4229 return TRUE; 4230 #endif 4231 return (*mb_off2cells)(LineOffset[row] + col, 4232 LineOffset[row] + screen_Columns) > 1; 4233 } 4234 4235 /* 4236 * Correct a position on the screen, if it's the right half of a double-wide 4237 * char move it to the left half. Returns the corrected column. 4238 */ 4239 int 4240 mb_fix_col(int col, int row) 4241 { 4242 col = check_col(col); 4243 row = check_row(row); 4244 if (has_mbyte && ScreenLines != NULL && col > 0 4245 && ((enc_dbcs 4246 && ScreenLines[LineOffset[row] + col] != NUL 4247 && dbcs_screen_head_off(ScreenLines + LineOffset[row], 4248 ScreenLines + LineOffset[row] + col)) 4249 || (enc_utf8 && ScreenLines[LineOffset[row] + col] == 0))) 4250 return col - 1; 4251 return col; 4252 } 4253 4254 static int enc_alias_search(char_u *name); 4255 4256 /* 4257 * Skip the Vim specific head of a 'encoding' name. 4258 */ 4259 char_u * 4260 enc_skip(char_u *p) 4261 { 4262 if (STRNCMP(p, "2byte-", 6) == 0) 4263 return p + 6; 4264 if (STRNCMP(p, "8bit-", 5) == 0) 4265 return p + 5; 4266 return p; 4267 } 4268 4269 /* 4270 * Find the canonical name for encoding "enc". 4271 * When the name isn't recognized, returns "enc" itself, but with all lower 4272 * case characters and '_' replaced with '-'. 4273 * Returns an allocated string. NULL for out-of-memory. 4274 */ 4275 char_u * 4276 enc_canonize(char_u *enc) 4277 { 4278 char_u *r; 4279 char_u *p, *s; 4280 int i; 4281 4282 if (STRCMP(enc, "default") == 0) 4283 { 4284 /* Use the default encoding as it's found by set_init_1(). */ 4285 r = get_encoding_default(); 4286 if (r == NULL) 4287 r = (char_u *)"latin1"; 4288 return vim_strsave(r); 4289 } 4290 4291 /* copy "enc" to allocated memory, with room for two '-' */ 4292 r = alloc((unsigned)(STRLEN(enc) + 3)); 4293 if (r != NULL) 4294 { 4295 /* Make it all lower case and replace '_' with '-'. */ 4296 p = r; 4297 for (s = enc; *s != NUL; ++s) 4298 { 4299 if (*s == '_') 4300 *p++ = '-'; 4301 else 4302 *p++ = TOLOWER_ASC(*s); 4303 } 4304 *p = NUL; 4305 4306 /* Skip "2byte-" and "8bit-". */ 4307 p = enc_skip(r); 4308 4309 /* Change "microsoft-cp" to "cp". Used in some spell files. */ 4310 if (STRNCMP(p, "microsoft-cp", 12) == 0) 4311 STRMOVE(p, p + 10); 4312 4313 /* "iso8859" -> "iso-8859" */ 4314 if (STRNCMP(p, "iso8859", 7) == 0) 4315 { 4316 STRMOVE(p + 4, p + 3); 4317 p[3] = '-'; 4318 } 4319 4320 /* "iso-8859n" -> "iso-8859-n" */ 4321 if (STRNCMP(p, "iso-8859", 8) == 0 && p[8] != '-') 4322 { 4323 STRMOVE(p + 9, p + 8); 4324 p[8] = '-'; 4325 } 4326 4327 /* "latin-N" -> "latinN" */ 4328 if (STRNCMP(p, "latin-", 6) == 0) 4329 STRMOVE(p + 5, p + 6); 4330 4331 if (enc_canon_search(p) >= 0) 4332 { 4333 /* canonical name can be used unmodified */ 4334 if (p != r) 4335 STRMOVE(r, p); 4336 } 4337 else if ((i = enc_alias_search(p)) >= 0) 4338 { 4339 /* alias recognized, get canonical name */ 4340 vim_free(r); 4341 r = vim_strsave((char_u *)enc_canon_table[i].name); 4342 } 4343 } 4344 return r; 4345 } 4346 4347 /* 4348 * Search for an encoding alias of "name". 4349 * Returns -1 when not found. 4350 */ 4351 static int 4352 enc_alias_search(char_u *name) 4353 { 4354 int i; 4355 4356 for (i = 0; enc_alias_table[i].name != NULL; ++i) 4357 if (STRCMP(name, enc_alias_table[i].name) == 0) 4358 return enc_alias_table[i].canon; 4359 return -1; 4360 } 4361 4362 4363 #ifdef HAVE_LANGINFO_H 4364 # include <langinfo.h> 4365 #endif 4366 4367 #ifndef FEAT_GUI_MSWIN 4368 /* 4369 * Get the canonicalized encoding from the specified locale string "locale" 4370 * or from the environment variables LC_ALL, LC_CTYPE and LANG. 4371 * Returns an allocated string when successful, NULL when not. 4372 */ 4373 char_u * 4374 enc_locale_env(char *locale) 4375 { 4376 char *s = locale; 4377 char *p; 4378 int i; 4379 char buf[50]; 4380 4381 if (s == NULL || *s == NUL) 4382 if ((s = getenv("LC_ALL")) == NULL || *s == NUL) 4383 if ((s = getenv("LC_CTYPE")) == NULL || *s == NUL) 4384 s = getenv("LANG"); 4385 4386 if (s == NULL || *s == NUL) 4387 return NULL; 4388 4389 /* The most generic locale format is: 4390 * language[_territory][.codeset][@modifier][+special][,[sponsor][_revision]] 4391 * If there is a '.' remove the part before it. 4392 * if there is something after the codeset, remove it. 4393 * Make the name lowercase and replace '_' with '-'. 4394 * Exception: "ja_JP.EUC" == "euc-jp", "zh_CN.EUC" = "euc-cn", 4395 * "ko_KR.EUC" == "euc-kr" 4396 */ 4397 if ((p = (char *)vim_strchr((char_u *)s, '.')) != NULL) 4398 { 4399 if (p > s + 2 && STRNICMP(p + 1, "EUC", 3) == 0 4400 && !isalnum((int)p[4]) && p[4] != '-' && p[-3] == '_') 4401 { 4402 /* copy "XY.EUC" to "euc-XY" to buf[10] */ 4403 STRCPY(buf + 10, "euc-"); 4404 buf[14] = p[-2]; 4405 buf[15] = p[-1]; 4406 buf[16] = 0; 4407 s = buf + 10; 4408 } 4409 else 4410 s = p + 1; 4411 } 4412 for (i = 0; i < (int)sizeof(buf) - 1 && s[i] != NUL; ++i) 4413 { 4414 if (s[i] == '_' || s[i] == '-') 4415 buf[i] = '-'; 4416 else if (isalnum((int)s[i])) 4417 buf[i] = TOLOWER_ASC(s[i]); 4418 else 4419 break; 4420 } 4421 buf[i] = NUL; 4422 4423 return enc_canonize((char_u *)buf); 4424 } 4425 #endif 4426 4427 /* 4428 * Get the canonicalized encoding of the current locale. 4429 * Returns an allocated string when successful, NULL when not. 4430 */ 4431 char_u * 4432 enc_locale(void) 4433 { 4434 #ifdef MSWIN 4435 char buf[50]; 4436 long acp = GetACP(); 4437 4438 if (acp == 1200) 4439 STRCPY(buf, "ucs-2le"); 4440 else if (acp == 1252) /* cp1252 is used as latin1 */ 4441 STRCPY(buf, "latin1"); 4442 else 4443 sprintf(buf, "cp%ld", acp); 4444 4445 return enc_canonize((char_u *)buf); 4446 #else 4447 char *s; 4448 4449 # ifdef HAVE_NL_LANGINFO_CODESET 4450 if ((s = nl_langinfo(CODESET)) == NULL || *s == NUL) 4451 # endif 4452 # if defined(HAVE_LOCALE_H) || defined(X_LOCALE) 4453 if ((s = setlocale(LC_CTYPE, NULL)) == NULL || *s == NUL) 4454 # endif 4455 s = NULL; 4456 4457 return enc_locale_env(s); 4458 #endif 4459 } 4460 4461 # if defined(MSWIN) || defined(PROTO) || defined(FEAT_CYGWIN_WIN32_CLIPBOARD) 4462 /* 4463 * Convert an encoding name to an MS-Windows codepage. 4464 * Returns zero if no codepage can be figured out. 4465 */ 4466 int 4467 encname2codepage(char_u *name) 4468 { 4469 int cp; 4470 char_u *p = name; 4471 int idx; 4472 4473 if (STRNCMP(p, "8bit-", 5) == 0) 4474 p += 5; 4475 else if (STRNCMP(p_enc, "2byte-", 6) == 0) 4476 p += 6; 4477 4478 if (p[0] == 'c' && p[1] == 'p') 4479 cp = atoi((char *)p + 2); 4480 else if ((idx = enc_canon_search(p)) >= 0) 4481 cp = enc_canon_table[idx].codepage; 4482 else 4483 return 0; 4484 if (IsValidCodePage(cp)) 4485 return cp; 4486 return 0; 4487 } 4488 # endif 4489 4490 # if defined(USE_ICONV) || defined(PROTO) 4491 4492 /* 4493 * Call iconv_open() with a check if iconv() works properly (there are broken 4494 * versions). 4495 * Returns (void *)-1 if failed. 4496 * (should return iconv_t, but that causes problems with prototypes). 4497 */ 4498 void * 4499 my_iconv_open(char_u *to, char_u *from) 4500 { 4501 iconv_t fd; 4502 #define ICONV_TESTLEN 400 4503 char_u tobuf[ICONV_TESTLEN]; 4504 char *p; 4505 size_t tolen; 4506 static int iconv_ok = -1; 4507 4508 if (iconv_ok == FALSE) 4509 return (void *)-1; /* detected a broken iconv() previously */ 4510 4511 #ifdef DYNAMIC_ICONV 4512 /* Check if the iconv.dll can be found. */ 4513 if (!iconv_enabled(TRUE)) 4514 return (void *)-1; 4515 #endif 4516 4517 fd = iconv_open((char *)enc_skip(to), (char *)enc_skip(from)); 4518 4519 if (fd != (iconv_t)-1 && iconv_ok == -1) 4520 { 4521 /* 4522 * Do a dummy iconv() call to check if it actually works. There is a 4523 * version of iconv() on Linux that is broken. We can't ignore it, 4524 * because it's wide-spread. The symptoms are that after outputting 4525 * the initial shift state the "to" pointer is NULL and conversion 4526 * stops for no apparent reason after about 8160 characters. 4527 */ 4528 p = (char *)tobuf; 4529 tolen = ICONV_TESTLEN; 4530 (void)iconv(fd, NULL, NULL, &p, &tolen); 4531 if (p == NULL) 4532 { 4533 iconv_ok = FALSE; 4534 iconv_close(fd); 4535 fd = (iconv_t)-1; 4536 } 4537 else 4538 iconv_ok = TRUE; 4539 } 4540 4541 return (void *)fd; 4542 } 4543 4544 /* 4545 * Convert the string "str[slen]" with iconv(). 4546 * If "unconvlenp" is not NULL handle the string ending in an incomplete 4547 * sequence and set "*unconvlenp" to the length of it. 4548 * Returns the converted string in allocated memory. NULL for an error. 4549 * If resultlenp is not NULL, sets it to the result length in bytes. 