|
|
| version 1.40, 2003/11/20 15:35:31 | version 1.113, 2024/09/28 14:37:53 |
|---|---|
| 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) 2001-2023 Art. Lebedev Studio (http://www.artlebedev.com) |
| Author: Alexander Petrosyan<paf@design.ru>(http://paf.design.ru) | Authors: Konstantin Morshnev <moko@design.ru>, Alexandr Petrosian <paf@design.ru> |
| */ | */ |
| static const char* IDENT_CHARSET_C="$Date$"; | |
| #include "pa_charset.h" | #include "pa_charset.h" |
| #include "pa_charsets.h" | #include "pa_charsets.h" |
| // we are using some pcre_internal.h stuff as well | |
| #include "../lib/pcre/pa_pcre_internal.h" | |
| volatile const char * IDENT_PA_CHARSET_C="$Id$" IDENT_PA_CHARSET_H; | |
| #ifdef XML | #ifdef XML |
| #include "libxml/encoding.h" | #include "libxml/encoding.h" |
| #endif | #endif |
| //#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 36 inline void prepare_case_tables(unsigned | Line 44 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) { | |
| unsigned char *ctypes_table=tables+ctypes_offset; | unsigned char *ctypes_table=tables+ctypes_offset; |
| ctypes_table[0]=bit; | ctypes_table[0]=bit; |
| for(; *cstr; cstr++) { | for(; *cstr; cstr++) { |
| Line 53 inline unsigned int to_wchar_code(const | Line 60 inline unsigned int to_wchar_code(const |
| 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; | return pa_atoui(cstr,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, unsigned char *tables, unsigned char bit, int group_offset=-1) { |
| unsigned char *tables, unsigned char bit, int group_offset=-1) { | |
| if(!belongs) | if(!belongs) |
| return; | return; |
| Line 70 static void element2ctypes(unsigned char | Line 75 static void element2ctypes(unsigned char |
| 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) { | |
| if(!to) | if(!to) |
| return; | return; |
| Line 81 static void element2case(unsigned char f | Line 85 static void element2case(unsigned char f |
| fcc_table[from]=to; fcc_table[to]=from; | fcc_table[from]=to; fcc_table[to]=from; |
| } | } |
| inline XMLByte *append_hex_8(XMLByte *dest, unsigned char c, const char* prefix=0) { | |
| if(prefix) { | |
| strcpy((char *)dest, prefix); | |
| dest+=strlen(prefix); | |
| } | |
| *dest++=hex_digits[c >> 4]; | |
| *dest++=hex_digits[c & 0x0F]; | |
| return dest; | |
| } | |
| inline XMLByte *append_hex_16(XMLByte *dest, unsigned int c, const char* prefix=0) { | |
| if(prefix) { | |
| strcpy((char *)dest, prefix); | |
| dest+=strlen(prefix); | |
| } | |
| *dest++=hex_digits[(c >> 12) & 0x0F]; | |
| *dest++=hex_digits[(c >> 8) & 0x0F]; | |
| *dest++=hex_digits[(c >> 4) & 0x0F]; | |
| *dest++=hex_digits[(c) & 0x0F]; | |
| return dest; | |
| } | |
| // methods | // methods |
| extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c | Charset::Charset(Request_charsets* acharsets, 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()) { |
| if(afile_spec) { | if(afile_spec) { |
| fisUTF8=false; | fisUTF8=false; |
| load_definition(*charsets, *afile_spec); | load_definition(*acharsets, *afile_spec); |
| #ifdef XML | #ifdef XML |
| addEncoding(FNAME_CSTR); | addEncoding(FNAME_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, pa_pcre_default_tables, sizeof(pcre_tables)); |
| } | } |
| #ifdef XML | #ifdef XML |
| Line 105 Charset::Charset(Request_charsets* chars | Line 130 Charset::Charset(Request_charsets* chars |
| #endif | #endif |
| } | } |
| void Charset::load_definition(Request_charsets& charsets, const String& afile_spec) { | void Charset::load_definition(Request_charsets& acharsets, const String& afile_spec) { |
