Mercurial > vim
annotate src/charset.c @ 5976:d18c63fd0c48
Added tag v7-4-328 for changeset 01d9ffdd6e6f
| author | Bram Moolenaar <bram@vim.org> |
|---|---|
| date | Tue, 17 Jun 2014 13:50:13 +0200 |
| parents | 50dbef5e774a |
| children | ef83b423ebf7 |
| rev | line source |
|---|---|
| 7 | 1 /* vi:set ts=8 sts=4 sw=4: |
| 2 * | |
| 3 * VIM - Vi IMproved by Bram Moolenaar | |
| 4 * | |
| 5 * Do ":help uganda" in Vim to read copying and usage conditions. | |
| 6 * Do ":help credits" in Vim to see a list of people who contributed. | |
| 7 * See README.txt for an overview of the Vim source code. | |
| 8 */ | |
| 9 | |
| 10 #include "vim.h" | |
| 11 | |
| 12 #ifdef FEAT_LINEBREAK | |
| 13 static int win_chartabsize __ARGS((win_T *wp, char_u *p, colnr_T col)); | |
| 14 #endif | |
| 15 | |
| 16 #ifdef FEAT_MBYTE | |
|
2541
8a156630208b
Include wchar.h in charset.c for towupper().
Bram Moolenaar <bram@vim.org>
parents:
2339
diff
changeset
|
17 # if defined(HAVE_WCHAR_H) |
|
8a156630208b
Include wchar.h in charset.c for towupper().
Bram Moolenaar <bram@vim.org>
parents:
2339
diff
changeset
|
18 # include <wchar.h> /* for towupper() and towlower() */ |
|
8a156630208b
Include wchar.h in charset.c for towupper().
Bram Moolenaar <bram@vim.org>
parents:
2339
diff
changeset
|
19 # endif |
| 7 | 20 static int win_nolbr_chartabsize __ARGS((win_T *wp, char_u *s, colnr_T col, int *headp)); |
| 21 #endif | |
| 22 | |
| 1869 | 23 static unsigned nr2hex __ARGS((unsigned c)); |
| 7 | 24 |
| 25 static int chartab_initialized = FALSE; | |
| 26 | |
| 27 /* b_chartab[] is an array of 32 bytes, each bit representing one of the | |
| 28 * characters 0-255. */ | |
| 29 #define SET_CHARTAB(buf, c) (buf)->b_chartab[(unsigned)(c) >> 3] |= (1 << ((c) & 0x7)) | |
| 30 #define RESET_CHARTAB(buf, c) (buf)->b_chartab[(unsigned)(c) >> 3] &= ~(1 << ((c) & 0x7)) | |
| 31 #define GET_CHARTAB(buf, c) ((buf)->b_chartab[(unsigned)(c) >> 3] & (1 << ((c) & 0x7))) | |
| 32 | |
| 33 /* | |
| 34 * Fill chartab[]. Also fills curbuf->b_chartab[] with flags for keyword | |
| 35 * characters for current buffer. | |
| 36 * | |
| 37 * Depends on the option settings 'iskeyword', 'isident', 'isfname', | |
| 38 * 'isprint' and 'encoding'. | |
| 39 * | |
| 40 * The index in chartab[] depends on 'encoding': | |
| 41 * - For non-multi-byte index with the byte (same as the character). | |
| 42 * - For DBCS index with the first byte. | |
| 43 * - For UTF-8 index with the character (when first byte is up to 0x80 it is | |
| 44 * the same as the character, if the first byte is 0x80 and above it depends | |
| 45 * on further bytes). | |
| 46 * | |
| 47 * The contents of chartab[]: | |
| 48 * - The lower two bits, masked by CT_CELL_MASK, give the number of display | |
| 49 * cells the character occupies (1 or 2). Not valid for UTF-8 above 0x80. | |
| 50 * - CT_PRINT_CHAR bit is set when the character is printable (no need to | |
| 51 * translate the character before displaying it). Note that only DBCS | |
| 52 * characters can have 2 display cells and still be printable. | |
| 53 * - CT_FNAME_CHAR bit is set when the character can be in a file name. | |
| 54 * - CT_ID_CHAR bit is set when the character can be in an identifier. | |
| 55 * | |
| 56 * Return FAIL if 'iskeyword', 'isident', 'isfname' or 'isprint' option has an | |
| 57 * error, OK otherwise. | |
| 58 */ | |
| 59 int | |
| 60 init_chartab() | |
| 61 { | |
| 62 return buf_init_chartab(curbuf, TRUE); | |
| 63 } | |
| 64 | |
| 65 int | |
| 66 buf_init_chartab(buf, global) | |
| 67 buf_T *buf; | |
| 68 int global; /* FALSE: only set buf->b_chartab[] */ | |
| 69 { | |
| 70 int c; | |
| 71 int c2; | |
| 72 char_u *p; | |
| 73 int i; | |
| 74 int tilde; | |
| 75 int do_isalpha; | |
| 76 | |
| 77 if (global) | |
| 78 { | |
| 79 /* | |
| 80 * Set the default size for printable characters: | |
| 81 * From <Space> to '~' is 1 (printable), others are 2 (not printable). | |
| 82 * This also inits all 'isident' and 'isfname' flags to FALSE. | |
| 83 * | |
| 84 * EBCDIC: all chars below ' ' are not printable, all others are | |
| 85 * printable. | |
| 86 */ | |
| 87 c = 0; | |
| 88 while (c < ' ') | |
| 89 chartab[c++] = (dy_flags & DY_UHEX) ? 4 : 2; | |
| 90 #ifdef EBCDIC | |
| 91 while (c < 255) | |
| 92 #else | |
| 93 while (c <= '~') | |
| 94 #endif | |
| 95 chartab[c++] = 1 + CT_PRINT_CHAR; | |
| 96 #ifdef FEAT_FKMAP | |
| 97 if (p_altkeymap) | |
| 98 { | |
| 99 while (c < YE) | |
| 100 chartab[c++] = 1 + CT_PRINT_CHAR; | |
| 101 } | |
| 102 #endif | |
| 103 while (c < 256) | |
| 104 { | |
| 105 #ifdef FEAT_MBYTE | |
| 106 /* UTF-8: bytes 0xa0 - 0xff are printable (latin1) */ | |
| 107 if (enc_utf8 && c >= 0xa0) | |
| 108 chartab[c++] = CT_PRINT_CHAR + 1; | |
| 109 /* euc-jp characters starting with 0x8e are single width */ | |
| 110 else if (enc_dbcs == DBCS_JPNU && c == 0x8e) | |
| 111 chartab[c++] = CT_PRINT_CHAR + 1; | |
| 112 /* other double-byte chars can be printable AND double-width */ | |
| 113 else if (enc_dbcs != 0 && MB_BYTE2LEN(c) == 2) | |
| 114 chartab[c++] = CT_PRINT_CHAR + 2; | |
| 115 else | |
| 116 #endif | |
| 117 /* the rest is unprintable by default */ | |
| 118 chartab[c++] = (dy_flags & DY_UHEX) ? 4 : 2; | |
| 119 } | |
| 120 | |
| 121 #ifdef FEAT_MBYTE | |
| 122 /* Assume that every multi-byte char is a filename character. */ | |
| 123 for (c = 1; c < 256; ++c) | |
| 124 if ((enc_dbcs != 0 && MB_BYTE2LEN(c) > 1) | |
| 125 || (enc_dbcs == DBCS_JPNU && c == 0x8e) | |
| 126 || (enc_utf8 && c >= 0xa0)) | |
| 127 chartab[c] |= CT_FNAME_CHAR; | |
| 128 #endif | |
| 129 } | |
| 130 | |
| 131 /* | |
| 132 * Init word char flags all to FALSE | |
| 133 */ | |
| 134 vim_memset(buf->b_chartab, 0, (size_t)32); | |
| 135 #ifdef FEAT_MBYTE | |
| 227 | 136 if (enc_dbcs != 0) |
| 137 for (c = 0; c < 256; ++c) | |
| 138 { | |
| 139 /* double-byte characters are probably word characters */ | |
| 140 if (MB_BYTE2LEN(c) == 2) | |
| 141 SET_CHARTAB(buf, c); | |
| 142 } | |
| 7 | 143 #endif |
| 144 | |
| 145 #ifdef FEAT_LISP | |
| 146 /* | |
| 147 * In lisp mode the '-' character is included in keywords. | |
| 148 */ | |
| 149 if (buf->b_p_lisp) | |
| 150 SET_CHARTAB(buf, '-'); | |
| 151 #endif | |
| 152 | |
| 153 /* Walk through the 'isident', 'iskeyword', 'isfname' and 'isprint' | |
| 154 * options Each option is a list of characters, character numbers or | |
| 155 * ranges, separated by commas, e.g.: "200-210,x,#-178,-" | |
| 156 */ | |
| 157 for (i = global ? 0 : 3; i <= 3; ++i) | |
| 158 { | |
| 159 if (i == 0) | |
| 160 p = p_isi; /* first round: 'isident' */ | |
| 161 else if (i == 1) | |
| 162 p = p_isp; /* second round: 'isprint' */ | |
| 163 else if (i == 2) | |
| 164 p = p_isf; /* third round: 'isfname' */ | |
| 165 else /* i == 3 */ | |
| 166 p = buf->b_p_isk; /* fourth round: 'iskeyword' */ | |
| 167 | |
| 168 while (*p) | |
| 169 { | |
| 170 tilde = FALSE; | |
| 171 do_isalpha = FALSE; | |
| 172 if (*p == '^' && p[1] != NUL) | |
| 173 { | |
| 174 tilde = TRUE; | |
| 175 ++p; | |
| 176 } | |
| 177 if (VIM_ISDIGIT(*p)) | |
| 178 c = getdigits(&p); | |
| 179 else | |
| 1955 | 180 #ifdef FEAT_MBYTE |
| 181 if (has_mbyte) | |
| 182 c = mb_ptr2char_adv(&p); | |
| 183 else | |
| 184 #endif | |
| 7 | 185 c = *p++; |
| 186 c2 = -1; | |
| 187 if (*p == '-' && p[1] != NUL) | |
| 188 { | |
| 189 ++p; | |
| 190 if (VIM_ISDIGIT(*p)) | |
| 191 c2 = getdigits(&p); | |
| 192 else | |
| 1979 | 193 #ifdef FEAT_MBYTE |
| 194 if (has_mbyte) | |
| 195 c2 = mb_ptr2char_adv(&p); | |
| 196 else | |
| 197 #endif | |
| 7 | 198 c2 = *p++; |
| 199 } | |
| 1979 | 200 if (c <= 0 || c >= 256 || (c2 < c && c2 != -1) || c2 >= 256 |
| 7 | 201 || !(*p == NUL || *p == ',')) |
| 202 return FAIL; | |
| 203 | |
| 204 if (c2 == -1) /* not a range */ | |
| 205 { | |
| 206 /* | |
| 207 * A single '@' (not "@-@"): | |
| 208 * Decide on letters being ID/printable/keyword chars with | |
| 209 * standard function isalpha(). This takes care of locale for | |
| 210 * single-byte characters). | |
| 211 */ | |
| 212 if (c == '@') | |
| 213 { | |
| 214 do_isalpha = TRUE; | |
| 215 c = 1; | |
| 216 c2 = 255; | |
| 217 } | |
| 218 else | |
| 219 c2 = c; | |
| 220 } | |
| 221 while (c <= c2) | |
| 222 { | |
