From e2900ce0a1b11ec212aeaf42bbefb26a54697c25 Mon Sep 17 00:00:00 2001 From: Martin Sustrik Date: Mon, 14 Sep 2009 14:30:15 +0200 Subject: xmlParser added --- foreign/xmlParser/xmlParser.cpp | 2888 +++++++++++++++++++++++++++++++++++++++ foreign/xmlParser/xmlParser.hpp | 762 +++++++++++ 2 files changed, 3650 insertions(+) create mode 100644 foreign/xmlParser/xmlParser.cpp create mode 100644 foreign/xmlParser/xmlParser.hpp (limited to 'foreign') diff --git a/foreign/xmlParser/xmlParser.cpp b/foreign/xmlParser/xmlParser.cpp new file mode 100644 index 0000000..2414c01 --- /dev/null +++ b/foreign/xmlParser/xmlParser.cpp @@ -0,0 +1,2888 @@ +/** + **************************************************************************** + *

XML.c - implementation file for basic XML parser written in ANSI C++ + * for portability. It works by using recursion and a node tree for breaking + * down the elements of an XML document.

+ * + * @version V2.39 + * @author Frank Vanden Berghen + * + * NOTE: + * + * If you add "#define STRICT_PARSING", on the first line of this file + * the parser will see the following XML-stream: + * some textother text + * as an error. Otherwise, this tring will be equivalent to: + * some textother text + * + * NOTE: + * + * If you add "#define APPROXIMATE_PARSING" on the first line of this file + * the parser will see the following XML-stream: + * + * + * + * as equivalent to the following XML-stream: + * + * + * + * This can be useful for badly-formed XML-streams but prevent the use + * of the following XML-stream (problem is: tags at contiguous levels + * have the same names): + * + * + * + * + * + * + * NOTE: + * + * If you add "#define _XMLPARSER_NO_MESSAGEBOX_" on the first line of this file + * the "openFileHelper" function will always display error messages inside the + * console instead of inside a message-box-window. Message-box-windows are + * available on windows 9x/NT/2000/XP/Vista only. + * + * Copyright (c) 2002, Frank Vanden Berghen + * All rights reserved. + * + * The following license terms apply to projects that are in some way related to + * the "ZeroMQ project", including applications + * using "ZeroMQ project" and tools developed + * for enhancing "ZeroMQ project". All other projects + * (not related to "ZeroMQ project") have to use this + * code under the Aladdin Free Public License (AFPL) + * See the file "AFPL-license.txt" for more informations about the AFPL license. + * (see http://www.artifex.com/downloads/doc/Public.htm for detailed AFPL terms) + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are met: + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * * Neither the name of Frank Vanden Berghen nor the + * names of its contributors may be used to endorse or promote products + * derived from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY Frank Vanden Berghen ``AS IS'' AND ANY + * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED + * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE + * DISCLAIMED. IN NO EVENT SHALL BE LIABLE FOR ANY + * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES + * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; + * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND + * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS + * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + **************************************************************************** + */ +#ifndef _CRT_SECURE_NO_DEPRECATE +#define _CRT_SECURE_NO_DEPRECATE +#endif +#include +#ifdef _XMLWINDOWS +//#ifdef _DEBUG +//#define _CRTDBG_MAP_ALLOC +//#include +//#endif +#define WIN32_LEAN_AND_MEAN +#include // to have IsTextUnicode, MultiByteToWideChar, WideCharToMultiByte to handle unicode files + // to have "MessageBoxA" to display error messages for openFilHelper +#endif + +#include +#include +#include +#include +#include + +XMLCSTR XMLNode::getVersion() { return _CXML("v2.39"); } +void freeXMLString(XMLSTR t){if(t)free(t);} + +static XMLNode::XMLCharEncoding characterEncoding=XMLNode::char_encoding_UTF8; +static char guessWideCharChars=1, dropWhiteSpace=1, removeCommentsInMiddleOfText=1; + +inline int mmin( const int t1, const int t2 ) { return t1 < t2 ? t1 : t2; } + +// You can modify the initialization of the variable "XMLClearTags" below +// to change the clearTags that are currently recognized by the library. +// The number on the second columns is the length of the string inside the +// first column. The "") }, + { _CXML("") }, + { _CXML("") }, + { _CXML("
")    ,5,  _CXML("
") }, +// { _CXML("")}, + { NULL ,0, NULL } +}; + +// You can modify the initialization of the variable "XMLEntities" below +// to change the character entities that are currently recognized by the library. +// The number on the second columns is the length of the string inside the +// first column. Additionally, the syntaxes " " and " " are recognized. +typedef struct { XMLCSTR s; int l; XMLCHAR c;} XMLCharacterEntity; +static XMLCharacterEntity XMLEntities[] = +{ + { _CXML("&" ), 5, _CXML('&' )}, + { _CXML("<" ), 4, _CXML('<' )}, + { _CXML(">" ), 4, _CXML('>' )}, + { _CXML("""), 6, _CXML('\"')}, + { _CXML("'"), 6, _CXML('\'')}, + { NULL , 0, '\0' } +}; + +// When rendering the XMLNode to a string (using the "createXMLString" function), +// you can ask for a beautiful formatting. This formatting is using the +// following indentation character: +#define INDENTCHAR _CXML('\t') + +// The following function parses the XML errors into a user friendly string. +// You can edit this to change the output language of the library to something else. +XMLCSTR XMLNode::getError(XMLError xerror) +{ + switch (xerror) + { + case eXMLErrorNone: return _CXML("No error"); + case eXMLErrorMissingEndTag: return _CXML("Warning: Unmatched end tag"); + case eXMLErrorNoXMLTagFound: return _CXML("Warning: No XML tag found"); + case eXMLErrorEmpty: return _CXML("Error: No XML data"); + case eXMLErrorMissingTagName: return _CXML("Error: Missing start tag name"); + case eXMLErrorMissingEndTagName: return _CXML("Error: Missing end tag name"); + case eXMLErrorUnmatchedEndTag: return _CXML("Error: Unmatched end tag"); + case eXMLErrorUnmatchedEndClearTag: return _CXML("Error: Unmatched clear tag end"); + case eXMLErrorUnexpectedToken: return _CXML("Error: Unexpected token found"); + case eXMLErrorNoElements: return _CXML("Error: No elements found"); + case eXMLErrorFileNotFound: return _CXML("Error: File not found"); + case eXMLErrorFirstTagNotFound: return _CXML("Error: First Tag not found"); + case eXMLErrorUnknownCharacterEntity:return _CXML("Error: Unknown character entity"); + case eXMLErrorCharacterCodeAbove255: return _CXML("Error: Character code above 255 is forbidden in MultiByte char mode."); + case eXMLErrorCharConversionError: return _CXML("Error: unable to convert between WideChar and MultiByte chars"); + case eXMLErrorCannotOpenWriteFile: return _CXML("Error: unable to open file for writing"); + case eXMLErrorCannotWriteFile: return _CXML("Error: cannot write into file"); + + case eXMLErrorBase64DataSizeIsNotMultipleOf4: return _CXML("Warning: Base64-string length is not a multiple of 4"); + case eXMLErrorBase64DecodeTruncatedData: return _CXML("Warning: Base64-string is truncated"); + case eXMLErrorBase64DecodeIllegalCharacter: return _CXML("Error: Base64-string contains an illegal character"); + case eXMLErrorBase64DecodeBufferTooSmall: return _CXML("Error: Base64 decode output buffer is too small"); + }; + return _CXML("Unknown"); +} + +///////////////////////////////////////////////////////////////////////// +// Here start the abstraction layer to be OS-independent // +///////////////////////////////////////////////////////////////////////// + +// Here is an abstraction layer to access some common string manipulation functions. +// The abstraction layer is currently working for gcc, Microsoft Visual Studio 6.0, +// Microsoft Visual Studio .NET, CC (sun compiler) and Borland C++. +// If you plan to "port" the library to a new system/compiler, all you have to do is +// to edit the following lines. +#ifdef XML_NO_WIDE_CHAR +char myIsTextWideChar(const void *b, int len) { return FALSE; } +#else + #if defined (UNDER_CE) || !defined(_XMLWINDOWS) + char myIsTextWideChar(const void *b, int len) // inspired by the Wine API: RtlIsTextUnicode + { +#ifdef sun + // for SPARC processors: wchar_t* buffers must always be alligned, otherwise it's a char* buffer. + if ((((unsigned long)b)%sizeof(wchar_t))!=0) return FALSE; +#endif + const wchar_t *s=(const wchar_t*)b; + + // buffer too small: + if (len<(int)sizeof(wchar_t)) return FALSE; + + // odd length test + if (len&1) return FALSE; + + /* only checks the first 256 characters */ + len=mmin(256,len/sizeof(wchar_t)); + + // Check for the special byte order: + if (*((unsigned short*)s) == 0xFFFE) return TRUE; // IS_TEXT_UNICODE_REVERSE_SIGNATURE; + if (*((unsigned short*)s) == 0xFEFF) return TRUE; // IS_TEXT_UNICODE_SIGNATURE + + // checks for ASCII characters in the UNICODE stream + int i,stats=0; + for (i=0; ilen/2) return TRUE; + + // Check for UNICODE NULL chars + for (i=0; i + static inline int xstrnicmp(XMLCSTR c1, XMLCSTR c2, int l) { return wsncasecmp(c1,c2,l);} + static inline int xstrncmp(XMLCSTR c1, XMLCSTR c2, int l) { return wsncmp(c1,c2,l);} + static inline int xstricmp(XMLCSTR c1, XMLCSTR c2) { return wscasecmp(c1,c2); } + #else + // for gcc + static inline int xstrnicmp(XMLCSTR c1, XMLCSTR c2, int l) { return wcsncasecmp(c1,c2,l);} + static inline int xstrncmp(XMLCSTR c1, XMLCSTR c2, int l) { return wcsncmp(c1,c2,l);} + static inline int xstricmp(XMLCSTR c1, XMLCSTR c2) { return wcscasecmp(c1,c2); } + #endif + static inline XMLSTR xstrstr(XMLCSTR c1, XMLCSTR c2) { return (XMLSTR)wcsstr(c1,c2); } + static inline XMLSTR xstrcpy(XMLSTR c1, XMLCSTR c2) { return (XMLSTR)wcscpy(c1,c2); } + static inline FILE *xfopen(XMLCSTR filename,XMLCSTR mode) + { + char *filenameAscii=myWideCharToMultiByte(filename); + FILE *f; + if (mode[0]==_CXML('r')) f=fopen(filenameAscii,"rb"); + else f=fopen(filenameAscii,"wb"); + free(filenameAscii); + return f; + } + #else + static inline FILE *xfopen(XMLCSTR filename,XMLCSTR mode) { return fopen(filename,mode); } + static inline int xstrlen(XMLCSTR c) { return strlen(c); } + static inline int xstrnicmp(XMLCSTR c1, XMLCSTR c2, int l) { return strncasecmp(c1,c2,l);} + static inline int xstrncmp(XMLCSTR c1, XMLCSTR c2, int l) { return strncmp(c1,c2,l);} + static inline int xstricmp(XMLCSTR c1, XMLCSTR c2) { return strcasecmp(c1,c2); } + static inline XMLSTR xstrstr(XMLCSTR c1, XMLCSTR c2) { return (XMLSTR)strstr(c1,c2); } + static inline XMLSTR xstrcpy(XMLSTR c1, XMLCSTR c2) { return (XMLSTR)strcpy(c1,c2); } + #endif + static inline int _strnicmp(const char *c1,const char *c2, int l) { return strncasecmp(c1,c2,l);} +#endif + + +/////////////////////////////////////////////////////////////////////////////// +// the "xmltoc,xmltob,xmltoi,xmltol,xmltof,xmltoa" functions // +/////////////////////////////////////////////////////////////////////////////// +// These 6 functions are not used inside the XMLparser. +// There are only here as "convenience" functions for the user. +// If you don't need them, you can delete them without any trouble. +#ifdef _XMLWIDECHAR + #ifdef _XMLWINDOWS + // for Microsoft Visual Studio 6.0 and Microsoft Visual Studio .NET and Borland C++ Builder 6.0 + char xmltob(XMLCSTR t,int v){ if (t&&(*t)) return (char)_wtoi(t); return v; } + int xmltoi(XMLCSTR t,int v){ if (t&&(*t)) return _wtoi(t); return v; } + long xmltol(XMLCSTR t,long v){ if (t&&(*t)) return _wtol(t); return v; } + double xmltof(XMLCSTR t,double v){ if (t&&(*t)) wscanf(t, "%f", &v); /*v=_wtof(t);*/ return v; } + #else + #ifdef sun + // for CC + #include + char xmltob(XMLCSTR t,int v){ if (t) return (char)wstol(t,NULL,10); return v; } + int xmltoi(XMLCSTR t,int v){ if (t) return (int)wstol(t,NULL,10); return v; } + long xmltol(XMLCSTR t,long v){ if (t) return wstol(t,NULL,10); return v; } + #else + // for gcc + char xmltob(XMLCSTR t,int v){ if (t) return (char)wcstol(t,NULL,10); return v; } + int xmltoi(XMLCSTR t,int v){ if (t) return (int)wcstol(t,NULL,10); return v; } + long xmltol(XMLCSTR t,long v){ if (t) return wcstol(t,NULL,10); return v; } + #endif + double xmltof(XMLCSTR t,double v){ if (t&&(*t)) wscanf(t, "%f", &v); /*v=_wtof(t);*/ return v; } + #endif +#else + char xmltob(XMLCSTR t,char v){ if (t&&(*t)) return (char)atoi(t); return v; } + int xmltoi(XMLCSTR t,int v){ if (t&&(*t)) return atoi(t); return v; } + long xmltol(XMLCSTR t,long v){ if (t&&(*t)) return atol(t); return v; } + double xmltof(XMLCSTR t,double v){ if (t&&(*t)) return atof(t); return v; } +#endif +XMLCSTR xmltoa(XMLCSTR t,XMLCSTR v){ if (t) return t; return v; } +XMLCHAR xmltoc(XMLCSTR t,XMLCHAR v){ if (t&&(*t)) return *t; return v; } + +///////////////////////////////////////////////////////////////////////// +// the "openFileHelper" function // +///////////////////////////////////////////////////////////////////////// + +// Since each application has its own way to report and deal with errors, you should modify & rewrite +// the following "openFileHelper" function to get an "error reporting mechanism" tailored to your needs. +XMLNode XMLNode::openFileHelper(XMLCSTR filename, XMLCSTR tag) +{ + // guess the value of the global parameter "characterEncoding" + // (the guess is based on the first 200 bytes of the file). + FILE *f=xfopen(filename,_CXML("rb")); + if (f) + { + char bb[205]; + int l=(int)fread(bb,1,200,f); + setGlobalOptions(guessCharEncoding(bb,l),guessWideCharChars,dropWhiteSpace,removeCommentsInMiddleOfText); + fclose(f); + } + + // parse the file + XMLResults pResults; + XMLNode xnode=XMLNode::parseFile(filename,tag,&pResults); + + // display error message (if any) + if (pResults.error != eXMLErrorNone) + { + // create message + char message[2000],*s1=(char*)"",*s3=(char*)""; XMLCSTR s2=_CXML(""); + if (pResults.error==eXMLErrorFirstTagNotFound) { s1=(char*)"First Tag should be '"; s2=tag; s3=(char*)"'.\n"; } + sprintf(message, +#ifdef _XMLWIDECHAR + "XML Parsing error inside file '%S'.