| 1 | // © 2016 and later: Unicode, Inc. and others. | 
|---|
| 2 | // License & terms of use: http://www.unicode.org/copyright.html | 
|---|
| 3 | /* | 
|---|
| 4 | ********************************************************************** | 
|---|
| 5 | *   Copyright (C) 2000-2015, International Business Machines | 
|---|
| 6 | *   Corporation and others.  All Rights Reserved. | 
|---|
| 7 | ********************************************************************** | 
|---|
| 8 | *   file name:  ucnvlat1.cpp | 
|---|
| 9 | *   encoding:   UTF-8 | 
|---|
| 10 | *   tab size:   8 (not used) | 
|---|
| 11 | *   indentation:4 | 
|---|
| 12 | * | 
|---|
| 13 | *   created on: 2000feb07 | 
|---|
| 14 | *   created by: Markus W. Scherer | 
|---|
| 15 | */ | 
|---|
| 16 |  | 
|---|
| 17 | #include "unicode/utypes.h" | 
|---|
| 18 |  | 
|---|
| 19 | #if !UCONFIG_NO_CONVERSION | 
|---|
| 20 |  | 
|---|
| 21 | #include "unicode/ucnv.h" | 
|---|
| 22 | #include "unicode/uset.h" | 
|---|
| 23 | #include "unicode/utf8.h" | 
|---|
| 24 | #include "ucnv_bld.h" | 
|---|
| 25 | #include "ucnv_cnv.h" | 
|---|
| 26 | #include "ustr_imp.h" | 
|---|
| 27 |  | 
|---|
| 28 | /* control optimizations according to the platform */ | 
|---|
| 29 | #define LATIN1_UNROLL_FROM_UNICODE 1 | 
|---|
| 30 |  | 
|---|
| 31 | /* ISO 8859-1 --------------------------------------------------------------- */ | 
|---|
| 32 |  | 
|---|
| 33 | /* This is a table-less and callback-less version of ucnv_MBCSSingleToBMPWithOffsets(). */ | 
|---|
| 34 | U_CDECL_BEGIN | 
|---|
| 35 | static void U_CALLCONV | 
|---|
| 36 | _Latin1ToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs, | 
|---|
| 37 | UErrorCode *pErrorCode) { | 
|---|
| 38 | const uint8_t *source; | 
|---|
| 39 | UChar *target; | 
|---|
| 40 | int32_t targetCapacity, length; | 
|---|
| 41 | int32_t *offsets; | 
|---|
| 42 |  | 
|---|
| 43 | int32_t sourceIndex; | 
|---|
| 44 |  | 
|---|
| 45 | /* set up the local pointers */ | 
|---|
| 46 | source=(const uint8_t *)pArgs->source; | 
|---|
| 47 | target=pArgs->target; | 
|---|
| 48 | targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target); | 
|---|
| 49 | offsets=pArgs->offsets; | 
|---|
| 50 |  | 
|---|
| 51 | sourceIndex=0; | 
|---|
| 52 |  | 
|---|
| 53 | /* | 
|---|
| 54 | * since the conversion here is 1:1 UChar:uint8_t, we need only one counter | 
|---|
| 55 | * for the minimum of the sourceLength and targetCapacity | 
|---|
| 56 | */ | 
|---|
| 57 | length=(int32_t)((const uint8_t *)pArgs->sourceLimit-source); | 
|---|
| 58 | if(length<=targetCapacity) { | 
|---|
| 59 | targetCapacity=length; | 
|---|
| 60 | } else { | 
|---|
| 61 | /* target will be full */ | 
|---|
| 62 | *pErrorCode=U_BUFFER_OVERFLOW_ERROR; | 
|---|
| 63 | length=targetCapacity; | 
|---|
| 64 | } | 
|---|
| 65 |  | 
|---|
| 66 | if(targetCapacity>=8) { | 
|---|
| 67 | /* This loop is unrolled for speed and improved pipelining. */ | 
|---|
| 68 | int32_t count, loops; | 
|---|
| 69 |  | 
|---|
| 70 | loops=count=targetCapacity>>3; | 
|---|
| 71 | length=targetCapacity&=0x7; | 
|---|
| 72 | do { | 
|---|
| 73 | target[0]=source[0]; | 
|---|
| 74 | target[1]=source[1]; | 
|---|
| 75 | target[2]=source[2]; | 
|---|
| 76 | target[3]=source[3]; | 
|---|
| 77 | target[4]=source[4]; | 
|---|
| 78 | target[5]=source[5]; | 
|---|
| 79 | target[6]=source[6]; | 
|---|
| 80 | target[7]=source[7]; | 
|---|
| 81 | target+=8; | 
|---|
| 82 | source+=8; | 
|---|
| 83 | } while(--count>0); | 
|---|
| 84 |  | 
|---|
| 85 | if(offsets!=NULL) { | 
|---|
| 86 | do { | 
|---|
| 87 | offsets[0]=sourceIndex++; | 
|---|
| 88 | offsets[1]=sourceIndex++; | 
|---|
| 89 | offsets[2]=sourceIndex++; | 
|---|
| 90 | offsets[3]=sourceIndex++; | 
|---|
| 91 | offsets[4]=sourceIndex++; | 