| 1365 | 223 /* Use the MB_ functions here, because isalpha() doesn't |
| 224 * work properly when 'encoding' is "latin1" and the locale is | |
| 225 * "C". */ | |
| 226 if (!do_isalpha || MB_ISLOWER(c) || MB_ISUPPER(c) | |
| 7 | 227 #ifdef FEAT_FKMAP |
| 228 || (p_altkeymap && (F_isalpha(c) || F_isdigit(c))) | |
| 229 #endif | |
| 230 ) | |
| 231 { | |
| 232 if (i == 0) /* (re)set ID flag */ | |
| 233 { | |
| 234 if (tilde) | |
| 235 chartab[c] &= ~CT_ID_CHAR; | |
| 236 else | |
| 237 chartab[c] |= CT_ID_CHAR; | |
| 238 } | |
| 239 else if (i == 1) /* (re)set printable */ | |
| 240 { | |
| 241 if ((c < ' ' | |
| 242 #ifndef EBCDIC | |
| 243 || c > '~' | |
| 244 #endif | |
| 245 #ifdef FEAT_FKMAP | |
| 246 || (p_altkeymap | |
| 247 && (F_isalpha(c) || F_isdigit(c))) | |
| 248 #endif | |
| 249 ) | |
| 250 #ifdef FEAT_MBYTE | |
| 251 /* For double-byte we keep the cell width, so | |
| 252 * that we can detect it from the first byte. */ | |
| 253 && !(enc_dbcs && MB_BYTE2LEN(c) == 2) | |
| 254 #endif | |
| 255 ) | |
| 256 { | |
| 257 if (tilde) | |
| 258 { | |
| 259 chartab[c] = (chartab[c] & ~CT_CELL_MASK) | |
| 260 + ((dy_flags & DY_UHEX) ? 4 : 2); | |
| 261 chartab[c] &= ~CT_PRINT_CHAR; | |
| 262 } | |
| 263 else | |
| 264 { | |
| 265 chartab[c] = (chartab[c] & ~CT_CELL_MASK) + 1; | |
| 266 chartab[c] |= CT_PRINT_CHAR; | |
| 267 } | |
| 268 } | |
| 269 } | |
| 270 else if (i == 2) /* (re)set fname flag */ | |
| 271 { | |
| 272 if (tilde) | |
| 273 chartab[c] &= ~CT_FNAME_CHAR; | |
| 274 else | |
| 275 chartab[c] |= CT_FNAME_CHAR; | |
| 276 } | |
| 277 else /* i == 3 */ /* (re)set keyword flag */ | |
| 278 { | |
| 279 if (tilde) | |
| 280 RESET_CHARTAB(buf, c); | |
| 281 else | |
| 282 SET_CHARTAB(buf, c); | |
| 283 } | |
| 284 } | |
| 285 ++c; | |
| 286 } | |
| 4096 | 287 |
| 288 c = *p; | |
| 7 | 289 p = skip_to_option_part(p); |
| 4096 | 290 if (c == ',' && *p == NUL) |
| 291 /* Trailing comma is not allowed. */ | |
| 292 return FAIL; | |
| 7 | 293 } |
| 294 } | |
| 295 chartab_initialized = TRUE; | |
| 296 return OK; | |
| 297 } | |
| 298 | |
| 299 /* | |
| 300 * Translate any special characters in buf[bufsize] in-place. | |
| 301 * The result is a string with only printable characters, but if there is not | |
| 302 * enough room, not all characters will be translated. | |
| 303 */ | |
| 304 void | |
| 305 trans_characters(buf, bufsize) | |
| 306 char_u *buf; | |
| 307 int bufsize; | |
| 308 { | |
| 309 int len; /* length of string needing translation */ | |
| 310 int room; /* room in buffer after string */ | |
| 311 char_u *trs; /* translated character */ | |
| 312 int trs_len; /* length of trs[] */ | |
| 313 | |
| 314 len = (int)STRLEN(buf); | |
| 315 room = bufsize - len; | |
| 316 while (*buf != 0) | |
| 317 { | |
| 318 # ifdef FEAT_MBYTE | |
| 319 /* Assume a multi-byte character doesn't need translation. */ | |
| 474 | 320 if (has_mbyte && (trs_len = (*mb_ptr2len)(buf)) > 1) |
| 7 | 321 len -= trs_len; |
| 322 else | |
| 323 # endif | |
| 324 { | |
| 325 trs = transchar_byte(*buf); | |
| 326 trs_len = (int)STRLEN(trs); | |
| 327 if (trs_len > 1) | |
| 328 { | |
| 329 room -= trs_len - 1; | |
| 330 if (room <= 0) | |
| 331 return; | |
| 332 mch_memmove(buf + trs_len, buf + 1, (size_t)len); | |
| 333 } | |
| 334 mch_memmove(buf, trs, (size_t)trs_len); | |
| 335 --len; | |
| 336 } | |
| 337 buf += trs_len; | |
| 338 } | |
| 339 } | |
| 340 | |
| 1097 | 341 #if defined(FEAT_EVAL) || defined(FEAT_TITLE) || defined(FEAT_INS_EXPAND) \ |
| 342 || defined(PROTO) | |
| 7 | 343 /* |
| 344 * Translate a string into allocated memory, replacing special chars with | |
| 345 * printable chars. Returns NULL when out of memory. | |
| 346 */ | |
| 347 char_u * | |
| 348 transstr(s) | |
| 349 char_u *s; | |
| 350 { | |
| 351 char_u *res; | |
| 352 char_u *p; | |
| 353 #ifdef FEAT_MBYTE | |
| 354 int l, len, c; | |
| 355 char_u hexbuf[11]; | |
| 356 #endif | |
| 357 | |
| 358 #ifdef FEAT_MBYTE | |
| 359 if (has_mbyte) | |
| 360 { | |
| 361 /* Compute the length of the result, taking account of unprintable | |
| 362 * multi-byte characters. */ | |
| 363 len = 0; | |
| 364 p = s; | |
| 365 while (*p != NUL) | |
| 366 { | |
| 474 | 367 if ((l = (*mb_ptr2len)(p)) > 1) |
| 7 | 368 { |
| 369 c = (*mb_ptr2char)(p); | |
| 370 p += l; | |
| 371 if (vim_isprintc(c)) | |
| 372 len += l; | |
| 373 else | |
| 374 { | |
| 375 transchar_hex(hexbuf, c); | |
| 835 | 376 len += (int)STRLEN(hexbuf); |
| 7 | 377 } |
| 378 } | |
| 379 else | |
| 380 { | |
| 381 l = byte2cells(*p++); | |
| 382 if (l > 0) | |
| 383 len += l; | |
| 384 else | |
| 385 len += 4; /* illegal byte sequence */ | |
| 386 } | |
| 387 } | |
| 388 res = alloc((unsigned)(len + 1)); | |
| 389 } | |
| 390 else | |
| 391 #endif | |
| 392 res = alloc((unsigned)(vim_strsize(s) + 1)); | |
| 393 if (res != NULL) | |
| 394 { | |
| 395 *res = NUL; | |
| 396 p = s; | |
| 397 while (*p != NUL) | |
| 398 { | |
| 399 #ifdef FEAT_MBYTE | |
| 474 | 400 if (has_mbyte && (l = (*mb_ptr2len)(p)) > 1) |
| 7 | 401 { |
| 402 c = (*mb_ptr2char)(p); | |
| 403 if (vim_isprintc(c)) | |
| 404 STRNCAT(res, p, l); /* append printable multi-byte char */ | |
| 405 else | |
| 406 transchar_hex(res + STRLEN(res), c); | |
| 407 p += l; | |
| 408 } | |
| 409 else | |
| 410 #endif | |
| 411 STRCAT(res, transchar_byte(*p++)); | |
| 412 } | |
| 413 } | |
| 414 return res; | |
| 415 } | |
| 416 #endif | |
| 417 | |
| 418 #if defined(FEAT_SYN_HL) || defined(FEAT_INS_EXPAND) || defined(PROTO) | |
| 419 /* | |
| 221 | 420 * Convert the string "str[orglen]" to do ignore-case comparing. Uses the |
| 421 * current locale. | |
| 130 | 422 * When "buf" is NULL returns an allocated string (NULL for out-of-memory). |
| 423 * Otherwise puts the result in "buf[buflen]". | |
| 7 | 424 */ |
| 425 char_u * | |
| 130 | 426 str_foldcase(str, orglen, buf, buflen) |
| 7 | 427 char_u *str; |
| 130 | 428 int orglen; |
| 429 char_u *buf; | |
| 430 int buflen; | |
| 7 | 431 { |
| 432 garray_T ga; | |
| 433 int i; | |
| 130 | 434 int len = orglen; |
| 7 | 435 |
| 436 #define GA_CHAR(i) ((char_u *)ga.ga_data)[i] | |
| 437 #define GA_PTR(i) ((char_u *)ga.ga_data + i) | |
| 130 | 438 #define STR_CHAR(i) (buf == NULL ? GA_CHAR(i) : buf[i]) |
| 439 #define STR_PTR(i) (buf == NULL ? GA_PTR(i) : buf + i) | |
| 7 | 440 |
| 130 | 441 /* Copy "str" into "buf" or allocated memory, unmodified. */ |
| 442 if (buf == NULL) | |
| 443 { | |
| 444 ga_init2(&ga, 1, 10); | |
| 445 if (ga_grow(&ga, len + 1) == FAIL) | |
| 446 return NULL; | |
| 447 mch_memmove(ga.ga_data, str, (size_t)len); | |
| 448 ga.ga_len = len; | |
| 449 } | |
| 450 else | |
| 451 { | |
| 452 if (len >= buflen) /* Ugly! */ | |
| 453 len = buflen - 1; | |
| 454 mch_memmove(buf, str, (size_t)len); | |
| 455 } | |
| 456 if (buf == NULL) | |
| 457 GA_CHAR(len) = NUL; | |
| 458 else | |
| 459 buf[len] = NUL; | |
| 7 | 460 |
| 461 /* Make each character lower case. */ | |
| 462 i = 0; | |
| 130 | 463 while (STR_CHAR(i) != NUL) |
| 7 | 464 { |
| 465 #ifdef FEAT_MBYTE | |
| 130 | 466 if (enc_utf8 || (has_mbyte && MB_BYTE2LEN(STR_CHAR(i)) > 1)) |
| 7 | 467 { |
| 468 if (enc_utf8) | |
| 469 { | |
| 1654 | 470 int c = utf_ptr2char(STR_PTR(i)); |
| 3263 | 471 int olen = utf_ptr2len(STR_PTR(i)); |
| 1654 | 472 int lc = utf_tolower(c); |
| 7 | 473 |
| 1654 | 474 /* Only replace the character when it is not an invalid |
| 475 * sequence (ASCII character or more than one byte) and | |
| 476 * utf_tolower() doesn't return the original character. */ | |
| 3263 | 477 if ((c < 0x80 || olen > 1) && c != lc) |
| 7 | 478 { |
| 3263 | 479 int nlen = utf_char2len(lc); |
| 7 | 480 |
| 481 /* If the byte length changes need to shift the following | |
| 482 * characters forward or backward. */ | |
| 3263 | 483 if (olen != nlen) |
| 7 | 484 { |
| 3263 | 485 if (nlen > olen) |
| 130 | 486 { |
| 3263 | 487 if (buf == NULL |
| 488 ? ga_grow(&ga, nlen - olen + 1) == FAIL | |
| 489 : len + nlen - olen >= buflen) | |
| 7 | 490 { |
| 491 /* out of memory, keep old char */ | |
| 492 lc = c; | |
| 3263 | 493 nlen = olen; |
| 7 | 494 } |
| 130 | 495 } |
| 3263 | 496 if (olen != nlen) |
| 7 | 497 { |
| 130 | 498 if (buf == NULL) |
| 499 { | |
| 3263 | 500 STRMOVE(GA_PTR(i) + nlen, GA_PTR(i) + olen); |
