added to HTMLFilter:

- now we can parse " and ' in html attributes
- we can skip html tags and commentaries, added method:
  void SkipAllTags(bool skip_all_tags, bool skip_commentaries);
- there is virtual method: virtual void ItemFound();
  which is called when a html tag is parsed




git-svn-id: svn://ttmath.org/publicrep/winix/trunk@1129 e52654a7-88a9-db11-a3e9-0013d4bc506e

winixd/core/htmlfilter.cpp

@@ -5,7 +5,7 @@
  */
 
 /* 
- * Copyright (c) 2008-2014, Tomasz Sowa
+ * Copyright (c) 2008-2018, Tomasz Sowa
  * All rights reserved.
  * 
  * Redistribution and use in source and binary forms, with or without
@@ -97,17 +97,25 @@ size_t out_projected_len = in.size() * 2 + 1;
 }
 
 
-HTMLFilter::HTMLFilter()
+void HTMLFilter::SetSomeDefaults()
 {
-	pstack = new Item[WINIX_HTMLFILTER_STACK_MAXLEN];
-	buffer = new wchar_t[WINIX_HTMLFILTER_BUFFER_MAXLEN];
-
 	tab_size    = 2;
 	trim_white  = false;
 	break_after = 0;
 	wrap_line   = 0;
 	orphan_mode = orphan_nbsp;
 	safe_mode   = false;
+	skip_all_tags = false;
+	skip_commentaries = false;
+}
+
+
+HTMLFilter::HTMLFilter()
+{
+	pstack = new Item[WINIX_HTMLFILTER_STACK_MAXLEN];
+	buffer = new wchar_t[WINIX_HTMLFILTER_BUFFER_MAXLEN];
+
+	SetSomeDefaults();
 }
 
 
@@ -116,6 +124,8 @@ HTMLFilter::HTMLFilter(const HTMLFilter & f)
 	// don't need to copy the stack
 	pstack = new Item[WINIX_HTMLFILTER_STACK_MAXLEN];
 	buffer = new wchar_t[WINIX_HTMLFILTER_BUFFER_MAXLEN];
+
+	SetSomeDefaults();
 }
 
 
@@ -125,6 +135,8 @@ HTMLFilter & HTMLFilter::operator=(const HTMLFilter & f)
 	pstack = new Item[WINIX_HTMLFILTER_STACK_MAXLEN];
 	buffer = new wchar_t[WINIX_HTMLFILTER_BUFFER_MAXLEN];
 
+	// we can copy some fields from f
+
 return *this;
 }
 
@@ -136,6 +148,8 @@ HTMLFilter::~HTMLFilter()
 }
 
 
+
+
 void HTMLFilter::BreakWord(size_t break_after_)
 {
 	break_after = break_after_;
@@ -224,6 +238,13 @@ void HTMLFilter::SafeMode(bool safe_mode_)
 }
 
 
+void HTMLFilter::SkipAllTags(bool skip_all_tags, bool skip_commentaries)
+{
+	this->skip_all_tags = skip_all_tags;
+	this->skip_commentaries = skip_commentaries;
+}
+
+
 void HTMLFilter::SetNoFilterTag(const std::wstring & tag_name)
 {
 	no_filter_tag = tag_name;
@@ -374,9 +395,10 @@ return false;
 
 
 // used for such tags as: script, pre, textarea
-void HTMLFilter::PutLastTagWithClosingTag()
+void HTMLFilter::PutEverythingUntilClosingTag(bool put_closing_tag_as_well)
 {
 const wchar_t * start = pchar;
+const wchar_t * end = pchar;
 
 	while( *pchar != 0 )
 	{
@@ -384,32 +406,9 @@ const wchar_t * start = pchar;
 		{
 			if( IsClosingTagForLastItem() )
 			{
-				PopStack();
+				if( put_closing_tag_as_well )
-				CheckNewLine();
+					end = pchar;
-				break;
-			}
-		}
-		else
-		{
-			pchar += 1;
-		}
-	}
 
