1 /* Multibyte character data type.
2 Copyright (C) 2001, 2005-2007, 2009-2023 Free Software Foundation, Inc.
3
4 This file is free software: you can redistribute it and/or modify
5 it under the terms of the GNU Lesser General Public License as
6 published by the Free Software Foundation; either version 2.1 of the
7 License, or (at your option) any later version.
8
9 This file is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU Lesser General Public License for more details.
13
14 You should have received a copy of the GNU Lesser General Public License
15 along with this program. If not, see <https://www.gnu.org/licenses/>. */
16
17 /* Written by Bruno Haible <bruno@clisp.org>. */
18
19 /* A multibyte character is a short subsequence of a char* string,
20 representing a single 32-bit wide character.
21
22 We use multibyte characters instead of 32-bit wide characters because
23 of the following goals:
24 1) correct multibyte handling, i.e. operate according to the LC_CTYPE
25 locale,
26 2) ease of maintenance, i.e. the maintainer needs not know all details
27 of the ISO C 99 standard,
28 3) don't fail grossly if the input is not in the encoding set by the
29 locale, because often different encodings are in use in the same
30 countries (ISO-8859-1/UTF-8, EUC-JP/Shift_JIS, ...),
31 4) fast in the case of ASCII characters.
32
33 Multibyte characters are only accessed through the mb* macros.
34
35 mb_ptr (mbc)
36 return a pointer to the beginning of the multibyte sequence.
37
38 mb_len (mbc)
39 returns the number of bytes occupied by the multibyte sequence.
40 Always > 0.
41
42 mb_iseq (mbc, sc)
43 returns true if mbc is the standard ASCII character sc.
44
45 mb_isnul (mbc)
46 returns true if mbc is the nul character.
47
48 mb_cmp (mbc1, mbc2)
49 returns a positive, zero, or negative value depending on whether mbc1
50 sorts after, same or before mbc2.
51
52 mb_casecmp (mbc1, mbc2)
53 returns a positive, zero, or negative value depending on whether mbc1
54 sorts after, same or before mbc2, modulo upper/lowercase conversion.
55
56 mb_equal (mbc1, mbc2)
57 returns true if mbc1 and mbc2 are equal.
58
59 mb_caseequal (mbc1, mbc2)
60 returns true if mbc1 and mbc2 are equal modulo upper/lowercase conversion.
61
62 mb_isalnum (mbc)
63 returns true if mbc is alphanumeric.
64
65 mb_isalpha (mbc)
66 returns true if mbc is alphabetic.
67
68 mb_isascii(mbc)
69 returns true if mbc is plain ASCII.
70
71 mb_isblank (mbc)
72 returns true if mbc is a blank.
73
74 mb_iscntrl (mbc)
75 returns true if mbc is a control character.
76
77 mb_isdigit (mbc)
78 returns true if mbc is a decimal digit.
79
80 mb_isgraph (mbc)
81 returns true if mbc is a graphic character.
82
83 mb_islower (mbc)
84 returns true if mbc is lowercase.
85
86 mb_isprint (mbc)
87 returns true if mbc is a printable character.
88
89 mb_ispunct (mbc)
90 returns true if mbc is a punctuation character.
91
92 mb_isspace (mbc)
93 returns true if mbc is a space character.
94
95 mb_isupper (mbc)
96 returns true if mbc is uppercase.
97
98 mb_isxdigit (mbc)
99 returns true if mbc is a hexadecimal digit.
100
101 mb_width (mbc)
102 returns the number of columns on the output device occupied by mbc.
103 Always >= 0.
104
105 mb_putc (mbc, stream)
106 outputs mbc on stream, a byte oriented FILE stream opened for output.
107
108 mb_setascii (&mbc, sc)
109 assigns the standard ASCII character sc to mbc.
110
111 mb_copy (&destmbc, &srcmbc)
112 copies srcmbc to destmbc.
