mbrtoc8.c revision 1.3
1 /* $NetBSD: mbrtoc8.c,v 1.3 2024/08/17 20:08:13 christos Exp $ */ 2 3 /*- 4 * Copyright (c) 2024 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 17 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 18 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 19 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 20 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 21 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 22 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 23 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 24 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 25 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 26 * POSSIBILITY OF SUCH DAMAGE. 27 */ 28 29 /* 30 * mbrtoc16(&c16, s, n, ps) 31 * 32 * Decode a Unicode scalar value from up to n bytes out of the 33 * multibyte string s, using multibyte encoding state ps, and 34 * store the next code unit in the UTF-8 representation of that 35 * scalar value at c8. 36 * 37 * If the UTF-8 representation of that scalar value is multiple 38 * bytes long, mbrtoc8 will yield leading byte in one call that 39 * consumes input, and will yield the trailing bytes in subsequent 40 * calls without consuming any input and returning (size_t)-3 41 * instead. 42 * 43 * Return the number of bytes consumed on success, or: 44 * 45 * - 0 if the code unit is NUL, or 46 * - (size_t)-3 if a trailing byte was returned without consuming 47 * any additional input, or 48 * - (size_t)-2 if the input is incomplete, or 49 * - (size_t)-1 on error with errno set to EILSEQ. 50 * 51 * In the case of incomplete input, the decoding state so far 52 * after processing s[0], s[1], ..., s[n - 1] is saved in ps, so 53 * subsequent calls to mbrtoc8 will pick up n bytes later into 54 * the input stream. 55 * 56 * References: 57 * 58 * The Unicode Standard, Version 15.0 -- Core Specification, The 59 * Unicode Consortium, Sec. 3.8 `Surrogates', p. 119. 60 * https://www.unicode.org/versions/Unicode15.0.0/UnicodeStandard-15.0.pdf#page=144 61 * https://web.archive.org/web/20240718101254/https://www.unicode.org/versions/Unicode15.0.0/UnicodeStandard-15.0.pdf#page=144 62 * 63 * The Unicode Standard, Version 15.0 -- Core Specification, The 64 * Unicode Consortium, Sec. 3.9 `Unicode Encoding Forms': UTF-16, 65 * p. 124. 66 * https://www.unicode.org/versions/Unicode15.0.0/UnicodeStandard-15.0.pdf#page=150 67 * https://web.archive.org/web/20240718101254/https://www.unicode.org/versions/Unicode15.0.0/UnicodeStandard-15.0.pdf#page=150 68 * 69 * F. Yergeau, `UTF-8, a transformation format of ISO 10646', 70 * RFC 3629, Internet Engineering Task Force, November 2003. 71 * https://datatracker.ietf.org/doc/html/rfc3629 72 */ 73 74 #include <sys/cdefs.h> 75 __RCSID("$NetBSD: mbrtoc8.c,v 1.3 2024/08/17 20:08:13 christos Exp $"); 76 77 #include "namespace.h" 78 79 #include <assert.h> 80 #include <errno.h> 81 #include <stdalign.h> 82 #include <stddef.h> 83 #include <uchar.h> 84 85 #include "mbrtoc32.h" 86 87 struct mbrtoc8state { 88 char8_t nleft; 89 char8_t buf[3]; 90 mbstate_t mbs; 91 }; 92 __CTASSERT(offsetof(struct mbrtoc8state, mbs) <= sizeof(mbstate_t)); 93 __CTASSERT(sizeof(struct mbrtoc32state) <= sizeof(mbstate_t) - 94 offsetof(struct mbrtoc8state, mbs)); 95 __CTASSERT(alignof(struct mbrtoc8state) <= alignof(mbstate_t)); 96 97 size_t 98 mbrtoc8(char8_t *restrict pc8, const char *restrict s, size_t n, 99 mbstate_t *restrict ps) 100 { 101 static mbstate_t psbuf; 102 struct mbrtoc8state *S; 103 char32_t c32; 104 size_t len; 105 106 /* 107 * `If ps is a null pointer, each function uses its own 108 * internal mbstate_t object instead, which is initialized at 109 * program startup to the initial conversion state; the 110 * functions are not required to avoid data races with other 111 * calls to the same function in this case. The 112 * implementation behaves as if no library function calls 113 * these functions with a null pointer for ps.' 114 */ 115 if (ps == NULL) 116 ps = &psbuf; 117 118 /* 119 * `If s is a null pointer, the mbrtoc8 function is equivalent 120 * to the call: 121 * 122 * mbrtoc8(NULL, "", 1, ps) 123 * 124 * In this case, the values of the parameters pc8 and n are 125 * ignored.' 126 */ 127 if (s == NULL) { 128 pc8 = NULL; 129 s = ""; 130 n = 1; 131 } 132 133 /* 134 * Get the private conversion state. 135 */ 136 S = (struct mbrtoc8state *)(void *)ps; 137 138 /* 139 * If there are pending trailing bytes, yield them and return 140 * (size_t)-3 to indicate that no bytes of input were consumed. 141 */ 142 if (S->nleft) { 143 if (pc8) 144 *pc8 = S->buf[sizeof(S->buf) - S->nleft]; 145 S->buf[sizeof(S->buf) - S->nleft] = 0; /* paranoia */ 146 S->nleft--; 147 return (size_t)-3; 148 } 149 150 /* 151 * Consume the next scalar value. If no full scalar value can 152 * be obtained, stop here. 153 */ 154 len = mbrtoc32(&c32, s, n, &S->mbs); 155 switch (len) { 156 case 0: /* NUL */ 157 if (pc8) 158 *pc8 = 0; 159 return 0; 160 case (size_t)-2: /* still incomplete after n bytes */ 161 case (size_t)-1: /* error */ 162 return len; 163 default: /* consumed len bytes of input */ 164 break; 165 } 166 167 /* 168 * We consumed a scalar value from the input. 169 * 170 * Encode it as UTF-8, yield the leading byte, and buffer the 171 * trailing bytes to yield later. 172 * 173 * Table 3-6: UTF-8 Bit Distribution 174 * Table 3-7: Well-Formed UTF-8 Byte Sequences 175 */ 176 switch (c32) { 177 case 0x00 ... 0x7f: 178 if (pc8) 179 *pc8 = c32; 180 _DIAGASSERT(S->nleft == 0); 181 break; 182 case 0x0080 ... 0x07ff: 183 if (pc8) 184 *pc8 = (char8_t)( 185 0xc0 | __SHIFTOUT(c32, __BITS(10,6))); 186 S->buf[2] = (char8_t)( 187 0x80 | __SHIFTOUT(c32, __BITS(5,0))); 188 S->nleft = 1; 189 break; 190 case 0x0800 ... 0xffff: 191 if (pc8) 192 *pc8 = (char8_t)( 193 0xe0 | __SHIFTOUT(c32, __BITS(15,12))); 194 S->buf[1] = (char8_t)( 195 0x80 | __SHIFTOUT(c32, __BITS(11,6))); 196 S->buf[2] = (char8_t)( 197 0x80 | __SHIFTOUT(c32, __BITS(5,0))); 198 S->nleft = 2; 199 break; 200 case 0x10000 ... 0x10ffff: 201 if (pc8) 202 *pc8 = (char8_t)( 203 0xf0 | __SHIFTOUT(c32, __BITS(20,18))); 204 S->buf[0] = (char8_t)( 205 0x80 | __SHIFTOUT(c32, __BITS(17,12))); 206 S->buf[1] = (char8_t)( 207 0x80 | __SHIFTOUT(c32, __BITS(11,6))); 208 S->buf[2] = (char8_t)( 209 0x80 | __SHIFTOUT(c32, __BITS(5,0))); 210 S->nleft = 3; 211 break; 212 default: 213 errno = EILSEQ; 214 return (size_t)-1; 215 } 216 217 /* 218 * Return the number of bytes consumed from the input. 219 */ 220 return len; 221 } 222
Indexes created Fri Sep 26 20:09:58 GMT 2025