renamed: config option 'html_filter_break_lines' to 'html_filter_break_word'

added: config option 'html_filter_wrap_line' this wraps the whole line (line calculated with html tags as well) changed: orphans (for html filter) are read from locale files now ('language_orphans' value ) git-svn-id: svn://ttmath.org/publicrep/winix/trunk@728 e52654a7-88a9-db11-a3e9-0013d4bc506e
2011-04-16 08:42:22 +00:00
parent aadf12c7b3
commit 426beae796
19 changed files with 1079 additions and 672 deletions
--- a/core/Makefile.dep
+++ b/core/Makefile.dep
@@ -114,12 +114,12 @@ misc.o: ../templates/patterncacher.h ../core/item.h
 misc.o: ../templates/ckeditorgetparser.h ../core/httpsimpleparser.h
 misc.o: ../core/log.h ../templates/indexpatterns.h
 misc.o: ../templates/localefilter.h ../templates/locale.h ../core/config.h
-misc.o: ../db/db.h ../db/dbbase.h ../db/dbconn.h ../db/dbtextstream.h
+misc.o: ../templates/htmltextstream.h ../core/cur.h ../core/system.h
+misc.o: ../core/sessionmanager.h ../core/htmlfilter.h ../db/db.h
+misc.o: ../db/dbbase.h ../db/dbconn.h ../db/dbtextstream.h
 misc.o: ../core/textstream.h ../core/error.h ../db/dbitemquery.h
 misc.o: ../db/dbitemcolumns.h ../core/user.h ../core/group.h
 misc.o: ../core/dircontainer.h ../core/ugcontainer.h
-misc.o: ../templates/htmltextstream.h ../core/cur.h ../core/system.h
-misc.o: ../core/sessionmanager.h
 mount.o: mount.h misc.h item.h
 mountparser.o: mountparser.h mount.h item.h error.h dirs.h dircontainer.h
 mountparser.o: ../db/db.h ../db/dbbase.h ../db/dbconn.h ../db/dbtextstream.h
@@ -372,8 +372,8 @@ system.o: thumb.h basethread.h ../templates/templates.h
 system.o: ../templates/patterncacher.h ../templates/ckeditorgetparser.h
 system.o: ../core/httpsimpleparser.h ../core/log.h
 system.o: ../templates/indexpatterns.h ../core/cur.h ../core/system.h
-system.o: ../core/sessionmanager.h ../functions/functionbase.h
-system.o: ../core/request.h ../core/synchro.h
+system.o: ../core/sessionmanager.h ../core/htmlfilter.h
+system.o: ../functions/functionbase.h ../core/request.h ../core/synchro.h
 textstream.o: textstream.h misc.h item.h
 thumb.o: thumb.h basethread.h textstream.h ../db/db.h ../db/dbbase.h
 thumb.o: ../db/dbconn.h ../db/dbtextstream.h ../core/textstream.h misc.h
--- a/core/app.cpp
+++ b/core/app.cpp
@@ -604,16 +604,6 @@ void App::SendHeaders(bool compressing, int compress_encoding, Header header)



-void App::SetHtmlFilterConf()
-{
-	html_filter.TrimWhite(config.html_filter_trim_white);
-	html_filter.BreakLines(config.html_filter_break_lines);
-	html_filter.InsertTabs(config.html_filter_tabs);
-
-	if( config.html_filter_orphans )
-		html_filter.CheckOrphans(config.html_filter_orphans_lang, config.html_filter_orphans_mode);
-}
-

 void App::FilterCompressSend(bool compressing, int compress_encoding, const std::wstring & source_ref)
 {
@@ -624,8 +614,7 @@ void App::FilterCompressSend(bool compressing, int compress_encoding, const std:

 	if( config.html_filter && !raw )
 	{
-		SetHtmlFilterConf();
-		html_filter.Filter(*source, clean_html);
+		TemplatesFunctions::html_filter.Filter(*source, clean_html);
 		AddDebugInfo(clean_html);
 		source = &clean_html;
 	}
--- a/core/app.h
+++ b/core/app.h
@@ -30,7 +30,6 @@
 #include "functions/functions.h"
 #include "templates/templates.h"
 #include "compress.h"
-#include "htmlfilter.h"
 #include "getparser.h"
 #include "postparser.h"
 #include "cookieparser.h"
@@ -124,7 +123,6 @@ private:
 	CookieParser cookie_parser;
 	AcceptEncodingParser accept_encoding_parser;	
 	Compress compress;
-	HTMLFilter html_filter;
 	std::wstring clean_html, html_with_debug;
 	FCGX_Request fcgi_request;
 	int fcgi_socket;
@@ -153,7 +151,6 @@ private:
 	void CheckRequestMethod();
 	void CheckFCGIRole();

