|
|
| version 1.33.2.2, 2003/01/28 11:38:50 | version 1.33.2.19.2.23, 2003/04/10 09:14:58 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: Charset connection implementation. | Parser: Charset connection implementation. |
| Copyright(c) 2001, 2003 ArtLebedev Group (http://www.artlebedev.com) | Copyright(c) 2001-2003 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) |
| */ | */ |
| static const char* IDENT_CHARSET_C="$Date$"; | static const char* IDENT_CHARSET_C="$Date$"; |
| #include "pa_charset.h" | #include "pa_charset.h" |
| //#include "pa_globals.h" | #include "pa_charsets.h" |
| #ifdef XML | #ifdef XML |
| #include "libxml/encoding.h" | #include "libxml/encoding.h" |
| #endif | #endif |
| // globals | |
| #define CHARSET_UTF8_NAME "UTF-8" | |
| CharsetPtr UTF8_charset(new Charset(ConstStringPtr(new String(CHARSET_UTF8_NAME)), | |
| ConstStringPtrZero/*no file=system*/, | |
| 0/* no file=no file read pool*/)); | |
| // consts | |
| #define MAX_CHARSET_UNI_CODES 500 | |
| // helpers | // helpers |
| inline void prepare_case_tables(unsigned char *tables) { | inline void prepare_case_tables(unsigned char *tables) { |
| Line 43 inline void cstr2ctypes(unsigned char *t | Line 31 inline void cstr2ctypes(unsigned char *t |
| ctypes_table[c]|=bit; | ctypes_table[c]|=bit; |
| } | } |
| } | } |
| inline unsigned int to_wchar_code(const char *cstr) { | inline unsigned int to_wchar_code(const char* cstr) { |
| if(!cstr || !*cstr) | if(!cstr || !*cstr) |
| return 0; | return 0; |
| if(cstr[1]==0) | if(cstr[1]==0) |
| Line 52 inline unsigned int to_wchar_code(const | Line 40 inline unsigned int to_wchar_code(const |
| 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; |
| } | } |
| static void element2ctypes(unsigned char c, bool belongs, | static void element2ctypes(unsigned char c, bool belongs, |
| Line 80 static void element2case(unsigned char f | Line 68 static void element2case(unsigned char f |
| // methods | // methods |
| extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c | extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c |
| Charset::Charset(ConstStringPtr aname, ConstStringPtr afile_spec, Pool* apool_for_load): | Charset::Charset(Request_charsets* charsets, const StringBody ANAME, const String* afile_spec): |
| fname(aname), | FNAME(ANAME), |
| fname_cstr(aname->cstr()) { | FNAME_CSTR(ANAME.cstrm()) { |
| for(char *c=fname_cstr; *c; c++) | if(afile_spec) { |
| *c = toupper(*c); | |
| if(afile_spec && apool_for_load) { | |
| fisUTF8=false; | fisUTF8=false; |
| loadDefinition(*apool_for_load, fname); | load_definition(*charsets, *afile_spec); |
| #ifdef XML | #ifdef XML |
| addEncoding(fname_cstr); | addEncoding(FNAME_CSTR); |
| #endif | #endif |
| } else { | } else { |
| fisUTF8=true; | fisUTF8=true; |
| Line 100 Charset::Charset(ConstStringPtr aname, C | Line 85 Charset::Charset(ConstStringPtr aname, C |
| } | } |
| #ifdef XML | #ifdef XML |
| initTranscoder(fname, fname_cstr); | initTranscoder(FNAME, FNAME_CSTR); |
| #endif | |
| } | |
| Charset::~Charset() { | |
| // @todonow unregister encodings | |
| #ifdef XML | |
| // not deleting transcoder, that's not our business | |
| #endif | #endif |
| } | } |
| void Charset::loadDefinition(Pool& apool_for_load, ConstStringPtr afile_spec) { | void Charset::load_definition(Request_charsets& charsets, const String& afile_spec) { |
