citrus_ues.c revision 219019
1/* $FreeBSD: head/lib/libiconv_modules/UES/citrus_ues.c 219019 2011-02-25 00:04:39Z gabor $ */
2/* $NetBSD: citrus_ues.c,v 1.1 2006/11/13 15:16:31 tnozaki Exp $ */
3
4/*-
5 * Copyright (c)2006 Citrus Project,
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
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.
16 *
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
27 * SUCH DAMAGE.
28 */
29
30#include <sys/cdefs.h>
31
32#include <assert.h>
33#include <errno.h>
34#include <limits.h>
35#include <stdio.h>
36#include <stdint.h>
37#include <stdlib.h>
38#include <string.h>
39#include <wchar.h>
40
41#include "citrus_namespace.h"
42#include "citrus_types.h"
43#include "citrus_bcs.h"
44#include "citrus_module.h"
45#include "citrus_stdenc.h"
46#include "citrus_ues.h"
47
48typedef struct {
49	size_t	 mb_cur_max;
50	int	 mode;
51#define MODE_C99	1
52} _UESEncodingInfo;
53
54typedef struct {
55	int	 chlen;
56	char	 ch[12];
57} _UESState;
58
59#define _CEI_TO_EI(_cei_)               (&(_cei_)->ei)
60#define _CEI_TO_STATE(_cei_, _func_)    (_cei_)->states.s_##_func_
61
62#define _FUNCNAME(m)			_citrus_UES_##m
63#define _ENCODING_INFO			_UESEncodingInfo
64#define _ENCODING_STATE			_UESState
65#define _ENCODING_MB_CUR_MAX(_ei_)	(_ei_)->mb_cur_max
66#define _ENCODING_IS_STATE_DEPENDENT		0
67#define _STATE_NEEDS_EXPLICIT_INIT(_ps_)	0
68
69static __inline void
70/*ARGSUSED*/
71_citrus_UES_init_state(_UESEncodingInfo * __restrict ei __unused,
72    _UESState * __restrict psenc)
73{
74
75	psenc->chlen = 0;
76}
77
78static __inline void
79/*ARGSUSED*/
80_citrus_UES_pack_state(_UESEncodingInfo * __restrict ei __unused,
81    void *__restrict pspriv, const _UESState * __restrict psenc)
82{
83
84	memcpy(pspriv, (const void *)psenc, sizeof(*psenc));
85}
86
87static __inline void
88/*ARGSUSED*/
89_citrus_UES_unpack_state(_UESEncodingInfo * __restrict ei __unused,
90    _UESState * __restrict psenc, const void * __restrict pspriv)
91{
92
93	memcpy((void *)psenc, pspriv, sizeof(*psenc));
94}
95
96static __inline int
97to_int(int ch)
98{
99
100	if (ch >= '0' && ch <= '9')
101		return (ch - '0');
102	else if (ch >= 'A' && ch <= 'F')
103		return ((ch - 'A') + 10);
104	else if (ch >= 'a' && ch <= 'f')
105		return ((ch - 'a') + 10);
106	return (-1);
107}
108
109#define ESCAPE		'\\'
110#define UCS2_ESC	'u'
111#define UCS4_ESC	'U'
112
113#define UCS2_BIT	16
114#define UCS4_BIT	32
115#define BMP_MAX		UINT32_C(0xFFFF)
116#define UCS2_MAX	UINT32_C(0x10FFFF)
117#define UCS4_MAX	UINT32_C(0x7FFFFFFF)
118
119static const char *xdig = "0123456789abcdef";
120
121static __inline int
122to_str(char *s, wchar_t wc, int bit)
123{
124	char *p;
125
126	p = s;
127	*p++ = ESCAPE;
128	switch (bit) {
129	case UCS2_BIT:
130		*p++ = UCS2_ESC;
131		break;
132	case UCS4_BIT:
133		*p++ = UCS4_ESC;
134		break;
135	default:
136		abort();
137	}
138	do {
139		*p++ = xdig[(wc >> (bit -= 4)) & 0xF];
140	} while (bit > 0);
141	return (p - s);
142}
143
144static __inline bool
145is_hi_surrogate(wchar_t wc)