| // pcre_tables | // pcre_tables |
| // lowcase, flipcase, bits digit+word+whitespace, masks | // lowcase, flipcase, bits digit+word+whitespace, masks |
| Line 119 void Charset::load_definition(Request_ch | Line 144 void Charset::load_definition(Request_ch |
| memset(&tables, 0, sizeof(tables)); | memset(&tables, 0, sizeof(tables)); |
| // loading text | // loading text |
| char *data=file_read_text(charsets, afile_spec); | char *data=file_read_text(acharsets, afile_spec); |
| // ignore header | // ignore header |
| getrow(&data); | getrow(&data); |
| // parse cells | // parse cells |
| char *row; | char *row; |
| while(row=getrow(&data)) { | while((row=getrow(&data))) { |
| // remove empty&comment lines | // remove empty&comment lines |
| if(!*row || *row=='#') | if(!*row || *row=='#') |
| 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, "charset must contain not more than %d unicode values", MAX_CHARSET_UNI_CODES); |
| &afile_spec, | |
| "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*/) |
| Line 167 void Charset::load_definition(Request_ch | Line 190 void Charset::load_definition(Request_ch |
| } | } |
| }; | }; |
| // parser charset tables declare only white-space before 0x20, thus adding the missing chars | |
| for(uint i=0; i<0x20; i++) | |
| if(!tables.fromTable[i]){ | |
| tables.fromTable[i]=i; | |
| tables.toTable[tables.toTableSize].intCh=i; | |
| tables.toTable[tables.toTableSize].extCh=(XMLByte)i; | |
| tables.toTableSize++; | |
| } | |
| // sort by the Unicode code point | // sort by the Unicode code point |
| sort_ToTable(); | sort_ToTable(); |
| } | } |
| Line 178 static int sort_cmp_Trans_rec_intCh(cons | Line 210 static int sort_cmp_Trans_rec_intCh(cons |
| } | } |
| void Charset::sort_ToTable() { | void Charset::sort_ToTable() { |
| _qsort(tables.toTable, tables.toTableSize, sizeof(*tables.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"); | |
| //fwrite(tables.toTable, tables.toTableSize, sizeof(*tables.toTable), f); | |
| //fclose(f); | |
| } | } |
| static XMLByte xlatOneTo(const XMLCh toXlat, | // @todo: precache for spedup searching |
| const Charset::Tables& tables, | static XMLByte xlatOneTo(const XMLCh toXlat, 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 206 static XMLByte xlatOneTo(const XMLCh toX | Line 233 static XMLByte xlatOneTo(const XMLCh toX |
| return not_found; | return not_found; |
| } | } |
| String::C Charset::transcode(const String::C src, | String::C Charset::transcode(const String::C src, const Charset& source_charset, const Charset& dest_charset) { |
| const Charset& source_charset, | |
| const Charset& dest_charset) { | |
| if(!src.length) | if(!src.length) |
| return String::C("", 0); | return String::C("", 0); |
| Line 258 static const XMLByte gUTFBytes[0x100] = | Line 283 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, int& toFillLen, const Charset::Tables& tables) { |
| XMLByte *toFill, size_t& toFillLen, | |
| 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 314 static int transcodeToUTF8(const XMLByte | Line 337 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 348 of ocetes consumed. | Line 371 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, int& toFillLen, const Charset::Tables& tables) { |
| XMLByte* toFill, size_t& toFillLen, | |
| 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 362 static int transcodeFromUTF8(const XMLBy | Line 383 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 389 static int transcodeFromUTF8(const XMLBy | Line 410 static int transcodeFromUTF8(const XMLBy |
| break; | break; |
| default: | default: |
