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
|
|
17 static int win_nolbr_chartabsize __ARGS((win_T *wp, char_u *s, colnr_T col, int *headp));
|
|
18 #endif
|
|
19
|
|
20 static int nr2hex __ARGS((int c));
|
|
21
|
|
22 static int chartab_initialized = FALSE;
|
|
23
|
|
24 /* b_chartab[] is an array of 32 bytes, each bit representing one of the
|
|
25 * characters 0-255. */
|
|
26 #define SET_CHARTAB(buf, c) (buf)->b_chartab[(unsigned)(c) >> 3] |= (1 << ((c) & 0x7))
|
|
27 #define RESET_CHARTAB(buf, c) (buf)->b_chartab[(unsigned)(c) >> 3] &= ~(1 << ((c) & 0x7))
|
|
28 #define GET_CHARTAB(buf, c) ((buf)->b_chartab[(unsigned)(c) >> 3] & (1 << ((c) & 0x7)))
|
|
29
|
|
30 /*
|
|
31 * Fill chartab[]. Also fills curbuf->b_chartab[] with flags for keyword
|
|
32 * characters for current buffer.
|
|
33 *
|
|
34 * Depends on the option settings 'iskeyword', 'isident', 'isfname',
|
|
35 * 'isprint' and 'encoding'.
|
|
36 *
|
|
37 * The index in chartab[] depends on 'encoding':
|
|
38 * - For non-multi-byte index with the byte (same as the character).
|
|
39 * - For DBCS index with the first byte.
|
|
40 * - For UTF-8 index with the character (when first byte is up to 0x80 it is
|
|
41 * the same as the character, if the first byte is 0x80 and above it depends
|
|
42 * on further bytes).
|
|
43 *
|
|
44 * The contents of chartab[]:
|
|
45 * - The lower two bits, masked by CT_CELL_MASK, give the number of display
|
|
46 * cells the character occupies (1 or 2). Not valid for UTF-8 above 0x80.
|
|
47 * - CT_PRINT_CHAR bit is set when the character is printable (no need to
|
|
48 * translate the character before displaying it). Note that only DBCS
|
|
49 * characters can have 2 display cells and still be printable.
|
|
50 * - CT_FNAME_CHAR bit is set when the character can be in a file name.
|
|
51 * - CT_ID_CHAR bit is set when the character can be in an identifier.
|
|
52 *
|
|
53 * Return FAIL if 'iskeyword', 'isident', 'isfname' or 'isprint' option has an
|
|
54 * error, OK otherwise.
|
|
55 */
|
|
56 int
|
|
57 init_chartab()
|
|
58 {
|
|
59 return buf_init_chartab(curbuf, TRUE);
|
|
60 }
|
|
61
|
|
62 int
|
|
63 buf_init_chartab(buf, global)
|
|
64 buf_T *buf;
|
|
65 int global; /* FALSE: only set buf->b_chartab[] */
|
|
66 {
|
|
67 int c;
|
|
68 int c2;
|
|
69 char_u *p;
|
|
70 int i;
|
|
71 int tilde;
|
|
72 int do_isalpha;
|
|
73
|
|
74 if (global)
|
|
75 {
|
|
76 /*
|
|
77 * Set the default size for printable characters:
|
|
78 * From <Space> to '~' is 1 (printable), others are 2 (not printable).
|
|
79 * This also inits all 'isident' and 'isfname' flags to FALSE.
|
|
80 *
|
|
81 * EBCDIC: all chars below ' ' are not printable, all others are
|
|
82 * printable.
|
|
83 */
|
|
84 c = 0;
|
|
85 while (c < ' ')
|
|
86 chartab[c++] = (dy_flags & DY_UHEX) ? 4 : 2;
|
|
87 #ifdef EBCDIC
|
|
88 while (c < 255)
|
|
89 #else
|
|
90 while (c <= '~')
|
|
91 #endif
|
|
92 chartab[c++] = 1 + CT_PRINT_CHAR;
|
|
93 #ifdef FEAT_FKMAP
|
|
94 if (p_altkeymap)
|
|
95 {
|
|
96 while (c < YE)
|
|
97 chartab[c++] = 1 + CT_PRINT_CHAR;
|
|
98 }
|
|
99 #endif
|
|
100 while (c < 256)
|
|
101 {
|
|
102 #ifdef FEAT_MBYTE
|
|
103 /* UTF-8: bytes 0xa0 - 0xff are printable (latin1) */
|
|
104 if (enc_utf8 && c >= 0xa0)
|
|
105 chartab[c++] = CT_PRINT_CHAR + 1;
|
|
106 /* euc-jp characters starting with 0x8e are single width */
|
|
107 else if (enc_dbcs == DBCS_JPNU && c == 0x8e)
|
|
108 chartab[c++] = CT_PRINT_CHAR + 1;
|
|
109 /* other double-byte chars can be printable AND double-width */
|
|
110 else if (enc_dbcs != 0 && MB_BYTE2LEN(c) == 2)
|
|
111 chartab[c++] = CT_PRINT_CHAR + 2;
|
|
112 else
|
|
113 #endif
|
|
114 /* the rest is unprintable by default */
|
|
115 chartab[c++] = (dy_flags & DY_UHEX) ? 4 : 2;
|
|
116 }
|
|
117
|
|
118 #ifdef FEAT_MBYTE
|
|
119 /* Assume that every multi-byte char is a filename character. */
|
|
120 for (c = 1; c < 256; ++c)
|
|
121 if ((enc_dbcs != 0 && MB_BYTE2LEN(c) > 1)
|
|
122 || (enc_dbcs == DBCS_JPNU && c == 0x8e)
|
|
123 || (enc_utf8 && c >= 0xa0))
|
|
124 chartab[c] |= CT_FNAME_CHAR;
|
|
125 #endif
|
|
126 }
|
|
127
|
|
128 /*
|
|
129 * Init word char flags all to FALSE
|
|
130 */
|
|
131 vim_memset(buf->b_chartab, 0, (size_t)32);
|
|
132 #ifdef FEAT_MBYTE
|
227
|
133 if (enc_dbcs != 0)
|
|
134 for (c = 0; c < 256; ++c)
|
|
135 {
|
|
136 /* double-byte characters are probably word characters */
|
|
137 if (MB_BYTE2LEN(c) == 2)
|
|
138 SET_CHARTAB(buf, c);
|
|
139 }
|
7
|
140 #endif
|
|
141
|
|
142 #ifdef FEAT_LISP
|
|
143 /*
|
|
144 * In lisp mode the '-' character is included in keywords.
|
|
145 */
|
|
146 if (buf->b_p_lisp)
|
|
147 SET_CHARTAB(buf, '-');
|
|
148 #endif
|
|
149
|
|
150 /* Walk through the 'isident', 'iskeyword', 'isfname' and 'isprint'
|
|
151 * options Each option is a list of characters, character numbers or
|
|
152 * ranges, separated by commas, e.g.: "200-210,x,#-178,-"
|
|
153 */
|
|
154 for (i = global ? 0 : 3; i <= 3; ++i)
|
|
155 {
|
|
156 if (i == 0)
|
|
157 p = p_isi; /* first round: 'isident' */
|
|
158 else if (i == 1)
|
|
159 p = p_isp; /* second round: 'isprint' */
|
|
160 else if (i == 2)
|
|
161 p = p_isf; /* third round: 'isfname' */
|
|
162 else /* i == 3 */
|
|
163 p = buf->b_p_isk; /* fourth round: 'iskeyword' */
|
|
164
|
|
165 while (*p)
|
|
166 {
|
|
167 tilde = FALSE;
|
|
168 do_isalpha = FALSE;
|
|
169 if (*p == '^' && p[1] != NUL)
|
|
170 {
|
|
171 tilde = TRUE;
|
|
172 ++p;
|
|
173 }
|
|
174 if (VIM_ISDIGIT(*p))
|
|
175 c = getdigits(&p);
|
|
176 else
|
|
177 c = *p++;
|
|
178 c2 = -1;
|
|
179 if (*p == '-' && p[1] != NUL)
|
|
180 {
|
|
181 ++p;
|
|
182 if (VIM_ISDIGIT(*p))
|
|
183 c2 = getdigits(&p);
|
|
184 else
|
|
185 c2 = *p++;
|
|
186 }
|
|
187 if (c <= 0 || (c2 < c && c2 != -1) || c2 >= 256
|
|
188 || !(*p == NUL || *p == ','))
|
|
189 return FAIL;
|
|
190
|
|
191 if (c2 == -1) /* not a range */
|
|
192 {
|
|
193 /*
|
|
194 * A single '@' (not "@-@"):
|
|
195 * Decide on letters being ID/printable/keyword chars with
|
|
196 * standard function isalpha(). This takes care of locale for
|
|
197 * single-byte characters).
|
|
198 */
|
|
199 if (c == '@')
|
|
200 {
|
|
201 do_isalpha = TRUE;
|
|
202 c = 1;
|
|
203 c2 = 255;
|
|
204 }
|
|
205 else
|
|
206 c2 = c;
|
|
207 }
|
|
208 while (c <= c2)
|
|
209 {
|
|
210 if (!do_isalpha || isalpha(c)
|
|
211 #ifdef FEAT_FKMAP
|
|
212 || (p_altkeymap && (F_isalpha(c) || F_isdigit(c)))
|
|
213 #endif
|
|
214 )
|
|
215 {
|
|
216 if (i == 0) /* (re)set ID flag */
|
|
217 {
|
|
218 if (tilde)
|
|
219 chartab[c] &= ~CT_ID_CHAR;
|
|
220 else
|
|
221 chartab[c] |= CT_ID_CHAR;
|
|
222 }
|
|
223 else if (i == 1) /* (re)set printable */
|
|
224 {
|
|
225 if ((c < ' '
|
|
226 #ifndef EBCDIC
|
|
227 || c > '~'
|
|
228 #endif
|
|
229 #ifdef FEAT_FKMAP
|
|
230 || (p_altkeymap
|
|
231 && (F_isalpha(c) || F_isdigit(c)))
|
|
232 #endif
|
|
233 )
|
|
234 #ifdef FEAT_MBYTE
|
|
235 /* For double-byte we keep the cell width, so
|
|
236 * that we can detect it from the first byte. */
|
|
237 && !(enc_dbcs && MB_BYTE2LEN(c) == 2)
|
|
238 #endif
|
|
239 )
|
|
240 {
|
|
241 if (tilde)
|
|
242 {
|
|
243 chartab[c] = (chartab[c] & ~CT_CELL_MASK)
|
|
244 + ((dy_flags & DY_UHEX) ? 4 : 2);
|
|
245 chartab[c] &= ~CT_PRINT_CHAR;
|
|
246 }
|
|
247 else
|
|
248 {
|
|
249 chartab[c] = (chartab[c] & ~CT_CELL_MASK) + 1;
|
|
250 chartab[c] |= CT_PRINT_CHAR;
|
|
251 }
|
|
252 }
|
|
253 }
|
|
254 else if (i == 2) /* (re)set fname flag */
|
|
255 {
|
|
256 if (tilde)
|
|
257 chartab[c] &= ~CT_FNAME_CHAR;
|
|
258 else
|
|
259 chartab[c] |= CT_FNAME_CHAR;
|
|
260 }
|
|
261 else /* i == 3 */ /* (re)set keyword flag */
|
|
262 {
|
|
263 if (tilde)
|
|
264 RESET_CHARTAB(buf, c);
|
|
265 else
|
|
266 SET_CHARTAB(buf, c);
|
|
267 }
|
|
268 }
|
|
269 ++c;
|
|
270 }
|
|
271 p = skip_to_option_part(p);
|
|
272 }
|
|
273 }
|
|
274 chartab_initialized = TRUE;
|
|
275 return OK;
|
|
276 }
|
|
277
|
|
278 /*
|
|
279 * Translate any special characters in buf[bufsize] in-place.