113
114 Here are the function prototypes of the macros.
115
116 extern const char * mb_ptr (const mbchar_t mbc);
117 extern size_t mb_len (const mbchar_t mbc);
118 extern bool mb_iseq (const mbchar_t mbc, char sc);
119 extern bool mb_isnul (const mbchar_t mbc);
120 extern int mb_cmp (const mbchar_t mbc1, const mbchar_t mbc2);
121 extern int mb_casecmp (const mbchar_t mbc1, const mbchar_t mbc2);
122 extern bool mb_equal (const mbchar_t mbc1, const mbchar_t mbc2);
123 extern bool mb_caseequal (const mbchar_t mbc1, const mbchar_t mbc2);
124 extern bool mb_isalnum (const mbchar_t mbc);
125 extern bool mb_isalpha (const mbchar_t mbc);
126 extern bool mb_isascii (const mbchar_t mbc);
127 extern bool mb_isblank (const mbchar_t mbc);
128 extern bool mb_iscntrl (const mbchar_t mbc);
129 extern bool mb_isdigit (const mbchar_t mbc);
130 extern bool mb_isgraph (const mbchar_t mbc);
131 extern bool mb_islower (const mbchar_t mbc);
132 extern bool mb_isprint (const mbchar_t mbc);
133 extern bool mb_ispunct (const mbchar_t mbc);
134 extern bool mb_isspace (const mbchar_t mbc);
135 extern bool mb_isupper (const mbchar_t mbc);
136 extern bool mb_isxdigit (const mbchar_t mbc);
137 extern int mb_width (const mbchar_t mbc);
138 extern void mb_putc (const mbchar_t mbc, FILE *stream);
139 extern void mb_setascii (mbchar_t *new, char sc);
140 extern void mb_copy (mbchar_t *new, const mbchar_t *old);
141 */
142
143 #ifndef _MBCHAR_H
144 #define _MBCHAR_H 1
145
146 /* This file uses _GL_INLINE_HEADER_BEGIN, _GL_INLINE. */
147 #if !_GL_CONFIG_H_INCLUDED
148 #error "Please include config.h first."
149 #endif
150
151 #include <string.h>
152 #include <uchar.h>
153
154 _GL_INLINE_HEADER_BEGIN
155 #ifndef MBCHAR_INLINE
156 # define MBCHAR_INLINE _GL_INLINE
157 #endif
158
159 #define MBCHAR_BUF_SIZE 24
160
161 struct mbchar
162 {
163 const char *ptr; /* pointer to current character */
164 size_t bytes; /* number of bytes of current character, > 0 */
165 bool wc_valid; /* true if wc is a valid 32-bit wide character */
166 char32_t wc; /* if wc_valid: the current character */
167 char buf[MBCHAR_BUF_SIZE]; /* room for the bytes, used for file input only */
168 };
169
170 /* EOF (not a real character) is represented with bytes = 0 and
171 wc_valid = false. */
172
173 typedef struct mbchar mbchar_t;
174
175 /* Access the current character. */
176 #define mb_ptr(mbc) ((mbc).ptr)
177 #define mb_len(mbc) ((mbc).bytes)
178
179 /* Comparison of characters. */
180 #define mb_iseq(mbc, sc) ((mbc).wc_valid && (mbc).wc == (sc))
181 #define mb_isnul(mbc) ((mbc).wc_valid && (mbc).wc == 0)
182 #define mb_cmp(mbc1, mbc2) \
183 ((mbc1).wc_valid \
184 ? ((mbc2).wc_valid \
185 ? _GL_CMP ((mbc1).wc, (mbc2).wc) \
186 : -1) \
187 : ((mbc2).wc_valid \
188 ? 1 \
189 : (mbc1).bytes == (mbc2).bytes \
190 ? memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) \
191 : (mbc1).bytes < (mbc2).bytes \
192 ? (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) > 0 ? 1 : -1) \
193 : (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc2).bytes) >= 0 ? 1 : -1)))
194 #define mb_casecmp(mbc1, mbc2) \
195 ((mbc1).wc_valid \
