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 * Copyright 2004 Sun Microsystems, Inc. All rights reserved. 23 * Use is subject to license terms. 24 * 25 * This particular file is to cover conversions from UCS-2, UCS-2BE, UCS-2LE, 26 * UCS-4, UCS-4BE, UCS-4LE, UTF-16, UTF-16BE, UTF-16LE, UTF-32, UTF-32BE, 27 * and UTF-32LE to various single byte codesets. 28 */ 29 30 31 #include <stdlib.h> 32 #include <errno.h> 33 #include <sys/types.h> 34 #include <sys/isa_defs.h> 35 #include "ucs_to_sb.h" 36 37 38 void * 39 _icv_open() 40 { 41 ucs_state_t *cd = (ucs_state_t *)calloc(1, sizeof(ucs_state_t)); 42 43 if (cd == (ucs_state_t *)NULL) { 44 errno = ENOMEM; 45 return((void *)-1); 46 } 47 48 #if defined(UTF_16BE) || defined(UCS_2BE) || defined(UCS_4BE) || \ 49 defined(UTF_32BE) 50 cd->little_endian = false; 51 cd->bom_written = true; 52 #elif defined(UTF_16LE) || defined(UCS_2LE) || defined(UCS_4LE) || \ 53 defined(UTF_32LE) 54 cd->little_endian = true; 55 cd->bom_written = true; 56 #elif defined(_LITTLE_ENDIAN) 57 cd->little_endian = true; 58 #endif 59 60 return((void *)cd); 61 } 62 63 64 void 65 _icv_close(ucs_state_t *cd) 66 { 67 if (! cd) 68 errno = EBADF; 69 else 70 free((void *)cd); 71 } 72 73 74 size_t 75 _icv_iconv(ucs_state_t *cd, char **inbuf, size_t *inbufleft, char **outbuf, 76 size_t *outbufleft) 77 { 78 size_t ret_val = 0; 79 unsigned char *ib; 80 unsigned char *ob; 81 unsigned char *ibtail; 82 unsigned char *obtail; 83 uint_t u4; 84 uint_t u4_2; 85 register int i; 86 register int l, h; 87 88 if (! cd) { 89 errno = EBADF; 90 return((size_t)-1); 91 } 92 93 if (!inbuf || !(*inbuf)) { 94 #if defined(UCS_2) || defined(UCS_4) || defined(UTF_16) || defined(UTF_32) 95 cd->bom_written = false; 96 #endif 97 return((size_t)0); 98 } 99 100 ib = (unsigned char *)*inbuf; 101 ob = (unsigned char *)*outbuf; 102 ibtail = ib + *inbufleft; 103 obtail = ob + *outbufleft; 104 105 #if defined(UCS_2) || defined(UCS_4) || defined(UTF_16) || defined(UTF_32) 106 if (! cd->bom_written) { 107 if ((ibtail - ib) < ICV_FETCH_UCS_SIZE) { 108 errno = EINVAL; 109 ret_val = (size_t)-1; 110 goto need_more_input_err; 111 } 112 113 for (u4 = 0, i = 0; i < ICV_FETCH_UCS_SIZE; i++) 114 u4 = (u4 << 8) | ((uint_t)(*(ib + i))); 115 116 if (u4 == ICV_BOM_IN_BIG_ENDIAN) { 117 ib += ICV_FETCH_UCS_SIZE; 118 cd->little_endian = false; 119 } else if (u4 == ICV_BOM_IN_LITTLE_ENDIAN) { 120 ib += ICV_FETCH_UCS_SIZE; 121 cd->little_endian = true; 122 } 123 } 124 cd->bom_written = true; 125 #endif 126 127 while (ib < ibtail) { 128 if ((ibtail - ib) < ICV_FETCH_UCS_SIZE) { 129 errno = EINVAL; 130 ret_val = (size_t)-1; 131 break; 132 } 133 134 u4 = u4_2 = 0; 135 if (cd->little_endian) { 136 for (i = ICV_FETCH_UCS_SIZE - 1; i >= 0; i--) 137 u4 = (u4 << 