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     // Returns TRUE if the parser is allowed to open given URL (may be forbidden
  52     // for security reasons)
  53     virtual bool CanOpenURL(const wxString& url) const { return TRUE; }
  54
  55     // You can simply call this method when you need parsed output.
  56     // This method does these things:
  57     // 1. call InitParser(source);
  58     // 2. call DoParsing();
  59     // 3. call GetProduct(); (it's return value is then returned)
  60     // 4. call DoneParser();
  61     wxObject* Parse(const wxString& source);
  62
  63     // Sets the source. This must be called before running Parse() method.
  64     virtual void InitParser(const wxString& source);
  65     // This must be called after Parse().
  66     virtual void DoneParser();
  67
  68     // Parses the m_Source from begin_pos to end_pos-1.
  69     // (in noparams version it parses whole m_Source)
  70     void DoParsing(int begin_pos, int end_pos);
  71     void DoParsing();
  72
  73     // Returns pointer to the tag at parser's current position
  74     wxHtmlTag *GetCurrentTag() const { return m_CurTag; }
  75
  76     // Returns product of parsing
  77     // Returned value is result of parsing of the part. The type of this result
  78     // depends on internal representation in derived parser
  79     // (see wxHtmlWinParser for details).
  80     virtual wxObject* GetProduct() = 0;
  81
  82     // adds handler to the list & hash table of handlers.
  83     virtual void AddTagHandler(wxHtmlTagHandler *handler);
  84
  85     // Forces the handler to handle additional tags (not returned by GetSupportedTags).
  86     // The handler should already be in use by this parser.
  87     // Example: you want to parse following pseudo-html structure:
  88     //   <myitems>
  89     //     <it name="one" value="1">
  90     //     <it name="two" value="2">
  91     //   </myitems>
  92     //   <it> This last it has different meaning, we don't want it to be parsed by myitems handler!
  93     // handler can handle only 'myitems' (e.g. it's GetSupportedTags returns "MYITEMS")
  94     // you can call PushTagHandler(handler, "IT") when you find <myitems>
  95     // and call PopTagHandler() when you find </myitems>
  96     void PushTagHandler(wxHtmlTagHandler *handler, wxString tags);
  97
  98     // Restores state before last call to PushTagHandler
  99     void PopTagHandler();
 100
 101     wxString* GetSource() {return &m_Source;}
 102     void SetSource(const wxString& src);
 103
 104     // Sets HTML source and remebers current parser's state so that it can
 105     // later be restored. This is useful for on-line modifications of
 106     // HTML source (for example, <pre> handler replaces spaces with &nbsp;
 107     // and newlines with <br>)
 108     virtual void SetSourceAndSaveState(const wxString& src);
 109     // Restores parser's state from stack or returns FALSE if the stack is
 110     // empty
 111     virtual bool RestoreState();
 112
 113 protected:
 114     // DOM structure
 115     void CreateDOMTree();
 116     void DestroyDOMTree();
 117     void CreateDOMSubTree(wxHtmlTag *cur,
 118                           int begin_pos, int end_pos,
 119                           wxHtmlTagsCache *cache);
 120
 121     // Adds text to the output.
 122     // This is called from Parse() and must be overriden in derived classes.
 123     // txt is not guaranteed to be only one word. It is largest continuous part of text
 124     // (= not broken by tags)
 125     // NOTE : using char* because of speed improvements
 126     virtual void AddText(const wxChar* txt) = 0;
 127
 128     // Adds tag and proceeds it. Parse() may (and usually is) called from this method.
 129     // This is called from Parse() and may be overriden.
 130     // Default behavior is that it looks for proper handler in m_Handlers. The tag is
 131     // ignored if no hander is found.
 132     // Derived class is *responsible* for filling in m_Handlers table.
 133     virtual void AddTag(const wxHtmlTag& tag);
 134
 135     // Returns entity parser object, used to substitute HTML &entities;
 136     wxHtmlEntitiesParser *GetEntitiesParser() const { return m_entitiesParser; }
 137
 138 protected:
 139     // DOM tree:
 140     wxHtmlTag *m_CurTag;
 141     wxHtmlTag *m_Tags;
 142     wxHtmlTextPieces *m_TextPieces;
 143     size_t m_CurTextPiece;
 144
 145     wxString m_Source;
 146
 147     wxHtmlParserState *m_SavedStates;
 148
 149     // handlers that handle particular tags. The table is accessed by
 150     // key = tag's name.