-	void SetHtmlFilterConf();
 	void PrepareSessionCookie();
 	void AddDebugInfo(std::wstring & out);
 	void FilterCompressSend(bool compressing, int compress_encoding, const std::wstring & source_ref);
--- a/core/bbcodeparser.cpp
+++ b/core/bbcodeparser.cpp
@@ -2,7 +2,7 @@
 * This file is a part of Winix
 * and is not publicly distributed
 * 
- * Copyright (c) 2008-2010, Tomasz Sowa
+ * Copyright (c) 2008-2011, Tomasz Sowa
 * All rights reserved.
 * 
 */
@@ -201,6 +201,12 @@ const BBCODEParser::Tags * BBCODEParser::FindTag(const wchar_t * tag)
 return 0;
 }

+const BBCODEParser::Tags * BBCODEParser::FindTag(const std::wstring & tag)
+{
+	return FindTag(tag.c_str());
+}
+
+

 void BBCODEParser::PrintArgumentCheckQuotes(const wchar_t * & start, const wchar_t * & end)
 {
--- a/core/bbcodeparser.h
+++ b/core/bbcodeparser.h
@@ -2,7 +2,7 @@
 * This file is a part of Winix
 * and is not publicly distributed
 * 
- * Copyright (c) 2008-2010, Tomasz Sowa
+ * Copyright (c) 2008-2011, Tomasz Sowa
 * All rights reserved.
 * 
 */
@@ -58,6 +58,7 @@ class BBCODEParser : public HTMLFilter
 	virtual void PutClosingTag(const wchar_t * tag);

 	const Tags * FindTag(const wchar_t * tag);
+	const Tags * FindTag(const std::wstring & tag);
 	void PrintArgumentCheckQuotes(const wchar_t * & start, const wchar_t * & end);


--- a/core/config.cpp
+++ b/core/config.cpp
@@ -159,10 +159,10 @@ void Config::AssignValues(bool stdout_is_closed)

 	html_filter             = Bool(L"html_filter", true);
 	html_filter_trim_white  = Bool(L"html_filter_trim_white", true);
-	html_filter_break_lines = Int(L"html_filter_break_lines", 60);
+	html_filter_break_word  = Int(L"html_filter_break_word", 60);
+	html_filter_wrap_line   = Int(L"html_filter_wrap_line", 110);
 	html_filter_tabs        = Size(L"html_filter_tabs", 2);
-	html_filter_orphans		= Bool(L"html_filter_orphans", false);
-	html_filter_orphans_lang_str = AText(L"html_filter_orphans_lang", L"pl");
+	html_filter_orphans		= Bool(L"html_filter_orphans", true);
 	html_filter_orphans_mode_str = AText(L"html_filter_orphans_mode_str", L"nbsp");

 	locale_dir         = Text(L"locale_dir");
@@ -200,17 +200,6 @@ void Config::SetAdditionalVariables()
 {
 	SetHttpHost(base_url, base_url_http_host);

-	if( html_filter_orphans_lang_str == "pl" )
-		html_filter_orphans_lang = HTMLFilter::lang_pl;
-	else
-	if( html_filter_orphans_lang_str == "cz" )
-		html_filter_orphans_lang = HTMLFilter::lang_cz;
-	else
-	if( html_filter_orphans_lang_str == "sk" )
-		html_filter_orphans_lang = HTMLFilter::lang_sk;
-	else
-		html_filter_orphans_lang = HTMLFilter::lang_none;
-
 	if( html_filter_orphans_mode_str == "160" )
 		html_filter_orphans_mode = HTMLFilter::orphan_160space;
 	else
--- a/core/config.h
+++ b/core/config.h
@@ -167,25 +167,24 @@ public:
 	// should white characters be trimmed
 	bool html_filter_trim_white;

