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 #ifndef _WX_HTMLPARS_H_
  11 #define _WX_HTMLPARS_H_
  12
  13 #include "wx/defs.h"
  14 #if wxUSE_HTML
  15
  16 #include "wx/html/htmltag.h"
  17 #include "wx/filesys.h"
  18 #include "wx/hash.h"
  19 #include "wx/fontenc.h"
  20
  21 class WXDLLIMPEXP_FWD_BASE wxMBConv;
  22 class WXDLLIMPEXP_FWD_HTML wxHtmlParser;
  23 class WXDLLIMPEXP_FWD_HTML wxHtmlTagHandler;
  24 class WXDLLIMPEXP_FWD_HTML wxHtmlEntitiesParser;
  25
  26 class wxHtmlTextPieces;
  27 class wxHtmlParserState;
  28
  29
  30 enum wxHtmlURLType
  31 {
  32     wxHTML_URL_PAGE,
  33     wxHTML_URL_IMAGE,
  34     wxHTML_URL_OTHER
  35 };
  36
  37 // This class handles generic parsing of HTML document : it scans
  38 // the document and divides it into blocks of tags (where one block
  39 // consists of starting and ending tag and of text between these
  40 // 2 tags.
  41 class WXDLLIMPEXP_HTML wxHtmlParser : public wxObject
  42 {
  43     DECLARE_ABSTRACT_CLASS(wxHtmlParser)
  44
  45 public:
  46     wxHtmlParser();
  47     virtual ~wxHtmlParser();
  48
  49     // Sets the class which will be used for opening files
  50     void SetFS(wxFileSystem *fs) { m_FS = fs; }
  51
  52     wxFileSystem* GetFS() const { return m_FS; }
  53
  54     // Opens file if the parser is allowed to open given URL (may be forbidden
  55     // for security reasons)
  56     virtual wxFSFile *OpenURL(wxHtmlURLType type, const wxString& url) const;
  57
  58     // You can simply call this method when you need parsed output.
  59     // This method does these things:
  60     // 1. call InitParser(source);
  61     // 2. call DoParsing();
  62     // 3. call GetProduct(); (its return value is then returned)
  63     // 4. call DoneParser();
  64     wxObject* Parse(const wxString& source);
  65
  66     // Sets the source. This must be called before running Parse() method.
  67     virtual void InitParser(const wxString& source);
  68     // This must be called after Parse().
  69     virtual void DoneParser();
  70
  71     // May be called during parsing to immediately return from Parse().
  72     virtual void StopParsing() { m_stopParsing = true; }
  73
  74     // Parses the m_Source from begin_pos to end_pos-1.
  75     // (in noparams version it parses whole m_Source)
  76     void DoParsing(const wxString::const_iterator& begin_pos,
  77                    const wxString::const_iterator& end_pos);
  78     void DoParsing();
  79
  80     // Returns pointer to the tag at parser's current position
  81     wxHtmlTag *GetCurrentTag() const { return m_CurTag; }
  82
  83     // Returns product of parsing
  84     // Returned value is result of parsing of the part. The type of this result
  85     // depends on internal representation in derived parser
  86     // (see wxHtmlWinParser for details).
  87     virtual wxObject* GetProduct() = 0;
  88
  89     // adds handler to the list & hash table of handlers.
  90     virtual void AddTagHandler(wxHtmlTagHandler *handler);
  91
  92     // Forces the handler to handle additional tags (not returned by GetSupportedTags).
  93     // The handler should already be in use by this parser.
  94     // Example: you want to parse following pseudo-html structure:
  95     //   <myitems>
  96     //     <it name="one" value="1">
  97     //     <it name="two" value="2">
  98     //   </myitems>
  99     //   <it> This last it has different meaning, we don't want it to be parsed by myitems handler!