| throw Exception(0, | throw Exception(0, 0, "transcodeFromUTF8 error: wrong trailingBytes value(%d)", trailingBytes); // never |
| 0, | |
| "transcodeFromUTF8 error: wrong trailingBytes value(%d)", trailingBytes); | |
| } | } |
| tmpVal-=gUTFOffsets[trailingBytes]; | tmpVal-=gUTFOffsets[trailingBytes]; |
| Line 401 static int transcodeFromUTF8(const XMLBy | Line 420 static int transcodeFromUTF8(const XMLBy |
| if(!(tmpVal & 0xFFFF0000)) { | if(!(tmpVal & 0xFFFF0000)) { |
| if(XMLByte xlat=xlatOneTo(tmpVal, tables, 0)) | if(XMLByte xlat=xlatOneTo(tmpVal, tables, 0)) |
| *outPtr++=xlat; | *outPtr++=xlat; |
| else | else { |
| outPtr+=sprintf((char *)outPtr, "&#%d;", tmpVal); // &#decimal; | outPtr+=sprintf((char *)outPtr, "&#%u;", tmpVal); // &#decimal; |
| } else | } |
| throw Exception(0, | } else { |
| 0, | const XMLByte* recoverPtr=srcPtr-trailingBytes-1; |
| "transcodeFromUTF8 error: too big tmpVal(0x%08X)", tmpVal); | for(uint i=0; i<=trailingBytes; i++) |
| outPtr+=sprintf((char*)outPtr, "%%%02X", *recoverPtr++); | |
| } | |
| } | } |
| // Update the bytes eaten | // Update the bytes eaten |
| Line 426 of ocetes consumed. | Line 447 of ocetes consumed. |
| return 0; | return 0; |
| } | } |
| /// @todo not so memory-hungry with prescan | static bool need_escape(XMLByte c){ |
| return | |
| !( | |
| (c<=127) | |
| && ( | |
| pa_isalnum((unsigned char)c) | |
| || 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 storing 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; | |
| } | |
| size_t Charset::calc_escaped_length_UTF8(XMLByte* src, size_t src_length){ | |
| size_t dest_length=0; | |
| for(UTF8_string_iterator i(src, src_length); i.has_next(); ){ | |
| if(i.getCharSize()==1) | |
| dest_length+=!need_escape(i.getFirstByte())?1/*as-is*/:3/*%XX*/; | |
| else | |
| dest_length+=6; // %uXXXX | |
| } | |
| return dest_length; | |
| } | |
| size_t Charset::calc_escaped_length(const XMLByte* src, size_t src_length, const Charset::Tables& tables){ | |
| const XMLByte* src_end=src+src_length; | |
| XMLByte first_byte; | |
| XMLCh UTF8_char; | |
| size_t dest_length=0; | |
| while(uint char_size=readChar(src, src_end, first_byte, UTF8_char, tables)){ | |
| if(char_size==1) | |
| dest_length+=(!first_byte/*replacement char '?'*/ || !need_escape(first_byte))?1:3/*'%XX'*/; | |
| else | |
| dest_length+=6; // %uXXXX | |
| } | |
| return dest_length; | |
| } | |
| size_t Charset::calc_escaped_length(const String::C src, const Charset& source_charset){ | |
| if(!src.length) | |
| return 0; | |
| #ifdef PRECALCULATE_DEST_LENGTH | |
| if(source_charset.isUTF8()) | |
| return calc_escaped_length_UTF8((XMLByte *)src.str, src.length); | |
| else | |
| return calc_escaped_length((XMLByte *)src.str, src.length, source_charset.tables); | |
| #else | |
| return src_length*6; // enough for %uXXXX but too memory-hungry | |
| #endif | |
| } | |
| #define escape_char(dest_ptr, char_size, first_byte, UTF8_char) \ | |
| if(char_size==1) \ | |
| if(first_byte){ \ | |
| if(need_escape(first_byte)) \ | |
| dest_ptr=append_hex_8(dest_ptr, first_byte, "%"); /* %XX */ \ | |
| else \ | |
| *dest_ptr++=first_byte; /*as is*/ \ | |
| } else \ | |
| *dest_ptr++='?'; /* replacement char '?' */ \ | |
| else \ | |
| dest_ptr=append_hex_16(dest_ptr, UTF8_char, "%u"); /* %uXXXX */ | |
| size_t Charset::escape_UTF8(const XMLByte* src, size_t src_length, XMLByte* dest) { | |
| XMLByte* dest_ptr=dest; | |
| // loop until we either run out of input data | |
| for(UTF8_string_iterator i((XMLByte *)src, src_length); i.has_next(); ) | |
