|
|
| version 1.1, 2001/12/15 21:28:21 | version 1.16, 2002/01/21 16:44:48 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: Charset connection implementation. | Parser: Charset connection implementation. |
| Copyright (c) 2001 ArtLebedev Group (http://www.artlebedev.com) | Copyright(c) 2001 ArtLebedev Group(http://www.artlebedev.com) |
| Author: Alexander Petrosyan <paf@design.ru> (http://paf.design.ru) | Author: Alexander Petrosyan<paf@design.ru>(http://paf.design.ru) |
| $Id$ | $Id$ |
| */ | */ |
| #include "pa_charset.h" | #include "pa_charset.h" |
| //#include "pa_exception.h" | |
| //#include "pa_common.h" | |
| //#include "pa_threads.h" | |
| #ifdef XML | #ifdef XML |
| # include <util/TransENameMap.hpp> | #include "libxml/encoding.h" |
| # include <util/XML256TableTranscoder.hpp> | |
| # include <util/PlatformUtils.hpp> | |
| # include <PlatformSupport/XalanTranscodingServices.hpp> | |
| #endif | #endif |
| // globals | // globals |
| Line 47 inline unsigned int to_wchar_code(const | Line 41 inline unsigned int to_wchar_code(const |
| if(!cstr || !*cstr) | if(!cstr || !*cstr) |
| return 0; | return 0; |
| if(cstr[1]==0) | if(cstr[1]==0) |
| return (unsigned int)(unsigned char)cstr[0]; | return(unsigned int)(unsigned char)cstr[0]; |
| char *error_pos; | char *error_pos; |
| return (unsigned int)strtol(cstr, &error_pos, 0); | return(unsigned int)strtol(cstr, &error_pos, 0); |
| } | } |
| inline bool to_bool(const char *cstr) { | inline bool to_bool(const char *cstr) { |
| return cstr && *cstr!=0; | return cstr && *cstr!=0; |
| Line 64 static void element2ctypes(unsigned char | Line 58 static void element2ctypes(unsigned char |
| ctypes_table[c]|=bit; | ctypes_table[c]|=bit; |
| if(group_offset>=0) | if(group_offset>=0) |
| tables[cbits_offset+group_offset+c/8] |= 1 << (c%8); | tables[cbits_offset+group_offset+c/8] |= 1<<(c%8); |
| } | } |
| static void element2case(unsigned char from, unsigned char to, | static void element2case(unsigned char from, unsigned char to, |
| unsigned char *tables) { | unsigned char *tables) { |
| Line 77 static void element2case(unsigned char f | Line 71 static void element2case(unsigned char f |
| fcc_table[from]=to; fcc_table[to]=from; | fcc_table[from]=to; fcc_table[to]=from; |
| } | } |
| #ifdef XML | |
| template <class TType> class ENameMapFor2 : public ENameMap { | |
| public : | |
| // ----------------------------------------------------------------------- | |
| // Constructors and Destructor | |
| // ----------------------------------------------------------------------- | |
| ENameMapFor2( | |
| const XMLCh* const encodingName | |
| , const XMLCh* const fromTable | |
| , const XMLTransService::TransRec* const toTable | |
| , const unsigned int toTableSize | |
| ) : ENameMap(encodingName), | |
| ffromTable(fromTable), | |
| ftoTable(toTable), | |
| ftoTableSize(toTableSize) {} | |
| // ----------------------------------------------------------------------- | |
| // Implementation of virtual factory method | |
| // ----------------------------------------------------------------------- | |
| virtual XMLTranscoder* makeNew(const unsigned int blockSize) const { | |
| return new TType( | |
| getKey(), | |
| blockSize, | |
| ffromTable, | |
| ftoTable, ftoTableSize); | |
| } | |
| private: | |
| const XMLCh* const ffromTable; | |
| const XMLTransService::TransRec* const ftoTable; | |
| const unsigned int ftoTableSize; | |
| private : | |
| // ----------------------------------------------------------------------- | |
