|
|
| version 1.60, 2008/07/15 12:53:10 | version 1.87, 2010/09/21 21:45:20 |
|---|---|
| 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 15 static const char * const IDENT_CHARSET_ | Line 15 static const char * const IDENT_CHARSET_ |
| #endif | #endif |
| //#define PA_PATCHED_LIBXML_BACKWARD | //#define PA_PATCHED_LIBXML_BACKWARD |
| // reduce memory usage by pre-calculation utf-8 string length | |
| #define PRECALCULATE_DEST_LENGTH | #define PRECALCULATE_DEST_LENGTH |
| // globals | // globals |
| Line 86 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 100 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 170 void Charset::load_definition(Request_ch | Line 171 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 192 void Charset::sort_ToTable() { | Line 202 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) { |
| 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 262 static const XMLByte gUTFBytes[0x100] = | Line 272 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 352 of ocetes consumed. | Line 362 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 432 of ocetes consumed. | Line 442 of ocetes consumed. |
| return 0; | return 0; |
| } | } |
| static bool is_escaped(char c){ | static bool need_escape(XMLByte c){ |
| return | return |
| !(c<=127 | !( |
| && ( | (c<=127) |
| ((c>='0') && (c<='9')) | && ( |
| || ((c>='A') && (c<='Z')) | ((c>='0') && (c<='9')) |
| || ((c>='a') && (c<='z')) | || ((c>='A') && (c<='Z')) |
| || strchr("*@-_+./", c)!=0 | || ((c>='a') && (c<='z')) |
| )); | || strchr("*@-_+./", c)!=0 |
| ) | |
| ); | |
| } | } |
| // read one utf8 character, return number of bytes needed for store it | // read one UTF8 char and return length of this char (in bytes) |
| static unsigned int readChar(const XMLByte*& srcPtr, const XMLByte*& srcEnd, XMLByte& firstByte, XMLCh& UTF8Char){ | static unsigned int readUTF8Char(const XMLByte*& srcPtr, const XMLByte* srcEnd, XMLByte& firstByte, XMLCh& UTF8Char){ |
| if(!srcPtr || !*srcPtr || srcPtr>=srcEnd) | if(!srcPtr || !*srcPtr || srcPtr>=srcEnd) |
| return 0; | return 0; |
| Line 478 static unsigned int readChar(const XMLBy | Line 490 static unsigned int readChar(const XMLBy |
| return trailingBytes+1; | return trailingBytes+1; |
| } | } |
| // read char, return number of bytes needed for store it as UTF8 | // skip UTF8 char and return length of this char (in bytes) |
| static unsigned int readChar(const XMLByte*& srcPtr, const XMLByte*& srcEnd, XMLByte& firstByte, XMLCh& UTF8Char, const Charset::Tables& tables){ | 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) | if(!srcPtr || !*srcPtr || srcPtr>=srcEnd) |
| return 0; | return 0; |
| Line 504 static unsigned int readChar(const XMLBy | Line 527 static unsigned int readChar(const XMLBy |
| return 1; | return 1; |
| } | } |
| static int escape(const XMLByte* srcData, size_t& srcLen, | size_t Charset::calc_escaped_length_UTF8(XMLByte* src, size_t src_length){ |
| XMLByte* toFill, size_t& toFillLen) { | size_t dest_length=0; |
| const XMLByte* srcPtr=srcData; | |
| const XMLByte* srcEnd=srcData+srcLen; | for(UTF8_string_iterator i(src, src_length); i.has_next(); ){ |
| XMLByte* outPtr=toFill; | if(i.getCharSize()==1) |
| XMLByte* outEnd=toFill+toFillLen; | dest_length+=!need_escape(i.getFirstByte())?1/*as-is*/:3/*%XX*/; |
| XMLByte firstByte; | else |
| XMLCh UTF8Char; | dest_length+=6; // %uXXXX |
| uint charSize; | |
| // loop until we either run out of input data, or room to store | |
| while((outPtr < outEnd) && (charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char))){ | |
| if(charSize==1){ | |
| if(is_escaped(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; | return dest_length; |
| } | } |
| static int escape(const XMLByte* srcData, size_t& srcLen, | size_t Charset::calc_escaped_length(const XMLByte* src, size_t src_length, const Charset::Tables& tables){ |
| XMLByte *toFill, size_t& toFillLen, | const XMLByte* src_end=src+src_length; |
| const Charset::Tables& tables) { | XMLByte first_byte; |
| const XMLByte* srcPtr=srcData; | XMLCh UTF8_char; |
| const XMLByte* srcEnd=srcData+srcLen; | size_t dest_length=0; |
| XMLByte* outPtr=toFill; | |
| //XMLByte* outEnd=toFill+toFillLen; | while(uint char_size=readChar(src, src_end, first_byte, UTF8_char, tables)){ |
