1 /* Character handling in C locale.
2
3 These functions work like the corresponding functions in <ctype.h>,
4 except that they have the C (POSIX) locale hardwired, whereas the
5 <ctype.h> functions' behaviour depends on the current locale set via
6 setlocale.
7
8 Copyright (C) 2000-2003, 2006, 2008-2022 Free Software Foundation, Inc.
9
10 This file is free software: you can redistribute it and/or modify
11 it under the terms of the GNU Lesser General Public License as
12 published by the Free Software Foundation; either version 2.1 of the
13 License, or (at your option) any later version.
14
15 This file is distributed in the hope that it will be useful,
16 but WITHOUT ANY WARRANTY; without even the implied warranty of
17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 GNU Lesser General Public License for more details.
19
20 You should have received a copy of the GNU Lesser General Public License
21 along with this program. If not, see <https://www.gnu.org/licenses/>. */
22
23 #ifndef C_CTYPE_H
24 #define C_CTYPE_H
25
26 #ifndef _GL_INLINE_HEADER_BEGIN
27 #error "Please include config.h first."
28 #endif
29 _GL_INLINE_HEADER_BEGIN
30 #ifndef C_CTYPE_INLINE
31 # define C_CTYPE_INLINE _GL_INLINE
32 #endif
33
34 #ifdef __cplusplus
35 extern "C" {
36 #endif
37
38
39 /* The functions defined in this file assume the "C" locale and a character
40 set without diacritics (ASCII-US or EBCDIC-US or something like that).
41 Even if the "C" locale on a particular system is an extension of the ASCII
42 character set (like on BeOS, where it is UTF-8, or on AmigaOS, where it
43 is ISO-8859-1), the functions in this file recognize only the ASCII
44 characters. */
45
46
47 #if (' ' == 32) && ('!' == 33) && ('"' == 34) && ('#' == 35) \
48 && ('%' == 37) && ('&' == 38) && ('\'' == 39) && ('(' == 40) \
49 && (')' == 41) && ('*' == 42) && ('+' == 43) && (',' == 44) \
50 && ('-' == 45) && ('.' == 46) && ('/' == 47) && ('0' == 48) \
51 && ('1' == 49) && ('2' == 50) && ('3' == 51) && ('4' == 52) \
52 && ('5' == 53) && ('6' == 54) && ('7' == 55) && ('8' == 56) \
53 && ('9' == 57) && (':' == 58) && (';' == 59) && ('<' == 60) \
54 && ('=' == 61) && ('>' == 62) && ('?' == 63) && ('A' == 65) \
55 && ('B' == 66) && ('C' == 67) && ('D' == 68) && ('E' == 69) \
56 && ('F' == 70) && ('G' == 71) && ('H' == 72) && ('I' == 73) \
57 && ('J' == 74) && ('K' == 75) && ('L' == 76) && ('M' == 77) \
58 && ('N' == 78) && ('O' == 79) && ('P' == 80) && ('Q' == 81) \
59 && ('R' == 82) && ('S' == 83) && ('T' == 84) && ('U' == 85) \
60 && ('V' == 86) && ('W' == 87) && ('X' == 88) && ('Y' == 89) \
61 && ('Z' == 90) && ('[' == 91) && ('\\' == 92) && (']' == 93) \
62 && ('^' == 94) && ('_' == 95) && ('a' == 97) && ('b' == 98) \
63 && ('c' == 99) && ('d' == 100) && ('e' == 101) && ('f' == 102) \
64 && ('g' == 103) && ('h' == 104) && ('i' == 105) && ('j' == 106) \
65 && ('k' == 107) && ('l' == 108) && ('m' == 109) && ('n' == 110) \
66 && ('o' == 111) && ('p' == 112) && ('q' == 113) && ('r' == 114) \
67 && ('s' == 115) && ('t' == 116) && ('u' == 117) && ('v' == 118) \
68 && ('w' == 119) && ('x' == 120) && ('y' == 121) && ('z' == 122) \
69 && ('{' == 123) && ('|' == 124) && ('}' == 125) && ('~' == 126)
70 /* The character set is ASCII or one of its variants or extensions, not EBCDIC.
