2 /* $NetBSD: citrus_utf7.c,v 1.5 2006/08/23 12:57:24 tnozaki Exp $ */
5 * Copyright (c)2004, 2005 Citrus Project,
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31 #include <sys/cdefs.h>
42 #include "citrus_namespace.h"
43 #include "citrus_types.h"
44 #include "citrus_module.h"
45 #include "citrus_stdenc.h"
46 #include "citrus_utf7.h"
48 /* ----------------------------------------------------------------------
49 * private stuffs used by templates
52 #define EI_MASK UINT16_C(0xff)
53 #define EI_DIRECT UINT16_C(0x100)
54 #define EI_OPTION UINT16_C(0x200)
55 #define EI_SPACE UINT16_C(0x400)
63 mode: 1, /* whether base64 mode */
64 bits: 4, /* need to hold 0 - 15 */
65 cache: 22, /* 22 = BASE64_BIT + UTF16_BIT */
66 surrogate: 1; /* whether surrogate pair or not */
68 char ch[4]; /* BASE64_IN, 3 * 6 = 18, most closed to UTF16_BIT */
71 #define _CEI_TO_EI(_cei_) (&(_cei_)->ei)
72 #define _CEI_TO_STATE(_cei_, _func_) (_cei_)->states.s_##_func_
74 #define _FUNCNAME(m) _citrus_UTF7_##m
75 #define _ENCODING_INFO _UTF7EncodingInfo
76 #define _ENCODING_STATE _UTF7State
77 #define _ENCODING_MB_CUR_MAX(_ei_) 4
78 #define _ENCODING_IS_STATE_DEPENDENT 1
79 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_) 0
83 _citrus_UTF7_init_state(_UTF7EncodingInfo * __restrict ei __unused,
84 _UTF7State * __restrict s)
87 memset((void *)s, 0, sizeof(*s));
92 _citrus_UTF7_pack_state(_UTF7EncodingInfo * __restrict ei __unused,
93 void *__restrict pspriv, const _UTF7State * __restrict s)
96 memcpy(pspriv, (const void *)s, sizeof(*s));
101 _citrus_UTF7_unpack_state(_UTF7EncodingInfo * __restrict ei __unused,
102 _UTF7State * __restrict s, const void * __restrict pspriv)
105 memcpy((void *)s, pspriv, sizeof(*s));
108 static const char base64[] =
109 "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
110 "abcdefghijklmnopqrstuvwxyz"
113 static const char direct[] =
114 "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
115 "abcdefghijklmnopqrstuvwxyz"
116 "0123456789(),-./:?";
118 static const char option[] = "!\"#$%&';<=>@[]^_`{|}";
119 static const char spaces[] = " \t\r\n";
124 #define BASE64_MAX 0x3f
125 #define UTF16_MAX UINT16_C(0xffff)
126 #define UTF32_MAX UINT32_C(0x10ffff)
128 #define BASE64_IN '+'
129 #define BASE64_OUT '-'
131 #define SHIFT7BIT(c) ((c) >> 7)
132 #define ISSPECIAL(c) ((c) == '\0' || (c) == BASE64_IN)
134 #define FINDLEN(ei, c) \
135 (SHIFT7BIT((c)) ? -1 : (((ei)->cell[(c)] & EI_MASK) - 1))
137 #define ISDIRECT(ei, c) (!SHIFT7BIT((c)) && (ISSPECIAL((c)) || \
138 ei->cell[(c)] & (EI_DIRECT | EI_OPTION | EI_SPACE)))
140 #define ISSAFE(ei, c) (!SHIFT7BIT((c)) && (ISSPECIAL((c)) || \
141 (c < 0x80 && ei->cell[(c)] & (EI_DIRECT | EI_SPACE))))
144 #define SRG_BASE UINT32_C(0x10000)
145 #define HISRG_MIN UINT16_C(0xd800)
146 #define HISRG_MAX UINT16_C(0xdbff)
