правка багов по результатам тестирования

DocFormatReader - картинки маркированных списков

ASCOfficeDocFile/Common/FormatUtils.h

@@ -77,22 +77,73 @@ namespace ASCDocFormatUtils
 	class FormatUtils
 	{
 	public:
-		static inline std::wstring XmlEncode(std::wstring data)
+		static inline bool IsUnicodeSymbol( wchar_t symbol )
+		{
+			bool result = false;
+
+			if ( ( 0x0009 == symbol ) || ( 0x000A == symbol ) || ( 0x000D == symbol ) ||
+				( ( 0x0020 <= symbol ) && ( 0xD7FF >= symbol ) ) || ( ( 0xE000 <= symbol ) && ( symbol <= 0xFFFD ) ) ||
+				( ( 0x10000 <= symbol ) && symbol ) )
+			{
+				result = true;
+			}
+
+			return result;		  
+		}
+		static inline std::wstring XmlEncode(std::wstring data, bool bDeleteNoUnicode = false)
 		{
 			std::wstring buffer;
 			buffer.reserve(data.size());
-			for(size_t pos = 0; pos != data.size(); ++pos)
+			
+			if(bDeleteNoUnicode)
 			{
-				switch(data[pos])
-				{
-				case '&':  buffer.append(_T("&amp;"));      break;
-				case '\"': buffer.append(_T("&quot;"));     break;
-				case '\'': buffer.append(_T("&apos;"));     break;
-				case '<':  buffer.append(_T("&lt;"));       break;
-				case '>':  buffer.append(_T("&gt;"));       break;
-				default:   buffer.append(&data[pos], 1);	break;
+				for(size_t pos = 0; pos != data.size(); ++pos)
+				{
+					switch(data[pos])
+					{
+						case '&':  buffer.append(_T("&amp;"));      break;
+						case '\"': buffer.append(_T("&quot;"));     break;
+						case '\'': buffer.append(_T("&apos;"));     break;
+						case '<':  buffer.append(_T("&lt;"));       break;
+						case '>':  buffer.append(_T("&gt;"));       break;
+						default:   
+						{
+							if ( false == IsUnicodeSymbol( data[pos] ) )
+							{
+								wchar_t symbol1 = data[pos];
+								if(0xD800 <= symbol1 && symbol1 <= 0xDFFF && pos + 1 < data.size())
+								{
+									pos++;
+									wchar_t symbol2 = data[pos];
+									if (symbol1 < 0xDC00 && symbol2 >= 0xDC00 && symbol2 <= 0xDFFF)
+									{
+										buffer.append(&data[pos-1], 2);
+									}
+								}
+							}
+							else
+								buffer.append(&data[pos], 1);	
+						}break;
+					}
+				}
+			}
+			else
+			{
+				for(size_t pos = 0; pos != data.size(); ++pos)
+				{
+					switch(data[pos])
+					{
+						case '&':  buffer.append(_T("&amp;"));      break;
+						case '\"': buffer.append(_T("&quot;"));     break;
+						case '\'': buffer.append(_T("&apos;"));     break;
+						case '<':  buffer.append(_T("&lt;"));       break;
+						case '>':  buffer.append(_T("&gt;"));       break;
+						default:   buffer.append(&data[pos], 1);	break;
+					}
 				}
 			}
+		
+
 			return buffer;
 		}
 

ASCOfficeDocFile/DocDocxConverter/FontTableMapping.cpp

@@ -35,14 +35,14 @@ namespace DocFileFormat
 			FontFamilyName* font = dynamic_cast<FontFamilyName*>( *iter );
 
 			m_pXmlWriter->WriteNodeBegin( _T( "w:font" ), TRUE );
-			m_pXmlWriter->WriteAttribute( _T( "w:name" ), FormatUtils::XmlEncode(font->xszFtn).c_str() );
+			m_pXmlWriter->WriteAttribute( _T( "w:name" ), FormatUtils::XmlEncode(font->xszFtn, true).c_str());
 			m_pXmlWriter->WriteNodeEnd( _T( "" ), TRUE, FALSE );
 
 			//alternative name
 			if ( ( font->xszAlt != wstring( _T( "" ) ) ) && ( font->xszAlt.length() > 0 ) )
 			{
 				m_pXmlWriter->WriteNodeBegin( _T( "w:altName" ), TRUE );
-				m_pXmlWriter->WriteAttribute( _T( "w:val" ), FormatUtils::XmlEncode(font->xszAlt).c_str() );
+				m_pXmlWriter->WriteAttribute( _T( "w:val" ), FormatUtils::XmlEncode(font->xszAlt, true).c_str() );
 				m_pXmlWriter->WriteNodeEnd( _T( "" ), TRUE, FALSE );
 				m_pXmlWriter->WriteNodeEnd( _T( "w:altName" ) );
 			}

