1 /*
2 * Copyright (C) 1984-2025 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 xbuffer user_omit_array;
132 static struct wchar_range_table user_wide_table;
133 static struct wchar_range_table user_ubin_table;
134 static struct wchar_range_table user_compose_table;
135 static struct wchar_range_table user_prt_table;
136 static struct wchar_range_table user_omit_table;
137
138 /*
139 * Set a wchar_range_table to the table in an xbuffer.
140 */
wchar_range_table_set(struct wchar_range_table * tbl,struct xbuffer * arr)141 static void wchar_range_table_set(struct wchar_range_table *tbl, struct xbuffer *arr)
142 {
143 tbl->table = (struct wchar_range *) arr->data;
144 tbl->count = (unsigned int) (arr->end / sizeof(struct wchar_range));
145 }
146
147 /*
148 * Skip over a "U" or "U+" prefix before a hex codepoint.
149 */
skip_uprefix(constant char * s)150 static constant char * skip_uprefix(constant char *s)
151 {
152 if (*s == 'U' || *s == 'u')
153 if (*++s == '+') ++s;
154 return s;
155 }
156
157 /*
158 * Parse a dash-separated range of hex values.
159 */
wchar_range_get(constant char ** ss,struct wchar_range * range)160 static void wchar_range_get(constant char **ss, struct wchar_range *range)
161 {
162 constant char *s = skip_uprefix(*ss);
163 range->first = lstrtoulc(s, &s, 16);
164 if (s[0] == '-')
165 {
166 s = skip_uprefix(&s[1]);
167 range->last = lstrtoulc(s, &s, 16);
168 } else
169 {
170 range->last = range->first;
171 }
172 *ss = s;
173 }
174
175 /*
176 * Parse the LESSUTFCHARDEF variable.
177 */
ichardef_utf(constant char * s)178 static void ichardef_utf(constant char *s)
179 {
180 xbuf_init(&user_wide_array);
181 xbuf_init(&user_ubin_array);
182 xbuf_init(&user_compose_array);
183 xbuf_init(&user_prt_array);
184 xbuf_init(&user_omit_array);
185
186 if (s != NULL)
187 {
188 while (s[0] != '\0')
189 {
190 struct wchar_range range;
191 wchar_range_get(&s, &range);
192 if (range.last == 0)
193 {
194 error("invalid hex number(s) in LESSUTFCHARDEF", NULL_PARG);
195 quit(QUIT_ERROR);
196 }
197 if (*s++ != ':')
198 {
199 error("missing colon in LESSUTFCHARDEF", NULL_PARG);
200 quit(QUIT_ERROR);
201 }
202 switch (*s++)
203 {
204 case 'b':
205 xbuf_add_data(&user_ubin_array, &range, sizeof(range));
206 break;
207 case 'c':
208 xbuf_add_data(&user_compose_array, &range, sizeof(range));
209 break;
210 case 'd':
211 xbuf_add_data(&user_omit_array, &range, sizeof(range));
212 break;
213 case 'w':
214 xbuf_add_data(&user_wide_array, &range, sizeof(range));
215 xbuf_add_data(&user_prt_array, &range, sizeof(range));
216 break;
217 case 'p': case '.':
218 xbuf_add_data(&user_prt_array, &range, sizeof(range));
219 break;
220 case '\0':
221 s--;
222 break;
223 default:
224 /* Ignore unknown character attribute. */
225 break;
226 }
227 if (s[0] == ',') ++s;
228 }
229 }
230 wchar_range_table_set(&user_wide_table, &user_wide_array);
231 wchar_range_table_set(&user_ubin_table, &user_ubin_array);
232 wchar_range_table_set(&user_compose_table, &user_compose_array);
233 wchar_range_table_set(&user_prt_table, &user_prt_array);
234 wchar_range_table_set(&user_omit_table, &user_omit_array);
235 }
236
237 /*
238 * Define a charset, given a description string.
239 * The string consists of 256 letters,
240 * one for each character in the charset.