|---|
| 92 | offsets[5]=sourceIndex++; | 
|---|
| 93 | offsets[6]=sourceIndex++; | 
|---|
| 94 | offsets[7]=sourceIndex++; | 
|---|
| 95 | offsets+=8; | 
|---|
| 96 | } while(--loops>0); | 
|---|
| 97 | } | 
|---|
| 98 | } | 
|---|
| 99 |  | 
|---|
| 100 | /* conversion loop */ | 
|---|
| 101 | while(targetCapacity>0) { | 
|---|
| 102 | *target++=*source++; | 
|---|
| 103 | --targetCapacity; | 
|---|
| 104 | } | 
|---|
| 105 |  | 
|---|
| 106 | /* write back the updated pointers */ | 
|---|
| 107 | pArgs->source=(const char *)source; | 
|---|
| 108 | pArgs->target=target; | 
|---|
| 109 |  | 
|---|
| 110 | /* set offsets */ | 
|---|
| 111 | if(offsets!=NULL) { | 
|---|
| 112 | while(length>0) { | 
|---|
| 113 | *offsets++=sourceIndex++; | 
|---|
| 114 | --length; | 
|---|
| 115 | } | 
|---|
| 116 | pArgs->offsets=offsets; | 
|---|
| 117 | } | 
|---|
| 118 | } | 
|---|
| 119 |  | 
|---|
| 120 | /* This is a table-less and callback-less version of ucnv_MBCSSingleGetNextUChar(). */ | 
|---|
| 121 | static UChar32 U_CALLCONV | 
|---|
| 122 | _Latin1GetNextUChar(UConverterToUnicodeArgs *pArgs, | 
|---|
| 123 | UErrorCode *pErrorCode) { | 
|---|
| 124 | const uint8_t *source=(const uint8_t *)pArgs->source; | 
|---|
| 125 | if(source<(const uint8_t *)pArgs->sourceLimit) { | 
|---|
| 126 | pArgs->source=(const char *)(source+1); | 
|---|
| 127 | return *source; | 
|---|
| 128 | } | 
|---|
| 129 |  | 
|---|
| 130 | /* no output because of empty input */ | 
|---|
| 131 | *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR; | 
|---|
| 132 | return 0xffff; | 
|---|
| 133 | } | 
|---|
| 134 |  | 
|---|
| 135 | /* This is a table-less version of ucnv_MBCSSingleFromBMPWithOffsets(). */ | 
|---|
| 136 | static void U_CALLCONV | 
|---|
| 137 | _Latin1FromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs, | 
|---|
| 138 | UErrorCode *pErrorCode) { | 
|---|
| 139 | UConverter *cnv; | 
|---|
| 140 | const UChar *source, *sourceLimit; | 
|---|
| 141 | uint8_t *target, *oldTarget; | 
|---|
| 142 | int32_t targetCapacity, length; | 
|---|
| 143 | int32_t *offsets; | 
|---|
| 144 |  | 
|---|
| 145 | UChar32 cp; | 
|---|
| 146 | UChar c, max; | 
|---|
| 147 |  | 
|---|
| 148 | int32_t sourceIndex; | 
|---|
| 149 |  | 
|---|
| 150 | /* set up the local pointers */ | 
|---|
| 151 | cnv=pArgs->converter; | 
|---|
| 152 | source=pArgs->source; | 
|---|
| 153 | sourceLimit=pArgs->sourceLimit; | 
|---|
| 154 | target=oldTarget=(uint8_t *)pArgs->target; | 
|---|
| 155 | targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target); | 
|---|
| 156 | offsets=pArgs->offsets; | 
|---|
| 157 |  | 
|---|
| 158 | if(cnv->sharedData==&_Latin1Data) { | 
|---|
| 159 | max=0xff; /* Latin-1 */ | 
|---|
| 160 | } else { | 
|---|
| 161 | max=0x7f; /* US-ASCII */ | 
|---|
| 162 | } | 
|---|
| 163 |  | 
|---|
| 164 | /* get the converter state from UConverter */ | 
|---|
| 165 | cp=cnv->fromUChar32; | 
|---|
| 166 |  | 
|---|
| 167 | /* sourceIndex=-1 if the current character began in the previous buffer */ | 
|---|
| 168 | sourceIndex= cp==0 ? 0 : -1; | 
|---|
| 169 |  | 
|---|
| 170 | /* | 
|---|
| 171 | * since the conversion here is 1:1 UChar:uint8_t, we need only one counter | 
|---|
| 172 | * for the minimum of the sourceLength and targetCapacity | 
|---|
| 173 | */ | 
|---|
| 174 | length=(int32_t)(sourceLimit-source); | 
|---|
| 175 | if(length<targetCapacity) { | 
|---|
| 176 | targetCapacity=length; | 
|---|
| 177 | } | 
|---|
| 178 |  | 
|---|
| 179 | /* conversion loop */ | 
|---|
| 180 | if(cp!=0 && targetCapacity>0) { | 
|---|
| 181 | goto getTrail; | 
|---|
| 182 | } | 
|---|
| 183 |  | 
|---|
| 184 | #if LATIN1_UNROLL_FROM_UNICODE | 
|---|
| 185 | /* unroll the loop with the most common case */ | 
|---|
