src/html/htmlpars.cpp

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        htmlpars.cpp
   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 #ifdef __GNUG__
  12 #pragma implementation "htmlpars.h"
  13 #endif
  14
  15 #include "wx/wxprec.h"
  16
  17 #include "wx/defs.h"
  18 #if wxUSE_HTML && wxUSE_STREAMS
  19
  20 #ifdef __BORLANDC__
  21 #pragma hdrstop
  22 #endif
  23
  24 #ifndef WXPRECOMP
  25     #include "wx/log.h"
  26     #include "wx/intl.h"
  27 #endif
  28
  29 #include "wx/tokenzr.h"
  30 #include "wx/wfstream.h"
  31 #include "wx/url.h"
  32 #include "wx/fontmap.h"
  33 #include "wx/html/htmldefs.h"
  34 #include "wx/html/htmlpars.h"
  35 #include "wx/dynarray.h"
  36 #include "wx/arrimpl.cpp"
  37
  38
  39 // DLL options compatibility check:
  40 #include "wx/app.h"
  41 WX_CHECK_BUILD_OPTIONS("wxHTML")
  42
  43 //-----------------------------------------------------------------------------
  44 // wxHtmlParser helpers
  45 //-----------------------------------------------------------------------------
  46
  47 class wxHtmlTextPiece
  48 {
  49 public:
  50     wxHtmlTextPiece(int pos, int lng) : m_pos(pos), m_lng(lng) {}
  51     int m_pos, m_lng;
  52 };
  53
  54 WX_DECLARE_OBJARRAY(wxHtmlTextPiece, wxHtmlTextPieces);
  55 WX_DEFINE_OBJARRAY(wxHtmlTextPieces);
  56
  57 class wxHtmlParserState
  58 {
  59 public:
  60     wxHtmlTag         *m_curTag;
  61     wxHtmlTag         *m_tags;
  62     wxHtmlTextPieces  *m_textPieces;
  63     int                m_curTextPiece;
  64     wxString           m_source;
  65     wxHtmlParserState *m_nextState;
  66 };
  67
  68 //-----------------------------------------------------------------------------
  69 // wxHtmlParser
  70 //-----------------------------------------------------------------------------
  71
  72 IMPLEMENT_ABSTRACT_CLASS(wxHtmlParser,wxObject)
  73
  74 wxHtmlParser::wxHtmlParser()
  75     : wxObject(), m_HandlersHash(wxKEY_STRING),
  76       m_FS(NULL), m_HandlersStack(NULL)
  77 {
  78     m_entitiesParser = new wxHtmlEntitiesParser;
  79     m_Tags = NULL;
  80     m_CurTag = NULL;
  81     m_TextPieces = NULL;
  82     m_CurTextPiece = 0;
  83     m_SavedStates = NULL;
  84 }
  85
  86 wxHtmlParser::~wxHtmlParser()
  87 {
  88     while (RestoreState()) {}
  89     DestroyDOMTree();
  90
  91     if (m_HandlersStack)
  92     {
  93         wxList& tmp = *m_HandlersStack;
  94         wxList::iterator it, en;
  95         for( it = tmp.begin(), en = tmp.end(); it != en; ++it )
  96             delete (wxHashTable*)*it;
  97         tmp.clear();
  98     }
  99     delete m_HandlersStack;
 100     m_HandlersHash.Clear();
 101     WX_CLEAR_LIST(wxList, m_HandlersList);
 102     delete m_entitiesParser;
 103 }
 104
 105 wxObject* wxHtmlParser::Parse(const wxString& source)
 106 {
 107     InitParser(source);
 108     DoParsing();
 109     wxObject *result = GetProduct();
 110     DoneParser();
 111     return result;
 112 }
 113
 114 void wxHtmlParser::InitParser(const wxString& source)
 115 {
 116     SetSource(source);
 117     m_stopParsing = FALSE;
 118 }
 119
 120 void wxHtmlParser::DoneParser()
 121 {
 122     DestroyDOMTree();
 123 }
 124
 125 void wxHtmlParser::SetSource(const wxString& src)
 126 {
 127     DestroyDOMTree();
 128     m_Source = src;
 129     CreateDOMTree();
 130     m_CurTag = NULL;
 131     m_CurTextPiece = 0;
 132 }
 133
 134 void wxHtmlParser::CreateDOMTree()
 135 {
 136     wxHtmlTagsCache cache(m_Source);
 137     m_TextPieces = new wxHtmlTextPieces;
 138     CreateDOMSubTree(NULL, 0, m_Source.Length(), &cache);
 139     m_CurTextPiece = 0;
 140 }
 141
 142 extern bool wxIsCDATAElement(const wxChar *tag);
 143
 144 void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
