--- parser3/src/main/pa_charset.C 2001/12/17 19:22:21 1.5 +++ parser3/src/main/pa_charset.C 2002/03/27 15:30:36 1.23 @@ -1,22 +1,16 @@ /** @file Parser: Charset connection implementation. - Copyright(c) 2001 ArtLebedev Group(http://www.artlebedev.com) + Copyright(c) 2001, 2002 ArtLebedev Group (http://www.artlebedev.com) Author: Alexander Petrosyan(http://paf.design.ru) - $Id: pa_charset.C,v 1.5 2001/12/17 19:22:21 paf Exp $ + $Id: pa_charset.C,v 1.23 2002/03/27 15:30:36 paf Exp $ */ #include "pa_charset.h" -//#include "pa_exception.h" -//#include "pa_common.h" -//#include "pa_threads.h" #ifdef XML -# include -# include -# include -# include +#include "libxml/encoding.h" #endif // globals @@ -77,76 +71,19 @@ static void element2case(unsigned char f fcc_table[from]=to; fcc_table[to]=from; } -#ifdef XML -template class ENameMapFor2 : public ENameMap { -public: - // ----------------------------------------------------------------------- - // Constructors and Destructor - // ----------------------------------------------------------------------- - ENameMapFor2( - const XMLCh* const encodingName - , const XMLCh* const fromTable - , const XMLTransService::TransRec* const toTable - , const unsigned int toTableSize - ): ENameMap(encodingName), - ffromTable(fromTable), - ftoTable(toTable), - ftoTableSize(toTableSize) {} - - // ----------------------------------------------------------------------- - // Implementation of virtual factory method - // ----------------------------------------------------------------------- - virtual XMLTranscoder* makeNew(const unsigned int blockSize) const { - return new TType( - getKey(), - blockSize, - ffromTable, - ftoTable, ftoTableSize); - } -private: - const XMLCh* const ffromTable; - const XMLTransService::TransRec* const ftoTable; - const unsigned int ftoTableSize; - -private: - // ----------------------------------------------------------------------- - // Unimplemented constructors and operators - // ----------------------------------------------------------------------- - ENameMapFor2(); - ENameMapFor2(const ENameMapFor2&); - void operator=(const ENameMapFor2&); -}; - -class XML256TableTranscoder2 : public XML256TableTranscoder { -public : - XML256TableTranscoder2( - const XMLCh* const encodingName - , const unsigned int blockSize - , const XMLCh* const fromTable - , const XMLTransService::TransRec* const toTable - , const unsigned int toTableSize - ) : XML256TableTranscoder(encodingName, blockSize, fromTable, toTable, toTableSize) {} - -private: - XML256TableTranscoder2(); - XML256TableTranscoder2(const XML256TableTranscoder2&); - void operator=(const XML256TableTranscoder2&); -}; -#endif - // methods extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c -Charset::Charset(Pool& apool, const String& aname, const String *file_spec): Pooled(apool), - fname(apool) { - // fname - char *name_cstr=(char *)malloc(aname.size()+1); - memcpy(name_cstr, aname.cstr(String::UL_AS_IS), aname.size()+1); - fname.APPEND_CLEAN(name_cstr, aname.size(), 0, 0); +Charset::Charset(Pool& apool, const String& aname, const String *request_file_spec): Pooled(apool), + fname(aname) { + + char *name_cstr=fname.cstr(); + for(char *c=name_cstr; *c; c++) + *c = toupper(*c); - if(file_spec) { + if(request_file_spec) { fisUTF8=false; - loadDefinition(*file_spec); + loadDefinition(*request_file_spec); #ifdef XML addEncoding(name_cstr); #endif @@ -163,11 +100,11 @@ Charset::Charset(Pool& apool, const Stri Charset::~Charset() { #ifdef XML - delete transcoder; + // not deleting transcoder, that's not our business #endif } -void Charset::loadDefinition(const String& file_spec) { +void Charset::loadDefinition(const String& request_file_spec) { // pcre_tables // lowcase, flipcase, bits digit+word+whitespace, masks @@ -178,16 +115,16 @@ void Charset::loadDefinition(const Strin cstr2ctypes(pcre_tables,(const unsigned char *)"*+?{^.