|
|
280 * The result is a string with only printable characters, but if there is not
|
|
281 * enough room, not all characters will be translated.
|
|
282 */
|
|
283 void
|
|
284 trans_characters(buf, bufsize)
|
|
285 char_u *buf;
|
|
286 int bufsize;
|
|
287 {
|
|
288 int len; /* length of string needing translation */
|
|
289 int room; /* room in buffer after string */
|
|
290 char_u *trs; /* translated character */
|
|
291 int trs_len; /* length of trs[] */
|
|
292
|
|
293 len = (int)STRLEN(buf);
|
|
294 room = bufsize - len;
|
|
295 while (*buf != 0)
|
|
296 {
|
|
297 # ifdef FEAT_MBYTE
|
|
298 /* Assume a multi-byte character doesn't need translation. */
|
|
299 if (has_mbyte && (trs_len = (*mb_ptr2len_check)(buf)) > 1)
|
|
300 len -= trs_len;
|
|
301 else
|
|
302 # endif
|
|
303 {
|
|
304 trs = transchar_byte(*buf);
|
|
305 trs_len = (int)STRLEN(trs);
|
|
306 if (trs_len > 1)
|
|
307 {
|
|
308 room -= trs_len - 1;
|
|
309 if (room <= 0)
|
|
310 return;
|
|
311 mch_memmove(buf + trs_len, buf + 1, (size_t)len);
|
|
312 }
|
|
313 mch_memmove(buf, trs, (size_t)trs_len);
|
|
314 --len;
|
|
315 }
|
|
316 buf += trs_len;
|
|
317 }
|
|
318 }
|
|
319
|
|
320 #if defined(FEAT_EVAL) || defined(FEAT_TITLE) || defined(PROTO)
|
|
321 /*
|
|
322 * Translate a string into allocated memory, replacing special chars with
|
|
323 * printable chars. Returns NULL when out of memory.
|
|
324 */
|
|
325 char_u *
|
|
326 transstr(s)
|
|
327 char_u *s;
|
|
328 {
|
|
329 char_u *res;
|
|
330 char_u *p;
|
|
331 #ifdef FEAT_MBYTE
|
|
332 int l, len, c;
|
|
333 char_u hexbuf[11];
|
|
334 #endif
|
|
335
|
|
336 #ifdef FEAT_MBYTE
|
|
337 if (has_mbyte)
|
|
338 {
|
|
339 /* Compute the length of the result, taking account of unprintable
|
|
340 * multi-byte characters. */
|
|
341 len = 0;
|
|
342 p = s;
|
|
343 while (*p != NUL)
|
|
344 {
|
|
345 if ((l = (*mb_ptr2len_check)(p)) > 1)
|
|
346 {
|
|
347 c = (*mb_ptr2char)(p);
|
|
348 p += l;
|
|
349 if (vim_isprintc(c))
|
|
350 len += l;
|
|
351 else
|
|
352 {
|
|
353 transchar_hex(hexbuf, c);
|
|
354 len += STRLEN(hexbuf);
|
|
355 }
|
|
356 }
|
|
357 else
|
|
358 {
|
|
359 l = byte2cells(*p++);
|
|
360 if (l > 0)
|
|
361 len += l;
|
|
362 else
|
|
363 len += 4; /* illegal byte sequence */
|
|
364 }
|
|
365 }
|
|
366 res = alloc((unsigned)(len + 1));
|
|
367 }
|
|
368 else
|
|
369 #endif
|
|
370 res = alloc((unsigned)(vim_strsize(s) + 1));
|
|
371 if (res != NULL)
|
|
372 {
|
|
373 *res = NUL;
|
|
374 p = s;
|
|
375 while (*p != NUL)
|
|
376 {
|
|
377 #ifdef FEAT_MBYTE
|
|
378 if (has_mbyte && (l = (*mb_ptr2len_check)(p)) > 1)
|
|
379 {
|
|
380 c = (*mb_ptr2char)(p);
|
|
381 if (vim_isprintc(c))
|
|
382 STRNCAT(res, p, l); /* append printable multi-byte char */
|
|
383 else
|
|
384 transchar_hex(res + STRLEN(res), c);
|
|
385 p += l;
|
|
386 }
|
|
387 else
|
|
388 #endif
|
|
389 STRCAT(res, transchar_byte(*p++));
|
|
390 }
|
|
391 }
|
|
392 return res;
|
|
393 }
|
|
394 #endif
|
|
395
|
|
396 #if defined(FEAT_SYN_HL) || defined(FEAT_INS_EXPAND) || defined(PROTO)
|
|
397 /*
|
221
|
398 * Convert the string "str[orglen]" to do ignore-case comparing. Uses the
|
|
399 * current locale.
|
130
|
400 * When "buf" is NULL returns an allocated string (NULL for out-of-memory).
|
|
401 * Otherwise puts the result in "buf[buflen]".
|
7
|
402 */
|
|
403 char_u *
|
130
|
404 str_foldcase(str, orglen, buf, buflen)
|
7
|
405 char_u *str;
|
130
|
406 int orglen;
|
|
407 char_u *buf;
|
|
408 int buflen;
|
7
|
409 {
|
|
410 garray_T ga;
|
|
411 int i;
|
130
|
412 int len = orglen;
|
7
|
413
|
|
414 #define GA_CHAR(i) ((char_u *)ga.ga_data)[i]
|
|
415 #define GA_PTR(i) ((char_u *)ga.ga_data + i)
|
130
|
416 #define STR_CHAR(i) (buf == NULL ? GA_CHAR(i) : buf[i])
|
|
417 #define STR_PTR(i) (buf == NULL ? GA_PTR(i) : buf + i)
|
7
|
418
|
130
|
419 /* Copy "str" into "buf" or allocated memory, unmodified. */
|
|
420 if (buf == NULL)
|
|
421 {
|
|
422 ga_init2(&ga, 1, 10);
|
|
423 if (ga_grow(&ga, len + 1) == FAIL)
|
|
424 return NULL;
|
|
425 mch_memmove(ga.ga_data, str, (size_t)len);
|
|
426 ga.ga_len = len;
|
|
427 }
|
|
428 else
|
|
429 {
|
|
430 if (len >= buflen) /* Ugly! */
|
|
431 len = buflen - 1;
|
|
432 mch_memmove(buf, str, (size_t)len);
|
|
433 }
|
|
434 if (buf == NULL)
|
|
435 GA_CHAR(len) = NUL;
|
|
436 else
|
|
437 buf[len] = NUL;
|
7
|
438
|
|
439 /* Make each character lower case. */
|
|
440 i = 0;
|
130
|
441 while (STR_CHAR(i) != NUL)
|
7
|
442 {
|
|
443 #ifdef FEAT_MBYTE
|
130
|
444 if (enc_utf8 || (has_mbyte && MB_BYTE2LEN(STR_CHAR(i)) > 1))
|
7
|
445 {
|
|
446 if (enc_utf8)
|
|
447 {
|
|
448 int c, lc;
|
|
449
|
130
|
450 c = utf_ptr2char(STR_PTR(i));
|
7
|
451 lc = utf_tolower(c);
|
|
452 if (c != lc)
|
|
453 {
|
|
454 int ol = utf_char2len(c);
|
|
455 int nl = utf_char2len(lc);
|
|
456
|
|
457 /* If the byte length changes need to shift the following
|
|
458 * characters forward or backward. */
|
|
459 if (ol != nl)
|
|
460 {
|
|
461 if (nl > ol)
|
130
|
462 {
|
|
463 if (buf == NULL ? ga_grow(&ga, nl - ol + 1) == FAIL
|
|
464 : len + nl - ol >= buflen)
|
7
|
465 {
|
|
466 /* out of memory, keep old char */
|
|
467 lc = c;
|
|
468 nl = ol;
|
|
469 }
|
130
|
470 }
|
7
|
471 if (ol != nl)
|
|
472 {
|
130
|
473 if (buf == NULL)
|
|
474 {
|
|
475 mch_memmove(GA_PTR(i) + nl, GA_PTR(i) + ol,
|
7
|
476 STRLEN(GA_PTR(i) + ol) + 1);
|
130
|
477 ga.ga_len += nl - ol;
|
|
478 }
|
|
479 else
|
|
480 {
|
|
481 mch_memmove(buf + i + nl, buf + i + ol,
|
|
482 STRLEN(buf + i + ol) + 1);
|
|
483 len += nl - ol;
|
|
484 }
|
7
|
485 }
|
|
486 }
|
130
|
487 (void)utf_char2bytes(lc, STR_PTR(i));
|
7
|
488 }
|
|
489 }
|
|
490 /* skip to next multi-byte char */
|
130
|
491 i += (*mb_ptr2len_check)(STR_PTR(i));
|
7
|
492 }
|
|
493 else
|
|
494 #endif
|
|
495 {
|
130
|
496 if (buf == NULL)
|
|
497 GA_CHAR(i) = TOLOWER_LOC(GA_CHAR(i));
|
|
498 else
|
|
499 buf[i] = TOLOWER_LOC(buf[i]);
|
7
|
500 ++i;
|
|
501 }
|
|
502 }
|
|
503
|
130
|
504 if (buf == NULL)
|
|
505 return (char_u *)ga.ga_data;
|
|
506 return buf;
|
7
|
507 }
|
|
508 #endif
|
|
509
|
|
510 /*
|
|
511 * Catch 22: chartab[] can't be initialized before the options are
|
|
512 * initialized, and initializing options may cause transchar() to be called!
|
|
513 * When chartab_initialized == FALSE don't use chartab[].
|
|
514 * Does NOT work for multi-byte characters, c must be <= 255.
|
|
515 * Also doesn't work for the first byte of a multi-byte, "c" must be a
|
|
516 * character!
|
|
517 */
|
|
518 static char_u transchar_buf[7];
|
|
519
|
|
520 char_u *
|
|
521 transchar(c)
|
|
522 int c;
|
|
523 {
|
|
524 int i;
|
|
525
|
|
526 i = 0;
|
|
527 if (IS_SPECIAL(c)) /* special key code, display as ~@ char */
|
|
528 {
|
|
529 transchar_buf[0] = '~';
|
|
530 transchar_buf[1] = '@';
|
|
531 i = 2;
|
|
532 c = K_SECOND(c);
|
|
533 }
|
|
534
|
|
535 if ((!chartab_initialized && (
|
|
536 #ifdef EBCDIC
|
|
537 (c >= 64 && c < 255)
|
|
538 #else
|
|
539 (c >= ' ' && c <= '~')
|
|
540 #endif
|
|
541 #ifdef FEAT_FKMAP
|
|
542 || F_ischar(c)
|
|
543 #endif
|
|
544 )) || (c < 256 && vim_isprintc_strict(c)))
|
|
545 {
|
|
546 /* printable character */
|
|
547 transchar_buf[i] = c;
|
|
548 transchar_buf[i + 1] = NUL;
|
|
549 }
|
|
550 else
|
|
551 transchar_nonprint(transchar_buf + i, c);
|
|
552 return transchar_buf;
|
|
553 }
|
|
554
|
|
555 #if defined(FEAT_MBYTE) || defined(PROTO)
|
|
556 /*
|
|
557 * Like transchar(), but called with a byte instead of a character. Checks
|
|
558 * for an illegal UTF-8 byte.
