|
|
| version 1.14, 2002/01/14 17:48:57 | version 1.33.2.12, 2003/02/26 10:17:30 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: Charset connection implementation. | Parser: Charset connection implementation. |
| Copyright(c) 2001 ArtLebedev Group(http://www.artlebedev.com) | Copyright(c) 2001-2003 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) |
| $Id$ | |
| */ | */ |
| static const char* IDENT_CHARSET_C="$Date$"; | |
| #include "pa_charset.h" | #include "pa_charset.h" |
| #ifdef XML | #ifdef XML |
| Line 15 | Line 15 |
| // globals | // globals |
| #define CHARSET_UTF8_NAME "UTF-8" | |
| // consts | CharsetPtr UTF8_charset(new Charset(StringPtr(new String(CHARSET_UTF8_NAME)), |
| StringPtr(0)/*no file=system*/)); | |
| #define MAX_CHARSET_UNI_CODES 500 | |
| // helpers | // helpers |
| Line 37 inline void cstr2ctypes(unsigned char *t | Line 37 inline void cstr2ctypes(unsigned char *t |
| ctypes_table[c]|=bit; | ctypes_table[c]|=bit; |
| } | } |
| } | } |
| inline unsigned int to_wchar_code(const char *cstr) { | inline unsigned int to_wchar_code(const char* cstr) { |
| if(!cstr || !*cstr) | if(!cstr || !*cstr) |
| return 0; | return 0; |
| if(cstr[1]==0) | if(cstr[1]==0) |
| Line 46 inline unsigned int to_wchar_code(const | Line 46 inline unsigned int to_wchar_code(const |
| char *error_pos; | char *error_pos; |
| return(unsigned int)strtol(cstr, &error_pos, 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, |
| Line 74 static void element2case(unsigned char f | Line 74 static void element2case(unsigned char f |
| // methods | // methods |
| extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c | extern "C" unsigned char pcre_default_tables[]; // pcre/chartables.c |
| Charset::Charset(Pool& apool, const String& aname, const String *request_file_spec): Pooled(apool), | Charset::Charset(StringPtr aname, StringPtr afile_spec): |
| fname(aname) { | fname(aname), |
| fname_cstr(aname->cstr()) { | |
| char *name_cstr=fname.cstr(); | for(char *c=fname_cstr; *c; c++) |
| for(char *c=name_cstr; *c; c++) | |
| *c = toupper(*c); | *c = toupper(*c); |
| if(request_file_spec) { | if(afile_spec) { |
| fisUTF8=false; | fisUTF8=false; |
| loadDefinition(*request_file_spec); | load_definition(afile_spec); |
| #ifdef XML | #ifdef XML |
| addEncoding(name_cstr); | addEncoding(fname_cstr); |
| #endif | #endif |
| } else { | } else { |
| fisUTF8=true; | fisUTF8=true; |
| Line 94 Charset::Charset(Pool& apool, const Stri | Line 94 Charset::Charset(Pool& apool, const Stri |
| } | } |
| #ifdef XML | #ifdef XML |
| initTranscoder(&aname, name_cstr); | initTranscoder(fname, fname_cstr); |
| #endif | #endif |
| } | } |
| Charset::~Charset() { | Charset::~Charset() { |
| // @todonow unregister encodings | |
| #ifdef XML | #ifdef XML |
| // not deleting transcoder, that's not our business | // not deleting transcoder, that's not our business |
| #endif | #endif |
| } | } |
| void Charset::loadDefinition(const String& request_file_spec) { | void Charset::load_definition(StringPtr afile_spec) { |
| // pcre_tables | // pcre_tables |
| // lowcase, flipcase, bits digit+word+whitespace, masks | // lowcase, flipcase, bits digit+word+whitespace, masks |
| Line 115 void Charset::loadDefinition(const Strin | Line 116 void Charset::loadDefinition(const Strin |
| cstr2ctypes(pcre_tables,(const unsigned char *)"*+?{^.$|()[", ctype_meta); | cstr2ctypes(pcre_tables,(const unsigned char *)"*+?{^.$|()[", ctype_meta); |
| // charset | // charset |
| memset(tables.fromTable, 0, sizeof(tables.fromTable)); | memset(&tables, 0, sizeof(tables)); |
