2 * Copyright (C) 1984-2015 Mark Nudelman
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.
7 * For more information, see the README file.
12 * Functions to define the character set
13 * and do things specific to the character set.
17 #if MSDOS_COMPILER==WIN32C
28 public int utf_mode = 0;
31 * Predefined character sets,
32 * selected by the LESSCHARSET environment variable.
39 { "ascii", NULL, "8bcccbcc18b95.b" },
40 { "utf-8", &utf_mode, "8bcccbcc18b95.b126.bb" },
41 { "iso8859", NULL, "8bcccbcc18b95.33b." },
42 { "latin3", NULL, "8bcccbcc18b95.33b5.b8.b15.b4.b12.b18.b12.b." },
43 { "arabic", NULL, "8bcccbcc18b95.33b.3b.7b2.13b.3b.b26.5b19.b" },
44 { "greek", NULL, "8bcccbcc18b95.33b4.2b4.b3.b35.b44.b" },
45 { "greek2005", NULL, "8bcccbcc18b95.33b14.b35.b44.b" },
46 { "hebrew", NULL, "8bcccbcc18b95.33b.b29.32b28.2b2.b" },
47 { "koi8-r", NULL, "8bcccbcc18b95.b." },
48 { "KOI8-T", NULL, "8bcccbcc18b95.b8.b6.b8.b.b.5b7.3b4.b4.b3.b.b.3b." },
49 { "georgianps", NULL, "8bcccbcc18b95.3b11.4b12.2b." },
50 { "tcvn", NULL, "b..b...bcccbccbbb7.8b95.b48.5b." },
51 { "TIS-620", NULL, "8bcccbcc18b95.b.4b.11b7.8b." },
52 { "next", NULL, "8bcccbcc18b95.bb125.bb" },
53 { "dos", NULL, "8bcccbcc12bc5b95.b." },
54 { "windows-1251", NULL, "8bcccbcc12bc5b95.b24.b." },
55 { "windows-1252", NULL, "8bcccbcc12bc5b95.b.b11.b.2b12.b." },
56 { "windows-1255", NULL, "8bcccbcc12bc5b95.b.b8.b.5b9.b.4b." },
57 { "ebcdic", NULL, "5bc6bcc7bcc41b.9b7.9b5.b..8b6.10b6.b9.7b9.8b8.17b3.3b9.7b9.8b8.6b10.b.b.b." },
58 { "IBM-1047", NULL, "4cbcbc3b9cbccbccbb4c6bcc5b3cbbc4bc4bccbc191.b" },
63 * Support "locale charmap"/nl_langinfo(CODESET) values, as well as others.
70 { "ANSI_X3.4-1968", "ascii" },
71 { "US-ASCII", "ascii" },
72 { "latin1", "iso8859" },
73 { "ISO-8859-1", "iso8859" },
74 { "latin9", "iso8859" },
75 { "ISO-8859-15", "iso8859" },
76 { "latin2", "iso8859" },
77 { "ISO-8859-2", "iso8859" },
78 { "ISO-8859-3", "latin3" },
79 { "latin4", "iso8859" },
80 { "ISO-8859-4", "iso8859" },
81 { "cyrillic", "iso8859" },
82 { "ISO-8859-5", "iso8859" },
83 { "ISO-8859-6", "arabic" },
84 { "ISO-8859-7", "greek" },
85 { "IBM9005", "greek2005" },
86 { "ISO-8859-8", "hebrew" },
87 { "latin5", "iso8859" },
88 { "ISO-8859-9", "iso8859" },
89 { "latin6", "iso8859" },
90 { "ISO-8859-10", "iso8859" },
91 { "latin7", "iso8859" },
92 { "ISO-8859-13", "iso8859" },
93 { "latin8", "iso8859" },
94 { "ISO-8859-14", "iso8859" },
95 { "latin10", "iso8859" },
96 { "ISO-8859-16", "iso8859" },
98 { "EBCDIC-US", "ebcdic" },
99 { "IBM1047", "IBM-1047" },
100 { "KOI8-R", "koi8-r" },
101 { "KOI8-U", "koi8-r" },
102 { "GEORGIAN-PS", "georgianps" },
103 { "TCVN5712-1", "tcvn" },
104 { "NEXTSTEP", "next" },
105 { "windows", "windows-1252" }, /* backward compatibility */
106 { "CP1251", "windows-1251" },
107 { "CP1252", "windows-1252" },
108 { "CP1255", "windows-1255" },
112 #define IS_BINARY_CHAR 01
113 #define IS_CONTROL_CHAR 02
115 static char chardef[256];
116 static char *binfmt = NULL;
117 static char *utfbinfmt = NULL;
118 public int binattr = AT_STANDOUT;
122 * Define a charset, given a description string.
