1 /*
2 * Copyright (C) 1984-2024 Mark Nudelman
3 *
4 * You may distribute under the terms of either the GNU General Public
5 * License or the Less License, as specified in the README file.
6 *
7 * For more information, see the README file.
8 */
9
10
11 /*
12 * Functions to define the character set
13 * and do things specific to the character set.
14 */
15
16 #include "less.h"
17 #if HAVE_LOCALE
18 #include <locale.h>
19 #include <ctype.h>
20 #include <langinfo.h>
21 #endif
22
23 #include "charset.h"
24 #include "xbuf.h"
25
26 #if MSDOS_COMPILER==WIN32C
27 #define WIN32_LEAN_AND_MEAN
28 #include <windows.h>
29 #endif
30
31 extern int bs_mode;
32
33 public int utf_mode = 0;
34
35 /*
36 * Predefined character sets,
37 * selected by the LESSCHARSET environment variable.
38 */
39 struct charset {
40 char *name;
41 int *p_flag;
42 char *desc;
43 } charsets[] = {
44 { "ascii", NULL, "8bcccbcc18b95.b" },
45 { "utf-8", &utf_mode, "8bcccbcc18b95.b126.bb" },
46 { "iso8859", NULL, "8bcccbcc18b95.33b." },
47 { "latin3", NULL, "8bcccbcc18b95.33b5.b8.b15.b4.b12.b18.b12.b." },
48 { "arabic", NULL, "8bcccbcc18b95.33b.3b.7b2.13b.3b.b26.5b19.b" },
49 { "greek", NULL, "8bcccbcc18b95.33b4.2b4.b3.b35.b44.b" },
50 { "greek2005", NULL, "8bcccbcc18b95.33b14.b35.b44.b" },
51 { "hebrew", NULL, "8bcccbcc18b95.33b.b29.32b28.2b2.b" },
52 { "koi8-r", NULL, "8bcccbcc18b95.b." },
53 { "KOI8-T", NULL, "8bcccbcc18b95.b8.b6.b8.b.b.5b7.3b4.b4.b3.b.b.3b." },
54 { "georgianps", NULL, "8bcccbcc18b95.3b11.4b12.2b." },
55 { "tcvn", NULL, "b..b...bcccbccbbb7.8b95.b48.5b." },
56 { "TIS-620", NULL, "8bcccbcc18b95.b.4b.11b7.8b." },
57 { "next", NULL, "8bcccbcc18b95.bb125.bb" },
58 { "dos", NULL, "8bcccbcc12bc5b95.b." },
59 { "windows-1251", NULL, "8bcccbcc12bc5b95.b24.b." },
60 { "windows-1252", NULL, "8bcccbcc12bc5b95.b.b11.b.2b12.b." },
61 { "windows-1255", NULL, "8bcccbcc12bc5b95.b.b8.b.5b9.b.4b." },
62 { "ebcdic", NULL, "5bc6bcc7bcc41b.9b7.9b5.b..8b6.10b6.b9.7b9.8b8.17b3.3b9.7b9.8b8.6b10.b.b.b." },
63 { "IBM-1047", NULL, "4cbcbc3b9cbccbccbb4c6bcc5b3cbbc4bc4bccbc191.b" },
64 { NULL, NULL, NULL }
65 };
66
67 /*
68 * Support "locale charmap"/nl_langinfo(CODESET) values, as well as others.
