|
|
| version 1.52, 2005/08/09 08:14:51 | version 1.80, 2009/10/03 02:18:11 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: Charset connection implementation. | Parser: Charset connection implementation. |
| Copyright(c) 2001-2005 ArtLebedev Group (http://www.artlebedev.com) | Copyright(c) 2001-2009 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) |
| */ | */ |
| Line 16 static const char * const IDENT_CHARSET_ | Line 16 static const char * const IDENT_CHARSET_ |
| //#define PA_PATCHED_LIBXML_BACKWARD | //#define PA_PATCHED_LIBXML_BACKWARD |
| // reduce memory usage by pre-calculation utf-8 string length | |
| #define PRECALCULATE_DEST_LENGTH | |
| // globals | // globals |
| Charset::UTF8CaseTable::Rec UTF8CaseToUpperRecords[]={ | Charset::UTF8CaseTable::Rec UTF8CaseToUpperRecords[]={ |
| Line 38 inline void prepare_case_tables(unsigned | Line 41 inline void prepare_case_tables(unsigned |
| unsigned char *lcc_table=tables+lcc_offset; | unsigned char *lcc_table=tables+lcc_offset; |
| unsigned char *fcc_table=tables+fcc_offset; | unsigned char *fcc_table=tables+fcc_offset; |
| for(int i=0; i<0x100; i++) | for(int i=0; i<0x100; i++) |
| lcc_table[i]=fcc_table[i]=i; | lcc_table[i]=fcc_table[i]=(unsigned char)i; |
| } | } |
| inline void cstr2ctypes(unsigned char *tables, const unsigned char *cstr, | inline void cstr2ctypes(unsigned char *tables, const unsigned char *cstr, |
| unsigned char bit) { | unsigned char bit) { |
| Line 85 static void element2case(unsigned char f | Line 88 static void element2case(unsigned char f |
| // methods | // methods |
| extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c | |
| Charset::Charset(Request_charsets* charsets, const String::Body ANAME, const String* afile_spec): | Charset::Charset(Request_charsets* charsets, const String::Body ANAME, const String* afile_spec): |
| FNAME(ANAME), | FNAME(ANAME), |
| FNAME_CSTR(ANAME.cstrm()) { | FNAME_CSTR(ANAME.cstrm()) { |
| Line 99 Charset::Charset(Request_charsets* chars | Line 101 Charset::Charset(Request_charsets* chars |
| } 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)); |
| } | } |
| #ifdef XML | #ifdef XML |
| Line 134 void Charset::load_definition(Request_ch | Line 136 void Charset::load_definition(Request_ch |
| continue; | continue; |
| // char white-space digit hex-digit letter word lowercase unicode1 unicode2 | // char white-space digit hex-digit letter word lowercase unicode1 unicode2 |
| unsigned int c=0; | unsigned char c=0; |
| char *cell; | char *cell; |
| for(int column=0; (cell=lsplit(&row, '\t')); column++) { | for(int column=0; (cell=lsplit(&row, '\t')); column++) { |
| switch(column) { | switch(column) { |
| case 0: c=to_wchar_code(cell); break; | case 0: c=(unsigned char)to_wchar_code(cell); break; |
| // pcre_tables | // pcre_tables |
| case 1: element2ctypes(c, to_bool(cell), pcre_tables, ctype_space, cbit_space); break; | case 1: element2ctypes(c, to_bool(cell), pcre_tables, ctype_space, cbit_space); break; |
| case 2: element2ctypes(c, to_bool(cell), pcre_tables, ctype_digit, cbit_digit); break; | case 2: element2ctypes(c, to_bool(cell), pcre_tables, ctype_digit, cbit_digit); break; |
| case 3: element2ctypes(c, to_bool(cell), pcre_tables, ctype_xdigit); break; | case 3: element2ctypes(c, to_bool(cell), pcre_tables, ctype_xdigit); break; |
