xref: /titanic_51/usr/src/lib/iconv_modules/common/ibm%utf8.c (revision 91e1e26ac6a73ce959289cf7d3d96c4baedbe0b8)
1 /*
2  * CDDL HEADER START
3  *
4  * The contents of this file are subject to the terms of the
5  * Common Development and Distribution License (the "License").
6  * You may not use this file except in compliance with the License.
7  *
8  * You can obtain a copy of the license at src/OPENSOLARIS.LICENSE
9  * or http://www.opensolaris.org/os/licensing.
10  * See the License for the specific language governing permissions
11  * and limitations under the License.
12  *
13  * When distributing Covered Code, include this CDDL HEADER in each
14  * file and include the License file at src/OPENSOLARIS.LICENSE.
15  * If applicable, add the following below this CDDL HEADER, with the
16  * fields enclosed by brackets "[]" replaced with your own identifying
17  * information: Portions Copyright [yyyy] [name of copyright owner]
18  *
19  * CDDL HEADER END
20  */
21 
22 /*
23  * Copyright (c) 1997, by Sun Microsystems, Inc.
24  * All rights reserved.
25  */
26 
27 #include <stdio.h>
28 #include <stdlib.h>
29 #include <errno.h>
30 #include <sys/types.h>
31 
32 #include "tab_lookup.h"   	/* table lookup data types */
33 
34 int bisearch(unsigned long val, _icv_state *st, int n);
35 
36 #define MSB     0x80    /* most significant bit */
37 #define MBYTE   0x8e    /* multi-byte (4 byte character) */
38 #define PMASK   0xa0    /* plane number mask */
39 #define ONEBYTE 0xff    /* right most byte */
40 
41 /* non-identified character */
42 #define UTF8_NON_ID_CHAR1 0xEF
43 #define UTF8_NON_ID_CHAR2 0xBF
44 #define UTF8_NON_ID_CHAR3 0xBD
45 
46 enum _USTATE    { C0, C1, C2 };
47 
48 
49 int ibm_to_utf8(_icv_state *st, char    *buf, size_t  buflen);
50 
51 
52 /*
53  * Actual conversion; called from iconv()
54  * Input is UTF-8 data.
55  * first convert to UCS2
56  */
57 size_t
58 _icv_iconv(_icv_state *st, char **inbuf, size_t *inbytesleft,
59                         char **outbuf, size_t *outbytesleft)
60 {
61 /*
62  * Actual conversion; called from iconv()
63  */
64 
65         int             n;
66 
67 #ifdef DEBUG
68     fprintf(stderr, "==========     iconv(): IBM --> UTF8     ==========\n");
69 #endif
70 
71         if (st == NULL) {
72                 errno = EBADF;
73                 return ((size_t) -1);
74         }
75 
76         if (inbuf == NULL || *inbuf == NULL) { /* Reset request. */
77                 st->ustate = C0;
78                 st->_errno = 0;
79 		st->shift = SHIFT_IN;
80                 return ((size_t) 0);
81         }
82 
83         st->_errno = 0;         /* reset internal errno */
84         errno = 0;              /* reset external errno */
85 
86         /* a state machine for interpreting UTF8 code */
87         while (*inbytesleft > 0 && *outbytesleft > 0) {
88 		switch (**inbuf) {
89 		case SHIFT_OUT :
90 			if (st->shift == SHIFT_IN) {
91 				st->shift = SHIFT_OUT;
92 				(*inbuf)++;
93 				(*inbytesleft)--;
94 				continue;
95 			}
96 			break;
97 		case SHIFT_IN :
98 			if (st->shift == SHIFT_OUT) {
99 				st->shift = SHIFT_IN;
100 				(*inbuf)++;
101 				(*inbytesleft)--;
102 				continue;
103 			}
104 			break;
105 		}
106 
107                 switch (st->ustate) {
108                 case C0 :
109 			/* the input is ascii, single byte, convert it */
110 			if (st->shift == SHIFT_IN) {
111 				st->keepc[0] = 0x0;
112 				st->keepc[1] = **inbuf;
113 				st->ustate = C2;
114 				continue;
115 			 }
116 
117 			/* two bytes character */
118 		        st->keepc[0] = (**inbuf);
119 			st->ustate = C1;
120 		        break;
121                 case C1 :
122 		        st->keepc[1] = (**inbuf);
123 			st->ustate = C2;
124 			continue;
125 		case C2 :
126                         n = ibm_to_utf8(st, *outbuf, *outbytesleft);
127                         if (n > 0) {
128                                 (*outbuf) += n;
129                                 (*outbytesleft) -= n;
130                         } else {
131                                 st->_errno = errno;
132                                 return((size_t)-1);
133                         }
134                         st->ustate = C0;