| // Unimplemented constructors and operators | |
| // ----------------------------------------------------------------------- | |
| ENameMapFor2(); | |
| ENameMapFor2(const ENameMapFor2<TType>&); | |
| void operator=(const ENameMapFor2<TType>&); | |
| }; | |
| class XML256TableTranscoder2 : public XML256TableTranscoder { | |
| public : | |
| XML256TableTranscoder2( | |
| const XMLCh* const encodingName | |
| , const unsigned int blockSize | |
| , const XMLCh* const fromTable | |
| , const XMLTransService::TransRec* const toTable | |
| , const unsigned int toTableSize | |
| ) : XML256TableTranscoder(encodingName, blockSize, fromTable, toTable, toTableSize) {} | |
| private : | |
| XML256TableTranscoder2(); | |
| XML256TableTranscoder2(const XML256TableTranscoder2&); | |
| void operator=(const XML256TableTranscoder2&); | |
| }; | |
| #endif | |
| // methods | // methods |
| extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c | extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c |
| Charset::Charset(Pool& apool, const String& aname, const String *file_spec) : Pooled(apool), | Charset::Charset(Pool& apool, const String& aname, const String *request_file_spec): Pooled(apool), |
| fname(apool) { | fname(aname) { |
| // fname | |
| char *name_cstr=(char *)malloc(aname.size()); | char *name_cstr=fname.cstr(); |
| memcpy(name_cstr, aname.cstr(String::UL_AS_IS), aname.size()); | for(char *c=name_cstr; *c; c++) |
| fname.APPEND_CLEAN(name_cstr, aname.size(), 0, 0); | *c = toupper(*c); |
| if(file_spec) { | if(request_file_spec) { |
| fisUTF8=false; | fisUTF8=false; |
| loadDefinition(*file_spec); | loadDefinition(*request_file_spec); |
| #ifdef XML | #ifdef XML |
| addEncoding(name_cstr); | addEncoding(name_cstr); |
| #endif | #endif |
| } else { | } else { |
| fisUTF8=true; | fisUTF8=true; |
| // grab default onces [for UTF-8 so to be able to make a-z => A-Z | // grab default onces [for UTF-8 so to be able to make a-z =>A-Z |
| memcpy(pcre_tables, pcre_default_tables, sizeof(pcre_tables)); | memcpy(pcre_tables, pcre_default_tables, sizeof(pcre_tables)); |
| } | } |
| Line 163 Charset::Charset(Pool& apool, const Stri | Line 100 Charset::Charset(Pool& apool, const Stri |
| Charset::~Charset() { | Charset::~Charset() { |
| #ifdef XML | #ifdef XML |
| delete transcoder; | // not deleting transcoder, that's not our business |
| #endif | #endif |
| } | } |
| void Charset::loadDefinition(const String& file_spec) { | void Charset::loadDefinition(const String& request_file_spec) { |
| // pcre_tables | // pcre_tables |
| // lowcase, flipcase, bits digit+word+whitespace, masks | // lowcase, flipcase, bits digit+word+whitespace, masks |
| Line 175 void Charset::loadDefinition(const Strin | Line 112 void Charset::loadDefinition(const Strin |
| // don't know the size there | // don't know the size there |
| memset(pcre_tables, 0, sizeof(pcre_tables)); | memset(pcre_tables, 0, sizeof(pcre_tables)); |
| prepare_case_tables(pcre_tables); | prepare_case_tables(pcre_tables); |
| cstr2ctypes(pcre_tables, (const unsigned char *)"*+?{^.$|()[", ctype_meta); | cstr2ctypes(pcre_tables,(const unsigned char *)"*+?{^.$|()[", ctype_meta); |
| // charset | // charset |
| memset(fromTable, 0, sizeof(fromTable)); | memset(tables.fromTable, 0, sizeof(tables.fromTable)); |
| toTable=(XMLTransService::TransRec *)calloc( | tables.toTable=(Charset_TransRec *)calloc(sizeof(Charset_TransRec)*MAX_CHARSET_UNI_CODES); |
| sizeof(XMLTransService::TransRec)*MAX_CHARSET_UNI_CODES); | tables.toTableSize=0; |