 100     // handler can handle only 'myitems' (e.g. its GetSupportedTags returns "MYITEMS")
 101     // you can call PushTagHandler(handler, "IT") when you find <myitems>
 102     // and call PopTagHandler() when you find </myitems>
 103     void PushTagHandler(wxHtmlTagHandler *handler, const wxString& tags);
 104
 105     // Restores state before last call to PushTagHandler
 106     void PopTagHandler();
 107
 108     const wxString* GetSource() {return m_Source;}
 109     void SetSource(const wxString& src);
 110
 111     // Sets HTML source and remembers current parser's state so that it can
 112     // later be restored. This is useful for on-line modifications of
 113     // HTML source (for example, <pre> handler replaces spaces with &nbsp;
 114     // and newlines with <br>)
 115     virtual void SetSourceAndSaveState(const wxString& src);
 116     // Restores parser's state from stack or returns false if the stack is
 117     // empty
 118     virtual bool RestoreState();
 119
 120     // Returns HTML source inside the element (i.e. between the starting
 121     // and ending tag)
 122     wxString GetInnerSource(const wxHtmlTag& tag);
 123
 124     // Parses HTML string 'markup' and extracts charset info from <meta> tag
 125     // if present. Returns empty string if the tag is missing.
 126     // For wxHTML's internal use.
 127     static wxString ExtractCharsetInformation(const wxString& markup);
 128
 129     // Returns entity parser object, used to substitute HTML &entities;
 130     wxHtmlEntitiesParser *GetEntitiesParser() const { return m_entitiesParser; }
 131
 132     // Returns true if the tag starting at the given position is a comment tag
 133     //
 134     // p should point to '<' character and is modified to point to the closing
 135     // '>' of the end comment tag if this is indeed a comment
 136     static bool
 137     SkipCommentTag(wxString::const_iterator& p, wxString::const_iterator end);
 138
 139 protected:
 140     // DOM structure
 141     void CreateDOMTree();
 142     void DestroyDOMTree();
 143     void CreateDOMSubTree(wxHtmlTag *cur,
 144                           const wxString::const_iterator& begin_pos,
 145                           const wxString::const_iterator& end_pos,
 146                           wxHtmlTagsCache *cache);
 147
 148     // Adds text to the output.
 149     // This is called from Parse() and must be overriden in derived classes.
 150     // txt is not guaranteed to be only one word. It is largest continuous part
 151     // of text (= not broken by tags)
 152     virtual void AddText(const wxString& txt) = 0;
 153
 154     // Adds tag and proceeds it. Parse() may (and usually is) called from this method.
 155     // This is called from Parse() and may be overriden.
 156     // Default behavior is that it looks for proper handler in m_Handlers. The tag is
 157     // ignored if no hander is found.
 158     // Derived class is *responsible* for filling in m_Handlers table.
 159     virtual void AddTag(const wxHtmlTag& tag);
 160
 161 protected:
 162     // DOM tree:
 163     wxHtmlTag *m_CurTag;
 164     wxHtmlTag *m_Tags;
 165     wxHtmlTextPieces *m_TextPieces;
 166     size_t m_CurTextPiece;
 167
 168     const wxString *m_Source;
 169
 170     wxHtmlParserState *m_SavedStates;
 171
 172     // handlers that handle particular tags. The table is accessed by
 173     // key = tag's name.
 174     // This attribute MUST be filled by derived class otherwise it would
 175     // be empty and no tags would be recognized
 176     // (see wxHtmlWinParser for details about filling it)
 177     // m_HandlersHash is for random access based on knowledge of tag name (BR, P, etc.)
 178     //      it may (and often does) contain more references to one object
 179     // m_HandlersList is list of all handlers and it is guaranteed to contain
 180     //      only one reference to each handler instance.
 181     wxList m_HandlersList;
 182     wxHashTable m_HandlersHash;
 183
 184     DECLARE_NO_COPY_CLASS(wxHtmlParser)
 185
 186     // class for opening files (file system)
 187     wxFileSystem *m_FS;
 188     // handlers stack used by PushTagHandler and PopTagHandler
 189     wxList *m_HandlersStack;
 190
 191     // entity parse
 192     wxHtmlEntitiesParser *m_entitiesParser;
 193
 194     // flag indicating that the parser should stop
 195     bool m_stopParsing;
 196 };
 197
 198
 199
 200 // This class (and derived classes) cooperates with wxHtmlParser.