| case 4: element2ctypes(c, to_bool(cell), pcre_tables, ctype_letter); break; | case 4: element2ctypes(c, to_bool(cell), pcre_tables, ctype_letter); break; |
| case 5: element2ctypes(c, to_bool(cell), pcre_tables, ctype_word, cbit_word); break; | case 5: element2ctypes(c, to_bool(cell), pcre_tables, ctype_word, cbit_word); break; |
| case 6: element2case(c, to_wchar_code(cell), pcre_tables); break; | case 6: element2case(c, (unsigned char)to_wchar_code(cell), pcre_tables); break; |
| case 7: | case 7: |
| case 8: | case 8: |
| // 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); |
| Line 187 void Charset::sort_ToTable() { | Line 189 void Charset::sort_ToTable() { |
| //fclose(f); | //fclose(f); |
| } | } |
| // @todo: precache for spedup searching | |
| 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) { |
| int lo = 0; | int lo = 0; |
| int hi = tables.toTableSize - 1; | int hi = tables.toTableSize - 1; |
| while(lo<=hi) { | while(lo<=hi) { |
| // Calc the mid point of the low and high offset. | // Calc the mid point of the low and high offset. |
| const unsigned int i = (lo + hi) / 2; | const unsigned int i = (lo + hi) / 2; |
| Line 260 static const XMLByte gUTFBytes[0x100] = | Line 263 static const XMLByte gUTFBytes[0x100] = |
| }; | }; |
| static const uint gUTFOffsets[6] = { | static const uint gUTFOffsets[6] = { |
| 0, 0x3080, 0xE2080, 0x3C82080, 0xFA082080, 0x82082080 | 0, 0x3080, 0xE2080, 0x3C82080, 0xFA082080, 0x82082080 |
| }; | }; |
| static const XMLByte gFirstByteMark[7] = { | static const XMLByte gFirstByteMark[7] = { |
| 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC | 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC |
| }; | }; |
| static int transcodeToUTF8(const XMLByte* srcData, size_t& srcLen, | static int transcodeToUTF8(const XMLByte* srcData, int& srcLen, |
| XMLByte *toFill, size_t& toFillLen, | XMLByte *toFill, int& 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; |
| Line 316 static int transcodeToUTF8(const XMLByte | Line 319 static int transcodeToUTF8(const XMLByte |
| // 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 |
| Line 350 of ocetes consumed. | Line 353 of ocetes consumed. |
| return 0; | return 0; |
| } | } |
| /// @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(const XMLByte* srcData, size_t& srcLen, | static int transcodeFromUTF8(const XMLByte* srcData, int& srcLen, |
| XMLByte* toFill, size_t& toFillLen, | XMLByte* toFill, int& 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; |
| Line 364 static int transcodeFromUTF8(const XMLBy | Line 367 static int transcodeFromUTF8(const XMLBy |
| 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 430 of ocetes consumed. | Line 433 of ocetes consumed. |
| return 0; | return 0; |
| } | } |
| /// @todo not so memory-hungry with prescan | static bool isEscaped(XMLByte c){ |
| const String::C Charset::transcodeToUTF8(const String::C src) const { | return |
| !( | |
| (c<=127) | |
| && ( | |
| ((c>='0') && (c<='9')) | |
| || ((c>='A') && (c<='Z')) | |
| || ((c>='a') && (c<='z')) | |
| || strchr("*@-_+./", c)!=0 | |
| ) | |
| ); | |
| } | |
| // read one UTF8 char and return length of this char (in bytes) | |
| static unsigned int readUTF8Char(const XMLByte*& srcPtr, const XMLByte* srcEnd, XMLByte& firstByte, XMLCh& UTF8Char){ | |