|
|
559 */
|
|
560 char_u *
|
|
561 transchar_byte(c)
|
|
562 int c;
|
|
563 {
|
|
564 if (enc_utf8 && c >= 0x80)
|
|
565 {
|
|
566 transchar_nonprint(transchar_buf, c);
|
|
567 return transchar_buf;
|
|
568 }
|
|
569 return transchar(c);
|
|
570 }
|
|
571 #endif
|
|
572
|
|
573 /*
|
|
574 * Convert non-printable character to two or more printable characters in
|
|
575 * "buf[]". "buf" needs to be able to hold five bytes.
|
|
576 * Does NOT work for multi-byte characters, c must be <= 255.
|
|
577 */
|
|
578 void
|
|
579 transchar_nonprint(buf, c)
|
|
580 char_u *buf;
|
|
581 int c;
|
|
582 {
|
|
583 if (c == NL)
|
|
584 c = NUL; /* we use newline in place of a NUL */
|
|
585 else if (c == CAR && get_fileformat(curbuf) == EOL_MAC)
|
|
586 c = NL; /* we use CR in place of NL in this case */
|
|
587
|
|
588 if (dy_flags & DY_UHEX) /* 'display' has "uhex" */
|
|
589 transchar_hex(buf, c);
|
|
590
|
|
591 #ifdef EBCDIC
|
|
592 /* For EBCDIC only the characters 0-63 and 255 are not printable */
|
|
593 else if (CtrlChar(c) != 0 || c == DEL)
|
|
594 #else
|
|
595 else if (c <= 0x7f) /* 0x00 - 0x1f and 0x7f */
|
|
596 #endif
|
|
597 {
|
|
598 buf[0] = '^';
|
|
599 #ifdef EBCDIC
|
|
600 if (c == DEL)
|
|
601 buf[1] = '?'; /* DEL displayed as ^? */
|
|
602 else
|
|
603 buf[1] = CtrlChar(c);
|
|
604 #else
|
|
605 buf[1] = c ^ 0x40; /* DEL displayed as ^? */
|
|
606 #endif
|
|
607
|
|
608 buf[2] = NUL;
|
|
609 }
|
|
610 #ifdef FEAT_MBYTE
|
|
611 else if (enc_utf8 && c >= 0x80)
|
|
612 {
|
|
613 transchar_hex(buf, c);
|
|
614 }
|
|
615 #endif
|
|
616 #ifndef EBCDIC
|
|
617 else if (c >= ' ' + 0x80 && c <= '~' + 0x80) /* 0xa0 - 0xfe */
|
|
618 {
|
|
619 buf[0] = '|';
|
|
620 buf[1] = c - 0x80;
|
|
621 buf[2] = NUL;
|
|
622 }
|
|
623 #else
|
|
624 else if (c < 64)
|
|
625 {
|
|
626 buf[0] = '~';
|
|
627 buf[1] = MetaChar(c);
|
|
628 buf[2] = NUL;
|
|
629 }
|
|
630 #endif
|
|
631 else /* 0x80 - 0x9f and 0xff */
|
|
632 {
|
|
633 /*
|
|
634 * TODO: EBCDIC I don't know what to do with this chars, so I display
|
|
635 * them as '~?' for now
|
|
636 */
|
|
637 buf[0] = '~';
|
|
638 #ifdef EBCDIC
|
|
639 buf[1] = '?'; /* 0xff displayed as ~? */
|
|
640 #else
|
|
641 buf[1] = (c - 0x80) ^ 0x40; /* 0xff displayed as ~? */
|
|
642 #endif
|
|
643 buf[2] = NUL;
|
|
644 }
|
|
645 }
|
|
646
|
|
647 void
|
|
648 transchar_hex(buf, c)
|
|
649 char_u *buf;
|
|
650 int c;
|
|
651 {
|
|
652 int i = 0;
|
|
653
|
|
654 buf[0] = '<';
|
|
655 #ifdef FEAT_MBYTE
|
|
656 if (c > 255)
|
|
657 {
|
|
658 buf[++i] = nr2hex((unsigned)c >> 12);
|
|
659 buf[++i] = nr2hex((unsigned)c >> 8);
|
|
660 }
|
|
661 #endif
|
|
662 buf[++i] = nr2hex((unsigned)c >> 4);
|
|
663 buf[++i] = nr2hex(c);
|
|
664 buf[++i] = '>';
|
|
665 buf[++i] = NUL;
|
|
666 }
|
|
667
|
|
668 /*
|
|
669 * Convert the lower 4 bits of byte "c" to its hex character.
|
|
670 * Lower case letters are used to avoid the confusion of <F1> being 0xf1 or
|
|
671 * function key 1.
|
|
672 */
|
|
673 static int
|
|
674 nr2hex(c)
|
|
675 int c;
|
|
676 {
|
|
677 if ((c & 0xf) <= 9)
|
|
678 return (c & 0xf) + '0';
|
|
679 return (c & 0xf) - 10 + 'a';
|
|
680 }
|
|
681
|
|
682 /*
|
|
683 * Return number of display cells occupied by byte "b".
|
|
684 * Caller must make sure 0 <= b <= 255.
|
|
685 * For multi-byte mode "b" must be the first byte of a character.
|
|
686 * A TAB is counted as two cells: "^I".
|
|
687 * For UTF-8 mode this will return 0 for bytes >= 0x80, because the number of
|
|
688 * cells depends on further bytes.
|
|
689 */
|
|
690 int
|
|
691 byte2cells(b)
|
|
692 int b;
|
|
693 {
|
|
694 #ifdef FEAT_MBYTE
|
|
695 if (enc_utf8 && b >= 0x80)
|
|
696 return 0;
|
|
697 #endif
|
|
698 return (chartab[b] & CT_CELL_MASK);
|
|
699 }
|
|
700
|
|
701 /*
|
|
702 * Return number of display cells occupied by character "c".
|
|
703 * "c" can be a special key (negative number) in which case 3 or 4 is returned.
|
|
704 * A TAB is counted as two cells: "^I" or four: "<09>".
|
|
705 */
|
|
706 int
|
|
707 char2cells(c)
|
|
708 int c;
|
|
709 {
|
|
710 if (IS_SPECIAL(c))
|
|
711 return char2cells(K_SECOND(c)) + 2;
|
|
712 #ifdef FEAT_MBYTE
|
|
713 if (c >= 0x80)
|
|
714 {
|
|
715 /* UTF-8: above 0x80 need to check the value */
|
|
716 if (enc_utf8)
|
|
717 return utf_char2cells(c);
|
|
718 /* DBCS: double-byte means double-width, except for euc-jp with first
|
|
719 * byte 0x8e */
|
|
720 if (enc_dbcs != 0 && c >= 0x100)
|
|
721 {
|
|
722 if (enc_dbcs == DBCS_JPNU && ((unsigned)c >> 8) == 0x8e)
|
|
723 return 1;
|
|
724 return 2;
|
|
725 }
|
|
726 }
|
|
727 #endif
|
|
728 return (chartab[c & 0xff] & CT_CELL_MASK);
|
|
729 }
|
|
730
|
|
731 /*
|
|
732 * Return number of display cells occupied by character at "*p".
|
|
733 * A TAB is counted as two cells: "^I" or four: "<09>".
|
|
734 */
|
|
735 int
|
|
736 ptr2cells(p)
|
|
737 char_u *p;
|
|
738 {
|
|
739 #ifdef FEAT_MBYTE
|
|
740 /* For UTF-8 we need to look at more bytes if the first byte is >= 0x80. */
|
|
741 if (enc_utf8 && *p >= 0x80)
|
|
742 return utf_ptr2cells(p);
|
|
743 /* For DBCS we can tell the cell count from the first byte. */
|
|
744 #endif
|
|
745 return (chartab[*p] & CT_CELL_MASK);
|
|
746 }
|
|
747
|
|
748 /*
|
|
749 * Return the number of characters string "s" will take on the screen,
|
|
750 * counting TABs as two characters: "^I".
|
|
751 */
|
|
752 int
|
|
753 vim_strsize(s)
|
|
754 char_u *s;
|
|
755 {
|
|
756 return vim_strnsize(s, (int)MAXCOL);
|
|
757 }
|
|
758
|
|
759 /*
|
|
760 * Return the number of characters string "s[len]" will take on the screen,
|
|
761 * counting TABs as two characters: "^I".
|
|
762 */
|
|
763 int
|
|
764 vim_strnsize(s, len)
|
|
765 char_u *s;
|
|
766 int len;
|
|
767 {
|
|
768 int size = 0;
|
|
769
|
|
770 while (*s != NUL && --len >= 0)
|
|
771 {
|
|
772 #ifdef FEAT_MBYTE
|
|
773 if (has_mbyte)
|
|
774 {
|
|
775 int l = (*mb_ptr2len_check)(s);
|
|
776
|
|
777 size += ptr2cells(s);
|
|
778 s += l;
|
|
779 len -= l - 1;
|
|
780 }
|
|
781 else
|
|
782 #endif
|
|
783 size += byte2cells(*s++);
|
|
784 }
|
|
785 return size;
|
|
786 }
|
|
787
|
|
788 /*
|
|
789 * Return the number of characters 'c' will take on the screen, taking
|
|
790 * into account the size of a tab.
|
|
791 * Use a define to make it fast, this is used very often!!!
|
|
792 * Also see getvcol() below.
|
|
793 */
|
|
794
|
|
795 #define RET_WIN_BUF_CHARTABSIZE(wp, buf, p, col) \
|
|
796 if (*(p) == TAB && (!(wp)->w_p_list || lcs_tab1)) \
|
|
797 { \
|
|
798 int ts; \
|
|
799 ts = (buf)->b_p_ts; \
|
|
800 return (int)(ts - (col % ts)); \
|
|
801 } \
|
|
802 else \
|
|
803 return ptr2cells(p);
|
|
804
|
|
805 #if defined(FEAT_VREPLACE) || defined(FEAT_EX_EXTRA) || defined(FEAT_GUI) \
|
|
806 || defined(FEAT_VIRTUALEDIT) || defined(PROTO)
|
|
807 int
|
|
808 chartabsize(p, col)
|
|
809 char_u *p;
|
|
810 colnr_T col;
|
|
811 {
|
|
812 RET_WIN_BUF_CHARTABSIZE(curwin, curbuf, p, col)
|
|
813 }
|
|
814 #endif
|
|
815
|
|
816 #ifdef FEAT_LINEBREAK
|
|
817 static int
|
|
818 win_chartabsize(wp, p, col)
|
|
819 win_T *wp;
|
|
820 char_u *p;
|
|
821 colnr_T col;
|
|
822 {
|
|
823 RET_WIN_BUF_CHARTABSIZE(wp, wp->w_buffer, p, col)
|
|
824 }
|
|
825 #endif
|
|
826
|
|
827 /*
|
|
828 * return the number of characters the string 's' will take on the screen,
|
|
829 * taking into account the size of a tab
|
|
830 */
|
|
831 int
|
|
832 linetabsize(s)
|
|
833 char_u *s;
|
|
834 {
|
|
835 colnr_T col = 0;
|
|
836
|
|
837 while (*s != NUL)
|
|
838 col += lbr_chartabsize_adv(&s, col);
|
|
839 return (int)col;
|
|
840 }
|
|
841
|
|
842 /*
|
|
843 * Like linetabsize(), but for a given window instead of the current one.