| // pcre_tables | // pcre_tables |
| // lowcase, flipcase, bits digit+word+whitespace, masks | // lowcase, flipcase, bits digit+word+whitespace, masks |
| Line 122 void Charset::loadDefinition(Pool& apool | Line 100 void Charset::loadDefinition(Pool& apool |
| cstr2ctypes(pcre_tables,(const unsigned char *)"*+?{^.$|()[", ctype_meta); | cstr2ctypes(pcre_tables,(const unsigned char *)"*+?{^.$|()[", ctype_meta); |
| // charset | // charset |
| memset(tables.fromTable, 0, sizeof(tables.fromTable)); | memset(&tables, 0, sizeof(tables)); |
| tables.toTable=(Charset_TransRec *)apool_for_load.calloc(sizeof(Charset_TransRec)*MAX_CHARSET_UNI_CODES); | |
| tables.toTableSize=0; | |
| // strangly vital | // strangly vital |
| tables.toTable[tables.toTableSize].intCh=0; | tables.toTable[tables.toTableSize].intCh=0; |
| tables.toTable[tables.toTableSize].extCh=(XMLByte)0; | tables.toTable[tables.toTableSize].extCh=(XMLByte)0; |
| tables.toTableSize++; | tables.toTableSize++; |
| // loading text | // loading text |
| char *data=file_read_text(apool_for_load, *UTF8_charset, afile_spec); | char *data=file_read_text(charsets, afile_spec); |
| // ignore header | // ignore header |
| getrow(&data); | getrow(&data); |
| Line 161 void Charset::loadDefinition(Pool& apool | Line 137 void Charset::loadDefinition(Pool& apool |
| // charset | // charset |
| if(tables.toTableSize>MAX_CHARSET_UNI_CODES) | if(tables.toTableSize>MAX_CHARSET_UNI_CODES) |
| throw Exception("parser.runtime", | throw Exception("parser.runtime", |
| afile_spec, | &afile_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); |
| Line 198 void Charset::sort_ToTable() { | Line 174 void Charset::sort_ToTable() { |
| } | } |
| static XMLByte xlatOneTo(const XMLCh toXlat, | static XMLByte xlatOneTo(const XMLCh toXlat, |
| const Charset::Tables& tables, | const Charset::Tables& tables, |
| XMLByte not_found) { | XMLByte not_found) { |
| unsigned int lowOfs = 0; | unsigned int lowOfs = 0; |
| unsigned int hiOfs = tables.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>tables.toTable[midOfs].intCh) | if(toXlat>tables.toTable[midOfs].intCh) |
| lowOfs = midOfs; | lowOfs = midOfs; |
| else if(toXlat<tables.toTable[midOfs].intCh) | else if(toXlat<tables.toTable[midOfs].intCh) |
| hiOfs = midOfs; | hiOfs = midOfs; |
| else | else |
| return tables.toTable[midOfs].extCh; | return tables.toTable[midOfs].extCh; |
| } while(lowOfs+1<hiOfs); | } while(lowOfs+1<hiOfs); |
| return not_found; | return not_found; |
| } | } |
| void Charset::transcode(Pool& pool, | String::C Charset::transcode(const String::C src, |
| const Charset& source_charset, const void *source_body, size_t source_content_length, | const Charset& source_charset, |
| const Charset& dest_charset, const void *& dest_body, size_t& dest_content_length | const Charset& dest_charset) { |
| ) { | if(!src.length) |
| if(!source_content_length) { | return String::C("", 0); |
| 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, | return source_charset.transcodeToCharset(src, dest_charset); |
| source_body, source_content_length, | |
| dest_body, dest_content_length); | |
| break; | |
| case 0x01: | case 0x01: |