69 */
70 struct cs_alias {
71 char *name;
72 char *oname;
73 } cs_aliases[] = {
74 { "UTF-8", "utf-8" },
75 { "utf8", "utf-8" },
76 { "UTF8", "utf-8" },
77 { "ANSI_X3.4-1968", "ascii" },
78 { "US-ASCII", "ascii" },
79 { "latin1", "iso8859" },
80 { "ISO-8859-1", "iso8859" },
81 { "latin9", "iso8859" },
82 { "ISO-8859-15", "iso8859" },
83 { "latin2", "iso8859" },
84 { "ISO-8859-2", "iso8859" },
85 { "ISO-8859-3", "latin3" },
86 { "latin4", "iso8859" },
87 { "ISO-8859-4", "iso8859" },
88 { "cyrillic", "iso8859" },
89 { "ISO-8859-5", "iso8859" },
90 { "ISO-8859-6", "arabic" },
91 { "ISO-8859-7", "greek" },
92 { "IBM9005", "greek2005" },
93 { "ISO-8859-8", "hebrew" },
94 { "latin5", "iso8859" },
95 { "ISO-8859-9", "iso8859" },
96 { "latin6", "iso8859" },
97 { "ISO-8859-10", "iso8859" },
98 { "latin7", "iso8859" },
99 { "ISO-8859-13", "iso8859" },
100 { "latin8", "iso8859" },
101 { "ISO-8859-14", "iso8859" },
102 { "latin10", "iso8859" },
103 { "ISO-8859-16", "iso8859" },
104 { "IBM437", "dos" },
105 { "EBCDIC-US", "ebcdic" },
106 { "IBM1047", "IBM-1047" },
107 { "KOI8-R", "koi8-r" },
108 { "KOI8-U", "koi8-r" },
109 { "GEORGIAN-PS", "georgianps" },
110 { "TCVN5712-1", "tcvn" },
111 { "NEXTSTEP", "next" },
112 { "windows", "windows-1252" }, /* backward compatibility */
113 { "CP1251", "windows-1251" },
114 { "CP1252", "windows-1252" },
115 { "CP1255", "windows-1255" },
116 { NULL, NULL }
117 };
118
119 #define IS_BINARY_CHAR 01
120 #define IS_CONTROL_CHAR 02
121
122 static char chardef[256];
123 static constant char *binfmt = NULL;
124 static constant char *utfbinfmt = NULL;
125 public int binattr = AT_STANDOUT|AT_COLOR_BIN;
126
127 static struct xbuffer user_wide_array;
128 static struct xbuffer user_ubin_array;
129 static struct xbuffer user_compose_array;
130 static struct xbuffer user_prt_array;
131 static struct wchar_range_table user_wide_table;
132 static struct wchar_range_table user_ubin_table;
133 static struct wchar_range_table user_compose_table;
134 static struct wchar_range_table user_prt_table;
135
136 /*
137 * Set a wchar_range_table to the table in an xbuffer.
138 */
wchar_range_table_set(struct wchar_range_table * tbl,struct xbuffer * arr)139 static void wchar_range_table_set(struct wchar_range_table *tbl, struct xbuffer *arr)
140 {
141 tbl->table = (struct wchar_range *) arr->data;
142 tbl->count = (unsigned int) (arr->end / sizeof(struct wchar_range));
143 }
144
145 /*
146 * Skip over a "U" or "U+" prefix before a hex codepoint.
147 */
skip_uprefix(constant char * s)148 static constant char * skip_uprefix(constant char *s)
149 {
150 if (*s == 'U' || *s == 'u')
151 if (*++s == '+') ++s;
152 return s;
153 }
154
155 /*
156 * Parse a dash-separated range of hex values.
157 */
wchar_range_get(constant char ** ss,struct wchar_range * range)158 static void wchar_range_get(constant char **ss, struct wchar_range *range)
159 {
160 constant char *s = skip_uprefix(*ss);
161 range->first = lstrtoulc(s, &s, 16);
162 if (s[0] == '-')
163 {
164 s = skip_uprefix(&s[1]);
165 range->last = lstrtoulc(s, &s, 16);
166 } else
167 {
168 range->last = range->first;
169 }
170 *ss = s;
171 }
172
173 /*
174 * Parse the LESSUTFCHARDEF variable.