 145                                     int begin_pos, int end_pos,
 146                                     wxHtmlTagsCache *cache)
 147 {
 148     if (end_pos <= begin_pos) return;
 149
 150     wxChar c;
 151     int i = begin_pos;
 152     int textBeginning = begin_pos;
 153
 154     // If the tag contains CDATA text, we include the text between beginning
 155     // and ending tag verbosely. Setting i=end_pos will skip to the very
 156     // end of this function where text piece is added, bypassing any child
 157     // tags parsing (CDATA element can't have child elements by definition):
 158     if (cur != NULL && wxIsCDATAElement(cur->GetName().c_str()))
 159     {
 160         i = end_pos;
 161     }
 162
 163     while (i < end_pos)
 164     {
 165         c = m_Source.GetChar(i);
 166
 167         if (c == wxT('<'))
 168         {
 169             // add text to m_TextPieces:
 170             if (i - textBeginning > 0)
 171                 m_TextPieces->Add(
 172                     wxHtmlTextPiece(textBeginning, i - textBeginning));
 173
 174             // if it is a comment, skip it:
 175             if (i < end_pos-6 && m_Source.GetChar(i+1) == wxT('!') &&
 176                                  m_Source.GetChar(i+2) == wxT('-') &&
 177                                  m_Source.GetChar(i+3) == wxT('-'))
 178             {
 179                 // Comments begin with "<!--" and end with "--[ \t\r\n]*>"
 180                 // according to HTML 4.0
 181                 int dashes = 0;
 182                 i += 4;
 183                 while (i < end_pos)
 184                 {
 185                     c = m_Source.GetChar(i++);
 186                     if ((c == wxT(' ') || c == wxT('\n') ||
 187                         c == wxT('\r') || c == wxT('\t')) && dashes >= 2) {}
 188                     else if (c == wxT('>') && dashes >= 2)
 189                     {
 190                         textBeginning = i;
 191                         break;
 192                     }
 193                     else if (c == wxT('-'))
 194                         dashes++;
 195                     else
 196                         dashes = 0;
 197                 }
 198             }
 199
 200             // add another tag to the tree:
 201             else if (i < end_pos-1 && m_Source.GetChar(i+1) != wxT('/'))
 202                 {
 203                 wxHtmlTag *chd;
 204                 if (cur)
 205                     chd = new wxHtmlTag(cur, m_Source,
 206                                         i, end_pos, cache, m_entitiesParser);
 207                 else
 208                 {
 209                     chd = new wxHtmlTag(NULL, m_Source,
 210                                         i, end_pos, cache, m_entitiesParser);
 211                     if (!m_Tags)
 212                     {
 213                         // if this is the first tag to be created make the root
 214                         // m_Tags point to it:
 215                         m_Tags = chd;
 216                     }
 217                     else
 218                     {
 219                         // if there is already a root tag add this tag as
 220                         // the last sibling:
 221                         chd->m_Prev = m_Tags->GetLastSibling();
 222                         chd->m_Prev->m_Next = chd;
 223                     }
 224                 }
 225
 226                 if (chd->HasEnding())
 227                 {
 228                     CreateDOMSubTree(chd,
 229                                      chd->GetBeginPos(), chd->GetEndPos1(),
 230                                      cache);
 231                     i = chd->GetEndPos2();
 232                 }
 233                 else
 234                     i = chd->GetBeginPos();
 235