-	// when long lines (lines without a white character) should be break (inserted a space)
+	// when long words should be broken (a space will be inserted)
 	// default: after 60 non white characters will be put a space
 	// set zero to turn off
-	int html_filter_break_lines;
+	size_t html_filter_break_word;
+
+	// when long lines should be broken (a new line character will be inserted)
+	// default: 80
+	// set zero to turn off
+	size_t html_filter_wrap_line;

 	// how many spaces will be put at one tree level
 	// default: 2
 	size_t html_filter_tabs;

 	// use checking for 'orphans' for a specicic language
-	// default: false
+	// default: true
 	bool html_filter_orphans;

-	// language for html orphans
-	// default: pl
-	// can be either: "pl" or "cz" or "sk"
-	std::string html_filter_orphans_lang_str;
-	HTMLFilter::Lang html_filter_orphans_lang;
-
 	// orphans mode
 	// either: "nbsp" or "160"
 	// default: "nbsp"
--- a/core/htmlfilter.cpp
+++ b/core/htmlfilter.cpp
--- a/core/htmlfilter.h
+++ b/core/htmlfilter.h
@@ -2,7 +2,7 @@
 * This file is a part of Winix
 * and is not publicly distributed
 * 
- * Copyright (c) 2008-2010, Tomasz Sowa
+ * Copyright (c) 2008-2011, Tomasz Sowa
 * All rights reserved.
 * 
 */
@@ -11,18 +11,31 @@
 #define headerfile_winix_core_htmlfilter

 #include <string>
+#include <map>
+#include <vector>
+#include <algorithm>




 // max length of a name of a html tag (with terminating null)
-#define WINIX_HTMLFILTER_ITEM_MAXLEN	30
+#define WINIX_HTMLFILTER_ITEM_NAME_MAXLEN	30
+
+// max length of a html lang attribute (e.g. "en", "pl")
+#define WINIX_HTMLFILTER_ITEM_LANG_MAXLEN	10
+
+
+#define WINIX_HTMLFILTER_ATTR_NAME_MAXLEN	40
+
+
+#define WINIX_HTMLFILTER_ATTR_VALUE_MAXLEN	500
+

 // depth of the html tree
-#define WINIX_HTMLFILTER_STACK_MAXLEN	100
+#define WINIX_HTMLFILTER_STACK_MAXLEN		100

 // length of a buffer used for printing
-// it should be at least: WINIX_HTMLFILTER_ITEM_MAXLEN+3
+// it should be at least: WINIX_HTMLFILTER_ITEM_NAME_MAXLEN+3
 #define WINIX_HTMLFILTER_BUFFER_MAXLEN	2048


@@ -48,23 +61,12 @@ class HTMLFilter
 {
 public:

-
-	// for checking orphans
-	enum Lang
-	{
-		lang_pl,
-		lang_cz,
-		lang_sk,
-		lang_none
-	};
-
 	enum OrphanMode
 	{
 		orphan_nbsp,		// putting "&nbsp;" string
 		orphan_160space		// putting 160 ascii code
 	};

-
 	HTMLFilter();
 	HTMLFilter(const HTMLFilter & f);
 	HTMLFilter & operator=(const HTMLFilter & f);
@@ -76,12 +78,18 @@ public:
 	void Filter(const std::wstring & in, std::wstring & out);


-	// insert a white space into long lines
-	// only between html tags
+	// insert a white space into long words
+	// (only between html tags)
 	// skipped in such tags: script, pre, textarea
 	// break_after - after how many characters insert a space (0 - off)
-	void BreakLines(size_t break_after_);
+	void BreakWord(size_t break_after_);

+	// insert a new line character into long lines
+	// (only between html tags)
+	// skipped in such tags: script, pre, textarea
+	// wrap_line - after how many characters wrap a line (0 - off)
+	// lines are wrapped only in 'body' tag (useful for text in 'title' tag which is in 'head' section)
+	void WrapLine(size_t wrap_line_);