241 * If the string is shorter than 256 letters, missing letters
242 * are taken to be identical to the last one.
243 * A decimal number followed by a letter is taken to be a
244 * repetition of the letter.
245 *
246 * Each letter is one of:
247 * . normal character
248 * b binary character
249 * c control character
250 */
ichardef(constant char * s)251 static void ichardef(constant char *s)
252 {
253 char *cp;
254 int n;
255 char v;
256
257 n = 0;
258 v = 0;
259 cp = chardef;
260 while (*s != '\0')
261 {
262 switch (*s++)
263 {
264 case '.':
265 v = 0;
266 break;
267 case 'c':
268 v = IS_CONTROL_CHAR;
269 break;
270 case 'b':
271 v = IS_BINARY_CHAR|IS_CONTROL_CHAR;
272 break;
273
274 case '0': case '1': case '2': case '3': case '4':
275 case '5': case '6': case '7': case '8': case '9':
276 if (ckd_mul(&n, n, 10) || ckd_add(&n, n, s[-1] - '0'))
277 goto invalid_chardef;
278 continue;
279
280 default:
281 invalid_chardef:
282 error("invalid chardef", NULL_PARG);
283 quit(QUIT_ERROR);
284 /*NOTREACHED*/
285 }
286
287 do
288 {
289 if (cp >= chardef + sizeof(chardef))
290 {
291 error("chardef longer than 256", NULL_PARG);
292 quit(QUIT_ERROR);
293 /*NOTREACHED*/
294 }
295 *cp++ = v;
296 } while (--n > 0);
297 n = 0;
298 }
299
300 while (cp < chardef + sizeof(chardef))
301 *cp++ = v;
302 }
303
304 /*
305 * Define a charset, given a charset name.
306 * The valid charset names are listed in the "charsets" array.
307 */
icharset(constant char * name,int no_error)308 static int icharset(constant char *name, int no_error)
309 {
310 struct charset *p;
311 struct cs_alias *a;
312
313 if (name == NULL || *name == '\0')
314 return (0);
315
316 /* First see if the name is an alias. */
317 for (a = cs_aliases; a->name != NULL; a++)
318 {
319 if (strcmp(name, a->name) == 0)
320 {
321 name = a->oname;
322 break;
323 }
324 }
325
326 for (p = charsets; p->name != NULL; p++)
327 {
328 if (strcmp(name, p->name) == 0)
329 {
330 ichardef(p->desc);
331 if (p->p_flag != NULL)
332 {
333 #if MSDOS_COMPILER==WIN32C
334 *(p->p_flag) = 1 + (GetConsoleOutputCP() != CP_UTF8);
335 #else
336 *(p->p_flag) = 1;
337 #endif
338 }
339 return (1);
340 }
341 }
342
343 if (!no_error) {
344 error("invalid charset name", NULL_PARG);
345 quit(QUIT_ERROR);
346 }
347 return (0);
348 }
349
350 #if HAVE_LOCALE
351 /*
352 * Define a charset, given a locale name.
353 */
ilocale(void)354 static void ilocale(void)
355 {
356 int c;
357
358 for (c = 0; c < (int) sizeof(chardef); c++)
359 {
360 if (isprint(c))
361 chardef[c] = 0;
362 else if (iscntrl(c))
363 chardef[c] = IS_CONTROL_CHAR;
364 else
365 chardef[c] = IS_BINARY_CHAR|IS_CONTROL_CHAR;
366 }
367 }
368 #endif
369
370 /*
371 * Define the printing format for control (or binary utf) chars.
372 */
setfmt(constant char * s,constant char ** fmtvarptr,int * attrptr,constant char * default_fmt,lbool for_printf)373 public void setfmt(constant char *s, constant char **fmtvarptr, int *attrptr, constant char *default_fmt, lbool for_printf)
374 {
375 if (s == NULL || *s == '\0')
376 s = default_fmt;
377 else if (for_printf &&
378 ((*s == '*' && (s[1] == '\0' || s[2] == '\0' || strchr(s + 2, 'n'))) ||
379 (*s != '*' && strchr(s, 'n'))))
380 /* %n is evil */
381 s = default_fmt;
382
383 /*
384 * Select the attributes if it starts with "*".