| 501 ga.ga_len += nlen - olen; | |
| 130 | 502 } |
| 503 else | |
| 504 { | |
| 3263 | 505 STRMOVE(buf + i + nlen, buf + i + olen); |
| 506 len += nlen - olen; | |
| 130 | 507 } |
| 7 | 508 } |
| 509 } | |
| 130 | 510 (void)utf_char2bytes(lc, STR_PTR(i)); |
| 7 | 511 } |
| 512 } | |
| 513 /* skip to next multi-byte char */ | |
| 474 | 514 i += (*mb_ptr2len)(STR_PTR(i)); |
| 7 | 515 } |
| 516 else | |
| 517 #endif | |
| 518 { | |
| 130 | 519 if (buf == NULL) |
| 520 GA_CHAR(i) = TOLOWER_LOC(GA_CHAR(i)); | |
| 521 else | |
| 522 buf[i] = TOLOWER_LOC(buf[i]); | |
| 7 | 523 ++i; |
| 524 } | |
| 525 } | |
| 526 | |
| 130 | 527 if (buf == NULL) |
| 528 return (char_u *)ga.ga_data; | |
| 529 return buf; | |
| 7 | 530 } |
| 531 #endif | |
| 532 | |
| 533 /* | |
| 534 * Catch 22: chartab[] can't be initialized before the options are | |
| 535 * initialized, and initializing options may cause transchar() to be called! | |
| 536 * When chartab_initialized == FALSE don't use chartab[]. | |
| 537 * Does NOT work for multi-byte characters, c must be <= 255. | |
| 538 * Also doesn't work for the first byte of a multi-byte, "c" must be a | |
| 539 * character! | |
| 540 */ | |
| 541 static char_u transchar_buf[7]; | |
| 542 | |
| 543 char_u * | |
| 544 transchar(c) | |
| 545 int c; | |
| 546 { | |
| 547 int i; | |
| 548 | |
| 549 i = 0; | |
| 550 if (IS_SPECIAL(c)) /* special key code, display as ~@ char */ | |
| 551 { | |
| 552 transchar_buf[0] = '~'; | |
| 553 transchar_buf[1] = '@'; | |
| 554 i = 2; | |
| 555 c = K_SECOND(c); | |
| 556 } | |
| 557 | |
| 558 if ((!chartab_initialized && ( | |
| 559 #ifdef EBCDIC | |
| 560 (c >= 64 && c < 255) | |
| 561 #else | |
| 562 (c >= ' ' && c <= '~') | |
| 563 #endif | |
| 564 #ifdef FEAT_FKMAP | |
| 565 || F_ischar(c) | |
| 566 #endif | |
| 567 )) || (c < 256 && vim_isprintc_strict(c))) | |
| 568 { | |
| 569 /* printable character */ | |
| 570 transchar_buf[i] = c; | |
| 571 transchar_buf[i + 1] = NUL; | |
| 572 } | |
| 573 else | |
| 574 transchar_nonprint(transchar_buf + i, c); | |
| 575 return transchar_buf; | |
| 576 } | |
| 577 | |
| 578 #if defined(FEAT_MBYTE) || defined(PROTO) | |
| 579 /* | |
| 580 * Like transchar(), but called with a byte instead of a character. Checks | |
| 581 * for an illegal UTF-8 byte. | |
| 582 */ | |
| 583 char_u * | |
| 584 transchar_byte(c) | |
| 585 int c; | |
| 586 { | |
| 587 if (enc_utf8 && c >= 0x80) | |
| 588 { | |
| 589 transchar_nonprint(transchar_buf, c); | |
| 590 return transchar_buf; | |
| 591 } | |
| 592 return transchar(c); | |
| 593 } | |
| 594 #endif | |
| 595 | |
| 596 /* | |
| 597 * Convert non-printable character to two or more printable characters in | |
| 598 * "buf[]". "buf" needs to be able to hold five bytes. | |
| 599 * Does NOT work for multi-byte characters, c must be <= 255. | |
| 600 */ | |
| 601 void | |
| 602 transchar_nonprint(buf, c) | |
| 603 char_u *buf; | |
| 604 int c; | |
| 605 { | |
| 606 if (c == NL) | |
| 607 c = NUL; /* we use newline in place of a NUL */ | |
| 608 else if (c == CAR && get_fileformat(curbuf) == EOL_MAC) | |
| 609 c = NL; /* we use CR in place of NL in this case */ | |
| 610 | |
| 611 if (dy_flags & DY_UHEX) /* 'display' has "uhex" */ | |
| 612 transchar_hex(buf, c); | |
| 613 | |
| 614 #ifdef EBCDIC | |
| 615 /* For EBCDIC only the characters 0-63 and 255 are not printable */ | |
| 616 else if (CtrlChar(c) != 0 || c == DEL) | |
| 617 #else | |
| 618 else if (c <= 0x7f) /* 0x00 - 0x1f and 0x7f */ | |
| 619 #endif | |
| 620 { | |
| 621 buf[0] = '^'; | |
| 622 #ifdef EBCDIC | |
| 623 if (c == DEL) | |
| 624 buf[1] = '?'; /* DEL displayed as ^? */ | |
| 625 else | |
| 626 buf[1] = CtrlChar(c); | |
| 627 #else | |
| 628 buf[1] = c ^ 0x40; /* DEL displayed as ^? */ | |
| 629 #endif | |
| 630 | |
| 631 buf[2] = NUL; | |
| 632 } | |
| 633 #ifdef FEAT_MBYTE | |
| 634 else if (enc_utf8 && c >= 0x80) | |
| 635 { | |
| 636 transchar_hex(buf, c); | |
| 637 } | |
| 638 #endif | |
| 639 #ifndef EBCDIC | |
| 640 else if (c >= ' ' + 0x80 && c <= '~' + 0x80) /* 0xa0 - 0xfe */ | |
| 641 { | |
| 642 buf[0] = '|'; | |
| 643 buf[1] = c - 0x80; | |
| 644 buf[2] = NUL; | |
| 645 } | |
| 646 #else | |
| 647 else if (c < 64) | |
| 648 { | |
| 649 buf[0] = '~'; | |
| 650 buf[1] = MetaChar(c); | |
| 651 buf[2] = NUL; | |
| 652 } | |
| 653 #endif | |
| 654 else /* 0x80 - 0x9f and 0xff */ | |
| 655 { | |
| 656 /* | |
| 657 * TODO: EBCDIC I don't know what to do with this chars, so I display | |
| 658 * them as '~?' for now | |
| 659 */ | |
| 660 buf[0] = '~'; | |
| 661 #ifdef EBCDIC | |
| 662 buf[1] = '?'; /* 0xff displayed as ~? */ | |
| 663 #else | |
| 664 buf[1] = (c - 0x80) ^ 0x40; /* 0xff displayed as ~? */ | |
| 665 #endif | |
| 666 buf[2] = NUL; | |
| 667 } | |
| 668 } | |
| 669 | |
| 670 void | |
| 671 transchar_hex(buf, c) | |
| 672 char_u *buf; | |
| 673 int c; | |
| 674 { | |
| 675 int i = 0; | |
| 676 | |
| 677 buf[0] = '<'; | |
| 678 #ifdef FEAT_MBYTE | |
| 679 if (c > 255) | |
| 680 { | |
| 681 buf[++i] = nr2hex((unsigned)c >> 12); | |
| 682 buf[++i] = nr2hex((unsigned)c >> 8); | |
| 683 } | |
| 684 #endif | |
| 685 buf[++i] = nr2hex((unsigned)c >> 4); | |
| 1869 | 686 buf[++i] = nr2hex((unsigned)c); |
| 7 | 687 buf[++i] = '>'; |
| 688 buf[++i] = NUL; | |
| 689 } | |
| 690 | |
| 691 /* | |
| 692 * Convert the lower 4 bits of byte "c" to its hex character. | |
| 693 * Lower case letters are used to avoid the confusion of <F1> being 0xf1 or | |
| 694 * function key 1. | |
| 695 */ | |
| 1869 | 696 static unsigned |
| 7 | 697 nr2hex(c) |
| 1869 | 698 unsigned c; |
| 7 | 699 { |
| 700 if ((c & 0xf) <= 9) | |
| 701 return (c & 0xf) + '0'; | |
| 702 return (c & 0xf) - 10 + 'a'; | |
| 703 } | |
| 704 | |
| 705 /* | |
| 706 * Return number of display cells occupied by byte "b". | |
| 707 * Caller must make sure 0 <= b <= 255. | |
| 708 * For multi-byte mode "b" must be the first byte of a character. | |
| 709 * A TAB is counted as two cells: "^I". | |
| 710 * For UTF-8 mode this will return 0 for bytes >= 0x80, because the number of | |
| 711 * cells depends on further bytes. | |
| 712 */ | |
| 713 int | |
| 714 byte2cells(b) | |
| 715 int b; | |
| 716 { | |
| 717 #ifdef FEAT_MBYTE | |
| 718 if (enc_utf8 && b >= 0x80) | |
| 719 return 0; | |
| 720 #endif | |
| 721 return (chartab[b] & CT_CELL_MASK); | |
| 722 } | |
| 723 | |
| 724 /* | |
| 725 * Return number of display cells occupied by character "c". | |
| 726 * "c" can be a special key (negative number) in which case 3 or 4 is returned. | |
| 727 * A TAB is counted as two cells: "^I" or four: "<09>". | |
| 728 */ | |
| 729 int | |
| 730 char2cells(c) | |
| 731 int c; | |
| 732 { | |
| 733 if (IS_SPECIAL(c)) | |
| 734 return char2cells(K_SECOND(c)) + 2; | |
| 735 #ifdef FEAT_MBYTE | |
| 736 if (c >= 0x80) | |
| 737 { | |
| 738 /* UTF-8: above 0x80 need to check the value */ | |
| 739 if (enc_utf8) | |
| 740 return utf_char2cells(c); | |
| 741 /* DBCS: double-byte means double-width, except for euc-jp with first | |
| 742 * byte 0x8e */ | |
| 743 if (enc_dbcs != 0 && c >= 0x100) | |
| 744 { | |
| 745 if (enc_dbcs == DBCS_JPNU && ((unsigned)c >> 8) == 0x8e) | |
| 746 return 1; | |
| 747 return 2; | |
| 748 } | |
| 749 } | |
| 750 #endif | |
| 751 return (chartab[c & 0xff] & CT_CELL_MASK); | |
| 752 } | |
| 753 | |
| 754 /* | |
| 755 * Return number of display cells occupied by character at "*p". | |
| 756 * A TAB is counted as two cells: "^I" or four: "<09>". | |
| 757 */ | |
| 758 int | |
| 759 ptr2cells(p) | |
| 760 char_u *p; | |
| 761 { | |
| 762 #ifdef FEAT_MBYTE | |
| 763 /* For UTF-8 we need to look at more bytes if the first byte is >= 0x80. */ | |
| 764 if (enc_utf8 && *p >= 0x80) | |
| 765 return utf_ptr2cells(p); | |
| 766 /* For DBCS we can tell the cell count from the first byte. */ | |
| 767 #endif | |
| 768 return (chartab[*p] & CT_CELL_MASK); | |
| 769 } | |
| 770 | |
| 771 /* | |
| 3292 | 772 * Return the number of character cells string "s" will take on the screen, |
| 7 | 773 * counting TABs as two characters: "^I". |
| 774 */ | |
| 775 int | |
| 776 vim_strsize(s) | |
| 777 char_u *s; | |
| 778 { | |
| 779 return vim_strnsize(s, (int)MAXCOL); | |
| 780 } | |
| 781 | |
| 782 /* | |