146{
147
148	return (wc >= 0xD800 && wc <= 0xDBFF);
149}
150
151static __inline bool
152is_lo_surrogate(wchar_t wc)
153{
154
155	return (wc >= 0xDC00 && wc <= 0xDFFF);
156}
157
158static __inline wchar_t
159surrogate_to_ucs(wchar_t hi, wchar_t lo)
160{
161
162	hi -= 0xD800;
163	lo -= 0xDC00;
164	return ((hi << 10 | lo) + 0x10000);
165}
166
167static __inline void
168ucs_to_surrogate(wchar_t wc, wchar_t * __restrict hi, wchar_t * __restrict lo)
169{
170
171	wc -= 0x10000;
172	*hi = (wc >> 10) + 0xD800;
173	*lo = (wc & 0x3FF) + 0xDC00;
174}
175
176static __inline bool
177is_basic(wchar_t wc)
178{
179
180	return ((uint32_t)wc <= 0x9F && wc != 0x24 && wc != 0x40 &&
181	    wc != 0x60);
182}
183
184static int
185_citrus_UES_mbrtowc_priv(_UESEncodingInfo * __restrict ei,
186    wchar_t * __restrict pwc, char ** __restrict s, size_t n,
187    _UESState * __restrict psenc, size_t * __restrict nresult)
188{
189	char *s0;
190	int ch, head, num, tail;
191	wchar_t hi, wc;
192
193	if (*s == NULL) {
194		_citrus_UES_init_state(ei, psenc);
195		*nresult = 0;
196		return (0);
197	}
198	s0 = *s;
199
200	hi = (wchar_t)0;
201	tail = 0;
202
203surrogate:
204	wc = (wchar_t)0;
205	head = tail;
206	if (psenc->chlen == head) {
207		if (n-- < 1)
208			goto restart;
209		psenc->ch[psenc->chlen++] = *s0++;
210	}
211	ch = (unsigned char)psenc->ch[head++];
212	if (ch == ESCAPE) {
213		if (psenc->chlen == head) {
214			if (n-- < 1)
215				goto restart;
216			psenc->ch[psenc->chlen++] = *s0++;
217		}
218		switch (psenc->ch[head]) {
219		case UCS2_ESC:
220			tail += 6;
221			break;
222		case UCS4_ESC:
223			if (ei->mode & MODE_C99) {
224				tail = 10;
225				break;
226			}
227		/*FALLTHROUGH*/
228		default:
229			tail = 0;
230		}
231		++head;
232	}
233	for (; head < tail; ++head) {
234		if (psenc->chlen == head) {
235			if (n-- < 1) {
236restart:
237				*s = s0;
238				*nresult = (size_t)-2;
239				return (0);
240			}
241			psenc->ch[psenc->chlen++] = *s0++;
242		}
243		num = to_int((int)(unsigned char)psenc->ch[head]);
244		if (num < 0) {
245			tail = 0;
246			break;
247		}
248		wc = (wc << 4) | num;
249	}
250	head = 0;
251	switch (tail) {
252	case 0:
253		break;
254	case 6:
255		if (hi != (wchar_t)0)
256			break;
257		if ((ei->mode & MODE_C99) == 0) {
258			if (is_hi_surrogate(wc) != 0) {
259				hi = wc;
260				goto surrogate;
261			}
262			if ((uint32_t)wc <= 0x7F /* XXX */ ||
263			    is_lo_surrogate(wc) != 0)
264				break;
265			goto done;
266		}
267	/*FALLTHROUGH*/
268	case 10:
269		if (is_basic(wc) == 0 && (uint32_t)wc <= UCS4_MAX &&
270		    is_hi_surrogate(wc) == 0 && is_lo_surrogate(wc) == 0)
271			goto done;
272		*nresult = (size_t)-1;
273		return (EILSEQ);
274	case 12:
275		if (is_lo_surrogate(wc) == 0)
276			break;
277		wc = surrogate_to_ucs(hi, wc);
278		goto done;
279	}
280	ch = (unsigned char)psenc->ch[0];
281	head = psenc->chlen;
282	if (--head > 0)
283		memmove(&psenc->ch[0], &psenc->ch[1], head);
284	wc = (wchar_t)ch;
285done:
286	psenc->chlen = head;
287	if (pwc != NULL)
288		*pwc = wc;