385 */
386 if (*s == '*' && s[1] != '\0')
387 {
388 switch (s[1])
389 {
390 case 'd': *attrptr = AT_BOLD; break;
391 case 'k': *attrptr = AT_BLINK; break;
392 case 's': *attrptr = AT_STANDOUT; break;
393 case 'u': *attrptr = AT_UNDERLINE; break;
394 default: *attrptr = AT_NORMAL; break;
395 }
396 s += 2;
397 }
398 *fmtvarptr = s;
399 }
400
401 /*
402 *
403 */
set_charset(void)404 static void set_charset(void)
405 {
406 constant char *s;
407
408 ichardef_utf(lgetenv("LESSUTFCHARDEF"));
409
410 /*
411 * See if environment variable LESSCHARSET is defined.
412 */
413 s = lgetenv("LESSCHARSET");
414 if (icharset(s, 0))
415 return;
416
417 /*
418 * LESSCHARSET is not defined: try LESSCHARDEF.
419 */
420 s = lgetenv("LESSCHARDEF");
421 if (!isnullenv(s))
422 {
423 ichardef(s);
424 return;
425 }
426
427 #if HAVE_LOCALE
428 #ifdef CODESET
429 /*
430 * Try using the codeset name as the charset name.
431 */
432 s = nl_langinfo(CODESET);
433 if (icharset(s, 1))
434 return;
435 #endif
436 #endif
437
438 /*
439 * Check whether LC_ALL, LC_CTYPE or LANG look like UTF-8 is used.
440 */
441 if ((s = lgetenv("LC_ALL")) != NULL ||
442 (s = lgetenv("LC_CTYPE")) != NULL ||
443 (s = lgetenv("LANG")) != NULL)
444 {
445 if ( strstr(s, "UTF-8") != NULL || strstr(s, "utf-8") != NULL
446 || strstr(s, "UTF8") != NULL || strstr(s, "utf8") != NULL)
447 if (icharset("utf-8", 1))
448 return;
449 }
450
451 #if HAVE_LOCALE
452 /*
453 * Get character definitions from locale functions,
454 * rather than from predefined charset entry.
455 */
456 ilocale();
457 #else
458 #if MSDOS_COMPILER
459 #if MSDOS_COMPILER==WIN32C
460 (void) icharset("utf-8", 1);
461 #else
462 (void) icharset("dos", 1);
463 #endif
464 #else
465 (void) icharset("utf-8", 1);
466 #endif
467 #endif
468 }
469
470 /*
471 * Initialize charset data structures.
472 */
init_charset(void)473 public void init_charset(void)
474 {
475 constant char *s;
476
477 #if HAVE_LOCALE
478 setlocale(LC_ALL, "");
479 #endif
480
481 set_charset();
482
483 s = lgetenv("LESSBINFMT");
484 setfmt(s, &binfmt, &binattr, "*s<%02X>", TRUE);
485
486 s = lgetenv("LESSUTFBINFMT");
487 setfmt(s, &utfbinfmt, &binattr, "<U+%04lX>", TRUE);
488 }
489
490 /*
491 * Is a given character a "binary" character?
492 */
binary_char(LWCHAR c)493 public lbool binary_char(LWCHAR c)
494 {
495 if (utf_mode)
496 return (is_ubin_char(c));
497 if (c >= sizeof(chardef))
498 return TRUE;
499 return ((chardef[c] & IS_BINARY_CHAR) != 0);
500 }
501
502 /*
503 * Is a given character a "control" character?
504 */
control_char(LWCHAR c)505 public lbool control_char(LWCHAR c)
506 {
507 if (c >= sizeof(chardef))
508 return TRUE;
509 return (chardef[c] & IS_CONTROL_CHAR);
510 }
511
512 /*
513 * Return the printable form of a character.