175 */
ichardef_utf(constant char * s)176 static void ichardef_utf(constant char *s)
177 {
178 xbuf_init(&user_wide_array);
179 xbuf_init(&user_ubin_array);
180 xbuf_init(&user_compose_array);
181 xbuf_init(&user_prt_array);
182
183 if (s != NULL)
184 {
185 while (s[0] != '\0')
186 {
187 struct wchar_range range;
188 wchar_range_get(&s, &range);
189 if (range.last == 0)
190 {
191 error("invalid hex number(s) in LESSUTFCHARDEF", NULL_PARG);
192 quit(QUIT_ERROR);
193 }
194 if (*s++ != ':')
195 {
196 error("missing colon in LESSUTFCHARDEF", NULL_PARG);
197 quit(QUIT_ERROR);
198 }
199 switch (*s++)
200 {
201 case 'b':
202 xbuf_add_data(&user_ubin_array, (unsigned char *) &range, sizeof(range));
203 break;
204 case 'c':
205 xbuf_add_data(&user_compose_array, (unsigned char *) &range, sizeof(range));
206 break;
207 case 'w':
208 xbuf_add_data(&user_wide_array, (unsigned char *) &range, sizeof(range));
209 xbuf_add_data(&user_prt_array, (unsigned char *) &range, sizeof(range));
210 break;
211 case 'p': case '.':
212 xbuf_add_data(&user_prt_array, (unsigned char *) &range, sizeof(range));
213 break;
214 case '\0':
215 s--;
216 break;
217 default:
218 /* Ignore unknown character attribute. */
219 break;
220 }
221 if (s[0] == ',') ++s;
222 }
223 }
224 wchar_range_table_set(&user_wide_table, &user_wide_array);
225 wchar_range_table_set(&user_ubin_table, &user_ubin_array);
226 wchar_range_table_set(&user_compose_table, &user_compose_array);
227 wchar_range_table_set(&user_prt_table, &user_prt_array);
228 }
229
230 /*
231 * Define a charset, given a description string.
232 * The string consists of 256 letters,
233 * one for each character in the charset.
234 * If the string is shorter than 256 letters, missing letters
235 * are taken to be identical to the last one.
236 * A decimal number followed by a letter is taken to be a
237 * repetition of the letter.
238 *
239 * Each letter is one of:
240 * . normal character
241 * b binary character
242 * c control character
243 */
ichardef(constant char * s)244 static void ichardef(constant char *s)
245 {
246 char *cp;
247 int n;
248 char v;
249
250 n = 0;
251 v = 0;
252 cp = chardef;
253 while (*s != '\0')
254 {
255 switch (*s++)
256 {
257 case '.':
258 v = 0;
259 break;
260 case 'c':
261 v = IS_CONTROL_CHAR;
262 break;
263 case 'b':
264 v = IS_BINARY_CHAR|IS_CONTROL_CHAR;
265 break;
266
267 case '0': case '1': case '2': case '3': case '4':
268 case '5': case '6': case '7': case '8': case '9':
269 if (ckd_mul(&n, n, 10) || ckd_add(&n, n, s[-1] - '0'))
270 goto invalid_chardef;
271 continue;
272
273 default:
274 invalid_chardef:
275 error("invalid chardef", NULL_PARG);
276 quit(QUIT_ERROR);
277 /*NOTREACHED*/
278 }
279
280 do
281 {
282 if (cp >= chardef + sizeof(chardef))
283 {
284 error("chardef longer than 256", NULL_PARG);
285 quit(QUIT_ERROR);
286 /*NOTREACHED*/
287 }
288 *cp++ = v;
289 } while (--n > 0);
290 n = 0;
291 }
292
293 while (cp < chardef + sizeof(chardef))
294 *cp++ = v;
295 }
296
297 /*
298 * Define a charset, given a charset name.
299 * The valid charset names are listed in the "charsets" array.