147 #define LOSRG_MIN UINT16_C(0xdc00)
148 #define LOSRG_MAX UINT16_C(0xdfff)
151 _citrus_UTF7_mbtoutf16(_UTF7EncodingInfo * __restrict ei,
152 uint16_t * __restrict u16, char ** __restrict s, size_t n,
153 _UTF7State * __restrict psenc, size_t * __restrict nresult)
162 for (i = 0, done = 0; done == 0; i++) {
163 if (i == psenc->chlen) {
165 *nresult = (size_t)-2;
167 sv.chlen = psenc->chlen;
171 psenc->ch[psenc->chlen++] = *s0++;
173 if (SHIFT7BIT((int)psenc->ch[i]))
176 if (psenc->bits > 0 || psenc->cache > 0)
178 if (psenc->ch[i] == BASE64_IN)
181 if (!ISDIRECT(ei, (int)psenc->ch[i]))
183 *u16 = (uint16_t)psenc->ch[i];
188 if (psenc->ch[i] == BASE64_OUT && psenc->cache == 0) {
190 *u16 = (uint16_t)BASE64_IN;
194 len = FINDLEN(ei, (int)psenc->ch[i]);
196 if (psenc->bits >= BASE64_BIT)
199 psenc->bits = psenc->cache = 0;
200 if (psenc->ch[i] != BASE64_OUT) {
201 if (!ISDIRECT(ei, (int)psenc->ch[i]))
203 *u16 = (uint16_t)psenc->ch[i];
208 (psenc->cache << BASE64_BIT) | len;
209 switch (psenc->bits) {
210 case 0: case 2: case 4: case 6: case 8:
211 psenc->bits += BASE64_BIT;
213 case 10: case 12: case 14:
214 psenc->bits -= (UTF16_BIT - BASE64_BIT);
215 *u16 = (psenc->cache >> psenc->bits) &
226 if (psenc->chlen > i)
229 *nresult = (size_t)((*u16 == 0) ? 0 : s0 - *s);
235 *nresult = (size_t)-1;
240 _citrus_UTF7_mbrtowc_priv(_UTF7EncodingInfo * __restrict ei,
241 wchar_t * __restrict pwc, char ** __restrict s, size_t n,
242 _UTF7State * __restrict psenc, size_t * __restrict nresult)
251 _citrus_UTF7_init_state(ei, psenc);
252 *nresult = (size_t)_ENCODING_IS_STATE_DEPENDENT;
256 if (psenc->surrogate) {
257 hi = (psenc->cache >> 2) & UTF16_MAX;
258 if (hi < HISRG_MIN || hi > HISRG_MAX)
262 err = _citrus_UTF7_mbtoutf16(ei, &hi, &s0, n, psenc, &nr);
263 if (nr == (size_t)-1 || nr == (size_t)-2) {
271 if (hi < HISRG_MIN || hi > HISRG_MAX) {
275 psenc->surrogate = 1;
277 err = _citrus_UTF7_mbtoutf16(ei, &lo, &s0, n, psenc, &nr);
278 if (nr == (size_t)-1 || nr == (size_t)-2) {
286 u32 = (hi << 10 | lo) + SRG_BASE;
292 if (u32 == (uint32_t)0) {
293 *nresult = (size_t)0;
294 _citrus_UTF7_init_state(ei, psenc);
297 psenc->surrogate = 0;
303 _citrus_UTF7_utf16tomb(_UTF7EncodingInfo * __restrict ei,
304 char * __restrict s, size_t n __unused, uint16_t u16,
305 _UTF7State * __restrict psenc, size_t * __restrict nresult)
309 if (psenc->chlen != 0 || psenc->bits > BASE64_BIT)
312 if (ISSAFE(ei, u16)) {
314 if (psenc->bits > 0) {
315 bits = BASE64_BIT - psenc->bits;
316 i = (psenc->cache << bits) & BASE64_MAX;
317 psenc->ch[psenc->chlen++] = base64[i];
318 psenc->bits = psenc->cache = 0;
320 if (u16 == BASE64_OUT || FINDLEN(ei, u16) >= 0)
321 psenc->ch[psenc->chlen++] = BASE64_OUT;
324 if (psenc->bits != 0)
326 psenc->ch[psenc->chlen++] = (char)u16;
327 if (u16 == BASE64_IN)
328 psenc->ch[psenc->chlen++] = BASE64_OUT;