| escape_char(dest_ptr, i.getCharSize(), i.getFirstByte(), i.next()) | |
| return dest_ptr - dest; | |
| } | |
| size_t Charset::escape(const XMLByte* src, size_t src_length, XMLByte* dest, const Charset::Tables& tables) { | |
| const XMLByte* src_end=src+src_length; | |
| XMLByte* dest_ptr=dest; | |
| XMLByte first_byte; | |
| XMLCh UTF8_char; | |
| uint char_size; | |
| while(char_size=readChar(src, src_end, first_byte, UTF8_char, tables)) | |
| escape_char(dest_ptr, char_size, first_byte, UTF8_char) | |
| return dest_ptr - dest; | |
| } | |
| String::C Charset::escape(const String::C src, const Charset& source_charset){ | |
| if(!src.length) | |
| return String::C("", 0); | |
| size_t dest_calculated_length=calc_escaped_length(src, source_charset); | |
| XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_calculated_length+1/*terminator*/]; | |
| size_t dest_length; | |
| if(source_charset.isUTF8()) | |
| dest_length=escape_UTF8((XMLByte *)src.str, src.length, dest_body); | |
| else | |
| dest_length=escape((XMLByte *)src.str, src.length, dest_body, source_charset.tables); | |
| if(dest_length>dest_calculated_length) | |
| throw Exception(0, 0, "Charset::escape buffer overflow"); | |
| 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) { | |
| String::C dest=Charset::escape(String::C(src.cstr(), src.length()), 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); | |
| } | |
| inline bool need_json_escape(unsigned char c){ | |
| return strchr("\n\"\\/\t\r\b\f", c)!=0; | |
| } | |
| size_t Charset::calc_JSON_escaped_length_UTF8(XMLByte* src, size_t src_length){ | |
| size_t dest_length=0; | |
| for(UTF8_string_iterator i(src, src_length); i.has_next(); ){ | |
| if(i.getCharSize()==1){ | |
| XMLByte first_byte=i.getFirstByte(); | |
| dest_length+=need_json_escape(first_byte) ? 2 : (first_byte < 0x20 && first_byte /* 0 replacement char is '?' */) ? 6 : 1; | |
| } else | |
| dest_length+=6; // \uXXXX | |
| } | |
| return dest_length; | |
| } | |
| size_t Charset::calc_JSON_escaped_length(const XMLByte* src, size_t src_length, const Charset::Tables& tables){ | |
| const XMLByte* src_end=src+src_length; | |
| XMLByte first_byte; | |
| XMLCh UTF8_char; | |
| size_t dest_length=0; | |
| while(uint char_size=readChar(src, src_end, first_byte, UTF8_char, tables)){ | |
| if(char_size==1) | |
| dest_length+=need_json_escape(first_byte) ? 2 : (first_byte < 0x20 && first_byte /* 0 replacement char is '?' */) ? 6 : 1; | |
| else | |
| dest_length+=6; // \uXXXX | |
| } | |
| return dest_length; | |
| } | |
| size_t Charset::calc_JSON_escaped_length(const String::C src, const Charset& source_charset){ | |
| if(!src.length) | |
| return 0; | |
| #ifdef PRECALCULATE_DEST_LENGTH | |
| if(source_charset.isUTF8()) | |
| return calc_JSON_escaped_length_UTF8((XMLByte *)src.str, src.length); | |
| else | |
| return calc_JSON_escaped_length((XMLByte *)src.str, src.length, source_charset.tables); | |
| #else | |
| return src_length*6; // enough for \uXXXX but too memory-hungry | |
| #endif | |
| } | |
| #define escape_char_JSON(dest_ptr, char_size, first_byte, UTF8_char) \ | |
| if(char_size==1) \ | |
| switch(first_byte){ \ | |
| case '\n': *dest_ptr++='\\'; *dest_ptr++='n'; break; \ | |
| case '"' : *dest_ptr++='\\'; *dest_ptr++='"'; break; \ | |
| case '\\': *dest_ptr++='\\'; *dest_ptr++='\\'; break; \ | |
| case '/' : *dest_ptr++='\\'; *dest_ptr++='/'; break; \ | |
| case '\t': *dest_ptr++='\\'; *dest_ptr++='t'; break; \ | |
| case '\r': *dest_ptr++='\\'; *dest_ptr++='r'; break; \ | |
| case '\b': *dest_ptr++='\\'; *dest_ptr++='b'; break; \ | |