300 */
icharset(constant char * name,int no_error)301 static int icharset(constant char *name, int no_error)
302 {
303 struct charset *p;
304 struct cs_alias *a;
305
306 if (name == NULL || *name == '\0')
307 return (0);
308
309 /* First see if the name is an alias. */
310 for (a = cs_aliases; a->name != NULL; a++)
311 {
312 if (strcmp(name, a->name) == 0)
313 {
314 name = a->oname;
315 break;
316 }
317 }
318
319 for (p = charsets; p->name != NULL; p++)
320 {
321 if (strcmp(name, p->name) == 0)
322 {
323 ichardef(p->desc);
324 if (p->p_flag != NULL)
325 {
326 #if MSDOS_COMPILER==WIN32C
327 *(p->p_flag) = 1 + (GetConsoleOutputCP() != CP_UTF8);
328 #else
329 *(p->p_flag) = 1;
330 #endif
331 }
332 return (1);
333 }
334 }
335
336 if (!no_error) {
337 error("invalid charset name", NULL_PARG);
338 quit(QUIT_ERROR);
339 }
340 return (0);
341 }
342
343 #if HAVE_LOCALE
344 /*
345 * Define a charset, given a locale name.
346 */
ilocale(void)347 static void ilocale(void)
348 {
349 int c;
350
351 for (c = 0; c < (int) sizeof(chardef); c++)
352 {
353 if (isprint(c))
354 chardef[c] = 0;
355 else if (iscntrl(c))
356 chardef[c] = IS_CONTROL_CHAR;
357 else
358 chardef[c] = IS_BINARY_CHAR|IS_CONTROL_CHAR;
359 }
360 }
361 #endif
362
363 /*
364 * Define the printing format for control (or binary utf) chars.
365 */
setfmt(constant char * s,constant char ** fmtvarptr,int * attrptr,constant char * default_fmt,lbool for_printf)366 public void setfmt(constant char *s, constant char **fmtvarptr, int *attrptr, constant char *default_fmt, lbool for_printf)
367 {
368 if (s == NULL || *s == '\0')
369 s = default_fmt;
370 else if (for_printf &&
371 ((*s == '*' && (s[1] == '\0' || s[2] == '\0' || strchr(s + 2, 'n'))) ||
372 (*s != '*' && strchr(s, 'n'))))
373 /* %n is evil */
374 s = default_fmt;
375
376 /*
377 * Select the attributes if it starts with "*".
378 */
379 if (*s == '*' && s[1] != '\0')
380 {
381 switch (s[1])
382 {
383 case 'd': *attrptr = AT_BOLD; break;
384 case 'k': *attrptr = AT_BLINK; break;
385 case 's': *attrptr = AT_STANDOUT; break;
386 case 'u': *attrptr = AT_UNDERLINE; break;
387 default: *attrptr = AT_NORMAL; break;
388 }
389 s += 2;
390 }
391 *fmtvarptr = s;
392 }
393
394 /*
395 *
396 */
set_charset(void)397 static void set_charset(void)
398 {
399 constant char *s;
400
401 ichardef_utf(lgetenv("LESSUTFCHARDEF"));
402
403 /*
404 * See if environment variable LESSCHARSET is defined.
405 */
406 s = lgetenv("LESSCHARSET");
407 if (icharset(s, 0))
408 return;
409
410 /*
411 * LESSCHARSET is not defined: try LESSCHARDEF.
412 */
413 s = lgetenv("LESSCHARDEF");
414 if (!isnullenv(s))
415 {
416 ichardef(s);
417 return;
418 }
419
420 #if HAVE_LOCALE
421 #ifdef CODESET
422 /*
423 * Try using the codeset name as the charset name.
424 */
425 s = nl_langinfo(CODESET);
426 if (icharset(s, 1))
427 return;
428 #endif
429 #endif
430
431 #if HAVE_STRSTR
432 /*
433 * Check whether LC_ALL, LC_CTYPE or LANG look like UTF-8 is used.