123 * The string consists of 256 letters,
124 * one for each character in the charset.
125 * If the string is shorter than 256 letters, missing letters
126 * are taken to be identical to the last one.
127 * A decimal number followed by a letter is taken to be a
128 * repetition of the letter.
130 * Each letter is one of:
133 * c control character
157 v = IS_BINARY_CHAR|IS_CONTROL_CHAR;
160 case '0': case '1': case '2': case '3': case '4':
161 case '5': case '6': case '7': case '8': case '9':
162 n = (10 * n) + (s[-1] - '0');
166 error("invalid chardef", NULL_PARG);
173 if (cp >= chardef + sizeof(chardef))
175 error("chardef longer than 256", NULL_PARG);
184 while (cp < chardef + sizeof(chardef))
189 * Define a charset, given a charset name.
190 * The valid charset names are listed in the "charsets" array.
193 icharset(name, no_error)
197 register struct charset *p;
198 register struct cs_alias *a;
200 if (name == NULL || *name == '\0')
203 /* First see if the name is an alias. */
204 for (a = cs_aliases; a->name != NULL; a++)
206 if (strcmp(name, a->name) == 0)
213 for (p = charsets; p->name != NULL; p++)
215 if (strcmp(name, p->name) == 0)
218 if (p->p_flag != NULL)
225 error("invalid charset name", NULL_PARG);
233 * Define a charset, given a locale name.
240 for (c = 0; c < (int) sizeof(chardef); c++)
245 chardef[c] = IS_CONTROL_CHAR;
247 chardef[c] = IS_BINARY_CHAR|IS_CONTROL_CHAR;
253 * Define the printing format for control (or binary utf) chars.
256 setbinfmt(s, fmtvarptr, default_fmt)
263 /* It would be too hard to account for width otherwise. */
267 if (*t < ' ' || *t > '~')
277 if (s == NULL || *s == '\0' ||
278 (*s == '*' && (s[1] == '\0' || s[2] == '\0' || strchr(s + 2, 'n'))) ||
279 (*s != '*' && strchr(s, 'n')))
283 * Select the attributes if it starts with "*".
290 case 'd': binattr = AT_BOLD; break;
291 case 'k': binattr = AT_BLINK; break;
292 case 's': binattr = AT_STANDOUT; break;
293 case 'u': binattr = AT_UNDERLINE; break;
294 default: binattr = AT_NORMAL; break;
310 * See if environment variable LESSCHARSET is defined.
312 s = lgetenv("LESSCHARSET");
317 * LESSCHARSET is not defined: try LESSCHARDEF.
319 s = lgetenv("LESSCHARDEF");
320 if (s != NULL && *s != '\0')
332 * Try using the codeset name as the charset name.
334 s = nl_langinfo(CODESET);
342 * Check whether LC_ALL, LC_CTYPE or LANG look like UTF-8 is used.
344 if ((s = lgetenv("LC_ALL")) != NULL ||
345 (s = lgetenv("LC_CTYPE")) != NULL ||
346 (s = lgetenv("LANG")) != NULL)
348 if ( strstr(s, "UTF-8") != NULL || strstr(s, "utf-8") != NULL
349 || strstr(s, "UTF8") != NULL || strstr(s, "utf8") != NULL)
350 if (icharset("utf-8", 1))
357 * Get character definitions from locale functions,
358 * rather than from predefined charset entry.
365 (void) icharset("dos", 1);
368 * Default to "latin1".
370 (void) icharset("latin1", 1);
377 * Initialize charset data structures.
