include/wx/html/htmlpars.h

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        wx/html/htmlpars.h
   3 // Purpose:     wxHtmlParser class (generic parser)
   4 // Author:      Vaclav Slavik
   5 // Copyright:   (c) 1999 Vaclav Slavik
   6 // Licence:     wxWindows licence
   7 /////////////////////////////////////////////////////////////////////////////
   8
   9 #ifndef _WX_HTMLPARS_H_
  10 #define _WX_HTMLPARS_H_
  11
  12 #include "wx/defs.h"
  13 #if wxUSE_HTML
  14
  15 #include "wx/html/htmltag.h"
  16 #include "wx/filesys.h"
  17 #include "wx/hashmap.h"
  18 #include "wx/hashset.h"
  19 #include "wx/vector.h"
  20 #include "wx/fontenc.h"
  21
  22 class WXDLLIMPEXP_FWD_BASE wxMBConv;
  23 class WXDLLIMPEXP_FWD_HTML wxHtmlParser;
  24 class WXDLLIMPEXP_FWD_HTML wxHtmlTagHandler;
  25 class WXDLLIMPEXP_FWD_HTML wxHtmlEntitiesParser;
  26
  27 class wxHtmlTextPieces;
  28 class wxHtmlParserState;
  29
  30 WX_DECLARE_HASH_SET_WITH_DECL_PTR(wxHtmlTagHandler*,
  31                                   wxPointerHash, wxPointerEqual,
  32                                   wxHtmlTagHandlersSet,
  33                                   class WXDLLIMPEXP_HTML);
  34 WX_DECLARE_STRING_HASH_MAP_WITH_DECL(wxHtmlTagHandler*,
  35                                      wxHtmlTagHandlersHash,
  36                                      class WXDLLIMPEXP_HTML);
  37
  38
  39 enum wxHtmlURLType
  40 {
  41     wxHTML_URL_PAGE,
  42     wxHTML_URL_IMAGE,
  43     wxHTML_URL_OTHER
  44 };
  45
  46 // This class handles generic parsing of HTML document : it scans
  47 // the document and divides it into blocks of tags (where one block
  48 // consists of starting and ending tag and of text between these
  49 // 2 tags.
  50 class WXDLLIMPEXP_HTML wxHtmlParser : public wxObject
  51 {
  52     DECLARE_ABSTRACT_CLASS(wxHtmlParser)
  53
  54 public:
  55     wxHtmlParser();
  56     virtual ~wxHtmlParser();
  57
  58     // Sets the class which will be used for opening files
  59     void SetFS(wxFileSystem *fs) { m_FS = fs; }
  60
  61     wxFileSystem* GetFS() const { return m_FS; }
  62
  63     // Opens file if the parser is allowed to open given URL (may be forbidden
  64     // for security reasons)
  65     virtual wxFSFile *OpenURL(wxHtmlURLType type, const wxString& url) const;
  66
  67     // You can simply call this method when you need parsed output.
  68     // This method does these things:
  69     // 1. call InitParser(source);
  70     // 2. call DoParsing();
  71     // 3. call GetProduct(); (its return value is then returned)
  72     // 4. call DoneParser();
  73     wxObject* Parse(const wxString& source);
  74
  75     // Sets the source. This must be called before running Parse() method.
  76     virtual void InitParser(const wxString& source);
  77     // This must be called after Parse().
  78     virtual void DoneParser();
  79
  80     // May be called during parsing to immediately return from Parse().
  81     virtual void StopParsing() { m_stopParsing = true; }
  82
  83     // Parses the m_Source from begin_pos to end_pos-1.
  84     // (in noparams version it parses whole m_Source)
  85     void DoParsing(const wxString::const_iterator& begin_pos,
  86                    const wxString::const_iterator& end_pos);
  87     void DoParsing();
  88
  89     // Returns pointer to the tag at parser's current position
  90     wxHtmlTag *GetCurrentTag() const { return m_CurTag; }
  91
  92     // Returns product of parsing
  93     // Returned value is result of parsing of the part. The type of this result
  94     // depends on internal representation in derived parser
  95     // (see wxHtmlWinParser for details).
  96     virtual wxObject* GetProduct() = 0;
  97
  98     // adds handler to the list & hash table of handlers.
  99     virtual void AddTagHandler(wxHtmlTagHandler *handler);
 100
 101     // Forces the handler to handle additional tags (not returned by GetSupportedTags).
 102     // The handler should already be in use by this parser.