71 Testing the value of '\n' and '\r' is not relevant. */
72 # define C_CTYPE_ASCII 1
73 #elif ! (' ' == '\x40' && '0' == '\xf0' \
74 && 'A' == '\xc1' && 'J' == '\xd1' && 'S' == '\xe2' \
75 && 'a' == '\x81' && 'j' == '\x91' && 's' == '\xa2')
76 # error "Only ASCII and EBCDIC are supported"
77 #endif
78
79 #if 'A' < 0
80 # error "EBCDIC and char is signed -- not supported"
81 #endif
82
83 /* Cases for control characters. */
84
85 #define _C_CTYPE_CNTRL \
86 case '\a': case '\b': case '\f': case '\n': \
87 case '\r': case '\t': case '\v': \
88 _C_CTYPE_OTHER_CNTRL
89
90 /* ASCII control characters other than those with \-letter escapes. */
91
92 #if C_CTYPE_ASCII
93 # define _C_CTYPE_OTHER_CNTRL \
94 case '\x00': case '\x01': case '\x02': case '\x03': \
95 case '\x04': case '\x05': case '\x06': case '\x0e': \
96 case '\x0f': case '\x10': case '\x11': case '\x12': \
97 case '\x13': case '\x14': case '\x15': case '\x16': \
98 case '\x17': case '\x18': case '\x19': case '\x1a': \
99 case '\x1b': case '\x1c': case '\x1d': case '\x1e': \
100 case '\x1f': case '\x7f'
101 #else
102 /* Use EBCDIC code page 1047's assignments for ASCII control chars;
103 assume all EBCDIC code pages agree about these assignments. */
104 # define _C_CTYPE_OTHER_CNTRL \
105 case '\x00': case '\x01': case '\x02': case '\x03': \
106 case '\x07': case '\x0e': case '\x0f': case '\x10': \
107 case '\x11': case '\x12': case '\x13': case '\x18': \
108 case '\x19': case '\x1c': case '\x1d': case '\x1e': \
109 case '\x1f': case '\x26': case '\x27': case '\x2d': \
110 case '\x2e': case '\x32': case '\x37': case '\x3c': \
111 case '\x3d': case '\x3f'
112 #endif
113
114 /* Cases for lowercase hex letters, and lowercase letters, all offset by N. */
115
116 #define _C_CTYPE_LOWER_A_THRU_F_N(N) \
117 case 'a' + (N): case 'b' + (N): case 'c' + (N): case 'd' + (N): \
118 case 'e' + (N): case 'f' + (N)
119 #define _C_CTYPE_LOWER_N(N) \
120 _C_CTYPE_LOWER_A_THRU_F_N(N): \
121 case 'g' + (N): case 'h' + (N): case 'i' + (N): case 'j' + (N): \
122 case 'k' + (N): case 'l' + (N): case 'm' + (N): case 'n' + (N): \
123 case 'o' + (N): case 'p' + (N): case 'q' + (N): case 'r' + (N): \
124 case 's' + (N): case 't' + (N): case 'u' + (N): case 'v' + (N): \
125 case 'w' + (N): case 'x' + (N): case 'y' + (N): case 'z' + (N)
126
127 /* Cases for hex letters, digits, lower, punct, and upper. */
128
129 #define _C_CTYPE_A_THRU_F \
130 _C_CTYPE_LOWER_A_THRU_F_N (0): \
131 _C_CTYPE_LOWER_A_THRU_F_N ('A' - 'a')
132 #define _C_CTYPE_DIGIT \
133 case '0': case '1': case '2': case '3': \
134 case '4': case '5': case '6': case '7': \
135 case '8': case '9'
136 #define _C_CTYPE_LOWER _C_CTYPE_LOWER_N (0)
137 #define _C_CTYPE_PUNCT \
138 case '!': case '"': case '#': case '$': \
139 case '%': case '&': case '\'': case '(': \
140 case ')': case '*': case '+': case ',': \
141 case '-': case '.': case '/': case ':': \
142 case ';': case '<': case '=': case '>': \
143 case '?': case '@': case '[': case '\\': \
144 case ']': case '^': case '_': case '`': \
145 case '{': case '|': case '}': case '~'
146 #define _C_CTYPE_UPPER _C_CTYPE_LOWER_N ('A' - 'a')
147
148
149 /* Function definitions. */
150
151 /* Unlike the functions in <ctype.h>, which require an argument in the range
152 of the 'unsigned char' type, the functions here operate on values that are
153 in the 'unsigned char' range or in the 'char' range. In other words,
154 when you have a 'char' value, you need to cast it before using it as
155 argument to a <ctype.h> function:
156
157 const char *s = ...;
158 if (isalpha ((unsigned char) *s)) ...