$|()[", ctype_meta); // charset - memset(fromTable, 0, sizeof(fromTable)); - toTable=(Charset_TransRec *)calloc(sizeof(Charset_TransRec)*MAX_CHARSET_UNI_CODES); - toTableSize=0; + memset(tables.fromTable, 0, sizeof(tables.fromTable)); + tables.toTable=(Charset_TransRec *)calloc(sizeof(Charset_TransRec)*MAX_CHARSET_UNI_CODES); + tables.toTableSize=0; // strangly vital - toTable[toTableSize].intCh=0; - toTable[toTableSize].extCh=(XMLByte)0; - toTableSize++; + tables.toTable[tables.toTableSize].intCh=0; + tables.toTable[tables.toTableSize].extCh=(XMLByte)0; + tables.toTableSize++; // loading text - char *data=file_read_text(pool(), file_spec); + char *data=file_read_text(pool(), request_file_spec); // ignore header getrow(&data); @@ -215,20 +152,20 @@ void Charset::loadDefinition(const Strin case 7: case 8: // charset - if(toTableSize>MAX_CHARSET_UNI_CODES) - throw Exception(0, 0, - &file_spec, + if(tables.toTableSize>MAX_CHARSET_UNI_CODES) + throw Exception("parser.runtime", + &request_file_spec, "charset must contain not more then %d unicode values", MAX_CHARSET_UNI_CODES); XMLCh unicode=(XMLCh)to_wchar_code(cell); if(!unicode && column==7/*unicode1 column*/) unicode=(XMLCh)c; if(unicode) { - if(!fromTable[c]) - fromTable[c]=unicode; - toTable[toTableSize].intCh=unicode; - toTable[toTableSize].extCh=(XMLByte)c; - toTableSize++; + if(!tables.fromTable[c]) + tables.fromTable[c]=unicode; + tables.toTable[tables.toTableSize].intCh=unicode; + tables.toTable[tables.toTableSize].extCh=(XMLByte)c; + tables.toTableSize++; } break; } @@ -239,36 +176,6 @@ void Charset::loadDefinition(const Strin sort_ToTable(); } -#ifdef XML -void Charset::addEncoding(const char *name_cstr) { - // addEncoding - XalanDOMString sencoding(name_cstr); - const XMLCh* const auto_encoding_cstr=sencoding.c_str(); - int size=sizeof(XMLCh)*(sencoding.size()+1); - XMLCh* pool_encoding_cstr=(XMLCh*)malloc(size); - memcpy(pool_encoding_cstr, auto_encoding_cstr, size); - XMLString::upperCase(pool_encoding_cstr); - - XMLPlatformUtils::fgTransService->addEncoding( - pool_encoding_cstr, - new ENameMapFor2( - pool_encoding_cstr - , fromTable - , toTable - , toTableSize - )); -} - -void Charset::initTranscoder(const String *source, const char *name_cstr) { - XMLTransService::Codes resValue; - transcoder=XMLPlatformUtils::fgTransService->makeNewTranscoderFor(name_cstr, resValue, 60); - if(!transcoder) - throw Exception(0, 0, - source, - "unsupported encoding"); -} -#endif - static int sort_cmp_Trans_rec_intCh(const void *a, const void *b) { return static_cast(a)->intCh- @@ -276,16 +183,17 @@ static int sort_cmp_Trans_rec_intCh(cons } void Charset::sort_ToTable() { - _qsort(toTable, toTableSize, sizeof(*toTable), + _qsort(tables.toTable, tables.toTableSize, sizeof(*tables.toTable), sort_cmp_Trans_rec_intCh); //FILE *f=fopen("c:\\temp\\a", "wb"); - //fwrite(toTable, toTableSize, sizeof(*toTable), f); + //fwrite(tables.toTable, tables.toTableSize, sizeof(*tables.toTable), f); //fclose(f); } -XMLByte Charset::xlatOneTo(const XMLCh toXlat) const { +static XMLByte xlatOneTo(const XMLCh toXlat, + const Charset::Tables& tables) { unsigned int lowOfs = 0; - unsigned int hiOfs = toTableSize - 1; + unsigned int hiOfs = tables.toTableSize - 1; XMLByte curByte = 0; do { // Calc the mid point of the low and high offset. @@ -294,12 +202,12 @@ XMLByte Charset::xlatOneTo(const XMLCh t // If our test char is greater than the mid point char, then // we move up to the upper half. Else we move to the lower // half. If its equal, then its our guy. - if(toXlat>toTable[midOfs].intCh) + if(toXlat>tables.toTable[midOfs].intCh) lowOfs = midOfs; - else if(toXlat outEnd) + break; // We can do it, so update the source index srcPtr++; @@ -448,22 +354,25 @@ void Charset::transcodeToUTF8(Pool& pool outPtr+= encodedBytes; } - // return - adest_body=dest_body; - adest_content_length=outPtr-dest_body; + // Update the bytes eaten + srcLen = srcPtr - srcData; + + // Return the characters read + toFillLen = outPtr - toFill; + + return srcPtr==srcEnd?