 236                 textBeginning = i;
 237             }
 238
 239             // ... or skip ending tag:
 240             else
 241             {
 242                 while (i < end_pos && m_Source.GetChar(i) != wxT('>')) i++;
 243                 textBeginning = i+1;
 244             }
 245         }
 246         else i++;
 247     }
 248
 249     // add remaining text to m_TextPieces:
 250     if (end_pos - textBeginning > 0)
 251         m_TextPieces->Add(
 252             wxHtmlTextPiece(textBeginning, end_pos - textBeginning));
 253 }
 254
 255 void wxHtmlParser::DestroyDOMTree()
 256 {
 257     wxHtmlTag *t1, *t2;
 258     t1 = m_Tags;
 259     while (t1)
 260     {
 261         t2 = t1->GetNextSibling();
 262         delete t1;
 263         t1 = t2;
 264     }
 265     m_Tags = m_CurTag = NULL;
 266
 267     delete m_TextPieces;
 268     m_TextPieces = NULL;
 269 }
 270
 271 void wxHtmlParser::DoParsing()
 272 {
 273     m_CurTag = m_Tags;
 274     m_CurTextPiece = 0;
 275     DoParsing(0, m_Source.Length());
 276 }
 277
 278 void wxHtmlParser::DoParsing(int begin_pos, int end_pos)
 279 {
 280     if (end_pos <= begin_pos) return;
 281
 282     wxHtmlTextPieces& pieces = *m_TextPieces;
 283     size_t piecesCnt = pieces.GetCount();
 284
 285     while (begin_pos < end_pos)
 286     {
 287         while (m_CurTag && m_CurTag->GetBeginPos() < begin_pos)
 288             m_CurTag = m_CurTag->GetNextTag();
 289         while (m_CurTextPiece < piecesCnt &&
 290                pieces[m_CurTextPiece].m_pos < begin_pos)
 291             m_CurTextPiece++;
 292
 293         if (m_CurTextPiece < piecesCnt &&
 294             (!m_CurTag ||
 295              pieces[m_CurTextPiece].m_pos < m_CurTag->GetBeginPos()))
 296         {
 297             // Add text:
 298             AddText(GetEntitiesParser()->Parse(
 299                        m_Source.Mid(pieces[m_CurTextPiece].m_pos,
 300                                     pieces[m_CurTextPiece].m_lng)));
 301             begin_pos = pieces[m_CurTextPiece].m_pos +
 302                         pieces[m_CurTextPiece].m_lng;
 303             m_CurTextPiece++;
 304         }
 305         else if (m_CurTag)
 306         {
 307             // Add tag:
 308             if (m_CurTag)
 309             {
 310                 if (m_CurTag->HasEnding())
 311                     begin_pos = m_CurTag->GetEndPos2();
 312                 else
 313                     begin_pos = m_CurTag->GetBeginPos();
 314             }
 315             wxHtmlTag *t = m_CurTag;
 316             m_CurTag = m_CurTag->GetNextTag();
 317             AddTag(*t);
 318             if (m_stopParsing)
 319                 return;
 320         }
 321         else break;
 322     }
 323 }
 324
 325 void wxHtmlParser::AddTag(const wxHtmlTag& tag)
 326 {
 327     wxHtmlTagHandler *h;
 328     bool inner = FALSE;
 329
 330     h = (wxHtmlTagHandler*) m_HandlersHash.Get(tag.GetName());
 331     if (h)
 332     {
 333         inner = h->HandleTag(tag);
 334         if (m_stopParsing)
 335             return;
 336     }
 337     if (!inner)
 338     {
 339         if (tag.HasEnding())
 340             DoParsing(tag.GetBeginPos(), tag.GetEndPos1());
 341     }
 342 }
 343
 344 void wxHtmlParser::AddTagHandler(wxHtmlTagHandler *handler)
 345 {
 346     wxString s(handler->GetSupportedTags());
 347     wxStringTokenizer tokenizer(s, wxT(", "));
 348
 349     while (tokenizer.HasMoreTokens())
 350         m_HandlersHash.Put(tokenizer.GetNextToken(), handler);
 351
 352     if (m_HandlersList.IndexOf(handler) == wxNOT_FOUND)
 353         m_HandlersList.Append(handler);
 354
 355     handler->SetParser(this);
 356 }
 357
 358 void wxHtmlParser::PushTagHandler(wxHtmlTagHandler *handler, wxString tags)
 359 {
 360     wxStringTokenizer tokenizer(tags, wxT(", "));