385 setlocale(LC_ALL, "");
390 s = lgetenv("LESSBINFMT");
391 setbinfmt(s, &binfmt, "*s<%02X>");
393 s = lgetenv("LESSUTFBINFMT");
394 setbinfmt(s, &utfbinfmt, "<U+%04lX>");
398 * Is a given character a "binary" character?
405 return (is_ubin_char(c));
407 return (chardef[c] & IS_BINARY_CHAR);
411 * Is a given character a "control" character?
418 return (chardef[c] & IS_CONTROL_CHAR);
422 * Return the printable form of a character.
423 * For example, in the "ascii" charset '\3' is printed as "^C".
429 /* {{ This buffer can be overrun if LESSBINFMT is a long string. }} */
433 if ((c < 128 || !utf_mode) && !control_char(c))
434 SNPRINTF1(buf, sizeof(buf), "%c", (int) c);
438 else if (!binary_char(c) && c < 64)
439 SNPRINTF1(buf, sizeof(buf), "^%c",
441 * This array roughly inverts CONTROL() #defined in less.h,
442 * and should be kept in sync with CONTROL() and IBM-1047.
448 "..V....D....TU.Z"[c]);
450 else if (c < 128 && !control_char(c ^ 0100))
451 SNPRINTF1(buf, sizeof(buf), "^%c", (int) (c ^ 0100));
454 SNPRINTF1(buf, sizeof(buf), binfmt, c);
459 * Return the printable form of a UTF-8 character.
469 else if (ch < 128 && control_char(ch))
471 if (!control_char(ch ^ 0100))
472 SNPRINTF1(buf, sizeof(buf), "^%c", ((char) ch) ^ 0100);
474 SNPRINTF1(buf, sizeof(buf), binfmt, (char) ch);
475 } else if (is_ubin_char(ch))
477 SNPRINTF1(buf, sizeof(buf), utfbinfmt, ch);
481 if (ch >= 0x80000000)
482 ch = 0xFFFD; /* REPLACEMENT CHARACTER */
490 * Get the length of a UTF-8 character in bytes.
496 if ((ch & 0x80) == 0)
498 if ((ch & 0xE0) == 0xC0)
500 if ((ch & 0xF0) == 0xE0)
502 if ((ch & 0xF8) == 0xF0)
504 if ((ch & 0xFC) == 0xF8)
506 if ((ch & 0xFE) == 0xFC)
508 /* Invalid UTF-8 encoding. */
513 * Does the parameter point to the lead byte of a well-formed UTF-8 character?
516 is_utf8_well_formed(s, slen)
523 if (IS_UTF8_INVALID(s[0]))
526 len = utf_len((char) s[0]);
538 mask = (~((1 << (8-len)) - 1)) & 0xFF;
539 if (s[0] == mask && (s[1] & mask) == 0x80)
543 for (i = 1; i < len; i++)
544 if (!IS_UTF8_TRAIL(s[i]))
550 * Return number of invalid UTF-8 sequences found in a buffer.
553 utf_bin_count(data, len)
560 if (is_utf8_well_formed(data, len))
562 int clen = utf_len(*data);
567 /* Skip to next lead byte. */
572 } while (len > 0 && !IS_UTF8_LEAD(*data));
579 * Get the value of a UTF-8 character.
585 switch (utf_len(p[0]))
593 /* 110xxxxx 10xxxxxx */
595 ((p[0] & 0x1F) << 6) |
598 /* 1110xxxx 10xxxxxx 10xxxxxx */
600 ((p[0] & 0x0F) << 12) |
601 ((p[1] & 0x3F) << 6) |
604 /* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
606 ((p[0] & 0x07) << 18) |
607 ((p[1] & 0x3F) << 12) |
608 ((p[2] & 0x3F) << 6) |
611 /* 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
613 ((p[0] & 0x03) << 24) |
614 ((p[1] & 0x3F) << 18) |
615 ((p[2] & 0x3F) << 12) |
616 ((p[3] & 0x3F) << 6) |
619 /* 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
621 ((p[0] & 0x01) << 30) |
622 ((p[1] & 0x3F) << 24) |
623 ((p[2] & 0x3F) << 18) |
624 ((p[3] & 0x3F) << 12) |
625 ((p[4] & 0x3F) << 6) |
631 * Store a character into a UTF-8 string.
