|
|
| version 1.55, 2006/04/09 13:38:47 | version 1.90, 2012/03/16 09:24:13 |
|---|---|
| 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) 2001-2012 Art. Lebedev Studio (http://www.artlebedev.com) |
| Author: Alexander Petrosyan<paf@design.ru>(http://paf.design.ru) | Author: Alexander Petrosyan<paf@design.ru>(http://paf.design.ru) |
| */ | */ |
| static const char * const IDENT_CHARSET_C="$Date$"; | |
| #include "pa_charset.h" | #include "pa_charset.h" |
| #include "pa_charsets.h" | #include "pa_charsets.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 | //#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 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 150 void Charset::load_definition(Request_ch | Line 152 void Charset::load_definition(Request_ch |
| 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 169 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 187 void Charset::sort_ToTable() { | Line 198 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 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 316 static int transcodeToUTF8(const XMLByte | Line 328 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 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 364 static int transcodeFromUTF8(const XMLBy | Line 376 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 442 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+=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 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) | |
| dest_length+=need_json_escape(i.getFirstByte()) ? 2 : 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+=(!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 | |
| 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 : *dest_ptr++=first_byte; \ | |
| } \ | |
| else \ | |
| dest_ptr+=sprintf((char*)dest_ptr, "\\u%04X", UTF8_char); // \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 447 const String::C Charset::transcodeToUTF8 | Line 786 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 811 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 876 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 920 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; | |
| 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 597 const String::C Charset::transcodeFromUT | Line 960 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 988 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 1006 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 1026 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 1091 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; |
| Line 760 String::C Charset::transcode_cstr(const | Line 1131 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 802 xmlChar* Charset::transcode_buf2xchar(co | Line 1173 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 819 String::Body Charset::transcode(const St | Line 1189 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 850 struct Transcode_pair_info { | Line 1218 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; | |
| } |