| source_charset.transcodeToUTF8(pool, | return source_charset.transcodeToUTF8(src); |
| source_body, source_content_length, | |
| dest_body, dest_content_length); | |
| break; | |
| case 0x10: | case 0x10: |
| dest_charset.transcodeFromUTF8(pool, | return dest_charset.transcodeFromUTF8(src); |
| source_body, source_content_length, | |
| dest_body, dest_content_length); | |
| break; | |
| case 0x11: | case 0x11: |
| dest_body=source_body; | return src; |
| dest_content_length=source_content_length; | |
| break; | |
| } | } |
| } | } |
| Line 296 static const XMLByte gFirstByteMark[7] = | Line 257 static const XMLByte gFirstByteMark[7] = |
| }; | }; |
| static int transcodeToUTF8( | static int transcodeToUTF8( |
| const XMLByte* srcData, size_t& srcLen, | const XMLByte* srcData, size_t& srcLen, |
| XMLByte *toFill, size_t& toFillLen, | XMLByte *toFill, size_t& toFillLen, |
| const Charset::Tables& tables) { | const Charset::Tables& tables) { |
| const XMLByte* srcPtr=srcData; | const XMLByte* srcPtr=srcData; |
| const XMLByte* srcEnd=srcData+srcLen; | const XMLByte* srcEnd=srcData+srcLen; |
| XMLByte* outPtr=toFill; | XMLByte* outPtr=toFill; |
| XMLByte* outEnd=toFill+toFillLen; | XMLByte* outEnd=toFill+toFillLen; |
| while(srcPtr<srcEnd) { | while(srcPtr<srcEnd) { |
| uint curVal = tables.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 | |
| unsigned int encodedBytes; | |
| if(curVal<0x80) | |
| encodedBytes = 1; | |
| else if(curVal<0x800) | |
| encodedBytes = 2; | |
| else if(curVal<0x10000) | |
| encodedBytes = 3; | |
| else if(curVal<0x200000) | |
| encodedBytes = 4; | |
| else if(curVal<0x4000000) | |
| encodedBytes = 5; | |
| else if(curVal<= 0x7FFFFFFF) | |
| encodedBytes = 6; | |
| else { | |
| // use the replacement character | |
| *outPtr++= '?'; | |
| srcPtr++; | |
| continue; | |
| } | |
| // If we cannot fully get this char into the output buffer | |
| if (outPtr + encodedBytes > outEnd) | |
| break; | |
| // We can do it, so update the source index | |
| srcPtr++; | |
| // 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. | |
| outPtr+= encodedBytes; | |
| switch(encodedBytes) { | |
| case 6: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 5: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 4: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 3: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 2: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 1: *--outPtr = XMLByte(curVal | gFirstByteMark[encodedBytes]); | |
| } | |
| // Add the encoded bytes back in again to indicate we've eaten them | |
| outPtr+= encodedBytes; | |
| } | |
| // Update the bytes eaten | |
| srcLen = srcPtr - srcData; | |
| // Return the characters read | // Figure out how many bytes we need |
| toFillLen = outPtr - toFill; | unsigned int encodedBytes; |
| if(curVal<0x80) | |
| encodedBytes = 1; | |
| else if(curVal<0x800) | |
| encodedBytes = 2; | |
| else if(curVal<0x10000) | |
| encodedBytes = 3; | |
| else if(curVal<0x200000) | |
| encodedBytes = 4; | |
| else if(curVal<0x4000000) | |
| encodedBytes = 5; | |
| else if(curVal<= 0x7FFFFFFF) | |
| encodedBytes = 6; | |
| else { | |
| // use the replacement character | |
| *outPtr++= '?'; | |
| srcPtr++; | |
| continue; | |
| } | |
| // If we cannot fully get this char into the output buffer | |
