mbrtoc8.c revision 1.1 1 /* $NetBSD: mbrtoc8.c,v 1.1 2024/08/15 21:19:45 riastradh 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 1.1 2024/08/15 21:19:45 riastradh Exp $");
76
77 #include <assert.h>
78 #include <errno.h>
79 #include <stdalign.h>
80 #include <stddef.h>
81 #include <uchar.h>
82
83 #include "mbrtoc32.h"
84
85 struct mbrtoc8state {
86 char8_t nleft;
87 char8_t buf[3];
88 mbstate_t mbs;
89 };
90 __CTASSERT(offsetof(struct mbrtoc8state, mbs) <= sizeof(mbstate_t));
91 __CTASSERT(sizeof(struct mbrtoc32state) <= sizeof(mbstate_t) -
92 offsetof(struct mbrtoc8state, mbs));
93 __CTASSERT(alignof(struct mbrtoc8state) <= alignof(mbstate_t));
94
95 size_t
96 mbrtoc8(char8_t *restrict pc8, const char *restrict s, size_t n,
97 mbstate_t *restrict ps)
98 {
99 static mbstate_t psbuf;
100 struct mbrtoc8state *S;
101 char32_t c32;
102 size_t len;
103
104 /*
105 * `If ps is a null pointer, each function uses its own
106 * internal mbstate_t object instead, which is initialized at
107 * program startup to the initial conversion state; the
108 * functions are not required to avoid data races with other
109 * calls to the same function in this case. The
110 * implementation behaves as if no library function calls
111 * these functions with a null pointer for ps.'
112 */
113 if (ps == NULL)
114 ps = &psbuf;
115
116 /*
117 * `If s is a null pointer, the mbrtoc8 function is equivalent
118 * to the call:
119 *
120 * mbrtoc8(NULL, "", 1, ps)
121 *
122 * In this case, the values of the parameters pc8 and n are
123 * ignored.'
124 */
125 if (s == NULL) {
126 pc8 = NULL;
127 s = "";
128 n = 1;
129 }
130
131 /*
132 * Get the private conversion state.
133 */
134 S = (struct mbrtoc8state *)ps;
135
136 /*
137 * If there are pending trailing bytes, yield them and return
138 * (size_t)-3 to indicate that no bytes of input were consumed.
139 */
140 if (S->nleft) {
141 if (pc8)
142 *pc8 = S->buf[sizeof(S->buf) - S->nleft];
143 S->buf[sizeof(S->buf) - S->nleft] = 0; /* paranoia */
144 S->nleft--;
145 return (size_t)-3;
146 }
147
148 /*
149 * Consume the next scalar value. If no full scalar value can
150 * be obtained, stop here.
151 */
152 len = mbrtoc32(&c32, s, n, &S->mbs);
153 switch (len) {
154 case 0: /* NUL */
155 if (pc8)
156 *pc8 = 0;
157 return 0;
158 case (size_t)-2: /* still incomplete after n bytes */
159 case (size_t)-1: /* error */
160 return len;
161 default: /* consumed len bytes of input */
162 break;
163 }
164
165 /*
166 * We consumed a scalar value from the input.
167 *
168 * Encode it as UTF-8, yield the leading byte, and buffer the
169 * trailing bytes to yield later.
170 *
171 * Table 3-6: UTF-8 Bit Distribution
172 * Table 3-7: Well-Formed UTF-8 Byte Sequences
173 */
174 switch (c32) {
175 case 0x00 ... 0x7f:
176 if (pc8)
177 *pc8 = c32;
178 _DIAGASSERT(S->nleft == 0);
179 break;
180 case 0x0080 ... 0x07ff:
181 if (pc8)
182 *pc8 = 0xc0 | __SHIFTOUT(c32, __BITS(10,6));
183 S->buf[2] = 0x80 | __SHIFTOUT(c32, __BITS(5,0));
184 S->nleft = 1;
185 break;
186 case 0x0800 ... 0xffff:
187 if (pc8)
188 *pc8 = 0xe0 | __SHIFTOUT(c32, __BITS(15,12));
189 S->buf[1] = 0x80 | __SHIFTOUT(c32, __BITS(11,6));
190 S->buf[2] = 0x80 | __SHIFTOUT(c32, __BITS(5,0));
191 S->nleft = 2;
192 break;
193 case 0x10000 ... 0x10ffff:
194 if (pc8)
195 *pc8 = 0xf0 | __SHIFTOUT(c32, __BITS(20,18));
196 S->buf[0] = 0x80 | __SHIFTOUT(c32, __BITS(17,12));
197 S->buf[1] = 0x80 | __SHIFTOUT(c32, __BITS(11,6));
198 S->buf[2] = 0x80 | __SHIFTOUT(c32, __BITS(5,0));
199 S->nleft = 3;
200 break;
201 default:
202 errno = EILSEQ;
203 return (size_t)-1;
204 }
205
206 /*
207 * Return the number of bytes consumed from the input.
208 */
209 return len;
210 }
211