(int)toFillLen:-1; } -void Charset::transcodeFromUTF8(Pool& pool, - const void *source_body, size_t source_content_length, - const void *& adest_body, size_t& adest_content_length) const { - size_t dest_content_length=0; - XMLByte *dest_body=(XMLByte*)pool.malloc(source_content_length/*surly enough*/); - - const XMLByte* srcPtr=(const XMLByte*)source_body; - const XMLByte* srcEnd=(const XMLByte*)source_body+source_content_length; - XMLByte* outPtr=dest_body; +static size_t transcodeFromUTF8( + const XMLByte *srcData, size_t& srcLen, + XMLByte* toFill, size_t& toFillLen, + const Charset::Tables& tables) { + const XMLByte* srcPtr=srcData; + const XMLByte* srcEnd=srcData+srcLen; + XMLByte* outPtr=toFill; + XMLByte* outEnd=toFill+toFillLen; - // We now loop until we either run out of input data - while(srcPtrtranscodeTo( - s.c_str(), s.length(), - dest, len, - charsEaten, - XMLTranscoder::UnRep_RepChar //UnRep_Throw - ); - dest[size]=0; - error=false; + for(XMLByte* outPtr=dest_body; srcPtrname=name_cstr; + handler->input=xml256CharEncodingInputFunc; handler->inputInfo=&tables; + handler->output=xml256CharEncodingOutputFunc; handler->outputInfo=&tables; + + xmlRegisterCharEncodingHandler(handler); +} + +void Charset::initTranscoder(const String *source, const char *name_cstr) { + ftranscoder=xmlFindCharEncodingHandler(name_cstr); + transcoder(source); // check right way +} + +xmlCharEncodingHandler *Charset::transcoder(const String *source) { + if(!ftranscoder) + throw Exception("parser.runtime", + source, + "unsupported encoding"); + return ftranscoder; +} + +const char *Charset::transcode_cstr(xmlChar *s) { + if(!s) + return ""; + + int inlen=strlen((const char *)s); + int outlen=inlen+1; // max + char *out=(char *)malloc(outlen*sizeof(char)); + + int size; + if(xmlCharEncodingOutputFunc output=transcoder(0)->output) { + size=output( + (unsigned char*)out, &outlen, + (const unsigned char*)s, &inlen, + transcoder(0)->outputInfo); + } else + memcpy(out, s, size=inlen); + if(size<0) + throw Exception(0, + 0, + "transcode_cstr failed (%d)", size); + + out[size]=0; + return out; +} +String& Charset::transcode(xmlChar *s) { + return *NEW String(pool(), transcode_cstr(s)); +} +const char *Charset::transcode_cstr(GdomeDOMString *s) { + return s?transcode_cstr(BAD_CAST s->str):""; } -String& Charset::transcode(const XalanDOMString& s) { +String& Charset::transcode(GdomeDOMString *s) { return *NEW String(pool(), transcode_cstr(s)); } -std::auto_ptrCharset::transcode_buf(const char *buf, size_t buf_size) { - unsigned int dest_size=0; - XMLCh* dest=(XMLCh *)malloc((buf_size+1)*sizeof(XMLCh)); - unsigned char *charSizes=(unsigned char *)malloc(buf_size*sizeof(unsigned char)); - XalanDOMString *result; - try { - if(transcoder) { - unsigned int bytesEaten; - unsigned int dest_size=transcoder->transcodeFrom( - (unsigned char *)buf, - (const unsigned int)buf_size, - dest,(const unsigned int)buf_size, - bytesEaten, - charSizes - ); - result=new XalanDOMString(dest, dest_size); - } - } catch(XMLException& e) { - Exception::provide_source(pool(), 0, e); - result=0; //calm, compiler - } +/// @test less memory using -maybe- xmlParserInputBufferCreateMem +GdomeDOMString_auto_ptr Charset::transcode_buf(const char *buf, size_t buf_size) { + int outlen=buf_size*6/*max*/+1; + unsigned char *out=(unsigned char*)malloc(outlen*sizeof(unsigned char)); + + int size; + if(xmlCharEncodingInputFunc input=transcoder(0)->input) { + size=input( + out, &outlen, + (const unsigned char *)buf, (int *)&buf_size, + transcoder(0)->inputInfo); + } else + memcpy(out, buf, size=buf_size); - return std::auto_ptr(result); + if(size<0) + throw Exception(0, + 0, + "transcode_buf failed (%d)", size); + + out[size]=0; + return GdomeDOMString_auto_ptr((gchar*)out); } -std::auto_ptrCharset::transcode(const String& s) { +GdomeDOMString_auto_ptr Charset::transcode(const String& s) { const char *cstr=s.cstr(String::UL_UNSPECIFIED); return transcode_buf(cstr, strlen(cstr));