| tables.toTable=(Charset_TransRec *)calloc(sizeof(Charset_TransRec)*MAX_CHARSET_UNI_CODES); | |
| tables.toTableSize=0; | |
| // strangly vital | // strangly vital |
| tables.toTable[tables.toTableSize].intCh=0; | tables.toTable[tables.toTableSize].intCh=0; |
| tables.toTable[tables.toTableSize].extCh=(XMLByte)0; | tables.toTable[tables.toTableSize].extCh=(XMLByte)0; |
| tables.toTableSize++; | tables.toTableSize++; |
| // loading text | // loading text |
| char *data=file_read_text(pool(), request_file_spec); | Pool pool_for_load_only; |
| char *data=file_read_text(pool_for_load_only, *UTF8_charset, afile_spec); | |
| // ignore header | // ignore header |
| getrow(&data); | getrow(&data); |
| Line 153 void Charset::loadDefinition(const Strin | Line 153 void Charset::loadDefinition(const Strin |
| case 8: | case 8: |
| // charset | // charset |
| if(tables.toTableSize>MAX_CHARSET_UNI_CODES) | if(tables.toTableSize>MAX_CHARSET_UNI_CODES) |
| throw Exception(0, 0, | throw Exception("parser.runtime", |
| &request_file_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); |
| XMLCh unicode=(XMLCh)to_wchar_code(cell); | XMLCh unicode=(XMLCh)to_wchar_code(cell); |
| Line 191 void Charset::sort_ToTable() { | Line 191 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) { | |
| unsigned int lowOfs = 0; | unsigned int lowOfs = 0; |
| unsigned int hiOfs = tables.toTableSize - 1; | unsigned int hiOfs = tables.toTableSize - 1; |
| XMLByte curByte = 0; | XMLByte curByte = 0; |
| Line 210 static XMLByte xlatOneTo(const XMLCh toX | Line 211 static XMLByte xlatOneTo(const XMLCh toX |
| return tables.toTable[midOfs].extCh; | return tables.toTable[midOfs].extCh; |
| } while(lowOfs+1<hiOfs); | } while(lowOfs+1<hiOfs); |
| return '?'; | return not_found; |
| } | } |
| void Charset::transcode(Pool& pool, | void Charset::transcode(Pool& pool, |
| Line 360 static int transcodeToUTF8( | Line 361 static int transcodeToUTF8( |
| // Return the characters read | // Return the characters read |
| toFillLen = outPtr - toFill; | toFillLen = outPtr - toFill; |
| return srcPtr==srcEnd?toFillLen:-1; | //return srcPtr==srcEnd?(int)toFillLen:-1; |
| /* | |
| xmlCharEncodingInputFunc | |
| Returns : | |
| the number of byte written, or -1 by lack of space, or -2 if the transcoding failed. The value of inlen after return is the | |
| number of octets consumed as the return value is positive, else unpredictiable. The value of outlen after return is the number | |
| of ocetes consumed. | |
| */ | |
| return 0; | |
| } | } |
| static size_t transcodeFromUTF8( | /// @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, | const XMLByte *srcData, size_t& srcLen, |
| XMLByte* toFill, size_t& toFillLen, | XMLByte* toFill, size_t& toFillLen, |
| const Charset::Tables& tables) { | const Charset::Tables& tables) { |
| Line 404 static size_t transcodeFromUTF8( | Line 414 static size_t transcodeFromUTF8( |
| break; | break; |
| default: | default: |
| throw Exception(0, 0, | throw Exception(0, |
| 0, | Exception::undefined_source, |
| "transcodeFromUTF8 error: wrong trailingBytes value(%d)", trailingBytes); | "transcodeFromUTF8 error: wrong trailingBytes value(%d)", trailingBytes); |
| } | } |
| tmpVal-=gUTFOffsets[trailingBytes]; | tmpVal-=gUTFOffsets[trailingBytes]; |
| Line 413 static size_t transcodeFromUTF8( | Line 423 static size_t transcodeFromUTF8( |
| // If it will fit into a single char, then put it in. Otherwise | // If it will fit into a single char, then put it in. Otherwise |
| // fail [*encode it as a surrogate pair. If its not valid, use the | // fail [*encode it as a surrogate pair. If its not valid, use the |
