fixed wxHTML parsing to run in O(n) even in UTF8 build

git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/trunk@48390 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775

diff --git a/include/wx/html/htmlpars.h b/include/wx/html/htmlpars.h
index 66f2de8db413d2670b152ad64fbff9c235a341bf..6fe3d7f78022d20c48ba84ec65d9685495c8fa97 100644 (file)
--- a/include/wx/html/htmlpars.h
+++ b/include/wx/html/htmlpars.h
@@ -73,7 +73,8 @@ public:
 
     // Parses the m_Source from begin_pos to end_pos-1.
     // (in noparams version it parses whole m_Source)
 
     // Parses the m_Source from begin_pos to end_pos-1.
     // (in noparams version it parses whole m_Source)

-    void DoParsing(int begin_pos, int end_pos);
+    void DoParsing(const wxString::const_iterator& begin_pos,
+                   const wxString::const_iterator& end_pos);

     void DoParsing();
 
     // Returns pointer to the tag at parser's current position
     void DoParsing();
 
     // Returns pointer to the tag at parser's current position


@@ -104,7 +105,7 @@ public:
     // Restores state before last call to PushTagHandler
     void PopTagHandler();
 
     // Restores state before last call to PushTagHandler
     void PopTagHandler();
 

-    wxString* GetSource() {return &m_Source;}
+    const wxString* GetSource() {return m_Source;}

     void SetSource(const wxString& src);
 
     // Sets HTML source and remembers current parser's state so that it can
     void SetSource(const wxString& src);
 
     // Sets HTML source and remembers current parser's state so that it can


@@ -140,7 +141,8 @@ protected:
     void CreateDOMTree();
     void DestroyDOMTree();
     void CreateDOMSubTree(wxHtmlTag *cur,
     void CreateDOMTree();
     void DestroyDOMTree();
     void CreateDOMSubTree(wxHtmlTag *cur,

-                          int begin_pos, int end_pos,
+                          const wxString::const_iterator& begin_pos,
+                          const wxString::const_iterator& end_pos,

                           wxHtmlTagsCache *cache);
 
     // Adds text to the output.
                           wxHtmlTagsCache *cache);
 
     // Adds text to the output.


@@ -163,7 +165,7 @@ protected:
     wxHtmlTextPieces *m_TextPieces;
     size_t m_CurTextPiece;
 
     wxHtmlTextPieces *m_TextPieces;
     size_t m_CurTextPiece;
 

-    wxString m_Source;
+    const wxString *m_Source;

 
     wxHtmlParserState *m_SavedStates;
 
 
     wxHtmlParserState *m_SavedStates;
 


@@ -232,7 +234,7 @@ protected:
     // parses input between beginning and ending tag.
     // m_Parser must be set.
     void ParseInner(const wxHtmlTag& tag)
     // parses input between beginning and ending tag.
     // m_Parser must be set.
     void ParseInner(const wxHtmlTag& tag)

-        { m_Parser->DoParsing(tag.GetBeginPos(), tag.GetEndPos1()); }
+        { m_Parser->DoParsing(tag.GetBeginIter(), tag.GetEndIter1()); }

 
     // Parses given source as if it was tag's inner code (see
     // wxHtmlParser::GetInnerSource).  Unlike ParseInner(), this method lets
 
     // Parses given source as if it was tag's inner code (see
     // wxHtmlParser::GetInnerSource).  Unlike ParseInner(), this method lets

diff --git a/include/wx/html/htmltag.h b/include/wx/html/htmltag.h
index e92bb40eba41a3ece5fed7354dbe720d87ae43da..56f042096ac9cbafed2a8604ea741329dc97f774 100644 (file)
--- a/include/wx/html/htmltag.h
+++ b/include/wx/html/htmltag.h
@@ -31,7 +31,7 @@ class WXDLLIMPEXP_HTML wxHtmlTagsCache
 {
 private:
     wxHtmlTagsCacheData *m_Cache;
 {
 private:
     wxHtmlTagsCacheData *m_Cache;

-    size_t m_CachePos;
+    int m_CachePos;

 
     wxHtmlTagsCacheData& Cache() { return *m_Cache; }
 
 
     wxHtmlTagsCacheData& Cache() { return *m_Cache; }
 


@@ -41,7 +41,11 @@ public:
     virtual ~wxHtmlTagsCache();
 
     // Finds parameters for tag starting at at and fills the variables
     virtual ~wxHtmlTagsCache();
 
     // Finds parameters for tag starting at at and fills the variables

-    void QueryTag(int at, int* end1, int* end2);
+    void QueryTag(const wxString::const_iterator& at,
+                  const wxString::const_iterator& inputEnd,
+                  wxString::const_iterator *end1,
+                  wxString::const_iterator *end2,
+                  bool *hasEnding);

 
     DECLARE_NO_COPY_CLASS(wxHtmlTagsCache)
 };
 
     DECLARE_NO_COPY_CLASS(wxHtmlTagsCache)
 };


@@ -60,7 +64,9 @@ protected:
     // The tag begins (with '<' character) at position pos in source
     // end_pos is position where parsing ends (usually end of document)
     wxHtmlTag(wxHtmlTag *parent,
     // The tag begins (with '<' character) at position pos in source
     // end_pos is position where parsing ends (usually end of document)
     wxHtmlTag(wxHtmlTag *parent,

