include/wx/html/htmlpars.h

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        htmlpars.h
   3 // Purpose:     wxHtmlParser class (generic parser)
   4 // Author:      Vaclav Slavik
   5 // RCS-ID:      $Id$
   6 // Copyright:   (c) 1999 Vaclav Slavik
   7 // Licence:     wxWindows Licence
   8 /////////////////////////////////////////////////////////////////////////////
   9
  10
  11 #ifndef _WX_HTMLPARS_H_
  12 #define _WX_HTMLPARS_H_
  13
  14 #ifdef __GNUG__
  15 #pragma interface "htmlpars.h"
  16 #endif
  17
  18 #include "wx/defs.h"
  19 #if wxUSE_HTML
  20
  21 #include "wx/html/htmltag.h"
  22 #include "wx/filesys.h"
  23 #include "wx/hash.h"
  24 #include "wx/fontenc.h"
  25
  26 class WXDLLEXPORT wxMBConv;
  27 class WXDLLEXPORT wxHtmlParser;
  28 class WXDLLEXPORT wxHtmlTagHandler;
  29 class WXDLLEXPORT wxHtmlEntitiesParser;
  30
  31 class wxHtmlTextPieces;
  32 class wxHtmlParserState;
  33
  34 // This class handles generic parsing of HTML document : it scans
  35 // the document and divide it into blocks of tags (where one block
  36 // consists of starting and ending tag and of text between these
  37 // 2 tags.
  38 class WXDLLEXPORT wxHtmlParser : public wxObject
  39 {
  40     DECLARE_ABSTRACT_CLASS(wxHtmlParser)
  41
  42 public:
  43     wxHtmlParser();
  44     virtual ~wxHtmlParser();
  45
  46     // Sets the class which will be used for opening files
  47     void SetFS(wxFileSystem *fs) { m_FS = fs; }
  48
  49     wxFileSystem* GetFS() const { return m_FS; }
  50
  51     // You can simply call this method when you need parsed output.
  52     // This method does these things:
  53     // 1. call InitParser(source);
  54     // 2. call DoParsing();
  55     // 3. call GetProduct(); (it's return value is then returned)
  56     // 4. call DoneParser();
  57     wxObject* Parse(const wxString& source);
  58
  59     // Sets the source. This must be called before running Parse() method.
  60     virtual void InitParser(const wxString& source);
  61     // This must be called after Parse().
  62     virtual void DoneParser();
  63
  64     // Parses the m_Source from begin_pos to end_pos-1.
  65     // (in noparams version it parses whole m_Source)
  66     void DoParsing(int begin_pos, int end_pos);
  67     void DoParsing();
  68
  69     // Returns pointer to the tag at parser's current position
  70     wxHtmlTag *GetCurrentTag() const { return m_CurTag; }
  71
  72     // Returns product of parsing
  73     // Returned value is result of parsing of the part. The type of this result
  74     // depends on internal representation in derived parser
  75     // (see wxHtmlWinParser for details).
  76     virtual wxObject* GetProduct() = 0;
  77
  78     // adds handler to the list & hash table of handlers.
  79     virtual void AddTagHandler(wxHtmlTagHandler *handler);
  80
  81     // Forces the handler to handle additional tags (not returned by GetSupportedTags).
  82     // The handler should already be in use by this parser.
  83     // Example: you want to parse following pseudo-html structure:
  84     //   <myitems>
  85     //     <it name="one" value="1">
  86     //     <it name="two" value="2">
  87     //   </myitems>
  88     //   <it> This last it has different meaning, we don't want it to be parsed by myitems handler!
  89     // handler can handle only 'myitems' (e.g. it's GetSupportedTags returns "MYITEMS")
  90     // you can call PushTagHandler(handler, "IT") when you find <myitems>
  91     // and call PopTagHandler() when you find </myitems>
  92     void PushTagHandler(wxHtmlTagHandler *handler, wxString tags);
  93
  94     // Restores state before last call to PushTagHandler
  95     void PopTagHandler();
  96
  97     wxString* GetSource() {return &m_Source;}
  98     void SetSource(const wxString& src);
  99
 100     // Sets HTML source and remebers current parser's state so that it can
 101     // later be restored. This is useful for on-line modifications of
 102     // HTML source (for example, <pre> handler replaces spaces with &nbsp;
 103     // and newlines with <br>)
 104     virtual void SetSourceAndSaveState(const wxString& src);
 105     // Restores parser's state from stack or returns FALSE if the stack is
 106     // empty
 107     virtual bool RestoreState();
 108
 109 protected:
 110     // DOM structure
 111     void CreateDOMTree();
 112     void DestroyDOMTree();
 113     void CreateDOMSubTree(wxHtmlTag *cur,
 114                           int begin_pos, int end_pos,
 115                           wxHtmlTagsCache *cache);
 116
 117     // Adds text to the output.
 118     // This is called from Parse() and must be overriden in derived classes.
 119     // txt is not guaranteed to be only one word. It is largest continuous part of text
 120     // (= not broken by tags)
 121     // NOTE : using char* because of speed improvements
 122     virtual void AddText(const wxChar* txt) = 0;
 123
 124     // Adds tag and proceeds it. Parse() may (and usually is) called from this method.