| case '\f': *dest_ptr++='\\'; *dest_ptr++='f'; break; \ | |
| case 0 : *dest_ptr++='?'; break; /*replacement char*/ \ | |
| default : if(first_byte < 0x20) dest_ptr=append_hex_16(dest_ptr, UTF8_char, "\\u"); \ | |
| else *dest_ptr++=first_byte; \ | |
| } \ | |
| else \ | |
| dest_ptr=append_hex_16(dest_ptr, UTF8_char, "\\u"); // \uXXXX | |
| size_t Charset::escape_JSON_UTF8(const XMLByte* src, size_t src_length, XMLByte* dest) { | |
| XMLByte* dest_ptr=dest; | |
| // loop until we either run out of input data | |
| for(UTF8_string_iterator i((XMLByte *)src, src_length); i.has_next(); ) | |
| escape_char_JSON(dest_ptr, i.getCharSize(), i.getFirstByte(), i.next()) | |
| return dest_ptr - dest; | |
| } | |
| size_t Charset::escape_JSON(const XMLByte* src, size_t src_length, XMLByte* dest, const Charset::Tables& tables) { | |
| const XMLByte* src_end=src+src_length; | |
| XMLByte* dest_ptr=dest; | |
| XMLByte first_byte; | |
| XMLCh UTF8_char; | |
| uint char_size; | |
| while(char_size=readChar(src, src_end, first_byte, UTF8_char, tables)) | |
| escape_char_JSON(dest_ptr, char_size, first_byte, UTF8_char) | |
| return dest_ptr - dest; | |
| } | |
| String::C Charset::escape_JSON(const String::C src, const Charset& source_charset){ | |
| if(!src.length) | |
| return String::C("", 0); | |
| size_t dest_calculated_length=calc_JSON_escaped_length(src, source_charset); | |
| XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_calculated_length+1/*terminator*/]; | |
| size_t dest_length; | |
| if(source_charset.isUTF8()) | |
| dest_length=escape_JSON_UTF8((XMLByte *)src.str, src.length, dest_body); | |
| else | |
| dest_length=escape_JSON((XMLByte *)src.str, src.length, dest_body, source_charset.tables); | |
| if(dest_length>dest_calculated_length) | |
| throw Exception(0, 0, "Charset::escape_JSON buffer overflow"); | |
| dest_body[dest_length]=0; // terminator | |
| return String::C((char*)dest_body, dest_length); | |
| } | |
| String::Body Charset::escape_JSON(const String::Body src, const Charset& source_charset) { | |
| String::C dest=Charset::escape_JSON(String::C(src.cstr(), src.length()), source_charset); | |
| return String::Body(dest.length ? dest.str:0); | |
| } | |
| String& Charset::escape_JSON(const String& src, const Charset& source_charset) { | |
| if(src.is_empty()) | |
| return *new String(); | |
| return *new String(escape_JSON((String::Body)src, source_charset), String::L_CLEAN); | |
| } | |
| const String::C Charset::transcodeToUTF8(const String::C src) const { | const String::C Charset::transcodeToUTF8(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, max utf8 seq len=6*/; | |
| #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-hungry | |
| #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 439 const String::C Charset::transcodeToUTF8 | Line 789 const String::C Charset::transcodeToUTF8 |
| (XMLByte *)src.str, src_length, | (XMLByte *)src.str, src_length, |
| dest_body, dest_length, | dest_body, dest_length, |
| tables)<0) | tables)<0) |
| throw(0, 0, | throw Exception(0, 0, "Charset::transcodeToUTF8 buffer overflow"); |
| 0, | |
| "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 467 static XMLCh change_case_UTF8(const XMLC | Line 816 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 515 static void store_UTF8(XMLCh src, XMLByt | Line 864 static void store_UTF8(XMLCh src, XMLByt |
| outPtr+= encodedBytes; | outPtr+= encodedBytes; |
| } | } |
| static void change_case_UTF8(XMLCh src, XMLByte*& outPtr, | static void change_case_UTF8(XMLCh src, XMLByte*& outPtr, const Charset::UTF8CaseTable& table) { |