\n%S\nAt line %i, column %i.\n%s%S%s" +#else + "XML Parsing error inside file '%s'.\n%s\nAt line %i, column %i.\n%s%s%s" +#endif + ,filename,XMLNode::getError(pResults.error),pResults.nLine,pResults.nColumn,s1,s2,s3); + + // display message +#if defined(_XMLWINDOWS) && !defined(UNDER_CE) && !defined(_XMLPARSER_NO_MESSAGEBOX_) + MessageBoxA(NULL,message,"XML Parsing error",MB_OK|MB_ICONERROR|MB_TOPMOST); +#else + printf("%s",message); +#endif + exit(255); + } + return xnode; +} + +///////////////////////////////////////////////////////////////////////// +// Here start the core implementation of the XMLParser library // +///////////////////////////////////////////////////////////////////////// + +// You should normally not change anything below this point. + +#ifndef _XMLWIDECHAR +// If "characterEncoding=ascii" then we assume that all characters have the same length of 1 byte. +// If "characterEncoding=UTF8" then the characters have different lengths (from 1 byte to 4 bytes). +// If "characterEncoding=ShiftJIS" then the characters have different lengths (from 1 byte to 2 bytes). +// This table is used as lookup-table to know the length of a character (in byte) based on the +// content of the first byte of the character. +// (note: if you modify this, you must always have XML_utf8ByteTable[0]=0 ). +static const char XML_utf8ByteTable[256] = +{ + // 0 1 2 3 4 5 6 7 8 9 a b c d e f + 0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x00 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x10 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x20 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x30 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x40 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x50 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x60 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x70 End of ASCII range + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x80 0x80 to 0xc1 invalid + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x90 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0xa0 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0xb0 + 1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xc0 0xc2 to 0xdf 2 byte + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xd0 + 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,// 0xe0 0xe0 to 0xef 3 byte + 4,4,4,4,4,1,1,1,1,1,1,1,1,1,1,1 // 0xf0 0xf0 to 0xf4 4 byte, 0xf5 and higher invalid +}; +static const char XML_legacyByteTable[256] = +{ + 0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1 +}; +static const char XML_sjisByteTable[256] = +{ + // 0 1 2 3 4 5 6 7 8 9 a b c d e f + 0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x00 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x10 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x20 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x30 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x40 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x50 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x60 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x70 + 1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0x80 0x81 to 0x9F 2 bytes + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0x90 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0xa0 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0xb0 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0xc0 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0xd0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xe0 0xe0 to 0xef 2 bytes + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1 // 0xf0 +}; +static const char XML_gb2312ByteTable[256] = +{ +// 0 1 2 3 4 5 6 7 8 9 a b c d e f + 0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x00 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x10 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x20 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x30 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x40 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x50 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x60 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x70 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x80 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x90 + 1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xa0 0xa1 to 0xf7 2 bytes + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xb0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xc0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xd0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xe0 + 2,2,2,2,2,2,2,2,1,1,1,1,1,1,1,1 // 0xf0 +}; +static const char XML_gbk_big5_ByteTable[256] = +{ + // 0 1 2 3 4 5 6 7 8 9 a b c d e f + 0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x00 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x10 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x20 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x30 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x40 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x50 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x60 + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,// 0x70 + 1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0x80 0x81 to 0xfe 2 bytes + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0x90 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xa0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xb0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xc0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xd0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,// 0xe0 + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,1 // 0xf0 +}; +static const char *XML_ByteTable=(const char *)XML_utf8ByteTable; // the default is "characterEncoding=XMLNode::encoding_UTF8" +#endif + + +XMLNode XMLNode::emptyXMLNode; +XMLClear XMLNode::emptyXMLClear={ NULL, NULL, NULL}; +XMLAttribute XMLNode::emptyXMLAttribute={ NULL, NULL}; + +// Enumeration used to decipher what type a token is +typedef enum XMLTokenTypeTag +{ + eTokenText = 0, + eTokenQuotedText, + eTokenTagStart, /* "<" */ + eTokenTagEnd, /* "" */ + eTokenEquals, /* "=" */ + eTokenDeclaration, /* "" */ + eTokenClear, + eTokenError +} XMLTokenType; + +// Main structure used for parsing XML +typedef struct XML +{ + XMLCSTR lpXML; + XMLCSTR lpszText; + int nIndex,nIndexMissigEndTag; + enum XMLError error; + XMLCSTR lpEndTag; + int cbEndTag; + XMLCSTR lpNewElement; + int cbNewElement; + int nFirst; +} XML; + +typedef struct +{ + ALLXMLClearTag *pClr; + XMLCSTR pStr; +} NextToken; + +// Enumeration used when parsing attributes +typedef enum Attrib +{ + eAttribName = 0, + eAttribEquals, + eAttribValue +} Attrib; + +// Enumeration used when parsing elements to dictate whether we are currently +// inside a tag +typedef enum Status +{ + eInsideTag = 0, + eOutsideTag +} Status; + +XMLError XMLNode::writeToFile(XMLCSTR filename, const char *encoding, char nFormat) const +{ + if (!d) return eXMLErrorNone; + FILE *f=xfopen(filename,_CXML("wb")); + if (!f) return eXMLErrorCannotOpenWriteFile; +#ifdef _XMLWIDECHAR + unsigned char h[2]={ 0xFF, 0xFE }; + if (!fwrite(h,2,1,f)) return eXMLErrorCannotWriteFile; + if ((!isDeclaration())&&((d->lpszName)||(!getChildNode().isDeclaration()))) + { + if (!fwrite(L"\n",sizeof(wchar_t)*40,1,f)) + return eXMLErrorCannotWriteFile; + } +#else + if ((!isDeclaration())&&((d->lpszName)||(!getChildNode().isDeclaration()))) + { + if (characterEncoding==char_encoding_UTF8) + { + // header so that windows recognize the file as UTF-8: + unsigned char h[3]={0xEF,0xBB,0xBF}; if (!fwrite(h,3,1,f)) return eXMLErrorCannotWriteFile; + encoding="utf-8"; + } else if (characterEncoding==char_encoding_ShiftJIS) encoding="SHIFT-JIS"; + + if (!encoding) encoding="ISO-8859-1"; + if (fprintf(f,"\n",encoding)<0) return eXMLErrorCannotWriteFile; + } else + { + if (characterEncoding==char_encoding_UTF8) + { + unsigned char h[3]={0xEF,0xBB,0xBF}; if (!fwrite(h,3,1,f)) return eXMLErrorCannotWriteFile; + } + } +#endif + int i; + XMLSTR t=createXMLString(nFormat,&i); + if (!fwrite(t,sizeof(XMLCHAR)*i,1,f)) return eXMLErrorCannotWriteFile; + if (fclose(f)!=0) return eXMLErrorCannotWriteFile; + free(t); + return eXMLErrorNone; +} + +// Duplicate a given string. +XMLSTR stringDup(XMLCSTR lpszData, int cbData) +{ + if (lpszData==NULL) return NULL; + + XMLSTR lpszNew; + if (cbData==-1) cbData=(int)xstrlen(lpszData); + lpszNew = (XMLSTR)malloc((cbData+1) * sizeof(XMLCHAR)); + if (lpszNew) + { + memcpy(lpszNew, lpszData, (cbData) * sizeof(XMLCHAR)); + lpszNew[cbData] = (XMLCHAR)NULL; + } + return lpszNew; +} + +XMLSTR ToXMLStringTool::toXMLUnSafe(XMLSTR dest,XMLCSTR source) +{ + XMLSTR dd=dest; + XMLCHAR ch; + XMLCharacterEntity *entity; + while ((ch=*source)) + { + entity=XMLEntities; + do + { + if (ch==entity->c) {xstrcpy(dest,entity->s); dest+=entity->l; source++; goto out_of_loop1; } + entity++; + } while(entity->s); +#ifdef _XMLWIDECHAR + *(dest++)=*(source++); +#else + switch(XML_ByteTable[(unsigned char)ch]) + { + case 4: *(dest++)=*(source++); + case 3: *(dest++)=*(source++); + case 2: *(dest++)=*(source++); + case 1: *(dest++)=*(source++); + } +#endif +out_of_loop1: + ; + } + *dest=0; + return dd; +} + +// private (used while rendering): +int ToXMLStringTool::lengthXMLString(XMLCSTR source) +{ + int r=0; + XMLCharacterEntity *entity; + XMLCHAR ch; + while ((ch=*source)) + { + entity=XMLEntities; + do + { + if (ch==entity->c) { r+=entity->l; source++; goto out_of_loop1; } + entity++; + } while(entity->s); +#ifdef _XMLWIDECHAR + r++; source++; +#else + ch=XML_ByteTable[(unsigned char)ch]; r+=ch; source+=ch; +#endif +out_of_loop1: + ; + } + return r; +} + +ToXMLStringTool::~ToXMLStringTool(){ freeBuffer(); } +void ToXMLStringTool::freeBuffer(){ if (buf) free(buf); buf=NULL; buflen=0; } +XMLSTR ToXMLStringTool::toXML(XMLCSTR source) +{ + int l=lengthXMLString(source)+1; + if (l>buflen) { buflen=l; buf=(XMLSTR)realloc(buf,l*sizeof(XMLCHAR)); } + return toXMLUnSafe(buf,source); +} + +// private: +XMLSTR fromXMLString(XMLCSTR s, int lo, XML *pXML) +{ + // This function is the opposite of the function "toXMLString". It decodes the escape + // sequences &, ", ', <, > and replace them by the characters + // &,",',<,>. This function is used internally by the XML Parser. All the calls to + // the XML library will always gives you back "decoded" strings. + // + // in: string (s) and length (lo) of string + // out: new allocated string converted from xml + if (!s) return NULL; + + int ll=0,j; + XMLSTR d; + XMLCSTR ss=s; + XMLCharacterEntity *entity; + while ((lo>0)&&(*s)) + { + if (*s==_CXML('&')) + { + if ((lo>2)&&(s[1]==_CXML('#'))) + { + s+=2; lo-=2; + if ((*s==_CXML('X'))||(*s==_CXML('x'))) { s++; lo--; } + while ((*s)&&(*s!=_CXML(';'))&&((lo--)>0)) s++; + if (*s!=_CXML(';')) + { + pXML->error=eXMLErrorUnknownCharacterEntity; + return NULL; + } + s++; lo--; + } else + { + entity=XMLEntities; + do + { + if ((lo>=entity->l)&&(xstrnicmp(s,entity->s,entity->l)==0)) { s+=entity->l; lo-=entity->l; break; } + entity++; + } while(entity->s); + if (!entity->s) + { + pXML->error=eXMLErrorUnknownCharacterEntity; + return NULL; + } + } + } else + { +#ifdef _XMLWIDECHAR + s++; lo--; +#else + j=XML_ByteTable[(unsigned char)*s]; s+=j; lo-=j; ll+=j-1; +#endif + } + ll++; + } + + d=(XMLSTR)malloc((ll+1)*sizeof(XMLCHAR)); + s=d; + while (ll-->0) + { + if (*ss==_CXML('&')) + { + if (ss[1]==_CXML('#')) + { + ss+=2; j=0; + if ((*ss==_CXML('X'))||(*ss==_CXML('x'))) + { + ss++; + while (*ss!=_CXML(';')) + { + if ((*ss>=_CXML('0'))&&(*ss<=_CXML('9'))) j=(j<<4)+*ss-_CXML('0'); + else if ((*ss>=_CXML('A'))&&(*ss<=_CXML('F'))) j=(j<<4)+*ss-_CXML('A')+10; + else if ((*ss>=_CXML('a'))&&(*ss<=_CXML('f'))) j=(j<<4)+*ss-_CXML('a')+10; + else { free((void*)s); pXML->error=eXMLErrorUnknownCharacterEntity;return NULL;} + ss++; + } + } else + { + while (*ss!=_CXML(';')) + { + if ((*ss>=_CXML('0'))&&(*ss<=_CXML('9'))) j=(j*10)+*ss-_CXML('0'); + else { free((void*)s); pXML->error=eXMLErrorUnknownCharacterEntity;return NULL;} + ss++; + } + } +#ifndef _XMLWIDECHAR + if (j>255) { free((void*)s); pXML->error=eXMLErrorCharacterCodeAbove255;return NULL;} +#endif + (*d++)=(XMLCHAR)j; ss++; + } else + { + entity=XMLEntities; + do + { + if (xstrnicmp(ss,entity->s,entity->l)==0) { *(d++)=entity->c; ss+=entity->l; break; } + entity++; + } while(entity->s); + } + } else + { +#ifdef _XMLWIDECHAR + *(d++)=*(ss++); +#else + switch(XML_ByteTable[(unsigned char)*ss]) + { + case 4: *(d++)=*(ss++); ll--; + case 3: *(d++)=*(ss++); ll--; + case 2: *(d++)=*(ss++); ll--; + case 1: *(d++)=*(ss++); + } +#endif + } + } + *d=0; + return (XMLSTR)s; +} + +#define XML_isSPACECHAR(ch) ((ch==_CXML('\n'))||(ch==_CXML(' '))||(ch== _CXML('\t'))||(ch==_CXML('\r'))) + +// private: +char myTagCompare(XMLCSTR cclose, XMLCSTR copen) +// !!!! WARNING strange convention&: +// return 0 if equals +// return 1 if different +{ + if (!cclose) return 1; + int l=(int)xstrlen(cclose); + if (xstrnicmp(cclose, copen, l)!=0) return 1; + const XMLCHAR c=copen[l]; + if (XML_isSPACECHAR(c)|| + (c==_CXML('/' ))|| + (c==_CXML('<' ))|| + (c==_CXML('>' ))|| + (c==_CXML('=' ))) return 0; + return 1; +} + +// Obtain the next character from the string. +static inline XMLCHAR getNextChar(XML *pXML) +{ + XMLCHAR ch = pXML->lpXML[pXML->nIndex]; +#ifdef _XMLWIDECHAR + if (ch!