434 */
435 if ((s = lgetenv("LC_ALL")) != NULL ||
436 (s = lgetenv("LC_CTYPE")) != NULL ||
437 (s = lgetenv("LANG")) != NULL)
438 {
439 if ( strstr(s, "UTF-8") != NULL || strstr(s, "utf-8") != NULL
440 || strstr(s, "UTF8") != NULL || strstr(s, "utf8") != NULL)
441 if (icharset("utf-8", 1))
442 return;
443 }
444 #endif
445
446 #if HAVE_LOCALE
447 /*
448 * Get character definitions from locale functions,
449 * rather than from predefined charset entry.
450 */
451 ilocale();
452 #else
453 #if MSDOS_COMPILER
454 #if MSDOS_COMPILER==WIN32C
455 (void) icharset("utf-8", 1);
456 #else
457 (void) icharset("dos", 1);
458 #endif
459 #else
460 (void) icharset("utf-8", 1);
461 #endif
462 #endif
463 }
464
465 /*
466 * Initialize charset data structures.
467 */
init_charset(void)468 public void init_charset(void)
469 {
470 constant char *s;
471
472 #if HAVE_LOCALE
473 setlocale(LC_ALL, "");
474 #endif
475
476 set_charset();
477
478 s = lgetenv("LESSBINFMT");
479 setfmt(s, &binfmt, &binattr, "*s<%02X>", TRUE);
480
481 s = lgetenv("LESSUTFBINFMT");
482 setfmt(s, &utfbinfmt, &binattr, "<U+%04lX>", TRUE);
483 }
484
485 /*
486 * Is a given character a "binary" character?
487 */
binary_char(LWCHAR c)488 public lbool binary_char(LWCHAR c)
489 {
490 if (utf_mode)
491 return (is_ubin_char(c));
492 if (c >= sizeof(chardef))
493 return TRUE;
494 return ((chardef[c] & IS_BINARY_CHAR) != 0);
495 }
496
497 /*
498 * Is a given character a "control" character?
499 */
control_char(LWCHAR c)500 public lbool control_char(LWCHAR c)
501 {
502 if (c >= sizeof(chardef))
503 return TRUE;
504 return (chardef[c] & IS_CONTROL_CHAR);
505 }
506
507 /*
508 * Return the printable form of a character.
509 * For example, in the "ascii" charset '\3' is printed as "^C".
510 */
prchar(LWCHAR c)511 public constant char * prchar(LWCHAR c)
512 {
513 /* {{ Fixed buffer size means LESSBINFMT etc can be truncated. }} */
514 static char buf[MAX_PRCHAR_LEN+1];
515
516 c &= 0377; /*{{type-issue}}*/
517 if ((c < 128 || !utf_mode) && !control_char(c))
518 SNPRINTF1(buf, sizeof(buf), "%c", (int) c);
519 else if (c == ESC)
520 strcpy(buf, "ESC");
521 #if IS_EBCDIC_HOST
522 else if (!binary_char(c) && c < 64)
523 SNPRINTF1(buf, sizeof(buf), "^%c",
524 /*
525 * This array roughly inverts CONTROL() #defined in less.h,
526 * and should be kept in sync with CONTROL() and IBM-1047.
527 */
528 "@ABC.I.?...KLMNO"
529 "PQRS.JH.XY.."
530 "\\]^_"
531 "......W[.....EFG"
532 "..V....D....TU.Z"[c]);
533 #else
534 else if (c < 128 && !control_char(c ^ 0100))
535 SNPRINTF1(buf, sizeof(buf), "^%c", (int) (c ^ 0100));
536 #endif
537 else
538 SNPRINTF1(buf, sizeof(buf), binfmt, c);
539 return (buf);
540 }
541
542 /*
543 * Return the printable form of a UTF-8 character.