|
|
844 */
|
|
845 int
|
|
846 win_linetabsize(wp, p, len)
|
|
847 win_T *wp;
|
|
848 char_u *p;
|
|
849 colnr_T len;
|
|
850 {
|
|
851 colnr_T col = 0;
|
|
852 char_u *s;
|
|
853
|
40
|
854 for (s = p; *s != NUL && (len == MAXCOL || s < p + len); mb_ptr_adv(s))
|
7
|
855 col += win_lbr_chartabsize(wp, s, col, NULL);
|
|
856 return (int)col;
|
|
857 }
|
|
858
|
|
859 /*
|
42
|
860 * Return TRUE if 'c' is a normal identifier character:
|
|
861 * Letters and characters from the 'isident' option.
|
7
|
862 */
|
|
863 int
|
|
864 vim_isIDc(c)
|
|
865 int c;
|
|
866 {
|
|
867 return (c > 0 && c < 0x100 && (chartab[c] & CT_ID_CHAR));
|
|
868 }
|
|
869
|
|
870 /*
|
|
871 * return TRUE if 'c' is a keyword character: Letters and characters from
|
|
872 * 'iskeyword' option for current buffer.
|
|
873 * For multi-byte characters mb_get_class() is used (builtin rules).
|
|
874 */
|
|
875 int
|
|
876 vim_iswordc(c)
|
|
877 int c;
|
|
878 {
|
|
879 #ifdef FEAT_MBYTE
|
|
880 if (c >= 0x100)
|
|
881 {
|
|
882 if (enc_dbcs != 0)
|
|
883 return dbcs_class((unsigned)c >> 8, c & 0xff) >= 2;
|
|
884 if (enc_utf8)
|
|
885 return utf_class(c) >= 2;
|
|
886 }
|
|
887 #endif
|
|
888 return (c > 0 && c < 0x100 && GET_CHARTAB(curbuf, c) != 0);
|
|
889 }
|
|
890
|
|
891 /*
|
|
892 * Just like vim_iswordc() but uses a pointer to the (multi-byte) character.
|
|
893 */
|
|
894 int
|
|
895 vim_iswordp(p)
|
|
896 char_u *p;
|
|
897 {
|
|
898 #ifdef FEAT_MBYTE
|
|
899 if (has_mbyte && MB_BYTE2LEN(*p) > 1)
|
|
900 return mb_get_class(p) >= 2;
|
|
901 #endif
|
|
902 return GET_CHARTAB(curbuf, *p) != 0;
|
|
903 }
|
|
904
|
|
905 #if defined(FEAT_SYN_HL) || defined(PROTO)
|
|
906 int
|
|
907 vim_iswordc_buf(p, buf)
|
|
908 char_u *p;
|
|
909 buf_T *buf;
|
|
910 {
|
|
911 # ifdef FEAT_MBYTE
|
|
912 if (has_mbyte && MB_BYTE2LEN(*p) > 1)
|
|
913 return mb_get_class(p) >= 2;
|
|
914 # endif
|
|
915 return (GET_CHARTAB(buf, *p) != 0);
|
|
916 }
|
227
|
917
|
258
|
918 /*
|
|
919 * The tables used for spelling. These are only used for the first 256
|
|
920 * characters.
|
|
921 */
|
|
922 typedef struct spelltab_S
|
|
923 {
|
|
924 char_u st_isw[256]; /* flags: is word char */
|
|
925 char_u st_isu[256]; /* flags: is uppercase char */
|
|
926 char_u st_fold[256]; /* chars: folded case */
|
|
927 } spelltab_T;
|
|
928
|
|
929 static spelltab_T spelltab;
|
|
930 static int did_set_spelltab;
|
|
931
|
|
932 #define SPELL_ISWORD 1
|
|
933 #define SPELL_ISUPPER 2
|
|
934
|
|
935 static void clear_spell_chartab __ARGS((spelltab_T *sp));
|
|
936 static int set_spell_finish __ARGS((spelltab_T *new_st));
|
|
937
|
|
938 /*
|
|
939 * Init the chartab used for spelling for ASCII.
|
|
940 * EBCDIC is not supported!
|
|
941 */
|
|
942 static void
|
|
943 clear_spell_chartab(sp)
|
|
944 spelltab_T *sp;
|
|
945 {
|
|
946 int i;
|
|
947
|
|
948 /* Init everything to FALSE. */
|
|
949 vim_memset(sp->st_isw, FALSE, sizeof(sp->st_isw));
|
|
950 vim_memset(sp->st_isu, FALSE, sizeof(sp->st_isu));
|
|
951 for (i = 0; i < 256; ++i)
|
|
952 sp->st_fold[i] = i;
|
|
953
|
|
954 /* We include digits. A word shouldn't start with a digit, but handling
|
|
955 * that is done separately. */
|
|
956 for (i = '0'; i <= '9'; ++i)
|
|
957 sp->st_isw[i] = TRUE;
|
|
958 for (i = 'A'; i <= 'Z'; ++i)
|
|
959 {
|
|
960 sp->st_isw[i] = TRUE;
|
|
961 sp->st_isu[i] = TRUE;
|
|
962 sp->st_fold[i] = i + 0x20;
|
|
963 }
|
|
964 for (i = 'a'; i <= 'z'; ++i)
|
|
965 sp->st_isw[i] = TRUE;
|
|
966 }
|
227
|
967
|
|
968 /*
|
|
969 * Init the chartab used for spelling. Only depends on 'encoding'.
|
258
|
970 * Called once while starting up and when 'encoding' changes.
|
|
971 * The default is to use isalpha(), but the spell file should define the word
|
|
972 * characters to make it possible that 'encoding' differs from the current
|
|
973 * locale.
|
227
|
974 */
|
|
975 void
|
|
976 init_spell_chartab()
|
|
977 {
|
|
978 int i;
|
|
979
|
258
|
980 did_set_spelltab = FALSE;
|
|
981 clear_spell_chartab(&spelltab);
|
|
982
|
227
|
983 #ifdef FEAT_MBYTE
|
|
984 if (enc_dbcs)
|
|
985 {
|
|
986 /* DBCS: assume double-wide characters are word characters. */
|
258
|
987 for (i = 128; i <= 255; ++i)
|
227
|
988 if (MB_BYTE2LEN(i) == 2)
|
258
|
989 spelltab.st_isw[i] = TRUE;
|
236
|
990 }
|
227
|
991 else
|
|
992 #endif
|
258
|
993 {
|
|
994 /* Rough guess: use isalpha() for characters above 128. */
|
|
995 for (i = 128; i < 256; ++i)
|
|
996 {
|
|
997 spelltab.st_isw[i] = isalpha(i);
|
|
998 if (isupper(i))
|
|
999 {
|
|
1000 spelltab.st_isu[i] = TRUE;
|
|
1001 spelltab.st_fold[i] = tolower(i);
|
|
1002 }
|
|
1003 }
|
|
1004 }
|
|
1005 }
|
|
1006
|
|
1007 static char *e_affform = N_("E761: Format error in affix file FOL, LOW or UPP");
|
|
1008 static char *e_affrange = N_("E762: Character in FOL, LOW or UPP is out of range");
|
|
1009
|
|
1010 /*
|
|
1011 * Set the spell character tables from strings in the affix file.
|
|
1012 */
|
|
1013 int
|
|
1014 set_spell_chartab(fol, low, upp)
|
|
1015 char_u *fol;
|
|
1016 char_u *low;
|
|
1017 char_u *upp;
|
|
1018 {
|
|
1019 /* We build the new tables here first, so that we can compare with the
|
|
1020 * previous one. */
|
|
1021 spelltab_T new_st;
|
|
1022 char_u *pf = fol, *pl = low, *pu = upp;
|
|
1023 int f, l, u;
|
|
1024
|
|
1025 clear_spell_chartab(&new_st);
|
|
1026
|
|
1027 while (*pf != NUL)
|
227
|
1028 {
|
258
|
1029 if (*pl == NUL || *pu == NUL)
|
|
1030 {
|
|
1031 EMSG(_(e_affform));
|
|
1032 return FAIL;
|
|
1033 }
|
|
1034 #ifdef FEAT_MBYTE
|
|
1035 f = mb_ptr2char_adv(&pf);
|
|
1036 l = mb_ptr2char_adv(&pl);
|
|
1037 u = mb_ptr2char_adv(&pu);
|
|
1038 #else
|
|
1039 f = *pf++;
|
|
1040 l = *pl++;
|
|
1041 u = *pu++;
|
|
1042 #endif
|
|
1043 /* Every character that appears is a word character. */
|
|
1044 if (f < 256)
|
|
1045 new_st.st_isw[f] = TRUE;
|
|
1046 if (l < 256)
|
|
1047 new_st.st_isw[l] = TRUE;
|
|
1048 if (u < 256)
|
|
1049 new_st.st_isw[u] = TRUE;
|
|
1050
|
|
1051 /* if "LOW" and "FOL" are not the same the "LOW" char needs
|
|
1052 * case-folding */
|
|
1053 if (l < 256 && l != f)
|
|
1054 {
|
|
1055 if (f >= 256)
|
|
1056 {
|
|
1057 EMSG(_(e_affrange));
|
|
1058 return FAIL;
|
|
1059 }
|
|
1060 new_st.st_fold[l] = f;
|
|
1061 }
|
|
1062
|
|
1063 /* if "UPP" and "FOL" are not the same the "UPP" char needs
|
|
1064 * case-folding and it's upper case. */
|
|
1065 if (u < 256 && u != f)
|
|
1066 {
|
|
1067 if (f >= 256)
|
|
1068 {
|
|
1069 EMSG(_(e_affrange));
|
|
1070 return FAIL;
|
|
1071 }
|
|
1072 new_st.st_fold[u] = f;
|
|
1073 new_st.st_isu[u] = TRUE;
|
|
1074 }
|
227
|
1075 }
|
258
|
1076
|
|
1077 if (*pl != NUL || *pu != NUL)
|
|
1078 {
|
|
1079 EMSG(_(e_affform));
|
|
1080 return FAIL;
|
|
1081 }
|
|
1082
|
|
1083 return set_spell_finish(&new_st);
|
227
|
1084 }
|
|
1085
|
|
1086 /*
|
258
|
1087 * Set the spell character tables from strings in the .spl file.
|
|
1088 */
|
|
1089 int
|
|
1090 set_spell_charflags(flags, cnt, upp)
|
|
1091 char_u *flags;
|
|
1092 int cnt;
|
|
1093 char_u *upp;
|
|
1094 {
|
|
1095 /* We build the new tables here first, so that we can compare with the
|
|
1096 * previous one. */
|
|
1097 spelltab_T new_st;
|
|
1098 int i;
|
|
1099 char_u *p = upp;
|
|
1100
|
|
1101 clear_spell_chartab(&new_st);
|
|
1102
|
|
1103 for (i = 0; i < cnt; ++i)
|
|
1104 {
|
|
1105 new_st.st_isw[i + 128] = (flags[i] & SPELL_ISWORD) != 0;
|
|
1106 new_st.st_isu[i + 128] = (flags[i] & SPELL_ISUPPER) != 0;
|
|
1107
|
|
1108 if (*p == NUL)
|
|
1109 return FAIL;
|
|
1110 #ifdef FEAT_MBYTE
|
|
1111 new_st.st_fold[i + 128] = mb_ptr2char_adv(&p);
|
|
1112 #else
|
|
1113 new_st.st_fold[i + 128] = *p++;
|
|
1114 #endif
|
|
1115 }
|
|
1116
|
|
1117 return set_spell_finish(&new_st);
|
|
1118 }
|
|
1119
|
|
1120 static int
|
|
1121 set_spell_finish(new_st)
|
|
1122 spelltab_T *new_st;
|
|
1123 {
|
|
1124 int i;
|
|
1125
|
|
1126 if (did_set_spelltab)
|
|
1127 {
|
|
1128 /* check that it's the same table */
|
|
1129 for (i = 0; i < 256; ++i)
|
|
1130 {
|
|
1131 if (spelltab.st_isw[i] != new_st->st_isw[i]
|
|
1132 || spelltab.st_isu[i] != new_st->st_isu[i]
|
|
1133 || spelltab.st_fold[i] != new_st->st_fold[i])
|
|
1134 {
|
|
1135 EMSG(_("E763: Word characters differ between spell files"));
|
|
1136 return FAIL;
|
|
1137 }
|
|
1138 }
|
|
1139 }
|
|
1140 else
|
|
1141 {
|
|
1142 /* copy the new spelltab into the one being used */
|
|
1143 spelltab = *new_st;
|
|
1144 did_set_spelltab = TRUE;
|
|
1145 }
|
|
1146
|
|
1147 return OK;
|
|
1148 }
|
|
1149
|
|
1150 #if defined(FEAT_MBYTE) || defined(PROTO)
|
|
1151 /*
|
|
1152 * Write the current tables into the .spl file.