8) | ((uint_t)(*(ib + i))); 138 } else { 139 for (i = 0; i < ICV_FETCH_UCS_SIZE; i++) 140 u4 = (u4 << 8) | ((uint_t)(*(ib + i))); 141 } 142 143 #if defined(UTF_16) || defined(UTF_16BE) || defined(UTF_16LE) 144 if ((u4 >= 0x00dc00 && u4 <= 0x00dfff) || u4 >= 0x00fffe) { 145 errno = EILSEQ; 146 ret_val = (size_t)-1; 147 break; 148 } 149 150 if (u4 >= 0x00d800 && u4 <= 0x00dbff) { 151 if ((ibtail - ib) < ICV_FETCH_UCS_SIZE_TWO) { 152 errno = EINVAL; 153 ret_val = (size_t)-1; 154 break; 155 } 156 157 if (cd->little_endian) { 158 for (i = ICV_FETCH_UCS_SIZE_TWO - 1; 159 i >= ICV_FETCH_UCS_SIZE; 160 i--) 161 u4_2 = (u4_2<<8)|((uint_t)(*(ib + i))); 162 } else { 163 for (i = ICV_FETCH_UCS_SIZE; 164 i < ICV_FETCH_UCS_SIZE_TWO; 165 i++) 166 u4_2 = (u4_2<<8)|((uint_t)(*(ib + i))); 167 } 168 169 if (u4_2 < 0x00dc00 || u4_2 > 0x00dfff) { 170 errno = EILSEQ; 171 ret_val = (size_t)-1; 172 break; 173 } 174 175 u4 = ((((u4 - 0x00d800) * 0x400) + 176 (u4_2 - 0x00dc00)) & 0x0fffff) + 0x010000; 177 } 178 #elif defined(UCS_2) || defined(UCS_2BE) || defined(UCS_2LE) 179 if (u4 >= 0x00fffe || (u4 >= 0x00d800 && u4 <= 0x00dfff)) { 180 errno = EILSEQ; 181 ret_val = (size_t)-1; 182 break; 183 } 184 #elif defined(UTF_32) || defined(UTF_32BE) || defined(UTF_32LE) 185 if (u4 == 0x00fffe || u4 == 0x00ffff || u4 > 0x10ffff || 186 (u4 >= 0x00d800 && u4 <= 0x00dfff)) { 187 errno = EILSEQ; 188 ret_val = (size_t)-1; 189 break; 190 } 191 #elif defined(UCS_4) || defined(UCS_4BE) || defined(UCS_4LE) 192 if (u4 == 0x00fffe || u4 == 0x00ffff || u4 > 0x7fffffff || 193 (u4 >= 0x00d800 && u4 <= 0x00dfff)) { 194 errno = EILSEQ; 195 ret_val = (size_t)-1; 196 break; 197 } 198 #else 199 #error "Fatal: one of the UCS macros need to be defined." 200 #endif 201 202 if (ob >= obtail) { 203 errno = E2BIG; 204 ret_val = (size_t)-1; 205 break; 206 } 207 208 if (u4 > 0x7f) { 209 i = l = 0; 210 h = (sizeof(u4_sb_tbl) / 211 sizeof(to_sb_table_component_t)) - 1; 212 while (l <= h) { 213 i = (l + h) / 2; 214 if (u4_sb_tbl[i].u8 == u4) 215 break; 216 else if (u4_sb_tbl[i].u8 < u4) 217 l = i + 1; 218 else 219 h = i - 1; 220 } 221 222 /* 223 * We just assume that either we found it or it is 224 * a non-identical character that we need to 225 * provide a replacement character. 226 */ 227 if (u4_sb_tbl[i].u8 == u4) { 228 u4 = u4_sb_tbl[i].sb; 229 } else { 230 u4 = ICV_CHAR_ASCII_REPLACEMENT; 231 ret_val++; 232 } 233 } 234 235 *ob++ = (unsigned char)u4; 236 ib += (u4_2) ? ICV_FETCH_UCS_SIZE_TWO : ICV_FETCH_UCS_SIZE; 237 } 238 #if defined(UCS_2) || defined(UCS_4) || defined(UTF_16) || defined(UTF_32) 239 need_more_input_err: 240 #endif 241 *inbuf = (char *)ib; 242 *inbufleft = ibtail - ib; 243 *outbuf = (char *)ob; 244 *outbufleft = obtail - ob; 245 246 return(ret_val); 247 } 248