4550 */ 4551 static char_u * 4552 iconv_string( 4553 vimconv_T *vcp, 4554 char_u *str, 4555 int slen, 4556 int *unconvlenp, 4557 int *resultlenp) 4558 { 4559 const char *from; 4560 size_t fromlen; 4561 char *to; 4562 size_t tolen; 4563 size_t len = 0; 4564 size_t done = 0; 4565 char_u *result = NULL; 4566 char_u *p; 4567 int l; 4568 4569 from = (char *)str; 4570 fromlen = slen; 4571 for (;;) 4572 { 4573 if (len == 0 || ICONV_ERRNO == ICONV_E2BIG) 4574 { 4575 /* Allocate enough room for most conversions. When re-allocating 4576 * increase the buffer size. */ 4577 len = len + fromlen * 2 + 40; 4578 p = alloc((unsigned)len); 4579 if (p != NULL && done > 0) 4580 mch_memmove(p, result, done); 4581 vim_free(result); 4582 result = p; 4583 if (result == NULL) /* out of memory */ 4584 break; 4585 } 4586 4587 to = (char *)result + done; 4588 tolen = len - done - 2; 4589 /* Avoid a warning for systems with a wrong iconv() prototype by 4590 * casting the second argument to void *. */ 4591 if (iconv(vcp->vc_fd, (void *)&from, &fromlen, &to, &tolen) 4592 != (size_t)-1) 4593 { 4594 /* Finished, append a NUL. */ 4595 *to = NUL; 4596 break; 4597 } 4598 4599 /* Check both ICONV_EINVAL and EINVAL, because the dynamically loaded 4600 * iconv library may use one of them. */ 4601 if (!vcp->vc_fail && unconvlenp != NULL 4602 && (ICONV_ERRNO == ICONV_EINVAL || ICONV_ERRNO == EINVAL)) 4603 { 4604 /* Handle an incomplete sequence at the end. */ 4605 *to = NUL; 4606 *unconvlenp = (int)fromlen; 4607 break; 4608 } 4609 4610 /* Check both ICONV_EILSEQ and EILSEQ, because the dynamically loaded 4611 * iconv library may use one of them. */ 4612 else if (!vcp->vc_fail 4613 && (ICONV_ERRNO == ICONV_EILSEQ || ICONV_ERRNO == EILSEQ 4614 || ICONV_ERRNO == ICONV_EINVAL || ICONV_ERRNO == EINVAL)) 4615 { 4616 /* Can't convert: insert a '?' and skip a character. This assumes 4617 * conversion from 'encoding' to something else. In other 4618 * situations we don't know what to skip anyway. */ 4619 *to++ = '?'; 4620 if ((*mb_ptr2cells)((char_u *)from) > 1) 4621 *to++ = '?'; 4622 if (enc_utf8) 4623 l = utfc_ptr2len_len((char_u *)from, (int)fromlen); 4624 else 4625 { 4626 l = (*mb_ptr2len)((char_u *)from); 4627 if (l > (int)fromlen) 4628 l = (int)fromlen; 4629 } 4630 from += l; 4631 fromlen -= l; 4632 } 4633 else if (ICONV_ERRNO != ICONV_E2BIG) 4634 { 4635 /* conversion failed */ 4636 VIM_CLEAR(result); 4637 break; 4638 } 4639 /* Not enough room or skipping illegal sequence. */ 4640 done = to - (char *)result; 4641 } 4642 4643 if (resultlenp != NULL && result != NULL) 4644 *resultlenp = (int)(to - (char *)result); 4645 return result; 4646 } 4647 4648 # if defined(DYNAMIC_ICONV) || defined(PROTO) 4649 /* 4650 * Dynamically load the "iconv.dll" on Win32. 4651 */ 4652 4653 # ifndef DYNAMIC_ICONV /* must be generating prototypes */ 4654 # define HINSTANCE int 4655 # endif 4656 static HINSTANCE hIconvDLL = 0; 4657 static HINSTANCE hMsvcrtDLL = 0; 4658 4659 # ifndef DYNAMIC_ICONV_DLL 4660 # define DYNAMIC_ICONV_DLL "iconv.dll" 4661 # define DYNAMIC_ICONV_DLL_ALT1 "libiconv.dll" 4662 # define DYNAMIC_ICONV_DLL_ALT2 "libiconv2.dll" 4663 # define DYNAMIC_ICONV_DLL_ALT3 "libiconv-2.dll" 4664 # endif 4665 # ifndef DYNAMIC_MSVCRT_DLL 4666 # define DYNAMIC_MSVCRT_DLL "msvcrt.dll" 4667 # endif 4668 4669 /* 4670 * Try opening the iconv.dll and return TRUE if iconv() can be used. 4671 */ 4672 int 4673 iconv_enabled(int verbose) 4674 { 4675 if (hIconvDLL != 0 && hMsvcrtDLL != 0) 4676 return TRUE; 4677 4678 /* The iconv DLL file goes under different names, try them all. 4679 * Do the "2" version first, it's newer. */ 4680 #ifdef DYNAMIC_ICONV_DLL_ALT2 4681 if (hIconvDLL == 0) 4682 hIconvDLL = vimLoadLib(DYNAMIC_ICONV_DLL_ALT2); 4683 #endif 4684 #ifdef DYNAMIC_ICONV_DLL_ALT3 4685 if (hIconvDLL == 0) 4686 hIconvDLL = vimLoadLib(DYNAMIC_ICONV_DLL_ALT3); 4687 #endif 4688 if (hIconvDLL == 0) 4689 hIconvDLL = vimLoadLib(DYNAMIC_ICONV_DLL); 4690 #ifdef DYNAMIC_ICONV_DLL_ALT1 4691 if (hIconvDLL == 0) 4692 hIconvDLL = vimLoadLib(DYNAMIC_ICONV_DLL_ALT1); 4693 #endif 4694 4695 if (hIconvDLL != 0) 4696 hMsvcrtDLL = vimLoadLib(DYNAMIC_MSVCRT_DLL); 4697 if (hIconvDLL == 0 || hMsvcrtDLL == 0) 4698 { 4699 /* Only give the message when 'verbose' is set, otherwise it might be 4700 * done whenever a conversion is attempted. */ 4701 if (verbose && p_verbose > 0) 4702 { 4703 verbose_enter(); 4704 semsg(_(e_loadlib), 4705 hIconvDLL == 0 ? DYNAMIC_ICONV_DLL : DYNAMIC_MSVCRT_DLL); 4706 verbose_leave(); 4707 } 4708 iconv_end(); 4709 return FALSE; 4710 } 4711 4712 iconv = (void *)GetProcAddress(hIconvDLL, "libiconv"); 4713 iconv_open = (void *)GetProcAddress(hIconvDLL, "libiconv_open"); 4714 iconv_close = (void *)GetProcAddress(hIconvDLL, "libiconv_close"); 4715 iconvctl = (void *)GetProcAddress(hIconvDLL, "libiconvctl"); 4716 iconv_errno = get_dll_import_func(hIconvDLL, "_errno"); 4717 if (iconv_errno == NULL) 4718 iconv_errno = (void *)GetProcAddress(hMsvcrtDLL, "_errno"); 4719 if (iconv == NULL || iconv_open == NULL || iconv_close == NULL 4720 || iconvctl == NULL || iconv_errno == NULL) 4721 { 4722 iconv_end(); 4723 if (verbose && p_verbose > 0) 4724 { 4725 verbose_enter(); 4726 semsg(_(e_loadfunc), "for libiconv"); 4727 verbose_leave(); 4728 } 4729 return FALSE; 4730 } 4731 return TRUE; 4732 } 4733 4734 void 4735 iconv_end(void) 4736 { 4737 /* Don't use iconv() when inputting or outputting characters. */ 4738 if (input_conv.vc_type == CONV_ICONV) 4739 convert_setup(&input_conv, NULL, NULL); 4740 if (output_conv.vc_type == CONV_ICONV) 4741 convert_setup(&output_conv, NULL, NULL); 4742 4743 if (hIconvDLL != 0) 4744 FreeLibrary(hIconvDLL); 4745 if (hMsvcrtDLL != 0) 4746 FreeLibrary(hMsvcrtDLL); 4747 hIconvDLL = 0; 4748 hMsvcrtDLL = 0; 4749 } 4750 # endif /* DYNAMIC_ICONV */ 4751 # endif /* USE_ICONV */ 4752 4753 4754 #ifdef FEAT_GUI 4755 # define USE_IMACTIVATEFUNC (!gui.in_use && *p_imaf != NUL) 4756 # define USE_IMSTATUSFUNC (!gui.in_use && *p_imsf != NUL) 4757 #else 4758 # define USE_IMACTIVATEFUNC (*p_imaf != NUL) 4759 # define USE_IMSTATUSFUNC (*p_imsf != NUL) 4760 #endif 4761 4762 #if defined(FEAT_EVAL) && (defined(FEAT_XIM) || defined(IME_WITHOUT_XIM)) 4763 static void 4764 call_imactivatefunc(int active) 4765 { 4766 typval_T argv[2]; 4767 4768 argv[0].v_type = VAR_NUMBER; 4769 argv[0].vval.v_number = active ? 1 : 0; 4770 argv[1].v_type = VAR_UNKNOWN; 4771 (void)call_func_retnr(p_imaf, 1, argv); 4772 } 4773 4774 static int 4775 call_imstatusfunc(void) 4776 { 4777 int is_active; 4778 4779 /* FIXME: Don't execute user function in unsafe situation. */ 4780 if (exiting || is_autocmd_blocked()) 4781 return FALSE; 4782 /* FIXME: :py print 'xxx' is shown duplicate result. 4783 * Use silent to avoid it. */ 4784 ++msg_silent; 4785 is_active = call_func_retnr(p_imsf, 0, NULL); 4786 --msg_silent; 4787 return (is_active > 0); 4788 } 4789 #endif 4790 4791 #if defined(FEAT_XIM) || defined(PROTO) 4792 4793 # if defined(FEAT_GUI_GTK) || defined(PROTO) 4794 static int xim_has_preediting INIT(= FALSE); /* IM current status */ 4795 4796 /* 4797 * Set preedit_start_col to the current cursor position. 4798 */ 4799 static void 4800 init_preedit_start_col(void) 4801 { 4802 if (State & CMDLINE) 4803 preedit_start_col = cmdline_getvcol_cursor(); 4804 else if (curwin != NULL && curwin->w_buffer != NULL) 4805 getvcol(curwin, &curwin->w_cursor, &preedit_start_col, NULL, NULL); 4806 /* Prevent that preediting marks the buffer as changed. */ 4807 xim_changed_while_preediting = curbuf->b_changed; 4808 } 4809 4810 static int im_is_active = FALSE; /* IM is enabled for current mode */ 4811 static int preedit_is_active = FALSE; 4812 static int im_preedit_cursor = 0; /* cursor offset in characters */ 4813 static int im_preedit_trailing = 0; /* number of characters after cursor */ 4814 4815 static unsigned long im_commit_handler_id = 0; 4816 static unsigned int im_activatekey_keyval = GDK_VoidSymbol; 4817 static unsigned int im_activatekey_state = 0; 4818 4819 static GtkWidget *preedit_window = NULL; 4820 static GtkWidget *preedit_label = NULL; 4821 4822 static void im_preedit_window_set_position(void); 4823 4824 void 4825 im_set_active(int active) 4826 { 4827 int was_active; 4828 4829 was_active = !!im_get_status(); 4830 im_is_active = (active && !p_imdisable); 4831 4832 if (im_is_active != was_active) 4833 xim_reset(); 4834 } 4835 4836 void 4837 xim_set_focus(int focus) 4838 { 4839 if (xic != NULL) 4840 { 4841 if (focus) 4842 gtk_im_context_focus_in(xic); 4843 else 4844 gtk_im_context_focus_out(xic); 4845 } 4846 } 4847 4848 void 4849 im_set_position(int row, int col) 4850 { 4851 if (xic != NULL) 4852 { 4853 GdkRectangle area; 4854 4855 area.x = FILL_X(col); 4856 area.y = FILL_Y(row); 4857 area.width = gui.char_width * (mb_lefthalve(row, col) ? 2 : 1); 4858 area.height = gui.char_height; 4859 4860 gtk_im_context_set_cursor_location(xic, &area); 4861 4862 if (p_imst == IM_OVER_THE_SPOT) 4863 im_preedit_window_set_position(); 4864 } 4865 } 4866 4867 # if 0 || defined(PROTO) /* apparently only used in gui_x11.c */ 4868 void 4869 xim_set_preedit(void) 4870 { 4871 im_set_position(gui.row, gui.col); 4872 } 4873 # endif 4874 4875 static void 4876 im_add_to_input(char_u *str, int len) 4877 { 4878 /* Convert from 'termencoding' (always "utf-8") to 'encoding' */ 4879 if (input_conv.vc_type != CONV_NONE) 4880 { 4881 str = string_convert(&input_conv, str, &len); 4882 g_return_if_fail(str != NULL); 4883 } 4884 4885 add_to_input_buf_csi(str, len); 4886 4887 if (input_conv.vc_type != CONV_NONE) 4888 vim_free(str); 4889 4890 if (p_mh) /* blank out the pointer if necessary */ 4891 gui_mch_mousehide(TRUE); 4892 } 4893 4894 static void 4895 im_preedit_window_set_position(void) 4896 { 4897 int x, y, width, height; 4898 int screen_x, screen_y, screen_width, screen_height; 4899 4900 if (preedit_window == NULL) 4901 return; 4902 4903 gui_gtk_get_screen_geom_of_win(gui.drawarea, 4904 &screen_x, &screen_y, &screen_width, &screen_height); 4905 gdk_window_get_origin(gtk_widget_get_window(gui.drawarea), &x, &y); 4906 gtk_window_get_size(GTK_WINDOW(preedit_window), &width, &height); 4907 x = x + FILL_X(gui.col); 4908 y = y + FILL_Y(gui.row); 4909 if (x + width > screen_x + screen_width) 4910 x = screen_x + screen_width - width; 4911 if (y + height > screen_y + screen_height) 4912 y = screen_y + screen_height - height; 4913 gtk_window_move(GTK_WINDOW(preedit_window), x, y); 4914 } 4915 4916 static void 4917 im_preedit_window_open() 4918 { 4919 char *preedit_string; 4920 #if !GTK_CHECK_VERSION(3,16,0) 4921 char buf[8]; 4922 #endif 4923 PangoAttrList *attr_list; 4924 PangoLayout *layout; 4925 #if GTK_CHECK_VERSION(3,0,0) 4926 # if !GTK_CHECK_VERSION(3,16,0) 4927 GdkRGBA color; 4928 # endif 4929 #else 4930 GdkColor color; 4931 #endif 4932 gint w, h; 4933 4934 if (preedit_window == NULL) 4935 { 4936 preedit_window = gtk_window_new(GTK_WINDOW_POPUP); 4937 gtk_window_set_transient_for(GTK_WINDOW(preedit_window), 4938 GTK_WINDOW(gui.mainwin)); 4939 preedit_label = gtk_label_new(""); 4940 gtk_widget_set_name(preedit_label, "vim-gui-preedit-area"); 4941 gtk_container_add(GTK_CONTAINER(preedit_window), preedit_label); 4942 } 4943 4944 #if GTK_CHECK_VERSION(3,16,0) 4945 { 4946 GtkStyleContext * const context 4947 = gtk_widget_get_style_context(gui.drawarea); 4948 GtkCssProvider * const provider = gtk_css_provider_new(); 4949 gchar *css = NULL; 4950 const char * const fontname 4951 = pango_font_description_get_family(gui.norm_font); 4952 gint fontsize 4953 = pango_font_description_get_size(gui.norm_font) / PANGO_SCALE; 4954 gchar *fontsize_propval = NULL; 4955 4956 if (!pango_font_description_get_size_is_absolute(gui.norm_font)) 4957 { 4958 /* fontsize was given in points. Convert it into that in pixels 4959 * to use with CSS. */ 4960 GdkScreen * const screen 4961 = gdk_window_get_screen(gtk_widget_get_window(gui.mainwin)); 4962 const gdouble dpi = gdk_screen_get_resolution(screen); 4963 fontsize = dpi * fontsize / 72; 4964 } 4965 if (fontsize > 0) 4966 fontsize_propval = g_strdup_printf("%dpx", fontsize); 4967 else 4968 fontsize_propval = g_strdup_printf("inherit"); 4969 4970 css = g_strdup_printf( 4971 "widget#vim-gui-preedit-area {\n" 4972 " font-family: %s,monospace;\n" 4973 " font-size: %s;\n" 4974 " color: #%.2lx%.2lx%.2lx;\n" 4975 " background-color: #%.2lx%.2lx%.2lx;\n" 4976 "}\n", 4977 fontname != NULL ? fontname : "inherit", 4978 fontsize_propval, 4979 (gui.norm_pixel >> 16) & 0xff, 4980 (gui.norm_pixel >> 8) & 0xff, 4981 gui.norm_pixel & 0xff, 4982 (gui.back_pixel >> 16) & 0xff, 4983 (gui.back_pixel >> 8) & 0xff, 4984 gui.back_pixel & 0xff); 4985 4986 gtk_css_provider_load_from_data(provider, css, -1, NULL); 4987 gtk_style_context_add_provider(context, 4988 GTK_STYLE_PROVIDER(provider), G_MAXUINT); 4989 4990 g_free(css); 4991 g_free(fontsize_propval); 4992 g_object_unref(provider); 4993 } 4994 #elif GTK_CHECK_VERSION(3,0,0) 4995 gtk_widget_override_font(preedit_label, gui.norm_font); 4996 4997 vim_snprintf(buf, sizeof(buf), "#%06X", gui.norm_pixel); 4998 gdk_rgba_parse(&color, buf); 4999 gtk_widget_override_color(preedit_label, GTK_STATE_FLAG_NORMAL, &color); 5000 5001 vim_snprintf(buf, sizeof(buf), "#%06X", gui.back_pixel); 5002 gdk_rgba_parse(&color, buf); 5003 gtk_widget_override_background_color(preedit_label, GTK_STATE_FLAG_NORMAL, 5004 &color); 5005 #else 5006 gtk_widget_modify_font(preedit_label, gui.norm_font); 5007 5008 vim_snprintf(buf, sizeof(buf), "#%06X", (unsigned)gui.norm_pixel); 5009 gdk_color_parse(buf, &color); 5010 gtk_widget_modify_fg(preedit_label, GTK_STATE_NORMAL, &color); 5011 5012 vim_snprintf(buf, sizeof(buf), "#%06X", (unsigned)gui.back_pixel); 5013 gdk_color_parse(buf, &color); 5014 gtk_widget_modify_bg(preedit_window, GTK_STATE_NORMAL, &color); 5015 #endif 5016 5017 gtk_im_context_get_preedit_string(xic, &preedit_string, &attr_list, NULL); 5018 5019 if (preedit_string[0] != NUL) 5020 { 5021 gtk_label_set_text(GTK_LABEL(preedit_label), preedit_string); 5022 gtk_label_set_attributes(GTK_LABEL(preedit_label), attr_list); 5023 5024 layout = gtk_label_get_layout(GTK_LABEL(preedit_label)); 5025 pango_layout_get_pixel_size(layout, &w, &h); 5026 h = MAX(h, gui.char_height); 5027 gtk_window_resize(GTK_WINDOW(preedit_window), w, h); 5028 5029 gtk_widget_show_all(preedit_window); 5030 5031 im_preedit_window_set_position(); 5032 } 5033 5034 g_free(preedit_string); 5035 pango_attr_list_unref(attr_list); 5036 } 5037 5038 static void 5039 im_preedit_window_close() 5040 { 5041 if (preedit_window != NULL) 5042 gtk_widget_hide(preedit_window); 5043 } 5044 5045 static void 5046 im_show_preedit() 5047 { 5048 im_preedit_window_open(); 5049 5050 if (p_mh) /* blank out the pointer if necessary */ 5051 gui_mch_mousehide(TRUE); 5052 } 5053 5054 static void 5055 im_delete_preedit(void) 5056 { 5057 char_u bskey[] = {CSI, 'k', 'b'}; 5058 char_u delkey[] = {CSI, 'k', 'D'}; 5059 5060 if (p_imst == IM_OVER_THE_SPOT) 5061 { 5062 im_preedit_window_close(); 5063 return; 5064 } 5065 5066 if (State & NORMAL 5067 #ifdef FEAT_TERMINAL 5068 && !term_use_loop() 5069 #endif 5070 ) 5071 { 5072 im_preedit_cursor = 0; 5073 return; 5074 } 5075 for (; im_preedit_cursor > 0; --im_preedit_cursor) 5076 add_to_input_buf(bskey, (int)sizeof(bskey)); 5077 5078 for (; im_preedit_trailing > 0; --im_preedit_trailing) 5079 add_to_input_buf(delkey, (int)sizeof(delkey)); 5080 } 5081 5082 /* 5083 * Move the cursor left by "num_move_back" characters. 5084 * Note that ins_left() checks im_is_preediting() to avoid breaking undo for 5085 * these K_LEFT keys. 5086 */ 5087 static void 5088 im_correct_cursor(int num_move_back) 5089 { 5090 char_u backkey[] = {CSI, 'k', 'l'}; 5091 5092 if (State & NORMAL) 5093 return; 5094 # ifdef FEAT_RIGHTLEFT 5095 if ((State & CMDLINE) == 0 && curwin != NULL && curwin->w_p_rl) 5096 backkey[2] = 'r'; 5097 # endif 5098 for (; num_move_back > 0; --num_move_back) 5099 add_to_input_buf(backkey, (int)sizeof(backkey)); 5100 } 5101 5102 static int xim_expected_char = NUL; 5103 static int xim_ignored_char = FALSE; 5104 5105 /* 5106 * Update the mode and cursor while in an IM callback. 5107 */ 5108 static void 5109 im_show_info(void) 5110 { 5111 int old_vgetc_busy; 5112 5113 old_vgetc_busy = vgetc_busy; 5114 vgetc_busy = TRUE; 5115 showmode(); 5116 vgetc_busy = old_vgetc_busy; 5117 if ((State & NORMAL) || (State & INSERT)) 5118 setcursor(); 5119 out_flush(); 5120 } 5121 5122 /* 5123 * Callback invoked when the user finished preediting. 5124 * Put the final string into the input buffer. 5125 */ 5126 static void 5127 im_commit_cb(GtkIMContext *context UNUSED, 5128 const gchar *str, 5129 gpointer data UNUSED) 5130 { 5131 int slen = (int)STRLEN(str); 5132 int add_to_input = TRUE; 5133 int clen; 5134 int len = slen; 5135 int commit_with_preedit = TRUE; 5136 char_u *im_str; 5137 5138 #ifdef XIM_DEBUG 5139 xim_log("im_commit_cb(): %s\n", str); 5140 #endif 5141 5142 if (p_imst == IM_ON_THE_SPOT) 5143 { 5144 /* The imhangul module doesn't reset the preedit string before 5145 * committing. Call im_delete_preedit() to work around that. */ 5146 im_delete_preedit(); 5147 5148 /* Indicate that preediting has finished. */ 5149 if (preedit_start_col == MAXCOL) 5150 { 5151 init_preedit_start_col(); 5152 commit_with_preedit = FALSE; 5153 } 5154 5155 /* The thing which setting "preedit_start_col" to MAXCOL means that 5156 * "preedit_start_col" will be set forcedly when calling 5157 * preedit_changed_cb() next time. 5158 * "preedit_start_col" should not reset with MAXCOL on this part. Vim 5159 * is simulating the preediting by using add_to_input_str(). when 5160 * preedit begin immediately before committed, the typebuf is not 5161 * flushed to screen, then it can't get correct "preedit_start_col". 5162 * Thus, it should calculate the cells by adding cells of the committed 5163 * string. */ 5164 if (input_conv.vc_type != CONV_NONE) 5165 { 5166 im_str = string_convert(&input_conv, (char_u *)str, &len); 5167 g_return_if_fail(im_str != NULL); 5168 } 5169 else 5170 im_str = (char_u *)str; 5171 5172 clen = mb_string2cells(im_str, len); 5173 5174 if (input_conv.vc_type != CONV_NONE) 5175 vim_free(im_str); 5176 preedit_start_col += clen; 5177 } 5178 5179 /* Is this a single character that matches a keypad key that's just 5180 * been pressed? If so, we don't want it to be entered as such - let 5181 * us carry on processing the raw keycode so that it may be used in 5182 * mappings as <kSomething>. */ 5183 if (xim_expected_char != NUL) 5184 { 5185 /* We're currently processing a keypad or other special key */ 5186 if (slen == 1 && str[0] == xim_expected_char) 5187 { 5188 /* It's a match - don't do it here */ 5189 xim_ignored_char = TRUE; 5190 add_to_input = FALSE; 5191 } 5192 else 5193 { 5194 /* Not a match */ 5195 xim_ignored_char = FALSE; 5196 } 5197 } 5198 5199 if (add_to_input) 5200 im_add_to_input((char_u *)str, slen); 5201 5202 if (p_imst == IM_ON_THE_SPOT) 5203 { 5204 /* Inserting chars while "im_is_active" is set does not cause a 5205 * change of buffer. When the chars are committed the buffer must be 5206 * marked as changed. */ 5207 if (!commit_with_preedit) 5208 preedit_start_col = MAXCOL; 5209 5210 /* This flag is used in changed() at next call. */ 5211 xim_changed_while_preediting = TRUE; 5212 } 5213 5214 if (gtk_main_level() > 0) 5215 gtk_main_quit(); 5216 } 5217 5218 /* 5219 * Callback invoked after start to the preedit. 