=0) pXML->nIndex++; +#else + pXML->nIndex+=XML_ByteTable[(unsigned char)ch]; +#endif + return ch; +} + +// Find the next token in a string. +// pcbToken contains the number of characters that have been read. +static NextToken GetNextToken(XML *pXML, int *pcbToken, enum XMLTokenTypeTag *pType) +{ + NextToken result; + XMLCHAR ch; + XMLCHAR chTemp; + int indexStart,nFoundMatch,nIsText=FALSE; + result.pClr=NULL; // prevent warning + + // Find next non-white space character + do { indexStart=pXML->nIndex; ch=getNextChar(pXML); } while XML_isSPACECHAR(ch); + + if (ch) + { + // Cache the current string pointer + result.pStr = &pXML->lpXML[indexStart]; + + // First check whether the token is in the clear tag list (meaning it + // does not need formatting). + ALLXMLClearTag *ctag=XMLClearTags; + do + { + if (xstrncmp(ctag->lpszOpen, result.pStr, ctag->openTagLen)==0) + { + result.pClr=ctag; + pXML->nIndex+=ctag->openTagLen-1; + *pType=eTokenClear; + return result; + } + ctag++; + } while(ctag->lpszOpen); + + // If we didn't find a clear tag then check for standard tokens + switch(ch) + { + // Check for quotes + case _CXML('\''): + case _CXML('\"'): + // Type of token + *pType = eTokenQuotedText; + chTemp = ch; + + // Set the size + nFoundMatch = FALSE; + + // Search through the string to find a matching quote + while((ch = getNextChar(pXML))) + { + if (ch==chTemp) { nFoundMatch = TRUE; break; } + if (ch==_CXML('<')) break; + } + + // If we failed to find a matching quote + if (nFoundMatch == FALSE) + { + pXML->nIndex=indexStart+1; + nIsText=TRUE; + break; + } + +// 4.02.2002 +// if (FindNonWhiteSpace(pXML)) pXML->nIndex--; + + break; + + // Equals (used with attribute values) + case _CXML('='): + *pType = eTokenEquals; + break; + + // Close tag + case _CXML('>'): + *pType = eTokenCloseTag; + break; + + // Check for tag start and tag end + case _CXML('<'): + + // Peek at the next character to see if we have an end tag 'lpXML[pXML->nIndex]; + + // If we have a tag end... + if (chTemp == _CXML('/')) + { + // Set the type and ensure we point at the next character + getNextChar(pXML); + *pType = eTokenTagEnd; + } + + // If we have an XML declaration tag + else if (chTemp == _CXML('?')) + { + + // Set the type and ensure we point at the next character + getNextChar(pXML); + *pType = eTokenDeclaration; + } + + // Otherwise we must have a start tag + else + { + *pType = eTokenTagStart; + } + break; + + // Check to see if we have a short hand type end tag ('/>'). + case _CXML('/'): + + // Peek at the next character to see if we have a short end tag '/>' + chTemp = pXML->lpXML[pXML->nIndex]; + + // If we have a short hand end tag... + if (chTemp == _CXML('>')) + { + // Set the type and ensure we point at the next character + getNextChar(pXML); + *pType = eTokenShortHandClose; + break; + } + + // If we haven't found a short hand closing tag then drop into the + // text process + + // Other characters + default: + nIsText = TRUE; + } + + // If this is a TEXT node + if (nIsText) + { + // Indicate we are dealing with text + *pType = eTokenText; + while((ch = getNextChar(pXML))) + { + if XML_isSPACECHAR(ch) + { + indexStart++; break; + + } else if (ch==_CXML('/')) + { + // If we find a slash then this maybe text or a short hand end tag + // Peek at the next character to see it we have short hand end tag + ch=pXML->lpXML[pXML->nIndex]; + // If we found a short hand end tag then we need to exit the loop + if (ch==_CXML('>')) { pXML->nIndex--; break; } + + } else if ((ch==_CXML('<'))||(ch==_CXML('>'))||(ch==_CXML('='))) + { + pXML->nIndex--; break; + } + } + } + *pcbToken = pXML->nIndex-indexStart; + } else + { + // If we failed to obtain a valid character + *pcbToken = 0; + *pType = eTokenError; + result.pStr=NULL; + } + + return result; +} + +XMLCSTR XMLNode::updateName_WOSD(XMLSTR lpszName) +{ + if (!d) { free(lpszName); return NULL; } + if (d->lpszName&&(lpszName!=d->lpszName)) free((void*)d->lpszName); + d->lpszName=lpszName; + return lpszName; +} + +// private: +XMLNode::XMLNode(struct XMLNodeDataTag *p){ d=p; (p->ref_count)++; } +XMLNode::XMLNode(XMLNodeData *pParent, XMLSTR lpszName, char isDeclaration) +{ + d=(XMLNodeData*)malloc(sizeof(XMLNodeData)); + d->ref_count=1; + + d->lpszName=NULL; + d->nChild= 0; + d->nText = 0; + d->nClear = 0; + d->nAttribute = 0; + + d->isDeclaration = isDeclaration; + + d->pParent = pParent; + d->pChild= NULL; + d->pText= NULL; + d->pClear= NULL; + d->pAttribute= NULL; + d->pOrder= NULL; + + updateName_WOSD(lpszName); +} + +XMLNode XMLNode::createXMLTopNode_WOSD(XMLSTR lpszName, char isDeclaration) { return XMLNode(NULL,lpszName,isDeclaration); } +XMLNode XMLNode::createXMLTopNode(XMLCSTR lpszName, char isDeclaration) { return XMLNode(NULL,stringDup(lpszName),isDeclaration); } + +#define MEMORYINCREASE 50 + +static inline void myFree(void *p) { if (p) free(p); } +static inline void *myRealloc(void *p, int newsize, int memInc, int sizeofElem) +{ + if (p==NULL) { if (memInc) return malloc(memInc*sizeofElem); return malloc(sizeofElem); } + if ((memInc==0)||((newsize%memInc)==0)) p=realloc(p,(newsize+memInc)*sizeofElem); +// if (!p) +// { +// printf("XMLParser Error: Not enough memory! Aborting...\n"); exit(220); +// } + return p; +} + +// private: +XMLElementPosition XMLNode::findPosition(XMLNodeData *d, int index, XMLElementType xxtype) +{ + if (index<0) return -1; + int i=0,j=(int)((index<<2)+xxtype),*o=d->pOrder; while (o[i]!=j) i++; return i; +} + +// private: +// update "order" information when deleting a content of a XMLNode +int XMLNode::removeOrderElement(XMLNodeData *d, XMLElementType t, int index) +{ + int n=d->nChild+d->nText+d->nClear, *o=d->pOrder,i=findPosition(d,index,t); + memmove(o+i, o+i+1, (n-i)*sizeof(int)); + for (;ipOrder=(int)realloc(d->pOrder,n*sizeof(int)); + // but we skip reallocation because it's too time consuming. + // Anyway, at the end, it will be free'd completely at once. + return i; +} + +void *XMLNode::addToOrder(int memoryIncrease,int *_pos, int nc, void *p, int size, XMLElementType xtype) +{ + // in: *_pos is the position inside d->pOrder ("-1" means "EndOf") + // out: *_pos is the index inside p + p=myRealloc(p,(nc+1),memoryIncrease,size); + int n=d->nChild+d->nText+d->nClear; + d->pOrder=(int*)myRealloc(d->pOrder,n+1,memoryIncrease*3,sizeof(int)); + int pos=*_pos,*o=d->pOrder; + + if ((pos<0)||(pos>=n)) { *_pos=nc; o[n]=(int)((nc<<2)+xtype); return p; } + + int i=pos; + memmove(o+i+1, o+i, (n-i)*sizeof(int)); + + while ((pos>2; + memmove(((char*)p)+(pos+1)*size,((char*)p)+pos*size,(nc-pos)*size); + + return p; +} + +// Add a child node to the given element. +XMLNode XMLNode::addChild_priv(int memoryIncrease, XMLSTR lpszName, char isDeclaration, int pos) +{ + if (!lpszName) return emptyXMLNode; + d->pChild=(XMLNode*)addToOrder(memoryIncrease,&pos,d->nChild,d->pChild,sizeof(XMLNode),eNodeChild); + d->pChild[pos].d=NULL; + d->pChild[pos]=XMLNode(d,lpszName,isDeclaration); + d->nChild++; + return d->pChild[pos]; +} + +// Add an attribute to an element. +XMLAttribute *XMLNode::addAttribute_priv(int memoryIncrease,XMLSTR lpszName, XMLSTR lpszValuev) +{ + if (!lpszName) return &emptyXMLAttribute; + if (!d) { myFree(lpszName); myFree(lpszValuev); return &emptyXMLAttribute; } + int nc=d->nAttribute; + d->pAttribute=(XMLAttribute*)myRealloc(d->pAttribute,(nc+1),memoryIncrease,sizeof(XMLAttribute)); + XMLAttribute *pAttr=d->pAttribute+nc; + pAttr->lpszName = lpszName; + pAttr->lpszValue = lpszValuev; + d->nAttribute++; + return pAttr; +} + +// Add text to the element. +XMLCSTR XMLNode::addText_priv(int memoryIncrease, XMLSTR lpszValue, int pos) +{ + if (!lpszValue) return NULL; + if (!d) { myFree(lpszValue); return NULL; } + d->pText=(XMLCSTR*)addToOrder(memoryIncrease,&pos,d->nText,d->pText,sizeof(XMLSTR),eNodeText); + d->pText[pos]=lpszValue; + d->nText++; + return lpszValue; +} + +// Add clear (unformatted) text to the element. +XMLClear *XMLNode::addClear_priv(int memoryIncrease, XMLSTR lpszValue, XMLCSTR lpszOpen, XMLCSTR lpszClose, int pos) +{ + if (!lpszValue) return &emptyXMLClear; + if (!d) { myFree(lpszValue); return &emptyXMLClear; } + d->pClear=(XMLClear *)addToOrder(memoryIncrease,&pos,d->nClear,d->pClear,sizeof(XMLClear),eNodeClear); + XMLClear *pNewClear=d->pClear+pos; + pNewClear->lpszValue = lpszValue; + if (!lpszOpen) lpszOpen=XMLClearTags->lpszOpen; + if (!lpszClose) lpszClose=XMLClearTags->lpszClose; + pNewClear->lpszOpenTag = lpszOpen; + pNewClear->lpszCloseTag = lpszClose; + d->nClear++; + return pNewClear; +} + +// private: +// Parse a clear (unformatted) type node. +char XMLNode::parseClearTag(void *px, void *_pClear) +{ + XML *pXML=(XML *)px; + ALLXMLClearTag pClear=*((ALLXMLClearTag*)_pClear); + int cbTemp=0; + XMLCSTR lpszTemp=NULL; + XMLCSTR lpXML=&pXML->lpXML[pXML->nIndex]; + static XMLCSTR docTypeEnd=_CXML("]>"); + + // Find the closing tag + // Seems the ')) { lpszTemp=pCh; break; } +#ifdef _XMLWIDECHAR + pCh++; +#else + pCh+=XML_ByteTable[(unsigned char)(*pCh)]; +#endif + } + } else lpszTemp=xstrstr(lpXML, pClear.lpszClose); + + if (lpszTemp) + { + // Cache the size and increment the index + cbTemp = (int)(lpszTemp - lpXML); + + pXML->nIndex += cbTemp+(int)xstrlen(pClear.lpszClose); + + // Add the clear node to the current element + addClear_priv(MEMORYINCREASE,stringDup(lpXML,cbTemp), pClear.lpszOpen, pClear.lpszClose,-1); + return 0; + } + + // If we failed to find the end tag + pXML->error = eXMLErrorUnmatchedEndClearTag; + return 1; +} + +void XMLNode::exactMemory(XMLNodeData *d) +{ + if (d->pOrder) d->pOrder=(int*)realloc(d->pOrder,(d->nChild+d->nText+d->nClear)*sizeof(int)); + if (d->pChild) d->pChild=(XMLNode*)realloc(d->pChild,d->nChild*sizeof(XMLNode)); + if (d->pAttribute) d->pAttribute=(XMLAttribute*)realloc(d->pAttribute,d->nAttribute*sizeof(XMLAttribute)); + if (d->pText) d->pText=(XMLCSTR*)realloc(d->pText,d->nText*sizeof(XMLSTR)); + if (d->pClear) d->pClear=(XMLClear *)realloc(d->pClear,d->nClear*sizeof(XMLClear)); +} + +char XMLNode::maybeAddTxT(void *pa, XMLCSTR tokenPStr) +{ + XML *pXML=(XML *)pa; + XMLCSTR lpszText=pXML->lpszText; + if (!lpszText) return 0; + if (dropWhiteSpace) while (XML_isSPACECHAR(*lpszText)&&(lpszText!=tokenPStr)) lpszText++; + int cbText = (int)(tokenPStr - lpszText); + if (!cbText) { pXML->lpszText=NULL; return 0; } + if (dropWhiteSpace) { cbText--; while ((cbText)&&XML_isSPACECHAR(lpszText[cbText])) cbText--; cbText++; } + if (!cbText) { pXML->lpszText=NULL; return 0; } + XMLSTR lpt=fromXMLString(lpszText,cbText,pXML); + if (!lpt) return 1; + pXML->lpszText=NULL; + if (removeCommentsInMiddleOfText && d->nText && d->nClear) + { + // if the previous insertion was a comment () AND + // if the previous previous insertion was a text then, delete the comment and append the text + int n=d->nChild+d->nText+d->nClear-1,*o=d->pOrder; + if (((o[n]&3)==eNodeClear)&&((o[n-1]&3)==eNodeText)) + { + int i=o[n]>>2; + if (d->pClear[i].lpszOpenTag==XMLClearTags[2].lpszOpen) + { + deleteClear(i); + i=o[n-1]>>2; + n=xstrlen(d->pText[i]); + int n2=xstrlen(lpt)+1; + d->pText[i]=(XMLSTR)realloc((void*)d->pText[i],(n+n2)*sizeof(XMLCHAR)); + if (!d->pText[i]) return 1; + memcpy((void*)(d->pText[i]+n),lpt,n2*sizeof(XMLCHAR)); + free(lpt); + return 0; + } + } + } + addText_priv(MEMORYINCREASE,lpt,-1); + return 0; +} +// private: +// Recursively parse an XML element. +int XMLNode::ParseXMLElement(void *pa) +{ + XML *pXML=(XML *)pa; + int cbToken; + enum XMLTokenTypeTag xtype; + NextToken token; + XMLCSTR lpszTemp=NULL; + int cbTemp=0; + char nDeclaration; + XMLNode pNew; + enum Status status; // inside or outside a tag + enum Attrib attrib = eAttribName; + + assert(pXML); + + // If this is the first call to the function + if (pXML->nFirst) + { + // Assume we are outside of a tag definition + pXML->nFirst = FALSE; + status = eOutsideTag; + } else + { + // If this is not the first call then we should only be called when inside a tag. + status = eInsideTag; + } + + // Iterate through the tokens in the document + for(;;) + { + // Obtain the next token + token = GetNextToken(pXML, &cbToken, &xtype); + + if (xtype != eTokenError) + { + // Check the current status + switch(status) + { + + // If we are outside of a tag definition + case eOutsideTag: + + // Check what type of token we obtained + switch(xtype) + { + // If we have found text or quoted text + case eTokenText: + case eTokenCloseTag: /* '>' */ + case eTokenShortHandClose: /* '/>' */ + case eTokenQuotedText: + case eTokenEquals: + break; + + // If we found a start tag '<' and declarations 'error = eXMLErrorMissingTagName; + return FALSE; + } + + // If we found a new element which is the same as this + // element then we need to pass this back to the caller.. + +#ifdef APPROXIMATE_PARSING + if (d->lpszName && + myTagCompare(d->lpszName, token.pStr) == 0) + { + // Indicate to the caller that it needs to create a + // new element. + pXML->lpNewElement = token.pStr; + pXML->cbNewElement = cbToken; + return TRUE; + } else +#endif + { + // If the name of the new element differs from the name of + // the current element we need to add the new element to + // the current one and recurse + pNew = addChild_priv(MEMORYINCREASE,stringDup(token.pStr,cbToken), nDeclaration,-1); + + while (!pNew.isEmpty()) + { + // Callself to process the new node. If we return + // FALSE this means we dont have any more + // processing to do... + + if (!pNew.ParseXMLElement(pXML)) return FALSE; + else + { + // If the call to recurse this function + // evented in a end tag specified in XML then + // we need to unwind the calls to this + // function until we find the appropriate node + // (the element name and end tag name must + // match) + if (pXML->cbEndTag) + { + // If we are back at the root node then we + // have an unmatched end tag + if (!d->lpszName) + { + pXML->error=eXMLErrorUnmatchedEndTag; + return FALSE; + } + + // If the end tag matches the name of this + // element then we only need to unwind + // once more... + + if (myTagCompare(d->lpszName, pXML->lpEndTag)==0) + { + pXML->cbEndTag = 0; + } + + return TRUE; + } else + if (pXML->cbNewElement) + { + // If the call indicated a new element is to + // be created on THIS element. + + // If the name of this element matches the + // name of the element we need to create + // then we need to return to the caller + // and let it process the element. + + if (myTagCompare(d->lpszName, pXML->lpNewElement)==0) + { + return TRUE; + } + + // Add the new element and recurse + pNew = addChild_priv(MEMORYINCREASE,stringDup(pXML->lpNewElement,pXML->cbNewElement),0,-1); + pXML->cbNewElement = 0; + } + else + { + // If we didn't have a new element to create + pNew = emptyXMLNode; + + } + } + } + } + break; + + // If we found an end tag + case eTokenTagEnd: + + // If we have node text then add this to the element + if (maybeAddTxT(pXML,token.pStr)) return FALSE; + + // Find the name of the end tag + token = GetNextToken(pXML, &cbTemp, &xtype); + + // The end tag should be text + if (xtype != eTokenText) + { + pXML->error = eXMLErrorMissingEndTagName; + return FALSE; + } + lpszTemp = token.pStr; + + // After the end tag we should find a closing tag + token = GetNextToken(pXML, &cbToken, &xtype); + if (xtype != eTokenCloseTag) + { + pXML->error = eXMLErrorMissingEndTagName; + return FALSE; + } + pXML->lpszText=pXML->lpXML+pXML->nIndex; + + // We need to return to the previous caller. If the name + // of the tag cannot be found we need to keep returning to + // caller until we find a match + if (myTagCompare(d->lpszName, lpszTemp) != 0) +#ifdef STRICT_PARSING + { + pXML->error=eXMLErrorUnmatchedEndTag; + pXML->nIndexMissigEndTag=pXML->nIndex; + return FALSE; + } +#else + { + pXML->error=eXMLErrorMissingEndTag; + pXML->nIndexMissigEndTag=pXML->nIndex; + pXML->lpEndTag = lpszTemp; + pXML->cbEndTag = cbTemp; + } +#endif + + // Return to the caller + exactMemory(d); + return TRUE; + + // If we found a clear (unformatted) token + case eTokenClear: + // If we have node text then add this to the element + if (maybeAddTxT(pXML,token.pStr)) return FALSE; + if (parseClearTag(pXML, token.pClr)) return FALSE; + pXML->lpszText=pXML->lpXML+pXML->nIndex; + break; + + default: + break; + } + break; + + // If we are inside a tag definition we need to search for attributes + case eInsideTag: + + // Check what part of the attribute (name, equals, value) we + // are looking for. + switch(attrib) + { + // If we are looking for a new attribute + case eAttribName: + + // Check what the current token type is + switch(xtype) + { + // If the current type is text... + // Eg. 