| 3292 | 783 * Return the number of character cells string "s[len]" will take on the |
| 784 * screen, counting TABs as two characters: "^I". | |
| 7 | 785 */ |
| 786 int | |
| 787 vim_strnsize(s, len) | |
| 788 char_u *s; | |
| 789 int len; | |
| 790 { | |
| 791 int size = 0; | |
| 792 | |
| 793 while (*s != NUL && --len >= 0) | |
| 794 { | |
| 795 #ifdef FEAT_MBYTE | |
| 796 if (has_mbyte) | |
| 797 { | |
| 474 | 798 int l = (*mb_ptr2len)(s); |
| 7 | 799 |
| 800 size += ptr2cells(s); | |
| 801 s += l; | |
| 802 len -= l - 1; | |
| 803 } | |
| 804 else | |
| 805 #endif | |
| 806 size += byte2cells(*s++); | |
| 807 } | |
| 808 return size; | |
| 809 } | |
| 810 | |
| 811 /* | |
| 812 * Return the number of characters 'c' will take on the screen, taking | |
| 813 * into account the size of a tab. | |
| 814 * Use a define to make it fast, this is used very often!!! | |
| 815 * Also see getvcol() below. | |
| 816 */ | |
| 817 | |
| 818 #define RET_WIN_BUF_CHARTABSIZE(wp, buf, p, col) \ | |
| 819 if (*(p) == TAB && (!(wp)->w_p_list || lcs_tab1)) \ | |
| 820 { \ | |
| 821 int ts; \ | |
| 822 ts = (buf)->b_p_ts; \ | |
| 823 return (int)(ts - (col % ts)); \ | |
| 824 } \ | |
| 825 else \ | |
| 826 return ptr2cells(p); | |
| 827 | |
| 828 #if defined(FEAT_VREPLACE) || defined(FEAT_EX_EXTRA) || defined(FEAT_GUI) \ | |
| 829 || defined(FEAT_VIRTUALEDIT) || defined(PROTO) | |
| 830 int | |
| 831 chartabsize(p, col) | |
| 832 char_u *p; | |
| 833 colnr_T col; | |
| 834 { | |
| 835 RET_WIN_BUF_CHARTABSIZE(curwin, curbuf, p, col) | |
| 836 } | |
| 837 #endif | |
| 838 | |
| 839 #ifdef FEAT_LINEBREAK | |
| 840 static int | |
| 841 win_chartabsize(wp, p, col) | |
| 842 win_T *wp; | |
| 843 char_u *p; | |
| 844 colnr_T col; | |
| 845 { | |
| 846 RET_WIN_BUF_CHARTABSIZE(wp, wp->w_buffer, p, col) | |
| 847 } | |
| 848 #endif | |
| 849 | |
| 850 /* | |
|
2339
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
851 * Return the number of characters the string 's' will take on the screen, |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
852 * taking into account the size of a tab. |
| 7 | 853 */ |
| 854 int | |
| 855 linetabsize(s) | |
| 856 char_u *s; | |
| 857 { | |
|
2339
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
858 return linetabsize_col(0, s); |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
859 } |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
860 |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
861 /* |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
862 * Like linetabsize(), but starting at column "startcol". |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
863 */ |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
864 int |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
865 linetabsize_col(startcol, s) |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
866 int startcol; |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
867 char_u *s; |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
868 { |
|
01e4b4d37842
Added strdisplaywidth() function.
Bram Moolenaar <bram@vim.org>
parents:
2108
diff
changeset
|
869 colnr_T col = startcol; |
| 7 | 870 |
| 871 while (*s != NUL) | |
| 872 col += lbr_chartabsize_adv(&s, col); | |
| 873 return (int)col; | |
| 874 } | |
| 875 | |
| 876 /* | |
| 877 * Like linetabsize(), but for a given window instead of the current one. | |
| 878 */ | |
| 879 int | |
| 880 win_linetabsize(wp, p, len) | |
| 881 win_T *wp; | |
| 882 char_u *p; | |
| 883 colnr_T len; | |
| 884 { | |
| 885 colnr_T col = 0; | |
| 886 char_u *s; | |
| 887 | |
| 40 | 888 for (s = p; *s != NUL && (len == MAXCOL || s < p + len); mb_ptr_adv(s)) |
| 7 | 889 col += win_lbr_chartabsize(wp, s, col, NULL); |
| 890 return (int)col; | |
| 891 } | |
| 892 | |
| 893 /* | |
| 42 | 894 * Return TRUE if 'c' is a normal identifier character: |
| 895 * Letters and characters from the 'isident' option. | |
| 7 | 896 */ |
| 897 int | |
| 898 vim_isIDc(c) | |
| 899 int c; | |
| 900 { | |
| 901 return (c > 0 && c < 0x100 && (chartab[c] & CT_ID_CHAR)); | |
| 902 } | |
| 903 | |
| 904 /* | |
| 905 * return TRUE if 'c' is a keyword character: Letters and characters from | |
| 906 * 'iskeyword' option for current buffer. | |
| 907 * For multi-byte characters mb_get_class() is used (builtin rules). | |
| 908 */ | |
| 909 int | |
| 910 vim_iswordc(c) | |
| 911 int c; | |
| 912 { | |
| 4043 | 913 return vim_iswordc_buf(c, curbuf); |
| 914 } | |
| 915 | |
| 916 int | |
| 917 vim_iswordc_buf(c, buf) | |
| 918 int c; | |
| 919 buf_T *buf; | |
| 920 { | |
| 7 | 921 #ifdef FEAT_MBYTE |
| 922 if (c >= 0x100) | |
| 923 { | |
| 924 if (enc_dbcs != 0) | |
| 1869 | 925 return dbcs_class((unsigned)c >> 8, (unsigned)(c & 0xff)) >= 2; |
| 7 | 926 if (enc_utf8) |
| 927 return utf_class(c) >= 2; | |
| 928 } | |
| 929 #endif | |
| 4043 | 930 return (c > 0 && c < 0x100 && GET_CHARTAB(buf, c) != 0); |
| 7 | 931 } |
| 932 | |
| 933 /* | |
| 934 * Just like vim_iswordc() but uses a pointer to the (multi-byte) character. | |
| 935 */ | |
| 936 int | |
| 937 vim_iswordp(p) | |
| 938 char_u *p; | |
| 939 { | |
| 940 #ifdef FEAT_MBYTE | |
| 941 if (has_mbyte && MB_BYTE2LEN(*p) > 1) | |
| 942 return mb_get_class(p) >= 2; | |
| 943 #endif | |
| 944 return GET_CHARTAB(curbuf, *p) != 0; | |
| 945 } | |
| 946 | |
| 947 int | |
| 4043 | 948 vim_iswordp_buf(p, buf) |
| 7 | 949 char_u *p; |
| 950 buf_T *buf; | |
| 951 { | |
| 4080 | 952 #ifdef FEAT_MBYTE |
| 7 | 953 if (has_mbyte && MB_BYTE2LEN(*p) > 1) |
| 954 return mb_get_class(p) >= 2; | |
| 4080 | 955 #endif |
| 7 | 956 return (GET_CHARTAB(buf, *p) != 0); |
| 957 } | |
| 958 | |
| 959 /* | |
| 960 * return TRUE if 'c' is a valid file-name character | |
| 961 * Assume characters above 0x100 are valid (multi-byte). | |
| 962 */ | |
| 963 int | |
| 964 vim_isfilec(c) | |
| 965 int c; | |
| 966 { | |
| 967 return (c >= 0x100 || (c > 0 && (chartab[c] & CT_FNAME_CHAR))); | |
| 968 } | |
| 969 | |
| 970 /* | |
| 1369 | 971 * return TRUE if 'c' is a valid file-name character or a wildcard character |
| 972 * Assume characters above 0x100 are valid (multi-byte). | |
| 973 * Explicitly interpret ']' as a wildcard character as mch_has_wildcard("]") | |
| 974 * returns false. | |
| 975 */ | |
| 976 int | |
| 977 vim_isfilec_or_wc(c) | |
| 978 int c; | |
| 979 { | |
| 980 char_u buf[2]; | |
| 981 | |
| 982 buf[0] = (char_u)c; | |
| 983 buf[1] = NUL; | |
| 984 return vim_isfilec(c) || c == ']' || mch_has_wildcard(buf); | |
| 985 } | |
| 986 | |
| 987 /* | |
| 7 | 988 * return TRUE if 'c' is a printable character |
| 989 * Assume characters above 0x100 are printable (multi-byte), except for | |
| 990 * Unicode. | |
| 991 */ | |
| 992 int | |
| 993 vim_isprintc(c) | |
| 994 int c; | |
| 995 { | |
| 996 #ifdef FEAT_MBYTE | |
| 997 if (enc_utf8 && c >= 0x100) | |
| 998 return utf_printable(c); | |
| 999 #endif | |
| 1000 return (c >= 0x100 || (c > 0 && (chartab[c] & CT_PRINT_CHAR))); | |
| 1001 } | |
| 1002 | |
| 1003 /* | |
| 1004 * Strict version of vim_isprintc(c), don't return TRUE if "c" is the head | |
| 1005 * byte of a double-byte character. | |
| 1006 */ | |
| 1007 int | |
| 1008 vim_isprintc_strict(c) | |
| 1009 int c; | |
| 1010 { | |
| 1011 #ifdef FEAT_MBYTE | |
| 1012 if (enc_dbcs != 0 && c < 0x100 && MB_BYTE2LEN(c) > 1) | |
| 1013 return FALSE; | |
| 1014 if (enc_utf8 && c >= 0x100) | |
| 1015 return utf_printable(c); | |
| 1016 #endif | |
| 1017 return (c >= 0x100 || (c > 0 && (chartab[c] & CT_PRINT_CHAR))); | |
| 1018 } | |
| 1019 | |
| 1020 /* | |
| 1021 * like chartabsize(), but also check for line breaks on the screen | |
| 1022 */ | |
| 1023 int | |
| 1024 lbr_chartabsize(s, col) | |
| 1025 unsigned char *s; | |
| 1026 colnr_T col; | |
| 1027 { | |
| 1028 #ifdef FEAT_LINEBREAK | |
| 1029 if (!curwin->w_p_lbr && *p_sbr == NUL) | |
| 1030 { | |
| 1031 #endif | |
| 1032 #ifdef FEAT_MBYTE | |
| 1033 if (curwin->w_p_wrap) | |
| 1034 return win_nolbr_chartabsize(curwin, s, col, NULL); | |
| 1035 #endif | |
| 1036 RET_WIN_BUF_CHARTABSIZE(curwin, curbuf, s, col) | |
| 1037 #ifdef FEAT_LINEBREAK | |
| 1038 } | |
| 1039 return win_lbr_chartabsize(curwin, s, col, NULL); | |
| 1040 #endif | |
| 1041 } | |
| 1042 | |
| 1043 /* | |