 361     wxString key;
 362
 363     if (m_HandlersStack == NULL)
 364     {
 365         m_HandlersStack = new wxList;
 366     }
 367
 368     m_HandlersStack->Insert((wxObject*)new wxHashTable(m_HandlersHash));
 369
 370     while (tokenizer.HasMoreTokens())
 371     {
 372         key = tokenizer.GetNextToken();
 373         m_HandlersHash.Delete(key);
 374         m_HandlersHash.Put(key, handler);
 375     }
 376 }
 377
 378 void wxHtmlParser::PopTagHandler()
 379 {
 380     wxList::compatibility_iterator first;
 381
 382     if ( !m_HandlersStack ||
 383 #if wxUSE_STL
 384          !(first = m_HandlersStack->GetFirst())
 385 #else // !wxUSE_STL
 386          ((first = m_HandlersStack->GetFirst()) == NULL)
 387 #endif // wxUSE_STL/!wxUSE_STL
 388         )
 389     {
 390         wxLogWarning(_("Warning: attempt to remove HTML tag handler from empty stack."));
 391         return;
 392     }
 393     m_HandlersHash = *((wxHashTable*) first->GetData());
 394     delete (wxHashTable*) first->GetData();
 395     m_HandlersStack->Erase(first);
 396 }
 397
 398 void wxHtmlParser::SetSourceAndSaveState(const wxString& src)
 399 {
 400     wxHtmlParserState *s = new wxHtmlParserState;
 401
 402     s->m_curTag = m_CurTag;
 403     s->m_tags = m_Tags;
 404     s->m_textPieces = m_TextPieces;
 405     s->m_curTextPiece = m_CurTextPiece;
 406     s->m_source = m_Source;
 407
 408     s->m_nextState = m_SavedStates;
 409     m_SavedStates = s;
 410
 411     m_CurTag = NULL;
 412     m_Tags = NULL;
 413     m_TextPieces = NULL;
 414     m_CurTextPiece = 0;
 415     m_Source = wxEmptyString;
 416
 417     SetSource(src);
 418 }
 419
 420 bool wxHtmlParser::RestoreState()
 421 {
 422     if (!m_SavedStates) return FALSE;
 423
 424     DestroyDOMTree();
 425
 426     wxHtmlParserState *s = m_SavedStates;
 427     m_SavedStates = s->m_nextState;
 428
 429     m_CurTag = s->m_curTag;
 430     m_Tags = s->m_tags;
 431     m_TextPieces = s->m_textPieces;
 432     m_CurTextPiece = s->m_curTextPiece;
 433     m_Source = s->m_source;
 434
 435     delete s;
 436     return TRUE;
 437 }
 438
 439 //-----------------------------------------------------------------------------
 440 // wxHtmlTagHandler
 441 //-----------------------------------------------------------------------------
 442
 443 IMPLEMENT_ABSTRACT_CLASS(wxHtmlTagHandler,wxObject)
 444
 445
 446 //-----------------------------------------------------------------------------
 447 // wxHtmlEntitiesParser
 448 //-----------------------------------------------------------------------------
 449
 450 IMPLEMENT_DYNAMIC_CLASS(wxHtmlEntitiesParser,wxObject)
 451
 452 wxHtmlEntitiesParser::wxHtmlEntitiesParser()
 453 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 454     : m_conv(NULL), m_encoding(wxFONTENCODING_SYSTEM)
 455 #endif
 456 {
 457 }
 458
 459 wxHtmlEntitiesParser::~wxHtmlEntitiesParser()
 460 {
 461 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 462     delete m_conv;
 463 #endif
 464 }
 465
 466 void wxHtmlEntitiesParser::SetEncoding(wxFontEncoding encoding)
 467 {
 468 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 469     if (encoding == m_encoding)
 470         return;
 471
 472     delete m_conv;
 473
 474     m_encoding = encoding;
 475     if (m_encoding == wxFONTENCODING_SYSTEM)
 476         m_conv = NULL;
 477     else
 478         m_conv = new wxCSConv(wxFontMapper::GetEncodingName(m_encoding));
 479 #else
 480     (void) encoding;
 481 #endif
 482 }
 483
 484 wxString wxHtmlEntitiesParser::Parse(const wxString& input)
 485 {
 486     const wxChar *c, *last;
 487     const wxChar *in_str = input.c_str();
 488     wxString output;
 489
 490     output.reserve(input.length());
 491
 492     for (c = in_str, last = in_str; *c != wxT('\0'); c++)