| // replacement char.*] | // replacement char.*] |
| if(!(tmpVal & 0xFFFF0000)) | if(!(tmpVal & 0xFFFF0000)) { |
| *outPtr++= xlatOneTo(tmpVal, tables); | if(XMLByte xlat=xlatOneTo(tmpVal, tables, 0)) |
| else | *outPtr++=xlat; |
| throw Exception(0, 0, | else |
| 0, | outPtr+=sprintf((char *)outPtr, "&#%d;", tmpVal); // &#decimal; |
| } else | |
| throw Exception(0, | |
| Exception::undefined_source, | |
| "transcodeFromUTF8 error: too big tmpVal(0x%08X)", tmpVal); | "transcodeFromUTF8 error: too big tmpVal(0x%08X)", tmpVal); |
| } | } |
| Line 427 static size_t transcodeFromUTF8( | Line 440 static size_t transcodeFromUTF8( |
| // Return the characters read | // Return the characters read |
| toFillLen = outPtr - toFill; | toFillLen = outPtr - toFill; |
| return srcPtr==srcEnd?toFillLen:-1; | //return srcPtr==srcEnd?(int)toFillLen:-1; |
| /* | |
| xmlCharEncodingOutputFunc | |
| Returns : | |
| the number of byte written, or -1 by lack of space, or -2 if the transcoding failed. The value of inlen after return is the | |
| number of octets consumed as the return value is positive, else unpredictiable. The value of outlen after return is the number | |
| of ocetes consumed. | |
| */ | |
| return 0; | |
| } | } |
| /// @todo not so memory-hungry with prescan | /// @todo not so memory-hungry with prescan |
| void Charset::transcodeToUTF8(Pool& pool, | void Charset::transcodeToUTF8(Pool& pool, |
| const void *source_body, size_t source_content_length, | const void *source_body, size_t source_content_length, |
| const void *& adest_body, size_t& dest_content_length) const { | const void *& adest_body, size_t& dest_content_length) const { |
| dest_content_length=source_content_length*6/*so that surly enough*/; | dest_content_length=source_content_length*6/*so that surly enough, max utf8 seq len=6*/; |
| XMLByte *dest_body=(XMLByte*)pool.malloc(dest_content_length); | XMLByte *dest_body=new(pool) XMLByte[dest_content_length]; |
| if(::transcodeToUTF8( | if(::transcodeToUTF8( |
| (XMLByte *)source_body, source_content_length, | (XMLByte *)source_body, source_content_length, |
| Line 448 void Charset::transcodeToUTF8(Pool& pool | Line 469 void Charset::transcodeToUTF8(Pool& pool |
| // return | // return |
| adest_body=dest_body; | adest_body=dest_body; |
| } | } |
| /// @test buf overflow | |
| void Charset::transcodeFromUTF8(Pool& pool, | void Charset::transcodeFromUTF8(Pool& pool, |
| const void *source_body, size_t source_content_length, | const void *source_body, size_t source_content_length, |
| const void *& adest_body, size_t& dest_content_length) const { | const void *& adest_body, size_t& dest_content_length) const { |
| dest_content_length=source_content_length/*surly enough*/; | dest_content_length=source_content_length*6/*so that surly enough, "ÿ" has max ratio */; |
| XMLByte *dest_body=(XMLByte*)pool.malloc(dest_content_length); | XMLByte *dest_body=new(pool) XMLByte[dest_content_length]; |
| if(::transcodeFromUTF8( | if(::transcodeFromUTF8( |
| (XMLByte *)source_body, source_content_length, | (XMLByte *)source_body, source_content_length, |
| Line 477 void Charset::transcodeToCharset(Pool& p | Line 497 void Charset::transcodeToCharset(Pool& p |
| adest_content_length=source_content_length; | adest_content_length=source_content_length; |
| } else { | } else { |
| size_t dest_content_length=source_content_length; | size_t dest_content_length=source_content_length; |
| unsigned char *dest_body=(unsigned char *)pool.malloc(dest_content_length); | unsigned char *dest_body=new(pool) unsigned char[dest_content_length]; |