| if(!srcPtr || !*srcPtr || srcPtr>=srcEnd) | |
| return 0; | |
| firstByte=*srcPtr; | |
| if(firstByte<=127){ | |
| UTF8Char=firstByte; | |
| srcPtr++; | |
| return 1; | |
| } | |
| unsigned int trailingBytes=gUTFBytes[firstByte]; | |
| if(srcPtr+trailingBytes>=srcEnd){ | |
| return 0; // not enough bytes in source string for reading | |
| } | |
| uint tmpVal=0; | |
| switch(trailingBytes){ | |
| case 5: tmpVal+=*srcPtr++; tmpVal<<=6; | |
| case 4: tmpVal+=*srcPtr++; tmpVal<<=6; | |
| case 3: tmpVal+=*srcPtr++; tmpVal<<=6; | |
| case 2: tmpVal+=*srcPtr++; tmpVal<<=6; | |
| case 1: tmpVal+=*srcPtr++; tmpVal<<=6; | |
| case 0: tmpVal+=*srcPtr++; | |
| } | |
| tmpVal-=gUTFOffsets[trailingBytes]; | |
| UTF8Char=tmpVal; | |
| return trailingBytes+1; | |
| } | |
| // skip UTF8 char and return length of this char (in bytes) | |
| static unsigned int skipUTF8Char(const XMLByte*& srcPtr, const XMLByte* srcEnd){ | |
| if(!srcPtr || !*srcPtr || srcPtr>=srcEnd) | |
| return 0; | |
| unsigned int trailingBytes=gUTFBytes[*srcPtr]+1; | |
| srcPtr+=trailingBytes; | |
| return trailingBytes; | |
| } | |
| // read non-UTF8 char, and return number of bytes needed for store this char in UTF8 | |
| static unsigned int readChar(const XMLByte*& srcPtr, const XMLByte* srcEnd, XMLByte& firstByte, XMLCh& UTF8Char, const Charset::Tables& tables){ | |
| if(!srcPtr || !*srcPtr || srcPtr>=srcEnd) | |
| return 0; | |
| firstByte=*srcPtr++; | |
| UTF8Char=tables.fromTable[firstByte]; | |
| if(UTF8Char<0x80) | |
| return 1; | |
| else if(UTF8Char<0x800) | |
| return 2; | |
| else if(UTF8Char<0x10000) | |
| return 3; | |
| else if(UTF8Char<0x200000) | |
| return 4; | |
| else if(UTF8Char<0x4000000) | |
| return 5; | |
| else if(UTF8Char<= 0x7FFFFFFF) | |
| return 6; | |
| // will use the replacement character '?' | |
| firstByte=0; | |
| return 1; | |
| } | |
| static int escapeUTF8(const XMLByte* srcData, size_t& srcLen, | |
| XMLByte* toFill, size_t& toFillLen) { | |
| const XMLByte* srcPtr=srcData; | |
| const XMLByte* srcEnd=srcData+srcLen; | |
| XMLByte* outPtr=toFill; | |
| XMLByte* outEnd=toFill+toFillLen; | |
| XMLByte firstByte; | |
| XMLCh UTF8Char; | |
| uint charSize; | |
| // loop until we either run out of input data, or room to store | |
| while((outPtr < outEnd) && (charSize=readUTF8Char(srcPtr, srcEnd, firstByte, UTF8Char))){ | |
| if(charSize==1){ | |
| if(isEscaped(firstByte)) // %XX | |
| outPtr+=sprintf((char*)outPtr, "%%%02X", firstByte); | |
| else | |
| *outPtr++=firstByte; | |
| } else | |
| outPtr+=sprintf((char*)outPtr, "%%u%04X", UTF8Char); // %uXXXX | |
| } | |
| // Update the bytes eaten | |
| srcLen=srcPtr-srcData; | |
| // Return the characters read | |
| toFillLen=outPtr-toFill; | |
| return 0; | |
| } | |
| static int escape(const XMLByte* srcData, size_t& srcLen, | |
| XMLByte *toFill, size_t& toFillLen, | |
| const Charset::Tables& tables) { | |
| const XMLByte* srcPtr=srcData; | |
| const XMLByte* srcEnd=srcData+srcLen; | |
| XMLByte* outPtr=toFill; | |
| XMLByte firstByte; | |
| XMLCh UTF8Char; | |
| uint charSize; | |
| while(charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, tables)){ | |
| if(charSize==1){ | |
| if(firstByte){ | |
| if(isEscaped(firstByte)) // %XX | |
| outPtr+=sprintf((char*)outPtr, "%%%02X", firstByte); | |
| else | |