| toTableSize=0; | |
| // strangly vital | // strangly vital |
| toTable[toTableSize].intCh=0; | tables.toTable[tables.toTableSize].intCh=0; |
| toTable[toTableSize].extCh=(XMLByte)0; | tables.toTable[tables.toTableSize].extCh=(XMLByte)0; |
| toTableSize++; | tables.toTableSize++; |
| // loading text | // loading text |
| char *data=file_read_text(pool(), file_spec); | char *data=file_read_text(pool(), request_file_spec); |
| // ignore header | // ignore header |
| getrow(&data); | getrow(&data); |
| Line 216 void Charset::loadDefinition(const Strin | Line 152 void Charset::loadDefinition(const Strin |
| case 7: | case 7: |
| case 8: | case 8: |
| // charset | // charset |
| if(toTableSize>MAX_CHARSET_UNI_CODES) | if(tables.toTableSize>MAX_CHARSET_UNI_CODES) |
| throw Exception(0, 0, | throw Exception(0, 0, |
| &file_spec, | &request_file_spec, |
| "charset must contain not more then %d unicode values", MAX_CHARSET_UNI_CODES); | "charset must contain not more then %d unicode values", MAX_CHARSET_UNI_CODES); |
| XMLCh unicode=(XMLCh)to_wchar_code(cell); | XMLCh unicode=(XMLCh)to_wchar_code(cell); |
| if(!unicode && column==7/*unicode1 column*/) | if(!unicode && column==7/*unicode1 column*/) |
| unicode=(XMLCh)c; | unicode=(XMLCh)c; |
| if(unicode) { | if(unicode) { |
| if(!fromTable[c]) | if(!tables.fromTable[c]) |
| fromTable[c]=unicode; | tables.fromTable[c]=unicode; |
| toTable[toTableSize].intCh=unicode; | tables.toTable[tables.toTableSize].intCh=unicode; |
| toTable[toTableSize].extCh=(XMLByte)c; | tables.toTable[tables.toTableSize].extCh=(XMLByte)c; |
| toTableSize++; | tables.toTableSize++; |
| } | } |
| break; | break; |
| } | } |
| Line 240 void Charset::loadDefinition(const Strin | Line 176 void Charset::loadDefinition(const Strin |
| sort_ToTable(); | sort_ToTable(); |
| } | } |
| #ifdef XML | |
| void Charset::addEncoding(const char *name_cstr) { | |
| // addEncoding | |
| XalanDOMString sencoding(name_cstr); | |
| const XMLCh* const auto_encoding_cstr=sencoding.c_str(); | |
| int size=sizeof(XMLCh)*(sencoding.size()+1); | |
| XMLCh* pool_encoding_cstr=(XMLCh*)malloc(size); | |
| memcpy(pool_encoding_cstr, auto_encoding_cstr, size); | |
| XMLString::upperCase(pool_encoding_cstr); | |
| XMLPlatformUtils::fgTransService->addEncoding( | |
| pool_encoding_cstr, | |
| new ENameMapFor2<XML256TableTranscoder2>( | |
| pool_encoding_cstr | |
| , fromTable | |
| , toTable | |
| , toTableSize | |
| )); | |
| } | |
| void Charset::initTranscoder(const String *source, const char *name_cstr) { | |
| XMLTransService::Codes resValue; | |
| transcoder=XMLPlatformUtils::fgTransService->makeNewTranscoderFor(name_cstr, resValue, 60); | |
| if(!transcoder) | |
| throw Exception(0, 0, | |
| source, | |
| "unsupported encoding"); | |
| } | |
| #endif | |
| static int sort_cmp_Trans_rec_intCh(const void *a, const void *b) { | static int sort_cmp_Trans_rec_intCh(const void *a, const void *b) { |
| return | return |
| static_cast<const Charset_TransRec *>(a)->intCh- | static_cast<const Charset_TransRec *>(a)->intCh- |
| Line 277 static int sort_cmp_Trans_rec_intCh(cons | Line 183 static int sort_cmp_Trans_rec_intCh(cons |
| } | } |
| void Charset::sort_ToTable() { | void Charset::sort_ToTable() { |
| _qsort(toTable, toTableSize, sizeof(*toTable), | _qsort(tables.toTable, tables.toTableSize, sizeof(*tables.toTable), |
| sort_cmp_Trans_rec_intCh); | sort_cmp_Trans_rec_intCh); |
| //FILE *f=fopen("c:\\temp\\a", "wb"); | //FILE *f=fopen("c:\\temp\\a", "wb"); |