| 186 | if(targetCapacity>=16) { | 
|---|
| 187 | int32_t count, loops; | 
|---|
| 188 | UChar u, oredChars; | 
|---|
| 189 |  | 
|---|
| 190 | loops=count=targetCapacity>>4; | 
|---|
| 191 | do { | 
|---|
| 192 | oredChars=u=*source++; | 
|---|
| 193 | *target++=(uint8_t)u; | 
|---|
| 194 | oredChars|=u=*source++; | 
|---|
| 195 | *target++=(uint8_t)u; | 
|---|
| 196 | oredChars|=u=*source++; | 
|---|
| 197 | *target++=(uint8_t)u; | 
|---|
| 198 | oredChars|=u=*source++; | 
|---|
| 199 | *target++=(uint8_t)u; | 
|---|
| 200 | oredChars|=u=*source++; | 
|---|
| 201 | *target++=(uint8_t)u; | 
|---|
| 202 | oredChars|=u=*source++; | 
|---|
| 203 | *target++=(uint8_t)u; | 
|---|
| 204 | oredChars|=u=*source++; | 
|---|
| 205 | *target++=(uint8_t)u; | 
|---|
| 206 | oredChars|=u=*source++; | 
|---|
| 207 | *target++=(uint8_t)u; | 
|---|
| 208 | oredChars|=u=*source++; | 
|---|
| 209 | *target++=(uint8_t)u; | 
|---|
| 210 | oredChars|=u=*source++; | 
|---|
| 211 | *target++=(uint8_t)u; | 
|---|
| 212 | oredChars|=u=*source++; | 
|---|
| 213 | *target++=(uint8_t)u; | 
|---|
| 214 | oredChars|=u=*source++; | 
|---|
| 215 | *target++=(uint8_t)u; | 
|---|
| 216 | oredChars|=u=*source++; | 
|---|
| 217 | *target++=(uint8_t)u; | 
|---|
| 218 | oredChars|=u=*source++; | 
|---|
| 219 | *target++=(uint8_t)u; | 
|---|
| 220 | oredChars|=u=*source++; | 
|---|
| 221 | *target++=(uint8_t)u; | 
|---|
| 222 | oredChars|=u=*source++; | 
|---|
| 223 | *target++=(uint8_t)u; | 
|---|
| 224 |  | 
|---|
| 225 | /* were all 16 entries really valid? */ | 
|---|
| 226 | if(oredChars>max) { | 
|---|
| 227 | /* no, return to the first of these 16 */ | 
|---|
| 228 | source-=16; | 
|---|
| 229 | target-=16; | 
|---|
| 230 | break; | 
|---|
| 231 | } | 
|---|
| 232 | } while(--count>0); | 
|---|
| 233 | count=loops-count; | 
|---|
| 234 | targetCapacity-=16*count; | 
|---|
| 235 |  | 
|---|
| 236 | if(offsets!=NULL) { | 
|---|
| 237 | oldTarget+=16*count; | 
|---|
| 238 | while(count>0) { | 
|---|
| 239 | *offsets++=sourceIndex++; | 
|---|
| 240 | *offsets++=sourceIndex++; | 
|---|
| 241 | *offsets++=sourceIndex++; | 
|---|
| 242 | *offsets++=sourceIndex++; | 
|---|
| 243 | *offsets++=sourceIndex++; | 
|---|
| 244 | *offsets++=sourceIndex++; | 
|---|
| 245 | *offsets++=sourceIndex++; | 
|---|
| 246 | *offsets++=sourceIndex++; | 
|---|
| 247 | *offsets++=sourceIndex++; | 
|---|
| 248 | *offsets++=sourceIndex++; | 
|---|
| 249 | *offsets++=sourceIndex++; | 
|---|
| 250 | *offsets++=sourceIndex++; | 
|---|
| 251 | *offsets++=sourceIndex++; | 
|---|
| 252 | *offsets++=sourceIndex++; | 
|---|
| 253 | *offsets++=sourceIndex++; | 
|---|
| 254 | *offsets++=sourceIndex++; | 
|---|
| 255 | --count; | 
|---|
| 256 | } | 
|---|
| 257 | } | 
|---|
| 258 | } | 
|---|
| 259 | #endif | 
|---|
| 260 |  | 
|---|
| 261 | /* conversion loop */ | 
|---|
| 262 | c=0; | 
|---|
| 263 | while(targetCapacity>0 && (c=*source++)<=max) { | 
|---|
| 264 | /* convert the Unicode code point */ | 
|---|
| 265 | *target++=(uint8_t)c; | 
|---|
| 266 | --targetCapacity; | 
|---|
| 267 | } | 
|---|
| 268 |  | 
|---|
| 269 | if(c>max) { | 
|---|
| 270 | cp=c; | 
|---|
| 271 | if(!U_IS_SURROGATE(cp)) { | 
|---|
| 272 | /* callback(unassigned) */ | 
|---|
| 273 | } else if(U_IS_SURROGATE_LEAD(cp)) { | 
|---|
| 274 | getTrail: | 
|---|
| 275 | if(source<sourceLimit) { | 
|---|
| 276 | /* test the following code unit */ | 
|---|
| 277 | UChar trail=*source; | 
|---|
| 278 | if(U16_IS_TRAIL(trail)) { | 
|---|
| 279 | ++source; | 
|---|
| 280 | cp=U16_GET_SUPPLEMENTARY(cp, trail); | 
|---|
| 281 | /* this codepage does not map supplementary code points */ | 
|---|
| 282 | /* callback(unassigned) */ | 
|---|
| 283 | } else { | 
|---|
| 284 | /* this is an unmatched lead code unit (1st surrogate) */ | 
