read CDATA as an ordinary text

src/html/htmlparser.cpp

@@ -73,6 +73,7 @@ void HTMLParser::Item::Clear()
 	name.clear();
 	type            = none;
 	is_commentary   = false;
+	is_cdata		= false;
 	porphans        = nullptr;
 	new_line_before = false;
 	new_line        = false;
@@ -646,7 +647,7 @@ bool HTMLParser::IsValidCharForName(int c)
 	if( (c>='a' && c<='z') ||
 		(c>='A' && c<='Z') ||
 		(c>='0' && c<='9') ||
-		c=='-' || c=='!' || c==':' || c=='-' || c=='_') // : is for a namespace character, - is for a commentary
+		c=='-' || c=='!' || c==':' || c=='-' || c=='_' || c=='[') // : is for a namespace character, - is for a commentary, [ is for CDATA
 		return true;
 
 return false;
@@ -696,6 +697,13 @@ size_t i;
 				read_char();
 				break;
 			}
+
+			if( LastItem().type == Item::special && name == L"![CDATA[" )
+			{
+				LastItem().is_cdata = true;
+				read_char();
+				break;
+			}
 		}
 
 		read_char();
@@ -1002,10 +1010,49 @@ return CheckOrphan(str, end, LastItem().porphans->tab);
 }
 
 
-void HTMLParser::PutNormalNonWhite(std::wstring & str, bool allow_put_new_line, bool allow_put_space)
+bool HTMLParser::PutNormalNonWhite(std::wstring & str, bool allow_put_new_line, bool allow_put_space, bool is_cdata)
 {
-	while( lastc != -1 && lastc != 10 && !IsWhite(lastc) && !(!char_was_escaped && IsOpeningTagMark(lastc)) )
+	bool was_closing_tag = false;
+
+	while( lastc != -1 && lastc != 10 && !IsWhite(lastc) )
 	{
+		if( is_cdata )
+		{
+			if( lastc == ']' )
+			{
+				read_char();
+
+				if( lastc == ']' )
+				{
+					read_char();
+
+					if( IsClosingTagMark(lastc) )
+					{
+						read_char();
+						was_closing_tag = true;
+						break;
+					}
+					else
+					{
+						str += ']';
+						str += ']';
+					}
+				}
+				else
+				{
+					str += ']';
+				}
+			}
+		}
+		else
+		{
+			if( !char_was_escaped && IsOpeningTagMark(lastc) )
+			{
+				was_closing_tag = true;
+				break;
+			}
+		}
+
 		str += lastc;
 		read_char();
 	}
@@ -1028,6 +1075,8 @@ void HTMLParser::PutNormalNonWhite(std::wstring & str, bool allow_put_new_line,
 		AnalyzeEntitiesAndPut(str.c_str(), str.c_str() + str.size(), nullptr);
 	else
 		Put(str);
+
+	return was_closing_tag;
 }
 
 
@@ -1290,6 +1339,7 @@ void HTMLParser::ReadTextUntilClosingCommentary()
 }
 
 
+
 bool HTMLParser::IsClosingTagForLastItem(bool put_closing_tag_as_well)
 {
 	tmp_text.clear();
@@ -1361,7 +1411,7 @@ void HTMLParser::ReadTextUntilClosingTag(bool put_closing_tag_as_well)
 
 
 // reading text between html tags
-void HTMLParser::ReadText()
+void HTMLParser::ReadText(bool is_cdata)
 {
 	bool was_white_char = false;
 	bool was_new_line = false;
@@ -1391,10 +1441,12 @@ void HTMLParser::ReadText()
 		text_space_wstr = &wstr_space.value.value_wstring;
 	}
 
-	while( lastc != -1 && !(!char_was_escaped && IsOpeningTagMark(lastc)) )
+	bool was_closing_tag = false;
+
+	while( lastc != -1 && !was_closing_tag )
 	{
 		tmp_text.clear();
-		PutNormalNonWhite(tmp_text, allow_put_new_line, allow_put_space);
+		was_closing_tag = PutNormalNonWhite(tmp_text, allow_put_new_line, allow_put_space, is_cdata);
 
 		if( !tmp_text.empty() )
 		{
@@ -1627,6 +1679,11 @@ void HTMLParser::ReadItemSpecial()
 			Put(LastItem().name);
 		}
 		else
+		if( LastItem().is_cdata )
+		{
+			// do nothing
+		}
+		else
 		{
 			tmp_text.clear();
 			SkipWhiteLines();
@@ -2158,6 +2215,8 @@ void HTMLParser::ReadLoop()
 {
 	while( status == ok && ReadItem() )
 	{
+		bool was_cdata = false;
+
 		if( LastItem().type == Item::opening )
 		{
 			if( parsing_html )
@@ -2174,6 +2233,9 @@ void HTMLParser::ReadLoop()
 			if( LastItem().is_commentary )
 				ReadTextUntilClosingCommentary();
 
+			if( LastItem().is_cdata )
+				was_cdata = true;
+
 			PopStack();
 		}
 		else
@@ -2199,7 +2261,7 @@ void HTMLParser::ReadLoop()
 
 		if( status == ok )
 		{
-			ReadText();
+			ReadText(was_cdata);
 		}
 
 		is_first_item = false;
@@ -2331,7 +2393,7 @@ void HTMLParser::Read()
 
 	// it can be some text or white lines before the first html tag (we print it if using filtering)
 	// but they are not added to the Space tree
-	ReadText();
+	ReadText(false);
 
 	// reading the whole html source
 	ReadLoop();

src/html/htmlparser.h

@@ -132,6 +132,8 @@ public:
 
 		bool is_commentary;
 
+		bool is_cdata;
+
 		bool new_line_before;
 
 		// is there a new line after this tag
@@ -392,7 +394,7 @@ protected:
 	void CheckStackPrintRest();
 	void AddForgottenTags();
 	void CheckClosingTags();
-	void ReadText();
+	void ReadText(bool is_cdata);
 	bool PrintRest();
 	bool PrintOpeningItem();
 	void ReadItemName(std::wstring & name, bool clear_name = true);
@@ -415,7 +417,7 @@ protected:
 
 	void CheckChar(wchar_t c);
 
-	void PutNormalNonWhite(std::wstring & str, bool allow_put_new_line, bool allow_put_space);
+	bool PutNormalNonWhite(std::wstring & str, bool allow_put_new_line, bool allow_put_space, bool is_cdata);
 	void PutNormalWhite(bool & was_white_char, bool & was_new_line, std::wstring * result_text = nullptr);
 
 	void PutTabs(size_t len);