| if (outPtr + encodedBytes > outEnd) | |
| break; | |
| // We can do it, so update the source index | |
| srcPtr++; | |
| // 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. | |
| outPtr+= encodedBytes; | |
| switch(encodedBytes) { | |
| case 6: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 5: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 4: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 3: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 2: *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL); | |
| curVal>>= 6; | |
| case 1: *--outPtr = XMLByte(curVal | gFirstByteMark[encodedBytes]); | |
| } | |
| // Add the encoded bytes back in again to indicate we've eaten them | |
| outPtr+= encodedBytes; | |
| } | |
| // Update the bytes eaten | |
| srcLen = srcPtr - srcData; | |
| // Return the characters read | |
| toFillLen = outPtr - toFill; | |
| //return srcPtr==srcEnd?(int)toFillLen:-1; | //return srcPtr==srcEnd?(int)toFillLen:-1; |
| /* | /* |
| xmlCharEncodingInputFunc | xmlCharEncodingInputFunc |
| Line 380 of ocetes consumed. | Line 341 of ocetes consumed. |
| } | } |
| /// @todo digital entites only when xml/html output [at output in html/xml mode, in html part of a letter] | /// @todo digital entites only when xml/html output [at output in html/xml mode, in html part of a letter] |
| static int transcodeFromUTF8( | static int transcodeFromUTF8( |
| const XMLByte *srcData, size_t& srcLen, | const XMLByte* srcData, size_t& srcLen, |
| XMLByte* toFill, size_t& toFillLen, | XMLByte* toFill, size_t& toFillLen, |
| const Charset::Tables& tables) { | const Charset::Tables& tables) { |
| const XMLByte* srcPtr=srcData; | const XMLByte* srcPtr=srcData; |
| Line 391 static int transcodeFromUTF8( | Line 352 static int transcodeFromUTF8( |
| // We now loop until we either run out of input data, or room to store | // We now loop until we either run out of input data, or room to store |
| while ((srcPtr < srcEnd) && (outPtr < outEnd)) { | 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) { |
| Line 422 static int transcodeFromUTF8( | Line 383 static int transcodeFromUTF8( |
| default: | default: |
| throw Exception(0, | throw Exception(0, |
| Exception::undefined_source, | 0, |
| "transcodeFromUTF8 error: wrong trailingBytes value(%d)", trailingBytes); | "transcodeFromUTF8 error: wrong trailingBytes value(%d)", trailingBytes); |
| } | } |
| tmpVal-=gUTFOffsets[trailingBytes]; | tmpVal-=gUTFOffsets[trailingBytes]; |
| Line 437 static int transcodeFromUTF8( | Line 398 static int transcodeFromUTF8( |
| outPtr+=sprintf((char *)outPtr, "&#%d;", tmpVal); // &#decimal; | outPtr+=sprintf((char *)outPtr, "&#%d;", tmpVal); // &#decimal; |
| } else | } else |
| throw Exception(0, | throw Exception(0, |
| Exception::undefined_source, | 0, |
| "transcodeFromUTF8 error: too big tmpVal(0x%08X)", tmpVal); | "transcodeFromUTF8 error: too big tmpVal(0x%08X)", tmpVal); |
| } | } |
| Line 459 of ocetes consumed. | Line 420 of ocetes consumed. |
| } | } |
| /// @todo not so memory-hungry with prescan | /// @todo not so memory-hungry with prescan |
| void Charset::transcodeToUTF8(Pool& pool, | const String::C Charset::transcodeToUTF8(const String::C src) const { |
| const void *source_body, size_t source_content_length, | size_t src_length=src.length; |