-              const wxString& source, int pos, int end_pos,
+              const wxString *source,
+              const wxString::const_iterator& pos,
+              const wxString::const_iterator& end_pos,

               wxHtmlTagsCache *cache,
               wxHtmlEntitiesParser *entParser);
     friend class wxHtmlParser;
               wxHtmlTagsCache *cache,
               wxHtmlEntitiesParser *entParser);
     friend class wxHtmlParser;


@@ -108,24 +114,37 @@ public:
     // Returns string containing all params.
     wxString GetAllParams() const;
 
     // Returns string containing all params.
     wxString GetAllParams() const;
 

-    // return true if this there is matching ending tag
-    inline bool HasEnding() const {return m_End1 >= 0;}
+    // return true if there is matching ending tag
+    inline bool HasEnding() const {return m_hasEnding;}

 
     // returns beginning position of _internal_ block of text
     // See explanation (returned value is marked with *):
     // bla bla bla <MYTAG>* bla bla intenal text</MYTAG> bla bla
 
     // returns beginning position of _internal_ block of text
     // See explanation (returned value is marked with *):
     // bla bla bla <MYTAG>* bla bla intenal text</MYTAG> bla bla

-    inline int GetBeginPos() const {return m_Begin;}
+    wxString::const_iterator GetBeginIter() const
+        { return m_Begin; }

     // returns ending position of _internal_ block of text.
     // bla bla bla <MYTAG> bla bla intenal text*</MYTAG> bla bla
     // returns ending position of _internal_ block of text.
     // bla bla bla <MYTAG> bla bla intenal text*</MYTAG> bla bla

-    inline int GetEndPos1() const {return m_End1;}
+    wxString::const_iterator GetEndIter1() const
+        { wxASSERT(m_hasEnding); return m_End1; }

     // returns end position 2 :
     // bla bla bla <MYTAG> bla bla internal text</MYTAG>* bla bla
     // returns end position 2 :
     // bla bla bla <MYTAG> bla bla internal text</MYTAG>* bla bla

-    inline int GetEndPos2() const {return m_End2;}
+    wxString::const_iterator GetEndIter2() const
+        { wxASSERT(m_hasEnding); return m_End2; }
+
+#if WXWIN_COMPATIBILITY_2_8
+    wxDEPRECATED( inline int GetBeginPos() const );
+    wxDEPRECATED( inline int GetEndPos1() const );
+    wxDEPRECATED( inline int GetEndPos2() const );
+#endif // WXWIN_COMPATIBILITY_2_8

 
 private:
     wxString m_Name;
 
 private:
     wxString m_Name;

-    int m_Begin, m_End1, m_End2;
+    bool m_hasEnding;
+    wxString::const_iterator m_Begin, m_End1, m_End2;

     wxArrayString m_ParamNames, m_ParamValues;
     wxArrayString m_ParamNames, m_ParamValues;

+#if WXWIN_COMPATIBILITY_2_8
+    wxString::const_iterator m_sourceStart;
+#endif

 
     // DOM tree relations:
     wxHtmlTag *m_Next;
 
     // DOM tree relations:
     wxHtmlTag *m_Next;


@@ -137,10 +156,16 @@ private:
 };
 
 
 };
 
 

+#if WXWIN_COMPATIBILITY_2_8
+inline int wxHtmlTag::GetBeginPos() const { return m_Begin - m_sourceStart; }
+inline int wxHtmlTag::GetEndPos1() const { return m_End1 - m_sourceStart; }
+inline int wxHtmlTag::GetEndPos2() const { return m_End2 - m_sourceStart; }
+#endif // WXWIN_COMPATIBILITY_2_8

 
 
 
 
 
 

-#endif
+
+#endif // wxUSE_HTML

 
 #endif // _WX_HTMLTAG_H_
 
 
 #endif // _WX_HTMLTAG_H_
 

diff --git a/src/html/htmlpars.cpp b/src/html/htmlpars.cpp
index c51181a3fa53682b30559f464159b0f2b3b46836..01e1b306719810dc819060174f26211d51dc7de1 100644 (file)
--- a/src/html/htmlpars.cpp
+++ b/src/html/htmlpars.cpp
@@ -48,8 +48,10 @@ class wxHtmlTextPiece
 {
 public:
     wxHtmlTextPiece() {}
 {
 public:
     wxHtmlTextPiece() {}

-    wxHtmlTextPiece(int pos, int lng) : m_pos(pos), m_lng(lng) {}
-    int m_pos, m_lng;
+    wxHtmlTextPiece(const wxString::const_iterator& start,
+                    const wxString::const_iterator& end)
+        : m_start(start), m_end(end) {}
+    wxString::const_iterator m_start, m_end;

 };
 
 // NB: this is an empty class and not typedef because of forward declaration
 };
 
 // NB: this is an empty class and not typedef because of forward declaration


@@ -64,7 +66,7 @@ public:
     wxHtmlTag         *m_tags;
     wxHtmlTextPieces  *m_textPieces;
     int                m_curTextPiece;
     wxHtmlTag         *m_tags;
     wxHtmlTextPieces  *m_textPieces;
     int                m_curTextPiece;

-    wxString           m_source;
+    const wxString    *m_source;

     wxHtmlParserState *m_nextState;
 };
 
     wxHtmlParserState *m_nextState;
 };
 