|
|
1153 */
|
|
1154 void
|
|
1155 write_spell_chartab(fd)
|
|
1156 FILE *fd;
|
|
1157 {
|
|
1158 char_u charbuf[256 * 4];
|
|
1159 int len = 0;
|
|
1160 int flags;
|
|
1161 int i;
|
|
1162
|
|
1163 if (!did_set_spelltab)
|
|
1164 {
|
|
1165 /* No character table specified, write zero counts. */
|
|
1166 fputc(0, fd);
|
|
1167 fputc(0, fd);
|
|
1168 fputc(0, fd);
|
|
1169 return;
|
|
1170 }
|
|
1171
|
|
1172 fputc(128, fd); /* <charflagslen> */
|
|
1173 for (i = 128; i < 256; ++i)
|
|
1174 {
|
|
1175 flags = 0;
|
|
1176 if (spelltab.st_isw[i])
|
|
1177 flags |= SPELL_ISWORD;
|
|
1178 if (spelltab.st_isu[i])
|
|
1179 flags |= SPELL_ISUPPER;
|
|
1180 fputc(flags, fd); /* <charflags> */
|
|
1181
|
|
1182 len += mb_char2bytes(spelltab.st_fold[i], charbuf + len);
|
|
1183 }
|
|
1184
|
|
1185 put_bytes(fd, (long_u)len, 2); /* <fcharlen> */
|
|
1186 fwrite(charbuf, (size_t)len, (size_t)1, fd); /* <fchars> */
|
|
1187 }
|
|
1188 #endif
|
|
1189
|
|
1190 /*
|
|
1191 * Return TRUE if "p" points to a word character for spelling.
|
227
|
1192 */
|
|
1193 int
|
|
1194 spell_iswordc(p)
|
|
1195 char_u *p;
|
|
1196 {
|
|
1197 # ifdef FEAT_MBYTE
|
|
1198 if (has_mbyte && MB_BYTE2LEN(*p) > 1)
|
|
1199 return mb_get_class(p) >= 2;
|
|
1200 # endif
|
258
|
1201 return spelltab.st_isw[*p];
|
|
1202 }
|
|
1203
|
|
1204 /*
|
|
1205 * Return TRUE if "c" is an upper-case character for spelling.
|
|
1206 */
|
|
1207 int
|
|
1208 spell_isupper(c)
|
|
1209 int c;
|
|
1210 {
|
|
1211 # ifdef FEAT_MBYTE
|
|
1212 if (enc_utf8)
|
|
1213 {
|
|
1214 /* For Unicode we can call utf_isupper(), but don't do that for ASCII,
|
|
1215 * because we don't want to use 'casemap' here. */
|
|
1216 if (c >= 128)
|
|
1217 return utf_isupper(c);
|
|
1218 }
|
|
1219 else if (has_mbyte && c > 256)
|
|
1220 {
|
|
1221 /* For characters above 255 we don't have something specfied.
|
|
1222 * Fall back to locale-dependent iswupper(). If not available
|
|
1223 * simply return FALSE. */
|
|
1224 # ifdef HAVE_ISWUPPER
|
|
1225 return iswupper(c);
|
|
1226 # else
|
|
1227 return FALSE;
|
|
1228 # endif
|
|
1229 }
|
|
1230 # endif
|
|
1231 return spelltab.st_isu[c];
|
227
|
1232 }
|
258
|
1233
|
|
1234 /*
|
|
1235 * case-fold "p[len]" into "buf[buflen]". Used for spell checking.
|
|
1236 * Returns FAIL when something wrong.
|
|
1237 */
|
|
1238 int
|
|
1239 spell_casefold(p, len, buf, buflen)
|
|
1240 char_u *p;
|
|
1241 int len;
|
|
1242 char_u *buf;
|
|
1243 int buflen;
|
|
1244 {
|
|
1245 int i;
|
|
1246
|
|
1247 if (len >= buflen)
|
|
1248 {
|
|
1249 buf[0] = NUL;
|
|
1250 return FAIL; /* result will not fit */
|
|
1251 }
|
|
1252
|
|
1253 #ifdef FEAT_MBYTE
|
|
1254 if (has_mbyte)
|
|
1255 {
|
|
1256 int c;
|
|
1257 int outi = 0;
|
|
1258
|
|
1259 /* Fold one character at a time. */
|
|
1260 for (i = 0; i < len; i += mb_ptr2len_check(p + i))
|
|
1261 {
|
|
1262 c = mb_ptr2char(p + i);
|
|
1263 if (enc_utf8)
|
|
1264 /* For Unicode case folding is always the same, no need to use
|
|
1265 * the table from the spell file. */
|
|
1266 c = utf_fold(c);
|
|
1267 else if (c < 256)
|
|
1268 /* Use the table from the spell file. */
|
|
1269 c = spelltab.st_fold[c];
|
|
1270 # ifdef HAVE_TOWLOWER
|
|
1271 else
|
|
1272 /* We don't know what to do, fall back to towlower(), it
|
|
1273 * depends on the current locale. */
|
|
1274 c = towlower(c);
|
|
1275 # endif
|
|
1276 if (outi + MB_MAXBYTES > buflen)
|
|
1277 {
|
|
1278 buf[outi] = NUL;
|
|
1279 return FAIL;
|
|
1280 }
|
|
1281 outi += mb_char2bytes(c, buf + outi);
|
|
1282 }
|
|
1283 buf[outi] = NUL;
|
|
1284 }
|
|
1285 else
|
7
|
1286 #endif
|
258
|
1287 {
|
|
1288 /* Be quick for non-multibyte encodings. */
|
|
1289 for (i = 0; i < len; ++i)
|
|
1290 buf[i] = spelltab.st_fold[p[i]];
|
|
1291 buf[i] = NUL;
|
|
1292 }
|
|
1293
|
|
1294 return OK;
|
|
1295 }
|
|
1296
|
|
1297 #endif /* FEAT_SYN_HL */
|
7
|
1298
|
|
1299 /*
|
|
1300 * return TRUE if 'c' is a valid file-name character
|
|
1301 * Assume characters above 0x100 are valid (multi-byte).
|
|
1302 */
|
|
1303 int
|
|
1304 vim_isfilec(c)
|
|
1305 int c;
|
|
1306 {
|
|
1307 return (c >= 0x100 || (c > 0 && (chartab[c] & CT_FNAME_CHAR)));
|
|
1308 }
|
|
1309
|
|
1310 /*
|
|
1311 * return TRUE if 'c' is a printable character
|
|
1312 * Assume characters above 0x100 are printable (multi-byte), except for
|
|
1313 * Unicode.
|
|
1314 */
|
|
1315 int
|
|
1316 vim_isprintc(c)
|
|
1317 int c;
|
|
1318 {
|
|
1319 #ifdef FEAT_MBYTE
|
|
1320 if (enc_utf8 && c >= 0x100)
|
|
1321 return utf_printable(c);
|
|
1322 #endif
|
|
1323 return (c >= 0x100 || (c > 0 && (chartab[c] & CT_PRINT_CHAR)));
|
|
1324 }
|
|
1325
|
|
1326 /*
|
|
1327 * Strict version of vim_isprintc(c), don't return TRUE if "c" is the head
|
|
1328 * byte of a double-byte character.
|
|
1329 */
|
|
1330 int
|
|
1331 vim_isprintc_strict(c)
|
|
1332 int c;
|
|
1333 {
|
|
1334 #ifdef FEAT_MBYTE
|
|
1335 if (enc_dbcs != 0 && c < 0x100 && MB_BYTE2LEN(c) > 1)
|
|
1336 return FALSE;
|
|
1337 if (enc_utf8 && c >= 0x100)
|
|
1338 return utf_printable(c);
|
|
1339 #endif
|
|
1340 return (c >= 0x100 || (c > 0 && (chartab[c] & CT_PRINT_CHAR)));
|
|
1341 }
|
|
1342
|
|
1343 /*
|
|
1344 * like chartabsize(), but also check for line breaks on the screen
|
|
1345 */
|
|
1346 int
|
|
1347 lbr_chartabsize(s, col)
|
|
1348 unsigned char *s;
|
|
1349 colnr_T col;
|
|
1350 {
|
|
1351 #ifdef FEAT_LINEBREAK
|
|
1352 if (!curwin->w_p_lbr && *p_sbr == NUL)
|
|
1353 {
|
|
1354 #endif
|
|
1355 #ifdef FEAT_MBYTE
|
|
1356 if (curwin->w_p_wrap)
|
|
1357 return win_nolbr_chartabsize(curwin, s, col, NULL);
|
|
1358 #endif
|
|
1359 RET_WIN_BUF_CHARTABSIZE(curwin, curbuf, s, col)
|
|
1360 #ifdef FEAT_LINEBREAK
|
|
1361 }
|
|
1362 return win_lbr_chartabsize(curwin, s, col, NULL);
|
|
1363 #endif
|
|
1364 }
|
|
1365
|
|
1366 /*
|
|
1367 * Call lbr_chartabsize() and advance the pointer.
|
|
1368 */
|
|
1369 int
|
|
1370 lbr_chartabsize_adv(s, col)
|
|
1371 char_u **s;
|
|
1372 colnr_T col;
|
|
1373 {
|
|
1374 int retval;
|
|
1375
|
|
1376 retval = lbr_chartabsize(*s, col);
|
39
|
1377 mb_ptr_adv(*s);
|
7
|
1378 return retval;
|
|
1379 }
|
|
1380
|
|
1381 /*
|
|
1382 * This function is used very often, keep it fast!!!!
|
|
1383 *
|
|
1384 * If "headp" not NULL, set *headp to the size of what we for 'showbreak'
|
|
1385 * string at start of line. Warning: *headp is only set if it's a non-zero
|
|
1386 * value, init to 0 before calling.
|
|
1387 */
|
|
1388 /*ARGSUSED*/
|
|
1389 int
|
|
1390 win_lbr_chartabsize(wp, s, col, headp)
|
|
1391 win_T *wp;
|
|
1392 char_u *s;
|
|
1393 colnr_T col;
|
|
1394 int *headp;
|
|
1395 {
|
|
1396 #ifdef FEAT_LINEBREAK
|
|
1397 int c;
|
|
1398 int size;
|
|
1399 colnr_T col2;
|
|
1400 colnr_T colmax;
|
|
1401 int added;
|
|
1402 # ifdef FEAT_MBYTE
|
|
1403 int mb_added = 0;
|
|
1404 # else
|
|
1405 # define mb_added 0
|
|
1406 # endif
|
|
1407 int numberextra;
|
|
1408 char_u *ps;
|
|
1409 int tab_corr = (*s == TAB);
|
236
|
1410 int n;
|
7
|
1411
|
|
1412 /*
|
|
1413 * No 'linebreak' and 'showbreak': return quickly.