5220 */ 5221 static void 5222 im_preedit_start_cb(GtkIMContext *context UNUSED, gpointer data UNUSED) 5223 { 5224 #ifdef XIM_DEBUG 5225 xim_log("im_preedit_start_cb()\n"); 5226 #endif 5227 5228 im_is_active = TRUE; 5229 preedit_is_active = TRUE; 5230 gui_update_cursor(TRUE, FALSE); 5231 im_show_info(); 5232 } 5233 5234 /* 5235 * Callback invoked after end to the preedit. 5236 */ 5237 static void 5238 im_preedit_end_cb(GtkIMContext *context UNUSED, gpointer data UNUSED) 5239 { 5240 #ifdef XIM_DEBUG 5241 xim_log("im_preedit_end_cb()\n"); 5242 #endif 5243 im_delete_preedit(); 5244 5245 /* Indicate that preediting has finished */ 5246 if (p_imst == IM_ON_THE_SPOT) 5247 preedit_start_col = MAXCOL; 5248 xim_has_preediting = FALSE; 5249 5250 #if 0 5251 /* Removal of this line suggested by Takuhiro Nishioka. Fixes that IM was 5252 * switched off unintentionally. We now use preedit_is_active (added by 5253 * SungHyun Nam). */ 5254 im_is_active = FALSE; 5255 #endif 5256 preedit_is_active = FALSE; 5257 gui_update_cursor(TRUE, FALSE); 5258 im_show_info(); 5259 } 5260 5261 /* 5262 * Callback invoked after changes to the preedit string. If the preedit 5263 * string was empty before, remember the preedit start column so we know 5264 * where to apply feedback attributes. Delete the previous preedit string 5265 * if there was one, save the new preedit cursor offset, and put the new 5266 * string into the input buffer. 5267 * 5268 * TODO: The pragmatic "put into input buffer" approach used here has 5269 * several fundamental problems: 5270 * 5271 * - The characters in the preedit string are subject to remapping. 5272 * That's broken, only the finally committed string should be remapped. 5273 * 5274 * - There is a race condition involved: The retrieved value for the 5275 * current cursor position will be wrong if any unprocessed characters 5276 * are still queued in the input buffer. 5277 * 5278 * - Due to the lack of synchronization between the file buffer in memory 5279 * and any typed characters, it's practically impossible to implement the 5280 * "retrieve_surrounding" and "delete_surrounding" signals reliably. IM 5281 * modules for languages such as Thai are likely to rely on this feature 5282 * for proper operation. 5283 * 5284 * Conclusions: I think support for preediting needs to be moved to the 5285 * core parts of Vim. Ideally, until it has been committed, the preediting 5286 * string should only be displayed and not affect the buffer content at all. 5287 * The question how to deal with the synchronization issue still remains. 5288 * Circumventing the input buffer is probably not desirable. Anyway, I think 5289 * implementing "retrieve_surrounding" is the only hard problem. 5290 * 5291 * One way to solve all of this in a clean manner would be to queue all key 5292 * press/release events "as is" in the input buffer, and apply the IM filtering 5293 * at the receiving end of the queue. This, however, would have a rather large 5294 * impact on the code base. If there is an easy way to force processing of all 5295 * remaining input from within the "retrieve_surrounding" signal handler, this 5296 * might not be necessary. Gotta ask on vim-dev for opinions. 5297 */ 5298 static void 5299 im_preedit_changed_cb(GtkIMContext *context, gpointer data UNUSED) 5300 { 5301 char *preedit_string = NULL; 5302 int cursor_index = 0; 5303 int num_move_back = 0; 5304 char_u *str; 5305 char_u *p; 5306 int i; 5307 5308 if (p_imst == IM_ON_THE_SPOT) 5309 gtk_im_context_get_preedit_string(context, 5310 &preedit_string, NULL, 5311 &cursor_index); 5312 else 5313 gtk_im_context_get_preedit_string(context, 5314 &preedit_string, NULL, 5315 NULL); 5316 5317 #ifdef XIM_DEBUG 5318 xim_log("im_preedit_changed_cb(): %s\n", preedit_string); 5319 #endif 5320 5321 g_return_if_fail(preedit_string != NULL); /* just in case */ 5322 5323 if (p_imst == IM_OVER_THE_SPOT) 5324 { 5325 if (preedit_string[0] == NUL) 5326 { 5327 xim_has_preediting = FALSE; 5328 im_delete_preedit(); 5329 } 5330 else 5331 { 5332 xim_has_preediting = TRUE; 5333 im_show_preedit(); 5334 } 5335 } 5336 else 5337 { 5338 /* If preedit_start_col is MAXCOL set it to the current cursor position. */ 5339 if (preedit_start_col == MAXCOL && preedit_string[0] != '\0') 5340 { 5341 xim_has_preediting = TRUE; 5342 5343 /* Urgh, this breaks if the input buffer isn't empty now */ 5344 init_preedit_start_col(); 5345 } 5346 else if (cursor_index == 0 && preedit_string[0] == '\0') 5347 { 5348 xim_has_preediting = FALSE; 5349 5350 /* If at the start position (after typing backspace) 5351 * preedit_start_col must be reset. */ 5352 preedit_start_col = MAXCOL; 5353 } 5354 5355 im_delete_preedit(); 5356 5357 /* 5358 * Compute the end of the preediting area: "preedit_end_col". 5359 * According to the documentation of gtk_im_context_get_preedit_string(), 5360 * the cursor_pos output argument returns the offset in bytes. This is 5361 * unfortunately not true -- real life shows the offset is in characters, 5362 * and the GTK+ source code agrees with me. Will file a bug later. 5363 */ 5364 if (preedit_start_col != MAXCOL) 5365 preedit_end_col = preedit_start_col; 5366 str = (char_u *)preedit_string; 5367 for (p = str, i = 0; *p != NUL; p += utf_byte2len(*p), ++i) 5368 { 5369 int is_composing; 5370 5371 is_composing = ((*p & 0x80) != 0 && utf_iscomposing(utf_ptr2char(p))); 5372 /* 5373 * These offsets are used as counters when generating <BS> and <Del> 5374 * to delete the preedit string. So don't count composing characters 5375 * unless 'delcombine' is enabled. 5376 */ 5377 if (!is_composing || p_deco) 5378 { 5379 if (i < cursor_index) 5380 ++im_preedit_cursor; 5381 else 5382 ++im_preedit_trailing; 5383 } 5384 if (!is_composing && i >= cursor_index) 5385 { 5386 /* This is essentially the same as im_preedit_trailing, except 5387 * composing characters are not counted even if p_deco is set. */ 5388 ++num_move_back; 5389 } 5390 if (preedit_start_col != MAXCOL) 5391 preedit_end_col += utf_ptr2cells(p); 5392 } 5393 5394 if (p > str) 5395 { 5396 im_add_to_input(str, (int)(p - str)); 5397 im_correct_cursor(num_move_back); 5398 } 5399 } 5400 5401 g_free(preedit_string); 5402 5403 if (gtk_main_level() > 0) 5404 gtk_main_quit(); 5405 } 5406 5407 /* 5408 * Translate the Pango attributes at iter to Vim highlighting attributes. 5409 * Ignore attributes not supported by Vim highlighting. This shouldn't have 5410 * too much impact -- right now we handle even more attributes than necessary 5411 * for the IM modules I tested with. 5412 */ 5413 static int 5414 translate_pango_attributes(PangoAttrIterator *iter) 5415 { 5416 PangoAttribute *attr; 5417 int char_attr = HL_NORMAL; 5418 5419 attr = pango_attr_iterator_get(iter, PANGO_ATTR_UNDERLINE); 5420 if (attr != NULL && ((PangoAttrInt *)attr)->value 5421 != (int)PANGO_UNDERLINE_NONE) 5422 char_attr |= HL_UNDERLINE; 5423 5424 attr = pango_attr_iterator_get(iter, PANGO_ATTR_WEIGHT); 5425 if (attr != NULL && ((PangoAttrInt *)attr)->value >= (int)PANGO_WEIGHT_BOLD) 5426 char_attr |= HL_BOLD; 5427 5428 attr = pango_attr_iterator_get(iter, PANGO_ATTR_STYLE); 5429 if (attr != NULL && ((PangoAttrInt *)attr)->value 5430 != (int)PANGO_STYLE_NORMAL) 5431 char_attr |= HL_ITALIC; 5432 5433 attr = pango_attr_iterator_get(iter, PANGO_ATTR_BACKGROUND); 5434 if (attr != NULL) 5435 { 5436 const PangoColor *color = &((PangoAttrColor *)attr)->color; 5437 5438 /* Assume inverse if black background is requested */ 5439 if ((color->red | color->green | color->blue) == 0) 5440 char_attr |= HL_INVERSE; 5441 } 5442 5443 return char_attr; 5444 } 5445 5446 /* 5447 * Retrieve the highlighting attributes at column col in the preedit string. 5448 * Return -1 if not in preediting mode or if col is out of range. 5449 */ 5450 int 5451 im_get_feedback_attr(int col) 5452 { 5453 char *preedit_string = NULL; 5454 PangoAttrList *attr_list = NULL; 5455 int char_attr = -1; 5456 5457 if (xic == NULL) 5458 return char_attr; 5459 5460 gtk_im_context_get_preedit_string(xic, &preedit_string, &attr_list, NULL); 5461 5462 if (preedit_string != NULL && attr_list != NULL) 5463 { 5464 int idx; 5465 5466 /* Get the byte index as used by PangoAttrIterator */ 5467 for (idx = 0; col > 0 && preedit_string[idx] != '\0'; --col) 5468 idx += utfc_ptr2len((char_u *)preedit_string + idx); 5469 5470 if (preedit_string[idx] != '\0') 5471 { 5472 PangoAttrIterator *iter; 5473 int start, end; 5474 5475 char_attr = HL_NORMAL; 5476 iter = pango_attr_list_get_iterator(attr_list); 5477 5478 /* Extract all relevant attributes from the list. */ 5479 do 5480 { 5481 pango_attr_iterator_range(iter, &start, &end); 5482 5483 if (idx >= start && idx < end) 5484 char_attr |= translate_pango_attributes(iter); 5485 } 5486 while (pango_attr_iterator_next(iter)); 5487 5488 pango_attr_iterator_destroy(iter); 5489 } 5490 } 5491 5492 if (attr_list != NULL) 5493 pango_attr_list_unref(attr_list); 5494 g_free(preedit_string); 5495 5496 return char_attr; 5497 } 5498 5499 void 5500 xim_init(void) 5501 { 5502 #ifdef XIM_DEBUG 5503 xim_log("xim_init()\n"); 5504 #endif 5505 5506 g_return_if_fail(gui.drawarea != NULL); 5507 g_return_if_fail(gtk_widget_get_window(gui.drawarea) != NULL); 5508 5509 xic = gtk_im_multicontext_new(); 5510 g_object_ref(xic); 5511 5512 im_commit_handler_id = g_signal_connect(G_OBJECT(xic), "commit", 5513 G_CALLBACK(&im_commit_cb), NULL); 5514 g_signal_connect(G_OBJECT(xic), "preedit_changed", 5515 G_CALLBACK(&im_preedit_changed_cb), NULL); 5516 g_signal_connect(G_OBJECT(xic), "preedit_start", 5517 G_CALLBACK(&im_preedit_start_cb), NULL); 5518 g_signal_connect(G_OBJECT(xic), "preedit_end", 5519 G_CALLBACK(&im_preedit_end_cb), NULL); 5520 5521 gtk_im_context_set_client_window(xic, gtk_widget_get_window(gui.drawarea)); 5522 } 5523 5524 void 5525 im_shutdown(void) 5526 { 5527 #ifdef XIM_DEBUG 5528 xim_log("im_shutdown()\n"); 5529 #endif 5530 5531 if (xic != NULL) 5532 { 5533 gtk_im_context_focus_out(xic); 5534 g_object_unref(xic); 5535 xic = NULL; 5536 } 5537 im_is_active = FALSE; 5538 im_commit_handler_id = 0; 5539 if (p_imst == IM_ON_THE_SPOT) 5540 preedit_start_col = MAXCOL; 5541 xim_has_preediting = FALSE; 5542 } 5543 5544 /* 5545 * Convert the string argument to keyval and state for GdkEventKey. 