| const XMLByte* srcPtr=(XMLByte *)source_body; | const XMLByte* srcPtr=(XMLByte *)source_body; |
| const XMLByte* srcEnd=(XMLByte *)source_body+source_content_length; | const XMLByte* srcEnd=(XMLByte *)source_body+source_content_length; |
| Line 485 void Charset::transcodeToCharset(Pool& p | Line 505 void Charset::transcodeToCharset(Pool& p |
| for(XMLByte* outPtr=dest_body; srcPtr<srcEnd; srcPtr++) { | for(XMLByte* outPtr=dest_body; srcPtr<srcEnd; srcPtr++) { |
| XMLCh curVal = tables.fromTable[*srcPtr]; | XMLCh curVal = tables.fromTable[*srcPtr]; |
| if(curVal) | if(curVal) |
| *outPtr++=xlatOneTo(curVal, dest_charset.tables); | *outPtr++=xlatOneTo(curVal, dest_charset.tables, '?'); |
| else { | else { |
| // use the replacement character | // use the replacement character |
| *outPtr++= '?'; | *outPtr++= '?'; |
| Line 505 static int xml256CharEncodingInp | Line 525 static int xml256CharEncodingInp |
| int *inlen, | int *inlen, |
| void *info) { | void *info) { |
| return transcodeToUTF8( | return transcodeToUTF8( |
| in, *(unsigned int*)inlen, | in, *(size_t*)inlen, |
| out, *(unsigned int*)outlen, | out, *(size_t*)outlen, |
| *(const Charset::Tables *)info); | *(const Charset::Tables *)info); |
| } | } |
| Line 517 static int xml256CharEncodingOut | Line 537 static int xml256CharEncodingOut |
| int *inlen, | int *inlen, |
| void *info) { | void *info) { |
| return transcodeFromUTF8( | return transcodeFromUTF8( |
| in, *(unsigned int*)inlen, | in, *(size_t*)inlen, |
| out, *(unsigned int*)outlen, | out, *(size_t*)outlen, |
| *(const Charset::Tables *)info); | *(const Charset::Tables *)info); |
| } | } |
| void Charset::addEncoding(char *name_cstr) { | void Charset::addEncoding(char *name_cstr) { |
| xmlCharEncodingHandler *handler= | xmlCharEncodingHandler *handler= |
| (xmlCharEncodingHandler *)malloc(sizeof(xmlCharEncodingHandler)); | (xmlCharEncodingHandler *)[ADD to local array of autoptrs to autofree]new ...(sizeof(xmlCharEncodingHandler)); |
| handler->name=name_cstr; | handler->name=name_cstr; |
| handler->input=xml256CharEncodingInputFunc; handler->inputInfo=&tables; | handler->input=xml256CharEncodingInputFunc; handler->inputInfo=&tables; |
| handler->output=xml256CharEncodingOutputFunc; handler->outputInfo=&tables; | handler->output=xml256CharEncodingOutputFunc; handler->outputInfo=&tables; |
| Line 533 void Charset::addEncoding(char *name_cst | Line 553 void Charset::addEncoding(char *name_cst |
| xmlRegisterCharEncodingHandler(handler); | xmlRegisterCharEncodingHandler(handler); |
| } | } |
| void Charset::initTranscoder(const String *source, const char *name_cstr) { | void Charset::initTranscoder(const String *source, const char* name_cstr) { |
| transcoder=xmlFindCharEncodingHandler(name_cstr); | ftranscoder=xmlFindCharEncodingHandler(name_cstr); |
| if(!transcoder) | transcoder(source); // check right way |
| throw Exception(0, 0, | } |
| xmlCharEncodingHandler *Charset::transcoder(const String *source) { | |
| if(!ftranscoder) | |
| throw Exception("parser.runtime", | |
| source, | source, |
| "unsupported encoding"); | "unsupported encoding"); |
| return ftranscoder; | |
| } | } |
| const char *Charset::transcode_cstr(xmlChar *s) { | CharPtr Charset::transcode_cstr(xmlChar *s) { |
| if(!transcoder) | |
| throw Exception(0, 0, | |
| 0, | |
| "transcode_cstr no transcoder"); | |
| if(!s) | if(!s) |
| return ""; | return ""; |
| int inlen=strlen((const char *)s); | int inlen=strlen((const char* )s); |
| int outlen=inlen+1; // max | int outlen=inlen+1; // max |
| char *out=(char *)malloc(outlen*sizeof(char)); | char *out=new char[outlen]; |