| const void *& adest_body, size_t& dest_content_length) const { | size_t dest_length=src.length*6/*so that surly enough, max utf8 seq len=6*/; |
| dest_content_length=source_content_length*6/*so that surly enough, max utf8 seq len=6*/; | XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; |
| XMLByte *dest_body=(XMLByte*)pool.malloc(dest_content_length); | |
| if(::transcodeToUTF8( | if(::transcodeToUTF8( |
| (XMLByte *)source_body, source_content_length, | (XMLByte *)src.str, src_length, |
| dest_body, dest_content_length, | dest_body, dest_length, |
| tables)<0) | tables)<0) |
| throw(0, 0, | throw(0, 0, |
| 0, | 0, |
| "Charset::transcodeToUTF8 buffer overflow"); | "Charset::transcodeToUTF8 buffer overflow"); |
| // return | dest_body[dest_length]=0; // terminator |
| adest_body=dest_body; | return String::C((char*)dest_body, dest_length); |
| } | } |
| void Charset::transcodeFromUTF8(Pool& pool, | const String::C Charset::transcodeFromUTF8(const String::C src) const { |
| const void *source_body, size_t source_content_length, | size_t src_length=src.length; |
| const void *& adest_body, size_t& dest_content_length) const { | size_t dest_length=src.length*6/*so that surly enough, "ÿ" has max ratio */; |
| dest_content_length=source_content_length*6/*so that surly enough, "ÿ" has max ratio */; | XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; |
| XMLByte *dest_body=(XMLByte*)pool.malloc(dest_content_length); | |
| if(::transcodeFromUTF8( | if(::transcodeFromUTF8( |
| (XMLByte *)source_body, source_content_length, | (XMLByte *)src.str, src_length, |
| dest_body, dest_content_length, | dest_body, dest_length, |
| tables)<0) | tables)<0) |
| throw(0, 0, | throw(0, 0, |
| 0, | 0, |
| "Charset::transcodeToUTF8 buffer overflow"); | "Charset::transcodeFromUTF8 buffer overflow"); |
| // return | dest_body[dest_length]=0; // terminator |
| adest_body=dest_body; | return String::C((char*)dest_body, dest_length); |
| } | } |
| /// transcode using both charsets | /// transcode using both charsets |
| void Charset::transcodeToCharset(Pool& pool, | const String::C Charset::transcodeToCharset(const String::C src, |
| const Charset& dest_charset, | const Charset& dest_charset) const { |
| const void *source_body, size_t source_content_length, | if(&dest_charset==this) |
| const void *& adest_body, size_t& adest_content_length) const { | return src; |
| if(&dest_charset==this) { | else { |
| adest_body=source_body; | size_t dest_length=src.length; |
| adest_content_length=source_content_length; | XMLByte* dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; |
| } else { | |
| size_t dest_content_length=source_content_length; | XMLByte* output=dest_body; |
| unsigned char *dest_body=(unsigned char *)pool.malloc(dest_content_length); | const XMLByte* input=(XMLByte *)src.str; |
| while(XMLCh c=*input++) { | |
| const XMLByte* srcPtr=(XMLByte *)source_body; | XMLCh curVal = tables.fromTable[c]; |
| const XMLByte* srcEnd=(XMLByte *)source_body+source_content_length; | *output++=curVal? |
| xlatOneTo(curVal, dest_charset.tables, '?') // OK | |
| for(XMLByte* outPtr=dest_body; srcPtr<srcEnd; srcPtr++) { | :'?'; // use the replacement character |
| XMLCh curVal = tables.fromTable[*srcPtr]; | |