5546 * If str is valid return TRUE, otherwise FALSE. 5547 * 5548 * See 'imactivatekey' for documentation of the format. 5549 */ 5550 static int 5551 im_string_to_keyval(const char *str, unsigned int *keyval, unsigned int *state) 5552 { 5553 const char *mods_end; 5554 unsigned tmp_keyval; 5555 unsigned tmp_state = 0; 5556 5557 mods_end = strrchr(str, '-'); 5558 mods_end = (mods_end != NULL) ? mods_end + 1 : str; 5559 5560 /* Parse modifier keys */ 5561 while (str < mods_end) 5562 switch (*str++) 5563 { 5564 case '-': break; 5565 case 'S': case 's': tmp_state |= (unsigned)GDK_SHIFT_MASK; break; 5566 case 'L': case 'l': tmp_state |= (unsigned)GDK_LOCK_MASK; break; 5567 case 'C': case 'c': tmp_state |= (unsigned)GDK_CONTROL_MASK;break; 5568 case '1': tmp_state |= (unsigned)GDK_MOD1_MASK; break; 5569 case '2': tmp_state |= (unsigned)GDK_MOD2_MASK; break; 5570 case '3': tmp_state |= (unsigned)GDK_MOD3_MASK; break; 5571 case '4': tmp_state |= (unsigned)GDK_MOD4_MASK; break; 5572 case '5': tmp_state |= (unsigned)GDK_MOD5_MASK; break; 5573 default: 5574 return FALSE; 5575 } 5576 5577 tmp_keyval = gdk_keyval_from_name(str); 5578 5579 if (tmp_keyval == 0 || tmp_keyval == GDK_VoidSymbol) 5580 return FALSE; 5581 5582 if (keyval != NULL) 5583 *keyval = tmp_keyval; 5584 if (state != NULL) 5585 *state = tmp_state; 5586 5587 return TRUE; 5588 } 5589 5590 /* 5591 * Return TRUE if p_imak is valid, otherwise FALSE. As a special case, an 5592 * empty string is also regarded as valid. 5593 * 5594 * Note: The numerical key value of p_imak is cached if it was valid; thus 5595 * boldly assuming im_xim_isvalid_imactivate() will always be called whenever 5596 * 'imak' changes. This is currently the case but not obvious -- should 5597 * probably rename the function for clarity. 5598 */ 5599 int 5600 im_xim_isvalid_imactivate(void) 5601 { 5602 if (p_imak[0] == NUL) 5603 { 5604 im_activatekey_keyval = GDK_VoidSymbol; 5605 im_activatekey_state = 0; 5606 return TRUE; 5607 } 5608 5609 return im_string_to_keyval((const char *)p_imak, 5610 &im_activatekey_keyval, 5611 &im_activatekey_state); 5612 } 5613 5614 static void 5615 im_synthesize_keypress(unsigned int keyval, unsigned int state) 5616 { 5617 GdkEventKey *event; 5618 5619 event = (GdkEventKey *)gdk_event_new(GDK_KEY_PRESS); 5620 g_object_ref(gtk_widget_get_window(gui.drawarea)); 5621 /* unreffed by gdk_event_free() */ 5622 event->window = gtk_widget_get_window(gui.drawarea); 5623 event->send_event = TRUE; 5624 event->time = GDK_CURRENT_TIME; 5625 event->state = state; 5626 event->keyval = keyval; 5627 event->hardware_keycode = /* needed for XIM */ 5628 XKeysymToKeycode(GDK_WINDOW_XDISPLAY(event->window), (KeySym)keyval); 5629 event->length = 0; 5630 event->string = NULL; 5631 5632 gtk_im_context_filter_keypress(xic, event); 5633 5634 /* For consistency, also send the corresponding release event. */ 5635 event->type = GDK_KEY_RELEASE; 5636 event->send_event = FALSE; 5637 gtk_im_context_filter_keypress(xic, event); 5638 5639 gdk_event_free((GdkEvent *)event); 5640 } 5641 5642 void 5643 xim_reset(void) 5644 { 5645 # ifdef FEAT_EVAL 5646 if (USE_IMACTIVATEFUNC) 5647 call_imactivatefunc(im_is_active); 5648 else 5649 # endif 5650 if (xic != NULL) 5651 { 5652 gtk_im_context_reset(xic); 5653 5654 if (p_imdisable) 5655 im_shutdown(); 5656 else 5657 { 5658 xim_set_focus(gui.in_focus); 5659 5660 if (im_activatekey_keyval != GDK_VoidSymbol) 5661 { 5662 if (im_is_active) 5663 { 5664 g_signal_handler_block(xic, im_commit_handler_id); 5665 im_synthesize_keypress(im_activatekey_keyval, 5666 im_activatekey_state); 5667 g_signal_handler_unblock(xic, im_commit_handler_id); 5668 } 5669 } 5670 else 5671 { 5672 im_shutdown(); 5673 xim_init(); 5674 xim_set_focus(gui.in_focus); 5675 } 5676 } 5677 } 5678 5679 if (p_imst == IM_ON_THE_SPOT) 5680 preedit_start_col = MAXCOL; 5681 xim_has_preediting = FALSE; 5682 } 5683 5684 int 5685 xim_queue_key_press_event(GdkEventKey *event, int down) 5686 { 5687 if (down) 5688 { 5689 /* 5690 * Workaround GTK2 XIM 'feature' that always converts keypad keys to 5691 * chars., even when not part of an IM sequence (ref. feature of 5692 * gdk/gdkkeyuni.c). 5693 * Flag any keypad keys that might represent a single char. 5694 * If this (on its own - i.e., not part of an IM sequence) is 5695 * committed while we're processing one of these keys, we can ignore 5696 * that commit and go ahead & process it ourselves. That way we can 5697 * still distinguish keypad keys for use in mappings. 5698 * Also add GDK_space to make <S-Space> work. 5699 */ 5700 switch (event->keyval) 5701 { 5702 case GDK_KP_Add: xim_expected_char = '+'; break; 5703 case GDK_KP_Subtract: xim_expected_char = '-'; break; 5704 case GDK_KP_Divide: xim_expected_char = '/'; break; 5705 case GDK_KP_Multiply: xim_expected_char = '*'; break; 5706 case GDK_KP_Decimal: xim_expected_char = '.'; break; 5707 case GDK_KP_Equal: xim_expected_char = '='; break; 5708 case GDK_KP_0: xim_expected_char = '0'; break; 5709 case GDK_KP_1: xim_expected_char = '1'; break; 5710 case GDK_KP_2: xim_expected_char = '2'; break; 5711 case GDK_KP_3: xim_expected_char = '3'; break; 5712 case GDK_KP_4: xim_expected_char = '4'; break; 5713 case GDK_KP_5: xim_expected_char = '5'; break; 5714 case GDK_KP_6: xim_expected_char = '6'; break; 5715 case GDK_KP_7: xim_expected_char = '7'; break; 5716 case GDK_KP_8: xim_expected_char = '8'; break; 5717 case GDK_KP_9: xim_expected_char = '9'; break; 5718 case GDK_space: xim_expected_char = ' '; break; 5719 default: xim_expected_char = NUL; 5720 } 5721 xim_ignored_char = FALSE; 5722 } 5723 5724 /* 5725 * When typing fFtT, XIM may be activated. Thus it must pass 5726 * gtk_im_context_filter_keypress() in Normal mode. 5727 * And while doing :sh too. 5728 */ 5729 if (xic != NULL && !p_imdisable 5730 && (State & (INSERT | CMDLINE | NORMAL | EXTERNCMD)) != 0) 5731 { 5732 /* 5733 * Filter 'imactivatekey' and map it to CTRL-^. This way, Vim is 5734 * always aware of the current status of IM, and can even emulate 5735 * the activation key for modules that don't support one. 5736 */ 5737 if (event->keyval == im_activatekey_keyval 5738 && (event->state & im_activatekey_state) == im_activatekey_state) 5739 { 5740 unsigned int state_mask; 5741 5742 /* Require the state of the 3 most used modifiers to match exactly. 5743 * Otherwise e.g. <S-C-space> would be unusable for other purposes 5744 * if the IM activate key is <S-space>. */ 5745 state_mask = im_activatekey_state; 5746 state_mask |= ((int)GDK_SHIFT_MASK | (int)GDK_CONTROL_MASK 5747 | (int)GDK_MOD1_MASK); 5748 5749 if ((event->state & state_mask) != im_activatekey_state) 5750 return FALSE; 5751 5752 /* Don't send it a second time on GDK_KEY_RELEASE. */ 5753 if (event->type != GDK_KEY_PRESS) 5754 return TRUE; 5755 5756 if (map_to_exists_mode((char_u *)"", LANGMAP, FALSE)) 5757 { 5758 im_set_active(FALSE); 5759 5760 /* ":lmap" mappings exists, toggle use of mappings. */ 5761 State ^= LANGMAP; 5762 if (State & LANGMAP) 5763 { 5764 curbuf->b_p_iminsert = B_IMODE_NONE; 5765 State &= ~LANGMAP; 5766 } 5767 else 5768 { 5769 curbuf->b_p_iminsert = B_IMODE_LMAP; 5770 State |= LANGMAP; 5771 } 5772 return TRUE; 5773 } 5774 5775 return gtk_im_context_filter_keypress(xic, event); 5776 } 5777 5778 /* Don't filter events through the IM context if IM isn't active 5779 * right now. Unlike with GTK+ 1.2 we cannot rely on the IM module 5780 * not doing anything before the activation key was sent. */ 5781 if (im_activatekey_keyval == GDK_VoidSymbol || im_is_active) 5782 { 5783 int imresult = gtk_im_context_filter_keypress(xic, event); 5784 5785 if (p_imst == IM_ON_THE_SPOT) 5786 { 5787 /* Some XIM send following sequence: 5788 * 1. preedited string. 5789 * 2. committed string. 5790 * 3. line changed key. 5791 * 4. preedited string. 5792 * 5. remove preedited string. 5793 * if 3, Vim can't move back the above line for 5. 