289	*nresult = (size_t)((wc == 0) ? 0 : (s0 - *s));
290	*s = s0;
291
292	return (0);
293}
294
295static int
296_citrus_UES_wcrtomb_priv(_UESEncodingInfo * __restrict ei,
297    char * __restrict s, size_t n, wchar_t wc,
298    _UESState * __restrict psenc, size_t * __restrict nresult)
299{
300	wchar_t hi, lo;
301
302	if (psenc->chlen != 0)
303		return (EINVAL);
304
305	if ((ei->mode & MODE_C99) ? is_basic(wc) : (uint32_t)wc <= 0x7F) {
306		if (n-- < 1)
307			goto e2big;
308		psenc->ch[psenc->chlen++] = (char)wc;
309	} else if ((uint32_t)wc <= BMP_MAX) {
310		if (n < 6)
311			goto e2big;
312		psenc->chlen = to_str(&psenc->ch[0], wc, UCS2_BIT);
313	} else if ((ei->mode & MODE_C99) == 0 && (uint32_t)wc <= UCS2_MAX) {
314		if (n < 12)
315			goto e2big;
316		ucs_to_surrogate(wc, &hi, &lo);
317		psenc->chlen += to_str(&psenc->ch[0], hi, UCS2_BIT);
318		psenc->chlen += to_str(&psenc->ch[6], lo, UCS2_BIT);
319	} else if ((ei->mode & MODE_C99) && (uint32_t)wc <= UCS4_MAX) {
320		if (n < 10)
321			goto e2big;
322		psenc->chlen = to_str(&psenc->ch[0], wc, UCS4_BIT);
323	} else {
324		*nresult = (size_t)-1;
325		return (EILSEQ);
326	}
327	memcpy(s, psenc->ch, psenc->chlen);
328	*nresult = psenc->chlen;
329	psenc->chlen = 0;
330
331	return (0);
332
333e2big:
334	*nresult = (size_t)-1;
335	return (E2BIG);
336}
337
338/*ARGSUSED*/
339static int
340_citrus_UES_stdenc_wctocs(_UESEncodingInfo * __restrict ei __unused,
341    _csid_t * __restrict csid, _index_t * __restrict idx, wchar_t wc)
342{
343
344	*csid = 0;
345	*idx = (_index_t)wc;
346
347	return (0);
348}
349
350static __inline int
351/*ARGSUSED*/
352_citrus_UES_stdenc_cstowc(_UESEncodingInfo * __restrict ei __unused,
353    wchar_t * __restrict wc, _csid_t csid, _index_t idx)
354{
355
356	if (csid != 0)
357		return (EILSEQ);
358	*wc = (wchar_t)idx;
359
360	return (0);
361}
362
363static __inline int
364/*ARGSUSED*/
365_citrus_UES_stdenc_get_state_desc_generic(_UESEncodingInfo * __restrict ei __unused,
366    _UESState * __restrict psenc, int * __restrict rstate)
367{
368
369	*rstate = (psenc->chlen == 0) ? _STDENC_SDGEN_INITIAL :
370	    _STDENC_SDGEN_INCOMPLETE_CHAR;
371	return (0);
372}
373
374static void
375/*ARGSUSED*/
376_citrus_UES_encoding_module_uninit(_UESEncodingInfo *ei __unused)
377{
378
379	/* ei seems to be unused */
380}
381
382static int
383/*ARGSUSED*/
384_citrus_UES_encoding_module_init(_UESEncodingInfo * __restrict ei,
385    const void * __restrict var, size_t lenvar)
386{
387	const char *p;
388
389	p = var;
390	memset((void *)ei, 0, sizeof(*ei));
391	while (lenvar > 0) {
392		switch (_bcs_toupper(*p)) {
393		case 'C':
394			MATCH(C99, ei->mode |= MODE_C99);
395			break;
396		}
397		++p;
398		--lenvar;
399	}
400	ei->mb_cur_max = (ei->mode & MODE_C99) ? 10 : 12;
401
402	return (0);
403}
404
405/* ----------------------------------------------------------------------
406 * public interface for stdenc
407 */
408
409_CITRUS_STDENC_DECLS(UES);
410_CITRUS_STDENC_DEF_OPS(UES);
411
412#include "citrus_stdenc_template.h"
413