159
160 but you don't need to cast it for the functions defined in this file:
161
162 const char *s = ...;
163 if (c_isalpha (*s)) ...
164 */
165
166 C_CTYPE_INLINE bool
167 c_isalnum (int c)
168 {
169 switch (c)
170 {
171 _C_CTYPE_DIGIT:
172 _C_CTYPE_LOWER:
173 _C_CTYPE_UPPER:
174 return true;
175 default:
176 return false;
177 }
178 }
179
180 C_CTYPE_INLINE bool
181 c_isalpha (int c)
182 {
183 switch (c)
184 {
185 _C_CTYPE_LOWER:
186 _C_CTYPE_UPPER:
187 return true;
188 default:
189 return false;
190 }
191 }
192
193 /* The function isascii is not locale dependent.
194 Its use in EBCDIC is questionable. */
195 C_CTYPE_INLINE bool
196 c_isascii (int c)
197 {
198 switch (c)
199 {
200 case ' ':
201 _C_CTYPE_CNTRL:
202 _C_CTYPE_DIGIT:
203 _C_CTYPE_LOWER:
204 _C_CTYPE_PUNCT:
205 _C_CTYPE_UPPER:
206 return true;
207 default:
208 return false;
209 }
210 }
211
212 C_CTYPE_INLINE bool
213 c_isblank (int c)
214 {
215 return c == ' ' || c == '\t';
216 }
217
218 C_CTYPE_INLINE bool
219 c_iscntrl (int c)
220 {
221 switch (c)
222 {
223 _C_CTYPE_CNTRL:
224 return true;
225 default:
226 return false;
227 }
228 }
229
230 C_CTYPE_INLINE bool
231 c_isdigit (int c)
232 {
233 switch (c)
234 {
235 _C_CTYPE_DIGIT:
236 return true;
237 default:
238 return false;
239 }
240 }
241
242 C_CTYPE_INLINE bool
243 c_isgraph (int c)
244 {
245 switch (c)
246 {
247 _C_CTYPE_DIGIT:
248 _C_CTYPE_LOWER:
249 _C_CTYPE_PUNCT:
250 _C_CTYPE_UPPER:
251 return true;
252 default:
253 return false;
254 }
255 }
256
257 C_CTYPE_INLINE bool
258 c_islower (int c)
259 {
260 switch (c)
261 {
262 _C_CTYPE_LOWER:
263 return true;
264 default:
265 return false;
266 }
267 }
268
269 C_CTYPE_INLINE bool
270 c_isprint (int c)
271 {
272 switch (c)
273 {
274 case ' ':
275 _C_CTYPE_DIGIT:
276 _C_CTYPE_LOWER:
277 _C_CTYPE_PUNCT:
278 _C_CTYPE_UPPER:
279 return true;
280 default:
281 return false;
282 }
283 }
284
285 C_CTYPE_INLINE bool
286 c_ispunct (int c)
287 {
288 switch (c)
289 {
290 _C_CTYPE_PUNCT:
291 return true;
292 default:
293 return false;
294 }
295 }
296
297 C_CTYPE_INLINE bool
298 c_isspace (int c)
299 {
300 switch (c)
301 {
302 case ' ': case '\t': case '\n': case '\v': case '\f': case '\r':
303 return true;
304 default:
305 return false;
306 }
307 }
308
309 C_CTYPE_INLINE bool
310 c_isupper (int c)
311 {
312 switch (c)
313 {
314 _C_CTYPE_UPPER:
315 return true;
316 default:
317 return false;
318 }
319 }
320
321 C_CTYPE_INLINE bool
322 c_isxdigit (int c)
323 {
324 switch (c)
325 {
326 _C_CTYPE_DIGIT:
327 _C_CTYPE_A_THRU_F:
328 return true;
329 default:
330 return false;
331 }
332 }
333
334 C_CTYPE_INLINE int
335 c_tolower (int c)
336 {
337 switch (c)
338 {
339 _C_CTYPE_UPPER:
340 return c - 'A' + 'a';
341 default:
342 return c;
343 }
344 }
345
346 C_CTYPE_INLINE int
347 c_toupper (int c)
348 {
349 switch (c)
350 {
351 _C_CTYPE_LOWER:
352 return c - 'a' + 'A';
353 default:
354 return c;
355 }
356 }
357
358 #ifdef __cplusplus
359 }
360 #endif
361
362 _GL_INLINE_HEADER_END
363
364 #endif /* C_CTYPE_H */