123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474 |
- // © 2016 and later: Unicode, Inc. and others.
- // License & terms of use: http://www.unicode.org/copyright.html
- /*
- *******************************************************************************
- *
- * Copyright (C) 2003-2014, International Business Machines
- * Corporation and others. All Rights Reserved.
- *
- *******************************************************************************
- * file name: udataswp.c
- * encoding: UTF-8
- * tab size: 8 (not used)
- * indentation:4
- *
- * created on: 2003jun05
- * created by: Markus W. Scherer
- *
- * Definitions for ICU data transformations for different platforms,
- * changing between big- and little-endian data and/or between
- * charset families (ASCII<->EBCDIC).
- */
- #include <stdarg.h>
- #include "unicode/utypes.h"
- #include "unicode/udata.h" /* UDataInfo */
- #include "ucmndata.h" /* DataHeader */
- #include "cmemory.h"
- #include "udataswp.h"
- /* swapping primitives ------------------------------------------------------ */
- static int32_t U_CALLCONV
- uprv_swapArray16(const UDataSwapper *ds,
- const void *inData, int32_t length, void *outData,
- UErrorCode *pErrorCode) {
- const uint16_t *p;
- uint16_t *q;
- int32_t count;
- uint16_t x;
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return 0;
- }
- if(ds==nullptr || inData==nullptr || length<0 || (length&1)!=0 || outData==nullptr) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return 0;
- }
- /* setup and swapping */
- p=(const uint16_t *)inData;
- q=(uint16_t *)outData;
- count=length/2;
- while(count>0) {
- x=*p++;
- *q++=(uint16_t)((x<<8)|(x>>8));
- --count;
- }
- return length;
- }
- static int32_t U_CALLCONV
- uprv_copyArray16(const UDataSwapper *ds,
- const void *inData, int32_t length, void *outData,
- UErrorCode *pErrorCode) {
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return 0;
- }
- if(ds==nullptr || inData==nullptr || length<0 || (length&1)!=0 || outData==nullptr) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return 0;
- }
- if(length>0 && inData!=outData) {
- uprv_memcpy(outData, inData, length);
- }
- return length;
- }
- static int32_t U_CALLCONV
- uprv_swapArray32(const UDataSwapper *ds,
- const void *inData, int32_t length, void *outData,
- UErrorCode *pErrorCode) {
- const uint32_t *p;
- uint32_t *q;
- int32_t count;
- uint32_t x;
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return 0;
- }
- if(ds==nullptr || inData==nullptr || length<0 || (length&3)!=0 || outData==nullptr) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return 0;
- }
- /* setup and swapping */
- p=(const uint32_t *)inData;
- q=(uint32_t *)outData;
- count=length/4;
- while(count>0) {
- x=*p++;
- *q++=(uint32_t)((x<<24)|((x<<8)&0xff0000)|((x>>8)&0xff00)|(x>>24));
- --count;
- }
- return length;
- }
- static int32_t U_CALLCONV
- uprv_copyArray32(const UDataSwapper *ds,
- const void *inData, int32_t length, void *outData,
- UErrorCode *pErrorCode) {
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return 0;
- }
- if(ds==nullptr || inData==nullptr || length<0 || (length&3)!=0 || outData==nullptr) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return 0;
- }
- if(length>0 && inData!=outData) {