 125     // This is called from Parse() and may be overriden.
 126     // Default behavior is that it looks for proper handler in m_Handlers. The tag is
 127     // ignored if no hander is found.
 128     // Derived class is *responsible* for filling in m_Handlers table.
 129     virtual void AddTag(const wxHtmlTag& tag);
 130
 131     // Returns entity parser object, used to substitute HTML &entities;
 132     wxHtmlEntitiesParser *GetEntitiesParser() const { return m_entitiesParser; }
 133
 134 protected:
 135     // DOM tree:
 136     wxHtmlTag *m_CurTag;
 137     wxHtmlTag *m_Tags;
 138     wxHtmlTextPieces *m_TextPieces;
 139     size_t m_CurTextPiece;
 140
 141     wxString m_Source;
 142
 143     wxHtmlParserState *m_SavedStates;
 144
 145     // handlers that handle particular tags. The table is accessed by
 146     // key = tag's name.
 147     // This attribute MUST be filled by derived class otherwise it would
 148     // be empty and no tags would be recognized
 149     // (see wxHtmlWinParser for details about filling it)
 150     // m_HandlersHash is for random access based on knowledge of tag name (BR, P, etc.)
 151     //      it may (and often does) contain more references to one object
 152     // m_HandlersList is list of all handlers and it is guaranteed to contain
 153     //      only one reference to each handler instance.
 154     wxList m_HandlersList;
 155     wxHashTable m_HandlersHash;
 156
 157     // class for opening files (file system)
 158     wxFileSystem *m_FS;
 159     // handlers stack used by PushTagHandler and PopTagHandler
 160     wxList *m_HandlersStack;
 161
 162     // entity parse
 163     wxHtmlEntitiesParser *m_entitiesParser;
 164 };
 165
 166
 167
 168 // This class (and derived classes) cooperates with wxHtmlParser.
 169 // Each recognized tag is passed to handler which is capable
 170 // of handling it. Each tag is handled in 3 steps:
 171 // 1. Handler will modifies state of parser
 172 //    (using it's public methods)
 173 // 2. Parser parses source between starting and ending tag
 174 // 3. Handler restores original state of the parser
 175 class WXDLLEXPORT wxHtmlTagHandler : public wxObject
 176 {
 177     DECLARE_ABSTRACT_CLASS(wxHtmlTagHandler)
 178
 179 public:
 180     wxHtmlTagHandler() : wxObject () { m_Parser = NULL; }
 181
 182     // Sets the parser.
 183     // NOTE : each _instance_ of handler is guaranteed to be called
 184     // only by one parser. This means you don't have to care about
 185     // reentrancy.
 186     virtual void SetParser(wxHtmlParser *parser)
 187         { m_Parser = parser; }
 188
 189     // Returns list of supported tags. The list is in uppercase and
 190     // tags are delimited by ','.
 191     // Example : "I,B,FONT,P"
 192     //   is capable of handling italic, bold, font and paragraph tags
 193     virtual wxString GetSupportedTags() = 0;
 194
 195     // This is hadling core method. It does all the Steps 1-3.
 196     // To process step 2, you can call ParseInner()
 197     // returned value : TRUE if it called ParseInner(),
 198     //                  FALSE etherwise
 199     virtual bool HandleTag(const wxHtmlTag& tag) = 0;
 200
 201 protected:
 202     // parses input between beginning and ending tag.
 203     // m_Parser must be set.
 204     void ParseInner(const wxHtmlTag& tag)
 205         { m_Parser->DoParsing(tag.GetBeginPos(), tag.GetEndPos1()); }
 206
 207     wxHtmlParser *m_Parser;
 208 };
 209
 210
 211 // This class is used to parse HTML entities in strings. It can handle
 212 // both named entities and &#xxxx entries where xxxx is Unicode code.
 213 class WXDLLEXPORT wxHtmlEntitiesParser : public wxObject
 214 {
 215     DECLARE_DYNAMIC_CLASS(wxHtmlEntitiesParser)
 216
 217 public:
 218     wxHtmlEntitiesParser();
 219     virtual ~wxHtmlEntitiesParser();
 220
 221     // Sets encoding of output string.
 222     // Has no effect if wxUSE_WCHAR_T==0 or wxUSE_UNICODE==1
 223     void SetEncoding(wxFontEncoding encoding);
 224
 225     // Parses entities in input and replaces them with respective characters
 226     // (with respect to output encoding)
 227     wxString Parse(const wxString& input);
 228
 229     // Returns character for given entity or 0 if the enity is unknown
 230     wxChar GetEntityChar(const wxString& entity);
 231
 232     // Returns character that represents given Unicode code
 233     wxChar GetCharForCode(unsigned code);
 234
 235 protected:
 236 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 237     wxMBConv *m_conv;
 238     wxFontEncoding m_encoding;
 239 #endif
 240 };
 241
 242
 243 #endif
 244
 245 #endif // _WX_HTMLPARS_H_