| *outPtr++=firstByte; | |
| } else // add replacement char '?' | |
| *outPtr++='?'; | |
| } else | |
| outPtr+=sprintf((char*)outPtr, "%%u%04X", UTF8Char); // %uXXXX | |
| } | |
| // Update the bytes eaten | |
| srcLen = srcPtr - srcData; | |
| // Return the characters read | |
| toFillLen = outPtr - toFill; | |
| return 0; | |
| } | |
| String::C Charset::escape(const String::C src, const Charset& source_charset){ | |
| size_t src_length=src.length; | size_t src_length=src.length; |
| size_t dest_length=src.length*6/*so that surly enough, max utf8 seq len=6*/; | if(!src_length) |
| return String::C("", 0); | |
| #ifdef PRECALCULATE_DEST_LENGTH | |
| size_t dest_length=0; | |
| const XMLByte* srcPtr=(XMLByte*)src.str; | |
| const XMLByte* srcEnd=srcPtr+src_length; | |
| XMLByte firstByte; | |
| XMLCh UTF8Char; | |
| if(source_charset.isUTF8()){ | |
| while(uint charSize=readUTF8Char(srcPtr, srcEnd, firstByte, UTF8Char)){ | |
| if(charSize==1) | |
| dest_length+=!isEscaped(firstByte)?1:3/*%XX*/; | |
| else | |
| dest_length+=6; // '%uXXXX' | |
| } | |
| } else { | |
| while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, source_charset.tables)){ | |
| if(charSize==1) | |
| dest_length+=(!firstByte/*replacement char '?'*/ || !isEscaped(firstByte))?1:3/*'%XX'*/; | |
| else | |
| dest_length+=6; // '%uXXXX' | |
| } | |
| } | |
| #else | |
| size_t dest_length=src_length*6; // enough for %uXXXX but too memory-hungry | |
| #endif | |
| //throw Exception(0,0,"%u",dest_length); | |
| #ifndef NDEBUG | #ifndef NDEBUG |
| size_t saved_dest_length=dest_length; | size_t saved_dest_length=dest_length; |
| #endif | #endif |
| XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; | XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; |
| int status; | |
| if(source_charset.isUTF8()){ | |
| status=::escapeUTF8((XMLByte *)src.str, src_length, dest_body, dest_length); | |
| } else { | |
| status=::escape((XMLByte *)src.str, src_length, dest_body, dest_length, source_charset.tables); | |
| } | |
| if(status<0) | |
| throw Exception(0, | |
| 0, | |
| "Charset::escape buffer overflow"); | |
| assert(dest_length<=saved_dest_length); | |
| dest_body[dest_length]=0; // terminator | |
| return String::C((char*)dest_body, dest_length); | |
| } | |
| String::Body Charset::escape(const String::Body src, const Charset& source_charset) { | |
| const char *src_ptr=src.cstr(); | |
| size_t src_size=strlen(src_ptr); | |
| String::C dest=Charset::escape(String::C(src_ptr, src_size), source_charset); | |
| return String::Body(dest.length ? dest.str:0); | |
| } | |
| String& Charset::escape(const String& src, const Charset& source_charset) { | |
| if(src.is_empty()) | |
| return *new String(); | |
| return *new String(escape((String::Body)src, source_charset), String::L_CLEAN); | |
| } | |
| const String::C Charset::transcodeToUTF8(const String::C src) const { | |
| int src_length=src.length; | |
| #ifdef PRECALCULATE_DEST_LENGTH | |
| int dest_length=0; | |
| const XMLByte* srcPtr=(XMLByte*)src.str; | |
| const XMLByte* srcEnd=srcPtr+src_length; | |
| XMLByte firstByte; | |
| XMLCh UTF8Char; | |
| while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, tables)) | |
| dest_length+=charSize; | |
| #else | |
| int dest_length=src_length*6; // so that surly enough (max utf8 seq len=6) but too memory-hyngry | |