196 ? ((mbc2).wc_valid \
197 ? _GL_CMP (c32tolower ((mbc1).wc), c32tolower ((mbc2).wc)) \
198 : -1) \
199 : ((mbc2).wc_valid \
200 ? 1 \
201 : (mbc1).bytes == (mbc2).bytes \
202 ? memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) \
203 : (mbc1).bytes < (mbc2).bytes \
204 ? (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) > 0 ? 1 : -1) \
205 : (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc2).bytes) >= 0 ? 1 : -1)))
206 #define mb_equal(mbc1, mbc2) \
207 ((mbc1).wc_valid && (mbc2).wc_valid \
208 ? (mbc1).wc == (mbc2).wc \
209 : (mbc1).bytes == (mbc2).bytes \
210 && memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) == 0)
211 #define mb_caseequal(mbc1, mbc2) \
212 ((mbc1).wc_valid && (mbc2).wc_valid \
213 ? c32tolower ((mbc1).wc) == c32tolower ((mbc2).wc) \
214 : (mbc1).bytes == (mbc2).bytes \
215 && memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) == 0)
216
217 /* <ctype.h>, <wctype.h> classification. */
218 #define mb_isascii(mbc) \
219 ((mbc).wc_valid && (mbc).wc >= 0 && (mbc).wc <= 127)
220 #define mb_isalnum(mbc) ((mbc).wc_valid && c32isalnum ((mbc).wc))
221 #define mb_isalpha(mbc) ((mbc).wc_valid && c32isalpha ((mbc).wc))
222 #define mb_isblank(mbc) ((mbc).wc_valid && c32isblank ((mbc).wc))
223 #define mb_iscntrl(mbc) ((mbc).wc_valid && c32iscntrl ((mbc).wc))
224 #define mb_isdigit(mbc) ((mbc).wc_valid && c32isdigit ((mbc).wc))
225 #define mb_isgraph(mbc) ((mbc).wc_valid && c32isgraph ((mbc).wc))
226 #define mb_islower(mbc) ((mbc).wc_valid && c32islower ((mbc).wc))
227 #define mb_isprint(mbc) ((mbc).wc_valid && c32isprint ((mbc).wc))
228 #define mb_ispunct(mbc) ((mbc).wc_valid && c32ispunct ((mbc).wc))
229 #define mb_isspace(mbc) ((mbc).wc_valid && c32isspace ((mbc).wc))
230 #define mb_isupper(mbc) ((mbc).wc_valid && c32isupper ((mbc).wc))
231 #define mb_isxdigit(mbc) ((mbc).wc_valid && c32isxdigit ((mbc).wc))
232
233 /* Extra <wchar.h> function. */
234
235 /* Unprintable characters appear as a small box of width 1. */
236 #define MB_UNPRINTABLE_WIDTH 1
237
238 MBCHAR_INLINE int
239 mb_width_aux (char32_t wc)
240 {
241 int w = c32width (wc);
242 /* For unprintable characters, arbitrarily return 0 for control characters
243 and MB_UNPRINTABLE_WIDTH otherwise. */
244 return (w >= 0 ? w : c32iscntrl (wc) ? 0 : MB_UNPRINTABLE_WIDTH);
245 }
246
247 #define mb_width(mbc) \
248 ((mbc).wc_valid ? mb_width_aux ((mbc).wc) : MB_UNPRINTABLE_WIDTH)
249
250 /* Output. */
251 #define mb_putc(mbc, stream) fwrite ((mbc).ptr, 1, (mbc).bytes, (stream))
252
253 /* Assignment. */
254 #define mb_setascii(mbc, sc) \
255 ((mbc)->ptr = (mbc)->buf, (mbc)->bytes = 1, (mbc)->wc_valid = 1, \
256 (mbc)->wc = (mbc)->buf[0] = (sc))
257
258 /* Copying a character. */
259 MBCHAR_INLINE void
260 mb_copy (mbchar_t *new_mbc, const mbchar_t *old_mbc)
261 {
262 if (old_mbc->ptr == &old_mbc->buf[0])
263 {
264 memcpy (&new_mbc->buf[0], &old_mbc->buf[0], old_mbc->bytes);
265 new_mbc->ptr = &new_mbc->buf[0];
266 }
267 else
268 new_mbc->ptr = old_mbc->ptr;
269 new_mbc->bytes = old_mbc->bytes;
270 if ((new_mbc->wc_valid = old_mbc->wc_valid))
271 new_mbc->wc = old_mbc->wc;