135                         st->_errno = 0;
136                         break;
137 
138                 default:                        /* should never come here */
139                         st->_errno = errno = EILSEQ;
140                         st->ustate = C0;        /* reset state */
141                         break;
142                 }
143 
144 
145                 (*inbuf)++;
146                 (*inbytesleft)--;
147 
148                 if (st->_errno) {
149 #ifdef DEBUG
150     fprintf(stderr, "!!!!!\tst->_errno = %d\tst->ustate = %d\n",
151                 st->_errno, st->ustate);
152 #endif
153                         break;
154                 }
155 
156                 if (errno)
157                         return((size_t)-1);
158         }
159 
160         if (*outbytesleft == 0) {
161                 errno = E2BIG;
162                 return((size_t)-1);
163         }
164         return (*inbytesleft);
165 }
166 
167 /*
168  * IBM code --> (Unicode)
169  * Unicode --> UTF8 (FSS-UTF)
170  *             (File System Safe Universal Character Set Transformation Format)
171  * Return: > 0 - converted with enough space in output buffer
172  *         = 0 - no space in outbuf
173  */
174 int ibm_to_utf8(st, buf, buflen)
175 _icv_state *st;
176 char    *buf;
177 size_t  buflen;
178 {
179         unsigned long   ibm_val;       /* Big-5 value */
180         int             unidx;          /* Unicode index */
181         unsigned long   uni_val;        /* Unicode */
182 
183         ibm_val = ((st->keepc[0]&ONEBYTE) << 8) + (st->keepc[1]&ONEBYTE);
184 #ifdef DEBUG
185     fprintf(stderr, "%x\t", ibm_val);
186 #endif
187 
188 
189         unidx = bisearch(ibm_val, st, st->table_size);
190 
191         if (unidx >= 0)
192 	{
193             if ( st->left_to_right )
194                 uni_val = st->table[unidx].right_code;
195             else
196                 uni_val = st->table[unidx].left_code;
197         }
198 
199 #ifdef DEBUG
200     fprintf(stderr, "unidx = %d, unicode = %x\t", unidx, uni_val);
201 #endif
202 
203         if (unidx >= 0) {       /* do Unicode to UTF8 conversion */
204 		if (uni_val <= 0x07f) {
205 			if (buflen < 1) {
206 				errno = E2BIG;
207 				return 0;
208 			}
209 			*buf = uni_val;
210 			return 1;
211 		}
212                 if (uni_val >= 0x0080 && uni_val <= 0x07ff) {
213                         if (buflen < 2) {
214 #ifdef DEBUG
215     fprintf(stderr, "outbuf overflow in ibm_to_utf8()!!\n");
216 #endif
217                                 errno = E2BIG;
218                                 return(0);
219                         }
220                         *buf = (char)((uni_val >> 6) & 0x1f) | 0xc0;
221                         *(buf+1) = (char)(uni_val & 0x3f) | 0x80;
222 #ifdef DEBUG
223     fprintf(stderr, "%x %x\n", *buf&ONEBYTE, *(buf+1)&ONEBYTE);
224 #endif
225                         return(2);
226                 }
227                 if (uni_val >= 0x0800 && uni_val <= 0xffff) {
228                         if (buflen < 3) {
229 #ifdef DEBUG
230     fprintf(stderr, "outbuf overflow in ibm_to_utf8()!!\n");
231 #endif
232                                 errno = E2BIG;
233                                 return(0);
234                         }
235                         *buf = (char)((uni_val >> 12) & 0xf) | 0xe0;
236                         *(buf+1) = (char)((uni_val >>6) & 0x3f) | 0x80;
237                         *(buf+2) = (char)(uni_val & 0x3f) | 0x80;
238 #ifdef DEBUG
239     fprintf(stderr, "%x %x %x\n", *buf&ONEBYTE, *(buf+1)&ONEBYTE, *(buf+2)&ONEBYTE);
240 #endif
241                         return(3);
242                 }
243         }
244 
245         /* can't find a match in IBM --> UTF8 table or illegal UTF8 code */
246         if (buflen < 3) {
247 #ifdef DEBUG
248     fprintf(stderr, "outbuf overflow in ibm_to_utf8()!!\n");
249 #endif
250                 errno = E2BIG;
251                 return(0);
252         }
253 
254         *buf     = (char)UTF8_NON_ID_CHAR1;
255         *(buf+1) = (char)UTF8_NON_ID_CHAR2;
256         *(buf+2) = (char)UTF8_NON_ID_CHAR3;
257 
258 #ifdef DEBUG
259     fprintf(stderr, "%c %c %c\n", *buf, *(buf+1), *(buf+2));
260 #endif
261         return(3);
262 }
263