xref: /titanic_50/usr/src/lib/iconv_modules/common/cnv_ibmutf8.c (revision 880d797826457b77414b37d531cc3e1aa166ecbe)
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 #define MSB     0x80    /* most significant bit */
35 #define MBYTE   0x8e    /* multi-byte (4 byte character) */
36 #define PMASK   0xa0    /* plane number mask */
37 #define ONEBYTE 0xff    /* right most byte */
38 
39 /* non-identified character */
40 #define UTF8_NON_ID_CHAR1 0xEF
41 #define UTF8_NON_ID_CHAR2 0xBF
42 #define UTF8_NON_ID_CHAR3 0xBD
43 
44 enum _USTATE    { C0, C1 };
45 
46 
47 
48 
49 /*
50  * Actual conversion; called from iconv()
51  * Input is UTF-8 data.
52  * first convert to UCS2
53  */
54 size_t
_icv_iconv(_icv_state * st,char ** inbuf,size_t * inbytesleft,char ** outbuf,size_t * outbytesleft)55 _icv_iconv(_icv_state *st, char **inbuf, size_t *inbytesleft,
56                         char **outbuf, size_t *outbytesleft)
57 {
58 /*
59  * Actual conversion; called from iconv()
60  */
61 
62         char            c1, c2;
63         int             n, unidx;
64         unsigned long   ibm_code;
65 
66 #ifdef DEBUG
67     fprintf(stderr, "==========     iconv(): IBM --> UTF8     ==========\n");
68 #endif
69 
70         if (st == NULL) {
71                 errno = EBADF;
72                 return ((size_t) -1);
73         }
74 
75         if (inbuf == NULL || *inbuf == NULL) { /* Reset request. */
76                 st->ustate = C0;
77                 st->_errno = 0;
78                 return ((size_t) 0);
79         }
80 
81         st->_errno = 0;         /* reset internal errno */
82         errno = 0;              /* reset external errno */
83 
84         /* a state machine for interpreting UTF8 code */
85         while (*inbytesleft > 0 && *outbytesleft > 0) {
86                 switch (st->ustate) {
87                 case C0 :
88 		        st->keepc[0] = (**inbuf);
89 			st->ustate = C1;
90 		        break;
91 
92                 case C1 :
93 		        st->keepc[1] = (**inbuf);
94                         n = ibm_to_utf8(st, *outbuf, *outbytesleft);
95                         if (n > 0) {
96                                 (*outbuf) += n;
97                                 (*outbytesleft) -= n;
98                         } else {
99                                 st->_errno = errno;
100                                 return((size_t)-1);
101                         }
102                         st->ustate = C0;
103                         st->_errno = 0;
104                         break;
105 
106                 default:                        /* should never come here */
107                         st->_errno = errno = EILSEQ;
108                         st->ustate = C0;        /* reset state */
109                         break;
110                 }
111 
112 
113                 (*inbuf)++;
114                 (*inbytesleft)--;
115 
116                 if (st->_errno) {
117 #ifdef DEBUG
118     fprintf(stderr, "!!!!!\tst->_errno = %d\tst->ustate = %d\n",
119                 st->_errno, st->ustate);
120 #endif
121                         break;
122                 }
123 
124                 if (errno)
125                         return((size_t)-1);
126         }
127 
128         if (*outbytesleft == 0) {
129                 errno = E2BIG;
130                 return((size_t)-1);
131         }
132         return (*inbytesleft);
133 }
134 
135 /*
136  * IBM code --> (Unicode)
137  * Unicode --> UTF8 (FSS-UTF)
138  *             (File System Safe Universal Character Set Transformation Format)
139  * Return: > 0 - converted with enough space in output buffer
140  *         = 0 - no space in outbuf
141  */
ibm_to_utf8(st,buf,buflen)142 int ibm_to_utf8(st, buf, buflen)
143 _icv_state *st;
144 char    *buf;
145 size_t  buflen;
146 {
147         unsigned long   ibm_val;       /* Big-5 value */
148         int             unidx;          /* Unicode index */
149         unsigned long   uni_val;        /* Unicode */
150 
151         ibm_val = ((st->keepc[0]&ONEBYTE) << 8) + (st->keepc[1]&ONEBYTE);
152 #ifdef DEBUG
153     fprintf(stderr, "%x\t", ibm_val);
154 #endif
155 
156 
157         unidx = bisearch(ibm_val, st, st->table_size);
158 
159         if (unidx >= 0)
160 	{
161             if ( st->left_to_right )
162                 uni_val = st->table[unidx].right_code;
163             else
164                 uni_val = st->table[unidx].left_code;
165         }
166 
167 #ifdef DEBUG
168     fprintf(stderr, "unidx = %d, unicode = %x\t", unidx, uni_val);
169 #endif
170 
171         if (unidx >= 0) {       /* do Unicode to UTF8 conversion */
172                 if (uni_val > 0x0000 && uni_val <= 0x07ff) {
173                         if (buflen < 2) {
174 #ifdef DEBUG
175     fprintf(stderr, "outbuf overflow in ibm_to_utf8()!!\n");
176 #endif
177                                 errno = E2BIG;
178                                 return(0);
179                         }
180                         *buf = (char)((uni_val >> 6) & 0x1f) | 0xc0;
181                         *(buf+1) = (char)(uni_val & 0x3f) | 0x80;
182 #ifdef DEBUG
183     fprintf(stderr, "%x %x\n", *buf&ONEBYTE, *(buf+1)&ONEBYTE);
184 #endif
185                         return(2);
186                 }
187                 if (uni_val > 0x0800 && uni_val <= 0xffff) {
188                         if (buflen < 3) {
189 #ifdef DEBUG
190     fprintf(stderr, "outbuf overflow in ibm_to_utf8()!!\n");
191 #endif
192                                 errno = E2BIG;
193                                 return(0);
194                         }
195                         *buf = (char)((uni_val >> 12) & 0xf) | 0xe0;
196                         *(buf+1) = (char)((uni_val >>6) & 0x3f) | 0x80;
197                         *(buf+2) = (char)(uni_val & 0x3f) | 0x80;
198 #ifdef DEBUG
199     fprintf(stderr, "%x %x %x\n", *buf&ONEBYTE, *(buf+1)&ONEBYTE, *(buf+2)&ONEBYTE);
200 #endif
201                         return(3);
202                 }
203         }
204 
205         /* can't find a match in IBM --> UTF8 table or illegal UTF8 code */
206         if (buflen < 3) {
207 #ifdef DEBUG
208     fprintf(stderr, "outbuf overflow in ibm_to_utf8()!!\n");
209 #endif
210                 errno = E2BIG;
211                 return(0);
212         }
213 
214         *buf     = (char)UTF8_NON_ID_CHAR1;
215         *(buf+1) = (char)UTF8_NON_ID_CHAR2;
216         *(buf+2) = (char)UTF8_NON_ID_CHAR3;
217 
218 #ifdef DEBUG
219     fprintf(stderr, "%c %c %c\n", *buf, *(buf+1), *(buf+2));
220 #endif
221         return(3);
222 }
223