| const Charset::UTF8CaseTable& table) { | |
| store_UTF8(change_case_UTF8(src, table), outPtr); | store_UTF8(change_case_UTF8(src, table), outPtr); |
| }; | } |
| void change_case_UTF8(const XMLByte* srcData, XMLByte* toFill, | |
| const Charset::UTF8CaseTable& table) { | void change_case_UTF8(const XMLByte* srcData, size_t srcLen, XMLByte* toFill, size_t toFillLen, const Charset::UTF8CaseTable& table) { |
| const XMLByte* srcPtr=srcData; | const XMLByte* srcPtr=srcData; |
| const XMLByte* srcEnd=srcData+srcLen; | |
| XMLByte* outPtr=toFill; | XMLByte* outPtr=toFill; |
| XMLByte* outEnd=toFill+toFillLen; | |
| // We now loop until we either run out of input data, or room to store | |
| while ((srcPtr < srcEnd) && (outPtr < outEnd)) { | |
| // Get the next leading byte out | |
| const XMLByte firstByte =* srcPtr; | |
| // Get the next leading byte out | if(firstByte<=127) { |
| while (const XMLByte firstByte = *srcPtr) { | |
| if(firstByte<= 127) { | |
| change_case_UTF8(firstByte, outPtr, table); | change_case_UTF8(firstByte, outPtr, table); |
| srcPtr++; | srcPtr++; |
| continue; | continue; |
| Line 547 void change_case_UTF8(const XMLByte* src | Line 900 void change_case_UTF8(const XMLByte* src |
| break; | break; |
| default: | default: |
| throw Exception(0, | throw Exception(0, 0, "change_case_UTF8 error: wrong trailingBytes value(%d)", trailingBytes); |
| 0, | |
| "change_case_UTF8 error: wrong trailingBytes value(%d)", trailingBytes); | |
| } | } |
| tmpVal-=gUTFOffsets[trailingBytes]; | tmpVal-=gUTFOffsets[trailingBytes]; |
| Line 559 void change_case_UTF8(const XMLByte* src | Line 910 void change_case_UTF8(const XMLByte* src |
| if(!(tmpVal & 0xFFFF0000)) | if(!(tmpVal & 0xFFFF0000)) |
| change_case_UTF8(tmpVal, outPtr, table); | change_case_UTF8(tmpVal, outPtr, table); |
| else | else |
| throw Exception(0, | throw Exception(0, 0, "change_case_UTF8 error: too big tmpVal(0x%08X)", tmpVal); |
| 0, | |
| "change_case_UTF8 error: too big tmpVal(0x%08X)", tmpVal); | |
| } | } |
| if(srcPtr!=outPtr) | if(srcPtr!=outPtr) |
| throw Exception(0, | throw Exception(0, 0, "change_case_UTF8 error: end pointers do not match"); |
| 0, | |
| "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; | |
| for(UTF8_string_iterator i((XMLByte *)src.str, src_length); i.has_next(); ){ | |
| dest_length += ( i.next() & 0xFFFF0000 ) | |
| ? 3*i.getCharSize() // %XX for each byte | |
| : ( xlatOneTo(i.next(), tables, 0) != 0 ) | |
| ? 1 // can convert it to a single char | |
| : 3+getDecNumLength( i.next() ); // print char as &#XX;, &#XXX;, &#XXXX; or &#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 583 const String::C Charset::transcodeFromUT | Line 953 const String::C Charset::transcodeFromUT |
| (XMLByte *)src.str, src_length, | (XMLByte *)src.str, src_length, |
| dest_body, dest_length, | dest_body, dest_length, |
| tables)<0) | tables)<0) |
| throw(0, 0, | throw Exception(0, 0, "Charset::transcodeFromUTF8 buffer overflow"); |
| 0, | |
| "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 614 const String::C Charset::transcodeToChar | Line 982 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]; |
| static xmlCharEncodingHandler xml_encoding_handlers[MAX_CHARSETS]; | |
| #ifdef PA_PATCHED_LIBXML_BACKWARD | |
| #define declareXml256ioFuncs(i) \ | |
| static int xml256CharEncodingInputFunc##i( \ | |
| unsigned char *out, int *outlen, \ | |
| const unsigned char *in, int *inlen, void*) { \ | |
| return transcodeToUTF8( \ | |
| in, *inlen, \ | |
| out, *outlen, \ | |
| *tables[i]); \ | |
| } \ | |
| static int xml256CharEncodingOutputFunc##i( \ | |