- uprv_memcpy(outData, inData, length);
- }
- return length;
- }
- static int32_t U_CALLCONV
- uprv_swapArray64(const UDataSwapper *ds,
- const void *inData, int32_t length, void *outData,
- UErrorCode *pErrorCode) {
- const uint64_t *p;
- uint64_t *q;
- int32_t count;
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return 0;
- }
- if(ds==nullptr || inData==nullptr || length<0 || (length&7)!=0 || outData==nullptr) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return 0;
- }
- /* setup and swapping */
- p=(const uint64_t *)inData;
- q=(uint64_t *)outData;
- count=length/8;
- while(count>0) {
- uint64_t x=*p++;
- x=(x<<56)|((x&0xff00)<<40)|((x&0xff0000)<<24)|((x&0xff000000)<<8)|
- ((x>>8)&0xff000000)|((x>>24)&0xff0000)|((x>>40)&0xff00)|(x>>56);
- *q++=x;
- --count;
- }
- return length;
- }
- static int32_t U_CALLCONV
- uprv_copyArray64(const UDataSwapper *ds,
- const void *inData, int32_t length, void *outData,
- UErrorCode *pErrorCode) {
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return 0;
- }
- if(ds==nullptr || inData==nullptr || length<0 || (length&7)!=0 || outData==nullptr) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return 0;
- }
- if(length>0 && inData!=outData) {
- uprv_memcpy(outData, inData, length);
- }
- return length;
- }
- static uint16_t U_CALLCONV
- uprv_readSwapUInt16(uint16_t x) {
- return (uint16_t)((x<<8)|(x>>8));
- }
- static uint16_t U_CALLCONV
- uprv_readDirectUInt16(uint16_t x) {
- return x;
- }
- static uint32_t U_CALLCONV
- uprv_readSwapUInt32(uint32_t x) {
- return (uint32_t)((x<<24)|((x<<8)&0xff0000)|((x>>8)&0xff00)|(x>>24));
- }
- static uint32_t U_CALLCONV
- uprv_readDirectUInt32(uint32_t x) {
- return x;
- }
- static void U_CALLCONV
- uprv_writeSwapUInt16(uint16_t *p, uint16_t x) {
- *p=(uint16_t)((x<<8)|(x>>8));
- }
- static void U_CALLCONV
- uprv_writeDirectUInt16(uint16_t *p, uint16_t x) {
- *p=x;
- }
- static void U_CALLCONV
- uprv_writeSwapUInt32(uint32_t *p, uint32_t x) {
- *p=(uint32_t)((x<<24)|((x<<8)&0xff0000)|((x>>8)&0xff00)|(x>>24));
- }
- static void U_CALLCONV
- uprv_writeDirectUInt32(uint32_t *p, uint32_t x) {
- *p=x;
- }
- U_CAPI int16_t U_EXPORT2
- udata_readInt16(const UDataSwapper *ds, int16_t x) {
- return (int16_t)ds->readUInt16((uint16_t)x);
- }
- U_CAPI int32_t U_EXPORT2
- udata_readInt32(const UDataSwapper *ds, int32_t x) {
- return (int32_t)ds->readUInt32((uint32_t)x);
- }
- /**
- * Swap a block of invariant, NUL-terminated strings, but not padding
- * bytes after the last string.
- * @internal
- */
- U_CAPI int32_t U_EXPORT2
- udata_swapInvStringBlock(const UDataSwapper *ds,
- const void *inData, int32_t length, void *outData,
- UErrorCode *pErrorCode) {
- const char *inChars;
- int32_t stringsLength;
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return 0;
- }
- if(ds==nullptr || inData==nullptr || length<0 || (length>0 && outData==nullptr)) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return 0;
- }
- /* reduce the strings length to not include bytes after the last NUL */
- inChars=(const char *)inData;
- stringsLength=length;
- while(stringsLength>0 && inChars[stringsLength-1]!=0) {
- --stringsLength;
- }
- /* swap up to the last NUL */
- ds->swapInvChars(ds, inData, stringsLength, outData, pErrorCode);
- /* copy the bytes after the last NUL */