|---|
| 285 | /* callback(illegal) */ | 
|---|
| 286 | } | 
|---|
| 287 | } else { | 
|---|
| 288 | /* no more input */ | 
|---|
| 289 | cnv->fromUChar32=cp; | 
|---|
| 290 | goto noMoreInput; | 
|---|
| 291 | } | 
|---|
| 292 | } else { | 
|---|
| 293 | /* this is an unmatched trail code unit (2nd surrogate) */ | 
|---|
| 294 | /* callback(illegal) */ | 
|---|
| 295 | } | 
|---|
| 296 |  | 
|---|
| 297 | *pErrorCode= U_IS_SURROGATE(cp) ? U_ILLEGAL_CHAR_FOUND : U_INVALID_CHAR_FOUND; | 
|---|
| 298 | cnv->fromUChar32=cp; | 
|---|
| 299 | } | 
|---|
| 300 | noMoreInput: | 
|---|
| 301 |  | 
|---|
| 302 | /* set offsets since the start */ | 
|---|
| 303 | if(offsets!=NULL) { | 
|---|
| 304 | size_t count=target-oldTarget; | 
|---|
| 305 | while(count>0) { | 
|---|
| 306 | *offsets++=sourceIndex++; | 
|---|
| 307 | --count; | 
|---|
| 308 | } | 
|---|
| 309 | } | 
|---|
| 310 |  | 
|---|
| 311 | if(U_SUCCESS(*pErrorCode) && source<sourceLimit && target>=(uint8_t *)pArgs->targetLimit) { | 
|---|
| 312 | /* target is full */ | 
|---|
| 313 | *pErrorCode=U_BUFFER_OVERFLOW_ERROR; | 
|---|
| 314 | } | 
|---|
| 315 |  | 
|---|
| 316 | /* write back the updated pointers */ | 
|---|
| 317 | pArgs->source=source; | 
|---|
| 318 | pArgs->target=(char *)target; | 
|---|
| 319 | pArgs->offsets=offsets; | 
|---|
| 320 | } | 
|---|
| 321 |  | 
|---|
| 322 | /* Convert UTF-8 to Latin-1. Adapted from ucnv_SBCSFromUTF8(). */ | 
|---|
| 323 | static void U_CALLCONV | 
|---|
| 324 | ucnv_Latin1FromUTF8(UConverterFromUnicodeArgs *pFromUArgs, | 
|---|
| 325 | UConverterToUnicodeArgs *pToUArgs, | 
|---|
| 326 | UErrorCode *pErrorCode) { | 
|---|
| 327 | UConverter *utf8; | 
|---|
| 328 | const uint8_t *source, *sourceLimit; | 
|---|
| 329 | uint8_t *target; | 
|---|
| 330 | int32_t targetCapacity; | 
|---|
| 331 |  | 
|---|
| 332 | UChar32 c; | 
|---|
| 333 | uint8_t b, t1; | 
|---|
| 334 |  | 
|---|
| 335 | /* set up the local pointers */ | 
|---|
| 336 | utf8=pToUArgs->converter; | 
|---|
| 337 | source=(uint8_t *)pToUArgs->source; | 
|---|
| 338 | sourceLimit=(uint8_t *)pToUArgs->sourceLimit; | 
|---|
| 339 | target=(uint8_t *)pFromUArgs->target; | 
|---|
| 340 | targetCapacity=(int32_t)(pFromUArgs->targetLimit-pFromUArgs->target); | 
|---|
| 341 |  | 
|---|
| 342 | /* get the converter state from the UTF-8 UConverter */ | 
|---|
| 343 | if (utf8->toULength > 0) { | 
|---|
| 344 | c=(UChar32)utf8->toUnicodeStatus; | 
|---|
| 345 | } else { | 
|---|
| 346 | c = 0; | 
|---|
| 347 | } | 
|---|
| 348 | if(c!=0 && source<sourceLimit) { | 
|---|
| 349 | if(targetCapacity==0) { | 
|---|
| 350 | *pErrorCode=U_BUFFER_OVERFLOW_ERROR; | 
|---|
| 351 | return; | 
|---|
| 352 | } else if(c>=0xc2 && c<=0xc3 && (t1=(uint8_t)(*source-0x80)) <= 0x3f) { | 
|---|
| 353 | ++source; | 
|---|
| 354 | *target++=(uint8_t)(((c&3)<<6)|t1); | 
|---|
| 355 | --targetCapacity; | 
|---|
| 356 |  | 
|---|
| 357 | utf8->toUnicodeStatus=0; | 
|---|
| 358 | utf8->toULength=0; | 
|---|
| 359 | } else { | 
|---|
| 360 | /* complicated, illegal or unmappable input: fall back to the pivoting implementation */ | 
|---|
| 361 | *pErrorCode=U_USING_DEFAULT_WARNING; | 
|---|
| 362 | return; | 
|---|
| 363 | } | 
|---|
| 364 | } | 
|---|
| 365 |  | 
|---|
| 366 | /* | 
|---|
| 367 | * Make sure that the last byte sequence before sourceLimit is complete | 
|---|
| 368 | * or runs into a lead byte. | 
|---|
| 369 | * In the conversion loop compare source with sourceLimit only once | 
|---|
| 370 | * per multi-byte character. | 
|---|
| 371 | * For Latin-1, adjust sourceLimit only for 1 trail byte because | 
|---|
| 372 | * the conversion loop handles at most 2-byte sequences. | 
|---|
| 373 | */ | 
|---|
| 374 | if(source<sourceLimit && U8_IS_LEAD(*(sourceLimit-1))) { | 
