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
  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
  24 class wxHtmlParser;
  25 class wxHtmlTagHandler;
  26
  27 //--------------------------------------------------------------------------------
  28 // wxHtmlParser
  29 //                  This class handles generic parsing of HTML document : it scans
  30 //                  the document and divide it into blocks of tags (where one block
  31 //                  consists of starting and ending tag and of text between these
  32 //                  2 tags.
  33 //--------------------------------------------------------------------------------
  34
  35 class WXDLLEXPORT wxHtmlParser : public wxObject
  36 {
  37     DECLARE_ABSTRACT_CLASS(wxHtmlParser)
  38
  39 public:
  40     wxHtmlParser() : wxObject(), m_HandlersHash(wxKEY_STRING)
  41         { m_FS = NULL; m_Cache = NULL; m_HandlersStack = NULL; }
  42     virtual ~wxHtmlParser();
  43
  44     // Sets the class which will be used for opening files
  45     void SetFS(wxFileSystem *fs) { m_FS = fs; }
  46
  47     wxFileSystem* GetFS() const { return m_FS; }
  48
  49     // You can simply call this method when you need parsed output.
  50     // This method does these things:
  51     // 1. call InitParser(source);
  52     // 2. call DoParsing();
  53     // 3. call GetProduct(); (it's return value is then returned)
  54     // 4. call DoneParser();
  55     wxObject* Parse(const wxString& source);
  56
  57     // Sets the source. This must be called before running Parse() method.
  58     virtual void InitParser(const wxString& source);
  59     // This must be called after Parse().
  60     virtual void DoneParser();
  61
  62     // Parses the m_Source from begin_pos to end_pos-1.
  63     // (in noparams version it parses whole m_Source)
  64     void DoParsing(int begin_pos, int end_pos);
  65     inline void DoParsing() {DoParsing(0, m_Source.Length());};
  66
  67     // Returns product of parsing
  68     // Returned value is result of parsing of the part. The type of this result
  69     // depends on internal representation in derived parser
  70     // (see wxHtmlWinParser for details).
  71     virtual wxObject* GetProduct() = 0;
  72
  73     // adds handler to the list & hash table of handlers.
  74     virtual void AddTagHandler(wxHtmlTagHandler *handler);
  75
  76     // Forces the handler to handle additional tags (not returned by GetSupportedTags).
  77     // The handler should already be in use by this parser.
  78     // Example: you want to parse following pseudo-html structure:
  79     //   <myitems>
  80     //     <it name="one" value="1">
  81     //     <it name="two" value="2">
  82     //   </myitems>
  83     //   <it> This last it has different meaning, we don't want it to be parsed by myitems handler!
  84     // handler can handle only 'myitems' (e.g. it's GetSupportedTags returns "MYITEMS")
  85     // you can call PushTagHandler(handler, "IT") when you find <myitems>
  86     // and call PopTagHandler() when you find </myitems>
  87     void PushTagHandler(wxHtmlTagHandler *handler, wxString tags);
  88
  89     // Restores state before last call to PushTagHandler
  90     void PopTagHandler();
  91
  92     wxString* GetSource() {return &m_Source;}
  93     void SetSource(const wxString& src);
  94
  95 protected:
  96     // Adds text to the output.
  97     // This is called from Parse() and must be overriden in derived classes.
  98     // txt is not guaranteed to be only one word. It is largest continuous part of text
  99     // (= not broken by tags)
 100     // NOTE : using char* because of speed improvements
 101     virtual void AddText(const char* txt) = 0;
 102
 103     // Adds tag and proceeds it. Parse() may (and usually is) called from this method.
 104     // This is called from Parse() and may be overriden.
 105     // Default behavior is that it looks for proper handler in m_Handlers. The tag is
 106     // ignored if no hander is found.
 107     // Derived class is *responsible* for filling in m_Handlers table.
 108     virtual void AddTag(const wxHtmlTag& tag);
 109
 110 protected:
 111     // source being parsed
 112     wxString m_Source;
 113
 114     // tags cache, used during parsing.
 115     wxHtmlTagsCache *m_Cache;
 116     wxHashTable m_HandlersHash;
 117
 118     // handlers that handle particular tags. The table is accessed by
 119     // key = tag's name.
 120     // This attribute MUST be filled by derived class otherwise it would
 121     // be empty and no tags would be recognized
 122     // (see wxHtmlWinParser for details about filling it)
 123     // m_HandlersHash is for random access based on knowledge of tag name (BR, P, etc.)
 124     //      it may (and often does) contain more references to one object
 125     // m_HandlersList is list of all handlers and it is guaranteed to contain
 126     //      only one reference to each handler instance.
 127     wxList m_HandlersList;
 128
 129     // class for opening files (file system)
 130     wxFileSystem *m_FS;
 131     // handlers stack used by PushTagHandler and PopTagHandler
 132     wxList *m_HandlersStack;
 133 };
 134
 135
 136
 137
 138
 139
 140 //--------------------------------------------------------------------------------
 141 // wxHtmlTagHandler
 142 //                  This class (and derived classes) cooperates with wxHtmlParser.
 143 //                  Each recognized tag is passed to handler which is capable
 144 //                  of handling it. Each tag is handled in 3 steps:
 145 //                  1. Handler will modifies state of parser
 146 //                    (using it's public methods)
 147 //                  2. Parser parses source between starting and ending tag
 148 //                  3. Handler restores original state of the parser
 149 //--------------------------------------------------------------------------------
 150
 151 class WXDLLEXPORT wxHtmlTagHandler : public wxObject
 152 {
 153     DECLARE_ABSTRACT_CLASS(wxHtmlTagHandler)
 154
 155 public:
 156     wxHtmlTagHandler() : wxObject () { m_Parser = NULL; }
 157
 158     // Sets the parser.
 159     // NOTE : each _instance_ of handler is guaranteed to be called
 160     // only by one parser. This means you don't have to care about
 161     // reentrancy.
 162     virtual void SetParser(wxHtmlParser *parser)
 163         { m_Parser = parser; }
 164
 165     // Returns list of supported tags. The list is in uppercase and
 166     // tags are delimited by ','.
 167     // Example : "I,B,FONT,P"
 168     //   is capable of handling italic, bold, font and paragraph tags
 169     virtual wxString GetSupportedTags() = 0;
 170
 171     // This is hadling core method. It does all the Steps 1-3.
 172     // To process step 2, you can call ParseInner()
 173     // returned value : TRUE if it called ParseInner(),
 174     //                  FALSE etherwise
 175     virtual bool HandleTag(const wxHtmlTag& tag) = 0;
 176
 177 protected:
 178     // parses input between beginning and ending tag.
 179     // m_Parser must be set.
 180     void ParseInner(const wxHtmlTag& tag)
 181         { m_Parser->DoParsing(tag.GetBeginPos(), tag.GetEndPos1()); }
 182
 183     wxHtmlParser *m_Parser;
 184 };
 185
 186
 187
 188
 189 #endif
 190
 191 #endif // _WX_HTMLPARS_H_