- if(inData!=outData && length>stringsLength) {
- uprv_memcpy((char *)outData+stringsLength, inChars+stringsLength, length-stringsLength);
- }
- /* return the length including padding bytes */
- if(U_SUCCESS(*pErrorCode)) {
- return length;
- } else {
- return 0;
- }
- }
- U_CAPI void U_EXPORT2
- udata_printError(const UDataSwapper *ds,
- const char *fmt,
- ...) {
- va_list args;
- if(ds->printError!=nullptr) {
- va_start(args, fmt);
- ds->printError(ds->printErrorContext, fmt, args);
- va_end(args);
- }
- }
- /* swap a data header ------------------------------------------------------- */
- U_CAPI int32_t U_EXPORT2
- udata_swapDataHeader(const UDataSwapper *ds,
- const void *inData, int32_t length, void *outData,
- UErrorCode *pErrorCode) {
- const DataHeader *pHeader;
- uint16_t headerSize, infoSize;
- /* argument checking */
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return 0;
- }
- if(ds==nullptr || inData==nullptr || length<-1 || (length>0 && outData==nullptr)) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return 0;
- }
- /* check minimum length and magic bytes */
- pHeader=(const DataHeader *)inData;
- if( (length>=0 && length<(int32_t)sizeof(DataHeader)) ||
- pHeader->dataHeader.magic1!=0xda ||
- pHeader->dataHeader.magic2!=0x27 ||
- pHeader->info.sizeofUChar!=2
- ) {
- udata_printError(ds, "udata_swapDataHeader(): initial bytes do not look like ICU data\n");
- *pErrorCode=U_UNSUPPORTED_ERROR;
- return 0;
- }
- headerSize=ds->readUInt16(pHeader->dataHeader.headerSize);
- infoSize=ds->readUInt16(pHeader->info.size);
- if( headerSize<sizeof(DataHeader) ||
- infoSize<sizeof(UDataInfo) ||
- headerSize<(sizeof(pHeader->dataHeader)+infoSize) ||
- (length>=0 && length<headerSize)
- ) {
- udata_printError(ds, "udata_swapDataHeader(): header size mismatch - headerSize %d infoSize %d length %d\n",
- headerSize, infoSize, length);
- *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
- return 0;
- }
- if(length>0) {
- DataHeader *outHeader;
- const char *s;
- int32_t maxLength;
- /* Most of the fields are just bytes and need no swapping. */
- if(inData!=outData) {
- uprv_memcpy(outData, inData, headerSize);
- }
- outHeader=(DataHeader *)outData;
- outHeader->info.isBigEndian = ds->outIsBigEndian;
- outHeader->info.charsetFamily = ds->outCharset;
- /* swap headerSize */
- ds->swapArray16(ds, &pHeader->dataHeader.headerSize, 2, &outHeader->dataHeader.headerSize, pErrorCode);
- /* swap UDataInfo size and reservedWord */
- ds->swapArray16(ds, &pHeader->info.size, 4, &outHeader->info.size, pErrorCode);
- /* swap copyright statement after the UDataInfo */
- infoSize+=sizeof(pHeader->dataHeader);
- s=(const char *)inData+infoSize;
- maxLength=headerSize-infoSize;
- /* get the length of the string */
- for(length=0; length<maxLength && s[length]!=0; ++length) {}
- /* swap the string contents */
- ds->swapInvChars(ds, s, length, (char *)outData+infoSize, pErrorCode);
- }
- return headerSize;
- }
- /* API functions ------------------------------------------------------------ */
- U_CAPI UDataSwapper * U_EXPORT2
- udata_openSwapper(UBool inIsBigEndian, uint8_t inCharset,
- UBool outIsBigEndian, uint8_t outCharset,
- UErrorCode *pErrorCode) {
- UDataSwapper *swapper;
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return nullptr;