544 */
prutfchar(LWCHAR ch)545 public constant char * prutfchar(LWCHAR ch)
546 {
547 static char buf[MAX_PRCHAR_LEN+1];
548
549 if (ch == ESC)
550 strcpy(buf, "ESC");
551 else if (ch < 128 && control_char(ch))
552 {
553 if (!control_char(ch ^ 0100))
554 SNPRINTF1(buf, sizeof(buf), "^%c", ((char) ch) ^ 0100);
555 else
556 SNPRINTF1(buf, sizeof(buf), binfmt, (char) ch);
557 } else if (is_ubin_char(ch))
558 {
559 SNPRINTF1(buf, sizeof(buf), utfbinfmt, ch);
560 } else
561 {
562 char *p = buf;
563 if (ch >= 0x80000000)
564 ch = 0xFFFD; /* REPLACEMENT CHARACTER */
565 put_wchar(&p, ch);
566 *p = '\0';
567 }
568 return (buf);
569 }
570
571 /*
572 * Get the length of a UTF-8 character in bytes.
573 */
utf_len(char ch)574 public int utf_len(char ch)
575 {
576 if ((ch & 0x80) == 0)
577 return 1;
578 if ((ch & 0xE0) == 0xC0)
579 return 2;
580 if ((ch & 0xF0) == 0xE0)
581 return 3;
582 if ((ch & 0xF8) == 0xF0)
583 return 4;
584 #if 0
585 if ((ch & 0xFC) == 0xF8)
586 return 5;
587 if ((ch & 0xFE) == 0xFC)
588 return 6;
589 #endif
590 /* Invalid UTF-8 encoding. */
591 return 1;
592 }
593
594 /*
595 * Does the parameter point to the lead byte of a well-formed UTF-8 character?
596 */
is_utf8_well_formed(constant char * ss,int slen)597 public lbool is_utf8_well_formed(constant char *ss, int slen)
598 {
599 int i;
600 int len;
601 unsigned char s0 = (unsigned char) ss[0];
602
603 if (IS_UTF8_INVALID(s0))
604 return (FALSE);
605
606 len = utf_len(ss[0]);
607 if (len > slen)
608 return (FALSE);
609 if (len == 1)
610 return (TRUE);
611 if (len == 2)
612 {
613 if (s0 < 0xC2)
614 return (FALSE);
615 } else
616 {
617 unsigned char mask = (unsigned char) (~((1 << (8-len)) - 1));
618 if (s0 == mask && (ss[1] & mask) == 0x80)
619 return (FALSE);
620 }
621
622 for (i = 1; i < len; i++)
623 if (!IS_UTF8_TRAIL(ss[i]))
624 return (FALSE);
625 return (TRUE);
626 }
627
628 /*
629 * Skip bytes until a UTF-8 lead byte (11xxxxxx) or ASCII byte (0xxxxxxx) is found.
630 */
utf_skip_to_lead(constant char ** pp,constant char * limit)631 public void utf_skip_to_lead(constant char **pp, constant char *limit)
632 {
633 do {
634 ++(*pp);
635 } while (*pp < limit && !IS_UTF8_LEAD((*pp)[0] & 0377) && !IS_ASCII_OCTET((*pp)[0]));
636 }
637
638
639 /*
640 * Get the value of a UTF-8 character.
641 */
get_wchar(constant char * sp)642 public LWCHAR get_wchar(constant char *sp)
643 {
644 constant unsigned char *p = (constant unsigned char *) sp;
645 switch (utf_len(sp[0]))
646 {
647 case 1:
648 default:
649 /* 0xxxxxxx */
650 return (LWCHAR)
651 (p[0] & 0xFF);
652 case 2:
653 /* 110xxxxx 10xxxxxx */
654 return (LWCHAR) (
655 ((p[0] & 0x1F) << 6) |
656 (p[1] & 0x3F));
657 case 3:
658 /* 1110xxxx 10xxxxxx 10xxxxxx */
659 return (LWCHAR) (
660 ((p[0] & 0x0F) << 12) |
661 ((p[1] & 0x3F) << 6) |
662 (p[2] & 0x3F));
663 case 4:
664 /* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
665 return (LWCHAR) (
666 ((p[0] & 0x07) << 18) |
667 ((p[1] & 0x3F) << 12) |
668 ((p[2] & 0x3F) << 6) |
669 (p[3] & 0x3F));
670 #if 0
671 case 5:
672 /* 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
673 return (LWCHAR) (
674 ((p[0] & 0x03) << 24) |
675 ((p[1] & 0x3F) << 18) |
676 ((p[2] & 0x3F) << 12) |
677 ((p[3] & 0x3F) << 6) |
678 (p[4] & 0x3F));
679 case 6:
680 /* 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
681 return (LWCHAR) (
682 ((p[0] & 0x01) << 30) |
683 ((p[1] & 0x3F) << 24) |
684 ((p[2] & 0x3F) << 18) |
685 ((p[3] & 0x3F) << 12) |
686 ((p[4] & 0x3F) << 6) |
687 (p[5] & 0x3F));
688 #endif
689 }
690 }
691
692 /*
693 * Store a character into a UTF-8 string.