 	// trimming white characters (with new lines)
 	// at the beginning, at the end and in the middle of a string
@@ -98,10 +106,16 @@ public:
 	void InsertTabs(size_t tabsize);


+	// orphans are checked only in 'body' tag
+	void AssignOrphans(const wchar_t * lang_code,      const std::vector<std::wstring> & otab);
+	void AssignOrphans(const std::wstring & lang_code, const std::vector<std::wstring> & otab);
+	void ClearOrphans();
+
+
 	// check 'orphans' for the specicic language
 	// if an orphan is detected then the non-break space ("&nbsp;" or ascii 160 code) will be put
 	// default disable (lang_none)
-	void CheckOrphans(Lang lang_, OrphanMode mode = orphan_nbsp);
+	void OrphansMode(OrphanMode mode = orphan_nbsp);


 	// skipping some unsafe tags
@@ -112,52 +126,80 @@ public:

 protected:

+	// orphans for one language
+	struct Orphans
+	{
+		std::vector<std::wstring> tab;
+		size_t max_len;
+	};
+
+
+	// orphans for all languages
+	// map<language_code, Orphans>
+	typedef std::map<std::wstring, Orphans> OrphansTab;
+	OrphansTab orphans_tab;
+
+
 	struct Item
 	{
-		wchar_t name[WINIX_HTMLFILTER_ITEM_MAXLEN];
-		size_t name_len;
+		std::wstring name; // max size: WINIX_HTMLFILTER_ITEM_NAME_MAXLEN

 		enum Type
 		{
-			opening,
-			closing,
-			simple,
-			special,
+			opening,		/* sample:  <h1>		*/
+			closing,		/* sample:  </h1>		*/
+			simple,			/* sample:  <br/>		*/
+			special,		/* sample:  <!doctype>	*/
 			none
 		} type;

 		// is there a new line after this tag
 		bool new_line;

+		// current orphans table
+		// (will be propagated)
+		Orphans * porphans;
+
+		// this item or one from its parents is a 'body' html tag
+		// (will be propagated)
+		bool has_body_tag;
+
 		void Clear();
 		Item();
 	};


-	// only this method have direct access to the output string
-	// you can easily change the output from a std::wstring to something else
-	virtual void Put(const wchar_t * str, const wchar_t * end);


 	Item & GetItem(size_t i);
 	Item & LastItem();

 	wchar_t ToLower(wchar_t c);
-	bool IsNameEqual(const wchar_t * name1, const wchar_t * name2);
-	bool IsNameEqual(const wchar_t * name1, const wchar_t * name2, size_t len);
+	void    ToLower(std::wstring & str);
+
+	bool IsNameEqual(const wchar_t * name1,      const wchar_t * name2);
+	bool IsNameEqual(const wchar_t * name1,      const std::wstring & name2);
+	bool IsNameEqual(const std::wstring & name1, const wchar_t * name2);
+	bool IsNameEqual(const std::wstring & name1, const std::wstring & name2);
+
+	bool IsNameEqual(const wchar_t * name1,      const wchar_t * name2,      size_t len);
+	bool IsNameEqual(const wchar_t * name1,      const std::wstring & name2, size_t len);
+	bool IsNameEqual(const std::wstring & name1, const wchar_t * name2,      size_t len);
+	bool IsNameEqual(const std::wstring & name1, const std::wstring & name2, size_t len);
+
 	bool IsLastTag(const wchar_t * name);
 	bool IsTagSafe(const wchar_t * tag);
+	bool IsTagSafe(const std::wstring & tag);

-	int CheckOrphan(const wchar_t * str, const wchar_t * end, const wchar_t * orphan);
-	bool CheckOrphanTable(const wchar_t * str, const wchar_t * end, const wchar_t ** table, size_t o1, size_t o2);
-	bool CheckOrphanLangPl(const wchar_t * str, const wchar_t * end);
-	bool CheckOrphanLangCz(const wchar_t * str, const wchar_t * end);
+	int  CheckOrphan(const wchar_t * str, const wchar_t * end, const std::wstring & orphan_str);
+	bool CheckOrphan(const wchar_t * str, const wchar_t * end, const std::vector<std::wstring> & orphans);
 	bool CheckOrphan(const wchar_t * str, const wchar_t * end);