'attribute' + case eTokenText: + // Cache the token then indicate that we are next to + // look for the equals + lpszTemp = token.pStr; + cbTemp = cbToken; + attrib = eAttribEquals; + break; + + // If we found a closing tag... + // Eg. '>' + case eTokenCloseTag: + // We are now outside the tag + status = eOutsideTag; + pXML->lpszText=pXML->lpXML+pXML->nIndex; + break; + + // If we found a short hand '/>' closing tag then we can + // return to the caller + case eTokenShortHandClose: + exactMemory(d); + pXML->lpszText=pXML->lpXML+pXML->nIndex; + return TRUE; + + // Errors... + case eTokenQuotedText: /* '"SomeText"' */ + case eTokenTagStart: /* '<' */ + case eTokenTagEnd: /* 'error = eXMLErrorUnexpectedToken; + return FALSE; + default: break; + } + break; + + // If we are looking for an equals + case eAttribEquals: + // Check what the current token type is + switch(xtype) + { + // If the current type is text... + // Eg. 'Attribute AnotherAttribute' + case eTokenText: + // Add the unvalued attribute to the list + addAttribute_priv(MEMORYINCREASE,stringDup(lpszTemp,cbTemp), NULL); + // Cache the token then indicate. We are next to + // look for the equals attribute + lpszTemp = token.pStr; + cbTemp = cbToken; + break; + + // If we found a closing tag 'Attribute >' or a short hand + // closing tag 'Attribute />' + case eTokenShortHandClose: + case eTokenCloseTag: + // If we are a declaration element 'lpszText=pXML->lpXML+pXML->nIndex; + + if (d->isDeclaration && + (lpszTemp[cbTemp-1]) == _CXML('?')) + { + cbTemp--; + if (d->pParent && d->pParent->pParent) xtype = eTokenShortHandClose; + } + + if (cbTemp) + { + // Add the unvalued attribute to the list + addAttribute_priv(MEMORYINCREASE,stringDup(lpszTemp,cbTemp), NULL); + } + + // If this is the end of the tag then return to the caller + if (xtype == eTokenShortHandClose) + { + exactMemory(d); + return TRUE; + } + + // We are now outside the tag + status = eOutsideTag; + break; + + // If we found the equals token... + // Eg. 'Attribute =' + case eTokenEquals: + // Indicate that we next need to search for the value + // for the attribute + attrib = eAttribValue; + break; + + // Errors... + case eTokenQuotedText: /* 'Attribute "InvalidAttr"'*/ + case eTokenTagStart: /* 'Attribute <' */ + case eTokenTagEnd: /* 'Attribute error = eXMLErrorUnexpectedToken; + return FALSE; + default: break; + } + break; + + // If we are looking for an attribute value + case eAttribValue: + // Check what the current token type is + switch(xtype) + { + // If the current type is text or quoted text... + // Eg. 'Attribute = "Value"' or 'Attribute = Value' or + // 'Attribute = 'Value''. + case eTokenText: + case eTokenQuotedText: + // If we are a declaration element 'isDeclaration && + (token.pStr[cbToken-1]) == _CXML('?')) + { + cbToken--; + } + + if (cbTemp) + { + // Add the valued attribute to the list + if (xtype==eTokenQuotedText) { token.pStr++; cbToken-=2; } + XMLSTR attrVal=(XMLSTR)token.pStr; + if (attrVal) + { + attrVal=fromXMLString(attrVal,cbToken,pXML); + if (!attrVal) return FALSE; + } + addAttribute_priv(MEMORYINCREASE,stringDup(lpszTemp,cbTemp),attrVal); + } + + // Indicate we are searching for a new attribute + attrib = eAttribName; + break; + + // Errors... + case eTokenTagStart: /* 'Attr = <' */ + case eTokenTagEnd: /* 'Attr = ' */ + case eTokenShortHandClose: /* "Attr = />" */ + case eTokenEquals: /* 'Attr = =' */ + case eTokenDeclaration: /* 'Attr = error = eXMLErrorUnexpectedToken; + return FALSE; + break; + default: break; + } + } + } + } + // If we failed to obtain the next token + else + { + if ((!d->isDeclaration)&&(d->pParent)) + { +#ifdef STRICT_PARSING + pXML->error=eXMLErrorUnmatchedEndTag; +#else + pXML->error=eXMLErrorMissingEndTag; +#endif + pXML->nIndexMissigEndTag=pXML->nIndex; + } + maybeAddTxT(pXML,pXML->lpXML+pXML->nIndex); + return FALSE; + } + } +} + +// Count the number of lines and columns in an XML string. +static void CountLinesAndColumns(XMLCSTR lpXML, int nUpto, XMLResults *pResults) +{ + XMLCHAR ch; + assert(lpXML); + assert(pResults); + + struct XML xml={ lpXML,lpXML, 0, 0, eXMLErrorNone, NULL, 0, NULL, 0, TRUE }; + + pResults->nLine = 1; + pResults->nColumn = 1; + while (xml.nIndexnColumn++; + else + { + pResults->nLine++; + pResults->nColumn=1; + } + } +} + +// Parse XML and return the root element. +XMLNode XMLNode::parseString(XMLCSTR lpszXML, XMLCSTR tag, XMLResults *pResults) +{ + if (!lpszXML) + { + if (pResults) + { + pResults->error=eXMLErrorNoElements; + pResults->nLine=0; + pResults->nColumn=0; + } + return emptyXMLNode; + } + + XMLNode xnode(NULL,NULL,FALSE); + struct XML xml={ lpszXML, lpszXML, 0, 0, eXMLErrorNone, NULL, 0, NULL, 0, TRUE }; + + // Create header element + xnode.ParseXMLElement(&xml); + enum XMLError error = xml.error; + if (!xnode.nChildNode()) error=eXMLErrorNoXMLTagFound; + if ((xnode.nChildNode()==1)&&(xnode.nElement()==1)) xnode=xnode.getChildNode(); // skip the empty node + + // If no error occurred + if ((error==eXMLErrorNone)||(error==eXMLErrorMissingEndTag)||(error==eXMLErrorNoXMLTagFound)) + { + XMLCSTR name=xnode.getName(); + if (tag&&(*tag)&&((!name)||(xstricmp(name,tag)))) + { + xnode=xnode.getChildNode(tag); + if (xnode.isEmpty()) + { + if (pResults) + { + pResults->error=eXMLErrorFirstTagNotFound; + pResults->nLine=0; + pResults->nColumn=0; + } + return emptyXMLNode; + } + } + } else + { + // Cleanup: this will destroy all the nodes + xnode = emptyXMLNode; + } + + + // If we have been given somewhere to place results + if (pResults) + { + pResults->error = error; + + // If we have an error + if (error!=eXMLErrorNone) + { + if (error==eXMLErrorMissingEndTag) xml.nIndex=xml.nIndexMissigEndTag; + // Find which line and column it starts on. + CountLinesAndColumns(xml.lpXML, xml.nIndex, pResults); + } + } + return xnode; +} + +XMLNode XMLNode::parseFile(XMLCSTR filename, XMLCSTR tag, XMLResults *pResults) +{ + if (pResults) { pResults->nLine=0; pResults->nColumn=0; } + FILE *f=xfopen(filename,_CXML("rb")); + if (f==NULL) { if (pResults) pResults->error=eXMLErrorFileNotFound; return emptyXMLNode; } + fseek(f,0,SEEK_END); + int l=ftell(f),headerSz=0; + if (!l) { if (pResults) pResults->error=eXMLErrorEmpty; fclose(f); return emptyXMLNode; } + fseek(f,0,SEEK_SET); + unsigned char *buf=(unsigned char*)malloc(l+4); + l=fread(buf,1,l,f); + fclose(f); + buf[l]=0;buf[l+1]=0;buf[l+2]=0;buf[l+3]=0; +#ifdef _XMLWIDECHAR + if (guessWideCharChars) + { + if (!myIsTextWideChar(buf,l)) + { + XMLNode::XMLCharEncoding ce=XMLNode::char_encoding_legacy; + if ((buf[0]==0xef)&&(buf[1]==0xbb)&&(buf[2]==0xbf)) { headerSz=3; ce=XMLNode::char_encoding_UTF8; } + XMLSTR b2=myMultiByteToWideChar((const char*)(buf+headerSz),ce); + free(buf); buf=(unsigned char*)b2; headerSz=0; + } else + { + if ((buf[0]==0xef)&&(buf[1]==0xff)) headerSz=2; + if ((buf[0]==0xff)&&(buf[1]==0xfe)) headerSz=2; + } + } +#else + if (guessWideCharChars) + { + if (myIsTextWideChar(buf,l)) + { + if ((buf[0]==0xef)&&(buf[1]==0xff)) headerSz=2; + if ((buf[0]==0xff)&&(buf[1]==0xfe)) headerSz=2; + char *b2=myWideCharToMultiByte((const wchar_t*)(buf+headerSz)); + free(buf); buf=(unsigned char*)b2; headerSz=0; + } else + { + if ((buf[0]==0xef)&&(buf[1]==0xbb)&&(buf[2]==0xbf)) headerSz=3; + } + } +#endif + + if (!buf) { if (pResults) pResults->error=eXMLErrorCharConversionError; return emptyXMLNode; } + XMLNode x=parseString((XMLSTR)(buf+headerSz),tag,pResults); + free(buf); + return x; +} + +static inline void charmemset(XMLSTR dest,XMLCHAR c,int l) { while (l--) *(dest++)=c; } +// private: +// Creates an user friendly XML string from a given element with +// appropriate white space and carriage returns. +// +// This recurses through all subnodes then adds contents of the nodes to the +// string. +int XMLNode::CreateXMLStringR(XMLNodeData *pEntry, XMLSTR lpszMarker, int nFormat) +{ + int nResult = 0; + int cb=nFormat<0?0:nFormat; + int cbElement; + int nChildFormat=-1; + int nElementI=pEntry->nChild+pEntry->nText+pEntry->nClear; + int i,j; + if ((nFormat>=0)&&(nElementI==1)&&(pEntry->nText==1)&&(!pEntry->isDeclaration)) nFormat=-2; + + assert(pEntry); + +#define LENSTR(lpsz) (lpsz ? xstrlen(lpsz) : 0) + + // If the element has no name then assume this is the head node. + cbElement = (int)LENSTR(pEntry->lpszName); + + if (cbElement) + { + // "isDeclaration) lpszMarker[nResult++]=_CXML('?'); + xstrcpy(&lpszMarker[nResult], pEntry->lpszName); + nResult+=cbElement; + lpszMarker[nResult++]=_CXML(' '); + + } else + { + nResult+=cbElement+2+cb; + if (pEntry->isDeclaration) nResult++; + } + + // Enumerate attributes and add them to the string + XMLAttribute *pAttr=pEntry->pAttribute; + for (i=0; inAttribute; i++) + { + // "Attrib + cb = (int)LENSTR(pAttr->lpszName); + if (cb) + { + if (lpszMarker) xstrcpy(&lpszMarker[nResult], pAttr->lpszName); + nResult += cb; + // "Attrib=Value " + if (pAttr->lpszValue) + { + cb=(int)ToXMLStringTool::lengthXMLString(pAttr->lpszValue); + if (lpszMarker) + { + lpszMarker[nResult]=_CXML('='); + lpszMarker[nResult+1]=_CXML('"'); + if (cb) ToXMLStringTool::toXMLUnSafe(&lpszMarker[nResult+2],pAttr->lpszValue); + lpszMarker[nResult+cb+2]=_CXML('"'); + } + nResult+=cb+3; + } + if (lpszMarker) lpszMarker[nResult] = _CXML(' '); + nResult++; + } + pAttr++; + } + + if (pEntry->isDeclaration) + { + if (lpszMarker) + { + lpszMarker[nResult-1]=_CXML('?'); + lpszMarker[nResult]=_CXML('>'); + } + nResult++; + if (nFormat!=-1) + { + if (lpszMarker) lpszMarker[nResult]=_CXML('\n'); + nResult++; + } + } else + // If there are child nodes we need to terminate the start tag + if (nElementI) + { + if (lpszMarker) lpszMarker[nResult-1]=_CXML('>'); + if (nFormat>=0) + { + if (lpszMarker) lpszMarker[nResult]=_CXML('\n'); + nResult++; + } + } else nResult--; + } + + // Calculate the child format for when we recurse. This is used to + // determine the number of spaces used for prefixes. + if (nFormat!=-1) + { + if (cbElement&&(!pEntry->isDeclaration)) nChildFormat=nFormat+1; + else nChildFormat=nFormat; + } + + // Enumerate through remaining children + for (i=0; ipOrder[i]; + switch((XMLElementType)(j&3)) + { + // Text nodes + case eNodeText: + { + // "Text" + XMLCSTR pChild=pEntry->pText[j>>2]; + cb = (int)ToXMLStringTool::lengthXMLString(pChild); + if (cb) + { + if (nFormat>=0) + { + if (lpszMarker) + { + charmemset(&lpszMarker[nResult],INDENTCHAR,nFormat+1); + ToXMLStringTool::toXMLUnSafe(&lpszMarker[nResult+nFormat+1],pChild); + lpszMarker[nResult+nFormat+1+cb]=_CXML('\n'); + } + nResult+=cb+nFormat+2; + } else + { + if (lpszMarker) ToXMLStringTool::toXMLUnSafe(&lpszMarker[nResult], pChild); + nResult += cb; + } + } + break; + } + + // Clear type nodes + case eNodeClear: + { + XMLClear *pChild=pEntry->pClear+(j>>2); + // "OpenTag" + cb = (int)LENSTR(pChild->lpszOpenTag); + if (cb) + { + if (nFormat!=-1) + { + if (lpszMarker) + { + charmemset(&lpszMarker[nResult], INDENTCHAR, nFormat+1); + xstrcpy(&lpszMarker[nResult+nFormat+1], pChild->lpszOpenTag); + } + nResult+=cb+nFormat+1; + } + else + { + if (lpszMarker)xstrcpy(&lpszMarker[nResult], pChild->lpszOpenTag); + nResult += cb; + } + } + + // "OpenTag Value" + cb = (int)LENSTR(pChild->lpszValue); + if (cb) + { + if (lpszMarker) xstrcpy(&lpszMarker[nResult], pChild->lpszValue); + nResult += cb; + } + + // "OpenTag Value CloseTag" + cb = (int)LENSTR(pChild->lpszCloseTag); + if (cb) + { + if (lpszMarker) xstrcpy(&lpszMarker[nResult], pChild->lpszCloseTag); + nResult += cb; + } + + if (nFormat!=-1) + { + if (lpszMarker) lpszMarker[nResult] = _CXML('\n'); + nResult++; + } + break; + } + + // Element nodes + case eNodeChild: + { + // Recursively add child nodes + nResult += CreateXMLStringR(pEntry->pChild[j>>2].d, lpszMarker ? lpszMarker + nResult : 0, nChildFormat); + break; + } + default: break; + } + } + + if ((cbElement)&&(!pEntry->isDeclaration)) + { + // If we have child entries we need to use long XML notation for + // closing the element - "blah blah blah" + if (nElementI) + { + // "\0" + if (lpszMarker) + { + if (nFormat >=0) + { + charmemset(&lpszMarker[nResult], INDENTCHAR,nFormat); + nResult+=nFormat; + } + + lpszMarker[nResult]=_CXML('<'); lpszMarker[nResult+1]=_CXML('/'); + nResult += 2; + xstrcpy(&lpszMarker[nResult], pEntry->lpszName); + nResult += cbElement; + + lpszMarker[nResult]=_CXML('>'); + if (nFormat == -1) nResult++; + else + { + lpszMarker[nResult+1]=_CXML('\n'); + nResult+=2; + } + } else + { + if (nFormat>=0) nResult+=cbElement+4+nFormat; + else if (nFormat==-1) nResult+=cbElement+3; + else nResult+=cbElement+4; + } + } else + { + // If there are no children we can use shorthand XML notation - + // "" + // "/>\0" + if (lpszMarker) + { + lpszMarker[nResult]=_CXML('/'); lpszMarker[nResult+1]=_CXML('>'); + if (nFormat != -1) lpszMarker[nResult+2]=_CXML('\n'); + } + nResult += nFormat == -1 ? 