| #endif | |
| //throw Exception(0,0,"%u",dest_length); | |
| #ifndef NDEBUG | |
| int saved_dest_length=dest_length; | |
| #endif | |
| XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; | |
| if(::transcodeToUTF8( | if(::transcodeToUTF8( |
| (XMLByte *)src.str, src_length, | (XMLByte *)src.str, src_length, |
| dest_body, dest_length, | dest_body, dest_length, |
| Line 447 const String::C Charset::transcodeToUTF8 | Line 682 const String::C Charset::transcodeToUTF8 |
| 0, | 0, |
| "Charset::transcodeToUTF8 buffer overflow"); | "Charset::transcodeToUTF8 buffer overflow"); |
| assert(dest_length<=saved_dest_length); dest_body[dest_length]=0; // terminator | assert(dest_length<=saved_dest_length); |
| dest_body[dest_length]=0; // terminator | |
| return String::C((char*)dest_body, dest_length); | return String::C((char*)dest_body, dest_length); |
| } | } |
| static XMLCh change_case_UTF8(const XMLCh src, const Charset::UTF8CaseTable& table) { | static XMLCh change_case_UTF8(const XMLCh src, const Charset::UTF8CaseTable& table) { |
| int lo = 0; | int lo = 0; |
| int hi = table.size - 1; | int hi = table.size - 1; |
| while(lo<=hi) { | while(lo<=hi) { |
| // Calc the mid point of the low and high offset. | // Calc the mid point of the low and high offset. |
| const unsigned int i = (lo + hi) / 2; | const unsigned int i = (lo + hi) / 2; |
| Line 471 static XMLCh change_case_UTF8(const XMLC | Line 707 static XMLCh change_case_UTF8(const XMLC |
| return src; | return src; |
| } | } |
| static void store_UTF8(XMLCh src, XMLByte*& outPtr ) { | static void store_UTF8(XMLCh src, XMLByte*& outPtr){ |
| if(!src) { | if(!src) { |
| // use the replacement character | // use the replacement character |
| *outPtr++= '?'; | *outPtr++= '?'; |
| Line 536 void change_case_UTF8(const XMLByte* src | Line 772 void change_case_UTF8(const XMLByte* src |
| // Get the next leading byte out | // Get the next leading byte out |
| const XMLByte firstByte =* srcPtr; | const XMLByte firstByte =* srcPtr; |
| if(firstByte<= 127) { | if(firstByte<=127) { |
| change_case_UTF8(firstByte, outPtr, table); | change_case_UTF8(firstByte, outPtr, table); |
| srcPtr++; | srcPtr++; |
| continue; | continue; |
| Line 580 void change_case_UTF8(const XMLByte* src | Line 816 void change_case_UTF8(const XMLByte* src |
| "change_case_UTF8 error: end pointers do not match"); | "change_case_UTF8 error: end pointers do not match"); |
| } | } |
| static size_t getDecNumLength(XMLCh UTF8Char){ | |
| return | |
| (UTF8Char < 100) | |
| ?2 | |
| :(UTF8Char < 1000) | |
| ?3 | |
| :(UTF8Char < 10000) | |
| ?4 | |
| :5; | |
| } | |
| const String::C Charset::transcodeFromUTF8(const String::C src) const { | const String::C Charset::transcodeFromUTF8(const String::C src) const { |
| size_t src_length=src.length; | int src_length=src.length; |
| size_t dest_length=src.length*6/*so that surly enough, "ÿ" has max ratio */; | |
| #ifdef PRECALCULATE_DEST_LENGTH | |
| int dest_length=0; | |
| const XMLByte* srcPtr=(XMLByte*)src.str; | |
| const XMLByte* srcEnd=srcPtr+src_length; | |
| XMLByte firstByte; | |
| XMLCh UTF8Char; | |
| while(uint charSize=readUTF8Char(srcPtr, srcEnd, firstByte, UTF8Char)){ | |
| if(charSize==1) | |
| dest_length++; | |
| else | |
| dest_length+=(UTF8Char & 0xFFFF0000) | |