5794 * thus, this part should not parse the key. */ 5795 if (!imresult && preedit_start_col != MAXCOL 5796 && event->keyval == GDK_Return) 5797 { 5798 im_synthesize_keypress(GDK_Return, 0U); 5799 return FALSE; 5800 } 5801 } 5802 5803 /* If XIM tried to commit a keypad key as a single char., 5804 * ignore it so we can use the keypad key 'raw', for mappings. */ 5805 if (xim_expected_char != NUL && xim_ignored_char) 5806 /* We had a keypad key, and XIM tried to thieve it */ 5807 return FALSE; 5808 5809 /* This is supposed to fix a problem with iBus, that space 5810 * characters don't work in input mode. */ 5811 xim_expected_char = NUL; 5812 5813 /* Normal processing */ 5814 return imresult; 5815 } 5816 } 5817 5818 return FALSE; 5819 } 5820 5821 int 5822 im_get_status(void) 5823 { 5824 # ifdef FEAT_EVAL 5825 if (USE_IMSTATUSFUNC) 5826 return call_imstatusfunc(); 5827 # endif 5828 return im_is_active; 5829 } 5830 5831 int 5832 preedit_get_status(void) 5833 { 5834 return preedit_is_active; 5835 } 5836 5837 int 5838 im_is_preediting(void) 5839 { 5840 return xim_has_preediting; 5841 } 5842 5843 # else /* !FEAT_GUI_GTK */ 5844 5845 static int xim_is_active = FALSE; /* XIM should be active in the current 5846 mode */ 5847 static int xim_has_focus = FALSE; /* XIM is really being used for Vim */ 5848 # ifdef FEAT_GUI_X11 5849 static XIMStyle input_style; 5850 static int status_area_enabled = TRUE; 5851 # endif 5852 5853 /* 5854 * Switch using XIM on/off. This is used by the code that changes "State". 5855 * When 'imactivatefunc' is defined use that function instead. 5856 */ 5857 void 5858 im_set_active(int active_arg) 5859 { 5860 int active = active_arg; 5861 5862 /* If 'imdisable' is set, XIM is never active. */ 5863 if (p_imdisable) 5864 active = FALSE; 5865 else if (input_style & XIMPreeditPosition) 5866 /* There is a problem in switching XIM off when preediting is used, 5867 * and it is not clear how this can be solved. For now, keep XIM on 5868 * all the time, like it was done in Vim 5.8. */ 5869 active = TRUE; 5870 5871 # if defined(FEAT_EVAL) 5872 if (USE_IMACTIVATEFUNC) 5873 { 5874 if (active != im_get_status()) 5875 { 5876 call_imactivatefunc(active); 5877 xim_has_focus = active; 5878 } 5879 return; 5880 } 5881 # endif 5882 5883 if (xic == NULL) 5884 return; 5885 5886 /* Remember the active state, it is needed when Vim gets keyboard focus. */ 5887 xim_is_active = active; 5888 xim_set_preedit(); 5889 } 5890 5891 /* 5892 * Adjust using XIM for gaining or losing keyboard focus. Also called when 5893 * "xim_is_active" changes. 5894 */ 5895 void 5896 xim_set_focus(int focus) 5897 { 5898 if (xic == NULL) 5899 return; 5900 5901 /* 5902 * XIM only gets focus when the Vim window has keyboard focus and XIM has 5903 * been set active for the current mode. 5904 */ 5905 if (focus && xim_is_active) 5906 { 5907 if (!xim_has_focus) 5908 { 5909 xim_has_focus = TRUE; 5910 XSetICFocus(xic); 5911 } 5912 } 5913 else 5914 { 5915 if (xim_has_focus) 5916 { 5917 xim_has_focus = FALSE; 5918 XUnsetICFocus(xic); 5919 } 5920 } 5921 } 5922 5923 void 5924 im_set_position(int row UNUSED, int col UNUSED) 5925 { 5926 xim_set_preedit(); 5927 } 5928 5929 /* 5930 * Set the XIM to the current cursor position. 5931 */ 5932 void 5933 xim_set_preedit(void) 5934 { 5935 XVaNestedList attr_list; 5936 XRectangle spot_area; 5937 XPoint over_spot; 5938 int line_space; 5939 5940 if (xic == NULL) 5941 return; 5942 5943 xim_set_focus(TRUE); 5944 5945 if (!xim_has_focus) 5946 { 5947 /* hide XIM cursor */ 5948 over_spot.x = 0; 5949 over_spot.y = -100; /* arbitrary invisible position */ 5950 attr_list = (XVaNestedList) XVaCreateNestedList(0, 5951 XNSpotLocation, 5952 &over_spot, 5953 NULL); 5954 XSetICValues(xic, XNPreeditAttributes, attr_list, NULL); 5955 XFree(attr_list); 5956 return; 5957 } 5958 5959 if (input_style & XIMPreeditPosition) 5960 { 5961 if (xim_fg_color == INVALCOLOR) 5962 { 5963 xim_fg_color = gui.def_norm_pixel; 5964 xim_bg_color = gui.def_back_pixel; 5965 } 5966 over_spot.x = TEXT_X(gui.col); 5967 over_spot.y = TEXT_Y(gui.row); 5968 spot_area.x = 0; 5969 spot_area.y = 0; 5970 spot_area.height = gui.char_height * Rows; 5971 spot_area.width = gui.char_width * Columns; 5972 line_space = gui.char_height; 5973 attr_list = (XVaNestedList) XVaCreateNestedList(0, 5974 XNSpotLocation, &over_spot, 5975 XNForeground, (Pixel) xim_fg_color, 5976 XNBackground, (Pixel) xim_bg_color, 5977 XNArea, &spot_area, 5978 XNLineSpace, line_space, 5979 NULL); 5980 if (XSetICValues(xic, XNPreeditAttributes, attr_list, NULL)) 5981 emsg(_("E284: Cannot set IC values")); 5982 XFree(attr_list); 5983 } 5984 } 5985 5986 # if defined(FEAT_GUI_X11) 5987 static char e_xim[] = N_("E285: Failed to create input context"); 5988 # endif 5989 5990 # if defined(FEAT_GUI_X11) || defined(PROTO) 5991 # if defined(XtSpecificationRelease) && XtSpecificationRelease >= 6 && !defined(SUN_SYSTEM) 5992 # define USE_X11R6_XIM 5993 # endif 5994 5995 static int xim_real_init(Window x11_window, Display *x11_display); 5996 5997 5998 # ifdef USE_X11R6_XIM 5999 static void 6000 xim_instantiate_cb( 6001 Display *display, 6002 XPointer client_data UNUSED, 6003 XPointer call_data UNUSED) 6004 { 6005 Window x11_window; 6006 Display *x11_display; 6007 6008 # ifdef XIM_DEBUG 6009 xim_log("xim_instantiate_cb()\n"); 6010 # endif 6011 6012 gui_get_x11_windis(&x11_window, &x11_display); 6013 if (display != x11_display) 6014 return; 6015 6016 xim_real_init(x11_window, x11_display); 6017 gui_set_shellsize(FALSE, FALSE, RESIZE_BOTH); 6018 if (xic != NULL) 6019 XUnregisterIMInstantiateCallback(x11_display, NULL, NULL, NULL, 6020 xim_instantiate_cb, NULL); 6021 } 6022 6023 static void 6024 xim_destroy_cb( 6025 XIM im UNUSED, 6026 XPointer client_data UNUSED, 6027 XPointer call_data UNUSED) 6028 { 6029 Window x11_window; 6030 Display *x11_display; 6031 6032 # ifdef XIM_DEBUG 6033 xim_log("xim_destroy_cb()\n"); 6034 #endif 6035 gui_get_x11_windis(&x11_window, &x11_display); 6036 6037 xic = NULL; 6038 status_area_enabled = FALSE; 6039 6040 gui_set_shellsize(FALSE, FALSE, RESIZE_BOTH); 6041 6042 XRegisterIMInstantiateCallback(x11_display, NULL, NULL, NULL, 6043 xim_instantiate_cb, NULL); 6044 } 6045 # endif 6046 6047 void 6048 xim_init(void) 6049 { 6050 Window x11_window; 6051 Display *x11_display; 6052 6053 # ifdef XIM_DEBUG 6054 xim_log("xim_init()\n"); 6055 # endif 6056 6057 gui_get_x11_windis(&x11_window, &x11_display); 6058 6059 xic = NULL; 6060 6061 if (xim_real_init(x11_window, x11_display)) 6062 return; 6063 6064 gui_set_shellsize(FALSE, FALSE, RESIZE_BOTH); 6065 6066 # ifdef USE_X11R6_XIM 6067 XRegisterIMInstantiateCallback(x11_display, NULL, NULL, NULL, 6068 xim_instantiate_cb, NULL); 6069 # endif 6070 } 6071 6072 static int 6073 xim_real_init(Window x11_window, Display *x11_display) 6074 { 6075 int i; 6076 char *p, 6077 *s, 6078 *ns, 6079 *end, 6080 tmp[1024]; 6081 # define IMLEN_MAX 40 6082 char buf[IMLEN_MAX + 7]; 6083 XIM xim = NULL; 6084 XIMStyles *xim_styles; 6085 XIMStyle this_input_style = 0; 6086 Boolean found; 6087 XPoint over_spot; 6088 XVaNestedList preedit_list, status_list; 6089 6090 input_style = 0; 6091 status_area_enabled = FALSE; 6092 6093 if (xic != NULL) 6094 return FALSE; 6095 6096 if (gui.rsrc_input_method != NULL && *gui.rsrc_input_method != NUL) 6097 { 6098 strcpy(tmp, gui.rsrc_input_method); 6099 for (ns = s = tmp; ns != NULL && *s != NUL;) 6100 { 6101 s = (char *)skipwhite((char_u *)s); 6102 if (*s == NUL) 6103 break; 6104 if ((ns = end = strchr(s, ',')) == NULL) 6105 end = s + strlen(s); 6106 while (isspace(((char_u *)end)[-1])) 6107 end--; 6108 *end = NUL; 6109 6110 if (strlen(s) <= IMLEN_MAX) 6111 { 6112 strcpy(buf, "@im="); 6113 strcat(buf, s); 6114 if ((p = XSetLocaleModifiers(buf)) != NULL && *p != NUL 6115 && (xim = XOpenIM(x11_display, NULL, NULL, NULL)) 6116 != NULL) 6117 break; 6118 } 6119 6120 s = ns + 1; 6121 } 6122 } 6123 6124 if (xim == NULL && (p = XSetLocaleModifiers("")) != NULL && *p != NUL) 6125 xim = XOpenIM(x11_display, NULL, NULL, NULL); 6126 6127 /* This is supposed to be useful to obtain characters through 6128 * XmbLookupString() without really using a XIM. */ 6129 if (xim == NULL && (p = XSetLocaleModifiers("@im=none")) != NULL 6130 && *p != NUL) 6131 xim = XOpenIM(x11_display, NULL, NULL, NULL); 6132 6133 if (xim == NULL) 6134 { 6135 /* Only give this message when verbose is set, because too many people 6136 * got this message when they didn't want to use a XIM. */ 6137 if (p_verbose > 0) 6138 { 6139 verbose_enter(); 6140 emsg(_("E286: Failed to open input method")); 6141 verbose_leave(); 6142 } 6143 return FALSE; 6144 } 6145 6146 # ifdef USE_X11R6_XIM 6147 { 6148 XIMCallback destroy_cb; 6149 6150 destroy_cb.callback = xim_destroy_cb; 6151 destroy_cb.client_data = NULL; 6152 if (XSetIMValues(xim, XNDestroyCallback, &destroy_cb, NULL)) 6153 emsg(_("E287: Warning: Could not set destroy callback to IM")); 6154 } 6155 # endif 6156 6157 if (XGetIMValues(xim, XNQueryInputStyle, &xim_styles, NULL) || !xim_styles) 6158 { 6159 emsg(_("E288: input method doesn't support any style")); 6160 XCloseIM(xim); 6161 return FALSE; 6162 } 6163 6164 found = False; 6165 strcpy(tmp, gui.rsrc_preedit_type_name); 6166 for (s = tmp; s && !found; ) 6167 { 6168 while (*s && isspace((unsigned char)*s)) 6169 s++; 6170 if (!