2 : 3; + } + } + + return nResult; +} + +#undef LENSTR + +// Create an XML string +// @param int nFormat - 0 if no formatting is required +// otherwise nonzero for formatted text +// with carriage returns and indentation. +// @param int *pnSize - [out] pointer to the size of the +// returned string not including the +// NULL terminator. +// @return XMLSTR - Allocated XML string, you must free +// this with free(). +XMLSTR XMLNode::createXMLString(int nFormat, int *pnSize) const +{ + if (!d) { if (pnSize) *pnSize=0; return NULL; } + + XMLSTR lpszResult = NULL; + int cbStr; + + // Recursively Calculate the size of the XML string + if (!dropWhiteSpace) nFormat=0; + nFormat = nFormat ? 0 : -1; + cbStr = CreateXMLStringR(d, 0, nFormat); + // Alllocate memory for the XML string + the NULL terminator and + // create the recursively XML string. + lpszResult=(XMLSTR)malloc((cbStr+1)*sizeof(XMLCHAR)); + CreateXMLStringR(d, lpszResult, nFormat); + lpszResult[cbStr]=_CXML('\0'); + if (pnSize) *pnSize = cbStr; + return lpszResult; +} + +int XMLNode::detachFromParent(XMLNodeData *d) +{ + XMLNode *pa=d->pParent->pChild; + int i=0; + while (((void*)(pa[i].d))!=((void*)d)) i++; + d->pParent->nChild--; + if (d->pParent->nChild) memmove(pa+i,pa+i+1,(d->pParent->nChild-i)*sizeof(XMLNode)); + else { free(pa); d->pParent->pChild=NULL; } + return removeOrderElement(d->pParent,eNodeChild,i); +} + +XMLNode::~XMLNode() +{ + if (!d) return; + d->ref_count--; + emptyTheNode(0); +} +void XMLNode::deleteNodeContent() +{ + if (!d) return; + if (d->pParent) { detachFromParent(d); d->pParent=NULL; d->ref_count--; } + emptyTheNode(1); +} +void XMLNode::emptyTheNode(char force) +{ + XMLNodeData *dd=d; // warning: must stay this way! + if ((dd->ref_count==0)||force) + { + if (d->pParent) detachFromParent(d); + int i; + XMLNode *pc; + for(i=0; inChild; i++) + { + pc=dd->pChild+i; + pc->d->pParent=NULL; + pc->d->ref_count--; + pc->emptyTheNode(force); + } + myFree(dd->pChild); + for(i=0; inText; i++) free((void*)dd->pText[i]); + myFree(dd->pText); + for(i=0; inClear; i++) free((void*)dd->pClear[i].lpszValue); + myFree(dd->pClear); + for(i=0; inAttribute; i++) + { + free((void*)dd->pAttribute[i].lpszName); + if (dd->pAttribute[i].lpszValue) free((void*)dd->pAttribute[i].lpszValue); + } + myFree(dd->pAttribute); + myFree(dd->pOrder); + myFree((void*)dd->lpszName); + dd->nChild=0; dd->nText=0; dd->nClear=0; dd->nAttribute=0; + dd->pChild=NULL; dd->pText=NULL; dd->pClear=NULL; dd->pAttribute=NULL; + dd->pOrder=NULL; dd->lpszName=NULL; dd->pParent=NULL; + } + if (dd->ref_count==0) + { + free(dd); + d=NULL; + } +} + +XMLNode& XMLNode::operator=( const XMLNode& A ) +{ + // shallow copy + if (this != &A) + { + if (d) { d->ref_count--; emptyTheNode(0); } + d=A.d; + if (d) (d->ref_count) ++ ; + } + return *this; +} + +XMLNode::XMLNode(const XMLNode &A) +{ + // shallow copy + d=A.d; + if (d) (d->ref_count)++ ; +} + +XMLNode XMLNode::deepCopy() const +{ + if (!d) return XMLNode::emptyXMLNode; + XMLNode x(NULL,stringDup(d->lpszName),d->isDeclaration); + XMLNodeData *p=x.d; + int n=d->nAttribute; + if (n) + { + p->nAttribute=n; p->pAttribute=(XMLAttribute*)malloc(n*sizeof(XMLAttribute)); + while (n--) + { + p->pAttribute[n].lpszName=stringDup(d->pAttribute[n].lpszName); + p->pAttribute[n].lpszValue=stringDup(d->pAttribute[n].lpszValue); + } + } + if (d->pOrder) + { + n=(d->nChild+d->nText+d->nClear)*sizeof(int); p->pOrder=(int*)malloc(n); memcpy(p->pOrder,d->pOrder,n); + } + n=d->nText; + if (n) + { + p->nText=n; p->pText=(XMLCSTR*)malloc(n*sizeof(XMLCSTR)); + while(n--) p->pText[n]=stringDup(d->pText[n]); + } + n=d->nClear; + if (n) + { + p->nClear=n; p->pClear=(XMLClear*)malloc(n*sizeof(XMLClear)); + while (n--) + { + p->pClear[n].lpszCloseTag=d->pClear[n].lpszCloseTag; + p->pClear[n].lpszOpenTag=d->pClear[n].lpszOpenTag; + p->pClear[n].lpszValue=stringDup(d->pClear[n].lpszValue); + } + } + n=d->nChild; + if (n) + { + p->nChild=n; p->pChild=(XMLNode*)malloc(n*sizeof(XMLNode)); + while (n--) + { + p->pChild[n].d=NULL; + p->pChild[n]=d->pChild[n].deepCopy(); + p->pChild[n].d->pParent=p; + } + } + return x; +} + +XMLNode XMLNode::addChild(XMLNode childNode, int pos) +{ + XMLNodeData *dc=childNode.d; + if ((!dc)||(!d)) return childNode; + if (!dc->lpszName) + { + // this is a root node: todo: correct fix + int j=pos; + while (dc->nChild) + { + addChild(dc->pChild[0],j); + if (pos>=0) j++; + } + return childNode; + } + if (dc->pParent) { if ((detachFromParent(dc)<=pos)&&(dc->pParent==d)) pos--; } else dc->ref_count++; + dc->pParent=d; +// int nc=d->nChild; +// d->pChild=(XMLNode*)myRealloc(d->pChild,(nc+1),memoryIncrease,sizeof(XMLNode)); + d->pChild=(XMLNode*)addToOrder(0,&pos,d->nChild,d->pChild,sizeof(XMLNode),eNodeChild); + d->pChild[pos].d=dc; + d->nChild++; + return childNode; +} + +void XMLNode::deleteAttribute(int i) +{ + if ((!d)||(i<0)||(i>=d->nAttribute)) return; + d->nAttribute--; + XMLAttribute *p=d->pAttribute+i; + free((void*)p->lpszName); + if (p->lpszValue) free((void*)p->lpszValue); + if (d->nAttribute) memmove(p,p+1,(d->nAttribute-i)*sizeof(XMLAttribute)); else { free(p); d->pAttribute=NULL; } +} + +void XMLNode::deleteAttribute(XMLAttribute *a){ if (a) deleteAttribute(a->lpszName); } +void XMLNode::deleteAttribute(XMLCSTR lpszName) +{ + int j=0; + getAttribute(lpszName,&j); + if (j) deleteAttribute(j-1); +} + +XMLAttribute *XMLNode::updateAttribute_WOSD(XMLSTR lpszNewValue, XMLSTR lpszNewName,int i) +{ + if (!d) { if (lpszNewValue) free(lpszNewValue); if (lpszNewName) free(lpszNewName); return NULL; } + if (i>=d->nAttribute) + { + if (lpszNewName) return addAttribute_WOSD(lpszNewName,lpszNewValue); + return NULL; + } + XMLAttribute *p=d->pAttribute+i; + if (p->lpszValue&&p->lpszValue!=lpszNewValue) free((void*)p->lpszValue); + p->lpszValue=lpszNewValue; + if (lpszNewName&&p->lpszName!=lpszNewName) { free((void*)p->lpszName); p->lpszName=lpszNewName; }; + return p; +} + +XMLAttribute *XMLNode::updateAttribute_WOSD(XMLAttribute *newAttribute, XMLAttribute *oldAttribute) +{ + if (oldAttribute) return updateAttribute_WOSD((XMLSTR)newAttribute->lpszValue,(XMLSTR)newAttribute->lpszName,oldAttribute->lpszName); + return addAttribute_WOSD((XMLSTR)newAttribute->lpszName,(XMLSTR)newAttribute->lpszValue); +} + +XMLAttribute *XMLNode::updateAttribute_WOSD(XMLSTR lpszNewValue, XMLSTR lpszNewName,XMLCSTR lpszOldName) +{ + int j=0; + getAttribute(lpszOldName,&j); + if (j) return updateAttribute_WOSD(lpszNewValue,lpszNewName,j-1); + else + { + if (lpszNewName) return addAttribute_WOSD(lpszNewName,lpszNewValue); + else return addAttribute_WOSD(stringDup(lpszOldName),lpszNewValue); + } +} + +int XMLNode::indexText(XMLCSTR lpszValue) const +{ + if (!d) return -1; + int i,l=d->nText; + if (!lpszValue) { if (l) return 0; return -1; } + XMLCSTR *p=d->pText; + for (i=0; i=d->nText)) return; + d->nText--; + XMLCSTR *p=d->pText+i; + free((void*)*p); + if (d->nText) memmove(p,p+1,(d->nText-i)*sizeof(XMLCSTR)); else { free(p); d->pText=NULL; } + removeOrderElement(d,eNodeText,i); +} + +void XMLNode::deleteText(XMLCSTR lpszValue) { deleteText(indexText(lpszValue)); } + +XMLCSTR XMLNode::updateText_WOSD(XMLSTR lpszNewValue, int i) +{ + if (!d) { if (lpszNewValue) free(lpszNewValue); return NULL; } + if (i>=d->nText) return addText_WOSD(lpszNewValue); + XMLCSTR *p=d->pText+i; + if (*p!=lpszNewValue) { free((void*)*p); *p=lpszNewValue; } + return lpszNewValue; +} + +XMLCSTR XMLNode::updateText_WOSD(XMLSTR lpszNewValue, XMLCSTR lpszOldValue) +{ + if (!d) { if (lpszNewValue) free(lpszNewValue); return NULL; } + int i=indexText(lpszOldValue); + if (i>=0) return updateText_WOSD(lpszNewValue,i); + return addText_WOSD(lpszNewValue); +} + +void XMLNode::deleteClear(int i) +{ + if ((!d)||(i<0)||(i>=d->nClear)) return; + d->nClear--; + XMLClear *p=d->pClear+i; + free((void*)p->lpszValue); + if (d->nClear) memmove(p,p+1,(d->nClear-i)*sizeof(XMLClear)); else { free(p); d->pClear=NULL; } + removeOrderElement(d,eNodeClear,i); +} + +int XMLNode::indexClear(XMLCSTR lpszValue) const +{ + if (!d) return -1; + int i,l=d->nClear; + if (!lpszValue) { if (l) return 0; return -1; } + XMLClear *p=d->pClear; + for (i=0; ilpszValue); } + +XMLClear *XMLNode::updateClear_WOSD(XMLSTR lpszNewContent, int i) +{ + if (!d) { if (lpszNewContent) free(lpszNewContent); return NULL; } + if (i>=d->nClear) return addClear_WOSD(lpszNewContent); + XMLClear *p=d->pClear+i; + if (lpszNewContent!=p->lpszValue) { free((void*)p->lpszValue); p->lpszValue=lpszNewContent; } + return p; +} + +XMLClear *XMLNode::updateClear_WOSD(XMLSTR lpszNewContent, XMLCSTR lpszOldValue) +{ + if (!d) { if (lpszNewContent) free(lpszNewContent); return NULL; } + int i=indexClear(lpszOldValue); + if (i>=0) return updateClear_WOSD(lpszNewContent,i); + return addClear_WOSD(lpszNewContent); +} + +XMLClear *XMLNode::updateClear_WOSD(XMLClear *newP,XMLClear *oldP) +{ + if (oldP) return updateClear_WOSD((XMLSTR)newP->lpszValue,(XMLSTR)oldP->lpszValue); + return NULL; +} + +int XMLNode::nChildNode(XMLCSTR name) const +{ + if (!d) return 0; + int i,j=0,n=d->nChild; + XMLNode *pc=d->pChild; + for (i=0; id->lpszName, name)==0) j++; + pc++; + } + return j; +} + +XMLNode XMLNode::getChildNode(XMLCSTR name, int *j) const +{ + if (!d) return emptyXMLNode; + int i=0,n=d->nChild; + if (j) i=*j; + XMLNode *pc=d->pChild+i; + for (; id->lpszName, name)) + { + if (j) *j=i+1; + return *pc; + } + pc++; + } + return emptyXMLNode; +} + +XMLNode XMLNode::getChildNode(XMLCSTR name, int j) const +{ + if (!d) return emptyXMLNode; + if (j>=0) + { + int i=0; + while (j-->0) getChildNode(name,&i); + return getChildNode(name,&i); + } + int i=d->nChild; + while (i--) if (!xstricmp(name,d->pChild[i].d->lpszName)) break; + if (i<0) return emptyXMLNode; + return getChildNode(i); +} + +XMLNode XMLNode::getChildNodeByPath(XMLCSTR _path, char createMissing, XMLCHAR sep) +{ + XMLSTR path=stringDup(_path); + XMLNode x=getChildNodeByPathNonConst(path,createMissing,sep); + if (path) free(path); + return x; +} + +XMLNode XMLNode::getChildNodeByPathNonConst(XMLSTR path, char createIfMissing, XMLCHAR sep) +{ + if ((!path)||(!(*path))) return *this; + XMLNode xn,xbase=*this; + XMLCHAR *tend1,sepString[2]; sepString[0]=sep; sepString[1]=0; + tend1=xstrstr(path,sepString); + while(tend1) + { + *tend1=0; + xn=xbase.getChildNode(path); + *tend1=sep; + if (xn.isEmpty()) + { + if (createIfMissing) xn=xbase.addChild(path); + else return XMLNode::emptyXMLNode; + } + xbase=xn; + path=tend1+1; + tend1=xstrstr(path,sepString); + } + xn=xbase.getChildNode(path); + if (xn.isEmpty()&&createIfMissing) xn=xbase.addChild(path); + return xn; +} + +XMLElementPosition XMLNode::positionOfText (int i) const { if (i>=d->nText ) i=d->nText-1; return findPosition(d,i,eNodeText ); } +XMLElementPosition XMLNode::positionOfClear (int i) const { if (i>=d->nClear) i=d->nClear-1; return findPosition(d,i,eNodeClear); } +XMLElementPosition XMLNode::positionOfChildNode(int i) const { if (i>=d->nChild) i=d->nChild-1; return findPosition(d,i,eNodeChild); } +XMLElementPosition XMLNode::positionOfText (XMLCSTR lpszValue) const { return positionOfText (indexText (lpszValue)); } +XMLElementPosition XMLNode::positionOfClear(XMLCSTR lpszValue) const { return positionOfClear(indexClear(lpszValue)); } +XMLElementPosition XMLNode::positionOfClear(XMLClear *a) const { if (a) return positionOfClear(a->lpszValue); return positionOfClear(); } +XMLElementPosition XMLNode::positionOfChildNode(XMLNode x) const +{ + if ((!d)||(!x.d)) return -1; + XMLNodeData *dd=x.d; + XMLNode *pc=d->pChild; + int i=d->nChild; + while (i--) if (pc[i].d==dd) return findPosition(d,i,eNodeChild); + return -1; +} +XMLElementPosition XMLNode::positionOfChildNode(XMLCSTR name, int count) const +{ + if (!name) return positionOfChildNode(count); + int j=0; + do { getChildNode(name,&j); if (j<0) return -1; } while (count--); + return findPosition(d,j-1,eNodeChild); +} + +XMLNode XMLNode::getChildNodeWithAttribute(XMLCSTR name,XMLCSTR attributeName,XMLCSTR attributeValue, int *k) const +{ + int i=0,j; + if (k) i=*k; + XMLNode x; + XMLCSTR t; + do + { + x=getChildNode(name,&i); + if (!x.isEmpty()) + { + if (attributeValue) + { + j=0; + do + { + t=x.getAttribute(attributeName,&j); + if (t&&(xstricmp(attributeValue,t)==0)) { if (k) *k=i; return x; } + } while (t); + } else + { + if (x.isAttributeSet(attributeName)) { if (k) *k=i; return x; } + } + } + } while (!x.isEmpty()); + return emptyXMLNode; +} + +// Find an attribute on an node. +XMLCSTR XMLNode::getAttribute(XMLCSTR lpszAttrib, int *j) const +{ + if (!d) return NULL; + int i=0,n=d->nAttribute; + if (j) i=*j; + XMLAttribute *pAttr=d->pAttribute+i; + for (; ilpszName, lpszAttrib)==0) + { + if (j) *j=i+1; + return pAttr->lpszValue; + } + pAttr++; + } + return NULL; +} + +char XMLNode::isAttributeSet(XMLCSTR lpszAttrib) const +{ + if (!d) return FALSE; + int i,n=d->nAttribute; + XMLAttribute *pAttr=d->pAttribute; + for (i=0; ilpszName, lpszAttrib)==0) + { + return TRUE; + } + pAttr++; + } + return FALSE; +} + +XMLCSTR XMLNode::getAttribute(XMLCSTR name, int j) const +{ + if (!d) return NULL; + int i=0; + while (j-->0) getAttribute(name,&i); + return getAttribute(name,&i); +} + +XMLNodeContents XMLNode::enumContents(int i) const +{ + XMLNodeContents c; + if (!d) { c.etype=eNodeNULL; return c; } + if (inAttribute) + { + c.etype=eNodeAttribute; + c.attrib=d->pAttribute[i]; + return c; + } + i-=d->nAttribute; + c.etype=(XMLElementType)(d->pOrder[i]&3); + i=(d->pOrder[i])>>2; + switch (c.etype) + { + case eNodeChild: c.child = d->pChild[i]; break; + case eNodeText: c.text = d->pText[i]; break; + case eNodeClear: c.clear = d->pClear[i]; break; + default: break; + } + return c; +} + +XMLCSTR XMLNode::getName() const { if (!d) return NULL; return d->lpszName; } +int XMLNode::nText() const { if (!d) return 0; return d->nText; } +int XMLNode::nChildNode() const { if (!d) return 0; return d->nChild; } +int XMLNode::nAttribute() const { if (!d) return 0; return d->nAttribute; } +int XMLNode::nClear() const { if (!d) return 0; return d->nClear; } +int XMLNode::nElement() const { if (!d) return 0; return d->nAttribute+d->nChild+d->nText+d->nClear; } +XMLClear XMLNode::getClear (int i) const { if ((!d)||(i>=d->nClear )) return emptyXMLClear; return d->pClear[i]; } +XMLAttribute XMLNode::getAttribute (int i) const { if ((!d)||(i>=d->nAttribute)) return emptyXMLAttribute; return d->pAttribute[i]; } +XMLCSTR XMLNode::getAttributeName (int i) const { if ((!d)||(i>=d->nAttribute)) return NULL; return d->pAttribute[i].lpszName; } +XMLCSTR XMLNode::getAttributeValue(int i) const { if ((!d)||(i>=d->nAttribute)) return NULL; return d->pAttribute[i].lpszValue; } +XMLCSTR XMLNode::getText (int i) const { if ((!d)||(i>=d->nText )) return NULL; return d->pText[i]; } +XMLNode XMLNode::getChildNode (int i) const { if ((!d)||(i>=d->nChild )) return emptyXMLNode; return d->pChild[i]; } +XMLNode XMLNode::getParentNode ( ) const { if ((!d)||(!d->pParent )) return emptyXMLNode; return XMLNode(d->pParent); } +char XMLNode::isDeclaration ( ) const { if (!d) return 0; return d->isDeclaration; } +char XMLNode::isEmpty ( ) const { return (d==NULL); } +XMLNode XMLNode::emptyNode ( ) { return XMLNode::emptyXMLNode; } + +XMLNode XMLNode::addChild(XMLCSTR lpszName, char isDeclaration, XMLElementPosition pos) + { return addChild_priv(0,stringDup(lpszName),isDeclaration,pos); } +XMLNode XMLNode::addChild_WOSD(XMLSTR lpszName, char isDeclaration, XMLElementPosition pos) + { return addChild_priv(0,lpszName,isDeclaration,pos); } +XMLAttribute *XMLNode::addAttribute(XMLCSTR lpszName, XMLCSTR lpszValue) + { return addAttribute_priv(0,stringDup(lpszName),stringDup(lpszValue)); } +XMLAttribute *XMLNode::addAttribute_WOSD(XMLSTR lpszName, XMLSTR lpszValuev) + { return addAttribute_priv(0,lpszName,lpszValuev); } +XMLCSTR XMLNode::addText(XMLCSTR lpszValue, XMLElementPosition pos) + { return addText_priv(0,stringDup(lpszValue),pos); } +XMLCSTR XMLNode::addText_WOSD(XMLSTR lpszValue, XMLElementPosition pos) + { return addText_priv(0,lpszValue,pos); } +XMLClear *XMLNode::addClear(XMLCSTR lpszValue, XMLCSTR lpszOpen, XMLCSTR lpszClose, XMLElementPosition pos) + { return addClear_priv(0,stringDup(lpszValue),lpszOpen,lpszClose,pos); } +XMLClear *XMLNode::addClear_WOSD(XMLSTR lpszValue, XMLCSTR lpszOpen, XMLCSTR lpszClose, XMLElementPosition pos) + { return addClear_priv(0,lpszValue,lpszOpen,lpszClose,pos); } +XMLCSTR XMLNode::updateName(XMLCSTR lpszName) + { return updateName_WOSD(stringDup(lpszName)); } +XMLAttribute *XMLNode::updateAttribute(XMLAttribute *newAttribute, XMLAttribute *oldAttribute) + { return updateAttribute_WOSD(stringDup(newAttribute->lpszValue),stringDup(newAttribute->lpszName),oldAttribute->lpszName); } +XMLAttribute *XMLNode::updateAttribute(XMLCSTR lpszNewValue, XMLCSTR lpszNewName,int i) + { return updateAttribute_WOSD(stringDup(lpszNewValue),stringDup(lpszNewName),i); } +XMLAttribute *XMLNode::updateAttribute(XMLCSTR lpszNewValue, XMLCSTR lpszNewName,XMLCSTR lpszOldName) + { return updateAttribute_WOSD(stringDup(lpszNewValue),stringDup(lpszNewName),lpszOldName); } +XMLCSTR XMLNode::updateText(XMLCSTR lpszNewValue, int i) + { return updateText_WOSD(stringDup(lpszNewValue),i); } +XMLCSTR XMLNode::updateText(XMLCSTR lpszNewValue, XMLCSTR lpszOldValue) + { return updateText_WOSD(stringDup(lpszNewValue),lpszOldValue); } +XMLClear *XMLNode::updateClear(XMLCSTR lpszNewContent, int i) + { return updateClear_WOSD(stringDup(lpszNewContent),i); } +XMLClear *XMLNode::updateClear(XMLCSTR lpszNewValue, XMLCSTR lpszOldValue) + { return updateClear_WOSD(stringDup(lpszNewValue),lpszOldValue); } +XMLClear *XMLNode::updateClear(XMLClear *newP,XMLClear *oldP) + { return updateClear_WOSD(stringDup(newP->lpszValue),oldP->lpszValue); } + +char XMLNode::setGlobalOptions(XMLCharEncoding _characterEncoding, char _guessWideCharChars, + char _dropWhiteSpace, char _removeCommentsInMiddleOfText) +{ + guessWideCharChars=_guessWideCharChars; dropWhiteSpace=_dropWhiteSpace; removeCommentsInMiddleOfText=_removeCommentsInMiddleOfText; +#ifdef _XMLWIDECHAR + if (_characterEncoding) characterEncoding=_characterEncoding; +#else + switch(_characterEncoding) + { + case char_encoding_UTF8: characterEncoding=_characterEncoding; XML_ByteTable=XML_utf8ByteTable; break; + case char_encoding_legacy: characterEncoding=_characterEncoding; XML_ByteTable=XML_legacyByteTable; break; + case char_encoding_ShiftJIS: characterEncoding=_characterEncoding; XML_ByteTable=XML_sjisByteTable; break; + case char_encoding_GB2312: characterEncoding=_characterEncoding; XML_ByteTable=XML_gb2312ByteTable; break; + case char_encoding_Big5: + case char_encoding_GBK: characterEncoding=_characterEncoding; XML_ByteTable=XML_gbk_big5_ByteTable; break; + default: return 1; + } +#endif + return 0; +} + +XMLNode::XMLCharEncoding XMLNode::guessCharEncoding(void *buf,int l, char useXMLEncodingAttribute) +{ +#ifdef _XMLWIDECHAR + return (XMLCharEncoding)0; +#else + if (l<25) return (XMLCharEncoding)0; + if (guessWideCharChars&&(myIsTextWideChar(buf,l))) return (XMLCharEncoding)0; + unsigned char *b=(unsigned char*)buf; + if ((b[0]==0xef)&&(b[1]==0xbb)&&(b[2]==0xbf)) return char_encoding_UTF8; + + // Match utf-8 model ? + XMLCharEncoding bestGuess=char_encoding_UTF8; + int i=0; + while (i>2 ]; + *(curr++)=base64EncodeTable[(inbuf[0]<<4)&0x3F]; + *(curr++)=base64Fillchar; + *(curr++)=base64Fillchar; + } else if (eLen==2) + { + j=(inbuf[0]<<8)|inbuf[1]; + *(curr++)=base64EncodeTable[ j>>10 ]; + *(curr++)=base64EncodeTable[(j>> 4)&0x3f]; + *(curr++)=base64EncodeTable[(j<< 2)&0x3f]; + *(curr++)=base64Fillchar; + } + *(curr++)=0; + return (XMLSTR)buf; +} + +unsigned int XMLParserBase64Tool::decodeSize(XMLCSTR data,XMLError *xe) +{ + if (xe) *xe=eXMLErrorNone; + int size=0; + unsigned char c; + //skip any extra characters (e.g. newlines or spaces) + while (*data) + { +#ifdef _XMLWIDECHAR + if (*data>255) { if (xe) *xe=eXMLErrorBase64DecodeIllegalCharacter; return 0; } +#endif + c=base64DecodeTable[(unsigned char)(*data)]; + if (c<97) size++; + else if (c==98) { if (xe) *xe=eXMLErrorBase64DecodeIllegalCharacter; return 0; } + data++; + } + if (xe&&(size%4!=0)) *xe=eXMLErrorBase64DataSizeIsNotMultipleOf4; + if (size==0) return 0; + do { data--; size--; } while(*data==base64Fillchar); size++; + return (unsigned int)((size*3)/4); +} + +unsigned char XMLParserBase64Tool::decode(XMLCSTR data, unsigned char *buf, int len, XMLError *xe) +{ + if (xe) *xe=eXMLErrorNone; + int i=0,p=0; + unsigned char d,c; + for(;;) + { + +#ifdef _XMLWIDECHAR +#define BASE64DECODE_READ_NEXT_CHAR(c) \ + do { \ + if (data[i]>255){ c=98; break; } \ + c=base64DecodeTable[(unsigned char)data[i++]]; \ + }while (c==97); \ + if(c==98){ if(xe)*xe=eXMLErrorBase64DecodeIllegalCharacter; return 0; } +#else +#define BASE64DECODE_READ_NEXT_CHAR(c) \ + do { c=base64DecodeTable[(unsigned char)data[i++]]; }while (c==97); \ + if(c==98){ if(xe)*xe=eXMLErrorBase64DecodeIllegalCharacter; return 0; } +#endif + + BASE64DECODE_READ_NEXT_CHAR(c) + if (c==99) { return 2; } + if (c==96) + { + if (p==(int)len) return 2; + if (xe) *xe=eXMLErrorBase64DecodeTruncatedData; + return 1; + } + + BASE64DECODE_READ_NEXT_CHAR(d) + if ((d==99)||(d==96)) { if (xe) *xe=eXMLErrorBase64DecodeTruncatedData; return 1; } + if (p==(int)len) { if (xe) *xe=eXMLErrorBase64DecodeBufferTooSmall; return 0; } + buf[p++]=(unsigned char)((c<<2)|((d>>4)&0x3)); + + BASE64DECODE_READ_NEXT_CHAR(c) + if (c==99) { if (xe) *xe=eXMLErrorBase64DecodeTruncatedData; return 1; } + if (p==(int)len) + { + if (c==96) return 2; + if (xe) *xe=eXMLErrorBase64DecodeBufferTooSmall; + return 0; + } + if (c==96) { if (xe) *xe=eXMLErrorBase64DecodeTruncatedData; return 1; } + buf[p++]=(unsigned char)(((d<<4)&0xf0)|((c>>2)&0xf)); + + BASE64DECODE_READ_NEXT_CHAR(d) + if (d==99 ) { if (xe) *xe=eXMLErrorBase64DecodeTruncatedData; return 1; } + if (p==(int)len) + { + if (d==96) return 2; + if (xe) *xe=eXMLErrorBase64DecodeBufferTooSmall; + return 0; + } + if (d==96) { if (xe) *xe=eXMLErrorBase64DecodeTruncatedData; return 1; } + buf[p++]=(unsigned char)(((c<<6)&0xc0)|d); + } +} +#undef BASE64DECODE_READ_NEXT_CHAR + +void XMLParserBase64Tool::alloc(int newsize) +{ + if ((!buf)&&(newsize)) { buf=malloc(newsize); buflen=newsize; return; } + if (newsize>buflen) { buf=realloc(buf,newsize); buflen=newsize; } +} + +unsigned char *XMLParserBase64Tool::decode(XMLCSTR data, int *outlen, XMLError *xe) +{ + if (xe) *xe=eXMLErrorNone; + unsigned int len=decodeSize(data,xe); + if (outlen) *outlen=len; + if (!len) return NULL; + alloc(len+1); + if(!decode(data,(unsigned char*)buf,len,xe)){ return NULL; } + return (unsigned char*)buf; +} + diff --git a/foreign/xmlParser/xmlParser.hpp b/foreign/xmlParser/xmlParser.hpp new file mode 100644 index 0000000..321eed3 --- /dev/null +++ b/foreign/xmlParser/xmlParser.hpp @@ -0,0 +1,762 @@ +/****************************************************************************/ +/*! \mainpage XMLParser library + * \section intro_sec Introduction + * + * This is a basic XML parser written in ANSI C++ for portability. + * It works by using recursion and a node tree for breaking + * down the elements of an XML document. + + * Copyright (c) 2002, Frank Vanden Berghen + * All rights reserved. + * + * The following license terms apply to projects that are in some way related to + * the "ZeroMQ project", including applications + * using "ZeroMQ project" and tools developed + * for enhancing "ZeroMQ project". All other projects + * (not related to "ZeroMQ project") have to use this + * code under the Aladdin Free Public License (AFPL) + * See the file "AFPL-license.txt" for more informations about the AFPL license. + * (see http://www.artifex.com/downloads/doc/Public.htm for detailed AFPL terms) + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are met: + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * * Neither the name of Frank Vanden Berghen nor the + * names of its contributors may be used to endorse or promote products + * derived from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY Frank Vanden Berghen ``AS IS'' AND ANY + * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED + * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE + * DISCLAIMED. IN NO EVENT SHALL BE LIABLE FOR ANY + * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES + * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; + * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND + * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS + * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * @version V2.39 + * @author Frank Vanden Berghen + * + * \section tutorial First Tutorial + * You can follow a simple Tutorial to know the basics... + * + * \section usage General usage: How to include the XMLParser library inside your project. + * + * The library is composed of two files: xmlParser.cpp and + * xmlParser.h. These are the ONLY 2 files that you need when + * using the library inside your own projects. + * + * All the functions of the library are documented inside the comments of the file + * xmlParser.h. These comments can be transformed in + * full-fledged HTML documentation using the DOXYGEN software: simply type: "doxygen doxy.cfg" + * + * By default, the XMLParser library uses (char*) for string representation.To use the (wchar_t*) + * version of the library, you need to define the "_UNICODE" preprocessor definition variable + * (this is usually done inside your project definition file) (This is done automatically for you + * when using Visual Studio). + * + * \section example Advanced Tutorial and Many Examples of usage. + * + * Some very small introductory examples are described inside the Tutorial file + * xmlParser.html + * + * Some additional small examples are also inside the file xmlTest.cpp + * (for the "char*" version of the library) and inside the file + * xmlTestUnicode.cpp (for the "wchar_t*" + * version of the library). If you have a question, please review these additionnal examples + * before sending an e-mail to the author. + * + * To build the examples: + * - linux/unix: type "make" + * - solaris: type "make -f makefile.solaris" + * - windows: Visual Studio: double-click on xmlParser.dsw + * (under Visual Studio .NET, the .dsp and .dsw files will be automatically converted to .vcproj and .sln files) + * + * In order to build the examples you need some additional files: + * - linux/unix: makefile + * - solaris: makefile.solaris + * - windows: Visual Studio: *.dsp, xmlParser.dsw and also xmlParser.lib and xmlParser.dll + * + * \section debugging Debugging with the XMLParser library + * + * \subsection debugwin Debugging under WINDOWS + * + * Inside Visual C++, the "debug versions" of the memory allocation functions are + * very slow: Do not forget to compile in "release mode" to get maximum speed. + * When I have to debug a software that is using the XMLParser Library, it was usually + * a nightmare because the library was sooOOOoooo slow in debug mode (because of the + * slow memory allocations in Debug mode). To solve this + * problem, during all the debugging session, I use a very fast DLL version of the + * XMLParser Library (the DLL is compiled in release mode). Using the DLL version of + * the XMLParser Library allows me to have lightening XML parsing speed even in debug! + * Other than that, the DLL version is useless: In the release version of my tool, + * I always use the normal, ".cpp"-based, XMLParser Library (I simply include the + * xmlParser.cpp and + * xmlParser.h files into the project). + * + * The file XMLNodeAutoexp.txt contains some + * "tweaks" that improve substancially the display of the content of the XMLNode objects + * inside the Visual Studio Debugger. Believe me, once you have seen inside the debugger + * the "smooth" display of the XMLNode objects, you cannot live without it anymore! + * + * \subsection debuglinux Debugging under LINUX/UNIX + * + * The speed of the debug version of the XMLParser library is tolerable so no extra + * work.has been done. + * + ****************************************************************************/ + +#ifndef __INCLUDE_XML_NODE__ +#define __INCLUDE_XML_NODE__ + +#include + +#ifdef _UNICODE +// If you comment the next "define" line then the library will never "switch to" _UNICODE (wchar_t*) mode (16/32 bits per characters). +// This is useful when you get error messages like: +// 'XMLNode::openFileHelper' : cannot convert parameter 2 from 'const char [5]' to 'const wchar_t *' +// The _XMLWIDECHAR preprocessor variable force the XMLParser library into either utf16/32-mode (the proprocessor variable +// must be defined) or utf8-mode(the pre-processor variable must be undefined). +#define _XMLWIDECHAR +#endif + +#if defined(WIN32) || defined(UNDER_CE) || defined(_WIN32) || defined(WIN64) || defined(__BORLANDC__) +// comment the next line if you are under windows and the compiler is not Microsoft Visual Studio (6.0 or .NET) or Borland +#define _XMLWINDOWS +#endif + +#ifdef XMLDLLENTRY +#undef XMLDLLENTRY +#endif +#ifdef _USE_XMLPARSER_DLL +#ifdef _DLL_EXPORTS_ +#define XMLDLLENTRY __declspec(dllexport) +#else +#define XMLDLLENTRY __declspec(dllimport) +#endif +#else +#define XMLDLLENTRY +#endif + +// uncomment the next line if you want no support for wchar_t* (no need for the or libraries anymore to compile) +//#define XML_NO_WIDE_CHAR + +#ifdef XML_NO_WIDE_CHAR +#undef _XMLWINDOWS +#undef _XMLWIDECHAR +#endif + +#ifdef _XMLWINDOWS +#include +#else +#define XMLDLLENTRY +#ifndef XML_NO_WIDE_CHAR +#include // to have 'wcsrtombs' for ANSI version + // to have 'mbsrtowcs' for WIDECHAR version +#endif +#endif + +// Some common types for char set portable code +#ifdef _XMLWIDECHAR + #define _CXML(c) L ## c + #define XMLCSTR const wchar_t * + #define XMLSTR wchar_t * + #define XMLCHAR wchar_t +#else + #define _CXML(c) c + #define XMLCSTR const char * + #define XMLSTR char * + #define XMLCHAR char +#endif +#ifndef FALSE + #define FALSE 0 +#endif /* FALSE */ +#ifndef TRUE + #define TRUE 1 +#endif /* TRUE */ + + +/// Enumeration for XML parse errors. +typedef enum XMLError +{ + eXMLErrorNone = 0, + eXMLErrorMissingEndTag, + eXMLErrorNoXMLTagFound, + eXMLErrorEmpty, + eXMLErrorMissingTagName, + eXMLErrorMissingEndTagName, + eXMLErrorUnmatchedEndTag, + eXMLErrorUnmatchedEndClearTag, + eXMLErrorUnexpectedToken, + eXMLErrorNoElements, + eXMLErrorFileNotFound, + eXMLErrorFirstTagNotFound, + eXMLErrorUnknownCharacterEntity, + eXMLErrorCharacterCodeAbove255, + eXMLErrorCharConversionError, + eXMLErrorCannotOpenWriteFile, + eXMLErrorCannotWriteFile, + + eXMLErrorBase64DataSizeIsNotMultipleOf4, + eXMLErrorBase64DecodeIllegalCharacter, + eXMLErrorBase64DecodeTruncatedData, + eXMLErrorBase64DecodeBufferTooSmall +} XMLError; + + +/// Enumeration used to manage type of data. Use in conjunction with structure XMLNodeContents +typedef enum XMLElementType +{ + eNodeChild=0, + eNodeAttribute=1, + eNodeText=2, + eNodeClear=3, + eNodeNULL=4 +} XMLElementType; + +/// Structure used to obtain error details if the parse fails. +typedef struct XMLResults +{ + enum XMLError error; + int nLine,nColumn; +} XMLResults; + +/// Structure for XML clear (unformatted) node (usually comments) +typedef struct XMLClear { + XMLCSTR lpszValue; XMLCSTR lpszOpenTag; XMLCSTR lpszCloseTag; +} XMLClear; + +/// Structure for XML attribute. +typedef struct XMLAttribute { + XMLCSTR lpszName; XMLCSTR lpszValue; +} XMLAttribute; + +/// XMLElementPosition are not interchangeable with simple indexes +typedef int XMLElementPosition; + +struct XMLNodeContents; + +/** @defgroup XMLParserGeneral The XML parser */ + +/// Main Class representing a XML node +/** + * All operations are performed using this class. + * \note The constructors of the XMLNode class are protected, so use instead one of these four methods to get your first instance of XMLNode: + *
    + *
  • XMLNode::parseString
    • + *
  • XMLNode::parseFile
    • + *
  • XMLNode::openFileHelper
    • + *
  • XMLNode::createXMLTopNode (or XMLNode::createXMLTopNode_WOSD)
    • + *
*/ +typedef struct XMLDLLENTRY XMLNode +{ + private: + + struct XMLNodeDataTag; + + /// Constructors are protected, so use instead one of: XMLNode::parseString, XMLNode::parseFile, XMLNode::openFileHelper, XMLNode::createXMLTopNode + XMLNode(struct XMLNodeDataTag *pParent, XMLSTR lpszName, char isDeclaration); + /// Constructors are protected, so use instead one of: XMLNode::parseString, XMLNode::parseFile, XMLNode::openFileHelper, XMLNode::createXMLTopNode + XMLNode(struct XMLNodeDataTag *p); + + public: + static XMLCSTR getVersion();///< Return the XMLParser library version number + + /** @defgroup conversions Parsing XML files/strings to an XMLNode structure and Rendering XMLNode's to files/string. + * @ingroup XMLParserGeneral + * @{ */ + + /// Parse an XML string and return the root of a XMLNode tree representing the string. + static XMLNode parseString (XMLCSTR lpXMLString, XMLCSTR tag=NULL, XMLResults *pResults=NULL); + /**< The "parseString" function parse an XML string and return the root of a XMLNode tree. The "opposite" of this function is + * the function "createXMLString" that re-creates an XML string from an XMLNode tree. If the XML document is corrupted, the + * "parseString" method will initialize the "pResults" variable with some information that can be used to trace the error. + * If you still want to parse the file, you can use the APPROXIMATE_PARSING option as explained inside the note at the + * beginning of the "xmlParser.cpp" file. + * + * @param lpXMLString the XML string to parse + * @param tag the name of the first tag inside the XML file. If the tag parameter is omitted, this function returns a node that represents the head of the xml document including the declaration term (). + * @param pResults a pointer to a XMLResults variable that will contain some information that can be used to trace the XML parsing error. You can have a user-friendly explanation of the parsing error with the "getError" function. + */ + + /// Parse an XML file and return the root of a XMLNode tree representing the file. + static XMLNode parseFile (XMLCSTR filename, XMLCSTR tag=NULL, XMLResults *pResults=NULL); + /**< The "parseFile" function parse an XML file and return the root of a XMLNode tree. The "opposite" of this function is + * the function "writeToFile" that re-creates an XML file from an XMLNode tree. If the XML document is corrupted, the + * "parseFile" method will initialize the "pResults" variable with some information that can be used to trace the error. + * If you still want to parse the file, you can use the APPROXIMATE_PARSING option as explained inside the note at the + * beginning of the "xmlParser.cpp" file. + * + * @param filename the path to the XML file to parse + * @param tag the name of the first tag inside the XML file. If the tag parameter is omitted, this function returns a node that represents the head of the xml document including the declaration term (). + * @param pResults a pointer to a XMLResults variable that will contain some information that can be used to trace the XML parsing error. You can have a user-friendly explanation of the parsing error with the "getError" function. + */ + + /// Parse an XML file and return the root of a XMLNode tree representing the file. A very crude error checking is made. An attempt to guess the Char Encoding used in the file is made. + static XMLNode openFileHelper(XMLCSTR filename, XMLCSTR tag=NULL); + /**< The "openFileHelper" function reports to the screen all the warnings and errors that occurred during parsing of the XML file. + * This function also tries to guess char Encoding (UTF-8, ASCII or SHIT-JIS) based on the first 200 bytes of the file. Since each + * application has its own way to report and deal with errors, you should rather use the "parseFile" function to parse XML files + * and program yourself thereafter an "error reporting" tailored for your needs (instead of using the very crude "error reporting" + * mechanism included inside the "openFileHelper" function). + * + * If the XML document is corrupted, the "openFileHelper" method will: + * - display an error message on the console (or inside a messageBox for windows). + * - stop execution (exit). + * + * I strongly suggest that you write your own "openFileHelper" method tailored to your needs. If you still want to parse + * the file, you can use the APPROXIMATE_PARSING option as explained inside the note at the beginning of the "xmlParser.cpp" file. + * + * @param filename the path of the XML file to parse. + * @param tag the name of the first tag inside the XML file. If the tag parameter is omitted, this function returns a node that represents the head of the xml document including the declaration term (). + */ + + static XMLCSTR getError(XMLError error); ///< this gives you a user-friendly explanation of the parsing error + + /// Create an XML string starting from the current XMLNode. + XMLSTR createXMLString(int nFormat=1, int *pnSize=NULL) const; + /**< The returned string should be free'd using the "freeXMLString" function. + * + * If nFormat==0, no formatting is required otherwise this returns an user friendly XML string from a given element + * with appropriate white spaces and carriage returns. if pnSize is given it returns the size in character of the string. */ + + /// Save the content of an xmlNode inside a file + XMLError writeToFile(XMLCSTR filename, + const char *encoding=NULL, + char nFormat=1) const; + /**< If nFormat==0, no formatting is required otherwise this returns an user friendly XML string from a given element with appropriate white spaces and carriage returns. + * If the global parameter "characterEncoding==encoding_UTF8", then the "encoding" parameter is ignored and always set to "utf-8". + * If the global parameter "characterEncoding==encoding_ShiftJIS", then the "encoding" parameter is ignored and always set to "SHIFT-JIS". + * If "_XMLWIDECHAR=1", then the "encoding" parameter is ignored and always set to "utf-16". + * If no "encoding" parameter is given the "ISO-8859-1" encoding is used. */ + /** @} */ + + /** @defgroup navigate Navigate the XMLNode structure + * @ingroup XMLParserGeneral + * @{ */ + XMLCSTR getName() const; ///< name of the node + XMLCSTR getText(int i=0) const; ///< return ith text field + int nText() const; ///< nbr of text field + XMLNode getParentNode() const; ///< return the parent node + XMLNode getChildNode(int i=0) const; ///< return ith child node + XMLNode getChildNode(XMLCSTR name, int i) const; ///< return ith child node with specific name (return an empty node if failing). If i==-1, this returns the last XMLNode with the given name. + XMLNode getChildNode(XMLCSTR name, int *i=NULL) const; ///< return next child node with specific name (return an empty node if failing) + XMLNode getChildNodeWithAttribute(XMLCSTR tagName, + XMLCSTR attributeName, + XMLCSTR attributeValue=NULL, + int *i=NULL) const; ///< return child node with specific name/attribute (return an empty node if failing) + XMLNode getChildNodeByPath(XMLCSTR path, char createNodeIfMissing=0, XMLCHAR sep='/'); + ///< return the first child node with specific path + XMLNode getChildNodeByPathNonConst(XMLSTR path, char createNodeIfMissing=0, XMLCHAR sep='/'); + ///< return the first child node with specific path. + + int nChildNode(XMLCSTR name) const; ///< return the number of child node with specific name + int nChildNode() const; ///< nbr of child node + XMLAttribute getAttribute(int i=0) const; ///< return ith attribute + XMLCSTR getAttributeName(int i=0) const; ///< return ith attribute name + XMLCSTR getAttributeValue(int i=0) const; ///< return ith attribute value + char isAttributeSet(XMLCSTR name) const; ///< test if an attribute with a specific name is given + XMLCSTR getAttribute(XMLCSTR name, int i) const; ///< return ith attribute content with specific name (return a NULL if failing) + XMLCSTR getAttribute(XMLCSTR name, int *i=NULL) const; ///< return next attribute content with specific name (return a NULL if failing) + int nAttribute() const; ///< nbr of attribute + XMLClear getClear(int i=0) const; ///< return ith clear field (comments) + int nClear() const; ///< nbr of clear field + XMLNodeContents enumContents(XMLElementPosition i) const; ///< enumerate all the different contents (attribute,child,text, clear) of the current XMLNode. The order is reflecting the order of the original file/string. NOTE: 0 <= i < nElement(); + int nElement() const; ///< nbr of different contents for current node + char isEmpty() const; ///< is this node Empty? + char isDeclaration() const; ///< is this node a declaration + XMLNode deepCopy() const; ///< deep copy (duplicate/clone) a XMLNode + static XMLNode emptyNode(); ///< return XMLNode::emptyXMLNode; + /** @} */ + + ~XMLNode(); + XMLNode(const XMLNode &A); ///< to allow shallow/fast copy: + XMLNode& operator=( const XMLNode& A ); ///< to allow shallow/fast copy: + + XMLNode(): d(NULL){}; + static XMLNode emptyXMLNode; + static XMLClear emptyXMLClear; + static XMLAttribute emptyXMLAttribute; + + /** @defgroup xmlModify Create or Update the XMLNode structure + * @ingroup XMLParserGeneral + * The functions in this group allows you to create from scratch (or update) a XMLNode structure. Start by creating your top + * node with the "createXMLTopNode" function and then add new nodes with the "addChild" function. The parameter 'pos' gives + * the position where the childNode, the text or the XMLClearTag will be inserted. The default value (pos=-1) inserts at the + * end. The value (pos=0) insert at the beginning (Insertion at the beginning is slower than at the end).