|
|
1414 */
|
|
1415 if (!wp->w_p_lbr && *p_sbr == NUL)
|
|
1416 #endif
|
|
1417 {
|
|
1418 #ifdef FEAT_MBYTE
|
|
1419 if (wp->w_p_wrap)
|
|
1420 return win_nolbr_chartabsize(wp, s, col, headp);
|
|
1421 #endif
|
|
1422 RET_WIN_BUF_CHARTABSIZE(wp, wp->w_buffer, s, col)
|
|
1423 }
|
|
1424
|
|
1425 #ifdef FEAT_LINEBREAK
|
|
1426 /*
|
|
1427 * First get normal size, without 'linebreak'
|
|
1428 */
|
|
1429 size = win_chartabsize(wp, s, col);
|
|
1430 c = *s;
|
|
1431
|
|
1432 /*
|
|
1433 * If 'linebreak' set check at a blank before a non-blank if the line
|
|
1434 * needs a break here
|
|
1435 */
|
|
1436 if (wp->w_p_lbr
|
|
1437 && vim_isbreak(c)
|
|
1438 && !vim_isbreak(s[1])
|
|
1439 && !wp->w_p_list
|
|
1440 && wp->w_p_wrap
|
|
1441 # ifdef FEAT_VERTSPLIT
|
|
1442 && wp->w_width != 0
|
|
1443 # endif
|
|
1444 )
|
|
1445 {
|
|
1446 /*
|
|
1447 * Count all characters from first non-blank after a blank up to next
|
|
1448 * non-blank after a blank.
|
|
1449 */
|
|
1450 numberextra = win_col_off(wp);
|
|
1451 col2 = col;
|
|
1452 colmax = W_WIDTH(wp) - numberextra;
|
|
1453 if (col >= colmax)
|
236
|
1454 {
|
|
1455 n = colmax + win_col_off2(wp);
|
|
1456 if (n > 0)
|
|
1457 colmax += (((col - colmax) / n) + 1) * n;
|
|
1458 }
|
|
1459
|
7
|
1460 for (;;)
|
|
1461 {
|
|
1462 ps = s;
|
39
|
1463 mb_ptr_adv(s);
|
7
|
1464 c = *s;
|
|
1465 if (!(c != NUL
|
|
1466 && (vim_isbreak(c)
|
|
1467 || (!vim_isbreak(c)
|
|
1468 && (col2 == col || !vim_isbreak(*ps))))))
|
|
1469 break;
|
|
1470
|
|
1471 col2 += win_chartabsize(wp, s, col2);
|
|
1472 if (col2 >= colmax) /* doesn't fit */
|
|
1473 {
|
|
1474 size = colmax - col;
|
|
1475 tab_corr = FALSE;
|
|
1476 break;
|
|
1477 }
|
|
1478 }
|
|
1479 }
|
|
1480 # ifdef FEAT_MBYTE
|
|
1481 else if (has_mbyte && size == 2 && MB_BYTE2LEN(*s) > 1
|
|
1482 && wp->w_p_wrap && in_win_border(wp, col))
|
|
1483 {
|
|
1484 ++size; /* Count the ">" in the last column. */
|
|
1485 mb_added = 1;
|
|
1486 }
|
|
1487 # endif
|
|
1488
|
|
1489 /*
|
|
1490 * May have to add something for 'showbreak' string at start of line
|
|
1491 * Set *headp to the size of what we add.
|
|
1492 */
|
|
1493 added = 0;
|
|
1494 if (*p_sbr != NUL && wp->w_p_wrap && col != 0)
|
|
1495 {
|
|
1496 numberextra = win_col_off(wp);
|
|
1497 col += numberextra + mb_added;
|
|
1498 if (col >= (colnr_T)W_WIDTH(wp))
|
|
1499 {
|
|
1500 col -= W_WIDTH(wp);
|
|
1501 numberextra = W_WIDTH(wp) - (numberextra - win_col_off2(wp));
|
|
1502 if (numberextra > 0)
|
|
1503 col = col % numberextra;
|
|
1504 }
|
|
1505 if (col == 0 || col + size > (colnr_T)W_WIDTH(wp))
|
|
1506 {
|
|
1507 added = vim_strsize(p_sbr);
|
|
1508 if (tab_corr)
|
|
1509 size += (added / wp->w_buffer->b_p_ts) * wp->w_buffer->b_p_ts;
|
|
1510 else
|
|
1511 size += added;
|
|
1512 if (col != 0)
|
|
1513 added = 0;
|
|
1514 }
|
|
1515 }
|
|
1516 if (headp != NULL)
|
|
1517 *headp = added + mb_added;
|
|
1518 return size;
|
|
1519 #endif
|
|
1520 }
|
|
1521
|
|
1522 #if defined(FEAT_MBYTE) || defined(PROTO)
|
|
1523 /*
|
|
1524 * Like win_lbr_chartabsize(), except that we know 'linebreak' is off and
|
|
1525 * 'wrap' is on. This means we need to check for a double-byte character that
|
|
1526 * doesn't fit at the end of the screen line.
|
|
1527 */
|
|
1528 static int
|
|
1529 win_nolbr_chartabsize(wp, s, col, headp)
|
|
1530 win_T *wp;
|
|
1531 char_u *s;
|
|
1532 colnr_T col;
|
|
1533 int *headp;
|
|
1534 {
|
|
1535 int n;
|
|
1536
|
|
1537 if (*s == TAB && (!wp->w_p_list || lcs_tab1))
|
|
1538 {
|
|
1539 n = wp->w_buffer->b_p_ts;
|
|
1540 return (int)(n - (col % n));
|
|
1541 }
|
|
1542 n = ptr2cells(s);
|
|
1543 /* Add one cell for a double-width character in the last column of the
|
|
1544 * window, displayed with a ">". */
|
|
1545 if (n == 2 && MB_BYTE2LEN(*s) > 1 && in_win_border(wp, col))
|
|
1546 {
|
|
1547 if (headp != NULL)
|
|
1548 *headp = 1;
|
|
1549 return 3;
|
|
1550 }
|
|
1551 return n;
|
|
1552 }
|
|
1553
|
|
1554 /*
|
|
1555 * Return TRUE if virtual column "vcol" is in the rightmost column of window
|
|
1556 * "wp".
|
|
1557 */
|
|
1558 int
|
|
1559 in_win_border(wp, vcol)
|
|
1560 win_T *wp;
|
|
1561 colnr_T vcol;
|
|
1562 {
|
|
1563 colnr_T width1; /* width of first line (after line number) */
|
|
1564 colnr_T width2; /* width of further lines */
|
|
1565
|
|
1566 #ifdef FEAT_VERTSPLIT
|
|
1567 if (wp->w_width == 0) /* there is no border */
|
|
1568 return FALSE;
|
|
1569 #endif
|
|
1570 width1 = W_WIDTH(wp) - win_col_off(wp);
|
|
1571 if (vcol < width1 - 1)
|
|
1572 return FALSE;
|
|
1573 if (vcol == width1 - 1)
|
|
1574 return TRUE;
|
|
1575 width2 = width1 + win_col_off2(wp);
|
|
1576 return ((vcol - width1) % width2 == width2 - 1);
|
|
1577 }
|
|
1578 #endif /* FEAT_MBYTE */
|
|
1579
|
|
1580 /*
|
|
1581 * Get virtual column number of pos.
|
|
1582 * start: on the first position of this character (TAB, ctrl)
|
|
1583 * cursor: where the cursor is on this character (first char, except for TAB)
|
|
1584 * end: on the last position of this character (TAB, ctrl)
|
|
1585 *
|
|
1586 * This is used very often, keep it fast!
|
|
1587 */
|
|
1588 void
|
|
1589 getvcol(wp, pos, start, cursor, end)
|
|
1590 win_T *wp;
|
|
1591 pos_T *pos;
|
|
1592 colnr_T *start;
|
|
1593 colnr_T *cursor;
|
|
1594 colnr_T *end;
|
|
1595 {
|
|
1596 colnr_T vcol;
|
|
1597 char_u *ptr; /* points to current char */
|
|
1598 char_u *posptr; /* points to char at pos->col */
|
|
1599 int incr;
|
|
1600 int head;
|
|
1601 int ts = wp->w_buffer->b_p_ts;
|
|
1602 int c;
|
|
1603
|
|
1604 vcol = 0;
|
|
1605 ptr = ml_get_buf(wp->w_buffer, pos->lnum, FALSE);
|
|
1606 posptr = ptr + pos->col;
|
|
1607
|
|
1608 /*
|
|
1609 * This function is used very often, do some speed optimizations.
|
|
1610 * When 'list', 'linebreak' and 'showbreak' are not set use a simple loop.
|
|
1611 * Also use this when 'list' is set but tabs take their normal size.
|
|
1612 */
|
|
1613 if ((!wp->w_p_list || lcs_tab1 != NUL)
|
|
1614 #ifdef FEAT_LINEBREAK
|
|
1615 && !wp->w_p_lbr && *p_sbr == NUL
|
|
1616 #endif
|
|
1617 )
|
|
1618 {
|
|
1619 #ifndef FEAT_MBYTE
|
|
1620 head = 0;
|
|
1621 #endif
|
|
1622 for (;;)
|
|
1623 {
|
|
1624 #ifdef FEAT_MBYTE
|
|
1625 head = 0;
|
|
1626 #endif
|
|
1627 c = *ptr;
|
|
1628 /* make sure we don't go past the end of the line */
|
|
1629 if (c == NUL)
|
|
1630 {
|
|
1631 incr = 1; /* NUL at end of line only takes one column */
|
|
1632 break;
|
|
1633 }
|
|
1634 /* A tab gets expanded, depending on the current column */
|
|
1635 if (c == TAB)
|
|
1636 incr = ts - (vcol % ts);
|
|
1637 else
|
|
1638 {
|
|
1639 #ifdef FEAT_MBYTE
|
|
1640 if (has_mbyte)
|
|
1641 {
|
|
1642 /* For utf-8, if the byte is >= 0x80, need to look at
|
|
1643 * further bytes to find the cell width. */
|
|
1644 if (enc_utf8 && c >= 0x80)
|
|
1645 incr = utf_ptr2cells(ptr);
|
|
1646 else
|
|
1647 incr = CHARSIZE(c);
|
|
1648
|
|
1649 /* If a double-cell char doesn't fit at the end of a line
|
|
1650 * it wraps to the next line, it's like this char is three
|
|
1651 * cells wide. */
|
|
1652 if (incr == 2 && wp->w_p_wrap && in_win_border(wp, vcol))
|
|
1653 {
|
|
1654 ++incr;
|
|
1655 head = 1;
|
|
1656 }
|
|
1657 }
|
|
1658 else
|
|
1659 #endif
|
|
1660 incr = CHARSIZE(c);
|
|
1661 }
|
|
1662
|
|
1663 if (ptr >= posptr) /* character at pos->col */
|
|
1664 break;
|
|
1665
|
|
1666 vcol += incr;
|
39
|
1667 mb_ptr_adv(ptr);
|
7
|
1668 }
|
|
1669 }
|
|
1670 else
|
|
1671 {
|
|
1672 for (;;)
|
|
1673 {
|
|
1674 /* A tab gets expanded, depending on the current column */
|
|
1675 head = 0;
|
|
1676 incr = win_lbr_chartabsize(wp, ptr, vcol, &head);
|
|
1677 /* make sure we don't go past the end of the line */
|
|
1678 if (*ptr == NUL)
|
|
1679 {
|
|
1680 incr = 1; /* NUL at end of line only takes one column */
|
|
1681 break;
|
|
1682 }
|
|
1683
|
|
1684 if (ptr >= posptr) /* character at pos->col */
|
|
1685 break;
|
|
1686
|
|
1687 vcol += incr;
|
39
|
1688 mb_ptr_adv(ptr);
|
7
|
1689 }
|
|
1690 }
|
|
1691 if (start != NULL)
|
|
1692 *start = vcol + head;
|
|
1693 if (end != NULL)
|
|
1694 *end = vcol + incr - 1;
|
|
1695 if (cursor != NULL)
|
|
1696 {
|
|
1697 if (*ptr == TAB
|
|
1698 && (State & NORMAL)
|
|
1699 && !wp->w_p_list
|
|
1700 && !virtual_active()
|
|
1701 #ifdef FEAT_VISUAL
|
|
1702 && !(VIsual_active
|
|
1703 && (*p_sel == 'e' || ltoreq(*pos, VIsual)))
|
|
1704 #endif
|
|
1705 )
|
|
1706 *cursor = vcol + incr - 1; /* cursor at end */
|
|
1707 else
|
|
1708 *cursor = vcol + head; /* cursor at start */
|
|
1709 }
|
|
1710 }
|
|
1711
|
|
1712 /*
|
|
1713 * Get virtual cursor column in the current window, pretending 'list' is off.