514 * For example, in the "ascii" charset '\3' is printed as "^C".
515 */
prchar(LWCHAR c)516 public constant char * prchar(LWCHAR c)
517 {
518 /* {{ Fixed buffer size means LESSBINFMT etc can be truncated. }} */
519 static char buf[MAX_PRCHAR_LEN+1];
520
521 c &= 0377; /*{{type-issue}}*/
522 if ((c < 128 || !utf_mode) && !control_char(c))
523 SNPRINTF1(buf, sizeof(buf), "%c", (int) c);
524 else if (c == ESC)
525 strcpy(buf, "ESC");
526 #if IS_EBCDIC_HOST
527 else if (!binary_char(c) && c < 64)
528 SNPRINTF1(buf, sizeof(buf), "^%c",
529 /*
530 * This array roughly inverts CONTROL() #defined in less.h,
531 * and should be kept in sync with CONTROL() and IBM-1047.
532 */
533 "@ABC.I.?...KLMNO"
534 "PQRS.JH.XY.."
535 "\\]^_"
536 "......W[.....EFG"
537 "..V....D....TU.Z"[c]);
538 #else
539 else if (c < 128 && !control_char(c ^ 0100))
540 SNPRINTF1(buf, sizeof(buf), "^%c", (int) (c ^ 0100));
541 #endif
542 else
543 SNPRINTF1(buf, sizeof(buf), binfmt, c);
544 return (buf);
545 }
546
547 /*
548 * Return the printable form of a UTF-8 character.
549 */
prutfchar(LWCHAR ch)550 public constant char * prutfchar(LWCHAR ch)
551 {
552 static char buf[MAX_PRCHAR_LEN+1];
553
554 if (ch == ESC)
555 strcpy(buf, "ESC");
556 else if (ch < 128 && control_char(ch))
557 {
558 if (!control_char(ch ^ 0100))
559 SNPRINTF1(buf, sizeof(buf), "^%c", ((char) ch) ^ 0100);
560 else
561 SNPRINTF1(buf, sizeof(buf), binfmt, (char) ch);
562 } else if (is_ubin_char(ch) || is_omit_char(ch))
563 {
564 SNPRINTF1(buf, sizeof(buf), utfbinfmt, ch);
565 } else
566 {
567 char *p = buf;
568 if (ch >= 0x80000000)
569 ch = 0xFFFD; /* REPLACEMENT CHARACTER */
570 put_wchar(&p, ch);
571 *p = '\0';
572 }
573 return (buf);
574 }
575
576 /*
577 * Get the length of a UTF-8 character in bytes.
578 */
utf_len(char ch)579 public int utf_len(char ch)
580 {
581 if ((ch & 0x80) == 0)
582 return 1;
583 if ((ch & 0xE0) == 0xC0)
584 return 2;
585 if ((ch & 0xF0) == 0xE0)
586 return 3;
587 if ((ch & 0xF8) == 0xF0)
588 return 4;
589 #if 0
590 if ((ch & 0xFC) == 0xF8)
591 return 5;
592 if ((ch & 0xFE) == 0xFC)
593 return 6;
594 #endif
595 /* Invalid UTF-8 encoding. */
596 return 1;
597 }
598
599 /*
600 * Does the parameter point to the lead byte of a well-formed UTF-8 character?