| //fwrite(toTable, toTableSize, sizeof(*toTable), f); | //fwrite(tables.toTable, tables.toTableSize, sizeof(*tables.toTable), f); |
| //fclose(f); | //fclose(f); |
| } | } |
| XMLByte Charset::xlatOneTo(const XMLCh toXlat) const { | static XMLByte xlatOneTo(const XMLCh toXlat, |
| const Charset::Tables& tables) { | |
| unsigned int lowOfs = 0; | unsigned int lowOfs = 0; |
| unsigned int hiOfs = toTableSize - 1; | unsigned int hiOfs = tables.toTableSize - 1; |
| XMLByte curByte = 0; | XMLByte curByte = 0; |
| do { | do { |
| // Calc the mid point of the low and high offset. | // Calc the mid point of the low and high offset. |
| const unsigned int midOfs = ((hiOfs - lowOfs) / 2) + lowOfs; | const unsigned int midOfs =((hiOfs - lowOfs) / 2)+lowOfs; |
| // If our test char is greater than the mid point char, then | // If our test char is greater than the mid point char, then |
| // we move up to the upper half. Else we move to the lower | // we move up to the upper half. Else we move to the lower |
| // half. If its equal, then its our guy. | // half. If its equal, then its our guy. |
| if (toXlat > toTable[midOfs].intCh) | if(toXlat>tables.toTable[midOfs].intCh) |
| lowOfs = midOfs; | lowOfs = midOfs; |
| else if (toXlat < toTable[midOfs].intCh) | else if(toXlat<tables.toTable[midOfs].intCh) |
| hiOfs = midOfs; | hiOfs = midOfs; |
| else | else |
| return toTable[midOfs].extCh; | return tables.toTable[midOfs].extCh; |
| } while(lowOfs + 1 < hiOfs); | } while(lowOfs+1<hiOfs); |
| return '?'; | return '?'; |
| } | } |
| Line 310 void Charset::transcode(Pool& pool, | Line 217 void Charset::transcode(Pool& pool, |
| const Charset& source_charset, const void *source_body, size_t source_content_length, | const Charset& source_charset, const void *source_body, size_t source_content_length, |
| const Charset& dest_charset, const void *& dest_body, size_t& dest_content_length | const Charset& dest_charset, const void *& dest_body, size_t& dest_content_length |
| ) { | ) { |
| if(!source_content_length) { | |
| dest_body=0; | |
| dest_content_length=0; | |
| return; | |
| } | |
| switch((source_charset.isUTF8()?0x10:0x00)|(dest_charset.isUTF8()?0x01:0x00)) { | switch((source_charset.isUTF8()?0x10:0x00)|(dest_charset.isUTF8()?0x01:0x00)) { |
| default: // 0x00 | default: // 0x00 |
| source_charset.transcodeToCharset(pool, dest_charset, | source_charset.transcodeToCharset(pool, dest_charset, |
| Line 374 static const XMLByte gFirstByteMark[7] = | Line 287 static const XMLByte gFirstByteMark[7] = |
| 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC | 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC |
| }; | }; |
| /// @todo not so memory-hungry with prescan | static int transcodeToUTF8( |
| void Charset::transcodeToUTF8(Pool& pool, | const XMLByte* srcData, size_t& srcLen, |
| const void *source_body, size_t source_content_length, | XMLByte *toFill, size_t& toFillLen, |
| const void *& adest_body, size_t& adest_content_length) const { | const Charset::Tables& tables) { |
| const XMLByte* srcPtr=srcData; | |
| size_t dest_content_length=0; | const XMLByte* srcEnd=srcData+srcLen; |
| XMLByte *dest_body=(XMLByte*)pool.malloc(source_content_length*6/*so that surly enough*/); | XMLByte* outPtr=toFill; |
| XMLByte* outEnd=toFill+toFillLen; | |
| const XMLByte* srcPtr=(const XMLByte*)source_body; | |
| const XMLByte* srcEnd=(const XMLByte*)source_body+source_content_length; | |
| XMLByte* outPtr=dest_body; | |
| while (srcPtr < srcEnd) { | while(srcPtr<srcEnd) { |
| uint curVal = fromTable[*srcPtr]; | uint curVal = tables.fromTable[*srcPtr]; |