| XMLByte firstByte; | if(char_size==1) |
| XMLCh UTF8Char; | dest_length+=(!first_byte/*replacement char '?'*/ || !need_escape(first_byte))?1:3/*'%XX'*/; |
| uint charSize; | else |
| dest_length+=6; // %uXXXX | |
| while(charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, tables)){ | |
| if(charSize==1){ | |
| if(firstByte){ | |
| if(is_escaped(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 | return dest_length; |
| srcLen = srcPtr - srcData; | } |
| // Return the characters read | size_t Charset::calc_escaped_length(const String::C src, const Charset& source_charset){ |
| toFillLen = outPtr - toFill; | 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+=sprintf((char*)dest_ptr, "%%%02X", first_byte); /* %XX */ \ | |
| else \ | |
| *dest_ptr++=first_byte; /*as is*/ \ | |
| } else \ | |
| *dest_ptr++='?'; /* replacement char '?' */ \ | |
| else \ | |
| dest_ptr+=sprintf((char*)dest_ptr, "%%u%04X", UTF8_char); // %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 0; | 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){ | String::C Charset::escape(const String::C src, const Charset& source_charset){ |
| size_t src_length=src.length; | if(!src.length) |
| if(!src_length) | |
| return String::C("", 0); | return String::C("", 0); |
| #ifdef PRECALCULATE_DEST_LENGTH | 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; | 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()){ | for(UTF8_string_iterator i(src, src_length); i.has_next(); ){ |
| while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char)){ | if(i.getCharSize()==1) |
| if(charSize==1) | dest_length+=need_json_escape(i.getFirstByte()) ? 2 : 1; |
| dest_length+=!is_escaped(firstByte)?1:3/*%XX*/; | else |
| else | dest_length+=6; // \uXXXX |
| dest_length+=6; // '%uXXXX' | } |
| } | |
| } else { | return dest_length; |
| while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, source_charset.tables)){ | } |
| if(charSize==1) | |
| dest_length+=(!firstByte/*replacement char '?'*/ || !is_escaped(firstByte))?1:3/*'%XX'*/; | size_t Charset::calc_JSON_escaped_length(const XMLByte* src, size_t src_length, const Charset::Tables& tables){ |
| else | const XMLByte* src_end=src+src_length; |
| dest_length+=6; // '%uXXXX' | 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_json_escape(first_byte))? 1 : 2; | |
| 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 | #else |
| size_t dest_length=src_length*6; // enough for %uXXXX but too memory-hungry | return src_length*6; // enough for \uXXXX but too memory-hungry |
| #endif | #endif |
| } | |
| //throw Exception(0,0,"%u",dest_length); | #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 : *dest_ptr++=first_byte; \ | |
| } \ | |
| else \ | |
| dest_ptr+=sprintf((char*)dest_ptr, "\\u%04X", UTF8_char); // \uXXXX | |
| #ifndef NDEBUG | |
| size_t saved_dest_length=dest_length; | |
| #endif | |
| XMLByte *dest_body=new(PointerFreeGC) XMLByte[dest_length+1/*for terminator*/]; | |
| int status; | size_t Charset::escape_JSON_UTF8(const XMLByte* src, size_t src_length, XMLByte* dest) { |
| if(source_charset.isUTF8()){ | XMLByte* dest_ptr=dest; |
| status=::escape((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) | // loop until we either run out of input data |
| throw Exception(0, | for(UTF8_string_iterator i((XMLByte *)src, src_length); i.has_next(); ) |
| 0, | escape_char_JSON(dest_ptr, i.getCharSize(), i.getFirstByte(), i.next()) |
| "Charset::escapeString buffer overflow"); | |
| 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"); | |
| assert(dest_length<=saved_dest_length); | |
| dest_body[dest_length]=0; // terminator | dest_body[dest_length]=0; // terminator |
| return String::C((char*)dest_body, dest_length); | 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; |
| #ifdef PRECALCULATE_DEST_LENGTH | #ifdef PRECALCULATE_DEST_LENGTH |
| size_t dest_length=0; | int dest_length=0; |
| const XMLByte* srcPtr=(XMLByte*)src.str; | const XMLByte* srcPtr=(XMLByte*)src.str; |
| const XMLByte* srcEnd=srcPtr+src_length; | const XMLByte* srcEnd=srcPtr+src_length; |
| XMLByte firstByte; | XMLByte firstByte; |
| XMLCh UTF8Char; | XMLCh UTF8Char; |
| while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, tables)) | while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char, tables)) |