@@ -78,6 +80,7 @@ wxHtmlParser::wxHtmlParser()
     : wxObject(), m_HandlersHash(wxKEY_STRING),
       m_FS(NULL), m_HandlersStack(NULL)
 {
     : wxObject(), m_HandlersHash(wxKEY_STRING),
       m_FS(NULL), m_HandlersStack(NULL)
 {

+    m_Source = NULL;

     m_entitiesParser = new wxHtmlEntitiesParser;
     m_Tags = NULL;
     m_CurTag = NULL;
     m_entitiesParser = new wxHtmlEntitiesParser;
     m_Tags = NULL;
     m_CurTag = NULL;


@@ -103,6 +106,7 @@ wxHtmlParser::~wxHtmlParser()
     m_HandlersHash.Clear();
     WX_CLEAR_LIST(wxList, m_HandlersList);
     delete m_entitiesParser;
     m_HandlersHash.Clear();
     WX_CLEAR_LIST(wxList, m_HandlersList);
     delete m_entitiesParser;

+    delete m_Source;

 }
 
 wxObject* wxHtmlParser::Parse(const wxString& source)
 }
 
 wxObject* wxHtmlParser::Parse(const wxString& source)


@@ -128,7 +132,10 @@ void wxHtmlParser::DoneParser()
 void wxHtmlParser::SetSource(const wxString& src)
 {
     DestroyDOMTree();
 void wxHtmlParser::SetSource(const wxString& src)
 {
     DestroyDOMTree();

-    m_Source = src;
+    // NB: this is allocated on heap because wxHtmlTag keeps a pointer to
+    //     this string if WXWIN_COMPATIBILITY_2_8
+    delete m_Source;
+    m_Source = new wxString(src);

     CreateDOMTree();
     m_CurTag = NULL;
     m_CurTextPiece = 0;
     CreateDOMTree();
     m_CurTag = NULL;
     m_CurTextPiece = 0;


@@ -136,54 +143,53 @@ void wxHtmlParser::SetSource(const wxString& src)
 
 void wxHtmlParser::CreateDOMTree()
 {
 
 void wxHtmlParser::CreateDOMTree()
 {

-    wxHtmlTagsCache cache(m_Source);
+    wxHtmlTagsCache cache(*m_Source);

     m_TextPieces = new wxHtmlTextPieces;
     m_TextPieces = new wxHtmlTextPieces;

-    CreateDOMSubTree(NULL, 0, m_Source.length(), &cache);
+    CreateDOMSubTree(NULL, m_Source->begin(), m_Source->end(), &cache);

     m_CurTextPiece = 0;
 }
 
     m_CurTextPiece = 0;
 }
 

-extern bool wxIsCDATAElement(const wxChar *tag);
+extern bool wxIsCDATAElement(const wxString& tag);

 
 void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
 
 void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,

-                                    int begin_pos, int end_pos,
+                                    const wxString::const_iterator& begin_pos,
+                                    const wxString::const_iterator& end_pos,

                                     wxHtmlTagsCache *cache)
 {
                                     wxHtmlTagsCache *cache)
 {

-    if (end_pos <= begin_pos) return;
+    if (end_pos <= begin_pos)
+        return;

 
     wxChar c;
 
     wxChar c;

-    int i = begin_pos;
-    int textBeginning = begin_pos;
+    wxString::const_iterator i = begin_pos;
+    wxString::const_iterator textBeginning = begin_pos;

 
     // If the tag contains CDATA text, we include the text between beginning
     // and ending tag verbosely. Setting i=end_pos will skip to the very
     // end of this function where text piece is added, bypassing any child
     // tags parsing (CDATA element can't have child elements by definition):
 
     // If the tag contains CDATA text, we include the text between beginning
     // and ending tag verbosely. Setting i=end_pos will skip to the very
     // end of this function where text piece is added, bypassing any child
     // tags parsing (CDATA element can't have child elements by definition):

-    if (cur != NULL && wxIsCDATAElement(cur->GetName().c_str()))
+    if (cur != NULL && wxIsCDATAElement(cur->GetName()))

     {
         i = end_pos;
     }
 
     while (i < end_pos)
     {
     {
         i = end_pos;
     }
 
     while (i < end_pos)
     {

-        c = m_Source.GetChar(i);
+        c = *i;

 
         if (c == wxT('<'))
         {
             // add text to m_TextPieces:
 
         if (c == wxT('<'))
         {
             // add text to m_TextPieces:

-            if (i - textBeginning > 0)
-                m_TextPieces->push_back(
-                    wxHtmlTextPiece(textBeginning, i - textBeginning));
+            if (i > textBeginning)
+                m_TextPieces->push_back(wxHtmlTextPiece(textBeginning, i));

 
             // if it is a comment, skip it:
 
             // if it is a comment, skip it:

-            wxString::const_iterator iter = m_Source.begin() + i;
-            if ( SkipCommentTag(iter, m_Source.end()) )
+            if ( SkipCommentTag(i, m_Source->end()) )

             {
             {

-                textBeginning =
-                i = iter - m_Source.begin() + 1; // skip closing '>' too
+                textBeginning = i = i + 1; // skip closing '>' too

             }
 
             // add another tag to the tree:
             }
 
             // add another tag to the tree:

-            else if (i < end_pos-1 && m_Source.GetChar(i+1) != wxT('/'))
+            else if (i < end_pos-1 && *(i+1) != wxT('/'))