| 1044 * Call lbr_chartabsize() and advance the pointer. | |
| 1045 */ | |
| 1046 int | |
| 1047 lbr_chartabsize_adv(s, col) | |
| 1048 char_u **s; | |
| 1049 colnr_T col; | |
| 1050 { | |
| 1051 int retval; | |
| 1052 | |
| 1053 retval = lbr_chartabsize(*s, col); | |
| 39 | 1054 mb_ptr_adv(*s); |
| 7 | 1055 return retval; |
| 1056 } | |
| 1057 | |
| 1058 /* | |
| 1059 * This function is used very often, keep it fast!!!! | |
| 1060 * | |
| 1061 * If "headp" not NULL, set *headp to the size of what we for 'showbreak' | |
| 1062 * string at start of line. Warning: *headp is only set if it's a non-zero | |
| 1063 * value, init to 0 before calling. | |
| 1064 */ | |
| 1065 int | |
| 1066 win_lbr_chartabsize(wp, s, col, headp) | |
| 1067 win_T *wp; | |
| 1068 char_u *s; | |
| 1069 colnr_T col; | |
| 1876 | 1070 int *headp UNUSED; |
| 7 | 1071 { |
| 1072 #ifdef FEAT_LINEBREAK | |
| 1073 int c; | |
| 1074 int size; | |
| 1075 colnr_T col2; | |
| 1076 colnr_T colmax; | |
| 1077 int added; | |
| 1078 # ifdef FEAT_MBYTE | |
| 1079 int mb_added = 0; | |
| 1080 # else | |
| 1081 # define mb_added 0 | |
| 1082 # endif | |
| 1083 int numberextra; | |
| 1084 char_u *ps; | |
| 1085 int tab_corr = (*s == TAB); | |
| 236 | 1086 int n; |
| 7 | 1087 |
| 1088 /* | |
| 1089 * No 'linebreak' and 'showbreak': return quickly. | |
| 1090 */ | |
| 1091 if (!wp->w_p_lbr && *p_sbr == NUL) | |
| 1092 #endif | |
| 1093 { | |
| 1094 #ifdef FEAT_MBYTE | |
| 1095 if (wp->w_p_wrap) | |
| 1096 return win_nolbr_chartabsize(wp, s, col, headp); | |
| 1097 #endif | |
| 1098 RET_WIN_BUF_CHARTABSIZE(wp, wp->w_buffer, s, col) | |
| 1099 } | |
| 1100 | |
| 1101 #ifdef FEAT_LINEBREAK | |
| 1102 /* | |
| 1103 * First get normal size, without 'linebreak' | |
| 1104 */ | |
| 1105 size = win_chartabsize(wp, s, col); | |
| 1106 c = *s; | |
| 1107 | |
| 1108 /* | |
| 1109 * If 'linebreak' set check at a blank before a non-blank if the line | |
| 1110 * needs a break here | |
| 1111 */ | |
| 1112 if (wp->w_p_lbr | |
| 1113 && vim_isbreak(c) | |
| 1114 && !vim_isbreak(s[1]) | |
| 1115 && !wp->w_p_list | |
| 1116 && wp->w_p_wrap | |
| 1117 # ifdef FEAT_VERTSPLIT | |
| 1118 && wp->w_width != 0 | |
| 1119 # endif | |
| 1120 ) | |
| 1121 { | |
| 1122 /* | |
| 1123 * Count all characters from first non-blank after a blank up to next | |
| 1124 * non-blank after a blank. | |
| 1125 */ | |
| 1126 numberextra = win_col_off(wp); | |
| 1127 col2 = col; | |
| 1869 | 1128 colmax = (colnr_T)(W_WIDTH(wp) - numberextra); |
| 7 | 1129 if (col >= colmax) |
| 236 | 1130 { |
| 1131 n = colmax + win_col_off2(wp); | |
| 1132 if (n > 0) | |
| 1133 colmax += (((col - colmax) / n) + 1) * n; | |
| 1134 } | |
| 1135 | |
| 7 | 1136 for (;;) |
| 1137 { | |
| 1138 ps = s; | |
| 39 | 1139 mb_ptr_adv(s); |
| 7 | 1140 c = *s; |
| 1141 if (!(c != NUL | |
| 1142 && (vim_isbreak(c) | |
| 1143 || (!vim_isbreak(c) | |
| 1144 && (col2 == col || !vim_isbreak(*ps)))))) | |
| 1145 break; | |
| 1146 | |
| 1147 col2 += win_chartabsize(wp, s, col2); | |
| 1148 if (col2 >= colmax) /* doesn't fit */ | |
| 1149 { | |
| 1150 size = colmax - col; | |
| 1151 tab_corr = FALSE; | |
| 1152 break; | |
| 1153 } | |
| 1154 } | |
| 1155 } | |
| 1156 # ifdef FEAT_MBYTE | |
| 1157 else if (has_mbyte && size == 2 && MB_BYTE2LEN(*s) > 1 | |
| 1158 && wp->w_p_wrap && in_win_border(wp, col)) | |
| 1159 { | |
| 1160 ++size; /* Count the ">" in the last column. */ | |
| 1161 mb_added = 1; | |
| 1162 } | |
| 1163 # endif | |
| 1164 | |
| 1165 /* | |
| 1166 * May have to add something for 'showbreak' string at start of line | |
| 1167 * Set *headp to the size of what we add. | |
| 1168 */ | |
| 1169 added = 0; | |
| 1170 if (*p_sbr != NUL && wp->w_p_wrap && col != 0) | |
| 1171 { | |
| 1172 numberextra = win_col_off(wp); | |
| 1173 col += numberextra + mb_added; | |
| 1174 if (col >= (colnr_T)W_WIDTH(wp)) | |
| 1175 { | |
| 1176 col -= W_WIDTH(wp); | |
| 1177 numberextra = W_WIDTH(wp) - (numberextra - win_col_off2(wp)); | |
| 1178 if (numberextra > 0) | |
| 1179 col = col % numberextra; | |
| 1180 } | |
| 1181 if (col == 0 || col + size > (colnr_T)W_WIDTH(wp)) | |
| 1182 { | |
| 1183 added = vim_strsize(p_sbr); | |
| 1184 if (tab_corr) | |
| 1185 size += (added / wp->w_buffer->b_p_ts) * wp->w_buffer->b_p_ts; | |
| 1186 else | |
| 1187 size += added; | |
| 1188 if (col != 0) | |
| 1189 added = 0; | |
| 1190 } | |
| 1191 } | |
| 1192 if (headp != NULL) | |
| 1193 *headp = added + mb_added; | |
| 1194 return size; | |
| 1195 #endif | |
| 1196 } | |
| 1197 | |
| 1198 #if defined(FEAT_MBYTE) || defined(PROTO) | |
| 1199 /* | |
| 1200 * Like win_lbr_chartabsize(), except that we know 'linebreak' is off and | |
| 1201 * 'wrap' is on. This means we need to check for a double-byte character that | |
| 1202 * doesn't fit at the end of the screen line. | |
| 1203 */ | |
| 1204 static int | |
| 1205 win_nolbr_chartabsize(wp, s, col, headp) | |
| 1206 win_T *wp; | |
| 1207 char_u *s; | |
| 1208 colnr_T col; | |
| 1209 int *headp; | |
| 1210 { | |
| 1211 int n; | |
| 1212 | |
| 1213 if (*s == TAB && (!wp->w_p_list || lcs_tab1)) | |
| 1214 { | |
| 1215 n = wp->w_buffer->b_p_ts; | |
| 1216 return (int)(n - (col % n)); | |
| 1217 } | |
| 1218 n = ptr2cells(s); | |
| 1219 /* Add one cell for a double-width character in the last column of the | |
| 1220 * window, displayed with a ">". */ | |
| 1221 if (n == 2 && MB_BYTE2LEN(*s) > 1 && in_win_border(wp, col)) | |
| 1222 { | |
| 1223 if (headp != NULL) | |
| 1224 *headp = 1; | |
| 1225 return 3; | |
| 1226 } | |
| 1227 return n; | |
| 1228 } | |
| 1229 | |
| 1230 /* | |
| 1231 * Return TRUE if virtual column "vcol" is in the rightmost column of window | |
| 1232 * "wp". | |
| 1233 */ | |
| 1234 int | |
| 1235 in_win_border(wp, vcol) | |
| 1236 win_T *wp; | |
| 1237 colnr_T vcol; | |
| 1238 { | |
| 1869 | 1239 int width1; /* width of first line (after line number) */ |
| 1240 int width2; /* width of further lines */ | |
| 7 | 1241 |
| 1242 #ifdef FEAT_VERTSPLIT | |
| 1243 if (wp->w_width == 0) /* there is no border */ | |
| 1244 return FALSE; | |
| 1245 #endif | |
| 1246 width1 = W_WIDTH(wp) - win_col_off(wp); | |
| 1869 | 1247 if ((int)vcol < width1 - 1) |
| 7 | 1248 return FALSE; |
| 1869 | 1249 if ((int)vcol == width1 - 1) |
| 7 | 1250 return TRUE; |
| 1251 width2 = width1 + win_col_off2(wp); | |
| 1970 | 1252 if (width2 <= 0) |
| 1253 return FALSE; | |
| 7 | 1254 return ((vcol - width1) % width2 == width2 - 1); |
| 1255 } | |
| 1256 #endif /* FEAT_MBYTE */ | |
| 1257 | |
| 1258 /* | |
| 1259 * Get virtual column number of pos. | |
| 1260 * start: on the first position of this character (TAB, ctrl) | |
| 1261 * cursor: where the cursor is on this character (first char, except for TAB) | |
| 1262 * end: on the last position of this character (TAB, ctrl) | |
| 1263 * | |
| 1264 * This is used very often, keep it fast! | |
| 1265 */ | |
| 1266 void | |
| 1267 getvcol(wp, pos, start, cursor, end) | |
| 1268 win_T *wp; | |
| 1269 pos_T *pos; | |
| 1270 colnr_T *start; | |
| 1271 colnr_T *cursor; | |
| 1272 colnr_T *end; | |
| 1273 { | |
| 1274 colnr_T vcol; | |
| 1275 char_u *ptr; /* points to current char */ | |
| 1276 char_u *posptr; /* points to char at pos->col */ | |
| 1277 int incr; | |
| 1278 int head; | |
| 1279 int ts = wp->w_buffer->b_p_ts; | |
| 1280 int c; | |
| 1281 | |
| 1282 vcol = 0; | |
| 1283 ptr = ml_get_buf(wp->w_buffer, pos->lnum, FALSE); | |
|
2108
3cdf2a653e00
updated for version 7.2.391
Bram Moolenaar <bram@zimbu.org>
parents:
1979
diff
changeset
|
1284 if (pos->col == MAXCOL) |
|
3cdf2a653e00
updated for version 7.2.391
Bram Moolenaar <bram@zimbu.org>
parents:
1979
diff
changeset
|
1285 posptr = NULL; /* continue until the NUL */ |
|
3cdf2a653e00
updated for version 7.2.391
Bram Moolenaar <bram@zimbu.org>
parents:
1979
diff
changeset
|
1286 else |
|
3cdf2a653e00
updated for version 7.2.391
Bram Moolenaar <bram@zimbu.org>
parents:
1979
diff
changeset
|
1287 posptr = ptr + pos->col; |
| 7 | 1288 |
| 1289 /* | |
| 1290 * This function is used very often, do some speed optimizations. | |
| 1291 * When 'list', 'linebreak' and 'showbreak' are not set use a simple loop. | |
| 1292 * Also use this when 'list' is set but tabs take their normal size. | |