| if(curVal) | |
| *outPtr++=xlatOneTo(curVal, dest_charset.tables, '?'); | |
| else { | |
| // use the replacement character | |
| *outPtr++= '?'; | |
| } | |
| } | } |
| adest_body=dest_body; | dest_body[dest_length]=0; // terminator |
| adest_content_length=dest_content_length; | return String::C((char*)dest_body, dest_length); |
| } | } |
| } | } |
| #ifdef XML | #ifdef XML |
| static int xml256CharEncodingInputFunc ( | |
| unsigned char *out, | |
| int *outlen, | |
| const unsigned char *in, | |
| int *inlen, | |
| void *info) { | |
| return transcodeToUTF8( | |
| in, *(size_t*)inlen, | |
| out, *(size_t*)outlen, | |
| *(const Charset::Tables *)info); | |
| } | |
| static int xml256CharEncodingOutputFunc ( | |
| unsigned char *out, | |
| int *outlen, | |
| const unsigned char *in, | |
| int *inlen, | |
| void *info) { | |
| return transcodeFromUTF8( | |
| in, *(size_t*)inlen, | |
| out, *(size_t*)outlen, | |
| *(const Charset::Tables *)info); | |
| } | |
| static const Charset::Tables* tables[MAX_CHARSETS]; | |
| #define declareXml256ioFuncs(i) \ | |
| static int xml256CharEncodingInputFunc##i( \ | |
| unsigned char *out, int *outlen, \ | |
| const unsigned char *in, int *inlen) { \ | |
| return transcodeToUTF8( \ | |
| in, *(size_t*)inlen, \ | |
| out, *(size_t*)outlen, \ | |
| *tables[i]); \ | |
| } \ | |
| static int xml256CharEncodingOutputFunc##i( \ | |
| unsigned char *out, int *outlen, \ | |
| const unsigned char *in, int *inlen) { \ | |
| return transcodeFromUTF8( \ | |
| in, *(size_t*)inlen, \ | |
| out, *(size_t*)outlen, \ | |
| *tables[i]); \ | |
| } | |
| declareXml256ioFuncs(0) declareXml256ioFuncs(1) | |
| declareXml256ioFuncs(2) declareXml256ioFuncs(3) | |
| declareXml256ioFuncs(4) declareXml256ioFuncs(5) | |
| declareXml256ioFuncs(6) declareXml256ioFuncs(7) | |
| declareXml256ioFuncs(8) declareXml256ioFuncs(9) | |
| static xmlCharEncodingInputFunc inputFuncs[MAX_CHARSETS]={ | |
| xml256CharEncodingInputFunc0, xml256CharEncodingInputFunc1, | |
| xml256CharEncodingInputFunc2, xml256CharEncodingInputFunc3, | |
| xml256CharEncodingInputFunc4, xml256CharEncodingInputFunc5, | |
| xml256CharEncodingInputFunc6, xml256CharEncodingInputFunc7, | |
| xml256CharEncodingInputFunc8, xml256CharEncodingInputFunc9 | |
| }; | |
| static xmlCharEncodingOutputFunc outputFuncs[MAX_CHARSETS]={ | |
| xml256CharEncodingOutputFunc0, xml256CharEncodingOutputFunc1, | |
| xml256CharEncodingOutputFunc2, xml256CharEncodingOutputFunc3, | |
| xml256CharEncodingOutputFunc4, xml256CharEncodingOutputFunc5, | |
| xml256CharEncodingOutputFunc6, xml256CharEncodingOutputFunc7, | |
| xml256CharEncodingOutputFunc8, xml256CharEncodingOutputFunc9 | |
| }; | |
| static size_t handlers_count=0; | |
| void Charset::addEncoding(char *name_cstr) { | void Charset::addEncoding(char *name_cstr) { |
| xmlCharEncodingHandler *handler= | if(handlers_count==MAX_CHARSETS) |
| (xmlCharEncodingHandler *)malloc(sizeof(xmlCharEncodingHandler)); | throw Exception(0, |
| handler->name=name_cstr; | 0, |
| handler->input=xml256CharEncodingInputFunc; handler->inputInfo=&tables; | "already allocated %d handlers, no space for new encoding '%s'", |
| handler->output=xml256CharEncodingOutputFunc; handler->outputInfo=&tables; | MAX_CHARSETS, name_cstr); |