 493     {
 494         if (*c == wxT('&'))
 495         {
 496             if (c - last > 0)
 497                 output.append(last, c - last);
 498             if (++c == wxT('\0')) break;
 499
 500             wxString entity;
 501             const wxChar *ent_s = c;
 502             wxChar entity_char;
 503
 504             for (; (*c >= wxT('a') && *c <= wxT('z')) ||
 505                    (*c >= wxT('A') && *c <= wxT('Z')) ||
 506                    (*c >= wxT('0') && *c <= wxT('9')) ||
 507                    *c == wxT('_') || *c == wxT('#'); c++) {}
 508             entity.append(ent_s, c - ent_s);
 509             if (*c != wxT(';')) c--;
 510             last = c+1;
 511             entity_char = GetEntityChar(entity);
 512             if (entity_char)
 513                 output << entity_char;
 514             else
 515             {
 516                 output.append(ent_s-1, c-ent_s+2);
 517                 wxLogDebug(wxT("Unrecognized HTML entity: '%s'"), entity.c_str());
 518             }
 519         }
 520     }
 521     if (*last != wxT('\0'))
 522         output.append(last);
 523     return output;
 524 }
 525
 526 struct wxHtmlEntityInfo
 527 {
 528     const wxChar *name;
 529     unsigned code;
 530 };
 531
 532 extern "C" int LINKAGEMODE wxHtmlEntityCompare(const void *key, const void *item)
 533 {
 534     return wxStrcmp((wxChar*)key, ((wxHtmlEntityInfo*)item)->name);
 535 }
 536
 537 #if !wxUSE_UNICODE
 538 wxChar wxHtmlEntitiesParser::GetCharForCode(unsigned code)
 539 {
 540 #if wxUSE_WCHAR_T
 541     char buf[2];
 542     wchar_t wbuf[2];
 543     wbuf[0] = (wchar_t)code;
 544     wbuf[1] = 0;
 545     wxMBConv *conv = m_conv ? m_conv : &wxConvLocal;
 546     if (conv->WC2MB(buf, wbuf, 2) == (size_t)-1)
 547         return '?';
 548     return buf[0];
 549 #else
 550     return (code < 256) ? (wxChar)code : '?';
 551 #endif
 552 }
 553 #endif
 554
 555 wxChar wxHtmlEntitiesParser::GetEntityChar(const wxString& entity)
 556 {
 557     unsigned code = 0;
 558
 559     if (entity[0] == wxT('#'))
 560     {
 561         const wxChar *ent_s = entity.c_str();
 562         const wxChar *format;
 563
 564         if (ent_s[1] == wxT('x') || ent_s[1] == wxT('X'))
 565         {
 566             format = wxT("%x");
 567             ent_s++;
 568         }
 569         else
 570             format = wxT("%u");
 571         ent_s++;
 572
 573         if (wxSscanf(ent_s, format, &code) != 1)
 574             code = 0;
 575     }
 576     else
 577     {
 578         static wxHtmlEntityInfo substitutions[] = {
 579             { wxT("AElig"),198 },
 580             { wxT("Aacute"),193 },
 581             { wxT("Acirc"),194 },
 582             { wxT("Agrave"),192 },
 583             { wxT("Alpha"),913 },
 584             { wxT("Aring"),197 },
 585             { wxT("Atilde"),195 },
 586             { wxT("Auml"),196 },
 587             { wxT("Beta"),914 },
 588             { wxT("Ccedil"),199 },
 589             { wxT("Chi"),935 },
 590             { wxT("Dagger"),8225 },
 591             { wxT("Delta"),916 },
 592             { wxT("ETH"),208 },
 593             { wxT("Eacute"),201 },
 594             { wxT("Ecirc"),202 },
 595             { wxT("Egrave"),200 },
 596             { wxT("Epsilon"),917 },
 597             { wxT("Eta"),919 },
 598             { wxT("Euml"),203 },
 599             { wxT("Gamma"),915 },
 600             { wxT("Iacute"),205 },
 601             { wxT("Icirc"),206 },
 602             { wxT("Igrave"),204 },
 603             { wxT("Iota"),921 },
 604             { wxT("Iuml"),207 },
 605             { wxT("Kappa"),922 },
 606             { wxT("Lambda"),923 },
 607             { wxT("Mu"),924 },
 608             { wxT("Ntilde"),209 },
 609             { wxT("Nu"),925 },