| 1293 */ | |
| 1294 if ((!wp->w_p_list || lcs_tab1 != NUL) | |
| 1295 #ifdef FEAT_LINEBREAK | |
| 1296 && !wp->w_p_lbr && *p_sbr == NUL | |
| 1297 #endif | |
| 1298 ) | |
| 1299 { | |
| 1300 #ifndef FEAT_MBYTE | |
| 1301 head = 0; | |
| 1302 #endif | |
| 1303 for (;;) | |
| 1304 { | |
| 1305 #ifdef FEAT_MBYTE | |
| 1306 head = 0; | |
| 1307 #endif | |
| 1308 c = *ptr; | |
| 1309 /* make sure we don't go past the end of the line */ | |
| 1310 if (c == NUL) | |
| 1311 { | |
| 1312 incr = 1; /* NUL at end of line only takes one column */ | |
| 1313 break; | |
| 1314 } | |
| 1315 /* A tab gets expanded, depending on the current column */ | |
| 1316 if (c == TAB) | |
| 1317 incr = ts - (vcol % ts); | |
| 1318 else | |
| 1319 { | |
| 1320 #ifdef FEAT_MBYTE | |
| 1321 if (has_mbyte) | |
| 1322 { | |
| 1323 /* For utf-8, if the byte is >= 0x80, need to look at | |
| 1324 * further bytes to find the cell width. */ | |
| 1325 if (enc_utf8 && c >= 0x80) | |
| 1326 incr = utf_ptr2cells(ptr); | |
| 1327 else | |
| 1328 incr = CHARSIZE(c); | |
| 1329 | |
| 1330 /* If a double-cell char doesn't fit at the end of a line | |
| 1331 * it wraps to the next line, it's like this char is three | |
| 1332 * cells wide. */ | |
| 1546 | 1333 if (incr == 2 && wp->w_p_wrap && MB_BYTE2LEN(*ptr) > 1 |
| 1334 && in_win_border(wp, vcol)) | |
| 7 | 1335 { |
| 1336 ++incr; | |
| 1337 head = 1; | |
| 1338 } | |
| 1339 } | |
| 1340 else | |
| 1341 #endif | |
| 1342 incr = CHARSIZE(c); | |
| 1343 } | |
| 1344 | |
|
2108
3cdf2a653e00
updated for version 7.2.391
Bram Moolenaar <bram@zimbu.org>
parents:
1979
diff
changeset
|
1345 if (posptr != NULL && ptr >= posptr) /* character at pos->col */ |
| 7 | 1346 break; |
| 1347 | |
| 1348 vcol += incr; | |
| 39 | 1349 mb_ptr_adv(ptr); |
| 7 | 1350 } |
| 1351 } | |
| 1352 else | |
| 1353 { | |
| 1354 for (;;) | |
| 1355 { | |
| 1356 /* A tab gets expanded, depending on the current column */ | |
| 1357 head = 0; | |
| 1358 incr = win_lbr_chartabsize(wp, ptr, vcol, &head); | |
| 1359 /* make sure we don't go past the end of the line */ | |
| 1360 if (*ptr == NUL) | |
| 1361 { | |
| 1362 incr = 1; /* NUL at end of line only takes one column */ | |
| 1363 break; | |
| 1364 } | |
| 1365 | |
|
2108
3cdf2a653e00
updated for version 7.2.391
Bram Moolenaar <bram@zimbu.org>
parents:
1979
diff
changeset
|
1366 if (posptr != NULL && ptr >= posptr) /* character at pos->col */ |
| 7 | 1367 break; |
| 1368 | |
| 1369 vcol += incr; | |
| 39 | 1370 mb_ptr_adv(ptr); |
| 7 | 1371 } |
| 1372 } | |
| 1373 if (start != NULL) | |
| 1374 *start = vcol + head; | |
| 1375 if (end != NULL) | |
| 1376 *end = vcol + incr - 1; | |
| 1377 if (cursor != NULL) | |
| 1378 { | |
| 1379 if (*ptr == TAB | |
| 1380 && (State & NORMAL) | |
| 1381 && !wp->w_p_list | |
| 1382 && !virtual_active() | |
| 5735 | 1383 && !(VIsual_active && (*p_sel == 'e' || ltoreq(*pos, VIsual))) |
| 7 | 1384 ) |
| 1385 *cursor = vcol + incr - 1; /* cursor at end */ | |
| 1386 else | |
| 1387 *cursor = vcol + head; /* cursor at start */ | |
| 1388 } | |
| 1389 } | |
| 1390 | |
| 1391 /* | |
| 1392 * Get virtual cursor column in the current window, pretending 'list' is off. | |
| 1393 */ | |
| 1394 colnr_T | |
| 1395 getvcol_nolist(posp) | |
| 1396 pos_T *posp; | |
| 1397 { | |
| 1398 int list_save = curwin->w_p_list; | |
| 1399 colnr_T vcol; | |
| 1400 | |
| 1401 curwin->w_p_list = FALSE; | |
| 1402 getvcol(curwin, posp, NULL, &vcol, NULL); | |
| 1403 curwin->w_p_list = list_save; | |
| 1404 return vcol; | |
| 1405 } | |
| 1406 | |
| 1407 #if defined(FEAT_VIRTUALEDIT) || defined(PROTO) | |
| 1408 /* | |
| 1409 * Get virtual column in virtual mode. | |
| 1410 */ | |
| 1411 void | |
| 1412 getvvcol(wp, pos, start, cursor, end) | |
| 1413 win_T *wp; | |
| 1414 pos_T *pos; | |
| 1415 colnr_T *start; | |
| 1416 colnr_T *cursor; | |
| 1417 colnr_T *end; | |
| 1418 { | |
| 1419 colnr_T col; | |
| 1420 colnr_T coladd; | |
| 1421 colnr_T endadd; | |
| 1422 # ifdef FEAT_MBYTE | |
| 1423 char_u *ptr; | |
| 1424 # endif | |
| 1425 | |
| 1426 if (virtual_active()) | |
| 1427 { | |
| 1428 /* For virtual mode, only want one value */ | |
| 1429 getvcol(wp, pos, &col, NULL, NULL); | |
| 1430 | |
| 1431 coladd = pos->coladd; | |
| 1432 endadd = 0; | |
| 1433 # ifdef FEAT_MBYTE | |
| 1434 /* Cannot put the cursor on part of a wide character. */ | |
| 1435 ptr = ml_get_buf(wp->w_buffer, pos->lnum, FALSE); | |
| 1869 | 1436 if (pos->col < (colnr_T)STRLEN(ptr)) |
| 7 | 1437 { |
| 1438 int c = (*mb_ptr2char)(ptr + pos->col); | |
| 1439 | |
| 1440 if (c != TAB && vim_isprintc(c)) | |
| 1441 { | |
| 1869 | 1442 endadd = (colnr_T)(char2cells(c) - 1); |
| 557 | 1443 if (coladd > endadd) /* past end of line */ |
| 1444 endadd = 0; | |
| 7 | 1445 else |
| 1446 coladd = 0; | |
| 1447 } | |
| 1448 } | |
| 1449 # endif | |
| 1450 col += coladd; | |
| 1451 if (start != NULL) | |
| 1452 *start = col; | |
| 1453 if (cursor != NULL) | |
| 1454 *cursor = col; | |
| 1455 if (end != NULL) | |
| 1456 *end = col + endadd; | |
| 1457 } | |
| 1458 else | |
| 1459 getvcol(wp, pos, start, cursor, end); | |
| 1460 } | |
| 1461 #endif | |
| 1462 | |
| 1463 /* | |
| 1464 * Get the leftmost and rightmost virtual column of pos1 and pos2. | |
| 1465 * Used for Visual block mode. | |
| 1466 */ | |
| 1467 void | |
| 1468 getvcols(wp, pos1, pos2, left, right) | |
| 1469 win_T *wp; | |
| 1470 pos_T *pos1, *pos2; | |
| 1471 colnr_T *left, *right; | |
| 1472 { | |
| 1473 colnr_T from1, from2, to1, to2; | |
| 1474 | |
| 1475 if (ltp(pos1, pos2)) | |
| 1476 { | |
| 1477 getvvcol(wp, pos1, &from1, NULL, &to1); | |
| 1478 getvvcol(wp, pos2, &from2, NULL, &to2); | |
| 1479 } | |
| 1480 else | |
| 1481 { | |
| 1482 getvvcol(wp, pos2, &from1, NULL, &to1); | |
| 1483 getvvcol(wp, pos1, &from2, NULL, &to2); | |
| 1484 } | |
| 1485 if (from2 < from1) | |
| 1486 *left = from2; | |
| 1487 else | |
| 1488 *left = from1; | |
| 1489 if (to2 > to1) | |
| 1490 { | |
| 1491 if (*p_sel == 'e' && from2 - 1 >= to1) | |
| 1492 *right = from2 - 1; | |
| 1493 else | |
| 1494 *right = to2; | |
| 1495 } | |
| 1496 else | |
| 1497 *right = to1; | |
| 1498 } | |
| 1499 | |
| 1500 /* | |
| 1501 * skipwhite: skip over ' ' and '\t'. | |
| 1502 */ | |
| 1503 char_u * | |
| 1687 | 1504 skipwhite(q) |
| 1505 char_u *q; | |
| 7 | 1506 { |
| 1687 | 1507 char_u *p = q; |
| 1508 | |
| 7 | 1509 while (vim_iswhite(*p)) /* skip to next non-white */ |
| 1510 ++p; | |
| 1511 return p; | |
| 1512 } | |
| 1513 | |
| 1514 /* | |
| 293 | 1515 * skip over digits |
| 7 | 1516 */ |
| 1517 char_u * | |
| 1687 | 1518 skipdigits(q) |
| 1519 char_u *q; | |
| 7 | 1520 { |
| 1687 | 1521 char_u *p = q; |
| 1522 | |
| 7 | 1523 while (VIM_ISDIGIT(*p)) /* skip to next non-digit */ |
| 1524 ++p; | |
| 1525 return p; | |
| 1526 } | |
| 1527 | |
| 741 | 1528 #if defined(FEAT_SYN_HL) || defined(FEAT_SPELL) || defined(PROTO) |
| 301 | 1529 /* |
| 1530 * skip over digits and hex characters | |
| 1531 */ | |
| 1532 char_u * | |
| 1687 | 1533 skiphex(q) |
| 1534 char_u *q; | |
| 301 | 1535 { |
| 1687 | 1536 char_u *p = q; |
| 1537 | |
| 301 | 1538 while (vim_isxdigit(*p)) /* skip to next non-digit */ |
| 1539 ++p; | |
| 1540 return p; | |
| 1541 } | |
| 1542 #endif | |
| 1543 | |
| 293 | 1544 #if defined(FEAT_EX_EXTRA) || defined(PROTO) |
| 1545 /* | |
| 1546 * skip to digit (or NUL after the string) | |
| 1547 */ | |
| 1548 char_u * | |
| 1687 | 1549 skiptodigit(q) |
| 1550 char_u *q; | |
| 293 | 1551 { |
| 1687 | 1552 char_u *p = q; |
| 1553 | |
| 293 | 1554 while (*p != NUL && !VIM_ISDIGIT(*p)) /* skip to next digit */ |
| 1555 ++p; | |
| 1556 return p; | |
| 1557 } | |
| 1558 | |
| 1559 /* | |
| 1560 * skip to hex character (or NUL after the string) | |
| 1561 */ | |
| 1562 char_u * | |
| 1687 | 1563 skiptohex(q) |
| 1564 char_u *q; | |
| 293 | 1565 { |
| 1687 | 1566 char_u *p = q; |
| 1567 | |
| 293 | 1568 while (*p != NUL && !vim_isxdigit(*p)) /* skip to next digit */ |
| 1569 ++p; | |
| 1570 return p; | |
| 1571 } | |
| 1572 #endif | |
| 1573 | |
| 7 | 1574 /* |
| 1575 * Variant of isdigit() that can handle characters > 0x100. | |
| 1576 * We don't use isdigit() here, because on some systems it also considers | |