| xmlCharEncodingHandler* handler=new(PointerFreeGC) xmlCharEncodingHandler; | |
| { | |
| handler->name=name_cstr; | |
| handler->input=inputFuncs[handlers_count]; | |
| handler->output=outputFuncs[handlers_count]; | |
| ::tables[handlers_count]=&tables; | |
| handlers_count++; | |
| } | |
| xmlRegisterCharEncodingHandler(handler); | xmlRegisterCharEncodingHandler(handler); |
| } | } |
| void Charset::initTranscoder(const String *source, const char *name_cstr) { | void Charset::initTranscoder(const StringBody NAME, const char* name_cstr) { |
| ftranscoder=xmlFindCharEncodingHandler(name_cstr); | ftranscoder=xmlFindCharEncodingHandler(name_cstr); |
| transcoder(source); // check right way | transcoder(NAME); // check right way |
| } | } |
| xmlCharEncodingHandler *Charset::transcoder(const String *source) { | xmlCharEncodingHandler& Charset::transcoder(const StringBody NAME) { |
| if(!ftranscoder) | if(!ftranscoder) |
| throw Exception("parser.runtime", | throw Exception("parser.runtime", |
| source, | new String(NAME, String::L_TAINTED), |
| "unsupported encoding"); | "unsupported encoding"); |
| return ftranscoder; | return *ftranscoder; |
| } | } |
| const char *Charset::transcode_cstr(xmlChar *s) { | String::C Charset::transcode_cstr(xmlChar* s) { |
| if(!s) | if(!s) |
| return ""; | return String::C("", 0); |
| int inlen=strlen((const char *)s); | int inlen=strlen((const char*)s); |
| int outlen=inlen+1; // max | int outlen=inlen; // max |
| char *out=(char *)malloc(outlen*sizeof(char)); | #ifndef NDEBUG |
| int saved_outlen=outlen; | |
| #endif | |
| char *out=new(PointerFreeGC) char[outlen+1]; | |
| int error; | int error; |
| if(xmlCharEncodingOutputFunc output=transcoder(0)->output) { | if(xmlCharEncodingOutputFunc output=transcoder(FNAME).output) { |
| error=output( | error=output( |
| (unsigned char*)out, &outlen, | (unsigned char*)out, &outlen, |
| (const unsigned char*)s, &inlen, | (const unsigned char*)s, &inlen); |
| transcoder(0)->outputInfo); | |
| } else { | } else { |
| memcpy(out, s, outlen=inlen); | memcpy(out, s, outlen=inlen); |
| error=0; | error=0; |
| Line 596 const char *Charset::transcode_cstr(xmlC | Line 578 const char *Charset::transcode_cstr(xmlC |
| 0, | 0, |
| "transcode_cstr failed (%d)", error); | "transcode_cstr failed (%d)", error); |
| out[outlen/*surely would be less then on input*/]=0; | assert(outlen<=saved_outlen); out[outlen]=0; |
| return out; | return String::C(out, outlen); |
| } | } |
| String& Charset::transcode(xmlChar *s | const String& Charset::transcode(xmlChar* s) { |
| #ifndef NO_STRING_ORIGIN | String::C cstr=transcode_cstr(s); |
| , const String *origin | return *new String(cstr.str, cstr.length, true); |
| #endif | } |
| ) { | String::C Charset::transcode_cstr(GdomeDOMString* s) { |
| String& result=*NEW String(pool()); | return s?transcode_cstr(BAD_CAST s->str) |
| result.APPEND_CLEAN(transcode_cstr(s), 0/*auto-size*/, origin->origin().file, origin->origin().line); | :String::C("", 0); |
| return result; | } |
| } | const String& Charset::transcode(GdomeDOMString* s) { |
| const char *Charset::transcode_cstr(GdomeDOMString *s) { | String::C cstr=transcode_cstr(s); |
| return s?transcode_cstr(BAD_CAST s->str):""; | return *new String(cstr.str, cstr.length, true); |
| } | |