601 */
is_utf8_well_formed(constant char * ss,int slen)602 public lbool is_utf8_well_formed(constant char *ss, int slen)
603 {
604 int i;
605 int len;
606 unsigned char s0 = (unsigned char) ss[0];
607
608 if (IS_UTF8_INVALID(s0))
609 return (FALSE);
610
611 len = utf_len(ss[0]);
612 if (len > slen)
613 return (FALSE);
614 if (len == 1)
615 return (TRUE);
616 if (len == 2)
617 {
618 if (s0 < 0xC2)
619 return (FALSE);
620 } else
621 {
622 unsigned char mask = (unsigned char) (~((1 << (8-len)) - 1));
623 if (s0 == mask && (ss[1] & mask) == 0x80)
624 return (FALSE);
625 }
626
627 for (i = 1; i < len; i++)
628 if (!IS_UTF8_TRAIL(ss[i]))
629 return (FALSE);
630 return (TRUE);
631 }
632
633 /*
634 * Skip bytes until a UTF-8 lead byte (11xxxxxx) or ASCII byte (0xxxxxxx) is found.
635 */
utf_skip_to_lead(constant char ** pp,constant char * limit)636 public void utf_skip_to_lead(constant char **pp, constant char *limit)
637 {
638 do {
639 ++(*pp);
640 } while (*pp < limit && !IS_UTF8_LEAD((*pp)[0] & 0377) && !IS_ASCII_OCTET((*pp)[0]));
641 }
642
643
644 /*
645 * Get the value of a UTF-8 character.
646 */
get_wchar(constant char * sp)647 public LWCHAR get_wchar(constant char *sp)
648 {
649 constant unsigned char *p = (constant unsigned char *) sp;
650 switch (utf_len(sp[0]))
651 {
652 case 1:
653 default:
654 /* 0xxxxxxx */
655 return (LWCHAR)
656 (p[0] & 0xFF);
657 case 2:
658 /* 110xxxxx 10xxxxxx */
659 return (LWCHAR) (
660 ((p[0] & 0x1F) << 6) |
661 (p[1] & 0x3F));
662 case 3:
663 /* 1110xxxx 10xxxxxx 10xxxxxx */
664 return (LWCHAR) (
665 ((p[0] & 0x0F) << 12) |
666 ((p[1] & 0x3F) << 6) |
667 (p[2] & 0x3F));
668 case 4:
669 /* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
670 return (LWCHAR) (
671 ((p[0] & 0x07) << 18) |
672 ((p[1] & 0x3F) << 12) |
673 ((p[2] & 0x3F) << 6) |
674 (p[3] & 0x3F));
675 #if 0
676 case 5:
677 /* 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
678 return (LWCHAR) (
679 ((p[0] & 0x03) << 24) |
680 ((p[1] & 0x3F) << 18) |
681 ((p[2] & 0x3F) << 12) |
682 ((p[3] & 0x3F) << 6) |
683 (p[4] & 0x3F));
684 case 6:
685 /* 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
686 return (LWCHAR) (
687 ((p[0] & 0x01) << 30) |
688 ((p[1] & 0x3F) << 24) |
689 ((p[2] & 0x3F) << 18) |
690 ((p[3] & 0x3F) << 12) |
691 ((p[4] & 0x3F) << 6) |
692 (p[5] & 0x3F));
693 #endif
694 }
695 }
696
697 /*
698 * Store a character into a UTF-8 string.
699 */
put_wchar(mutable char ** pp,LWCHAR ch)700 public void put_wchar(mutable char **pp, LWCHAR ch)
701 {
702 if (!utf_mode || ch < 0x80)
703 {
704 /* 0xxxxxxx */
705 *(*pp)++ = (char) ch;
706 } else if (ch < 0x800)
707 {
708 /* 110xxxxx 10xxxxxx */
709 *(*pp)++ = (char) (0xC0 | ((ch >> 6) & 0x1F));
710 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
711 } else if (ch < 0x10000)
712 {
713 /* 1110xxxx 10xxxxxx 10xxxxxx */
714 *(*pp)++ = (char) (0xE0 | ((ch >> 12) & 0x0F));
715 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
716 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
717 } else if (ch < 0x200000)
718 {
719 /* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
720 *(*pp)++ = (char) (0xF0 | ((ch >> 18) & 0x07));
721 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
722 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
723 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
724 #if 0
725 } else if (ch < 0x4000000)
726 {
727 /* 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
728 *(*pp)++ = (char) (0xF0 | ((ch >> 24) & 0x03));
729 *(*pp)++ = (char) (0x80 | ((ch >> 18) & 0x3F));
730 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
731 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
732 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
733 } else
734 {
735 /* 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
736 *(*pp)++ = (char) (0xF0 | ((ch >> 30) & 0x01));
737 *(*pp)++ = (char) (0x80 | ((ch >> 24) & 0x3F));
738 *(*pp)++ = (char) (0x80 | ((ch >> 18) & 0x3F));
739 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
740 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
741 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
742 #endif
743 }
744 }
745
746 /*
747 * Step forward or backward one character in a string.