ASCOfficeDocFile/DocDocxConverter/OleObject.h

@@ -187,9 +187,16 @@ namespace DocFileFormat
 
 			if (sz_obj > 4)
 			{
-				UserType		= reader.ReadLengthPrefixedAnsiString();
-				ClipboardFormat	= reader.ReadLengthPrefixedAnsiString();
-				Program			= reader.ReadLengthPrefixedAnsiString();
+				//todooo сделать по нормальному CompObjHeader - psc3a.doc
+				//UserType		= reader.ReadLengthPrefixedAnsiString();
+
+				//sz_obj = reader.GetSize() - reader.GetPosition();
+				//if (sz_obj > 4)
+				//	ClipboardFormat	= reader.ReadLengthPrefixedAnsiString();
+
+				//sz_obj = reader.GetSize() - reader.GetPosition();
+				//if (sz_obj > 4)
+				//	Program			= reader.ReadLengthPrefixedAnsiString();
 			}
 			delete pCompStream;
 		  }

ASCOfficeDocFile/DocDocxConverter/PictureDescriptor.cpp

@@ -46,7 +46,7 @@ namespace DocFileFormat
 		if (lcb > 10000000) 
 			return;
 
-		if (lcb > sz) 
+		if (lcb > sz && sz != 2) //bullet picture   
 		{
 			unsigned char* bytes = reader.ReadBytes(sz - fc - 4, false);
 			if ( bytes )
@@ -56,7 +56,7 @@ namespace DocFileFormat
 			return;
 		}
 
-		if (lcb > 0)
+		if (lcb >= 10)
 		{
 			unsigned short cbHeader	=	reader.ReadUInt16();
 

ASCOfficeDocFile/DocDocxConverter/StyleSheetMapping.cpp

@@ -69,7 +69,7 @@ namespace DocFileFormat
 
 				// <w:name val="" />
 				m_pXmlWriter->WriteNodeBegin( _T( "w:name" ), TRUE );  
-				m_pXmlWriter->WriteAttribute( _T( "w:val" ), FormatUtils::XmlEncode(getStyleName( *iter )).c_str() );
+				m_pXmlWriter->WriteAttribute( _T( "w:val" ), FormatUtils::XmlEncode(getStyleName( *iter ), true ).c_str());
 				m_pXmlWriter->WriteNodeEnd( _T( "" ), TRUE );
 
 				// <w:basedOn val="" />

ASCOfficeDocFile/DocDocxConverter/VMLPictureMapping.cpp

@@ -140,6 +140,10 @@ namespace DocFileFormat
 			{
 				m_pXmlWriter->WriteAttribute( _T( "o:ole" ), _T( "" ) );
 			}
+			else if (m_isBulletPicture)
+			{
+				m_pXmlWriter->WriteAttribute( _T( "o:bullet" ), true );
+			}
 			
 			std::list<OptionEntry>::iterator end = options.end();
 			for (std::list<OptionEntry>::iterator iter = options.begin(); iter != end; ++iter)

ASCOfficeDocFile/DocDocxConverter/VMLShapeTypeMapping.cpp

@@ -41,6 +41,8 @@ namespace DocFileFormat
 			// Path
 			if (!pShape->Path.empty())
 				m_pXmlWriter->WriteAttribute( _T("path"), pShape->Path.c_str() );
+			else if (_isBulletPicture)
+				m_pXmlWriter->WriteAttribute( _T("path"), _T("m@4@5l@4@11@9@11@9@5xe"));
 
 
 			//Default fill / stroke
@@ -87,6 +89,16 @@ namespace DocFileFormat
 
 				m_pXmlWriter->WriteNodeEnd( _T( "v:formulas" ) );
 			}
+			else if (_isBulletPicture)
+			{
+				m_pXmlWriter->WriteString(_T("<v:formulas><v:f eqn=\"if lineDrawn pixelLineWidth 0\"/>\
+											 <v:f eqn=\"sum @0 1 0\"/><v:f eqn=\"sum 0 0 @1\"/>\
+											 <v:f eqn=\"prod @2 1 2\"/><v:f eqn=\"prod @3 21600 pixelWidth\"/>\
+											 <v:f eqn=\"prod @3 21600 pixelHeight\"/><v:f eqn=\"sum @0 0 1\"/>\
+											 <v:f eqn=\"prod @6 1 2\"/><v:f eqn=\"prod @7 21600 pixelWidth\"/>\
+											 <v:f eqn=\"sum @8 21600 0\"/><v:f eqn=\"prod @7 21600 pixelHeight\"/>\
+											 <v:f eqn=\"sum @10 21600 0\"/></v:formulas>"));
+			}
 
 			// Path
 			m_pXmlWriter->WriteNodeBegin( _T( "v:path" ), true );