*s) 6171 break; 6172 if ((ns = end = strchr(s, ',')) != 0) 6173 ns++; 6174 else 6175 end = s + strlen(s); 6176 while (isspace((unsigned char)*end)) 6177 end--; 6178 *end = '\0'; 6179 6180 if (!strcmp(s, "OverTheSpot")) 6181 this_input_style = (XIMPreeditPosition | XIMStatusArea); 6182 else if (!strcmp(s, "OffTheSpot")) 6183 this_input_style = (XIMPreeditArea | XIMStatusArea); 6184 else if (!strcmp(s, "Root")) 6185 this_input_style = (XIMPreeditNothing | XIMStatusNothing); 6186 6187 for (i = 0; (unsigned short)i < xim_styles->count_styles; i++) 6188 { 6189 if (this_input_style == xim_styles->supported_styles[i]) 6190 { 6191 found = True; 6192 break; 6193 } 6194 } 6195 if (!found) 6196 for (i = 0; (unsigned short)i < xim_styles->count_styles; i++) 6197 { 6198 if ((xim_styles->supported_styles[i] & this_input_style) 6199 == (this_input_style & ~XIMStatusArea)) 6200 { 6201 this_input_style &= ~XIMStatusArea; 6202 found = True; 6203 break; 6204 } 6205 } 6206 6207 s = ns; 6208 } 6209 XFree(xim_styles); 6210 6211 if (!found) 6212 { 6213 /* Only give this message when verbose is set, because too many people 6214 * got this message when they didn't want to use a XIM. */ 6215 if (p_verbose > 0) 6216 { 6217 verbose_enter(); 6218 emsg(_("E289: input method doesn't support my preedit type")); 6219 verbose_leave(); 6220 } 6221 XCloseIM(xim); 6222 return FALSE; 6223 } 6224 6225 over_spot.x = TEXT_X(gui.col); 6226 over_spot.y = TEXT_Y(gui.row); 6227 input_style = this_input_style; 6228 6229 /* A crash was reported when trying to pass gui.norm_font as XNFontSet, 6230 * thus that has been removed. Hopefully the default works... */ 6231 # ifdef FEAT_XFONTSET 6232 if (gui.fontset != NOFONTSET) 6233 { 6234 preedit_list = XVaCreateNestedList(0, 6235 XNSpotLocation, &over_spot, 6236 XNForeground, (Pixel)gui.def_norm_pixel, 6237 XNBackground, (Pixel)gui.def_back_pixel, 6238 XNFontSet, (XFontSet)gui.fontset, 6239 NULL); 6240 status_list = XVaCreateNestedList(0, 6241 XNForeground, (Pixel)gui.def_norm_pixel, 6242 XNBackground, (Pixel)gui.def_back_pixel, 6243 XNFontSet, (XFontSet)gui.fontset, 6244 NULL); 6245 } 6246 else 6247 # endif 6248 { 6249 preedit_list = XVaCreateNestedList(0, 6250 XNSpotLocation, &over_spot, 6251 XNForeground, (Pixel)gui.def_norm_pixel, 6252 XNBackground, (Pixel)gui.def_back_pixel, 6253 NULL); 6254 status_list = XVaCreateNestedList(0, 6255 XNForeground, (Pixel)gui.def_norm_pixel, 6256 XNBackground, (Pixel)gui.def_back_pixel, 6257 NULL); 6258 } 6259 6260 xic = XCreateIC(xim, 6261 XNInputStyle, input_style, 6262 XNClientWindow, x11_window, 6263 XNFocusWindow, gui.wid, 6264 XNPreeditAttributes, preedit_list, 6265 XNStatusAttributes, status_list, 6266 NULL); 6267 XFree(status_list); 6268 XFree(preedit_list); 6269 if (xic != NULL) 6270 { 6271 if (input_style & XIMStatusArea) 6272 { 6273 xim_set_status_area(); 6274 status_area_enabled = TRUE; 6275 } 6276 else 6277 gui_set_shellsize(FALSE, FALSE, RESIZE_BOTH); 6278 } 6279 else 6280 { 6281 if (!is_not_a_term()) 6282 emsg(_(e_xim)); 6283 XCloseIM(xim); 6284 return FALSE; 6285 } 6286 6287 return TRUE; 6288 } 6289 6290 # endif /* FEAT_GUI_X11 */ 6291 6292 /* 6293 * Get IM status. When IM is on, return TRUE. Else return FALSE. 6294 * FIXME: This doesn't work correctly: Having focus doesn't always mean XIM is 6295 * active, when not having focus XIM may still be active (e.g., when using a 6296 * tear-off menu item). 6297 */ 6298 int 6299 im_get_status(void) 6300 { 6301 # ifdef FEAT_EVAL 6302 if (USE_IMSTATUSFUNC) 6303 return call_imstatusfunc(); 6304 # endif 6305 return xim_has_focus; 6306 } 6307 6308 # endif /* !FEAT_GUI_GTK */ 6309 6310 # if !defined(FEAT_GUI_GTK) || defined(PROTO) 6311 /* 6312 * Set up the status area. 6313 * 6314 * This should use a separate Widget, but that seems not possible, because 6315 * preedit_area and status_area should be set to the same window as for the 6316 * text input. Unfortunately this means the status area pollutes the text 6317 * window... 6318 */ 6319 void 6320 xim_set_status_area(void) 6321 { 6322 XVaNestedList preedit_list = 0, status_list = 0, list = 0; 6323 XRectangle pre_area, status_area; 6324 6325 if (xic == NULL) 6326 return; 6327 6328 if (input_style & XIMStatusArea) 6329 { 6330 if (input_style & XIMPreeditArea) 6331 { 6332 XRectangle *needed_rect; 6333 6334 /* to get status_area width */ 6335 status_list = XVaCreateNestedList(0, XNAreaNeeded, 6336 &needed_rect, NULL); 6337 XGetICValues(xic, XNStatusAttributes, status_list, NULL); 6338 XFree(status_list); 6339 6340 status_area.width = needed_rect->width; 6341 } 6342 else 6343 status_area.width = gui.char_width * Columns; 6344 6345 status_area.x = 0; 6346 status_area.y = gui.char_height * Rows + gui.border_offset; 6347 if (gui.which_scrollbars[SBAR_BOTTOM]) 6348 status_area.y += gui.scrollbar_height; 6349 #ifdef FEAT_MENU 6350 if (gui.menu_is_active) 6351 status_area.y += gui.menu_height; 6352 #endif 6353 status_area.height = gui.char_height; 6354 status_list = XVaCreateNestedList(0, XNArea, &status_area, NULL); 6355 } 6356 else 6357 { 6358 status_area.x = 0; 6359 status_area.y = gui.char_height * Rows + gui.border_offset; 6360 if (gui.which_scrollbars[SBAR_BOTTOM]) 6361 status_area.y += gui.scrollbar_height; 6362 #ifdef FEAT_MENU 6363 if (gui.menu_is_active) 6364 status_area.y += gui.menu_height; 6365 #endif 6366 status_area.width = 0; 6367 status_area.height = gui.char_height; 6368 } 6369 6370 if (input_style & XIMPreeditArea) /* off-the-spot */ 6371 { 6372 pre_area.x = status_area.x + status_area.width; 6373 pre_area.y = gui.char_height * Rows + gui.border_offset; 6374 pre_area.width = gui.char_width * Columns - pre_area.x; 6375 if (gui.which_scrollbars[SBAR_BOTTOM]) 6376 pre_area.y += gui.scrollbar_height; 6377 #ifdef FEAT_MENU 6378 if (gui.menu_is_active) 6379 pre_area.y += gui.menu_height; 6380 #endif 6381 pre_area.height = gui.char_height; 6382 preedit_list = XVaCreateNestedList(0, XNArea, &pre_area, NULL); 6383 } 6384 else if (input_style & XIMPreeditPosition) /* over-the-spot */ 6385 { 6386 pre_area.x = 0; 6387 pre_area.y = 0; 6388 pre_area.height = gui.char_height * Rows; 6389 pre_area.width = gui.char_width * Columns; 6390 preedit_list = XVaCreateNestedList(0, XNArea, &pre_area, NULL); 6391 } 6392 6393 if (preedit_list && status_list) 6394 list = XVaCreateNestedList(0, XNPreeditAttributes, preedit_list, 6395 XNStatusAttributes, status_list, NULL); 6396 else if (preedit_list) 6397 list = XVaCreateNestedList(0, XNPreeditAttributes, preedit_list, 6398 NULL); 6399 else if (status_list) 6400 list = XVaCreateNestedList(0, XNStatusAttributes, status_list, 6401 NULL); 6402 else 6403 list = NULL; 6404 6405 if (list) 6406 { 6407 XSetICValues(xic, XNVaNestedList, list, NULL); 6408 XFree(list); 6409 } 6410 if (status_list) 6411 XFree(status_list); 6412 if (preedit_list) 6413 XFree(preedit_list); 6414 } 6415 6416 int 6417 xim_get_status_area_height(void) 6418 { 6419 if (status_area_enabled) 6420 return gui.char_height; 6421 return 0; 6422 } 6423 # endif 6424 6425 #else /* !defined(FEAT_XIM) */ 6426 6427 # ifdef IME_WITHOUT_XIM 6428 static int im_was_set_active = FALSE; 6429 6430 int 6431 im_get_status(void) 6432 { 6433 # if defined(FEAT_EVAL) 6434 if (USE_IMSTATUSFUNC) 6435 return call_imstatusfunc(); 6436 # endif 6437 return im_was_set_active; 6438 } 6439 6440 void 6441 im_set_active(int active_arg) 6442 { 6443 # if defined(FEAT_EVAL) 6444 int active = !p_imdisable && active_arg; 6445 6446 if (USE_IMACTIVATEFUNC && active != im_get_status()) 6447 { 6448 call_imactivatefunc(active); 6449 im_was_set_active = active; 6450 } 6451 # endif 6452 } 6453 6454 # ifdef FEAT_GUI 6455 void 6456 im_set_position(int row UNUSED, int col UNUSED) 6457 { 6458 } 6459 # endif 6460 # endif 6461 6462 #endif /* FEAT_XIM */ 6463 6464 6465 /* 6466 * Setup "vcp" for conversion from "from" to "to". 6467 * The names must have been made canonical with enc_canonize(). 6468 * vcp->vc_type must have been initialized to CONV_NONE. 6469 * Note: cannot be used for conversion from/to ucs-2 and ucs-4 (will use utf-8 6470 * instead). 6471 * Afterwards invoke with "from" and "to" equal to NULL to cleanup. 6472 * Return FAIL when conversion is not supported, OK otherwise. 6473 */ 6474 int 6475 convert_setup(vimconv_T *vcp, char_u *from, char_u *to) 6476 { 6477 return convert_setup_ext(vcp, from, TRUE, to, TRUE); 6478 } 6479 6480 /* 6481 * As convert_setup(), but only when from_unicode_is_utf8 is TRUE will all 6482 * "from" unicode charsets be considered utf-8. Same for "to". 6483 */ 6484 int 6485 convert_setup_ext( 6486 vimconv_T *vcp, 6487 char_u *from, 6488 int from_unicode_is_utf8, 6489 char_u *to, 6490 int to_unicode_is_utf8) 6491 { 6492 int from_prop; 6493 int to_prop; 6494 int from_is_utf8; 6495 int to_is_utf8; 6496 6497 /* Reset to no conversion. */ 6498 #ifdef USE_ICONV 6499 if (vcp->vc_type == CONV_ICONV && vcp->vc_fd != (iconv_t)-1) 6500 iconv_close(vcp->vc_fd); 6501 #endif 6502 vcp->vc_type = CONV_NONE; 6503 vcp->vc_factor = 1; 6504 vcp->vc_fail = FALSE; 6505 6506 /* No conversion when one of the names is empty or they are equal. */ 6507 if (from == NULL || *from == NUL || to == NULL || *to == NUL 6508 || STRCMP(from, to) == 0) 6509 return OK; 6510 6511 from_prop = enc_canon_props(from); 6512 to_prop = enc_canon_props(to); 6513 if (from_unicode_is_utf8) 6514 from_is_utf8 = from_prop & ENC_UNICODE; 6515 else 6516 from_is_utf8 = from_prop == ENC_UNICODE; 6517 if (to_unicode_is_utf8) 6518 to_is_utf8 = to_prop & ENC_UNICODE; 6519 else 6520 to_is_utf8 = to_prop == ENC_UNICODE; 6521 6522 if ((from_prop & ENC_LATIN1) && to_is_utf8) 6523 { 6524 /* Internal latin1 -> utf-8 conversion. */ 6525 vcp->vc_type = CONV_TO_UTF8; 6526 vcp->vc_factor = 2; /* up to twice as long */ 6527 } 6528 else if ((from_prop & ENC_LATIN9) && to_is_utf8) 6529 { 6530 /* Internal latin9 -> utf-8 conversion. */ 6531 vcp->vc_type = CONV_9_TO_UTF8; 6532 vcp->vc_factor = 3; /* up to three as long (euro sign) */ 6533 } 6534 else if (from_is_utf8 && (to_prop & ENC_LATIN1)) 6535 { 6536 /* Internal utf-8 -> latin1 conversion. */ 6537 vcp->vc_type = CONV_TO_LATIN1; 6538 } 6539 else if (from_is_utf8 && (to_prop & ENC_LATIN9)) 6540 { 6541 /* Internal utf-8 -> latin9 conversion. */ 6542 vcp->vc_type = CONV_TO_LATIN9; 6543 } 6544 #ifdef MSWIN 6545 /* Win32-specific codepage <-> codepage conversion without iconv. */ 6546 else if ((from_is_utf8 || encname2codepage(from) > 0) 6547 && (to_is_utf8 || encname2codepage(to) > 0)) 6548 { 6549 vcp->vc_type = CONV_CODEPAGE; 6550 vcp->vc_factor = 2; /* up to twice as long */ 6551 vcp->vc_cpfrom = from_is_utf8 ? 