             {
                 wxHtmlTag *chd;
                 if (cur)
             {
                 wxHtmlTag *chd;
                 if (cur)


@@ -211,12 +217,12 @@ void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
                 if (chd->HasEnding())
                 {
                     CreateDOMSubTree(chd,
                 if (chd->HasEnding())
                 {
                     CreateDOMSubTree(chd,

-                                     chd->GetBeginPos(), chd->GetEndPos1(),
+                                     chd->GetBeginIter(), chd->GetEndIter1(),

                                      cache);
                                      cache);

-                    i = chd->GetEndPos2();
+                    i = chd->GetEndIter2();

                 }
                 else
                 }
                 else

-                    i = chd->GetBeginPos();
+                    i = chd->GetBeginIter();

 
                 textBeginning = i;
             }
 
                 textBeginning = i;
             }


@@ -224,17 +230,16 @@ void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
             // ... or skip ending tag:
             else
             {
             // ... or skip ending tag:
             else
             {

-                while (i < end_pos && m_Source.GetChar(i) != wxT('>')) i++;
+                while (i < end_pos && *i != wxT('>')) ++i;

                 textBeginning = i+1;
             }
         }
                 textBeginning = i+1;
             }
         }

-        else i++;
+        else ++i;

     }
 
     // add remaining text to m_TextPieces:
     }
 
     // add remaining text to m_TextPieces:

-    if (end_pos - textBeginning > 0)
-        m_TextPieces->push_back(
-            wxHtmlTextPiece(textBeginning, end_pos - textBeginning));
+    if (end_pos > textBeginning)
+        m_TextPieces->push_back(wxHtmlTextPiece(textBeginning, end_pos));

 }
 
 void wxHtmlParser::DestroyDOMTree()
 }
 
 void wxHtmlParser::DestroyDOMTree()


@@ -257,42 +262,45 @@ void wxHtmlParser::DoParsing()
 {
     m_CurTag = m_Tags;
     m_CurTextPiece = 0;
 {
     m_CurTag = m_Tags;
     m_CurTextPiece = 0;

-    DoParsing(0, m_Source.length());
+    DoParsing(m_Source->begin(), m_Source->end());

 }
 
 }
 

-void wxHtmlParser::DoParsing(int begin_pos, int end_pos)
+void wxHtmlParser::DoParsing(const wxString::const_iterator& begin_pos_,
+                             const wxString::const_iterator& end_pos)

 {
 {

-    if (end_pos <= begin_pos) return;
+    wxString::const_iterator begin_pos(begin_pos_);
+
+    if (end_pos <= begin_pos)
+        return;

 
     wxHtmlTextPieces& pieces = *m_TextPieces;
     size_t piecesCnt = pieces.size();
 
     while (begin_pos < end_pos)
     {
 
     wxHtmlTextPieces& pieces = *m_TextPieces;
     size_t piecesCnt = pieces.size();
 
     while (begin_pos < end_pos)
     {

-        while (m_CurTag && m_CurTag->GetBeginPos() < begin_pos)
+        while (m_CurTag && m_CurTag->GetBeginIter() < begin_pos)

             m_CurTag = m_CurTag->GetNextTag();
         while (m_CurTextPiece < piecesCnt &&
             m_CurTag = m_CurTag->GetNextTag();
         while (m_CurTextPiece < piecesCnt &&

-               pieces[m_CurTextPiece].m_pos < begin_pos)
+               pieces[m_CurTextPiece].m_start < begin_pos)

             m_CurTextPiece++;
 
         if (m_CurTextPiece < piecesCnt &&
             (!m_CurTag ||
             m_CurTextPiece++;
 
         if (m_CurTextPiece < piecesCnt &&
             (!m_CurTag ||

-             pieces[m_CurTextPiece].m_pos < m_CurTag->GetBeginPos()))
+             pieces[m_CurTextPiece].m_start < m_CurTag->GetBeginIter()))

         {
             // Add text:
             AddText(GetEntitiesParser()->Parse(
         {
             // Add text:
             AddText(GetEntitiesParser()->Parse(

-                       m_Source.Mid(pieces[m_CurTextPiece].m_pos,
-                                    pieces[m_CurTextPiece].m_lng)));
-            begin_pos = pieces[m_CurTextPiece].m_pos +
-                        pieces[m_CurTextPiece].m_lng;
+                       wxString(pieces[m_CurTextPiece].m_start,
+                                pieces[m_CurTextPiece].m_end)));
+            begin_pos = pieces[m_CurTextPiece].m_end;

             m_CurTextPiece++;
         }
         else if (m_CurTag)
         {
             if (m_CurTag->HasEnding())
             m_CurTextPiece++;
         }
         else if (m_CurTag)
         {
             if (m_CurTag->HasEnding())

-                begin_pos = m_CurTag->GetEndPos2();
+                begin_pos = m_CurTag->GetEndIter2();

             else
             else

-                begin_pos = m_CurTag->GetBeginPos();
+                begin_pos = m_CurTag->GetBeginIter();

             wxHtmlTag *t = m_CurTag;
             m_CurTag = m_CurTag->GetNextTag();
             AddTag(*t);
             wxHtmlTag *t = m_CurTag;
             m_CurTag = m_CurTag->GetNextTag();
             AddTag(*t);


@@ -318,7 +326,7 @@ void wxHtmlParser::AddTag(const wxHtmlTag& tag)
     if (!inner)
     {
         if (tag.HasEnding())
     if (!inner)
     {
         if (tag.HasEnding())