|
|
1714 */
|
|
1715 colnr_T
|
|
1716 getvcol_nolist(posp)
|
|
1717 pos_T *posp;
|
|
1718 {
|
|
1719 int list_save = curwin->w_p_list;
|
|
1720 colnr_T vcol;
|
|
1721
|
|
1722 curwin->w_p_list = FALSE;
|
|
1723 getvcol(curwin, posp, NULL, &vcol, NULL);
|
|
1724 curwin->w_p_list = list_save;
|
|
1725 return vcol;
|
|
1726 }
|
|
1727
|
|
1728 #if defined(FEAT_VIRTUALEDIT) || defined(PROTO)
|
|
1729 /*
|
|
1730 * Get virtual column in virtual mode.
|
|
1731 */
|
|
1732 void
|
|
1733 getvvcol(wp, pos, start, cursor, end)
|
|
1734 win_T *wp;
|
|
1735 pos_T *pos;
|
|
1736 colnr_T *start;
|
|
1737 colnr_T *cursor;
|
|
1738 colnr_T *end;
|
|
1739 {
|
|
1740 colnr_T col;
|
|
1741 colnr_T coladd;
|
|
1742 colnr_T endadd;
|
|
1743 # ifdef FEAT_MBYTE
|
|
1744 char_u *ptr;
|
|
1745 # endif
|
|
1746
|
|
1747 if (virtual_active())
|
|
1748 {
|
|
1749 /* For virtual mode, only want one value */
|
|
1750 getvcol(wp, pos, &col, NULL, NULL);
|
|
1751
|
|
1752 coladd = pos->coladd;
|
|
1753 endadd = 0;
|
|
1754 # ifdef FEAT_MBYTE
|
|
1755 /* Cannot put the cursor on part of a wide character. */
|
|
1756 ptr = ml_get_buf(wp->w_buffer, pos->lnum, FALSE);
|
|
1757 if (pos->col < STRLEN(ptr))
|
|
1758 {
|
|
1759 int c = (*mb_ptr2char)(ptr + pos->col);
|
|
1760
|
|
1761 if (c != TAB && vim_isprintc(c))
|
|
1762 {
|
|
1763 endadd = char2cells(c) - 1;
|
|
1764 if (coladd >= endadd)
|
|
1765 coladd -= endadd;
|
|
1766 else
|
|
1767 coladd = 0;
|
|
1768 }
|
|
1769 }
|
|
1770 # endif
|
|
1771 col += coladd;
|
|
1772 if (start != NULL)
|
|
1773 *start = col;
|
|
1774 if (cursor != NULL)
|
|
1775 *cursor = col;
|
|
1776 if (end != NULL)
|
|
1777 *end = col + endadd;
|
|
1778 }
|
|
1779 else
|
|
1780 getvcol(wp, pos, start, cursor, end);
|
|
1781 }
|
|
1782 #endif
|
|
1783
|
|
1784 #if defined(FEAT_VISUAL) || defined(PROTO)
|
|
1785 /*
|
|
1786 * Get the leftmost and rightmost virtual column of pos1 and pos2.
|
|
1787 * Used for Visual block mode.
|
|
1788 */
|
|
1789 void
|
|
1790 getvcols(wp, pos1, pos2, left, right)
|
|
1791 win_T *wp;
|
|
1792 pos_T *pos1, *pos2;
|
|
1793 colnr_T *left, *right;
|
|
1794 {
|
|
1795 colnr_T from1, from2, to1, to2;
|
|
1796
|
|
1797 if (ltp(pos1, pos2))
|
|
1798 {
|
|
1799 getvvcol(wp, pos1, &from1, NULL, &to1);
|
|
1800 getvvcol(wp, pos2, &from2, NULL, &to2);
|
|
1801 }
|
|
1802 else
|
|
1803 {
|
|
1804 getvvcol(wp, pos2, &from1, NULL, &to1);
|
|
1805 getvvcol(wp, pos1, &from2, NULL, &to2);
|
|
1806 }
|
|
1807 if (from2 < from1)
|
|
1808 *left = from2;
|
|
1809 else
|
|
1810 *left = from1;
|
|
1811 if (to2 > to1)
|
|
1812 {
|
|
1813 if (*p_sel == 'e' && from2 - 1 >= to1)
|
|
1814 *right = from2 - 1;
|
|
1815 else
|
|
1816 *right = to2;
|
|
1817 }
|
|
1818 else
|
|
1819 *right = to1;
|
|
1820 }
|
|
1821 #endif
|
|
1822
|
|
1823 /*
|
|
1824 * skipwhite: skip over ' ' and '\t'.
|
|
1825 */
|
|
1826 char_u *
|
|
1827 skipwhite(p)
|
|
1828 char_u *p;
|
|
1829 {
|
|
1830 while (vim_iswhite(*p)) /* skip to next non-white */
|
|
1831 ++p;
|
|
1832 return p;
|
|
1833 }
|
|
1834
|
|
1835 /*
|
|
1836 * skipdigits: skip over digits;
|
|
1837 */
|
|
1838 char_u *
|
|
1839 skipdigits(p)
|
|
1840 char_u *p;
|
|
1841 {
|
|
1842 while (VIM_ISDIGIT(*p)) /* skip to next non-digit */
|
|
1843 ++p;
|
|
1844 return p;
|
|
1845 }
|
|
1846
|
|
1847 /*
|
|
1848 * Variant of isdigit() that can handle characters > 0x100.
|
|
1849 * We don't use isdigit() here, because on some systems it also considers
|
|
1850 * superscript 1 to be a digit.
|
|
1851 * Use the VIM_ISDIGIT() macro for simple arguments.
|
|
1852 */
|
|
1853 int
|
|
1854 vim_isdigit(c)
|
|
1855 int c;
|
|
1856 {
|
|
1857 return (c >= '0' && c <= '9');
|
|
1858 }
|
|
1859
|
|
1860 /*
|
|
1861 * Variant of isxdigit() that can handle characters > 0x100.
|
|
1862 * We don't use isxdigit() here, because on some systems it also considers
|
|
1863 * superscript 1 to be a digit.
|
|
1864 */
|
|
1865 int
|
|
1866 vim_isxdigit(c)
|
|
1867 int c;
|
|
1868 {
|
|
1869 return (c >= '0' && c <= '9')
|
|
1870 || (c >= 'a' && c <= 'f')
|
|
1871 || (c >= 'A' && c <= 'F');
|
|
1872 }
|
|
1873
|
|
1874 /*
|
|
1875 * skiptowhite: skip over text until ' ' or '\t' or NUL.
|
|
1876 */
|
|
1877 char_u *
|
|
1878 skiptowhite(p)
|
|
1879 char_u *p;
|
|
1880 {
|
|
1881 while (*p != ' ' && *p != '\t' && *p != NUL)
|
|
1882 ++p;
|
|
1883 return p;
|
|
1884 }
|
|
1885
|
|
1886 #if defined(FEAT_LISTCMDS) || defined(FEAT_SIGNS) || defined(FEAT_SNIFF) \
|
|
1887 || defined(PROTO)
|
|
1888 /*
|
|
1889 * skiptowhite_esc: Like skiptowhite(), but also skip escaped chars
|
|
1890 */
|
|
1891 char_u *
|
|
1892 skiptowhite_esc(p)
|
|
1893 char_u *p;
|
|
1894 {
|
|
1895 while (*p != ' ' && *p != '\t' && *p != NUL)
|
|
1896 {
|
|
1897 if ((*p == '\\' || *p == Ctrl_V) && *(p + 1) != NUL)
|
|
1898 ++p;
|
|
1899 ++p;
|
|
1900 }
|
|
1901 return p;
|
|
1902 }
|
|
1903 #endif
|
|
1904
|
|
1905 /*
|
|
1906 * Getdigits: Get a number from a string and skip over it.
|
|
1907 * Note: the argument is a pointer to a char_u pointer!
|
|
1908 */
|
|
1909 long
|
|
1910 getdigits(pp)
|
|
1911 char_u **pp;
|
|
1912 {
|
|
1913 char_u *p;
|
|
1914 long retval;
|
|
1915
|
|
1916 p = *pp;
|
|
1917 retval = atol((char *)p);
|
|
1918 if (*p == '-') /* skip negative sign */
|
|
1919 ++p;
|
|
1920 p = skipdigits(p); /* skip to next non-digit */
|
|
1921 *pp = p;
|
|
1922 return retval;
|
|
1923 }
|
|
1924
|
|
1925 /*
|
|
1926 * Return TRUE if "lbuf" is empty or only contains blanks.
|
|
1927 */
|
|
1928 int
|
|
1929 vim_isblankline(lbuf)
|
|
1930 char_u *lbuf;
|
|
1931 {
|
|
1932 char_u *p;
|
|
1933
|
|
1934 p = skipwhite(lbuf);
|
|
1935 return (*p == NUL || *p == '\r' || *p == '\n');
|
|
1936 }
|
|
1937
|
|
1938 /*
|
|
1939 * Convert a string into a long and/or unsigned long, taking care of
|
16
|
1940 * hexadecimal and octal numbers. Accepts a '-' sign.
|
7
|
1941 * If "hexp" is not NULL, returns a flag to indicate the type of the number:
|
|
1942 * 0 decimal
|
|
1943 * '0' octal
|
|
1944 * 'X' hex
|
|
1945 * 'x' hex
|
|
1946 * If "len" is not NULL, the length of the number in characters is returned.
|
|
1947 * If "nptr" is not NULL, the signed result is returned in it.
|
|
1948 * If "unptr" is not NULL, the unsigned result is returned in it.