333 psenc->ch[psenc->chlen++] = BASE64_IN;
336 psenc->cache = (psenc->cache << UTF16_BIT) | u16;
337 bits = UTF16_BIT + psenc->bits;
338 psenc->bits = bits % BASE64_BIT;
339 while ((bits -= BASE64_BIT) >= 0) {
340 i = (psenc->cache >> bits) & BASE64_MAX;
341 psenc->ch[psenc->chlen++] = base64[i];
344 memcpy(s, psenc->ch, psenc->chlen);
345 *nresult = psenc->chlen;
352 _citrus_UTF7_wcrtomb_priv(_UTF7EncodingInfo * __restrict ei,
353 char * __restrict s, size_t n, wchar_t wchar,
354 _UTF7State * __restrict psenc, size_t * __restrict nresult)
361 u32 = (uint32_t)wchar;
362 if (u32 <= UTF16_MAX) {
363 u16[0] = (uint16_t)u32;
365 } else if (u32 <= UTF32_MAX) {
367 u16[0] = (u32 >> 10) + HISRG_MIN;
368 u16[1] = ((uint16_t)(u32 & UINT32_C(0x3ff))) + LOSRG_MIN;
371 *nresult = (size_t)-1;
375 for (i = 0; i < len; ++i) {
376 err = _citrus_UTF7_utf16tomb(ei, s, n, u16[i], psenc, &nr);
378 return (err); /* XXX: state has been modified */
390 _citrus_UTF7_put_state_reset(_UTF7EncodingInfo * __restrict ei __unused,
391 char * __restrict s, size_t n, _UTF7State * __restrict psenc,
392 size_t * __restrict nresult)
396 if (psenc->chlen != 0 || psenc->bits > BASE64_BIT || psenc->surrogate)
400 if (psenc->bits > 0) {
403 bits = BASE64_BIT - psenc->bits;
404 pos = (psenc->cache << bits) & BASE64_MAX;
405 psenc->ch[psenc->chlen++] = base64[pos];
406 psenc->ch[psenc->chlen++] = BASE64_OUT;
407 psenc->bits = psenc->cache = 0;
411 if (psenc->bits != 0)
416 *nresult = (size_t)psenc->chlen;
417 if (psenc->chlen > 0) {
418 memcpy(s, psenc->ch, psenc->chlen);
427 _citrus_UTF7_stdenc_wctocs(_UTF7EncodingInfo * __restrict ei __unused,
428 _csid_t * __restrict csid, _index_t * __restrict idx, wchar_t wc)
439 _citrus_UTF7_stdenc_cstowc(_UTF7EncodingInfo * __restrict ei __unused,
440 wchar_t * __restrict wc, _csid_t csid, _index_t idx)
452 _citrus_UTF7_stdenc_get_state_desc_generic(_UTF7EncodingInfo * __restrict ei __unused,
453 _UTF7State * __restrict psenc, int * __restrict rstate)
456 *rstate = (psenc->chlen == 0) ? _STDENC_SDGEN_INITIAL :
457 _STDENC_SDGEN_INCOMPLETE_CHAR;
463 _citrus_UTF7_encoding_module_uninit(_UTF7EncodingInfo *ei __unused)
466 /* ei seems to be unused */
471 _citrus_UTF7_encoding_module_init(_UTF7EncodingInfo * __restrict ei,
472 const void * __restrict var __unused, size_t lenvar __unused)
476 memset(ei, 0, sizeof(*ei));
478 #define FILL(str, flag) \
480 for (s = str; *s != '\0'; s++) \
481 ei->cell[*s & 0x7f] |= flag; \
482 } while (/*CONSTCOND*/0)
484 FILL(base64, (s - base64) + 1);
485 FILL(direct, EI_DIRECT);
486 FILL(option, EI_OPTION);
487 FILL(spaces, EI_SPACE);
492 /* ----------------------------------------------------------------------
493 * public interface for stdenc
496 _CITRUS_STDENC_DECLS(UTF7);
497 _CITRUS_STDENC_DEF_OPS(UTF7);
499 #include "citrus_stdenc_template.h"