| dest_length+=charSize; | dest_length+=charSize; |
| #else | #else |
| size_t dest_length=src_length*6; // so that surly enough (max utf8 seq len=6) but too memory-hyngry | int dest_length=src_length*6; // so that surly enough (max utf8 seq len=6) but too memory-hungry |
| #endif | #endif |
| //throw Exception(0,0,"%u",dest_length); | |
| #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 660 const String::C Charset::transcodeToUTF8 | Line 794 const String::C Charset::transcodeToUTF8 |
| } | } |
| 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 800 static size_t getDecNumLength(XMLCh UTF8 | Line 934 static size_t getDecNumLength(XMLCh UTF8 |
| } | } |
| 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; |
| #ifdef PRECALCULATE_DEST_LENGTH | #ifdef PRECALCULATE_DEST_LENGTH |
| size_t dest_length=0; | int dest_length=0; |
| const XMLByte* srcPtr=(XMLByte*)src.str; | for(UTF8_string_iterator i((XMLByte *)src.str, src_length); i.has_next(); ){ |
| const XMLByte* srcEnd=srcPtr+src_length; | if(i.getCharSize()==1) |
| XMLByte firstByte; | |
| XMLCh UTF8Char; | |
| while(uint charSize=readChar(srcPtr, srcEnd, firstByte, UTF8Char)){ | |
| if(charSize==1) | |
| dest_length++; | dest_length++; |
| else | else |
| dest_length+=(UTF8Char & 0xFFFF0000) | dest_length+=(i.next() & 0xFFFF0000) |
| ?charSize*3 // '%XX' for each byte | ?i.getCharSize()*3 // '%XX' for each byte |
| :(xlatOneTo(UTF8Char, tables, 0)!=0) | :(xlatOneTo(i.next(), tables, 0)!=0) |
| ?1 // can convert it to single char | ?1 // can convert it to single char |
| :getDecNumLength(UTF8Char)+3; // &#XX; - &#XXXXX; | :getDecNumLength(i.next())+3; // &#XX; - &#XXXXX; |
| } | } |
| #else | #else |
| // so that surly enough, "&#XXX;" has max ratio (huh? 8 bytes needed for '&#XXXXX;') | // so that surly enough, "&#XXX;" has max ratio (huh? 8 bytes needed for '&#XXXXX;') |
| size_t dest_length=src_length*6; | int dest_length=src_length*6; |
| #endif | #endif |
| //throw Exception(0,0,"%u",dest_length); | |
| #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 845 const String::C Charset::transcodeFromUT | Line 972 const String::C Charset::transcodeFromUT |
| /// 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 884 static const Charset::Tables* tables[MAX | Line 1011 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 904 static const Charset::Tables* tables[MAX | Line 1031 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 1009 String::C Charset::transcode_cstr(const | Line 1136 String::C Charset::transcode_cstr(const |
| } | } |
| const String& Charset::transcode(const xmlChar* s) { | const String& Charset::transcode(const xmlChar* 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 1051 xmlChar* Charset::transcode_buf2xchar(co | Line 1178 xmlChar* Charset::transcode_buf2xchar(co |
| assert(outlen<=saved_outlen); out[outlen]=0; | assert(outlen<=saved_outlen); out[outlen]=0; |
| return out; | return out; |
| } | } |
| xmlChar* Charset::transcode(const String& s) { | |
| const char* cstr=s.cstr(String::L_UNSPECIFIED); | |
| return transcode_buf2xchar(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()); | |
| } | } |
| xmlChar* Charset::transcode(const String::Body s) { | |
| const char* cstr=s.cstr(); | |
| return transcode_buf2xchar(cstr, s.length()); | xmlChar* Charset::transcode(const String::Body s) { |
| return transcode_buf2xchar(s.cstr(), s.length()); | |
| } | } |
| #endif | #endif |
| Line 1068 String::Body Charset::transcode(const St | Line 1194 String::Body Charset::transcode(const St |
| const Charset& dest_transcoder) { | const Charset& dest_transcoder) { |
| const char *src_ptr=src.cstr(); | const char *src_ptr=src.cstr(); |
| size_t src_size=strlen(src_ptr); | size_t src_size=src.length(); |
| 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 1099 struct Transcode_pair_info { | Line 1223 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*>(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; | |
| } | |
| unsigned int lengthUTF8Char(const XMLByte c){ | |
| return gUTFBytes[c]+1; | |
| } | |
| bool UTF8_string_iterator::has_next(){ | |
| fcharSize=readUTF8Char(fsrcPtr, fsrcEnd, ffirstByte, fUTF8Char); | |
| return fcharSize!=0; | |
| } |