272 }
273
274
275 /* is_basic(c) tests whether the single-byte character c is in the
276 ISO C "basic character set".
277 This is a convenience function, and is in this file only to share code
278 between mbiter_multi.h and mbfile_multi.h. */
279 #if (' ' == 32) && ('!' == 33) && ('"' == 34) && ('#' == 35) \
280 && ('%' == 37) && ('&' == 38) && ('\'' == 39) && ('(' == 40) \
281 && (')' == 41) && ('*' == 42) && ('+' == 43) && (',' == 44) \
282 && ('-' == 45) && ('.' == 46) && ('/' == 47) && ('0' == 48) \
283 && ('1' == 49) && ('2' == 50) && ('3' == 51) && ('4' == 52) \
284 && ('5' == 53) && ('6' == 54) && ('7' == 55) && ('8' == 56) \
285 && ('9' == 57) && (':' == 58) && (';' == 59) && ('<' == 60) \
286 && ('=' == 61) && ('>' == 62) && ('?' == 63) && ('A' == 65) \
287 && ('B' == 66) && ('C' == 67) && ('D' == 68) && ('E' == 69) \
288 && ('F' == 70) && ('G' == 71) && ('H' == 72) && ('I' == 73) \
289 && ('J' == 74) && ('K' == 75) && ('L' == 76) && ('M' == 77) \
290 && ('N' == 78) && ('O' == 79) && ('P' == 80) && ('Q' == 81) \
291 && ('R' == 82) && ('S' == 83) && ('T' == 84) && ('U' == 85) \
292 && ('V' == 86) && ('W' == 87) && ('X' == 88) && ('Y' == 89) \
293 && ('Z' == 90) && ('[' == 91) && ('\\' == 92) && (']' == 93) \
294 && ('^' == 94) && ('_' == 95) && ('a' == 97) && ('b' == 98) \
295 && ('c' == 99) && ('d' == 100) && ('e' == 101) && ('f' == 102) \
296 && ('g' == 103) && ('h' == 104) && ('i' == 105) && ('j' == 106) \
297 && ('k' == 107) && ('l' == 108) && ('m' == 109) && ('n' == 110) \
298 && ('o' == 111) && ('p' == 112) && ('q' == 113) && ('r' == 114) \
299 && ('s' == 115) && ('t' == 116) && ('u' == 117) && ('v' == 118) \
300 && ('w' == 119) && ('x' == 120) && ('y' == 121) && ('z' == 122) \
301 && ('{' == 123) && ('|' == 124) && ('}' == 125) && ('~' == 126)
302 /* The character set is ISO-646, not EBCDIC. */
303 # define IS_BASIC_ASCII 1
304
305 extern const unsigned int is_basic_table[];
306
307 MBCHAR_INLINE bool
308 is_basic (char c)
309 {
310 return (is_basic_table [(unsigned char) c >> 5] >> ((unsigned char) c & 31))
311 & 1;
312 }
313
314 #else
315
316 MBCHAR_INLINE bool
317 is_basic (char c)
318 {
319 switch (c)
320 {
321 case '\t': case '\v': case '\f':
322 case ' ': case '!': case '"': case '#': case '%':
323 case '&': case '\'': case '(': case ')': case '*':
324 case '+': case ',': case '-': case '.': case '/':
325 case '0': case '1': case '2': case '3': case '4':
326 case '5': case '6': case '7': case '8': case '9':
327 case ':': case ';': case '<': case '=': case '>':
328 case '?':
329 case 'A': case 'B': case 'C': case 'D': case 'E':
330 case 'F': case 'G': case 'H': case 'I': case 'J':
331 case 'K': case 'L': case 'M': case 'N': case 'O':
332 case 'P': case 'Q': case 'R': case 'S': case 'T':
333 case 'U': case 'V': case 'W': case 'X': case 'Y':
334 case 'Z':
335 case '[': case '\\': case ']': case '^': case '_':
336 case 'a': case 'b': case 'c': case 'd': case 'e':
337 case 'f': case 'g': case 'h': case 'i': case 'j':
338 case 'k': case 'l': case 'm': case 'n': case 'o':
339 case 'p': case 'q': case 'r': case 's': case 't':
340 case 'u': case 'v': case 'w': case 'x': case 'y':
341 case 'z': case '{': case '|': case '}': case '~':
342 return 1;
343 default:
344 return 0;
345 }
346 }
347
348 #endif
349
350 _GL_INLINE_HEADER_END
351
352 #endif /* _MBCHAR_H */