 201 // Each recognized tag is passed to handler which is capable
 202 // of handling it. Each tag is handled in 3 steps:
 203 // 1. Handler will modifies state of parser
 204 //    (using its public methods)
 205 // 2. Parser parses source between starting and ending tag
 206 // 3. Handler restores original state of the parser
 207 class WXDLLIMPEXP_HTML wxHtmlTagHandler : public wxObject
 208 {
 209     DECLARE_ABSTRACT_CLASS(wxHtmlTagHandler)
 210
 211 public:
 212     wxHtmlTagHandler() : wxObject () { m_Parser = NULL; }
 213
 214     // Sets the parser.
 215     // NOTE : each _instance_ of handler is guaranteed to be called
 216     // only by one parser. This means you don't have to care about
 217     // reentrancy.
 218     virtual void SetParser(wxHtmlParser *parser)
 219         { m_Parser = parser; }
 220
 221     // Returns list of supported tags. The list is in uppercase and
 222     // tags are delimited by ','.
 223     // Example : "I,B,FONT,P"
 224     //   is capable of handling italic, bold, font and paragraph tags
 225     virtual wxString GetSupportedTags() = 0;
 226
 227     // This is hadling core method. It does all the Steps 1-3.
 228     // To process step 2, you can call ParseInner()
 229     // returned value : true if it called ParseInner(),
 230     //                  false etherwise
 231     virtual bool HandleTag(const wxHtmlTag& tag) = 0;
 232
 233 protected:
 234     // parses input between beginning and ending tag.
 235     // m_Parser must be set.
 236     void ParseInner(const wxHtmlTag& tag)
 237         { m_Parser->DoParsing(tag.GetBeginIter(), tag.GetEndIter1()); }
 238
 239     // Parses given source as if it was tag's inner code (see
 240     // wxHtmlParser::GetInnerSource).  Unlike ParseInner(), this method lets
 241     // you specify the source code to parse. This is useful when you need to
 242     // modify the inner text before parsing.
 243     void ParseInnerSource(const wxString& source);
 244
 245     wxHtmlParser *m_Parser;
 246
 247     DECLARE_NO_COPY_CLASS(wxHtmlTagHandler)
 248 };
 249
 250
 251 // This class is used to parse HTML entities in strings. It can handle
 252 // both named entities and &#xxxx entries where xxxx is Unicode code.
 253 class WXDLLIMPEXP_HTML wxHtmlEntitiesParser : public wxObject
 254 {
 255     DECLARE_DYNAMIC_CLASS(wxHtmlEntitiesParser)
 256
 257 public:
 258     wxHtmlEntitiesParser();
 259     virtual ~wxHtmlEntitiesParser();
 260
 261     // Sets encoding of output string.
 262     // Has no effect if wxUSE_WCHAR_T==0 or wxUSE_UNICODE==1
 263     void SetEncoding(wxFontEncoding encoding);
 264
 265     // Parses entities in input and replaces them with respective characters
 266     // (with respect to output encoding)
 267     wxString Parse(const wxString& input) const;
 268
 269     // Returns character for given entity or 0 if the enity is unknown
 270     wxChar GetEntityChar(const wxString& entity) const;
 271
 272     // Returns character that represents given Unicode code
 273 #if wxUSE_UNICODE
 274     wxChar GetCharForCode(unsigned code) const { return (wxChar)code; }
 275 #else
 276     wxChar GetCharForCode(unsigned code) const;
 277 #endif
 278
 279 protected:
 280 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 281     wxMBConv *m_conv;
 282     wxFontEncoding m_encoding;
 283 #endif
 284
 285     DECLARE_NO_COPY_CLASS(wxHtmlEntitiesParser)
 286 };
 287
 288
 289 #endif
 290
 291 #endif // _WX_HTMLPARS_H_