638 if (!utf_mode || ch < 0x80)
641 *(*pp)++ = (char) ch;
642 } else if (ch < 0x800)
644 /* 110xxxxx 10xxxxxx */
645 *(*pp)++ = (char) (0xC0 | ((ch >> 6) & 0x1F));
646 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
647 } else if (ch < 0x10000)
649 /* 1110xxxx 10xxxxxx 10xxxxxx */
650 *(*pp)++ = (char) (0xE0 | ((ch >> 12) & 0x0F));
651 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
652 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
653 } else if (ch < 0x200000)
655 /* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
656 *(*pp)++ = (char) (0xF0 | ((ch >> 18) & 0x07));
657 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
658 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
659 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
660 } else if (ch < 0x4000000)
662 /* 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
663 *(*pp)++ = (char) (0xF0 | ((ch >> 24) & 0x03));
664 *(*pp)++ = (char) (0x80 | ((ch >> 18) & 0x3F));
665 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
666 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
667 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
670 /* 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx */
671 *(*pp)++ = (char) (0xF0 | ((ch >> 30) & 0x01));
672 *(*pp)++ = (char) (0x80 | ((ch >> 24) & 0x3F));
673 *(*pp)++ = (char) (0x80 | ((ch >> 18) & 0x3F));
674 *(*pp)++ = (char) (0x80 | ((ch >> 12) & 0x3F));
675 *(*pp)++ = (char) (0x80 | ((ch >> 6) & 0x3F));
676 *(*pp)++ = (char) (0x80 | (ch & 0x3F));
681 * Step forward or backward one character in a string.
684 step_char(pp, dir, limit)
695 /* It's easy if chars are one byte. */
697 ch = (LWCHAR) ((p < limit) ? *p++ : 0);
699 ch = (LWCHAR) ((p > limit) ? *--p : 0);
714 while (p > limit && IS_UTF8_TRAIL(p[-1]))
726 * Unicode characters data
727 * Actual data is in the generated *.uni files.
730 #define DECLARE_RANGE_TABLE_START(name) \
731 static struct wchar_range name##_array[] = {
732 #define DECLARE_RANGE_TABLE_END(name) \
733 }; struct wchar_range_table name##_table = { name##_array, sizeof(name##_array)/sizeof(*name##_array) };
735 DECLARE_RANGE_TABLE_START(compose)
736 #include "compose.uni"
737 DECLARE_RANGE_TABLE_END(compose)
739 DECLARE_RANGE_TABLE_START(ubin)
741 DECLARE_RANGE_TABLE_END(ubin)
743 DECLARE_RANGE_TABLE_START(wide)
745 DECLARE_RANGE_TABLE_END(wide)
747 /* comb_table is special pairs, not ranges. */
748 static struct wchar_range comb_table[] = {
749 {0x0644,0x0622}, {0x0644,0x0623}, {0x0644,0x0625}, {0x0644,0x0627},
754 is_in_table(ch, table)
756 struct wchar_range_table *table;
761 /* Binary search in the table. */
762 if (ch < table->table[0].first)
765 hi = table->count - 1;
768 int mid = (lo + hi) / 2;
769 if (ch > table->table[mid].last)
771 else if (ch < table->table[mid].first)
780 * Is a character a UTF-8 composing character?
781 * If a composing character follows any char, the two combine into one glyph.
784 is_composing_char(ch)
787 return is_in_table(ch, &compose_table);
791 * Should this UTF-8 character be treated as binary?
797 return is_in_table(ch, &ubin_table);
801 * Is this a double width UTF-8 character?
807 return is_in_table(ch, &wide_table);
811 * Is a character a UTF-8 combining character?
812 * A combining char acts like an ordinary char, but if it follows
813 * a specific char (not any char), the two combine into one glyph.
816 is_combining_char(ch1, ch2)
820 /* The table is small; use linear search. */
822 for (i = 0; i < sizeof(comb_table)/sizeof(*comb_table); i++)
824 if (ch1 == comb_table[i].first &&
825 ch2 == comb_table[i].last)