| 1577 * superscript 1 to be a digit. | |
| 1578 * Use the VIM_ISDIGIT() macro for simple arguments. | |
| 1579 */ | |
| 1580 int | |
| 1581 vim_isdigit(c) | |
| 1582 int c; | |
| 1583 { | |
| 1584 return (c >= '0' && c <= '9'); | |
| 1585 } | |
| 1586 | |
| 1587 /* | |
| 1588 * Variant of isxdigit() that can handle characters > 0x100. | |
| 1589 * We don't use isxdigit() here, because on some systems it also considers | |
| 1590 * superscript 1 to be a digit. | |
| 1591 */ | |
| 1592 int | |
| 1593 vim_isxdigit(c) | |
| 1594 int c; | |
| 1595 { | |
| 1596 return (c >= '0' && c <= '9') | |
| 1597 || (c >= 'a' && c <= 'f') | |
| 1598 || (c >= 'A' && c <= 'F'); | |
| 1599 } | |
| 1600 | |
| 492 | 1601 #if defined(FEAT_MBYTE) || defined(PROTO) |
| 1602 /* | |
| 1603 * Vim's own character class functions. These exist because many library | |
| 1604 * islower()/toupper() etc. do not work properly: they crash when used with | |
| 1605 * invalid values or can't handle latin1 when the locale is C. | |
| 1606 * Speed is most important here. | |
| 1607 */ | |
| 1608 #define LATIN1LOWER 'l' | |
| 1609 #define LATIN1UPPER 'U' | |
| 1610 | |
| 497 | 1611 static char_u latin1flags[257] = " UUUUUUUUUUUUUUUUUUUUUUUUUU llllllllllllllllllllllllll UUUUUUUUUUUUUUUUUUUUUUU UUUUUUUllllllllllllllllllllllll llllllll"; |
| 3533 | 1612 static char_u latin1upper[257] = " !\"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`ABCDEFGHIJKLMNOPQRSTUVWXYZ{|}~\x7f\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8a\x8b\x8c\x8d\x8e\x8f\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9a\x9b\x9c\x9d\x9e\x9f\xa0\xa1\xa2\xa3\xa4\xa5\xa6\xa7\xa8\xa9\xaa\xab\xac\xad\xae\xaf\xb0\xb1\xb2\xb3\xb4\xb5\xb6\xb7\xb8\xb9\xba\xbb\xbc\xbd\xbe\xbf\xc0\xc1\xc2\xc3\xc4\xc5\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf\xd0\xd1\xd2\xd3\xd4\xd5\xd6\xd7\xd8\xd9\xda\xdb\xdc\xdd\xde\xdf\xc0\xc1\xc2\xc3\xc4\xc5\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf\xd0\xd1\xd2\xd3\xd4\xd5\xd6\xf7\xd8\xd9\xda\xdb\xdc\xdd\xde\xff"; |
| 1613 static char_u latin1lower[257] = " !\"#$%&'()*+,-./0123456789:;<=>?@abcdefghijklmnopqrstuvwxyz[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\x7f\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8a\x8b\x8c\x8d\x8e\x8f\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9a\x9b\x9c\x9d\x9e\x9f\xa0\xa1\xa2\xa3\xa4\xa5\xa6\xa7\xa8\xa9\xaa\xab\xac\xad\xae\xaf\xb0\xb1\xb2\xb3\xb4\xb5\xb6\xb7\xb8\xb9\xba\xbb\xbc\xbd\xbe\xbf\xe0\xe1\xe2\xe3\xe4\xe5\xe6\xe7\xe8\xe9\xea\xeb\xec\xed\xee\xef\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xd7\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xdf\xe0\xe1\xe2\xe3\xe4\xe5\xe6\xe7\xe8\xe9\xea\xeb\xec\xed\xee\xef\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf7\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff"; | |
| 492 | 1614 |
| 1615 int | |
| 1616 vim_islower(c) | |
| 1617 int c; | |
| 1618 { | |
| 1619 if (c <= '@') | |
| 1620 return FALSE; | |
| 1621 if (c >= 0x80) | |
| 1622 { | |
| 1623 if (enc_utf8) | |
| 1624 return utf_islower(c); | |
| 1625 if (c >= 0x100) | |
| 1626 { | |
| 1627 #ifdef HAVE_ISWLOWER | |
| 1628 if (has_mbyte) | |
| 1629 return iswlower(c); | |
| 1630 #endif | |
| 1631 /* islower() can't handle these chars and may crash */ | |
| 1632 return FALSE; | |
| 1633 } | |
| 1634 if (enc_latin1like) | |
| 1635 return (latin1flags[c] & LATIN1LOWER) == LATIN1LOWER; | |
| 1636 } | |
| 1637 return islower(c); | |
| 1638 } | |
| 1639 | |
| 1640 int | |
| 1641 vim_isupper(c) | |
| 1642 int c; | |
| 1643 { | |
| 1644 if (c <= '@') | |
| 1645 return FALSE; | |
| 1646 if (c >= 0x80) | |
| 1647 { | |
| 1648 if (enc_utf8) | |
| 1649 return utf_isupper(c); | |
| 1650 if (c >= 0x100) | |
| 1651 { | |
| 1652 #ifdef HAVE_ISWUPPER | |
| 1653 if (has_mbyte) | |
| 1654 return iswupper(c); | |
| 1655 #endif | |
| 1656 /* islower() can't handle these chars and may crash */ | |
| 1657 return FALSE; | |
| 1658 } | |
| 1659 if (enc_latin1like) | |
| 1660 return (latin1flags[c] & LATIN1UPPER) == LATIN1UPPER; | |
| 1661 } | |
| 1662 return isupper(c); | |
| 1663 } | |
| 1664 | |
| 1665 int | |
| 1666 vim_toupper(c) | |
| 1667 int c; | |
| 1668 { | |
| 1669 if (c <= '@') | |
| 1670 return c; | |
| 1671 if (c >= 0x80) | |
| 1672 { | |
| 1673 if (enc_utf8) | |
| 1674 return utf_toupper(c); | |
| 1675 if (c >= 0x100) | |
| 1676 { | |
| 1677 #ifdef HAVE_TOWUPPER | |
| 1678 if (has_mbyte) | |
| 1679 return towupper(c); | |
| 1680 #endif | |
| 1681 /* toupper() can't handle these chars and may crash */ | |
| 1682 return c; | |
| 1683 } | |
| 1684 if (enc_latin1like) | |
| 1685 return latin1upper[c]; | |
| 1686 } | |
| 1687 return TOUPPER_LOC(c); | |
| 1688 } | |
| 1689 | |
| 1690 int | |
| 1691 vim_tolower(c) | |
| 1692 int c; | |
| 1693 { | |
| 1694 if (c <= '@') | |
| 1695 return c; | |
| 1696 if (c >= 0x80) | |
| 1697 { | |
| 1698 if (enc_utf8) | |
| 1699 return utf_tolower(c); | |
| 1700 if (c >= 0x100) | |
| 1701 { | |
| 1702 #ifdef HAVE_TOWLOWER | |
| 1703 if (has_mbyte) | |
| 1704 return towlower(c); | |
| 1705 #endif | |
| 1706 /* tolower() can't handle these chars and may crash */ | |
| 1707 return c; | |
| 1708 } | |
| 1709 if (enc_latin1like) | |
| 1710 return latin1lower[c]; | |
| 1711 } | |
| 1712 return TOLOWER_LOC(c); | |
| 1713 } | |
| 1714 #endif | |
| 1715 | |
| 7 | 1716 /* |
| 1717 * skiptowhite: skip over text until ' ' or '\t' or NUL. | |
| 1718 */ | |
| 1719 char_u * | |
| 1720 skiptowhite(p) | |
| 1721 char_u *p; | |
| 1722 { | |
| 1723 while (*p != ' ' && *p != '\t' && *p != NUL) | |
| 1724 ++p; | |
| 1725 return p; | |
| 1726 } | |
| 1727 | |
| 1728 #if defined(FEAT_LISTCMDS) || defined(FEAT_SIGNS) || defined(FEAT_SNIFF) \ | |
| 1729 || defined(PROTO) | |
| 1730 /* | |
| 1731 * skiptowhite_esc: Like skiptowhite(), but also skip escaped chars | |
| 1732 */ | |
| 1733 char_u * | |
| 1734 skiptowhite_esc(p) | |
| 1735 char_u *p; | |
| 1736 { | |
| 1737 while (*p != ' ' && *p != '\t' && *p != NUL) | |
| 1738 { | |
| 1739 if ((*p == '\\' || *p == Ctrl_V) && *(p + 1) != NUL) | |
| 1740 ++p; | |
| 1741 ++p; | |
| 1742 } | |
| 1743 return p; | |
| 1744 } | |
| 1745 #endif | |
| 1746 | |
| 1747 /* | |
| 1748 * Getdigits: Get a number from a string and skip over it. | |
| 1749 * Note: the argument is a pointer to a char_u pointer! | |
| 1750 */ | |
| 1751 long | |
| 1752 getdigits(pp) | |
| 1753 char_u **pp; | |
| 1754 { | |
| 1755 char_u *p; | |
| 1756 long retval; | |
| 1757 | |
| 1758 p = *pp; | |
| 1759 retval = atol((char *)p); | |
| 1760 if (*p == '-') /* skip negative sign */ | |
| 1761 ++p; | |
| 1762 p = skipdigits(p); /* skip to next non-digit */ | |
| 1763 *pp = p; | |
| 1764 return retval; | |
| 1765 } | |
| 1766 | |
| 1767 /* | |
| 1768 * Return TRUE if "lbuf" is empty or only contains blanks. | |
| 1769 */ | |
| 1770 int | |
| 1771 vim_isblankline(lbuf) | |
| 1772 char_u *lbuf; | |
| 1773 { | |
| 1774 char_u *p; | |
| 1775 | |
| 1776 p = skipwhite(lbuf); | |
| 1777 return (*p == NUL || *p == '\r' || *p == '\n'); | |
| 1778 } | |
| 1779 | |
| 1780 /* | |
| 1781 * Convert a string into a long and/or unsigned long, taking care of | |
| 16 | 1782 * hexadecimal and octal numbers. Accepts a '-' sign. |
| 7 | 1783 * If "hexp" is not NULL, returns a flag to indicate the type of the number: |
| 1784 * 0 decimal | |
| 1785 * '0' octal | |
| 1786 * 'X' hex | |
| 1787 * 'x' hex | |
| 1788 * If "len" is not NULL, the length of the number in characters is returned. | |
| 1789 * If "nptr" is not NULL, the signed result is returned in it. | |
| 1790 * If "unptr" is not NULL, the unsigned result is returned in it. | |
| 293 | 1791 * If "dooct" is non-zero recognize octal numbers, when > 1 always assume |
| 1792 * octal number. | |
| 782 | 1793 * If "dohex" is non-zero recognize hex numbers, when > 1 always assume |
| 293 | 1794 * hex number. |
| 7 | 1795 */ |
| 1796 void | |
| 1797 vim_str2nr(start, hexp, len, dooct, dohex, nptr, unptr) | |
| 1798 char_u *start; | |
| 1799 int *hexp; /* return: type of number 0 = decimal, 'x' | |
| 1800 or 'X' is hex, '0' = octal */ | |
| 1801 int *len; /* return: detected length of number */ | |
| 1802 int dooct; /* recognize octal number */ | |
| 1803 int dohex; /* recognize hex number */ | |
