xlat16_iconv.c revision 126949 
1/*-
2 * Copyright (c) 2003 Ryuichiro Imura
3 * All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 *    notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 *    notice, this list of conditions and the following disclaimer in the
12 *    documentation and/or other materials provided with the distribution.
13 *
14 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
15 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
16 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
17 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
18 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
19 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
20 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
21 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
22 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
23 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
24 * SUCH DAMAGE.
25 *
26 * $FreeBSD: head/lib/libkiconv/xlat16_iconv.c 126949 2004-03-14 05:14:00Z bde $
27 */
28
29/*
30 * kiconv(3) requires shared linked, and reduce module size
31 * when statically linked.
32 */
33
34#ifdef PIC
35
36#include <sys/types.h>
37#include <sys/iconv.h>
38#include <sys/sysctl.h>
39
40#include <ctype.h>
41#include <dlfcn.h>
42#include <err.h>
43#include <errno.h>
44#include <stdio.h>
45#include <stdlib.h>
46#include <string.h>
47
48#include "quirks.h"
49
50typedef void *iconv_t;
51
52struct xlat16_table {
53	uint32_t *	idx[0x200];
54	void *		data;
55	size_t		size;
56};
57
58static struct xlat16_table kiconv_xlat16_open(const char *, const char *, int);
59
60static int my_iconv_init(void);
61static iconv_t (*my_iconv_open)(const char *, const char *);
62static size_t (*my_iconv)(iconv_t, const char **, size_t *, char **, size_t *);
63static int (*my_iconv_close)(iconv_t);
64static size_t my_iconv_char(iconv_t, const u_char **, size_t *, u_char **, size_t *);
65
66int
67kiconv_add_xlat16_cspair(const char *tocode, const char *fromcode, int flag)
68{
69	int error;
70	size_t i, size, idxsize;
71	struct iconv_cspair_info *csi;
72	struct xlat16_table xt;
73	void *data, *p;
74
75	if (sysctlbyname("kern.iconv.cslist", NULL, &size, NULL, 0) == -1)
76		return (-1);
77	if (size > 0) {
78		csi = malloc(size);
79		if (csi == NULL)
80			return (-1);
81		if (sysctlbyname("kern.iconv.cslist", csi, &size, NULL, 0) == -1) {
82			free(csi);
83			return (-1);
84		}
85		for (i = 0; i < (size/sizeof(*csi)); i++, csi++){
86			if (strcmp(csi->cs_to, tocode) == 0 &&
87			    strcmp(csi->cs_from, fromcode) == 0)
88				return (0);
89		}
90	}
91
92	xt = kiconv_xlat16_open(tocode, fromcode, flag);
93	if (xt.size == 0)
94		return (-1);
95
96	idxsize = sizeof(xt.idx);
97
98	if ((idxsize + xt.size) > ICONV_CSMAXDATALEN) {
99		errno = E2BIG;
100		return (-1);
101	}
102
103	if ((data = malloc(idxsize + xt.size)) != NULL) {
104		p = data;
105		memcpy(p, xt.idx, idxsize);
106		p += idxsize;
107		memcpy(p, xt.data, xt.size);
108		error = kiconv_add_xlat16_table(tocode, fromcode, data,
109		    (int)(idxsize + xt.size));
110		return (error);
111	}
112
113	return (-1);
114}
115
116int
117kiconv_add_xlat16_cspairs(const char *foreigncode, const char *localcode)