| if(!curVal) { | if(!curVal) { |
| // use the replacement character | // use the replacement character |
| *outPtr++ = '?'; | *outPtr++= '?'; |
| srcPtr ++; | srcPtr++; |
| continue; | continue; |
| } | } |
| // Figure out how many bytes we need | // Figure out how many bytes we need |
| unsigned int encodedBytes; | unsigned int encodedBytes; |
| if (curVal < 0x80) | if(curVal<0x80) |
| encodedBytes = 1; | encodedBytes = 1; |
| else if (curVal < 0x800) | else if(curVal<0x800) |
| encodedBytes = 2; | encodedBytes = 2; |
| else if (curVal < 0x10000) | else if(curVal<0x10000) |
| encodedBytes = 3; | encodedBytes = 3; |
| else if (curVal < 0x200000) | else if(curVal<0x200000) |
| encodedBytes = 4; | encodedBytes = 4; |
| else if (curVal < 0x4000000) | else if(curVal<0x4000000) |
| encodedBytes = 5; | encodedBytes = 5; |
| else if (curVal <= 0x7FFFFFFF) | else if(curVal<= 0x7FFFFFFF) |
| encodedBytes = 6; | encodedBytes = 6; |
| else { | else { |
| // use the replacement character | // use the replacement character |
| *outPtr++ = '?'; | *outPtr++= '?'; |
| srcPtr ++; | srcPtr++; |
| continue; | continue; |
| } | } |
| // If we cannot fully get this char into the output buffer, | // If we cannot fully get this char into the output buffer |
| // never | if (outPtr + encodedBytes > outEnd) |
| break; | |
| // We can do it, so update the source index | // We can do it, so update the source index |
| srcPtr++; | srcPtr++; |
| // And spit out the bytes. We spit them out in reverse order | // And spit out the bytes. We spit them out in reverse order |
| // here, so bump up the output pointer and work down as we go. | // here, so bump up the output pointer and work down as we go. |
| outPtr += encodedBytes; | outPtr+= encodedBytes; |
| switch(encodedBytes) { | switch(encodedBytes) { |
| case 6 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | case 6: *--outPtr = (XMLByte)((curVal | 0x80UL) & 0xBFUL); |
| curVal >>= 6; | curVal>>= 6; |
| case 5 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | case 5: *--outPtr = (XMLByte)((curVal | 0x80UL) & 0xBFUL); |
| curVal >>= 6; | curVal>>= 6; |
| case 4 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | case 4: *--outPtr = (XMLByte)((curVal | 0x80UL) & 0xBFUL); |
| curVal >>= 6; | curVal>>= 6; |
| case 3 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | case 3: *--outPtr = (XMLByte)((curVal | 0x80UL) & 0xBFUL); |
| curVal >>= 6; | curVal>>= 6; |
| case 2 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | case 2: *--outPtr = (XMLByte)((curVal | 0x80UL) & 0xBFUL); |
| curVal >>= 6; | curVal>>= 6; |
| case 1 : *--outPtr = XMLByte(curVal | gFirstByteMark[encodedBytes]); | case 1: *--outPtr = (XMLByte)(curVal | gFirstByteMark[encodedBytes]); |
| } | } |
| // Add the encoded bytes back in again to indicate we've eaten them | // Add the encoded bytes back in again to indicate we've eaten them |
| outPtr += encodedBytes; | outPtr+= encodedBytes; |
| } | } |
| // return | // Update the bytes eaten |
| adest_body=dest_body; | srcLen = srcPtr - srcData; |
| adest_content_length=outPtr-dest_body; | |
| // Return the characters read | |
| toFillLen = outPtr - toFill; | |
| return srcPtr==srcEnd?(int)toFillLen:-1; | |
| } | } |
| void Charset::transcodeFromUTF8(Pool& pool, | static size_t transcodeFromUTF8( |
| const void *source_body, size_t source_content_length, | const XMLByte *srcData, size_t& srcLen, |
| const void *& adest_body, size_t& adest_content_length) const { | XMLByte* toFill, size_t& toFillLen, |