694 */
put_wchar(mutable char ** pp,LWCHAR ch)695 public void put_wchar(mutable char **pp, LWCHAR ch)
696 {
697 if (!utf_mode || ch < 0x80)
698 {
699 /* 0xxxxxxx */
700 *(*pp)++ = (char) ch;
701 } else if (ch < 0x800)
702 {
703 /* 110xxxxx 10xxxxxx */
704 *(*pp)++ = (char) (0xC0 | ((ch >> 6) & 0x1F));
705 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
706 } else if (ch < 0x10000)
707 {
708 /* 1110xxxx 10xxxxxx 10xxxxxx */
709 *(*pp)++ = (char) (0xE0 | ((ch >> 12) & 0x0F));
710 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
711 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
712 } else if (ch < 0x200000)
713 {
714 /* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
715 *(*pp)++ = (char) (0xF0 | ((ch >> 18) & 0x07));
716 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
717 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
718 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
719 #if 0
720 } else if (ch < 0x4000000)
721 {
722 /* 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
723 *(*pp)++ = (char) (0xF0 | ((ch >> 24) & 0x03));
724 *(*pp)++ = (char) (0x80 | ((ch >> 18) & 0x3F));
725 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
726 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
727 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
728 } else
729 {
730 /* 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
731 *(*pp)++ = (char) (0xF0 | ((ch >> 30) & 0x01));
732 *(*pp)++ = (char) (0x80 | ((ch >> 24) & 0x3F));
733 *(*pp)++ = (char) (0x80 | ((ch >> 18) & 0x3F));
734 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
735 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
736 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
737 #endif
738 }
739 }
740
741 /*
742 * Step forward or backward one character in a string.
743 */
step_charc(constant char ** pp,signed int dir,constant char * limit)744 public LWCHAR step_charc(constant char **pp, signed int dir, constant char *limit)
745 {
746 LWCHAR ch;
747 int len;
748 constant char *p = *pp;
749
750 if (!utf_mode)
751 {
752 /* It's easy if chars are one byte. */
753 if (dir > 0)
754 ch = (LWCHAR) (unsigned char) ((p < limit) ? *p++ : 0);
755 else
756 ch = (LWCHAR) (unsigned char) ((p > limit) ? *--p : 0);
757 } else if (dir > 0)
758 {
759 len = utf_len(*p);
760 if (p + len > limit)
761 {
762 ch = 0;
763 p = (char *) limit;
764 } else
765 {
766 ch = get_wchar(p);
767 p += len;
768 }
769 } else
770 {
771 while (p > limit && IS_UTF8_TRAIL(p[-1]))
772 p--;
773 if (p > limit)
774 ch = get_wchar(--p);
775 else
776 ch = 0;
777 }
778 *pp = p;
779 return ch;
780 }
781
step_char(char ** pp,signed int dir,constant char * limit)782 public LWCHAR step_char(char **pp, signed int dir, constant char *limit)
783 {
784 constant char *p = (constant char *) *pp;
785 LWCHAR ch = step_charc(&p, dir, limit);
786 *pp = (char *) p;
787 return ch;
788 }
789
790 /*
791 * Unicode characters data
792 * Actual data is in the generated *.uni files.