 610             { wxT("OElig"),338 },
 611             { wxT("Oacute"),211 },
 612             { wxT("Ocirc"),212 },
 613             { wxT("Ograve"),210 },
 614             { wxT("Omega"),937 },
 615             { wxT("Omicron"),927 },
 616             { wxT("Oslash"),216 },
 617             { wxT("Otilde"),213 },
 618             { wxT("Ouml"),214 },
 619             { wxT("Phi"),934 },
 620             { wxT("Pi"),928 },
 621             { wxT("Prime"),8243 },
 622             { wxT("Psi"),936 },
 623             { wxT("Rho"),929 },
 624             { wxT("Scaron"),352 },
 625             { wxT("Sigma"),931 },
 626             { wxT("THORN"),222 },
 627             { wxT("Tau"),932 },
 628             { wxT("Theta"),920 },
 629             { wxT("Uacute"),218 },
 630             { wxT("Ucirc"),219 },
 631             { wxT("Ugrave"),217 },
 632             { wxT("Upsilon"),933 },
 633             { wxT("Uuml"),220 },
 634             { wxT("Xi"),926 },
 635             { wxT("Yacute"),221 },
 636             { wxT("Yuml"),376 },
 637             { wxT("Zeta"),918 },
 638             { wxT("aacute"),225 },
 639             { wxT("acirc"),226 },
 640             { wxT("acute"),180 },
 641             { wxT("aelig"),230 },
 642             { wxT("agrave"),224 },
 643             { wxT("alefsym"),8501 },
 644             { wxT("alpha"),945 },
 645             { wxT("amp"),38 },
 646             { wxT("and"),8743 },
 647             { wxT("ang"),8736 },
 648             { wxT("aring"),229 },
 649             { wxT("asymp"),8776 },
 650             { wxT("atilde"),227 },
 651             { wxT("auml"),228 },
 652             { wxT("bdquo"),8222 },
 653             { wxT("beta"),946 },
 654             { wxT("brvbar"),166 },
 655             { wxT("bull"),8226 },
 656             { wxT("cap"),8745 },
 657             { wxT("ccedil"),231 },
 658             { wxT("cedil"),184 },
 659             { wxT("cent"),162 },
 660             { wxT("chi"),967 },
 661             { wxT("circ"),710 },
 662             { wxT("clubs"),9827 },
 663             { wxT("cong"),8773 },
 664             { wxT("copy"),169 },
 665             { wxT("crarr"),8629 },
 666             { wxT("cup"),8746 },
 667             { wxT("curren"),164 },
 668             { wxT("dArr"),8659 },
 669             { wxT("dagger"),8224 },
 670             { wxT("darr"),8595 },
 671             { wxT("deg"),176 },
 672             { wxT("delta"),948 },
 673             { wxT("diams"),9830 },
 674             { wxT("divide"),247 },
 675             { wxT("eacute"),233 },
 676             { wxT("ecirc"),234 },
 677             { wxT("egrave"),232 },
 678             { wxT("empty"),8709 },
 679             { wxT("emsp"),8195 },
 680             { wxT("ensp"),8194 },
 681             { wxT("epsilon"),949 },
 682             { wxT("equiv"),8801 },
 683             { wxT("eta"),951 },
 684             { wxT("eth"),240 },
 685             { wxT("euml"),235 },
 686             { wxT("euro"),8364 },
 687             { wxT("exist"),8707 },
 688             { wxT("fnof"),402 },
 689             { wxT("forall"),8704 },
 690             { wxT("frac12"),189 },
 691             { wxT("frac14"),188 },
 692             { wxT("frac34"),190 },
 693             { wxT("frasl"),8260 },
 694             { wxT("gamma"),947 },
 695             { wxT("ge"),8805 },
 696             { wxT("gt"),62 },
 697             { wxT("hArr"),8660 },
 698             { wxT("harr"),8596 },
 699             { wxT("hearts"),9829 },
 700             { wxT("hellip"),8230 },
 701             { wxT("iacute"),237 },
 702             { wxT("icirc"),238 },
 703             { wxT("iexcl"),161 },
 704             { wxT("igrave"),236 },
 705             { wxT("image"),8465 },
 706             { wxT("infin"),8734 },