| size_t dest_content_length=0; | const Charset::Tables& tables) { |
| XMLByte *dest_body=(XMLByte*)pool.malloc(source_content_length/*surly enough*/); | const XMLByte* srcPtr=srcData; |
| const XMLByte* srcEnd=srcData+srcLen; | |
| const XMLByte* srcPtr=(const XMLByte*)source_body; | XMLByte* outPtr=toFill; |
| const XMLByte* srcEnd=(const XMLByte*)source_body+source_content_length; | XMLByte* outEnd=toFill+toFillLen; |
| XMLByte* outPtr=dest_body; | |
| // We now loop until we either run out of input data | // We now loop until we either run out of input data, or room to store |
| while (srcPtr < srcEnd) { | while ((srcPtr < srcEnd) && (outPtr < outEnd)) { |
| // Get the next leading byte out | // Get the next leading byte out |
| const XMLByte firstByte = *srcPtr; | const XMLByte firstByte = *srcPtr; |
| // Special-case ASCII, which is a leading byte value of <= 127 | // Special-case ASCII, which is a leading byte value of<= 127 |
| if (firstByte <= 127) { | if(firstByte<= 127) { |
| *outPtr++ = firstByte; | *outPtr++= firstByte; |
| srcPtr++; | srcPtr++; |
| continue; | continue; |
| } | } |
| Line 473 void Charset::transcodeFromUTF8(Pool& po | Line 387 void Charset::transcodeFromUTF8(Pool& po |
| const unsigned int trailingBytes = gUTFBytes[firstByte]; | const unsigned int trailingBytes = gUTFBytes[firstByte]; |
| // If there are not enough source bytes to do this one, then we | // If there are not enough source bytes to do this one, then we |
| // are done. Note that we done >= here because we are implicitly | // are done. Note that we done>= here because we are implicitly |
| // counting the 1 byte we get no matter what. | // counting the 1 byte we get no matter what. |
| if (srcPtr + trailingBytes >= srcEnd) | if(srcPtr+trailingBytes>= srcEnd) |
| break; | break; |
| // Looks ok, so lets build up the value | // Looks ok, so lets build up the value |
| Line 492 void Charset::transcodeFromUTF8(Pool& po | Line 406 void Charset::transcodeFromUTF8(Pool& po |
| default: | default: |
| throw Exception(0, 0, | throw Exception(0, 0, |
| 0, | 0, |
| "transcodeFromUTF8 error: wrong trailingBytes value (%d)", trailingBytes); | "transcodeFromUTF8 error: wrong trailingBytes value(%d)", trailingBytes); |
| } | } |
| tmpVal-=gUTFOffsets[trailingBytes]; | tmpVal-=gUTFOffsets[trailingBytes]; |
| // If it will fit into a single char, then put it in. Otherwise | // If it will fit into a single char, then put it in. Otherwise |
| // fail [*encode it as a surrogate pair. If its not valid, use the | // fail [*encode it as a surrogate pair. If its not valid, use the |
| // replacement char.*] | // replacement char.*] |
| if (!(tmpVal & 0xFFFF0000)) | if(!(tmpVal & 0xFFFF0000)) |
| *outPtr++ = xlatOneTo(tmpVal); | *outPtr++= xlatOneTo(tmpVal, tables); |
| else | else |
| throw Exception(0, 0, | throw Exception(0, 0, |
| 0, | 0, |
| "transcodeFromUTF8 error: too big tmpVal (0x%08X)", tmpVal); | "transcodeFromUTF8 error: too big tmpVal(0x%08X)", tmpVal); |
| } | } |
| // Update the bytes eaten | |
| srcLen = srcPtr - srcData; | |
| // Return the characters read | |
| toFillLen = outPtr - toFill; | |
| return srcPtr==srcEnd?(int)toFillLen:-1; | |
| } | |
| /// @todo not so memory-hungry with prescan | |
| void Charset::transcodeToUTF8(Pool& pool, | |
| const void *source_body, size_t source_content_length, | |
| const void *& adest_body, size_t& dest_content_length) const { | |
| dest_content_length=source_content_length*6/*so that surly enough*/; | |