-	Put(start, pchar);
-}
-
-
-// used with <nofilter> </nofilter> tags
-void HTMLFilter::PutTextBetweenLastTagWithClosingTag()
-{
-const wchar_t * start = pchar, * end = pchar;
-
-	while( *pchar != 0 )
-	{
-		if( IsOpeningTagMark() )
-		{
-			if( IsClosingTagForLastItem() )
-			{
 				PopStack();
 				CheckNewLine();
 				break;
@@ -427,16 +426,36 @@ const wchar_t * start = pchar, * end = pchar;
 
 
 
+
 void HTMLFilter::SkipAndCheckClosingTag()
 {
+	bool is_quoted = false;
+	wchar_t quote_char = 0;
+
 	for( ; *pchar ; ++pchar )
 	{
-		if( LastItem().type == Item::opening && IsClosingXmlSimpleTagMark() ) // closing xml tag: default '/'
+		if( *pchar == '"' || *pchar == '\'' )
+		{
+			if( is_quoted )
+			{
+				if( *pchar == quote_char )
+				{
+					is_quoted = false;
+				}
+			}
+			else
+			{
+				is_quoted = true;
+				quote_char = *pchar;
+			}
+		}
+		else
+		if( !is_quoted && LastItem().type == Item::opening && IsClosingXmlSimpleTagMark() ) // closing xml tag: default '/'
 		{
 			LastItem().type = Item::simple;
 		}
-
+		else
-		if( IsClosingTagMark() )
+		if( !is_quoted && IsClosingTagMark() )
 		{
 			++pchar;
 			break;
@@ -502,18 +521,26 @@ size_t i;
 
 
 
-void HTMLFilter::ReadItemAttrValue(bool has_quote)
+void HTMLFilter::ReadItemAttrValue(bool has_quote, wchar_t quote_char)
 {
 size_t i;
 
-	// sprawdzic to wszedzie bo teraz jest tablica
 	attr_value.clear();
 	attr_value_temp.clear();
 
-	// !! dodac obsluge pojedynczego cudzyslowu
+	for(i=0 ; *pchar ; ++i, ++pchar )
-
-	for(i=0 ; *pchar && *pchar != '\"' && !IsClosingTagMark() && (has_quote || (*pchar!=10 && !IsWhite(*pchar)) ); ++i )
 	{
+		if( has_quote )
+		{
+			if( *pchar == quote_char )
+				break;
+		}
+		else
+		{
+			if( IsClosingTagMark() || *pchar == 10 || IsWhite(*pchar) )
+				break;
+		}
+
 		if( *pchar==10 || IsWhite(*pchar) )
 		{
 			if( !attr_value_temp.empty() )
@@ -524,9 +551,9 @@ size_t i;
 		}
 		else
 		if( i < WINIX_HTMLFILTER_ATTR_VALUE_MAXLEN )
+		{
 			attr_value_temp += *pchar;
-
+		}
-		++pchar;
 	}
 
 	if( !attr_value_temp.empty() )
@@ -795,8 +822,8 @@ void HTMLFilter::PutClosingTagMark()
 
 
 
-// !! zmienic na lepsza nazwe
+// !! IMPROVE ME change to a better name
-// bo to nie zwraca true jesli tag jest safe
+// this functions does not return true when the tag is safe
 bool HTMLFilter::IsTagSafe(const wchar_t * tag)
 {
 	if( !safe_mode )
@@ -838,9 +865,10 @@ bool HTMLFilter::IsTagSafe(const std::wstring & tag)
 bool HTMLFilter::PutOpeningTag()
 {
 	if( !IsTagSafe(LastItem().name) )
-		// !! IMPROVE ME
+	{
-		// !! dodac tutaj skipniecie calego tagu
+		SkipAndCheckClosingTag();
 		return false;
+	}
 
 	PutOpeningTagMark();
 	Put(LastItem().name);
@@ -852,7 +880,7 @@ return true;
 
 void HTMLFilter::PutClosingTag(const wchar_t * tag)
 {
-	if( !IsTagSafe(tag) )
+	if( skip_all_tags || !IsTagSafe(tag) )
 		return;
 