|---|
| 375 | --sourceLimit; | 
|---|
| 376 | } | 
|---|
| 377 |  | 
|---|
| 378 | /* conversion loop */ | 
|---|
| 379 | while(source<sourceLimit) { | 
|---|
| 380 | if(targetCapacity>0) { | 
|---|
| 381 | b=*source++; | 
|---|
| 382 | if(U8_IS_SINGLE(b)) { | 
|---|
| 383 | /* convert ASCII */ | 
|---|
| 384 | *target++=(uint8_t)b; | 
|---|
| 385 | --targetCapacity; | 
|---|
| 386 | } else if( /* handle U+0080..U+00FF inline */ | 
|---|
| 387 | b>=0xc2 && b<=0xc3 && | 
|---|
| 388 | (t1=(uint8_t)(*source-0x80)) <= 0x3f | 
|---|
| 389 | ) { | 
|---|
| 390 | ++source; | 
|---|
| 391 | *target++=(uint8_t)(((b&3)<<6)|t1); | 
|---|
| 392 | --targetCapacity; | 
|---|
| 393 | } else { | 
|---|
| 394 | /* complicated, illegal or unmappable input: fall back to the pivoting implementation */ | 
|---|
| 395 | pToUArgs->source=(char *)(source-1); | 
|---|
| 396 | pFromUArgs->target=(char *)target; | 
|---|
| 397 | *pErrorCode=U_USING_DEFAULT_WARNING; | 
|---|
| 398 | return; | 
|---|
| 399 | } | 
|---|
| 400 | } else { | 
|---|
| 401 | /* target is full */ | 
|---|
| 402 | *pErrorCode=U_BUFFER_OVERFLOW_ERROR; | 
|---|
| 403 | break; | 
|---|
| 404 | } | 
|---|
| 405 | } | 
|---|
| 406 |  | 
|---|
| 407 | /* | 
|---|
| 408 | * The sourceLimit may have been adjusted before the conversion loop | 
|---|
| 409 | * to stop before a truncated sequence. | 
|---|
| 410 | * If so, then collect the truncated sequence now. | 
|---|
| 411 | * For Latin-1, there is at most exactly one lead byte because of the | 
|---|
| 412 | * smaller sourceLimit adjustment logic. | 
|---|
| 413 | */ | 
|---|
| 414 | if(U_SUCCESS(*pErrorCode) && source<(sourceLimit=(uint8_t *)pToUArgs->sourceLimit)) { | 
|---|
| 415 | utf8->toUnicodeStatus=utf8->toUBytes[0]=b=*source++; | 
|---|
| 416 | utf8->toULength=1; | 
|---|
| 417 | utf8->mode=U8_COUNT_BYTES(b); | 
|---|
| 418 | } | 
|---|
| 419 |  | 
|---|
| 420 | /* write back the updated pointers */ | 
|---|
| 421 | pToUArgs->source=(char *)source; | 
|---|
| 422 | pFromUArgs->target=(char *)target; | 
|---|
| 423 | } | 
|---|
| 424 |  | 
|---|
| 425 | static void U_CALLCONV | 
|---|
| 426 | _Latin1GetUnicodeSet(const UConverter *cnv, | 
|---|
| 427 | const USetAdder *sa, | 
|---|
| 428 | UConverterUnicodeSet which, | 
|---|
| 429 | UErrorCode *pErrorCode) { | 
|---|
| 430 | (void)cnv; | 
|---|
| 431 | (void)which; | 
|---|
| 432 | (void)pErrorCode; | 
|---|
| 433 | sa->addRange(sa->set, 0, 0xff); | 
|---|
| 434 | } | 
|---|
| 435 | U_CDECL_END | 
|---|
| 436 |  | 
|---|
| 437 |  | 
|---|
| 438 | static const UConverterImpl _Latin1Impl={ | 
|---|
| 439 | UCNV_LATIN_1, | 
|---|
| 440 |  | 
|---|
| 441 | NULL, | 
|---|
| 442 | NULL, | 
|---|
| 443 |  | 
|---|
| 444 | NULL, | 
|---|
| 445 | NULL, | 
|---|
| 446 | NULL, | 
|---|
| 447 |  | 
|---|
| 448 | _Latin1ToUnicodeWithOffsets, | 
|---|
| 449 | _Latin1ToUnicodeWithOffsets, | 
|---|
| 450 | _Latin1FromUnicodeWithOffsets, | 
|---|
| 451 | _Latin1FromUnicodeWithOffsets, | 
|---|
| 452 | _Latin1GetNextUChar, | 
|---|
| 453 |  | 
|---|
| 454 | NULL, | 
|---|
| 455 | NULL, | 
|---|
| 456 | NULL, | 
|---|
| 457 | NULL, | 
|---|
| 458 | _Latin1GetUnicodeSet, | 
|---|
| 459 |  | 
|---|
| 460 | NULL, | 
|---|
| 461 | ucnv_Latin1FromUTF8 | 
|---|
| 462 | }; | 
|---|
| 463 |  | 
|---|
| 464 | static const UConverterStaticData _Latin1StaticData={ | 
|---|
| 465 | sizeof(UConverterStaticData), | 
|---|
| 466 | "ISO-8859-1", | 
|---|
| 467 | 819, UCNV_IBM, UCNV_LATIN_1, 1, 1, | 
|---|
| 468 | { 0x1a, 0, 0, 0 }, 1, FALSE, FALSE, | 
|---|
| 469 | 0, | 
|---|
| 470 | 0, | 
|---|
| 471 | { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */ | 
|---|
| 472 | }; | 
|---|
| 473 |  | 
|---|
| 474 | const UConverterSharedData _Latin1Data= | 