118{
119	int error;
120
121	error = kiconv_add_xlat16_cspair(foreigncode, localcode,
122	    KICONV_FROM_LOWER | KICONV_FROM_UPPER);
123	if (error)
124		return (error);
125	error = kiconv_add_xlat16_cspair(localcode, foreigncode,
126	    KICONV_LOWER | KICONV_UPPER);
127	if (error)
128		return (error);
129
130	return (0);
131}
132
133static struct xlat16_table
134kiconv_xlat16_open(const char *tocode, const char *fromcode, int lcase)
135{
136	u_char src[3], dst[4], *srcp, *dstp, ud, ld;
137	int us, ls, ret;
138	uint16_t c;
139	uint32_t table[0x80];
140	size_t inbytesleft, outbytesleft, pre_q_size, post_q_size;
141	struct xlat16_table xt;
142	struct quirk_replace_list *pre_q_list, *post_q_list;
143	iconv_t cd;
144	void *p;
145
146	xt.data = NULL;
147	xt.size = 0;
148
149	src[2] = '\0';
150	dst[3] = '\0';
151
152	ret = my_iconv_init();
153	if (ret)
154		return (xt);
155
156	cd = my_iconv_open(search_quirk(tocode, fromcode, &pre_q_list, &pre_q_size),
157	    search_quirk(fromcode, tocode, &post_q_list, &post_q_size));
158	if (cd == (iconv_t) (-1))
159		return (xt);
160
161	if ((xt.data = malloc(0x200 * 0x80 * sizeof(uint32_t))) == NULL)
162		return (xt);
163
164	p = xt.data;
165
166	for (ls = 0 ; ls < 0x200 ; ls++) {
167		xt.idx[ls] = NULL;
168		for (us = 0 ; us < 0x80 ; us++) {
169			srcp = src;
170			dstp = dst;
171
172			inbytesleft = 2;
173			outbytesleft = 3;
174			bzero(dst, outbytesleft);
175
176			c = ((ls & 0x100 ? us | 0x80 : us) << 8) | (u_char)ls;
177			c = quirk_vendor2unix(c, pre_q_list, pre_q_size);
178			src[0] = (u_char)(c >> 8);
179			src[1] = (u_char)c;
180
181			ret = my_iconv_char(cd, (const u_char **)&srcp,
182			    &inbytesleft, &dstp, &outbytesleft);
183			if (ret == -1) {
184				table[us] = 0;
185				continue;
186			}
187
188			ud = (u_char)dst[0];
189			ld = (u_char)dst[1];
190
191			switch(outbytesleft) {
192			case 0:
193#ifdef XLAT16_ACCEPT_3BYTE_CHR
194				table[us] = (ud << 8) | ld;
195				table[us] |= (u_char)dst[2] << 16;
196				table[us] |= XLAT16_IS_3BYTE_CHR;
197#else
198				table[us] = 0;
199				continue;
200#endif
201				break;
202			case 1:
203				table[us] = quirk_unix2vendor((ud << 8) | ld,
204				    post_q_list, post_q_size);
205				if ((table[us] >> 8) == 0)
206					table[us] |= XLAT16_ACCEPT_NULL_OUT;
207				break;
208			case 2:
209				table[us] = ud;
210				if (lcase & KICONV_LOWER && ud != tolower(ud)) {
211					table[us] |= (u_char)tolower(ud) << 16;
212					table[us] |= XLAT16_HAS_LOWER_CASE;
213				}
214				if (lcase & KICONV_UPPER && ud != toupper(ud)) {
215					table[us] |= (u_char)toupper(ud) << 16;
216					table[us] |= XLAT16_HAS_UPPER_CASE;
217				}
218				break;
219			}
220
221			switch(inbytesleft) {
222			case 0:
223				if ((ls & 0xff) == 0)
224					table[us] |= XLAT16_ACCEPT_NULL_IN;
225				break;
226			case 1:
227				c = ls > 0xff ? us | 0x80 : us;
228				if (lcase & KICONV_FROM_LOWER && c != tolower(c)) {
229					table[us] |= (u_char)tolower(c) << 16;
230					table[us] |= XLAT16_HAS_FROM_LOWER_CASE;
231				}
232				if (lcase & KICONV_FROM_UPPER && c != toupper(c)) {
233					table[us] |= (u_char)toupper(c) << 16;
234					table[us] |= XLAT16_HAS_FROM_UPPER_CASE;
235				}
236				break;
237			}