748 */
step_charc(constant char ** pp,signed int dir,constant char * limit)749 public LWCHAR step_charc(constant char **pp, signed int dir, constant char *limit)
750 {
751 LWCHAR ch;
752 int len;
753 constant char *p = *pp;
754
755 if (!utf_mode)
756 {
757 /* It's easy if chars are one byte. */
758 if (dir > 0)
759 ch = (LWCHAR) (unsigned char) ((p < limit) ? *p++ : 0);
760 else
761 ch = (LWCHAR) (unsigned char) ((p > limit) ? *--p : 0);
762 } else if (dir > 0)
763 {
764 if (p >= limit)
765 ch = 0;
766 else
767 {
768 len = utf_len(*p);
769 if (p + len > limit || !is_utf8_well_formed(p, len))
770 {
771 ch = (LWCHAR) (unsigned char) *p++;
772 } else
773 {
774 ch = get_wchar(p);
775 p += len;
776 }
777 }
778 } else
779 {
780 while (p > limit && IS_UTF8_TRAIL(p[-1]))
781 p--;
782 if (p <= limit)
783 ch = 0;
784 else
785 {
786 len = utf_len(*--p);
787 if (p + len != *pp || !is_utf8_well_formed(p, len))
788 {
789 p = *pp - 1;
790 ch = (LWCHAR) (unsigned char) *p;
791 } else
792 {
793 ch = get_wchar(p);
794 }
795 }
796 }
797 *pp = p;
798 return ch;
799 }
800
step_char(char ** pp,signed int dir,constant char * limit)801 public LWCHAR step_char(char **pp, signed int dir, constant char *limit)
802 {
803 constant char *p = (constant char *) *pp;
804 LWCHAR ch = step_charc(&p, dir, limit);
805 *pp = (char *) p;
806 return ch;
807 }
808
809 /*
810 * Unicode characters data
811 * Actual data is in the generated *.uni files.
812 */
813
814 #define DECLARE_RANGE_TABLE_START(name) \
815 static struct wchar_range name##_array[] = {
816 #define DECLARE_RANGE_TABLE_END(name) \
817 }; struct wchar_range_table name##_table = { name##_array, countof(name##_array) };
818
819 DECLARE_RANGE_TABLE_START(compose)
820 #include "compose.uni"
821 DECLARE_RANGE_TABLE_END(compose)
822
823 DECLARE_RANGE_TABLE_START(ubin)
824 #include "ubin.uni"
825 DECLARE_RANGE_TABLE_END(ubin)
826
827 DECLARE_RANGE_TABLE_START(wide)
828 #include "wide.uni"
829 DECLARE_RANGE_TABLE_END(wide)
830
831 DECLARE_RANGE_TABLE_START(fmt)
832 #include "fmt.uni"
833 DECLARE_RANGE_TABLE_END(fmt)
834
835 DECLARE_RANGE_TABLE_START(omit)
836 #include "omit.uni"
837 DECLARE_RANGE_TABLE_END(omit)
838
839 /* comb_table is special pairs, not ranges. */
840 static struct wchar_range comb_table[] = {
841 {0x0644,0x0622}, {0x0644,0x0623}, {0x0644,0x0625}, {0x0644,0x0627},
842 };
843
844
is_in_table(LWCHAR ch,struct wchar_range_table * table)845 static lbool is_in_table(LWCHAR ch, struct wchar_range_table *table)
846 {
847 unsigned int hi;
848 unsigned int lo;
849
850 /* Binary search in the table. */
851 if (table->table == NULL || table->count == 0 || ch < table->table[0].first)
852 return FALSE;
853 lo = 0;
854 hi = table->count - 1;
855 while (lo <= hi)
856 {
857 unsigned int mid = (lo + hi) / 2;
858 if (ch > table->table[mid].last)
859 lo = mid + 1;
860 else if (ch < table->table[mid].first)
861 hi = mid - 1;
862 else
863 return TRUE;
864 }
865 return FALSE;
866 }
867
868 /*
869 * Is a character in none of a set of specified user tables?