 	bool IsWhite(int c);
 	void SkipWhite();
 	void SkipWhiteLines();
 	void SkipWhiteWithFirstNewLine();
+	void SkipWhiteLines(const wchar_t * & str, const wchar_t * end);
 	bool IsClosingTagForLastItem();
 	virtual bool IsOpeningTagMark();
 	virtual bool IsOpeningCommentaryTagMark();
@@ -165,11 +207,12 @@ protected:
 	virtual bool IsClosingTagMark();
 	virtual bool IsClosingXmlSimpleTagMark();
 	bool SkipCommentaryTagIfExists();
-	const wchar_t * SkipItemCheckXmlSimple();
+	void SkipAndCheckClosingTag();

 	void PopStack();
 	bool PushStack();
 	virtual bool IsValidCharForName(int c);
+	virtual bool IsValidCharForAttrName(int c);
 	void CheckNewLine();
 	virtual void CheckExceptions();
 	void CheckStackPrintRest();
@@ -178,41 +221,68 @@ protected:
 	virtual void ReadNormalTextSkipWhite(const wchar_t * & start, const wchar_t * & last_non_white);
 	void ReadNormalText();
 	bool PrintRest();
-	void PrintItem(const wchar_t * start, const wchar_t * end);
+	void PrintOpeningItem();
 	void ReadItemName();
+	void ReadItemAttrName();
+	void ReadItemAttrValue(bool has_quote);
+
+	bool ReadItemAttr();
+	bool CheckItemAttr();
+	void PrinItemAttr();
+
+	void ReadItemClosing();
+	void ReadItemSpecial();
+	void ReadItemOpening();
 	bool ReadItem();
 	virtual void Init();
 	virtual void Deinit();
+	void ReadLoop();
 	void Read();

-	size_t PutNormalTextTrimFillBuffer(const wchar_t * & str, const wchar_t * & end);
-	size_t PutNormalTextFillBuffer(const wchar_t * & str, const wchar_t * & end);
+	void CheckChar(wchar_t c);
+
+	virtual void Put(wchar_t c);
+	virtual void Put(const wchar_t * str);
+	virtual void Put(const wchar_t * str, const wchar_t * end);
+	virtual void Put(const std::wstring & str);
+
+	void CheckLineWrap();
+	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);
 	virtual void PutNormalText(const wchar_t * str, const wchar_t * end);
-	virtual void PutNormalTextTrim(const wchar_t * str, const wchar_t * end);
 	void PutLastTagWithClosingTag();
 	virtual void PutOpeningTagMark();
 	virtual void PutClosingTagMark();
 	virtual void PutTagName(const wchar_t * name);
-	virtual void PutOpeningTag(const wchar_t * start, const wchar_t * end);
+	void PutTagName(const std::wstring & name);
+	virtual void PutOpeningTag(bool close_tag = true);
 	virtual void PutClosingTag(const wchar_t * tag);
-	size_t PutTabsToBuffer(size_t index, size_t len);
-	size_t PutNonBreakSpaceToBuffer(size_t index);
 	void PutTabs(size_t len);
+	void PutNonBreakingSpace();
 	void PutNewLine();
+	void CalcOrphansMaxLen(Orphans & orphans);

 	const wchar_t * pchar;
 	Item empty;
-	Item * pstack;		// stack pointer
-	size_t stack_len;	// length of the stack
+	Item * pstack;			// stack pointer
+	size_t stack_len;		// length of the stack
 	wchar_t * buffer;		// buffer used when printing
 	std::wstring * out_string;
 	bool last_new_line;
-	size_t break_after;	// insert a space into long lines after break_after characters
-	bool trim_white;	// trimming white characters
+	size_t break_after;		// insert a space into long words after 'break_after' characters
+	size_t wrap_line;		// insert a new line character into long lines
+	bool trim_white;		// trimming white characters
 	size_t tab_size;
-	Lang lang;          // current language for checking orphans
 	OrphanMode orphan_mode;
-	bool safe_mode;     // skipping some unsafe tags
+	std::wstring attr_name;
+	std::wstring attr_value;
+	std::wstring attr_value_lower;
+	bool attr_has_value;
+	std::wstring lang_code_lower;
+	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;
 };