238
239			if (table[us] == 0)
240				continue;
241
242			/*
243			 * store not NULL
244			 */
245			xt.idx[ls] = table;
246		}
247		if (xt.idx[ls]) {
248			memcpy(p, table, sizeof(table));
249			p += sizeof(table);
250		}
251	}
252	my_iconv_close(cd);
253
254	xt.size = p - xt.data;
255	xt.data = realloc(xt.data, xt.size);
256	return (xt);
257}
258
259static int
260my_iconv_init(void)
261{
262	void *iconv_lib;
263
264	iconv_lib = dlopen("libiconv.so", RTLD_LAZY | RTLD_GLOBAL);
265	if (iconv_lib == NULL) {
266		warn("Unable to load iconv library: %s\n", dlerror());
267		errno = ENOENT;
268		return (-1);
269	}
270	my_iconv_open = dlsym(iconv_lib, "iconv_open");
271	my_iconv = dlsym(iconv_lib, "iconv");
272	my_iconv_close = dlsym(iconv_lib, "iconv_close");
273
274	return (0);
275}
276
277static size_t
278my_iconv_char(iconv_t cd, const u_char **ibuf, size_t * ilen, u_char **obuf,
279	size_t * olen)
280{
281	const u_char *sp;
282	u_char *dp, ilocal[3], olocal[3];
283	u_char c1, c2;
284	int ret;
285	size_t ir, or;
286
287	sp = *ibuf;
288	dp = *obuf;
289	ir = *ilen;
290
291	bzero(*obuf, *olen);
292	ret = my_iconv(cd, (const char **)&sp, ilen, (char **)&dp, olen);
293	c1 = (*obuf)[0];
294	c2 = (*obuf)[1];
295
296	if (ret == -1) {
297		if (*ilen == ir - 1 && (*ibuf)[1] == '\0' && (c1 || c2))
298			return (0);
299		else
300			return (-1);
301	}
302
303	/*
304	 * We must judge if inbuf is a single byte char or double byte char.
305	 * Here, to judge, try first byte(*sp) conversion and compare.
306	 */
307	ir = 1;
308	or = 3;
309
310	bzero(olocal, or);
311	memcpy(ilocal, *ibuf, sizeof(ilocal));
312	sp = ilocal;
313	dp = olocal;
314
315	if ((my_iconv(cd,(const char **)&sp, &ir, (char **)&dp, &or)) != -1) {
316		if (olocal[0] != c1)
317			return (ret);
318
319		if (olocal[1] == c2 && (*ibuf)[1] == '\0') {
320			/*
321			 * inbuf is a single byte char
322			 */
323			*ilen = 1;
324			*olen = or;
325			return (ret);
326		}
327
328		switch(or) {
329		case 0:
330		case 1:
331			if (olocal[1] == c2) {
332				/*
333				 * inbuf is a single byte char,
334				 * so return false here.
335				 */
336				return (-1);
337			} else {
338				/*
339				 * inbuf is a double byte char
340				 */
341				return (ret);
342			}
343			break;
344		case 2:
345			/*
346			 * should compare second byte of inbuf
347			 */
348			break;
349		}
350	} else {
351		/*
352		 * inbuf clould not be splitted, so inbuf is
353		 * a double byte char.
354		 */
355		return (ret);
356	}
357
358	/*
359	 * try second byte(*(sp+1)) conversion, and compare
360	 */
361	ir = 1;
362	or = 3;
363
364	bzero(olocal, or);
365
366	sp = ilocal + 1;
367	dp = olocal;
368
369	if ((my_iconv(cd,(const char **)&sp, &ir, (char **)&dp, &or)) != -1) {
370		if (olocal[0] == c2)
371			/*
372			 * inbuf is a single byte char
373			 */
374			return (-1);
375	}
376
377	return (ret);
378}
379
380#else /* statically linked */
381
382#include <errno.h>
383
384int
385kiconv_add_xlat16_cspair(const char *tocode, const char *fromcode, int flag)
386{
387	errno = EINVAL;
388	return (-1);
389}
390
391int
392kiconv_add_xlat16_cspairs(const char *tocode, const char *fromcode)
393{
394	errno = EINVAL;
395	return (-1);
396}
397
398#endif /* PIC */
399