 707             { wxT("int"),8747 },
 708             { wxT("iota"),953 },
 709             { wxT("iquest"),191 },
 710             { wxT("isin"),8712 },
 711             { wxT("iuml"),239 },
 712             { wxT("kappa"),954 },
 713             { wxT("lArr"),8656 },
 714             { wxT("lambda"),955 },
 715             { wxT("lang"),9001 },
 716             { wxT("laquo"),171 },
 717             { wxT("larr"),8592 },
 718             { wxT("lceil"),8968 },
 719             { wxT("ldquo"),8220 },
 720             { wxT("le"),8804 },
 721             { wxT("lfloor"),8970 },
 722             { wxT("lowast"),8727 },
 723             { wxT("loz"),9674 },
 724             { wxT("lrm"),8206 },
 725             { wxT("lsaquo"),8249 },
 726             { wxT("lsquo"),8216 },
 727             { wxT("lt"),60 },
 728             { wxT("macr"),175 },
 729             { wxT("mdash"),8212 },
 730             { wxT("micro"),181 },
 731             { wxT("middot"),183 },
 732             { wxT("minus"),8722 },
 733             { wxT("mu"),956 },
 734             { wxT("nabla"),8711 },
 735             { wxT("nbsp"),160 },
 736             { wxT("ndash"),8211 },
 737             { wxT("ne"),8800 },
 738             { wxT("ni"),8715 },
 739             { wxT("not"),172 },
 740             { wxT("notin"),8713 },
 741             { wxT("nsub"),8836 },
 742             { wxT("ntilde"),241 },
 743             { wxT("nu"),957 },
 744             { wxT("oacute"),243 },
 745             { wxT("ocirc"),244 },
 746             { wxT("oelig"),339 },
 747             { wxT("ograve"),242 },
 748             { wxT("oline"),8254 },
 749             { wxT("omega"),969 },
 750             { wxT("omicron"),959 },
 751             { wxT("oplus"),8853 },
 752             { wxT("or"),8744 },
 753             { wxT("ordf"),170 },
 754             { wxT("ordm"),186 },
 755             { wxT("oslash"),248 },
 756             { wxT("otilde"),245 },
 757             { wxT("otimes"),8855 },
 758             { wxT("ouml"),246 },
 759             { wxT("para"),182 },
 760             { wxT("part"),8706 },
 761             { wxT("permil"),8240 },
 762             { wxT("perp"),8869 },
 763             { wxT("phi"),966 },
 764             { wxT("pi"),960 },
 765             { wxT("piv"),982 },
 766             { wxT("plusmn"),177 },
 767             { wxT("pound"),163 },
 768             { wxT("prime"),8242 },
 769             { wxT("prod"),8719 },
 770             { wxT("prop"),8733 },
 771             { wxT("psi"),968 },
 772             { wxT("quot"),34 },
 773             { wxT("rArr"),8658 },
 774             { wxT("radic"),8730 },
 775             { wxT("rang"),9002 },
 776             { wxT("raquo"),187 },
 777             { wxT("rarr"),8594 },
 778             { wxT("rceil"),8969 },
 779             { wxT("rdquo"),8221 },
 780             { wxT("real"),8476 },
 781             { wxT("reg"),174 },
 782             { wxT("rfloor"),8971 },
 783             { wxT("rho"),961 },
 784             { wxT("rlm"),8207 },
 785             { wxT("rsaquo"),8250 },
 786             { wxT("rsquo"),8217 },
 787             { wxT("sbquo"),8218 },
 788             { wxT("scaron"),353 },
 789             { wxT("sdot"),8901 },
 790             { wxT("sect"),167 },
 791             { wxT("shy"),173 },
 792             { wxT("sigma"),963 },
 793             { wxT("sigmaf"),962 },
 794             { wxT("sim"),8764 },
 795             { wxT("spades"),9824 },
 796             { wxT("sub"),8834 },
 797             { wxT("sube"),8838 },
 798             { wxT("sum"),8721 },
 799             { wxT("sup"),8835 },
 800             { wxT("sup1"),185 },
 801             { wxT("sup2"),178 },
 802             { wxT("sup3"),179 },
 803             { wxT("supe"),8839 },