 	PutOpeningTagMark();
@@ -1005,9 +1033,19 @@ const wchar_t * last_non_white = pchar;
 
 	while( *pchar != 0 )
 	{
+		const wchar_t * commentary_start = pchar;
+
 		if( SkipCommentaryTagIfExists() )
 		{
 			last_non_white = pchar - 1; // pointing at the last '>' from a commentary
+			PutNormalText(start, commentary_start);
+
+			if( !skip_commentaries )
+			{
+				PutNormalText(commentary_start, pchar);
+			}
+
+			start = pchar;
 		}
 		else
 		{
@@ -1029,7 +1067,7 @@ const wchar_t * last_non_white = pchar;
 
 bool HTMLFilter::PrintOpeningItem()
 {
-	if( IsNameEqual(no_filter_tag, LastItem().name) )
+	if( skip_all_tags || IsNameEqual(no_filter_tag, LastItem().name) )
 		return true;
 
 	if( last_new_line )
@@ -1068,16 +1106,15 @@ bool HTMLFilter::ReadItemAttr()
 	pchar += 1;				// skipping '='
 	SkipWhiteLines();
 
-
+	bool has_quote = (*pchar == '\"' || *pchar == '\'');
-	// !! dodac obsluge pojedynczego cudzyslowu
+	wchar_t quote_char = *pchar;
-	bool has_quote = (*pchar == '\"');
 
 	if( has_quote )
 		pchar += 1;			// skipping the first quote mark
 
-	ReadItemAttrValue(has_quote);
+	ReadItemAttrValue(has_quote, quote_char);
 
-	if( *pchar == '\"' )
+	if( has_quote && *pchar == quote_char )
 		pchar += 1;			// skipping the last quote mark
 
 return true;
@@ -1112,7 +1149,7 @@ void HTMLFilter::PrintItemAttr()
 {
 size_t i;
 
-	if( IsNameEqual(no_filter_tag, LastItem().name) )
+	if( skip_all_tags || IsNameEqual(no_filter_tag, LastItem().name) )
 		return;
 
 	Put(' ');
@@ -1150,11 +1187,17 @@ void HTMLFilter::ReadItemClosing()
 void HTMLFilter::ReadItemSpecial()
 {
 	LastItem().type = Item::special;
-	PutOpeningTagMark();
+
+	if( !skip_all_tags )
+		PutOpeningTagMark();
+
 	const wchar_t * start = pchar;
+	pchar += 1; // skipping '!'
+
+	ReadItemName();
 	SkipAndCheckClosingTag();
 
-	if( pchar > start )
+	if( !skip_all_tags && pchar > start )
 		Put(start, pchar);
 
 	// closing tag mark is printed directly from the source
@@ -1176,7 +1219,7 @@ void HTMLFilter::ReadItemOpening()
 
 		SkipAndCheckClosingTag(); // here LastItem().type can be changed to 'simple'
 
-		if( !IsNameEqual(no_filter_tag, LastItem().name) )
+		if( !skip_all_tags && !IsNameEqual(no_filter_tag, LastItem().name) )
 		{
 			if( LastItem().type == Item::simple )
 				Put(L" /");
@@ -1187,6 +1230,11 @@ void HTMLFilter::ReadItemOpening()
 }
 
 
+void HTMLFilter::ItemFound()
+{
+}
+
+
 bool HTMLFilter::ReadItem()
 {
 	if( *pchar == 0 )
@@ -1209,6 +1257,8 @@ bool HTMLFilter::ReadItem()
 	CheckNewLine();
 	LastItem().new_line = last_new_line;
 
+	ItemFound();
+
 return true;
 }	
 
@@ -1332,13 +1382,13 @@ void HTMLFilter::CheckExceptions()
 
 	// in safe_mode the script tag is ignored
 	if( !safe_mode && IsLastTag(L"script") )
-		PutLastTagWithClosingTag();
+		PutEverythingUntilClosingTag(!skip_all_tags);
 
 	if( IsLastTag(L"pre") || IsLastTag(L"textarea") )
-		PutLastTagWithClosingTag();
+		PutEverythingUntilClosingTag(!skip_all_tags);
 
 	if( IsLastTag(no_filter_tag) )
-		PutTextBetweenLastTagWithClosingTag();
+		PutEverythingUntilClosingTag(false);
 
 	if( IsLastTag(L"body") )
 		LastItem().has_body_tag = true;
@@ -1371,7 +1421,7 @@ int i;
 