| XMLByte *dest_body=(XMLByte*)pool.malloc(dest_content_length); | |
| if(::transcodeToUTF8( | |
| (XMLByte *)source_body, source_content_length, | |
| dest_body, dest_content_length, | |
| tables)<0) | |
| throw(0, 0, | |
| 0, | |
| "Charset::transcodeToUTF8 buffer overflow"); | |
| // return | |
| adest_body=dest_body; | |
| } | |
| /// @test buf overflow | |
| void Charset::transcodeFromUTF8(Pool& pool, | |
| const void *source_body, size_t source_content_length, | |
| const void *& adest_body, size_t& dest_content_length) const { | |
| dest_content_length=source_content_length/*surly enough*/; | |
| XMLByte *dest_body=(XMLByte*)pool.malloc(dest_content_length); | |
| if(::transcodeFromUTF8( | |
| (XMLByte *)source_body, source_content_length, | |
| dest_body, dest_content_length, | |
| tables)<0) | |
| throw(0, 0, | |
| 0, | |
| "Charset::transcodeToUTF8 buffer overflow"); | |
| // return | // return |
| adest_body=dest_body; | adest_body=dest_body; |
| adest_content_length=outPtr-dest_body; | |
| } | } |
| /// transcode using both charsets | /// transcode using both charsets |
| void Charset::transcodeToCharset(Pool& pool, | void Charset::transcodeToCharset(Pool& pool, |
| const Charset& dest_charset, | const Charset& dest_charset, |
| const void *source_body, size_t source_content_length, | const void *source_body, size_t source_content_length, |
| const void *& adest_body, size_t& dest_content_length) const { | const void *& adest_body, size_t& adest_content_length) const { |
| throw Exception(0, 0, | if(&dest_charset==this) { |
| 0, | adest_body=source_body; |
| "transcodeToCharset not supported(yet)"); | adest_content_length=source_content_length; |
| /* | } else { |
| void *dest_body; | size_t dest_content_length=source_content_length; |
| unsigned char *dest_body=(unsigned char *)pool.malloc(dest_content_length); | |
| dest_body=pool.malloc(dest_content_length=source_content_length); | |
| // dummy | |
| memset(dest_body, '?', dest_content_length); | |
| adest_body=dest_body;*/ | const XMLByte* srcPtr=(XMLByte *)source_body; |
| } | const XMLByte* srcEnd=(XMLByte *)source_body+source_content_length; |
| #ifdef XML | for(XMLByte* outPtr=dest_body; srcPtr<srcEnd; srcPtr++) { |
| const char *Charset::transcode_cstr(const XalanDOMString& s) { | XMLCh curVal = tables.fromTable[*srcPtr]; |
| const unsigned int len=s.size()*2; | if(curVal) |
| XMLByte* dest=(XMLByte *)malloc((len+1)*sizeof(XMLByte)); | *outPtr++=xlatOneTo(curVal, dest_charset.tables); |
| bool error=true; | else { |
| try { | // use the replacement character |
| if(transcoder) { | *outPtr++= '?'; |
| unsigned int charsEaten; | } |
| unsigned int size=transcoder->transcodeTo( | |
| s.c_str(), s.length(), | |
| dest, len, | |
| charsEaten, | |
| XMLTranscoder::UnRep_RepChar //UnRep_Throw | |
| ); | |
| dest[size]=0; | |
| error=false; | |
| } | } |
| } catch(XMLException& e) { | |
| Exception::provide_source(pool(), 0, e); | adest_body=dest_body; |
| adest_content_length=dest_content_length; | |
| } | } |
| return (const char *)dest; | } |
| #ifdef XML | |
| static int xml256CharEncodingInputFunc ( | |
| unsigned char *out, | |
| int *outlen, | |
| const unsigned char *in, | |
| int *inlen, | |
| void *info) { | |
| return transcodeToUTF8( | |
| in, *(unsigned int*)inlen, | |
| out, *(unsigned int*)outlen, | |
| *(const Charset::Tables *)info); | |
| } | |
| static int xml256CharEncodingOutputFunc ( | |
| unsigned char *out, | |
| int *outlen, | |
| const unsigned char *in, | |
| int *inlen, | |
| void *info) { | |