-            DoParsing(tag.GetBeginPos(), tag.GetEndPos1());
+            DoParsing(tag.GetBeginIter(), tag.GetEndIter1());

     }
 }
 
     }
 }
 


@@ -393,7 +401,7 @@ void wxHtmlParser::SetSourceAndSaveState(const wxString& src)
     m_Tags = NULL;
     m_TextPieces = NULL;
     m_CurTextPiece = 0;
     m_Tags = NULL;
     m_TextPieces = NULL;
     m_CurTextPiece = 0;

-    m_Source = wxEmptyString;
+    m_Source = NULL;

 
     SetSource(src);
 }
 
     SetSource(src);
 }


@@ -419,8 +427,7 @@ bool wxHtmlParser::RestoreState()
 
 wxString wxHtmlParser::GetInnerSource(const wxHtmlTag& tag)
 {
 
 wxString wxHtmlParser::GetInnerSource(const wxHtmlTag& tag)
 {

-    return GetSource()->Mid(tag.GetBeginPos(),
-                            tag.GetEndPos1() - tag.GetBeginPos());
+    return wxString(tag.GetBeginIter(), tag.GetEndIter1());

 }
 
 //-----------------------------------------------------------------------------
 }
 
 //-----------------------------------------------------------------------------


@@ -501,11 +508,16 @@ wxString wxHtmlEntitiesParser::Parse(const wxString& input) const
             const wxString::const_iterator ent_s = c;
             wxChar entity_char;
 
             const wxString::const_iterator ent_s = c;
             wxChar entity_char;
 

-            for (; c != end &&
-                   ((*c >= wxT('a') && *c <= wxT('z')) ||
-                    (*c >= wxT('A') && *c <= wxT('Z')) ||
-                    (*c >= wxT('0') && *c <= wxT('9')) ||
-                    *c == wxT('_') || *c == wxT('#')); ++c) {}
+            for ( ; c != end; ++c )
+            {
+                wxChar ch = *c;
+                if ( !((ch >= wxT('a') && ch <= wxT('z')) ||
+                       (ch >= wxT('A') && ch <= wxT('Z')) ||
+                       (ch >= wxT('0') && ch <= wxT('9')) ||
+                        ch == wxT('_') || ch == wxT('#')) )
+                    break;
+            }
+

             entity.append(ent_s, c);
             if (c == end || *c != wxT(';')) --c;
             last = c+1;
             entity.append(ent_s, c);
             if (c == end || *c != wxT(';')) --c;
             last = c+1;

diff --git a/src/html/htmltag.cpp b/src/html/htmltag.cpp
index a0f9b6af379ea5bd993cdb3ee02c385da76fc574..4365a26554316d4165255e2fb0322b4403650d9c 100644 (file)
--- a/src/html/htmltag.cpp
+++ b/src/html/htmltag.cpp
@@ -36,14 +36,21 @@ struct wxHtmlCacheItem
 {
     // this is "pos" value passed to wxHtmlTag's constructor.
     // it is position of '<' character of the tag
 {
     // this is "pos" value passed to wxHtmlTag's constructor.
     // it is position of '<' character of the tag

-    int Key;
+    wxString::const_iterator Key;
+
+    // Tag type
+    enum Type
+    {
+        Type_Normal, // normal tag with a matching ending tag
+        Type_NoMatchingEndingTag, // there's no ending tag for this tag
+        Type_EndingTag // this is ending tag </..>
+    };
+    Type type;

 
     // end positions for the tag:
     // end1 is '<' of ending tag,
     // end2 is '>' or both are
 
     // end positions for the tag:
     // end1 is '<' of ending tag,
     // end2 is '>' or both are

-    // -1 if there is no ending tag for this one...
-    // or -2 if this is ending tag  </...>
-    int End1, End2;
+    wxString::const_iterator End1, End2;

 
     // name of this tag
     wxChar *Name;
 
     // name of this tag
     wxChar *Name;


@@ -60,95 +67,108 @@ bool wxIsCDATAElement(const wxChar *tag)
            (wxStrcmp(tag, _T("STYLE")) == 0);
 }
 
            (wxStrcmp(tag, _T("STYLE")) == 0);
 }
 

+bool wxIsCDATAElement(const wxString& tag)
+{
+    return (wxStrcmp(tag.wx_str(), wxSTRING_TEXT("SCRIPT")) == 0) ||
+           (wxStrcmp(tag.wx_str(), wxSTRING_TEXT("STYLE")) == 0);
+}
+

 wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
 {
     m_Cache = new wxHtmlTagsCacheData;
     m_CachePos = 0;
 
 wxHtmlTagsCache::wxHtmlTagsCache(const wxString& source)
 {
     m_Cache = new wxHtmlTagsCacheData;
     m_CachePos = 0;
 

-    const wxChar *src = source.c_str();
-    int lng = source.length();

     wxChar tagBuffer[256];
 
     wxChar tagBuffer[256];
 

-    for ( int pos = 0; pos < lng; pos++ )
+    const wxString::const_iterator end = source.end();
+    for ( wxString::const_iterator pos = source.begin(); pos < end; ++pos )

     {
     {

-        if (src[pos] == wxT('<'))   // tag found:
+        if (*pos == wxT('<'))   // tag found:

         {
             // don't cache comment tags
         {
             // don't cache comment tags