 103     // Example: you want to parse following pseudo-html structure:
 104     //   <myitems>
 105     //     <it name="one" value="1">
 106     //     <it name="two" value="2">
 107     //   </myitems>
 108     //   <it> This last it has different meaning, we don't want it to be parsed by myitems handler!
 109     // handler can handle only 'myitems' (e.g. its GetSupportedTags returns "MYITEMS")
 110     // you can call PushTagHandler(handler, "IT") when you find <myitems>
 111     // and call PopTagHandler() when you find </myitems>
 112     void PushTagHandler(wxHtmlTagHandler *handler, const wxString& tags);
 113
 114     // Restores state before last call to PushTagHandler
 115     void PopTagHandler();
 116
 117     const wxString* GetSource() {return m_Source;}
 118     void SetSource(const wxString& src);
 119
 120     // Sets HTML source and remembers current parser's state so that it can
 121     // later be restored. This is useful for on-line modifications of
 122     // HTML source (for example, <pre> handler replaces spaces with &nbsp;
 123     // and newlines with <br>)
 124     virtual void SetSourceAndSaveState(const wxString& src);
 125     // Restores parser's state from stack or returns false if the stack is
 126     // empty
 127     virtual bool RestoreState();
 128
 129     // Returns HTML source inside the element (i.e. between the starting
 130     // and ending tag)
 131     wxString GetInnerSource(const wxHtmlTag& tag);
 132
 133     // Parses HTML string 'markup' and extracts charset info from <meta> tag
 134     // if present. Returns empty string if the tag is missing.
 135     // For wxHTML's internal use.
 136     static wxString ExtractCharsetInformation(const wxString& markup);
 137
 138     // Returns entity parser object, used to substitute HTML &entities;
 139     wxHtmlEntitiesParser *GetEntitiesParser() const { return m_entitiesParser; }
 140
 141     // Returns true if the tag starting at the given position is a comment tag
 142     //
 143     // p should point to '<' character and is modified to point to the closing
 144     // '>' of the end comment tag if this is indeed a comment
 145     static bool
 146     SkipCommentTag(wxString::const_iterator& p, wxString::const_iterator end);
 147
 148 protected:
 149     // DOM structure
 150     void CreateDOMTree();
 151     void DestroyDOMTree();
 152     void CreateDOMSubTree(wxHtmlTag *cur,
 153                           const wxString::const_iterator& begin_pos,
 154                           const wxString::const_iterator& end_pos,
 155                           wxHtmlTagsCache *cache);
 156
 157     // Adds text to the output.
 158     // This is called from Parse() and must be overridden in derived classes.
 159     // txt is not guaranteed to be only one word. It is largest continuous part
 160     // of text (= not broken by tags)
 161     virtual void AddText(const wxString& txt) = 0;
 162
 163     // Adds tag and proceeds it. Parse() may (and usually is) called from this method.
 164     // This is called from Parse() and may be overridden.
 165     // Default behaviour is that it looks for proper handler in m_Handlers. The tag is
 166     // ignored if no hander is found.
 167     // Derived class is *responsible* for filling in m_Handlers table.
 168     virtual void AddTag(const wxHtmlTag& tag);
 169
 170 protected:
 171     // DOM tree:
 172     wxHtmlTag *m_CurTag;
 173     wxHtmlTag *m_Tags;
 174     wxHtmlTextPieces *m_TextPieces;
 175     size_t m_CurTextPiece;
 176
 177     const wxString *m_Source;
 178
 179     wxHtmlParserState *m_SavedStates;
 180
 181     // handlers that handle particular tags. The table is accessed by
 182     // key = tag's name.
 183     // This attribute MUST be filled by derived class otherwise it would
 184     // be empty and no tags would be recognized
 185     // (see wxHtmlWinParser for details about filling it)
 186     // m_HandlersHash is for random access based on knowledge of tag name (BR, P, etc.)
 187     //      it may (and often does) contain more references to one object
 188     // m_HandlersList is list of all handlers and it is guaranteed to contain
 189     //      only one reference to each handler instance.
 190     wxHtmlTagHandlersSet m_HandlersSet;
 191     wxHtmlTagHandlersHash m_HandlersHash;
 192
 193     wxDECLARE_NO_COPY_CLASS(wxHtmlParser);
 194
 195     // class for opening files (file system)
 196     wxFileSystem *m_FS;
 197     // handlers stack used by PushTagHandler and PopTagHandler
 198     wxVector<wxHtmlTagHandlersHash*> m_HandlersStack;
 199
 200     // entity parse
 201     wxHtmlEntitiesParser *m_entitiesParser;
 202
 203     // flag indicating that the parser should stop
 204     bool m_stopParsing;
 205 };
 206
 207
 208
 209 // This class (and derived classes) cooperates with wxHtmlParser.