| unsigned char *out, int *outlen, \ | |
| const unsigned char *in, int *inlen, void*) { \ | |
| return transcodeFromUTF8( \ | |
| in, *inlen, \ | |
| out, *outlen, \ | |
| *tables[i]); \ | |
| } | |
| #else | |
| #define declareXml256ioFuncs(i) \ | #define declareXml256ioFuncs(i) \ |
| static int xml256CharEncodingInputFunc##i( \ | static int xml256CharEncodingInputFunc##i( \ |
| 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]); \ |
| } | } |
| #endif | |
| declareXml256ioFuncs(0) declareXml256ioFuncs(1) | declareXml256ioFuncs(0) declareXml256ioFuncs(1) |
| declareXml256ioFuncs(2) declareXml256ioFuncs(3) | declareXml256ioFuncs(2) declareXml256ioFuncs(3) |
| declareXml256ioFuncs(4) declareXml256ioFuncs(5) | declareXml256ioFuncs(4) declareXml256ioFuncs(5) |
| Line 660 static size_t handlers_count=0; | Line 1061 static size_t handlers_count=0; |
| void Charset::addEncoding(char *name_cstr) { | void Charset::addEncoding(char *name_cstr) { |
| if(handlers_count==MAX_CHARSETS) | if(handlers_count==MAX_CHARSETS) |
| throw Exception(0, | throw Exception(0, 0, "already allocated %d handlers, no space for new encoding '%s'", MAX_CHARSETS, name_cstr); |
| 0, | |
| "already allocated %d handlers, no space for new encoding '%s'", | |
| MAX_CHARSETS, name_cstr); | |
| xmlCharEncodingHandler* handler=new(PointerFreeGC) xmlCharEncodingHandler; | xmlCharEncodingHandler* handler=&xml_encoding_handlers[handlers_count]; |
| { | { |
| handler->name=name_cstr; | handler->name=name_cstr; |
| handler->input=inputFuncs[handlers_count]; | handler->input=inputFuncs[handlers_count]; |
| Line 685 void Charset::initTranscoder(const Strin | Line 1083 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), "unsupported encoding"); |
| new String(NAME, String::L_TAINTED), | |
| "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); |
| int inlen=strlen((const char*)s); | int inlen=strlen((const char*)s); |
| int outlen=inlen; // max | int outlen=inlen*6/*strlen("ÿ")*/; // max |
| #ifndef NDEBUG | #ifndef NDEBUG |
| int saved_outlen=outlen; | int saved_outlen=outlen; |
| #endif | #endif |
| Line 706 String::C Charset::transcode_cstr(xmlCha | Line 1102 String::C Charset::transcode_cstr(xmlCha |
| if(xmlCharEncodingOutputFunc output=transcoder(FNAME).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 |
| #ifdef PA_PATCHED_LIBXML_BACKWARD | |
| ,0 | |
| #endif | |
| ); | |
| } else { | } else { |
| memcpy(out, s, outlen=inlen); | memcpy(out, s, outlen=inlen); |
| error=0; | error=0; |
| } | } |
| if(error<0) | if(error<0) |
| throw Exception(0, | throw Exception(0, 0, "transcode_cstr failed (%d)", error); |
| 0, | |
| "transcode_cstr failed (%d)", error); | |
| 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 741 xmlChar* Charset::transcode_buf2xchar(co | Line 1131 xmlChar* Charset::transcode_buf2xchar(co |
| int saved_outlen; | int saved_outlen; |
| #endif | #endif |
| if(xmlCharEncodingInputFunc input=transcoder(FNAME).input) { | if(xmlCharEncodingInputFunc input=transcoder(FNAME).input) { |
| outlen=buf_size*6/*max*/; | outlen=buf_size*6/*max UTF8 bytes per char*/; |
| #ifndef NDEBUG | #ifndef NDEBUG |
| saved_outlen=outlen; | saved_outlen=outlen; |
| #endif | #endif |
| out=(xmlChar*)xmlMalloc(outlen+1); | out=(xmlChar*)xmlMalloc(outlen+1); |
| error=input( | error=input( |
| out, &outlen, | out, &outlen, |
| (const unsigned char*)buf, (int*)&buf_size); | (const unsigned char*)buf, (int*)&buf_size |
| #ifdef PA_PATCHED_LIBXML_BACKWARD | |
| ,0 | |
| #endif | |
| ); | |
| } else { | } else { |
| outlen=buf_size; | outlen=buf_size; |