-            wxString::const_iterator iter = source.begin() + pos;
-            if ( wxHtmlParser::SkipCommentTag(iter, source.end()) )
-            {
-                pos = iter - source.begin();
+            if ( wxHtmlParser::SkipCommentTag(pos, source.end()) )

                 continue;
                 continue;

-            }

 
             size_t tg = Cache().size();
             Cache().push_back(wxHtmlCacheItem());
 
 
             size_t tg = Cache().size();
             Cache().push_back(wxHtmlCacheItem());
 

-            int stpos = pos++;
+            wxString::const_iterator stpos = pos++;

             Cache()[tg].Key = stpos;
 
             int i;
             for ( i = 0;
             Cache()[tg].Key = stpos;
 
             int i;
             for ( i = 0;

-                  pos < lng && i < (int)WXSIZEOF(tagBuffer) - 1 &&
-                  src[pos] != wxT('>') && !wxIsspace(src[pos]);
-                  i++, pos++ )
+                  pos < end && i < (int)WXSIZEOF(tagBuffer) - 1 &&
+                  *pos != wxT('>') && !wxIsspace(*pos);
+                  ++i, ++pos )

             {
             {

-                tagBuffer[i] = (wxChar)wxToupper(src[pos]);
+                tagBuffer[i] = (wxChar)wxToupper(*pos);

             }
             tagBuffer[i] = _T('\0');
 
             Cache()[tg].Name = new wxChar[i+1];
             memcpy(Cache()[tg].Name, tagBuffer, (i+1)*sizeof(wxChar));
 
             }
             tagBuffer[i] = _T('\0');
 
             Cache()[tg].Name = new wxChar[i+1];
             memcpy(Cache()[tg].Name, tagBuffer, (i+1)*sizeof(wxChar));
 

-            while (pos < lng && src[pos] != wxT('>')) pos++;
+            while (pos < end && *pos != wxT('>'))
+                ++pos;

 
 

-            if (src[stpos+1] == wxT('/')) // ending tag:
+            if ((stpos+1) < end && *(stpos+1) == wxT('/')) // ending tag:

             {
             {

-                Cache()[tg].End1 = Cache()[tg].End2 = -2;
+                Cache()[tg].type = wxHtmlCacheItem::Type_EndingTag;

                 // find matching begin tag:
                 for (i = tg; i >= 0; i--)
                 // find matching begin tag:
                 for (i = tg; i >= 0; i--)

-                    if ((Cache()[i].End1 == -1) && (wxStrcmp(Cache()[i].Name, tagBuffer+1) == 0))
+                {
+                    if ((Cache()[i].type == wxHtmlCacheItem::Type_NoMatchingEndingTag) && (wxStrcmp(Cache()[i].Name, tagBuffer+1) == 0))

                     {
                     {

+                        Cache()[i].type = wxHtmlCacheItem::Type_Normal;

                         Cache()[i].End1 = stpos;
                         Cache()[i].End2 = pos + 1;
                         break;
                     }
                         Cache()[i].End1 = stpos;
                         Cache()[i].End2 = pos + 1;
                         break;
                     }

+                }

             }
             else
             {
             }
             else
             {

-                Cache()[tg].End1 = Cache()[tg].End2 = -1;
+                Cache()[tg].type = wxHtmlCacheItem::Type_NoMatchingEndingTag;

 
                 if (wxIsCDATAElement(tagBuffer))
                 {
                     // store the orig pos in case we are missing the closing
                     // tag (see below)
 
                 if (wxIsCDATAElement(tagBuffer))
                 {
                     // store the orig pos in case we are missing the closing
                     // tag (see below)

-                    wxInt32 old_pos = pos;
+                    const wxString::const_iterator old_pos = pos;

                     bool foundCloseTag = false;
 
                     // find next matching tag
                     int tag_len = wxStrlen(tagBuffer);
                     bool foundCloseTag = false;
 
                     // find next matching tag
                     int tag_len = wxStrlen(tagBuffer);

-                    while (pos < lng)
+                    while (pos < end)

                     {
                         // find the ending tag
                     {
                         // find the ending tag

-                        while (pos + 1 < lng &&
-                               (src[pos] != '<' || src[pos+1] != '/'))
+                        while (pos + 1 < end &&
+                               (*pos != '<' || *(pos+1) != '/'))

                             ++pos;
                             ++pos;

-                        if (src[pos] == '<')
+                        if (*pos == '<')

                             ++pos;
 
                         // see if it matches
                         int match_pos = 0;
                             ++pos;
 
                         // see if it matches
                         int match_pos = 0;

-                        while (pos < lng && match_pos < tag_len && src[pos] != '>' && src[pos] != '<') {
+                        while (pos < end && match_pos < tag_len )
+                        {
+                            wxChar c = *pos;
+                            if ( c == '>' || c == '<' )
+                                break;
+

                             // cast to wxChar needed to suppress warning in
                             // Unicode build
                             // cast to wxChar needed to suppress warning in
                             // Unicode build

-                            if ((wxChar)wxToupper(src[pos]) == tagBuffer[match_pos]) {
+                            if ((wxChar)wxToupper(c) == tagBuffer[match_pos])
+                            {

                                 ++match_pos;
                             }
                                 ++match_pos;
                             }