| String& Charset::transcode(GdomeDOMString *s | |
| #ifndef NO_STRING_ORIGIN | |
| , const String *origin | |
| #endif | |
| ) { | |
| String& result=*NEW String(pool()); | |
| result.APPEND_CLEAN(transcode_cstr(s), 0/*auto-size*/, origin->origin().file, origin->origin().line); | |
| return result; | |
| } | } |
| /// @test less memory using -maybe- xmlParserInputBufferCreateMem | /// @test less memory using -maybe- xmlParserInputBufferCreateMem |
| xmlChar *Charset::transcode_buf2xchar(const char *buf, size_t buf_size) { | void* Charset::transcode_buf2mchar(transcode_buf_malloc_func malloc_func, |
| unsigned char *out; | const char* buf, size_t buf_size) { |
| unsigned char* out; | |
| int outlen; | int outlen; |
| int error; | int error; |
| if(xmlCharEncodingInputFunc input=transcoder(0)->input) { | #ifndef NDEBUG |
| int saved_outlen; | |
| #endif | |
| if(xmlCharEncodingInputFunc input=transcoder(FNAME).input) { | |
| outlen=buf_size*6/*max*/; | outlen=buf_size*6/*max*/; |
| out=(unsigned char*)malloc((outlen+1)*sizeof(unsigned char)); | #ifndef NDEBUG |
| saved_outlen=outlen; | |
| #endif | |
| out=(unsigned char*)malloc_func(outlen+1); | |
| error=input( | error=input( |
| out, &outlen, | out, &outlen, |
| (const unsigned char *)buf, (int *)&buf_size, | (const unsigned char*)buf, (int*)&buf_size); |
| transcoder(0)->inputInfo); | |
| } else { | } else { |
| outlen=buf_size; | outlen=buf_size; |
| out=(unsigned char*)malloc((outlen+1)*sizeof(unsigned char)); | #ifndef NDEBUG |
| saved_outlen=outlen; | |
| #endif | |
| out=(unsigned char*)malloc_func(outlen+1); | |
| memcpy(out, buf, outlen); | memcpy(out, buf, outlen); |
| error=0; | error=0; |
| } | } |
| Line 645 xmlChar *Charset::transcode_buf2xchar(co | Line 627 xmlChar *Charset::transcode_buf2xchar(co |
| 0, | 0, |
| "transcode_buf failed (%d)", error); | "transcode_buf failed (%d)", error); |
| out[outlen/*surely would be less then on input*/]=0; | assert(outlen<=saved_outlen); out[outlen]=0; |
| return (xmlChar *)out; | return out; |
| } | } |
| GdomeDOMString_auto_ptr Charset::transcode_buf2dom(const char *buf, size_t buf_size) { | |
| return GdomeDOMString_auto_ptr((gchar*)transcode_buf2xchar(buf, buf_size)); | xmlChar* Charset::transcode_buf2xchar(const char* buf, size_t buf_size) { |
| return static_cast<xmlChar*>(transcode_buf2mchar(xmlMalloc, buf, buf_size)); | |
| } | |
| static void* g_malloc_wrapper(size_t size) { | |
| if(void* out=g_malloc(size)) | |
| return out; | |
| return pa_fail_alloc("g_malloc_wrapper", size); | |
| } | |
| gchar* Charset::transcode_buf2gchar(const char* buf, size_t buf_size) { | |
| return static_cast<gchar*>(transcode_buf2mchar(g_malloc_wrapper, buf, buf_size)); | |
| } | |
| GdomeDOMString_auto_ptr Charset::transcode_buf2dom(const char* buf, size_t buf_size) { | |
| return GdomeDOMString_auto_ptr(transcode_buf2gchar(buf, buf_size)); | |
| } | } |
| GdomeDOMString_auto_ptr 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::L_UNSPECIFIED); |
| return transcode_buf2dom(cstr, strlen(cstr)); | return transcode_buf2dom(cstr, strlen(cstr)); |
| } | } |
| GdomeDOMString_auto_ptr Charset::transcode(const StringBody s) { | |
| const char* cstr=s.cstr(); | |
| return transcode_buf2dom(cstr, s.length()); | |
| } | |
| #endif | #endif |