793 */
794
795 #define DECLARE_RANGE_TABLE_START(name) \
796 static struct wchar_range name##_array[] = {
797 #define DECLARE_RANGE_TABLE_END(name) \
798 }; struct wchar_range_table name##_table = { name##_array, countof(name##_array) };
799
800 DECLARE_RANGE_TABLE_START(compose)
801 #include "compose.uni"
802 DECLARE_RANGE_TABLE_END(compose)
803
804 DECLARE_RANGE_TABLE_START(ubin)
805 #include "ubin.uni"
806 DECLARE_RANGE_TABLE_END(ubin)
807
808 DECLARE_RANGE_TABLE_START(wide)
809 #include "wide.uni"
810 DECLARE_RANGE_TABLE_END(wide)
811
812 DECLARE_RANGE_TABLE_START(fmt)
813 #include "fmt.uni"
814 DECLARE_RANGE_TABLE_END(fmt)
815
816 /* comb_table is special pairs, not ranges. */
817 static struct wchar_range comb_table[] = {
818 {0x0644,0x0622}, {0x0644,0x0623}, {0x0644,0x0625}, {0x0644,0x0627},
819 };
820
821
is_in_table(LWCHAR ch,struct wchar_range_table * table)822 static lbool is_in_table(LWCHAR ch, struct wchar_range_table *table)
823 {
824 unsigned int hi;
825 unsigned int lo;
826
827 /* Binary search in the table. */
828 if (table->table == NULL || table->count == 0 || ch < table->table[0].first)
829 return FALSE;
830 lo = 0;
831 hi = table->count - 1;
832 while (lo <= hi)
833 {
834 unsigned int mid = (lo + hi) / 2;
835 if (ch > table->table[mid].last)
836 lo = mid + 1;
837 else if (ch < table->table[mid].first)
838 hi = mid - 1;
839 else
840 return TRUE;
841 }
842 return FALSE;
843 }
844
845 /*
846 * Is a character a UTF-8 composing character?
847 * If a composing character follows any char, the two combine into one glyph.
848 */
is_composing_char(LWCHAR ch)849 public lbool is_composing_char(LWCHAR ch)
850 {
851 if (is_in_table(ch, &user_prt_table)) return FALSE;
852 return is_in_table(ch, &user_compose_table) ||
853 is_in_table(ch, &compose_table) ||
854 (bs_mode != BS_CONTROL && is_in_table(ch, &fmt_table));
855 }
856
857 /*
858 * Should this UTF-8 character be treated as binary?
859 */
is_ubin_char(LWCHAR ch)860 public lbool is_ubin_char(LWCHAR ch)
861 {
862 if (is_in_table(ch, &user_prt_table)) return FALSE;
863 return is_in_table(ch, &user_ubin_table) ||
864 is_in_table(ch, &ubin_table) ||
865 (bs_mode == BS_CONTROL && is_in_table(ch, &fmt_table));
866 }
867
868 /*
869 * Is this a double width UTF-8 character?
870 */
is_wide_char(LWCHAR ch)871 public lbool is_wide_char(LWCHAR ch)
872 {
873 return is_in_table(ch, &user_wide_table) ||
874 is_in_table(ch, &wide_table);
875 }
876
877 /*
878 * Is a character a UTF-8 combining character?
879 * A combining char acts like an ordinary char, but if it follows
880 * a specific char (not any char), the two combine into one glyph.
881 */
is_combining_char(LWCHAR ch1,LWCHAR ch2)882 public lbool is_combining_char(LWCHAR ch1, LWCHAR ch2)
883 {
884 /* The table is small; use linear search. */
885 int i;
886 for (i = 0; i < countof(comb_table); i++)
887 {
888 if (ch1 == comb_table[i].first &&
889 ch2 == comb_table[i].last)
890 return TRUE;
891 }
892 return FALSE;
893 }
894
895