|---|
| 475 | UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_Latin1StaticData, &_Latin1Impl); | 
|---|
| 476 |  | 
|---|
| 477 | /* US-ASCII ----------------------------------------------------------------- */ | 
|---|
| 478 |  | 
|---|
| 479 | U_CDECL_BEGIN | 
|---|
| 480 | /* This is a table-less version of ucnv_MBCSSingleToBMPWithOffsets(). */ | 
|---|
| 481 | static void U_CALLCONV | 
|---|
| 482 | _ASCIIToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs, | 
|---|
| 483 | UErrorCode *pErrorCode) { | 
|---|
| 484 | const uint8_t *source, *sourceLimit; | 
|---|
| 485 | UChar *target, *oldTarget; | 
|---|
| 486 | int32_t targetCapacity, length; | 
|---|
| 487 | int32_t *offsets; | 
|---|
| 488 |  | 
|---|
| 489 | int32_t sourceIndex; | 
|---|
| 490 |  | 
|---|
| 491 | uint8_t c; | 
|---|
| 492 |  | 
|---|
| 493 | /* set up the local pointers */ | 
|---|
| 494 | source=(const uint8_t *)pArgs->source; | 
|---|
| 495 | sourceLimit=(const uint8_t *)pArgs->sourceLimit; | 
|---|
| 496 | target=oldTarget=pArgs->target; | 
|---|
| 497 | targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target); | 
|---|
| 498 | offsets=pArgs->offsets; | 
|---|
| 499 |  | 
|---|
| 500 | /* sourceIndex=-1 if the current character began in the previous buffer */ | 
|---|
| 501 | sourceIndex=0; | 
|---|
| 502 |  | 
|---|
| 503 | /* | 
|---|
| 504 | * since the conversion here is 1:1 UChar:uint8_t, we need only one counter | 
|---|
| 505 | * for the minimum of the sourceLength and targetCapacity | 
|---|
| 506 | */ | 
|---|
| 507 | length=(int32_t)(sourceLimit-source); | 
|---|
| 508 | if(length<targetCapacity) { | 
|---|
| 509 | targetCapacity=length; | 
|---|
| 510 | } | 
|---|
| 511 |  | 
|---|
| 512 | if(targetCapacity>=8) { | 
|---|
| 513 | /* This loop is unrolled for speed and improved pipelining. */ | 
|---|
| 514 | int32_t count, loops; | 
|---|
| 515 | UChar oredChars; | 
|---|
| 516 |  | 
|---|
| 517 | loops=count=targetCapacity>>3; | 
|---|
| 518 | do { | 
|---|
| 519 | oredChars=target[0]=source[0]; | 
|---|
| 520 | oredChars|=target[1]=source[1]; | 
|---|
| 521 | oredChars|=target[2]=source[2]; | 
|---|
| 522 | oredChars|=target[3]=source[3]; | 
|---|
| 523 | oredChars|=target[4]=source[4]; | 
|---|
| 524 | oredChars|=target[5]=source[5]; | 
|---|
| 525 | oredChars|=target[6]=source[6]; | 
|---|
| 526 | oredChars|=target[7]=source[7]; | 
|---|
| 527 |  | 
|---|
| 528 | /* were all 16 entries really valid? */ | 
|---|
| 529 | if(oredChars>0x7f) { | 
|---|
| 530 | /* no, return to the first of these 16 */ | 
|---|
| 531 | break; | 
|---|
| 532 | } | 
|---|
| 533 | source+=8; | 
|---|
| 534 | target+=8; | 
|---|
| 535 | } while(--count>0); | 
|---|
| 536 | count=loops-count; | 
|---|
| 537 | targetCapacity-=count*8; | 
|---|
| 538 |  | 
|---|
| 539 | if(offsets!=NULL) { | 
|---|
| 540 | oldTarget+=count*8; | 
|---|
| 541 | while(count>0) { | 
|---|
| 542 | offsets[0]=sourceIndex++; | 
|---|
| 543 | offsets[1]=sourceIndex++; | 
|---|
| 544 | offsets[2]=sourceIndex++; | 
|---|
| 545 | offsets[3]=sourceIndex++; | 
|---|
| 546 | offsets[4]=sourceIndex++; | 
|---|
| 547 | offsets[5]=sourceIndex++; | 
|---|
| 548 | offsets[6]=sourceIndex++; | 
|---|
| 549 | offsets[7]=sourceIndex++; | 
|---|
| 550 | offsets+=8; | 
|---|
| 551 | --count; | 
|---|
| 552 | } | 
|---|
| 553 | } | 
|---|
| 554 | } | 
|---|
| 555 |  | 
|---|
| 556 | /* conversion loop */ | 
|---|
| 557 | c=0; | 
|---|
| 558 | while(targetCapacity>0 && (c=*source++)<=0x7f) { | 
|---|
| 559 | *target++=c; | 
|---|
| 560 | --targetCapacity; | 
|---|
| 561 | } | 
|---|
| 562 |  | 
|---|
| 563 | if(c>0x7f) { | 
|---|
| 564 | /* callback(illegal); copy the current bytes to toUBytes[] */ | 
|---|
| 565 | UConverter *cnv=pArgs->converter; | 