-                            else if (src[pos] == wxT(' ') || src[pos] == wxT('\n') ||
-                                src[pos] == wxT('\r') || src[pos] == wxT('\t')) {
+                            else if (c == wxT(' ') || c == wxT('\n') ||
+                                c == wxT('\r') || c == wxT('\t'))
+                            {

                                 // need to skip over these
                             }
                                 // need to skip over these
                             }

-                            else {
+                            else
+                            {

                                 match_pos = 0;
                             }
                             ++pos;
                                 match_pos = 0;
                             }
                             ++pos;


@@ -193,7 +213,11 @@ wxHtmlTagsCache::~wxHtmlTagsCache()
     delete m_Cache;
 }
 
     delete m_Cache;
 }
 

-void wxHtmlTagsCache::QueryTag(int at, int* end1, int* end2)
+void wxHtmlTagsCache::QueryTag(const wxString::const_iterator& at,
+                               const wxString::const_iterator& inputEnd,
+                               wxString::const_iterator *end1,
+                               wxString::const_iterator *end2,
+                               bool *hasEnding)

 {
     if (Cache().empty())
         return;
 {
     if (Cache().empty())
         return;


@@ -203,22 +227,28 @@ void wxHtmlTagsCache::QueryTag(int at, int* end1, int* end2)
         int delta = (at < Cache()[m_CachePos].Key) ? -1 : 1;
         do
         {
         int delta = (at < Cache()[m_CachePos].Key) ? -1 : 1;
         do
         {

-            if ( m_CachePos < 0 || m_CachePos == Cache().size() )
+            m_CachePos += delta;
+
+            if ( m_CachePos < 0 || m_CachePos >= (int)Cache().size() )

             {
             {

+                if ( m_CachePos < 0 )
+                    m_CachePos = 0;
+                else
+                    m_CachePos = Cache().size() - 1;

                 // something is very wrong with HTML, give up by returning an
                 // impossibly large value which is going to be ignored by the
                 // caller
                 *end1 =
                 // something is very wrong with HTML, give up by returning an
                 // impossibly large value which is going to be ignored by the
                 // caller
                 *end1 =

-                *end2 = INT_MAX;
+                *end2 = inputEnd;
+                *hasEnding = true;

                 return;
             }
                 return;
             }

-
-            m_CachePos += delta;

         }
         while (Cache()[m_CachePos].Key != at);
     }
     *end1 = Cache()[m_CachePos].End1;
     *end2 = Cache()[m_CachePos].End2;
         }
         while (Cache()[m_CachePos].Key != at);
     }
     *end1 = Cache()[m_CachePos].End1;
     *end2 = Cache()[m_CachePos].End2;

+    *hasEnding = (Cache()[m_CachePos].type == wxHtmlCacheItem::Type_Normal);

 }
 
 
 }
 
 


@@ -229,7 +259,9 @@ void wxHtmlTagsCache::QueryTag(int at, int* end1, int* end2)
 //-----------------------------------------------------------------------------
 
 wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
 //-----------------------------------------------------------------------------
 
 wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,

-                     const wxString& source, int pos, int end_pos,
+                     const wxString *source,
+                     const wxString::const_iterator& pos,
+                     const wxString::const_iterator& end_pos,

                      wxHtmlTagsCache *cache,
                      wxHtmlEntitiesParser *entParser)
 {
                      wxHtmlTagsCache *cache,
                      wxHtmlEntitiesParser *entParser)
 {


@@ -252,15 +284,14 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
 
     /* Find parameters and their values: */
 
 
     /* Find parameters and their values: */
 

-    int i;

     wxChar c;
 
     // fill-in name, params and begin pos:
     wxChar c;
 
     // fill-in name, params and begin pos:

-    i = pos+1;
+    wxString::const_iterator i(pos+1);

 
     // find tag's name and convert it to uppercase:
     while ((i < end_pos) &&
 
     // find tag's name and convert it to uppercase:
     while ((i < end_pos) &&

-           ((c = source[i++]) != wxT(' ') && c != wxT('\r') &&
+           ((c = *(i++)) != wxT(' ') && c != wxT('\r') &&

              c != wxT('\n') && c != wxT('\t') &&
              c != wxT('>')))
     {
              c != wxT('\n') && c != wxT('\t') &&
              c != wxT('>')))
     {


@@ -272,7 +303,7 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
     // if the tag has parameters, read them and "normalize" them,
     // i.e. convert to uppercase, replace whitespaces by spaces and
     // remove whitespaces around '=':
     // if the tag has parameters, read them and "normalize" them,
     // i.e. convert to uppercase, replace whitespaces by spaces and
     // remove whitespaces around '=':

-    if (source[i-1] != wxT('>'))
+    if (*(i-1) != wxT('>'))

     {
         #define IS_WHITE(c) (c == wxT(' ') || c == wxT('\r') || \
                              c == wxT('\n') || c == wxT('\t'))
     {
         #define IS_WHITE(c) (c == wxT(' ') || c == wxT('\r') || \
                              c == wxT('\n') || c == wxT('\t'))


@@ -291,14 +322,14 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
         state = ST_BEFORE_NAME;
         while (i < end_pos)
         {
         state = ST_BEFORE_NAME;
         while (i < end_pos)
         {

-            c = source[i++];
+            c = *(i++);

 
             if (c == wxT('>') && !(state == ST_VALUE && quote != 0))
             {
                 if (state == ST_BEFORE_EQ || state == ST_NAME)
                 {
                     m_ParamNames.Add(pname);
 
             if (c == wxT('>') && !(state == ST_VALUE && quote != 0))
             {
                 if (state == ST_BEFORE_EQ || state == ST_NAME)
                 {
                     m_ParamNames.Add(pname);