| ?charSize*3 // '%XX' for each byte | |
| :(xlatOneTo(UTF8Char, tables, 0)!=0) | |
| ?1 // can convert it to single char | |
| :getDecNumLength(UTF8Char)+3; // &#XX; - &#XXXXX; | |
| } | |
| #else | |
| // so that surly enough, "&#XXX;" has max ratio (huh? 8 bytes needed for '&#XXXXX;') | |
| int dest_length=src_length*6; | |
| #endif | |
| #ifndef NDEBUG | #ifndef NDEBUG |
| size_t saved_dest_length=dest_length; | int saved_dest_length=dest_length; |
| #endif | #endif |
| XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; | XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; |
| Line 597 const String::C Charset::transcodeFromUT | Line 864 const String::C Charset::transcodeFromUT |
| 0, | 0, |
| "Charset::transcodeFromUTF8 buffer overflow"); | "Charset::transcodeFromUTF8 buffer overflow"); |
| assert(dest_length<=saved_dest_length); dest_body[dest_length]=0; // terminator | assert(dest_length<=saved_dest_length); |
| dest_body[dest_length]=0; // terminator | |
| return String::C((char*)dest_body, dest_length); | return String::C((char*)dest_body, dest_length); |
| } | } |
| /// transcode using both charsets | /// transcode using both charsets |
| const String::C Charset::transcodeToCharset(const String::C src, | const String::C Charset::transcodeToCharset(const String::C src, |
| const Charset& dest_charset) const { | const Charset& dest_charset) const { |
| if(&dest_charset==this) | if(&dest_charset==this) |
| return src; | return src; |
| else { | else { |
| Line 624 const String::C Charset::transcodeToChar | Line 892 const String::C Charset::transcodeToChar |
| } | } |
| } | } |
| void Charset::store_Char(XMLByte*& outPtr, XMLCh src, XMLByte not_found){ | |
| if(isUTF8()) | |
| store_UTF8(src, outPtr); | |
| else if(char ch=xlatOneTo(src, tables, not_found)) | |
| *outPtr++=ch; | |
| } | |
| #ifdef XML | #ifdef XML |
| static const Charset::Tables* tables[MAX_CHARSETS]; | static const Charset::Tables* tables[MAX_CHARSETS]; |
| Line 635 static const Charset::Tables* tables[MAX | Line 910 static const Charset::Tables* tables[MAX |
| unsigned char *out, int *outlen, \ | unsigned char *out, int *outlen, \ |
| const unsigned char *in, int *inlen, void*) { \ | const unsigned char *in, int *inlen, void*) { \ |
| return transcodeToUTF8( \ | return transcodeToUTF8( \ |
| in, *(size_t*)inlen, \ | in, *inlen, \ |
| out, *(size_t*)outlen, \ | out, *outlen, \ |
| *tables[i]); \ | *tables[i]); \ |
| } \ | } \ |
| static int xml256CharEncodingOutputFunc##i( \ | static int xml256CharEncodingOutputFunc##i( \ |
| unsigned char *out, int *outlen, \ | unsigned char *out, int *outlen, \ |
| const unsigned char *in, int *inlen, void*) { \ | const unsigned char *in, int *inlen, void*) { \ |
| return transcodeFromUTF8( \ | return transcodeFromUTF8( \ |
| in, *(size_t*)inlen, \ | in, *inlen, \ |
| out, *(size_t*)outlen, \ | out, *outlen, \ |
| *tables[i]); \ | *tables[i]); \ |
| } | } |
| Line 655 static const Charset::Tables* tables[MAX | Line 930 static const Charset::Tables* tables[MAX |
| unsigned char *out, int *outlen, \ | unsigned char *out, int *outlen, \ |
| const unsigned char *in, int *inlen) { \ | const unsigned char *in, int *inlen) { \ |
| return transcodeToUTF8( \ | return transcodeToUTF8( \ |
| in, *(size_t*)inlen, \ | in, *inlen, \ |
| out, *(size_t*)outlen, \ | out, *outlen, \ |