|
|
1949 */
|
|
1950 void
|
|
1951 vim_str2nr(start, hexp, len, dooct, dohex, nptr, unptr)
|
|
1952 char_u *start;
|
|
1953 int *hexp; /* return: type of number 0 = decimal, 'x'
|
|
1954 or 'X' is hex, '0' = octal */
|
|
1955 int *len; /* return: detected length of number */
|
|
1956 int dooct; /* recognize octal number */
|
|
1957 int dohex; /* recognize hex number */
|
|
1958 long *nptr; /* return: signed result */
|
|
1959 unsigned long *unptr; /* return: unsigned result */
|
|
1960 {
|
|
1961 char_u *ptr = start;
|
|
1962 int hex = 0; /* default is decimal */
|
|
1963 int negative = FALSE;
|
|
1964 unsigned long un = 0;
|
39
|
1965 int n;
|
7
|
1966
|
|
1967 if (ptr[0] == '-')
|
|
1968 {
|
|
1969 negative = TRUE;
|
|
1970 ++ptr;
|
|
1971 }
|
|
1972
|
39
|
1973 /* Recognize hex and octal. */
|
|
1974 if (ptr[0] == '0' && ptr[1] != '8' && ptr[1] != '9')
|
7
|
1975 {
|
|
1976 hex = ptr[1];
|
|
1977 if (dohex && (hex == 'X' || hex == 'x') && vim_isxdigit(ptr[2]))
|
|
1978 ptr += 2; /* hexadecimal */
|
|
1979 else
|
|
1980 {
|
39
|
1981 hex = 0; /* default is decimal */
|
|
1982 if (dooct)
|
|
1983 {
|
|
1984 /* Don't interpret "0", "08" or "0129" as octal. */
|
|
1985 for (n = 1; VIM_ISDIGIT(ptr[n]); ++n)
|
|
1986 {
|
|
1987 if (ptr[n] > '7')
|
|
1988 {
|
|
1989 hex = 0; /* can't be octal */
|
|
1990 break;
|
|
1991 }
|
|
1992 if (ptr[n] > '0')
|
|
1993 hex = '0'; /* assume octal */
|
|
1994 }
|
|
1995 }
|
7
|
1996 }
|
|
1997 }
|
|
1998
|
|
1999 /*
|
|
2000 * Do the string-to-numeric conversion "manually" to avoid sscanf quirks.
|
|
2001 */
|
|
2002 if (hex)
|
|
2003 {
|
|
2004 if (hex == '0')
|
|
2005 {
|
|
2006 /* octal */
|
|
2007 while ('0' <= *ptr && *ptr <= '7')
|
|
2008 {
|
|
2009 un = 8 * un + (unsigned long)(*ptr - '0');
|
|
2010 ++ptr;
|
|
2011 }
|
|
2012 }
|
|
2013 else
|
|
2014 {
|
|
2015 /* hex */
|
|
2016 while (vim_isxdigit(*ptr))
|
|
2017 {
|
|
2018 un = 16 * un + (unsigned long)hex2nr(*ptr);
|
|
2019 ++ptr;
|
|
2020 }
|
|
2021 }
|
|
2022 }
|
|
2023 else
|
|
2024 {
|
|
2025 /* decimal */
|
|
2026 while (VIM_ISDIGIT(*ptr))
|
|
2027 {
|
|
2028 un = 10 * un + (unsigned long)(*ptr - '0');
|
|
2029 ++ptr;
|
|
2030 }
|
|
2031 }
|
|
2032
|
|
2033 if (hexp != NULL)
|
|
2034 *hexp = hex;
|
|
2035 if (len != NULL)
|
|
2036 *len = (int)(ptr - start);
|
|
2037 if (nptr != NULL)
|
16
|
2038 {
|
|
2039 if (negative) /* account for leading '-' for decimal numbers */
|
|
2040 *nptr = -(long)un;
|
|
2041 else
|
|
2042 *nptr = (long)un;
|
|
2043 }
|
7
|
2044 if (unptr != NULL)
|
|
2045 *unptr = un;
|
|
2046 }
|
|
2047
|
|
2048 /*
|
|
2049 * Return the value of a single hex character.
|
|
2050 * Only valid when the argument is '0' - '9', 'A' - 'F' or 'a' - 'f'.
|
|
2051 */
|
|
2052 int
|
|
2053 hex2nr(c)
|
|
2054 int c;
|
|
2055 {
|
|
2056 if (c >= 'a' && c <= 'f')
|
|
2057 return c - 'a' + 10;
|
|
2058 if (c >= 'A' && c <= 'F')
|
|
2059 return c - 'A' + 10;
|
|
2060 return c - '0';
|
|
2061 }
|
|
2062
|
|
2063 #if defined(FEAT_TERMRESPONSE) \
|
|
2064 || (defined(FEAT_GUI_GTK) && defined(FEAT_WINDOWS)) || defined(PROTO)
|
|
2065 /*
|
|
2066 * Convert two hex characters to a byte.
|
|
2067 * Return -1 if one of the characters is not hex.
|
|
2068 */
|
|
2069 int
|
|
2070 hexhex2nr(p)
|
|
2071 char_u *p;
|
|
2072 {
|
|
2073 if (!vim_isxdigit(p[0]) || !vim_isxdigit(p[1]))
|
|
2074 return -1;
|
|
2075 return (hex2nr(p[0]) << 4) + hex2nr(p[1]);
|
|
2076 }
|
|
2077 #endif
|
|
2078
|
|
2079 /*
|
|
2080 * Return TRUE if "str" starts with a backslash that should be removed.
|
|
2081 * For MS-DOS, WIN32 and OS/2 this is only done when the character after the
|
|
2082 * backslash is not a normal file name character.
|
|
2083 * '$' is a valid file name character, we don't remove the backslash before
|
|
2084 * it. This means it is not possible to use an environment variable after a
|
|
2085 * backslash. "C:\$VIM\doc" is taken literally, only "$VIM\doc" works.
|
|
2086 * Although "\ name" is valid, the backslash in "Program\ files" must be
|
|
2087 * removed. Assume a file name doesn't start with a space.
|
|
2088 * For multi-byte names, never remove a backslash before a non-ascii
|
|
2089 * character, assume that all multi-byte characters are valid file name
|
|
2090 * characters.
|
|
2091 */
|
|
2092 int
|
|
2093 rem_backslash(str)
|
|
2094 char_u *str;
|
|
2095 {
|
|
2096 #ifdef BACKSLASH_IN_FILENAME
|
|
2097 return (str[0] == '\\'
|
|
2098 # ifdef FEAT_MBYTE
|
|
2099 && str[1] < 0x80
|
|
2100 # endif
|
|
2101 && (str[1] == ' '
|
|
2102 || (str[1] != NUL
|
|
2103 && str[1] != '*'
|
|
2104 && str[1] != '?'
|
|
2105 && !vim_isfilec(str[1]))));
|
|
2106 #else
|
|
2107 return (str[0] == '\\' && str[1] != NUL);
|
|
2108 #endif
|
|
2109 }
|
|
2110
|
|
2111 /*
|
|
2112 * Halve the number of backslashes in a file name argument.
|
|
2113 * For MS-DOS we only do this if the character after the backslash
|
|
2114 * is not a normal file character.
|
|
2115 */
|
|
2116 void
|
|
2117 backslash_halve(p)
|
|
2118 char_u *p;
|
|
2119 {
|
|
2120 for ( ; *p; ++p)
|
|
2121 if (rem_backslash(p))
|
|
2122 STRCPY(p, p + 1);
|
|
2123 }
|
|
2124
|
|
2125 /*
|
|
2126 * backslash_halve() plus save the result in allocated memory.
|
|
2127 */
|
|
2128 char_u *
|
|
2129 backslash_halve_save(p)
|
|
2130 char_u *p;
|
|
2131 {
|
|
2132 char_u *res;
|
|
2133
|
|
2134 res = vim_strsave(p);
|
|
2135 if (res == NULL)
|
|
2136 return p;
|
|
2137 backslash_halve(res);
|
|
2138 return res;
|
|
2139 }
|
|
2140
|
|
2141 #if (defined(EBCDIC) && defined(FEAT_POSTSCRIPT)) || defined(PROTO)
|
|
2142 /*
|
|
2143 * Table for EBCDIC to ASCII conversion unashamedly taken from xxd.c!
|
|
2144 * The first 64 entries have been added to map control characters defined in
|
|
2145 * ascii.h
|
|
2146 */
|
|
2147 static char_u ebcdic2ascii_tab[256] =
|
|
2148 {
|
|
2149 0000, 0001, 0002, 0003, 0004, 0011, 0006, 0177,
|
|
2150 0010, 0011, 0012, 0013, 0014, 0015, 0016, 0017,
|
|
2151 0020, 0021, 0022, 0023, 0024, 0012, 0010, 0027,
|
|
2152 0030, 0031, 0032, 0033, 0033, 0035, 0036, 0037,
|
|
2153 0040, 0041, 0042, 0043, 0044, 0045, 0046, 0047,
|
|
2154 0050, 0051, 0052, 0053, 0054, 0055, 0056, 0057,
|
|
2155 0060, 0061, 0062, 0063, 0064, 0065, 0066, 0067,
|
|
2156 0070, 0071, 0072, 0073, 0074, 0075, 0076, 0077,
|
|
2157 0040, 0240, 0241, 0242, 0243, 0244, 0245, 0246,
|
|
2158 0247, 0250, 0325, 0056, 0074, 0050, 0053, 0174,
|
|
2159 0046, 0251, 0252, 0253, 0254, 0255, 0256, 0257,
|
|
2160 0260, 0261, 0041, 0044, 0052, 0051, 0073, 0176,
|
|
2161 0055, 0057, 0262, 0263, 0264, 0265, 0266, 0267,
|
|
2162 0270, 0271, 0313, 0054, 0045, 0137, 0076, 0077,
|
|
2163 0272, 0273, 0274, 0275, 0276, 0277, 0300, 0301,
|
|
2164 0302, 0140, 0072, 0043, 0100, 0047, 0075, 0042,
|
|
2165 0303, 0141, 0142, 0143, 0144, 0145, 0146, 0147,
|
|
2166 0150, 0151, 0304, 0305, 0306, 0307, 0310, 0311,
|
|
2167 0312, 0152, 0153, 0154, 0155, 0156, 0157, 0160,
|
|
2168 0161, 0162, 0136, 0314, 0315, 0316, 0317, 0320,
|
|
2169 0321, 0345, 0163, 0164, 0165, 0166, 0167, 0170,
|
|
2170 0171, 0172, 0322, 0323, 0324, 0133, 0326, 0327,
|
|
2171 0330, 0331, 0332, 0333, 0334, 0335, 0336, 0337,
|
|
2172 0340, 0341, 0342, 0343, 0344, 0135, 0346, 0347,
|
|
2173 0173, 0101, 0102, 0103, 0104, 0105, 0106, 0107,
|
|
2174 0110, 0111, 0350, 0351, 0352, 0353, 0354, 0355,
|
|
2175 0175, 0112, 0113, 0114, 0115, 0116, 0117, 0120,
|
|
2176 0121, 0122, 0356, 0357, 0360, 0361, 0362, 0363,
|
|
2177 0134, 0237, 0123, 0124, 0125, 0126, 0127, 0130,
|
|
2178 0131, 0132, 0364, 0365, 0366, 0367, 0370, 0371,
|
|
2179 0060, 0061, 0062, 0063, 0064, 0065, 0066, 0067,
|
|
2180 0070, 0071, 0372, 0373, 0374, 0375, 0376, 0377
|
|
2181 };
|
|
2182
|
|
2183 /*
|
|
2184 * Convert a buffer worth of characters from EBCDIC to ASCII. Only useful if
|
|
2185 * wanting 7-bit ASCII characters out the other end.
|
|
2186 */
|
|
2187 void
|
|
2188 ebcdic2ascii(buffer, len)
|
|
2189 char_u *buffer;
|
|
2190 int len;
|
|
2191 {
|
|
2192 int i;
|
|
2193
|
|
2194 for (i = 0; i < len; i++)
|
|
2195 buffer[i] = ebcdic2ascii_tab[buffer[i]];
|
|
2196 }
|
|
2197 #endif
|