-                    m_ParamValues.Add(wxEmptyString);
+                    m_ParamValues.Add(wxGetEmptyString());

                 }
                 else if (state == ST_VALUE && quote == 0)
                 {
                 }
                 else if (state == ST_VALUE && quote == 0)
                 {


@@ -333,7 +364,7 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
                     else if (!IS_WHITE(c))
                     {
                         m_ParamNames.Add(pname);
                     else if (!IS_WHITE(c))
                     {
                         m_ParamNames.Add(pname);

-                        m_ParamValues.Add(wxEmptyString);
+                        m_ParamValues.Add(wxGetEmptyString());

                         pname = c;
                         state = ST_NAME;
                     }
                         pname = c;
                         state = ST_NAME;
                     }


@@ -342,7 +373,7 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
                     if (!IS_WHITE(c))
                     {
                         if (c == wxT('"') || c == wxT('\''))
                     if (!IS_WHITE(c))
                     {
                         if (c == wxT('"') || c == wxT('\''))

-                            quote = c, pvalue = wxEmptyString;
+                            quote = c, pvalue = wxGetEmptyString();

                         else
                             quote = 0, pvalue = c;
                         state = ST_VALUE;
                         else
                             quote = 0, pvalue = c;
                         state = ST_VALUE;


@@ -374,10 +405,13 @@ wxHtmlTag::wxHtmlTag(wxHtmlTag *parent,
         #undef IS_WHITE
     }
     m_Begin = i;
         #undef IS_WHITE
     }
     m_Begin = i;

-
-    cache->QueryTag(pos, &m_End1, &m_End2);
+    cache->QueryTag(pos, source->end(), &m_End1, &m_End2, &m_hasEnding);

     if (m_End1 > end_pos) m_End1 = end_pos;
     if (m_End2 > end_pos) m_End2 = end_pos;
     if (m_End1 > end_pos) m_End1 = end_pos;
     if (m_End2 > end_pos) m_End2 = end_pos;

+
+#if WXWIN_COMPATIBILITY_2_8
+    m_sourceStart = source->begin();
+#endif

 }
 
 wxHtmlTag::~wxHtmlTag()
 }
 
 wxHtmlTag::~wxHtmlTag()


@@ -401,7 +435,7 @@ wxString wxHtmlTag::GetParam(const wxString& par, bool with_commas) const
 {
     int index = m_ParamNames.Index(par, false);
     if (index == wxNOT_FOUND)
 {
     int index = m_ParamNames.Index(par, false);
     if (index == wxNOT_FOUND)

-        return wxEmptyString;
+        return wxGetEmptyString();

     if (with_commas)
     {
         // VS: backward compatibility, seems to be never used by wxHTML...
     if (with_commas)
     {
         // VS: backward compatibility, seems to be never used by wxHTML...


@@ -438,8 +472,8 @@ bool wxHtmlTag::GetParamAsColour(const wxString& par, wxColour *clr) const
     // handle colours defined in HTML 4.0 first:
     if (str.length() > 1 && str[0] != _T('#'))
     {
     // handle colours defined in HTML 4.0 first:
     if (str.length() > 1 && str[0] != _T('#'))
     {

-        #define HTML_COLOUR(name, r, g, b)              \
-            if (str.IsSameAs(wxT(name), false))         \
+        #define HTML_COLOUR(name, r, g, b)                        \
+            if (str.IsSameAs(wxSTRING_TEXT(name), false))         \

                 { clr->Set(r, g, b); return true; }
         HTML_COLOUR("black",   0x00,0x00,0x00)
         HTML_COLOUR("silver",  0xC0,0xC0,0xC0)
                 { clr->Set(r, g, b); return true; }
         HTML_COLOUR("black",   0x00,0x00,0x00)
         HTML_COLOUR("silver",  0xC0,0xC0,0xC0)

diff --git a/src/html/m_layout.cpp b/src/html/m_layout.cpp
index 1bea3402014246c47a75f8026e0bba6eeb65bd4c..2451f2fb539a50e2d16aa30a9acf1e6489e839a7 100644 (file)
--- a/src/html/m_layout.cpp
+++ b/src/html/m_layout.cpp
@@ -289,9 +289,7 @@ TAG_HANDLER_BEGIN(TITLE, "TITLE")
         wxHtmlWindowInterface *winIface = m_WParser->GetWindowInterface();
         if (winIface)
         {
         wxHtmlWindowInterface *winIface = m_WParser->GetWindowInterface();
         if (winIface)
         {

-            wxString title = m_WParser->GetSource()->Mid(
-                                    tag.GetBeginPos(),
-                                    tag.GetEndPos1()-tag.GetBeginPos());
+            wxString title(tag.GetBeginIter(), tag.GetEndIter1());

 #if !wxUSE_UNICODE && wxUSE_WCHAR_T
             const wxFontEncoding enc = m_WParser->GetInputEncoding();
             if ( enc != wxFONTENCODING_DEFAULT )
 #if !wxUSE_UNICODE && wxUSE_WCHAR_T
             const wxFontEncoding enc = m_WParser->GetInputEncoding();
             if ( enc != wxFONTENCODING_DEFAULT )