 151     // This attribute MUST be filled by derived class otherwise it would
 152     // be empty and no tags would be recognized
 153     // (see wxHtmlWinParser for details about filling it)
 154     // m_HandlersHash is for random access based on knowledge of tag name (BR, P, etc.)
 155     //      it may (and often does) contain more references to one object
 156     // m_HandlersList is list of all handlers and it is guaranteed to contain
 157     //      only one reference to each handler instance.
 158     wxList m_HandlersList;
 159     wxHashTable m_HandlersHash;
 160
 161     // class for opening files (file system)
 162     wxFileSystem *m_FS;
 163     // handlers stack used by PushTagHandler and PopTagHandler
 164     wxList *m_HandlersStack;
 165
 166     // entity parse
 167     wxHtmlEntitiesParser *m_entitiesParser;
 168 };
 169
 170
 171
 172 // This class (and derived classes) cooperates with wxHtmlParser.
 173 // Each recognized tag is passed to handler which is capable
 174 // of handling it. Each tag is handled in 3 steps:
 175 // 1. Handler will modifies state of parser
 176 //    (using it's public methods)
 177 // 2. Parser parses source between starting and ending tag
 178 // 3. Handler restores original state of the parser
 179 class WXDLLEXPORT wxHtmlTagHandler : public wxObject
 180 {
 181     DECLARE_ABSTRACT_CLASS(wxHtmlTagHandler)
 182
 183 public:
 184     wxHtmlTagHandler() : wxObject () { m_Parser = NULL; }
 185
 186     // Sets the parser.
 187     // NOTE : each _instance_ of handler is guaranteed to be called
 188     // only by one parser. This means you don't have to care about
 189     // reentrancy.
 190     virtual void SetParser(wxHtmlParser *parser)
 191         { m_Parser = parser; }
 192
 193     // Returns list of supported tags. The list is in uppercase and
 194     // tags are delimited by ','.
 195     // Example : "I,B,FONT,P"
 196     //   is capable of handling italic, bold, font and paragraph tags
 197     virtual wxString GetSupportedTags() = 0;
 198
 199     // This is hadling core method. It does all the Steps 1-3.
 200     // To process step 2, you can call ParseInner()
 201     // returned value : TRUE if it called ParseInner(),
 202     //                  FALSE etherwise
 203     virtual bool HandleTag(const wxHtmlTag& tag) = 0;
 204
 205 protected:
 206     // parses input between beginning and ending tag.
 207     // m_Parser must be set.
 208     void ParseInner(const wxHtmlTag& tag)
 209         { m_Parser->DoParsing(tag.GetBeginPos(), tag.GetEndPos1()); }
 210
 211     wxHtmlParser *m_Parser;
 212 };
 213
 214
 215 // This class is used to parse HTML entities in strings. It can handle
 216 // both named entities and &#xxxx entries where xxxx is Unicode code.
 217 class WXDLLEXPORT wxHtmlEntitiesParser : public wxObject
 218 {
 219     DECLARE_DYNAMIC_CLASS(wxHtmlEntitiesParser)
 220
 221 public:
 222     wxHtmlEntitiesParser();
 223     virtual ~wxHtmlEntitiesParser();
 224
 225     // Sets encoding of output string.
 226     // Has no effect if wxUSE_WCHAR_T==0 or wxUSE_UNICODE==1
 227     void SetEncoding(wxFontEncoding encoding);
 228
 229     // Parses entities in input and replaces them with respective characters
 230     // (with respect to output encoding)
 231     wxString Parse(const wxString& input);
 232
 233     // Returns character for given entity or 0 if the enity is unknown
 234     wxChar GetEntityChar(const wxString& entity);
 235
 236     // Returns character that represents given Unicode code
 237     wxChar GetCharForCode(unsigned code);
 238
 239 protected:
 240 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 241     wxMBConv *m_conv;
 242     wxFontEncoding m_encoding;
 243 #endif
 244 };
 245
 246
 247 #endif
 248
 249 #endif // _WX_HTMLPARS_H_