| int size=transcoder->output( | int error; |
| (unsigned char*)out, &outlen, | if(xmlCharEncodingOutputFunc output=transcoder(0)->output) { |
| (const unsigned char*)s, &inlen, | error=output( |
| transcoder->outputInfo); | (unsigned char*)out, &outlen, |
| if(size<0) | (const unsigned char*)s, &inlen, |
| throw Exception(0, 0, | transcoder(0)->outputInfo); |
| } else { | |
| memcpy(out, s, outlen=inlen); | |
| error=0; | |
| } | |
| if(error<0) | |
| throw Exception(0, | |
| 0, | 0, |
| "transcode_cstr failed (%d)", size); | "transcode_cstr failed (%d)", error); |
| out[size]=0; | out[outlen/*surely would be less then on input*/]=0; |
| return out; | return out; |
| } | } |
| String& Charset::transcode(xmlChar *s) { | String& Charset::transcode(xmlChar *s |
| return *NEW String(pool(), transcode_cstr(s)); | #ifndef NO_STRING_ORIGIN |
| , const String *origin | |
| #endif | |
| ) { | |
| String& result=*NEW String(pool()); | |
| result.APPEND_CLEAN(transcode_cstr(s), 0/*auto-size*/, origin->origin().file, origin->origin().line); | |
| return result; | |
| } | } |
| const char *Charset::transcode_cstr(GdomeDOMString *s) { | const char* Charset::transcode_cstr(GdomeDOMString *s) { |
| return s?transcode_cstr(BAD_CAST s->str):""; | return s?transcode_cstr(BAD_CAST s->str):""; |
| } | } |
| String& Charset::transcode(GdomeDOMString *s) { | String& Charset::transcode(GdomeDOMString *s |
| return *NEW String(pool(), transcode_cstr(s)); | #ifndef NO_STRING_ORIGIN |
| , const String *origin | |
| #endif | |
| ) { | |
| String& result=*NEW String(pool()); | |
| result.APPEND_CLEAN(transcode_cstr(s), 0/*auto-size*/, origin->origin().file, origin->origin().line); | |
| return result; | |
| } | } |
| /// @test less memory using -maybe- xmlParserInputBufferCreateMem | /// @test less memory using -maybe- xmlParserInputBufferCreateMem |
| GdomeDOMString_auto_ptr Charset::transcode_buf(const char *buf, size_t buf_size) { | xmlChar *Charset::transcode_buf2xchar(Pool& pool, const char* buf, size_t buf_size) { |
| if(!transcoder) | unsigned char *out; |
| throw Exception(0, 0, | int outlen; |
| 0, | int error; |
| "transcode_buf no transcoder"); | if(xmlCharEncodingInputFunc input=transcoder(0)->input) { |
| outlen=buf_size*6/*max*/; | |
| int outlen=buf_size*6/*max*/+1; | out=new(pool) unsigned char[outlen+1]; |
| unsigned char *out=(unsigned char*)malloc(outlen*sizeof(unsigned char)); | error=input( |
| int size=transcoder->input( | out, &outlen, |
| out, &outlen, | (const unsigned char *)buf, (int *)&buf_size, |
| (const unsigned char *)buf, (int *)&buf_size, | transcoder(0)->inputInfo); |
| transcoder->inputInfo); | } else { |
| if(size<0) | outlen=buf_size; |
| throw Exception(0, 0, | out=new(pool) unsigned char[outlen+1]; |
| memcpy(out, buf, outlen); | |
| error=0; | |
| } | |
| if(error<0) | |
| throw Exception(0, | |
| 0, | 0, |
| "transcode_buf failed (%d)", size); | "transcode_buf failed (%d)", error); |
| out[size]=0; | out[outlen/*surely would be less then on input*/]=0; |
| return (gchar*)out; | return (xmlChar *)out; |
| } | |
| GdomeDOMString_auto_ptr Charset::transcode_buf2dom(const char* buf, size_t buf_size) { | |
| return GdomeDOMString_auto_ptr((gchar*)transcode_buf2xchar(buf, buf_size)); | |
| } | } |
| GdomeDOMString_auto_ptr Charset::transcode(const String& s) { | GdomeDOMString_auto_ptr Charset::transcode(const String& s) { |
| const char *cstr=s.cstr(String::UL_UNSPECIFIED); | CharPtr cstr=s.cstr(String::UL_UNSPECIFIED); |
| return transcode_buf(cstr, strlen(cstr)); | return transcode_buf2dom(cstr, strlen(cstr)); |
| } | } |
| #endif | #endif |