|---|
| 566 | cnv->toUBytes[0]=c; | 
|---|
| 567 | cnv->toULength=1; | 
|---|
| 568 | *pErrorCode=U_ILLEGAL_CHAR_FOUND; | 
|---|
| 569 | } else if(source<sourceLimit && target>=pArgs->targetLimit) { | 
|---|
| 570 | /* target is full */ | 
|---|
| 571 | *pErrorCode=U_BUFFER_OVERFLOW_ERROR; | 
|---|
| 572 | } | 
|---|
| 573 |  | 
|---|
| 574 | /* set offsets since the start */ | 
|---|
| 575 | if(offsets!=NULL) { | 
|---|
| 576 | size_t count=target-oldTarget; | 
|---|
| 577 | while(count>0) { | 
|---|
| 578 | *offsets++=sourceIndex++; | 
|---|
| 579 | --count; | 
|---|
| 580 | } | 
|---|
| 581 | } | 
|---|
| 582 |  | 
|---|
| 583 | /* write back the updated pointers */ | 
|---|
| 584 | pArgs->source=(const char *)source; | 
|---|
| 585 | pArgs->target=target; | 
|---|
| 586 | pArgs->offsets=offsets; | 
|---|
| 587 | } | 
|---|
| 588 |  | 
|---|
| 589 | /* This is a table-less version of ucnv_MBCSSingleGetNextUChar(). */ | 
|---|
| 590 | static UChar32 U_CALLCONV | 
|---|
| 591 | _ASCIIGetNextUChar(UConverterToUnicodeArgs *pArgs, | 
|---|
| 592 | UErrorCode *pErrorCode) { | 
|---|
| 593 | const uint8_t *source; | 
|---|
| 594 | uint8_t b; | 
|---|
| 595 |  | 
|---|
| 596 | source=(const uint8_t *)pArgs->source; | 
|---|
| 597 | if(source<(const uint8_t *)pArgs->sourceLimit) { | 
|---|
| 598 | b=*source++; | 
|---|
| 599 | pArgs->source=(const char *)source; | 
|---|
| 600 | if(b<=0x7f) { | 
|---|
| 601 | return b; | 
|---|
| 602 | } else { | 
|---|
| 603 | UConverter *cnv=pArgs->converter; | 
|---|
| 604 | cnv->toUBytes[0]=b; | 
|---|
| 605 | cnv->toULength=1; | 
|---|
| 606 | *pErrorCode=U_ILLEGAL_CHAR_FOUND; | 
|---|
| 607 | return 0xffff; | 
|---|
| 608 | } | 
|---|
| 609 | } | 
|---|
| 610 |  | 
|---|
| 611 | /* no output because of empty input */ | 
|---|
| 612 | *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR; | 
|---|
| 613 | return 0xffff; | 
|---|
| 614 | } | 
|---|
| 615 |  | 
|---|
| 616 | /* "Convert" UTF-8 to US-ASCII: Validate and copy. */ | 
|---|
| 617 | static void U_CALLCONV | 
|---|
| 618 | ucnv_ASCIIFromUTF8(UConverterFromUnicodeArgs *pFromUArgs, | 
|---|
| 619 | UConverterToUnicodeArgs *pToUArgs, | 
|---|
| 620 | UErrorCode *pErrorCode) { | 
|---|
| 621 | const uint8_t *source, *sourceLimit; | 
|---|
| 622 | uint8_t *target; | 
|---|
| 623 | int32_t targetCapacity, length; | 
|---|
| 624 |  | 
|---|
| 625 | uint8_t c; | 
|---|
| 626 |  | 
|---|
| 627 | if(pToUArgs->converter->toULength > 0) { | 
|---|
| 628 | /* no handling of partial UTF-8 characters here, fall back to pivoting */ | 
|---|
| 629 | *pErrorCode=U_USING_DEFAULT_WARNING; | 
|---|
| 630 | return; | 
|---|
| 631 | } | 
|---|
| 632 |  | 
|---|
| 633 | /* set up the local pointers */ | 
|---|
| 634 | source=(const uint8_t *)pToUArgs->source; | 
|---|
| 635 | sourceLimit=(const uint8_t *)pToUArgs->sourceLimit; | 
|---|
| 636 | target=(uint8_t *)pFromUArgs->target; | 
|---|
| 637 | targetCapacity=(int32_t)(pFromUArgs->targetLimit-pFromUArgs->target); | 
|---|
| 638 |  | 
|---|
| 639 | /* | 
|---|
| 640 | * since the conversion here is 1:1 uint8_t:uint8_t, we need only one counter | 
|---|
| 641 | * for the minimum of the sourceLength and targetCapacity | 
|---|
| 642 | */ | 
|---|
| 643 | length=(int32_t)(sourceLimit-source); | 
|---|
| 644 | if(length<targetCapacity) { | 
|---|
| 645 | targetCapacity=length; | 
|---|
| 646 | } | 
|---|
| 647 |  | 
|---|
| 648 | /* unroll the loop with the most common case */ | 
|---|
| 649 | if(targetCapacity>=16) { | 
|---|
| 650 | int32_t count, loops; | 
|---|
| 651 | uint8_t oredChars; | 
|---|
| 652 |  | 
|---|
| 653 | loops=count=targetCapacity>>4; | 
|---|
| 654 | do { | 
|---|
| 655 | oredChars=*target++=*source++; | 
|---|
| 656 | oredChars|=*target++=*source++; | 