- }
- if(inCharset>U_EBCDIC_FAMILY || outCharset>U_EBCDIC_FAMILY) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return nullptr;
- }
- /* allocate the swapper */
- swapper=(UDataSwapper *)uprv_malloc(sizeof(UDataSwapper));
- if(swapper==nullptr) {
- *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
- return nullptr;
- }
- uprv_memset(swapper, 0, sizeof(UDataSwapper));
- /* set values and functions pointers according to in/out parameters */
- swapper->inIsBigEndian=inIsBigEndian;
- swapper->inCharset=inCharset;
- swapper->outIsBigEndian=outIsBigEndian;
- swapper->outCharset=outCharset;
- swapper->readUInt16= inIsBigEndian==U_IS_BIG_ENDIAN ? uprv_readDirectUInt16 : uprv_readSwapUInt16;
- swapper->readUInt32= inIsBigEndian==U_IS_BIG_ENDIAN ? uprv_readDirectUInt32 : uprv_readSwapUInt32;
- swapper->writeUInt16= outIsBigEndian==U_IS_BIG_ENDIAN ? uprv_writeDirectUInt16 : uprv_writeSwapUInt16;
- swapper->writeUInt32= outIsBigEndian==U_IS_BIG_ENDIAN ? uprv_writeDirectUInt32 : uprv_writeSwapUInt32;
- swapper->compareInvChars= outCharset==U_ASCII_FAMILY ? uprv_compareInvAscii : uprv_compareInvEbcdic;
- if(inIsBigEndian==outIsBigEndian) {
- swapper->swapArray16=uprv_copyArray16;
- swapper->swapArray32=uprv_copyArray32;
- swapper->swapArray64=uprv_copyArray64;
- } else {
- swapper->swapArray16=uprv_swapArray16;
- swapper->swapArray32=uprv_swapArray32;
- swapper->swapArray64=uprv_swapArray64;
- }
- if(inCharset==U_ASCII_FAMILY) {
- swapper->swapInvChars= outCharset==U_ASCII_FAMILY ? uprv_copyAscii : uprv_ebcdicFromAscii;
- } else /* U_EBCDIC_FAMILY */ {
- swapper->swapInvChars= outCharset==U_EBCDIC_FAMILY ? uprv_copyEbcdic : uprv_asciiFromEbcdic;
- }
- return swapper;
- }
- U_CAPI UDataSwapper * U_EXPORT2
- udata_openSwapperForInputData(const void *data, int32_t length,
- UBool outIsBigEndian, uint8_t outCharset,
- UErrorCode *pErrorCode) {
- const DataHeader *pHeader;
- uint16_t headerSize, infoSize;
- UBool inIsBigEndian;
- int8_t inCharset;
- if(pErrorCode==nullptr || U_FAILURE(*pErrorCode)) {
- return nullptr;
- }
- if( data==nullptr ||
- (length>=0 && length<(int32_t)sizeof(DataHeader)) ||
- outCharset>U_EBCDIC_FAMILY
- ) {
- *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
- return nullptr;
- }
- pHeader=(const DataHeader *)data;
- if( (length>=0 && length<(int32_t)sizeof(DataHeader)) ||
- pHeader->dataHeader.magic1!=0xda ||
- pHeader->dataHeader.magic2!=0x27 ||
- pHeader->info.sizeofUChar!=2
- ) {
- *pErrorCode=U_UNSUPPORTED_ERROR;
- return 0;
- }
- inIsBigEndian=(UBool)pHeader->info.isBigEndian;
- inCharset=pHeader->info.charsetFamily;
- if(inIsBigEndian==U_IS_BIG_ENDIAN) {
- headerSize=pHeader->dataHeader.headerSize;
- infoSize=pHeader->info.size;
- } else {
- headerSize=uprv_readSwapUInt16(pHeader->dataHeader.headerSize);
- infoSize=uprv_readSwapUInt16(pHeader->info.size);
- }
- if( headerSize<sizeof(DataHeader) ||
- infoSize<sizeof(UDataInfo) ||
- headerSize<(sizeof(pHeader->dataHeader)+infoSize) ||
- (length>=0 && length<headerSize)
- ) {
- *pErrorCode=U_UNSUPPORTED_ERROR;
- return 0;
- }
- return udata_openSwapper(inIsBigEndian, inCharset, outIsBigEndian, outCharset, pErrorCode);
- }
- U_CAPI void U_EXPORT2
- udata_closeSwapper(UDataSwapper *ds) {
- uprv_free(ds);
- }
|