 	for(int z=(int)stack_len-2 ; z>=i ; --z)
 	{
-		if( pstack[z].new_line )
+		if( !skip_all_tags && pstack[z].new_line )
 		{
 			PutNewLine();
 			PutTabs(z);
@@ -1421,7 +1471,7 @@ void HTMLFilter::CheckClosingTags()
 	if( IsNameEqual(pstack[stack_len-1].name, pstack[stack_len-2].name) )
 	{
 		// last closing tag is from the previous one
-		if( pstack[stack_len-2].new_line )
+		if( !skip_all_tags && pstack[stack_len-2].new_line )
 		{
 			PutNewLine();
 			PutTabs(stack_len-2);
@@ -1444,7 +1494,7 @@ bool HTMLFilter::PrintRest()
 const wchar_t * start = pchar;
 
 	// in safe mode we do not print the rest html code
-	if( safe_mode )
+	if( safe_mode || skip_all_tags )
 		return false;
 
 	while( *pchar )
@@ -1474,7 +1524,7 @@ void HTMLFilter::ReadLoop()
 		{
 			if( stack_len > 1 )
 			{
-				pstack[stack_len-2].new_line = LastItem().new_line;
+				//pstack[stack_len-2].new_line = LastItem().new_line;
 			}
 			else
 			if( trim_white )
@@ -1492,6 +1542,10 @@ void HTMLFilter::ReadLoop()
 		{
 			CheckClosingTags();
 		}
+		else
+		{
+			PopStack();
+		}
 
 		ReadNormalText();
 	}

winixd/core/htmlfilter.h

@@ -5,7 +5,7 @@
  */
 
 /* 
- * Copyright (c) 2008-2014, Tomasz Sowa
+ * Copyright (c) 2008-2018, Tomasz Sowa
  * All rights reserved.
  * 
  * Redistribution and use in source and binary forms, with or without
@@ -99,7 +99,7 @@ public:
 	HTMLFilter();
 	HTMLFilter(const HTMLFilter & f);
 	HTMLFilter & operator=(const HTMLFilter & f);
-	~HTMLFilter();
+	virtual ~HTMLFilter();
 
 
 	// main methods used for filtering
@@ -156,6 +156,9 @@ public:
 	// (script, iframe, frame, frameset, applet, head, meta, html, link, body, ...)
 	void SafeMode(bool safe_mode_);
 
+	// skip all html tags
+	// gives only text without markup
+	void SkipAllTags(bool skip_all_tags, bool skip_commentaries);
 
 
 protected:
@@ -239,10 +242,13 @@ protected:
 	virtual void PutNormalText(const wchar_t * str, const wchar_t * end);
 	virtual void ReadNormalTextSkipWhite(const wchar_t * & start, const wchar_t * & last_non_white);
 
+	virtual void ItemFound();
 
 	/*
 		others
 	*/
+	void SetSomeDefaults();
+
 	Item & GetItem(size_t i);
 	Item & LastItem();
 
@@ -288,7 +294,7 @@ protected:
 	bool PrintOpeningItem();
 	void ReadItemName();
 	void ReadItemAttrName();
-	void ReadItemAttrValue(bool has_quote);
+	void ReadItemAttrValue(bool has_quote, wchar_t quote_char);
 
 	bool ReadItemAttr();
 	bool CheckItemAttr();
@@ -307,8 +313,7 @@ protected:
 	bool HasSemiloconAround(const wchar_t * str, const wchar_t * end);
 	void PutNormalNonWhite(const wchar_t * & str, const wchar_t * end);
 	void PutNormalWhite(const wchar_t * & str, const wchar_t * end);
-	void PutLastTagWithClosingTag();
+	void PutEverythingUntilClosingTag(bool put_closing_tag_as_well);
-	void PutTextBetweenLastTagWithClosingTag();
 	void PutTabs(size_t len);
 	void PutNonBreakingSpace();
 	void PutNewLine();
@@ -335,6 +340,8 @@ protected:
 	size_t line_len;		//length of the current line (without first spaces which create the html tree)
 	bool safe_mode;			// skipping some unsafe tags
 	Orphans orphans_temp;
+	bool skip_all_tags;
+	bool skip_commentaries;
 };