870 */
not_user_defined(LWCHAR ch,struct wchar_range_table * tbl1,struct wchar_range_table * tbl2,struct wchar_range_table * tbl3)871 static lbool not_user_defined(LWCHAR ch, struct wchar_range_table *tbl1, struct wchar_range_table *tbl2, struct wchar_range_table *tbl3)
872 {
873 if (is_in_table(ch, tbl1)) return FALSE;
874 if (is_in_table(ch, tbl2)) return FALSE;
875 if (is_in_table(ch, tbl3)) return FALSE;
876 return TRUE;
877 }
878
879 /*
880 * Is a character a UTF-8 composing character?
881 * If a composing character follows any char, the two combine into one glyph.
882 */
is_composing_char(LWCHAR ch)883 public lbool is_composing_char(LWCHAR ch)
884 {
885 if (is_in_table(ch, &user_prt_table)) return FALSE;
886 return is_in_table(ch, &user_compose_table) ||
887 (is_in_table(ch, &compose_table) ||
888 (bs_mode != BS_CONTROL && is_in_table(ch, &fmt_table) &&
889 not_user_defined(ch, &user_prt_table, &user_ubin_table, &user_omit_table)));
890 }
891
892 /*
893 * Should this UTF-8 character be treated as binary?
894 */
is_ubin_char(LWCHAR ch)895 public lbool is_ubin_char(LWCHAR ch)
896 {
897 if (is_in_table(ch, &user_prt_table)) return FALSE;
898 return is_in_table(ch, &user_ubin_table) ||
899 (is_in_table(ch, &ubin_table) ||
900 (bs_mode == BS_CONTROL && is_in_table(ch, &fmt_table) &&
901 not_user_defined(ch, &user_prt_table, &user_compose_table, &user_omit_table)));
902 }
903
904 /*
905 * Is this a double width UTF-8 character?
906 */
is_wide_char(LWCHAR ch)907 public lbool is_wide_char(LWCHAR ch)
908 {
909 return is_in_table(ch, &user_wide_table) ||
910 (is_in_table(ch, &wide_table) &&
911 not_user_defined(ch, &user_compose_table, &user_ubin_table, &user_omit_table));
912 }
913
914 /*
915 * Is this an omittable character?
916 */
is_omit_char(LWCHAR ch)917 public lbool is_omit_char(LWCHAR ch)
918 {
919 return is_in_table(ch, &user_omit_table) ||
920 (is_in_table(ch, &omit_table) &&
921 not_user_defined(ch, &user_prt_table, &user_compose_table, &user_ubin_table));
922 }
923
924 /*
925 * Is a character a UTF-8 combining character?
926 * A combining char acts like an ordinary char, but if it follows
927 * a specific char (not any char), the two combine into one glyph.
928 */
is_combining_char(LWCHAR ch1,LWCHAR ch2)929 public lbool is_combining_char(LWCHAR ch1, LWCHAR ch2)
930 {
931 /* The table is small; use linear search. */
932 int i;
933 for (i = 0; i < countof(comb_table); i++)
934 {
935 if (ch1 == comb_table[i].first &&
936 ch2 == comb_table[i].last)
937 return TRUE;
938 }
939 return FALSE;
940 }
941