|---|
| 657 | oredChars|=*target++=*source++; | 
|---|
| 658 | oredChars|=*target++=*source++; | 
|---|
| 659 | oredChars|=*target++=*source++; | 
|---|
| 660 | oredChars|=*target++=*source++; | 
|---|
| 661 | oredChars|=*target++=*source++; | 
|---|
| 662 | oredChars|=*target++=*source++; | 
|---|
| 663 | oredChars|=*target++=*source++; | 
|---|
| 664 | oredChars|=*target++=*source++; | 
|---|
| 665 | oredChars|=*target++=*source++; | 
|---|
| 666 | oredChars|=*target++=*source++; | 
|---|
| 667 | oredChars|=*target++=*source++; | 
|---|
| 668 | oredChars|=*target++=*source++; | 
|---|
| 669 | oredChars|=*target++=*source++; | 
|---|
| 670 | oredChars|=*target++=*source++; | 
|---|
| 671 |  | 
|---|
| 672 | /* were all 16 entries really valid? */ | 
|---|
| 673 | if(oredChars>0x7f) { | 
|---|
| 674 | /* no, return to the first of these 16 */ | 
|---|
| 675 | source-=16; | 
|---|
| 676 | target-=16; | 
|---|
| 677 | break; | 
|---|
| 678 | } | 
|---|
| 679 | } while(--count>0); | 
|---|
| 680 | count=loops-count; | 
|---|
| 681 | targetCapacity-=16*count; | 
|---|
| 682 | } | 
|---|
| 683 |  | 
|---|
| 684 | /* conversion loop */ | 
|---|
| 685 | c=0; | 
|---|
| 686 | while(targetCapacity>0 && (c=*source)<=0x7f) { | 
|---|
| 687 | ++source; | 
|---|
| 688 | *target++=c; | 
|---|
| 689 | --targetCapacity; | 
|---|
| 690 | } | 
|---|
| 691 |  | 
|---|
| 692 | if(c>0x7f) { | 
|---|
| 693 | /* non-ASCII character, handle in standard converter */ | 
|---|
| 694 | *pErrorCode=U_USING_DEFAULT_WARNING; | 
|---|
| 695 | } else if(source<sourceLimit && target>=(const uint8_t *)pFromUArgs->targetLimit) { | 
|---|
| 696 | /* target is full */ | 
|---|
| 697 | *pErrorCode=U_BUFFER_OVERFLOW_ERROR; | 
|---|
| 698 | } | 
|---|
| 699 |  | 
|---|
| 700 | /* write back the updated pointers */ | 
|---|
| 701 | pToUArgs->source=(const char *)source; | 
|---|
| 702 | pFromUArgs->target=(char *)target; | 
|---|
| 703 | } | 
|---|
| 704 |  | 
|---|
| 705 | static void U_CALLCONV | 
|---|
| 706 | _ASCIIGetUnicodeSet(const UConverter *cnv, | 
|---|
| 707 | const USetAdder *sa, | 
|---|
| 708 | UConverterUnicodeSet which, | 
|---|
| 709 | UErrorCode *pErrorCode) { | 
|---|
| 710 | (void)cnv; | 
|---|
| 711 | (void)which; | 
|---|
| 712 | (void)pErrorCode; | 
|---|
| 713 | sa->addRange(sa->set, 0, 0x7f); | 
|---|
| 714 | } | 
|---|
| 715 | U_CDECL_END | 
|---|
| 716 |  | 
|---|
| 717 | static const UConverterImpl _ASCIIImpl={ | 
|---|
| 718 | UCNV_US_ASCII, | 
|---|
| 719 |  | 
|---|
| 720 | NULL, | 
|---|
| 721 | NULL, | 
|---|
| 722 |  | 
|---|
| 723 | NULL, | 
|---|
| 724 | NULL, | 
|---|
| 725 | NULL, | 
|---|
| 726 |  | 
|---|
| 727 | _ASCIIToUnicodeWithOffsets, | 
|---|
| 728 | _ASCIIToUnicodeWithOffsets, | 
|---|
| 729 | _Latin1FromUnicodeWithOffsets, | 
|---|
| 730 | _Latin1FromUnicodeWithOffsets, | 
|---|
| 731 | _ASCIIGetNextUChar, | 
|---|
| 732 |  | 
|---|
| 733 | NULL, | 
|---|
| 734 | NULL, | 
|---|
| 735 | NULL, | 
|---|
| 736 | NULL, | 
|---|
| 737 | _ASCIIGetUnicodeSet, | 
|---|
| 738 |  | 
|---|
| 739 | NULL, | 
|---|
| 740 | ucnv_ASCIIFromUTF8 | 
|---|
| 741 | }; | 
|---|
| 742 |  | 
|---|
| 743 | static const UConverterStaticData _ASCIIStaticData={ | 
|---|
| 744 | sizeof(UConverterStaticData), | 
|---|
| 745 | "US-ASCII", | 
|---|
| 746 | 367, UCNV_IBM, UCNV_US_ASCII, 1, 1, | 
|---|
| 747 | { 0x1a, 0, 0, 0 }, 1, FALSE, FALSE, | 
|---|
| 748 | 0, | 
|---|
| 749 | 0, | 
|---|
| 750 | { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */ | 
|---|
| 751 | }; | 
|---|
| 752 |  | 
|---|
| 753 | const UConverterSharedData _ASCIIData= | 
|---|
| 754 | UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_ASCIIStaticData, &_ASCIIImpl); | 
|---|
| 755 |  | 
|---|
| 756 | #endif | 
|---|
| 757 |  | 
|---|