 210 // Each recognized tag is passed to handler which is capable
 211 // of handling it. Each tag is handled in 3 steps:
 212 // 1. Handler will modifies state of parser
 213 //    (using its public methods)
 214 // 2. Parser parses source between starting and ending tag
 215 // 3. Handler restores original state of the parser
 216 class WXDLLIMPEXP_HTML wxHtmlTagHandler : public wxObject
 217 {
 218     DECLARE_ABSTRACT_CLASS(wxHtmlTagHandler)
 219
 220 public:
 221     wxHtmlTagHandler() : wxObject () { m_Parser = NULL; }
 222
 223     // Sets the parser.
 224     // NOTE : each _instance_ of handler is guaranteed to be called
 225     // only by one parser. This means you don't have to care about
 226     // reentrancy.
 227     virtual void SetParser(wxHtmlParser *parser)
 228         { m_Parser = parser; }
 229
 230     // Get the parser associated with this tag handler.
 231     wxHtmlParser* GetParser() const { return m_Parser; }
 232
 233     // Returns list of supported tags. The list is in uppercase and
 234     // tags are delimited by ','.
 235     // Example : "I,B,FONT,P"
 236     //   is capable of handling italic, bold, font and paragraph tags
 237     virtual wxString GetSupportedTags() = 0;
 238
 239     // This is hadling core method. It does all the Steps 1-3.
 240     // To process step 2, you can call ParseInner()
 241     // returned value : true if it called ParseInner(),
 242     //                  false etherwise
 243     virtual bool HandleTag(const wxHtmlTag& tag) = 0;
 244
 245 protected:
 246     // parses input between beginning and ending tag.
 247     // m_Parser must be set.
 248     void ParseInner(const wxHtmlTag& tag)
 249         { m_Parser->DoParsing(tag.GetBeginIter(), tag.GetEndIter1()); }
 250
 251     // Parses given source as if it was tag's inner code (see
 252     // wxHtmlParser::GetInnerSource).  Unlike ParseInner(), this method lets
 253     // you specify the source code to parse. This is useful when you need to
 254     // modify the inner text before parsing.
 255     void ParseInnerSource(const wxString& source);
 256
 257     wxHtmlParser *m_Parser;
 258
 259     wxDECLARE_NO_COPY_CLASS(wxHtmlTagHandler);
 260 };
 261
 262
 263 // This class is used to parse HTML entities in strings. It can handle
 264 // both named entities and &#xxxx entries where xxxx is Unicode code.
 265 class WXDLLIMPEXP_HTML wxHtmlEntitiesParser : public wxObject
 266 {
 267     DECLARE_DYNAMIC_CLASS(wxHtmlEntitiesParser)
 268
 269 public:
 270     wxHtmlEntitiesParser();
 271     virtual ~wxHtmlEntitiesParser();
 272
 273     // Sets encoding of output string.
 274     // Has no effect if wxUSE_UNICODE==1
 275 #if wxUSE_UNICODE
 276     void SetEncoding(wxFontEncoding WXUNUSED(encoding)) {}
 277 #else
 278     void SetEncoding(wxFontEncoding encoding);
 279 #endif
 280
 281     // Parses entities in input and replaces them with respective characters
 282     // (with respect to output encoding)
 283     wxString Parse(const wxString& input) const;
 284
 285     // Returns character for given entity or 0 if the enity is unknown
 286     wxChar GetEntityChar(const wxString& entity) const;
 287
 288     // Returns character that represents given Unicode code
 289 #if wxUSE_UNICODE
 290     wxChar GetCharForCode(unsigned code) const { return (wxChar)code; }
 291 #else
 292     wxChar GetCharForCode(unsigned code) const;
 293 #endif
 294
 295 protected:
 296 #if !wxUSE_UNICODE
 297     wxMBConv *m_conv;
 298     wxFontEncoding m_encoding;
 299 #endif
 300
 301     wxDECLARE_NO_COPY_CLASS(wxHtmlEntitiesParser);
 302 };
 303
 304
 305 #endif
 306
 307 #endif // _WX_HTMLPARS_H_