| *tables[i]); \ | *tables[i]); \ |
| } \ | } \ |
| static int xml256CharEncodingOutputFunc##i( \ | static int xml256CharEncodingOutputFunc##i( \ |
| unsigned char *out, int *outlen, \ | unsigned char *out, int *outlen, \ |
| const unsigned char *in, int *inlen) { \ | const unsigned char *in, int *inlen) { \ |
| return transcodeFromUTF8( \ | return transcodeFromUTF8( \ |
| in, *(size_t*)inlen, \ | in, *inlen, \ |
| out, *(size_t*)outlen, \ | out, *outlen, \ |
| *tables[i]); \ | *tables[i]); \ |
| } | } |
| Line 720 void Charset::initTranscoder(const Strin | Line 995 void Charset::initTranscoder(const Strin |
| xmlCharEncodingHandler& Charset::transcoder(const String::Body NAME) { | xmlCharEncodingHandler& Charset::transcoder(const String::Body NAME) { |
| if(!ftranscoder) | if(!ftranscoder) |
| throw Exception("parser.runtime", | throw Exception(PARSER_RUNTIME, |
| new String(NAME, String::L_TAINTED), | new String(NAME, String::L_TAINTED), |
| "unsupported encoding"); | "unsupported encoding"); |
| return *ftranscoder; | return *ftranscoder; |
| } | } |
| String::C Charset::transcode_cstr(xmlChar* s) { | String::C Charset::transcode_cstr(const xmlChar* s) { |
| if(!s) | if(!s) |
| return String::C("", 0); | return String::C("", 0); |
| Line 758 String::C Charset::transcode_cstr(xmlCha | Line 1033 String::C Charset::transcode_cstr(xmlCha |
| assert(outlen<=saved_outlen); out[outlen]=0; | assert(outlen<=saved_outlen); out[outlen]=0; |
| return String::C(out, outlen); | return String::C(out, outlen); |
| } | } |
| const String& Charset::transcode(xmlChar* s) { | const String& Charset::transcode(const xmlChar* s) { |
| String::C cstr=transcode_cstr(s); | |
| return *new String(cstr.str, cstr.length, true); | |
| } | |
| String::C Charset::transcode_cstr(GdomeDOMString* s) { | |
| return s?transcode_cstr(BAD_CAST s->str) | |
| :String::C("", 0); | |
| } | |
| const String& Charset::transcode(GdomeDOMString* s) { | |
| String::C cstr=transcode_cstr(s); | String::C cstr=transcode_cstr(s); |
| return *new String(cstr.str, cstr.length, true); | return *new String(cstr.str, String::L_TAINTED); |
| } | } |
| /// @test less memory using -maybe- xmlParserInputBufferCreateMem | /// @test less memory using -maybe- xmlParserInputBufferCreateMem |
| Line 810 xmlChar* Charset::transcode_buf2xchar(co | Line 1077 xmlChar* Charset::transcode_buf2xchar(co |
| assert(outlen<=saved_outlen); out[outlen]=0; | assert(outlen<=saved_outlen); out[outlen]=0; |
| return out; | return out; |
| } | } |
| GdomeDOMString_auto_ptr Charset::transcode_buf2dom(const char* buf, size_t buf_size) { | |
| return GdomeDOMString_auto_ptr(transcode_buf2xchar(buf, buf_size)); | |
| } | |
| GdomeDOMString_auto_ptr Charset::transcode(const String& s) { | |
| const char* cstr=s.cstr(String::L_UNSPECIFIED); | |
| return transcode_buf2dom(cstr, strlen(cstr)); | xmlChar* Charset::transcode(const String& s) { |
| String::Body sbody=s.cstr_to_string_body_untaint(String::L_AS_IS); | |
| return transcode_buf2xchar(sbody.cstr(), sbody.length()); | |
| } | } |
| GdomeDOMString_auto_ptr Charset::transcode(const String::Body s) { | |
| const char* cstr=s.cstr(); | |
| return transcode_buf2dom(cstr, s.length()); | xmlChar* Charset::transcode(const String::Body s) { |
| return transcode_buf2xchar(s.cstr(), s.length()); | |