| return transcodeFromUTF8( | |
| in, *(unsigned int*)inlen, | |
| out, *(unsigned int*)outlen, | |
| *(const Charset::Tables *)info); | |
| } | } |
| String& Charset::transcode(const XalanDOMString& s) { | |
| return *NEW String(pool(), transcode_cstr(s)); | |
| void Charset::addEncoding(char *name_cstr) { | |
| xmlCharEncodingHandler *handler= | |
| (xmlCharEncodingHandler *)malloc(sizeof(xmlCharEncodingHandler)); | |
| handler->name=name_cstr; | |
| handler->input=xml256CharEncodingInputFunc; handler->inputInfo=&tables; | |
| handler->output=xml256CharEncodingOutputFunc; handler->outputInfo=&tables; | |
| xmlRegisterCharEncodingHandler(handler); | |
| } | } |
| std::auto_ptr<XalanDOMString> Charset::transcode_buf(const char *buf, size_t buf_size) { | void Charset::initTranscoder(const String *source, const char *name_cstr) { |
| unsigned int dest_size=0; | ftranscoder=xmlFindCharEncodingHandler(name_cstr); |
| XMLCh* dest=(XMLCh *)malloc((buf_size+1)*sizeof(XMLCh)); | transcoder(source); // check right way |
| unsigned char *charSizes=(unsigned char *)malloc(buf_size*sizeof(unsigned char)); | } |
| XalanDOMString *result; | |
| try { | xmlCharEncodingHandler *Charset::transcoder(const String *source) { |
| if(transcoder) { | if(!ftranscoder) |
| unsigned int bytesEaten; | throw Exception(0, 0, |
| unsigned int dest_size=transcoder->transcodeFrom( | source, |
| (unsigned char *)buf, | "unsupported encoding"); |
| (const unsigned int)buf_size, | return ftranscoder; |
| dest, (const unsigned int)buf_size, | } |
| bytesEaten, | |
| charSizes | const char *Charset::transcode_cstr(xmlChar *s) { |
| ); | if(!s) |
| result=new XalanDOMString(dest, dest_size); | return ""; |
| } | |
| } catch(XMLException& e) { | int inlen=strlen((const char *)s); |
| Exception::provide_source(pool(), 0, e); | int outlen=inlen+1; // max |
| result=0; //calm, compiler | char *out=(char *)malloc(outlen*sizeof(char)); |
| } | |
| return std::auto_ptr<XalanDOMString>(result); | int size=transcoder(0)->output( |
| (unsigned char*)out, &outlen, | |
| (const unsigned char*)s, &inlen, | |
| transcoder(0)->outputInfo); | |
| if(size<0) | |
| throw Exception(0, 0, | |
| 0, | |
| "transcode_cstr failed (%d)", size); | |
| out[size]=0; | |
| return out; | |
| } | |
| String& Charset::transcode(xmlChar *s) { | |
| return *NEW String(pool(), transcode_cstr(s)); | |
| } | |
| const char *Charset::transcode_cstr(GdomeDOMString *s) { | |
| return s?transcode_cstr(BAD_CAST s->str):""; | |
| } | |
| String& Charset::transcode(GdomeDOMString *s) { | |
| return *NEW String(pool(), transcode_cstr(s)); | |
| } | |
| /// @test less memory using -maybe- xmlParserInputBufferCreateMem | |
| GdomeDOMString_auto_ptr Charset::transcode_buf(const char *buf, size_t buf_size) { | |
| int outlen=buf_size*6/*max*/+1; | |
| unsigned char *out=(unsigned char*)malloc(outlen*sizeof(unsigned char)); | |
| int size=transcoder(0)->input( | |
| out, &outlen, | |
| (const unsigned char *)buf, (int *)&buf_size, | |
| transcoder(0)->inputInfo); | |
| if(size<0) | |
| throw Exception(0, 0, | |
| 0, | |
| "transcode_buf failed (%d)", size); | |
| out[size]=0; | |
| return GdomeDOMString_auto_ptr((gchar*)out); | |
| } | } |
| std::auto_ptr<XalanDOMString> Charset::transcode(const String& s) { | GdomeDOMString_auto_ptr Charset::transcode(const String& s) { |
| const char *cstr=s.cstr(String::UL_UNSPECIFIED); | const char *cstr=s.cstr(String::UL_UNSPECIFIED); |
| return transcode_buf(cstr, strlen(cstr)); | return transcode_buf(cstr, strlen(cstr)); |