| 1804 long *nptr; /* return: signed result */ | |
| 1805 unsigned long *unptr; /* return: unsigned result */ | |
| 1806 { | |
| 1807 char_u *ptr = start; | |
| 1808 int hex = 0; /* default is decimal */ | |
| 1809 int negative = FALSE; | |
| 1810 unsigned long un = 0; | |
| 39 | 1811 int n; |
| 7 | 1812 |
| 1813 if (ptr[0] == '-') | |
| 1814 { | |
| 1815 negative = TRUE; | |
| 1816 ++ptr; | |
| 1817 } | |
| 1818 | |
| 39 | 1819 /* Recognize hex and octal. */ |
| 1820 if (ptr[0] == '0' && ptr[1] != '8' && ptr[1] != '9') | |
| 7 | 1821 { |
| 1822 hex = ptr[1]; | |
| 1823 if (dohex && (hex == 'X' || hex == 'x') && vim_isxdigit(ptr[2])) | |
| 1824 ptr += 2; /* hexadecimal */ | |
| 1825 else | |
| 1826 { | |
| 39 | 1827 hex = 0; /* default is decimal */ |
| 1828 if (dooct) | |
| 1829 { | |
| 1830 /* Don't interpret "0", "08" or "0129" as octal. */ | |
| 1831 for (n = 1; VIM_ISDIGIT(ptr[n]); ++n) | |
| 1832 { | |
| 1833 if (ptr[n] > '7') | |
| 1834 { | |
| 1835 hex = 0; /* can't be octal */ | |
| 1836 break; | |
| 1837 } | |
| 3292 | 1838 if (ptr[n] >= '0') |
| 39 | 1839 hex = '0'; /* assume octal */ |
| 1840 } | |
| 1841 } | |
| 7 | 1842 } |
| 1843 } | |
| 1844 | |
| 1845 /* | |
| 1846 * Do the string-to-numeric conversion "manually" to avoid sscanf quirks. | |
| 1847 */ | |
| 293 | 1848 if (hex == '0' || dooct > 1) |
| 7 | 1849 { |
| 293 | 1850 /* octal */ |
| 1851 while ('0' <= *ptr && *ptr <= '7') | |
| 7 | 1852 { |
| 293 | 1853 un = 8 * un + (unsigned long)(*ptr - '0'); |
| 1854 ++ptr; | |
| 7 | 1855 } |
| 293 | 1856 } |
| 1857 else if (hex != 0 || dohex > 1) | |
| 1858 { | |
| 1859 /* hex */ | |
| 1860 while (vim_isxdigit(*ptr)) | |
| 7 | 1861 { |
| 293 | 1862 un = 16 * un + (unsigned long)hex2nr(*ptr); |
| 1863 ++ptr; | |
| 7 | 1864 } |
| 1865 } | |
| 1866 else | |
| 1867 { | |
| 1868 /* decimal */ | |
| 1869 while (VIM_ISDIGIT(*ptr)) | |
| 1870 { | |
| 1871 un = 10 * un + (unsigned long)(*ptr - '0'); | |
| 1872 ++ptr; | |
| 1873 } | |
| 1874 } | |
| 1875 | |
| 1876 if (hexp != NULL) | |
| 1877 *hexp = hex; | |
| 1878 if (len != NULL) | |
| 1879 *len = (int)(ptr - start); | |
| 1880 if (nptr != NULL) | |
| 16 | 1881 { |
| 1882 if (negative) /* account for leading '-' for decimal numbers */ | |
| 1883 *nptr = -(long)un; | |
| 1884 else | |
| 1885 *nptr = (long)un; | |
| 1886 } | |
| 7 | 1887 if (unptr != NULL) |
| 1888 *unptr = un; | |
| 1889 } | |
| 1890 | |
| 1891 /* | |
| 1892 * Return the value of a single hex character. | |
| 1893 * Only valid when the argument is '0' - '9', 'A' - 'F' or 'a' - 'f'. | |
| 1894 */ | |
| 1895 int | |
| 1896 hex2nr(c) | |
| 1897 int c; | |
| 1898 { | |
| 1899 if (c >= 'a' && c <= 'f') | |
| 1900 return c - 'a' + 10; | |
| 1901 if (c >= 'A' && c <= 'F') | |
| 1902 return c - 'A' + 10; | |
| 1903 return c - '0'; | |
| 1904 } | |
| 1905 | |
| 1906 #if defined(FEAT_TERMRESPONSE) \ | |
| 1907 || (defined(FEAT_GUI_GTK) && defined(FEAT_WINDOWS)) || defined(PROTO) | |
| 1908 /* | |
| 1909 * Convert two hex characters to a byte. | |
| 1910 * Return -1 if one of the characters is not hex. | |
| 1911 */ | |
| 1912 int | |
| 1913 hexhex2nr(p) | |
| 1914 char_u *p; | |
| 1915 { | |
| 1916 if (!vim_isxdigit(p[0]) || !vim_isxdigit(p[1])) | |
| 1917 return -1; | |
| 1918 return (hex2nr(p[0]) << 4) + hex2nr(p[1]); | |
| 1919 } | |
| 1920 #endif | |
| 1921 | |
| 1922 /* | |
| 1923 * Return TRUE if "str" starts with a backslash that should be removed. | |
| 1924 * For MS-DOS, WIN32 and OS/2 this is only done when the character after the | |
| 1925 * backslash is not a normal file name character. | |
| 1926 * '$' is a valid file name character, we don't remove the backslash before | |
| 1927 * it. This means it is not possible to use an environment variable after a | |
| 1928 * backslash. "C:\$VIM\doc" is taken literally, only "$VIM\doc" works. | |
| 1929 * Although "\ name" is valid, the backslash in "Program\ files" must be | |
| 1930 * removed. Assume a file name doesn't start with a space. | |
| 1931 * For multi-byte names, never remove a backslash before a non-ascii | |
| 1932 * character, assume that all multi-byte characters are valid file name | |
| 1933 * characters. | |
| 1934 */ | |
| 1935 int | |
| 1936 rem_backslash(str) | |
| 1937 char_u *str; | |
| 1938 { | |
| 1939 #ifdef BACKSLASH_IN_FILENAME | |
| 1940 return (str[0] == '\\' | |
| 1941 # ifdef FEAT_MBYTE | |
| 1942 && str[1] < 0x80 | |
| 1943 # endif | |
| 1944 && (str[1] == ' ' | |
| 1945 || (str[1] != NUL | |
| 1946 && str[1] != '*' | |
| 1947 && str[1] != '?' | |
| 1948 && !vim_isfilec(str[1])))); | |
| 1949 #else | |
| 1950 return (str[0] == '\\' && str[1] != NUL); | |
| 1951 #endif | |
| 1952 } | |
| 1953 | |
| 1954 /* | |
| 1955 * Halve the number of backslashes in a file name argument. | |
| 1956 * For MS-DOS we only do this if the character after the backslash | |
| 1957 * is not a normal file character. | |
| 1958 */ | |
| 1959 void | |
| 1960 backslash_halve(p) | |
| 1961 char_u *p; | |
| 1962 { | |
| 1963 for ( ; *p; ++p) | |
| 1964 if (rem_backslash(p)) | |
| 1621 | 1965 STRMOVE(p, p + 1); |
| 7 | 1966 } |
| 1967 | |
| 1968 /* | |
| 1969 * backslash_halve() plus save the result in allocated memory. | |
| 1970 */ | |
| 1971 char_u * | |
| 1972 backslash_halve_save(p) | |
| 1973 char_u *p; | |
| 1974 { | |
| 1975 char_u *res; | |
| 1976 | |
| 1977 res = vim_strsave(p); | |
| 1978 if (res == NULL) | |
| 1979 return p; | |
| 1980 backslash_halve(res); | |
| 1981 return res; | |
| 1982 } | |
| 1983 | |
| 1984 #if (defined(EBCDIC) && defined(FEAT_POSTSCRIPT)) || defined(PROTO) | |
| 1985 /* | |
| 1986 * Table for EBCDIC to ASCII conversion unashamedly taken from xxd.c! | |
| 1987 * The first 64 entries have been added to map control characters defined in | |
| 1988 * ascii.h | |
| 1989 */ | |
| 1990 static char_u ebcdic2ascii_tab[256] = | |
| 1991 { | |
| 1992 0000, 0001, 0002, 0003, 0004, 0011, 0006, 0177, | |
| 1993 0010, 0011, 0012, 0013, 0014, 0015, 0016, 0017, | |
| 1994 0020, 0021, 0022, 0023, 0024, 0012, 0010, 0027, | |
| 1995 0030, 0031, 0032, 0033, 0033, 0035, 0036, 0037, | |
| 1996 0040, 0041, 0042, 0043, 0044, 0045, 0046, 0047, | |
| 1997 0050, 0051, 0052, 0053, 0054, 0055, 0056, 0057, | |
| 1998 0060, 0061, 0062, 0063, 0064, 0065, 0066, 0067, | |
| 1999 0070, 0071, 0072, 0073, 0074, 0075, 0076, 0077, | |
| 2000 0040, 0240, 0241, 0242, 0243, 0244, 0245, 0246, | |
| 2001 0247, 0250, 0325, 0056, 0074, 0050, 0053, 0174, | |
| 2002 0046, 0251, 0252, 0253, 0254, 0255, 0256, 0257, | |
| 2003 0260, 0261, 0041, 0044, 0052, 0051, 0073, 0176, | |
| 2004 0055, 0057, 0262, 0263, 0264, 0265, 0266, 0267, | |
| 2005 0270, 0271, 0313, 0054, 0045, 0137, 0076, 0077, | |
| 2006 0272, 0273, 0274, 0275, 0276, 0277, 0300, 0301, | |
| 2007 0302, 0140, 0072, 0043, 0100, 0047, 0075, 0042, | |
| 2008 0303, 0141, 0142, 0143, 0144, 0145, 0146, 0147, | |
| 2009 0150, 0151, 0304, 0305, 0306, 0307, 0310, 0311, | |
| 2010 0312, 0152, 0153, 0154, 0155, 0156, 0157, 0160, | |
| 2011 0161, 0162, 0136, 0314, 0315, 0316, 0317, 0320, | |
| 2012 0321, 0345, 0163, 0164, 0165, 0166, 0167, 0170, | |
| 2013 0171, 0172, 0322, 0323, 0324, 0133, 0326, 0327, | |
| 2014 0330, 0331, 0332, 0333, 0334, 0335, 0336, 0337, | |
| 2015 0340, 0341, 0342, 0343, 0344, 0135, 0346, 0347, | |
| 2016 0173, 0101, 0102, 0103, 0104, 0105, 0106, 0107, | |
| 2017 0110, 0111, 0350, 0351, 0352, 0353, 0354, 0355, | |
| 2018 0175, 0112, 0113, 0114, 0115, 0116, 0117, 0120, | |
| 2019 0121, 0122, 0356, 0357, 0360, 0361, 0362, 0363, | |
| 2020 0134, 0237, 0123, 0124, 0125, 0126, 0127, 0130, | |
| 2021 0131, 0132, 0364, 0365, 0366, 0367, 0370, 0371, | |
| 2022 0060, 0061, 0062, 0063, 0064, 0065, 0066, 0067, | |
| 2023 0070, 0071, 0372, 0373, 0374, 0375, 0376, 0377 | |
| 2024 }; | |
| 2025 | |
| 2026 /* | |
| 2027 * Convert a buffer worth of characters from EBCDIC to ASCII. Only useful if | |
| 2028 * wanting 7-bit ASCII characters out the other end. | |
| 2029 */ | |
| 2030 void | |
| 2031 ebcdic2ascii(buffer, len) | |
| 2032 char_u *buffer; | |
| 2033 int len; | |
| 2034 { | |
| 2035 int i; | |
| 2036 | |
| 2037 for (i = 0; i < len; i++) | |
| 2038 buffer[i] = ebcdic2ascii_tab[buffer[i]]; | |
| 2039 } | |
| 2040 #endif |