| } | } |
| #endif | #endif |
| Line 832 String::Body Charset::transcode(const St | Line 1095 String::Body Charset::transcode(const St |
| const char *src_ptr=src.cstr(); | const char *src_ptr=src.cstr(); |
| size_t src_size=strlen(src_ptr); | size_t src_size=strlen(src_ptr); |
| String::C dest=Charset::transcode(String::C(src_ptr, src_size), | String::C dest=Charset::transcode(String::C(src_ptr, src_size), source_transcoder, dest_transcoder); |
| source_transcoder, | |
| dest_transcoder); | |
| return String::Body(dest.str, dest.length); | return String::Body(dest.length ? dest.str:0); |
| } | } |
| String& Charset::transcode(const String& src, | String& Charset::transcode(const String& src, |
| const Charset& source_transcoder, | const Charset& source_transcoder, |
| const Charset& dest_transcoder) { | const Charset& dest_transcoder) { |
| if(!src.length()) | if(src.is_empty()) |
| return *new String("", 0, false); | return *new String(); |
| return *new String(transcode((String::Body)src, source_transcoder, dest_transcoder), String::L_CLEAN); | return *new String(transcode((String::Body)src, source_transcoder, dest_transcoder), String::L_CLEAN); |
| } | } |
| Line 861 struct Transcode_pair_info { | Line 1122 struct Transcode_pair_info { |
| const Charset* dest_transcoder; | const Charset* dest_transcoder; |
| }; | }; |
| #endif | #endif |
| static void transcode_pair(const String::Body /*akey*/, | static void transcode_pair(HashStringValue::key_type /*akey*/, |
| String::Body& avalue, | String::Body& avalue, |
| Transcode_pair_info* info) { | Transcode_pair_info* info) { |
| avalue=Charset::transcode(avalue, | avalue=Charset::transcode(avalue, |
| *info->source_transcoder, | *info->source_transcoder, |
| *info->dest_transcoder); | *info->dest_transcoder); |
| } | } |
| void Charset::transcode(HashStringString& src, | void Charset::transcode(HashStringString& src, |
| const Charset& source_transcoder, | const Charset& source_transcoder, |
| const Charset& dest_transcoder) { | const Charset& dest_transcoder) { |
| Transcode_pair_info info={&source_transcoder, &dest_transcoder}; | Transcode_pair_info info={&source_transcoder, &dest_transcoder}; |
| src.for_each_ref(transcode_pair, &info); | src.for_each_ref<Transcode_pair_info*>(transcode_pair, &info); |
| } | |
| size_t getUTF8BytePos(const XMLByte* srcBegin, const XMLByte* srcEnd, size_t charPos){ | |
| const XMLByte* ptr=srcBegin; | |
| while(charPos-- && skipUTF8Char(ptr, srcEnd)); | |
| return ptr-srcBegin; | |
| } | |
| size_t getUTF8CharPos(const XMLByte* srcBegin, const XMLByte* srcEnd, size_t bytePos){ | |
| size_t charPos=0; | |
| const XMLByte* ptr=srcBegin; | |
| const XMLByte* ptrEnd=srcBegin+bytePos; | |
| while(skipUTF8Char(ptr, srcEnd)){ | |
| if(ptr>ptrEnd) | |
| return charPos; | |
| charPos++; | |
| } | |
| // scan till end but position in bytes still too low | |
| throw Exception(0, | |
| 0, | |
| "Error convertion byte pos to char pos"); | |
| } | |
| size_t lengthUTF8(const XMLByte* srcBegin, const XMLByte* srcEnd){ | |
| size_t size=0; | |
| while(skipUTF8Char(srcBegin, srcEnd)) | |
| size++; | |
| return size; | |
| } | |
| bool UTF8_string_iterator::has_next(){ | |
| fcharSize=readUTF8Char(fsrcPtr, fsrcEnd, ffirstByte, fUTF8Char); | |
| return fcharSize!=0; | |
| } | } |