0 : encname2codepage(from); 6552 vcp->vc_cpto = to_is_utf8 ? 0 : encname2codepage(to); 6553 } 6554 #endif 6555 #ifdef MACOS_CONVERT 6556 else if ((from_prop & ENC_MACROMAN) && (to_prop & ENC_LATIN1)) 6557 { 6558 vcp->vc_type = CONV_MAC_LATIN1; 6559 } 6560 else if ((from_prop & ENC_MACROMAN) && to_is_utf8) 6561 { 6562 vcp->vc_type = CONV_MAC_UTF8; 6563 vcp->vc_factor = 2; /* up to twice as long */ 6564 } 6565 else if ((from_prop & ENC_LATIN1) && (to_prop & ENC_MACROMAN)) 6566 { 6567 vcp->vc_type = CONV_LATIN1_MAC; 6568 } 6569 else if (from_is_utf8 && (to_prop & ENC_MACROMAN)) 6570 { 6571 vcp->vc_type = CONV_UTF8_MAC; 6572 } 6573 #endif 6574 #ifdef USE_ICONV 6575 else 6576 { 6577 /* Use iconv() for conversion. */ 6578 vcp->vc_fd = (iconv_t)my_iconv_open( 6579 to_is_utf8 ? (char_u *)"utf-8" : to, 6580 from_is_utf8 ? (char_u *)"utf-8" : from); 6581 if (vcp->vc_fd != (iconv_t)-1) 6582 { 6583 vcp->vc_type = CONV_ICONV; 6584 vcp->vc_factor = 4; /* could be longer too... */ 6585 } 6586 } 6587 #endif 6588 if (vcp->vc_type == CONV_NONE) 6589 return FAIL; 6590 6591 return OK; 6592 } 6593 6594 #if defined(FEAT_GUI) || defined(AMIGA) || defined(MSWIN) \ 6595 || defined(PROTO) 6596 /* 6597 * Do conversion on typed input characters in-place. 6598 * The input and output are not NUL terminated! 6599 * Returns the length after conversion. 6600 */ 6601 int 6602 convert_input(char_u *ptr, int len, int maxlen) 6603 { 6604 return convert_input_safe(ptr, len, maxlen, NULL, NULL); 6605 } 6606 #endif 6607 6608 /* 6609 * Like convert_input(), but when there is an incomplete byte sequence at the 6610 * end return that as an allocated string in "restp" and set "*restlenp" to 6611 * the length. If "restp" is NULL it is not used. 6612 */ 6613 int 6614 convert_input_safe( 6615 char_u *ptr, 6616 int len, 6617 int maxlen, 6618 char_u **restp, 6619 int *restlenp) 6620 { 6621 char_u *d; 6622 int dlen = len; 6623 int unconvertlen = 0; 6624 6625 d = string_convert_ext(&input_conv, ptr, &dlen, 6626 restp == NULL ? NULL : &unconvertlen); 6627 if (d != NULL) 6628 { 6629 if (dlen <= maxlen) 6630 { 6631 if (unconvertlen > 0) 6632 { 6633 /* Move the unconverted characters to allocated memory. */ 6634 *restp = alloc(unconvertlen); 6635 if (*restp != NULL) 6636 mch_memmove(*restp, ptr + len - unconvertlen, unconvertlen); 6637 *restlenp = unconvertlen; 6638 } 6639 mch_memmove(ptr, d, dlen); 6640 } 6641 else 6642 /* result is too long, keep the unconverted text (the caller must 6643 * have done something wrong!) */ 6644 dlen = len; 6645 vim_free(d); 6646 } 6647 return dlen; 6648 } 6649 6650 /* 6651 * Convert text "ptr[*lenp]" according to "vcp". 6652 * Returns the result in allocated memory and sets "*lenp". 6653 * When "lenp" is NULL, use NUL terminated strings. 6654 * Illegal chars are often changed to "?", unless vcp->vc_fail is set. 6655 * When something goes wrong, NULL is returned and "*lenp" is unchanged. 6656 */ 6657 char_u * 6658 string_convert( 6659 vimconv_T *vcp, 6660 char_u *ptr, 6661 int *lenp) 6662 { 6663 return string_convert_ext(vcp, ptr, lenp, NULL); 6664 } 6665 6666 /* 6667 * Like string_convert(), but when "unconvlenp" is not NULL and there are is 6668 * an incomplete sequence at the end it is not converted and "*unconvlenp" is 6669 * set to the number of remaining bytes. 6670 */ 6671 char_u * 6672 string_convert_ext( 6673 vimconv_T *vcp, 6674 char_u *ptr, 6675 int *lenp, 6676 int *unconvlenp) 6677 { 6678 char_u *retval = NULL; 6679 char_u *d; 6680 int len; 6681 int i; 6682 int l; 6683 int c; 6684 6685 if (lenp == NULL) 6686 len = (int)STRLEN(ptr); 6687 else 6688 len = *lenp; 6689 if (len == 0) 6690 return vim_strsave((char_u *)""); 6691 6692 switch (vcp->vc_type) 6693 { 6694 case CONV_TO_UTF8: /* latin1 to utf-8 conversion */ 6695 retval = alloc(len * 2 + 1); 6696 if (retval == NULL) 6697 break; 6698 d = retval; 6699 for (i = 0; i < len; ++i) 6700 { 6701 c = ptr[i]; 6702 if (c < 0x80) 6703 *d++ = c; 6704 else 6705 { 6706 *d++ = 0xc0 + ((unsigned)c >> 6); 6707 *d++ = 0x80 + (c & 0x3f); 6708 } 6709 } 6710 *d = NUL; 6711 if (lenp != NULL) 6712 *lenp = (int)(d - retval); 6713 break; 6714 6715 case CONV_9_TO_UTF8: /* latin9 to utf-8 conversion */ 6716 retval = alloc(len * 3 + 1); 6717 if (retval == NULL) 6718 break; 6719 d = retval; 6720 for (i = 0; i < len; ++i) 6721 { 6722 c = ptr[i]; 6723 switch (c) 6724 { 6725 case 0xa4: c = 0x20ac; break; /* euro */ 6726 case 0xa6: c = 0x0160; break; /* S hat */ 6727 case 0xa8: c = 0x0161; break; /* S -hat */ 6728 case 0xb4: c = 0x017d; break; /* Z hat */ 6729 case 0xb8: c = 0x017e; break; /* Z -hat */ 6730 case 0xbc: c = 0x0152; break; /* OE */ 6731 case 0xbd: c = 0x0153; break; /* oe */ 6732 case 0xbe: c = 0x0178; break; /* Y */ 6733 } 6734 d += utf_char2bytes(c, d); 6735 } 6736 *d = NUL; 6737 if (lenp != NULL) 6738 *lenp = (int)(d - retval); 6739 break; 6740 6741 case CONV_TO_LATIN1: /* utf-8 to latin1 conversion */ 6742 case CONV_TO_LATIN9: /* utf-8 to latin9 conversion */ 6743 retval = alloc(len + 1); 6744 if (retval == NULL) 6745 break; 6746 d = retval; 6747 for (i = 0; i < len; ++i) 6748 { 6749 l = utf_ptr2len_len(ptr + i, len - i); 6750 if (l == 0) 6751 *d++ = NUL; 6752 else if (l == 1) 6753 { 6754 int l_w = utf8len_tab_zero[ptr[i]]; 6755 6756 if (l_w == 0) 6757 { 6758 /* Illegal utf-8 byte cannot be converted */ 6759 vim_free(retval); 6760 return NULL; 6761 } 6762 if (unconvlenp != NULL && l_w > len - i) 6763 { 6764 /* Incomplete sequence at the end. */ 6765 *unconvlenp = len - i; 6766 break; 6767 } 6768 *d++ = ptr[i]; 6769 } 6770 else 6771 { 6772 c = utf_ptr2char(ptr + i); 6773 if (vcp->vc_type == CONV_TO_LATIN9) 6774 switch (c) 6775 { 6776 case 0x20ac: c = 0xa4; break; /* euro */ 6777 case 0x0160: c = 0xa6; break; /* S hat */ 6778 case 0x0161: c = 0xa8; break; /* S -hat */ 6779 case 0x017d: c = 0xb4; break; /* Z hat */ 6780 case 0x017e: c = 0xb8; break; /* Z -hat */ 6781 case 0x0152: c = 0xbc; break; /* OE */ 6782 case 0x0153: c = 0xbd; break; /* oe */ 6783 case 0x0178: c = 0xbe; break; /* Y */ 6784 case 0xa4: 6785 case 0xa6: 6786 case 0xa8: 6787 case 0xb4: 6788 case 0xb8: 6789 case 0xbc: 6790 case 0xbd: 6791 case 0xbe: c = 0x100; break; /* not in latin9 */ 6792 } 6793 if (!utf_iscomposing(c)) /* skip composing chars */ 6794 { 6795 if (c < 0x100) 6796 *d++ = c; 6797 else if (vcp->vc_fail) 6798 { 6799 vim_free(retval); 6800 return NULL; 6801 } 6802 else 6803 { 6804 *d++ = 0xbf; 6805 if (utf_char2cells(c) > 1) 6806 *d++ = '?'; 6807 } 6808 } 6809 i += l - 1; 6810 } 6811 } 6812 *d = NUL; 6813 if (lenp != NULL) 6814 *lenp = (int)(d - retval); 6815 break; 6816 6817 # ifdef MACOS_CONVERT 6818 case CONV_MAC_LATIN1: 6819 retval = mac_string_convert(ptr, len, lenp, vcp->vc_fail, 6820 'm', 'l', unconvlenp); 6821 break; 6822 6823 case CONV_LATIN1_MAC: 6824 retval = mac_string_convert(ptr, len, lenp, vcp->vc_fail, 6825 'l', 'm', unconvlenp); 6826 break; 6827 6828 case CONV_MAC_UTF8: 6829 retval = mac_string_convert(ptr, len, lenp, vcp->vc_fail, 6830 'm', 'u', unconvlenp); 6831 break; 6832 6833 case CONV_UTF8_MAC: 6834 retval = mac_string_convert(ptr, len, lenp, vcp->vc_fail, 6835 'u', 'm', unconvlenp); 6836 break; 6837 # endif 6838 6839 # ifdef USE_ICONV 6840 case CONV_ICONV: /* conversion with output_conv.vc_fd */ 6841 retval = iconv_string(vcp, ptr, len, unconvlenp, lenp); 6842 break; 6843 # endif 6844 # ifdef MSWIN 6845 case CONV_CODEPAGE: /* codepage -> codepage */ 6846 { 6847 int retlen; 6848 int tmp_len; 6849 short_u *tmp; 6850 6851 /* 1. codepage/UTF-8 -> ucs-2. */ 6852 if (vcp->vc_cpfrom == 0) 6853 tmp_len = utf8_to_utf16(ptr, len, NULL, NULL); 6854 else 6855 { 6856 tmp_len = MultiByteToWideChar(vcp->vc_cpfrom, 6857 unconvlenp ? MB_ERR_INVALID_CHARS : 0, 6858 (char *)ptr, len, 0, 0); 6859 if (tmp_len == 0 6860 && GetLastError() == ERROR_NO_UNICODE_TRANSLATION) 6861 { 6862 if (lenp != NULL) 6863 *lenp = 0; 6864 if (unconvlenp != NULL) 6865 *unconvlenp = len; 6866 retval = alloc(1); 6867 if (retval) 6868 retval[0] = NUL; 6869 return retval; 6870 } 6871 } 6872 tmp = (short_u *)alloc(sizeof(short_u) * tmp_len); 6873 if (tmp == NULL) 6874 break; 6875 if (vcp->vc_cpfrom == 0) 6876 utf8_to_utf16(ptr, len, tmp, unconvlenp); 6877 else 6878 MultiByteToWideChar(vcp->vc_cpfrom, 0, 6879 (char *)ptr, len, tmp, tmp_len); 6880 6881 /* 2. ucs-2 -> codepage/UTF-8. */ 6882 if (vcp->vc_cpto == 0) 6883 retlen = utf16_to_utf8(tmp, tmp_len, NULL); 6884 else 6885 retlen = WideCharToMultiByte(vcp->vc_cpto, 0, 6886 tmp, tmp_len, 0, 0, 0, 0); 6887 retval = alloc(retlen + 1); 6888 if (retval != NULL) 6889 { 6890 if (vcp->vc_cpto == 0) 6891 utf16_to_utf8(tmp, tmp_len, retval); 6892 else 6893 WideCharToMultiByte(vcp->vc_cpto, 0, 6894 tmp, tmp_len, 6895 (char *)retval, retlen, 0, 0); 6896 retval[retlen] = NUL; 6897 if (lenp != NULL) 6898 *lenp = retlen; 6899 } 6900 vim_free(tmp); 6901 break; 6902 } 6903 # endif 6904 } 6905 6906 return retval; 6907 } 6908