| #ifndef NDEBUG | #ifndef NDEBUG |
| Line 760 xmlChar* Charset::transcode_buf2xchar(co | Line 1154 xmlChar* Charset::transcode_buf2xchar(co |
| } | } |
| if(error<0) | if(error<0) |
| throw Exception(0, | throw Exception(0, 0, "transcode_buf failed (%d)", error); |
| 0, | |
| "transcode_buf failed (%d)", error); | |
| 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 |
| String::Body Charset::transcode(const String::Body src, | String::Body Charset::transcode(const String::Body src, const Charset& source_transcoder, const Charset& dest_transcoder) { |
| const Charset& source_transcoder, | return String::Body(Charset::transcode(String::C(src.cstr(), src.length()), source_transcoder, dest_transcoder)); |
| const Charset& dest_transcoder) { | |
| const char *src_ptr=src.cstr(); | |
| size_t src_size=strlen(src_ptr); | |
| String::C dest=Charset::transcode(String::C(src_ptr, src_size), | |
| source_transcoder, | |
| dest_transcoder); | |
| return String::Body(dest.str, dest.length); | |
| } | } |
| String& Charset::transcode(const String& src, | String& Charset::transcode(const String& src, const Charset& source_transcoder, const Charset& dest_transcoder) { |
| const Charset& source_transcoder, | if(src.is_empty()) |
| const Charset& dest_transcoder) { | return *new String(); |
| if(!src.length()) | |
| return *new String("", 0, false); | |
| 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); |
| } | } |
| void Charset::transcode(ArrayString& src, | void Charset::transcode(ArrayString& src, const Charset& source_transcoder, const Charset& dest_transcoder) { |
| const Charset& source_transcoder, | |
| const Charset& dest_transcoder) { | |
| for(size_t i=0; i<src.count(); i++) | for(size_t i=0; i<src.count(); i++) |
| src.put(i, &transcode(*src[i], source_transcoder, dest_transcoder)); | src.put(i, &transcode(*src[i], source_transcoder, dest_transcoder)); |
| } | } |
| Line 818 struct Transcode_pair_info { | Line 1192 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, Transcode_pair_info* info) { |
| String::Body& avalue, | avalue=Charset::transcode(avalue, *info->source_transcoder, *info->dest_transcoder); |
| Transcode_pair_info* info) { | } |
| avalue=Charset::transcode(avalue, | |
| *info->source_transcoder, | void Charset::transcode(HashStringString& src, const Charset& source_transcoder, const Charset& dest_transcoder) { |
| *info->dest_transcoder); | |
| } | |
| void Charset::transcode(HashStringString& src, | |
| const Charset& source_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 conversion 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; | |
| } | |
| unsigned int lengthUTF8Char(const XMLByte c){ | |
| return gUTFBytes[c]+1; | |
| } | |
| const char *fixUTF8(const char *src){ | |
| if(src && *src){ | |
| size_t length=strlen(src); | |
| int error_offset; | |
| if(pa_pcre_valid_utf((unsigned char *)src, length, &error_offset)){ | |
| char *result=(char *)pa_malloc_atomic(length+1); | |
| char *dst=result; | |
| do { | |
| if(error_offset){ | |
| memcpy(dst, src, error_offset); | |
| dst+=error_offset; | |
| src+=error_offset; | |
| length-=error_offset; | |
| } | |
| *dst++='?'; | |
| src++; | |
| length--; | |
| } while (length && pa_pcre_valid_utf((unsigned char *)src, length, &error_offset)); | |
| if(length){ | |
| strcpy(dst, src); | |
| } else { | |
| *dst='\0'; | |
| } | |
| return result; | |
| } | |
| } | |
| return src; | |
| } | |
| bool UTF8_string_iterator::has_next(){ | |
| fcharSize=readUTF8Char(fsrcPtr, fsrcEnd, ffirstByte, fUTF8Char); | |
| return fcharSize!=0; | |
| } | } |