+ * + * REMARK: 0 <= pos < nChild()+nText()+nClear()
+ */ + + /** @defgroup creation Creating from scratch a XMLNode structure + * @ingroup xmlModify + * @{ */ + static XMLNode createXMLTopNode(XMLCSTR lpszName, char isDeclaration=FALSE); ///< Create the top node of an XMLNode structure + XMLNode addChild(XMLCSTR lpszName, char isDeclaration=FALSE, XMLElementPosition pos=-1); ///< Add a new child node + XMLNode addChild(XMLNode nodeToAdd, XMLElementPosition pos=-1); ///< If the "nodeToAdd" has some parents, it will be detached from it's parents before being attached to the current XMLNode + XMLAttribute *addAttribute(XMLCSTR lpszName, XMLCSTR lpszValuev); ///< Add a new attribute + XMLCSTR addText(XMLCSTR lpszValue, XMLElementPosition pos=-1); ///< Add a new text content + XMLClear *addClear(XMLCSTR lpszValue, XMLCSTR lpszOpen=NULL, XMLCSTR lpszClose=NULL, XMLElementPosition pos=-1); + /**< Add a new clear tag + * @param lpszOpen default value "" + */ + /** @} */ + + /** @defgroup xmlUpdate Updating Nodes + * @ingroup xmlModify + * Some update functions: + * @{ + */ + XMLCSTR updateName(XMLCSTR lpszName); ///< change node's name + XMLAttribute *updateAttribute(XMLAttribute *newAttribute, XMLAttribute *oldAttribute); ///< if the attribute to update is missing, a new one will be added + XMLAttribute *updateAttribute(XMLCSTR lpszNewValue, XMLCSTR lpszNewName=NULL,int i=0); ///< if the attribute to update is missing, a new one will be added + XMLAttribute *updateAttribute(XMLCSTR lpszNewValue, XMLCSTR lpszNewName,XMLCSTR lpszOldName);///< set lpszNewName=NULL if you don't want to change the name of the attribute if the attribute to update is missing, a new one will be added + XMLCSTR updateText(XMLCSTR lpszNewValue, int i=0); ///< if the text to update is missing, a new one will be added + XMLCSTR updateText(XMLCSTR lpszNewValue, XMLCSTR lpszOldValue); ///< if the text to update is missing, a new one will be added + XMLClear *updateClear(XMLCSTR lpszNewContent, int i=0); ///< if the clearTag to update is missing, a new one will be added + XMLClear *updateClear(XMLClear *newP,XMLClear *oldP); ///< if the clearTag to update is missing, a new one will be added + XMLClear *updateClear(XMLCSTR lpszNewValue, XMLCSTR lpszOldValue); ///< if the clearTag to update is missing, a new one will be added + /** @} */ + + /** @defgroup xmlDelete Deleting Nodes or Attributes + * @ingroup xmlModify + * Some deletion functions: + * @{ + */ + /// The "deleteNodeContent" function forces the deletion of the content of this XMLNode and the subtree. + void deleteNodeContent(); + /**< \note The XMLNode instances that are referring to the part of the subtree that has been deleted CANNOT be used anymore!!. Unexpected results will occur if you continue using them. */ + void deleteAttribute(int i=0); ///< Delete the ith attribute of the current XMLNode + void deleteAttribute(XMLCSTR lpszName); ///< Delete the attribute with the given name (the "strcmp" function is used to find the right attribute) + void deleteAttribute(XMLAttribute *anAttribute); ///< Delete the attribute with the name "anAttribute->lpszName" (the "strcmp" function is used to find the right attribute) + void deleteText(int i=0); ///< Delete the Ith text content of the current XMLNode + void deleteText(XMLCSTR lpszValue); ///< Delete the text content "lpszValue" inside the current XMLNode (direct "pointer-to-pointer" comparison is used to find the right text) + void deleteClear(int i=0); ///< Delete the Ith clear tag inside the current XMLNode + void deleteClear(XMLCSTR lpszValue); ///< Delete the clear tag "lpszValue" inside the current XMLNode (direct "pointer-to-pointer" comparison is used to find the clear tag) + void deleteClear(XMLClear *p); ///< Delete the clear tag "p" inside the current XMLNode (direct "pointer-to-pointer" comparison on the lpszName of the clear tag is used to find the clear tag) + /** @} */ + + /** @defgroup xmlWOSD ???_WOSD functions. + * @ingroup xmlModify + * The strings given as parameters for the "add" and "update" methods that have a name with + * the postfix "_WOSD" (that means "WithOut String Duplication")(for example "addText_WOSD") + * will be free'd by the XMLNode class. For example, it means that this is incorrect: + * \code + * xNode.addText_WOSD("foo"); + * xNode.updateAttribute_WOSD("#newcolor" ,NULL,"color"); + * \endcode + * In opposition, this is correct: + * \code + * xNode.addText("foo"); + * xNode.addText_WOSD(stringDup("foo")); + * xNode.updateAttribute("#newcolor" ,NULL,"color"); + * xNode.updateAttribute_WOSD(stringDup("#newcolor"),NULL,"color"); + * \endcode + * Typically, you will never do: + * \code + * char *b=(char*)malloc(...); + * xNode.addText(b); + * free(b); + * \endcode + * ... but rather: + * \code + * char *b=(char*)malloc(...); + * xNode.addText_WOSD(b); + * \endcode + * ('free(b)' is performed by the XMLNode class) + * @{ */ + static XMLNode createXMLTopNode_WOSD(XMLSTR lpszName, char isDeclaration=FALSE); ///< Create the top node of an XMLNode structure + XMLNode addChild_WOSD(XMLSTR lpszName, char isDeclaration=FALSE, XMLElementPosition pos=-1); ///< Add a new child node + XMLAttribute *addAttribute_WOSD(XMLSTR lpszName, XMLSTR lpszValue); ///< Add a new attribute + XMLCSTR addText_WOSD(XMLSTR lpszValue, XMLElementPosition pos=-1); ///< Add a new text content + XMLClear *addClear_WOSD(XMLSTR lpszValue, XMLCSTR lpszOpen=NULL, XMLCSTR lpszClose=NULL, XMLElementPosition pos=-1); ///< Add a new clear Tag + + XMLCSTR updateName_WOSD(XMLSTR lpszName); ///< change node's name + XMLAttribute *updateAttribute_WOSD(XMLAttribute *newAttribute, XMLAttribute *oldAttribute); ///< if the attribute to update is missing, a new one will be added + XMLAttribute *updateAttribute_WOSD(XMLSTR lpszNewValue, XMLSTR lpszNewName=NULL,int i=0); ///< if the attribute to update is missing, a new one will be added + XMLAttribute *updateAttribute_WOSD(XMLSTR lpszNewValue, XMLSTR lpszNewName,XMLCSTR lpszOldName); ///< set lpszNewName=NULL if you don't want to change the name of the attribute if the attribute to update is missing, a new one will be added + XMLCSTR updateText_WOSD(XMLSTR lpszNewValue, int i=0); ///< if the text to update is missing, a new one will be added + XMLCSTR updateText_WOSD(XMLSTR lpszNewValue, XMLCSTR lpszOldValue); ///< if the text to update is missing, a new one will be added + XMLClear *updateClear_WOSD(XMLSTR lpszNewContent, int i=0); ///< if the clearTag to update is missing, a new one will be added + XMLClear *updateClear_WOSD(XMLClear *newP,XMLClear *oldP); ///< if the clearTag to update is missing, a new one will be added + XMLClear *updateClear_WOSD(XMLSTR lpszNewValue, XMLCSTR lpszOldValue); ///< if the clearTag to update is missing, a new one will be added + /** @} */ + + /** @defgroup xmlPosition Position helper functions (use in conjunction with the update&add functions + * @ingroup xmlModify + * These are some useful functions when you want to insert a childNode, a text or a XMLClearTag in the + * middle (at a specified position) of a XMLNode tree already constructed. The value returned by these + * methods is to be used as last parameter (parameter 'pos') of addChild, addText or addClear. + * @{ */ + XMLElementPosition positionOfText(int i=0) const; + XMLElementPosition positionOfText(XMLCSTR lpszValue) const; + XMLElementPosition positionOfClear(int i=0) const; + XMLElementPosition positionOfClear(XMLCSTR lpszValue) const; + XMLElementPosition positionOfClear(XMLClear *a) const; + XMLElementPosition positionOfChildNode(int i=0) const; + XMLElementPosition positionOfChildNode(XMLNode x) const; + XMLElementPosition positionOfChildNode(XMLCSTR name, int i=0) const; ///< return the position of the ith childNode with the specified name if (name==NULL) return the position of the ith childNode + /** @} */ + + /// Enumeration for XML character encoding. + typedef enum XMLCharEncoding + { + char_encoding_error=0, + char_encoding_UTF8=1, + char_encoding_legacy=2, + char_encoding_ShiftJIS=3, + char_encoding_GB2312=4, + char_encoding_Big5=5, + char_encoding_GBK=6 // this is actually the same as Big5 + } XMLCharEncoding; + + /** \addtogroup conversions + * @{ */ + + /// Sets the global options for the conversions + static char setGlobalOptions(XMLCharEncoding characterEncoding=XMLNode::char_encoding_UTF8, char guessWideCharChars=1, + char dropWhiteSpace=1, char removeCommentsInMiddleOfText=1); + /**< The "setGlobalOptions" function allows you to change four global parameters that affect string & file + * parsing. First of all, you most-probably will never have to change these 3 global parameters. + * + * @param guessWideCharChars If "guessWideCharChars"=1 and if this library is compiled in WideChar mode, then the + * XMLNode::parseFile and XMLNode::openFileHelper functions will test if the file contains ASCII + * characters. If this is the case, then the file will be loaded and converted in memory to + * WideChar before being parsed. If 0, no conversion will be performed. + * + * @param guessWideCharChars If "guessWideCharChars"=1 and if this library is compiled in ASCII/UTF8/char* mode, then the + * XMLNode::parseFile and XMLNode::openFileHelper functions will test if the file contains WideChar + * characters. If this is the case, then the file will be loaded and converted in memory to + * ASCII/UTF8/char* before being parsed. If 0, no conversion will be performed. + * + * @param characterEncoding This parameter is only meaningful when compiling in char* mode (multibyte character mode). + * In wchar_t* (wide char mode), this parameter is ignored. This parameter should be one of the + * three currently recognized encodings: XMLNode::encoding_UTF8, XMLNode::encoding_ascii, + * XMLNode::encoding_ShiftJIS. + * + * @param dropWhiteSpace In most situations, text fields containing only white spaces (and carriage returns) + * are useless. Even more, these "empty" text fields are annoying because they increase the + * complexity of the user's code for parsing. So, 99% of the time, it's better to drop + * the "empty" text fields. However The XML specification indicates that no white spaces + * should be lost when parsing the file. So to be perfectly XML-compliant, you should set + * dropWhiteSpace=0. A note of caution: if you set "dropWhiteSpace=0", the parser will be + * slower and your code will be more complex. + * + * @param removeCommentsInMiddleOfText To explain this parameter, let's consider this code: + * \code + * XMLNode x=XMLNode::parseString("foobarchu","a"); + * \endcode + * If removeCommentsInMiddleOfText=0, then we will have: + * \code + * x.getText(0) -> "foo" + * x.getText(1) -> "bar" + * x.getText(2) -> "chu" + * x.getClear(0) --> "" + * x.getClear(1) --> "" + * \endcode + * If removeCommentsInMiddleOfText=1, then we will have: + * \code + * x.getText(0) -> "foobar" + * x.getText(1) -> "chu" + * x.getClear(0) --> "" + * \endcode + * + * \return "0" when there are no errors. If you try to set an unrecognized encoding then the return value will be "1" to signal an error. + * + * \note Sometime, it's useful to set "guessWideCharChars=0" to disable any conversion + * because the test to detect the file-type (ASCII/UTF8/char* or WideChar) may fail (rarely). */ + + /// Guess the character encoding of the string (ascii, utf8 or shift-JIS) + static XMLCharEncoding guessCharEncoding(void *buffer, int bufLen, char useXMLEncodingAttribute=1); + /**< The "guessCharEncoding" function try to guess the character encoding. You most-probably will never + * have to use this function. It then returns the appropriate value of the global parameter + * "characterEncoding" described in the XMLNode::setGlobalOptions. The guess is based on the content of a buffer of length + * "bufLen" bytes that contains the first bytes (minimum 25 bytes; 200 bytes is a good value) of the + * file to be parsed. The XMLNode::openFileHelper function is using this function to automatically compute + * the value of the "characterEncoding" global parameter. There are several heuristics used to do the + * guess. One of the heuristic is based on the "encoding" attribute. The original XML specifications + * forbids to use this attribute to do the guess but you can still use it if you set + * "useXMLEncodingAttribute" to 1 (this is the default behavior and the behavior of most parsers). + * If an inconsistency in the encoding is detected, then the return value is "0". */ + /** @} */ + + private: + // these are functions and structures used internally by the XMLNode class (don't bother about them): + + typedef struct XMLNodeDataTag // to allow shallow copy and "intelligent/smart" pointers (automatic delete): + { + XMLCSTR lpszName; // Element name (=NULL if root) + int nChild, // Number of child nodes + nText, // Number of text fields + nClear, // Number of Clear fields (comments) + nAttribute; // Number of attributes + char isDeclaration; // Whether node is an XML declaration - '' + struct XMLNodeDataTag *pParent; // Pointer to parent element (=NULL if root) + XMLNode *pChild; // Array of child nodes + XMLCSTR *pText; // Array of text fields + XMLClear *pClear; // Array of clear fields + XMLAttribute *pAttribute; // Array of attributes + int *pOrder; // order of the child_nodes,text_fields,clear_fields + int ref_count; // for garbage collection (smart pointers) + } XMLNodeData; + XMLNodeData *d; + + char parseClearTag(void *px, void *pa); + char maybeAddTxT(void *pa, XMLCSTR tokenPStr); + int ParseXMLElement(void *pXML); + void *addToOrder(int memInc, int *_pos, int nc, void *p, int size, XMLElementType xtype); + int indexText(XMLCSTR lpszValue) const; + int indexClear(XMLCSTR lpszValue) const; + XMLNode addChild_priv(int,XMLSTR,char,int); + XMLAttribute *addAttribute_priv(int,XMLSTR,XMLSTR); + XMLCSTR addText_priv(int,XMLSTR,int); + XMLClear *addClear_priv(int,XMLSTR,XMLCSTR,XMLCSTR,int); + void emptyTheNode(char force); + static inline XMLElementPosition findPosition(XMLNodeData *d, int index, XMLElementType xtype); + static int CreateXMLStringR(XMLNodeData *pEntry, XMLSTR lpszMarker, int nFormat); + static int removeOrderElement(XMLNodeData *d, XMLElementType t, int index); + static void exactMemory(XMLNodeData *d); + static int detachFromParent(XMLNodeData *d); +} XMLNode; + +/// This structure is given by the function XMLNode::enumContents. +typedef struct XMLNodeContents +{ + /// This dictates what's the content of the XMLNodeContent + enum XMLElementType etype; + /**< should be an union to access the appropriate data. Compiler does not allow union of object with constructor... too bad. */ + XMLNode child; + XMLAttribute attrib; + XMLCSTR text; + XMLClear clear; + +} XMLNodeContents; + +/** @defgroup StringAlloc String Allocation/Free functions + * @ingroup xmlModify + * @{ */ +/// Duplicate (copy in a new allocated buffer) the source string. +XMLDLLENTRY XMLSTR stringDup(XMLCSTR source, int cbData=-1); +/**< This is + * a very handy function when used with all the "XMLNode::*_WOSD" functions (\link xmlWOSD \endlink). + * @param cbData If !=0 then cbData is the number of chars to duplicate. New strings allocated with + * this function should be free'd using the "freeXMLString" function. */ + +/// to free the string allocated inside the "stringDup" function or the "createXMLString" function. +XMLDLLENTRY void freeXMLString(XMLSTR t); // {free(t);} +/** @} */ + +/** @defgroup atoX ato? like functions + * @ingroup XMLParserGeneral + * The "xmlto?" functions are equivalents to the atoi, atol, atof functions. + * The only difference is: If the variable "xmlString" is NULL, than the return value + * is "defautValue". These 6 functions are only here as "convenience" functions for the + * user (they are not used inside the XMLparser). If you don't need them, you can + * delete them without any trouble. + * + * @{ */ +XMLDLLENTRY char xmltob(XMLCSTR xmlString,char defautValue=0); +XMLDLLENTRY int xmltoi(XMLCSTR xmlString,int defautValue=0); +XMLDLLENTRY long xmltol(XMLCSTR xmlString,long defautValue=0); +XMLDLLENTRY double xmltof(XMLCSTR xmlString,double defautValue=.0); +XMLDLLENTRY XMLCSTR xmltoa(XMLCSTR xmlString,XMLCSTR defautValue=_CXML("")); +XMLDLLENTRY XMLCHAR xmltoc(XMLCSTR xmlString,XMLCHAR defautValue=_CXML('\0')); +/** @} */ + +/** @defgroup ToXMLStringTool Helper class to create XML files using "printf", "fprintf", "cout",... functions. + * @ingroup XMLParserGeneral + * @{ */ +/// Helper class to create XML files using "printf", "fprintf", "cout",... functions. +/** The ToXMLStringTool class helps you creating XML files using "printf", "fprintf", "cout",... functions. + * The "ToXMLStringTool" class is processing strings so that all the characters + * &,",',<,> are replaced by their XML equivalent: + * \verbatim &, ", ', <, > \endverbatim + * Using the "ToXMLStringTool class" and the "fprintf function" is THE most efficient + * way to produce VERY large XML documents VERY fast. + * \note If you are creating from scratch an XML file using the provided XMLNode class + * you must not use the "ToXMLStringTool" class (because the "XMLNode" class does the + * processing job for you during rendering).*/ +typedef struct XMLDLLENTRY ToXMLStringTool +{ +public: + ToXMLStringTool(): buf(NULL),buflen(0){} + ~ToXMLStringTool(); + void freeBuffer();///