 804             { wxT("szlig"),223 },
 805             { wxT("tau"),964 },
 806             { wxT("there4"),8756 },
 807             { wxT("theta"),952 },
 808             { wxT("thetasym"),977 },
 809             { wxT("thinsp"),8201 },
 810             { wxT("thorn"),254 },
 811             { wxT("tilde"),732 },
 812             { wxT("times"),215 },
 813             { wxT("trade"),8482 },
 814             { wxT("uArr"),8657 },
 815             { wxT("uacute"),250 },
 816             { wxT("uarr"),8593 },
 817             { wxT("ucirc"),251 },
 818             { wxT("ugrave"),249 },
 819             { wxT("uml"),168 },
 820             { wxT("upsih"),978 },
 821             { wxT("upsilon"),965 },
 822             { wxT("uuml"),252 },
 823             { wxT("weierp"),8472 },
 824             { wxT("xi"),958 },
 825             { wxT("yacute"),253 },
 826             { wxT("yen"),165 },
 827             { wxT("yuml"),255 },
 828             { wxT("zeta"),950 },
 829             { wxT("zwj"),8205 },
 830             { wxT("zwnj"),8204 },
 831             {NULL, 0}};
 832         static size_t substitutions_cnt = 0;
 833
 834         if (substitutions_cnt == 0)
 835             while (substitutions[substitutions_cnt].code != 0)
 836                 substitutions_cnt++;
 837
 838         wxHtmlEntityInfo *info;
 839         info = (wxHtmlEntityInfo*) bsearch(entity.c_str(), substitutions,
 840                                            substitutions_cnt,
 841                                            sizeof(wxHtmlEntityInfo),
 842                                            wxHtmlEntityCompare);
 843         if (info)
 844             code = info->code;
 845     }
 846
 847     if (code == 0)
 848         return 0;
 849     else
 850         return GetCharForCode(code);
 851 }
 852
 853 wxFSFile *wxHtmlParser::OpenURL(wxHtmlURLType WXUNUSED(type),
 854                                 const wxString& url) const
 855 {
 856     return GetFS()->OpenFile(url);
 857 }
 858
 859
 860 //-----------------------------------------------------------------------------
 861 // wxHtmlParser::ExtractCharsetInformation
 862 //-----------------------------------------------------------------------------
 863
 864 class wxMetaTagParser : public wxHtmlParser
 865 {
 866 public:
 867     wxMetaTagParser() { }
 868
 869     wxObject* GetProduct() { return NULL; }
 870
 871 protected:
 872     virtual void AddText(const wxChar* WXUNUSED(txt)) {}
 873
 874     DECLARE_NO_COPY_CLASS(wxMetaTagParser)
 875 };
 876
 877 class wxMetaTagHandler : public wxHtmlTagHandler
 878 {
 879 public:
 880     wxMetaTagHandler(wxString *retval) : wxHtmlTagHandler(), m_retval(retval) {}
 881     wxString GetSupportedTags() { return wxT("META,BODY"); }
 882     bool HandleTag(const wxHtmlTag& tag);
 883
 884 private:
 885     wxString *m_retval;
 886
 887     DECLARE_NO_COPY_CLASS(wxMetaTagHandler)
 888 };
 889
 890 bool wxMetaTagHandler::HandleTag(const wxHtmlTag& tag)
 891 {
 892     if (tag.GetName() == _T("BODY"))
 893     {
 894         m_Parser->StopParsing();
 895         return FALSE;
 896     }
 897
 898     if (tag.HasParam(_T("HTTP-EQUIV")) &&
 899         tag.GetParam(_T("HTTP-EQUIV")).IsSameAs(_T("Content-Type"), false) &&
 900         tag.HasParam(_T("CONTENT")))
 901     {
 902         wxString content = tag.GetParam(_T("CONTENT"));
 903         if (content.Left(19) == _T("text/html; charset="))
 904         {
 905             *m_retval = content.Mid(19);
 906             m_Parser->StopParsing();
 907         }
 908     }
 909     return FALSE;
 910 }
 911
 912
 913 /*static*/
 914 wxString wxHtmlParser::ExtractCharsetInformation(const wxString& markup)
 915 {
 916     wxString charset;
 917     wxMetaTagParser parser;
 918     parser.AddTagHandler(new wxMetaTagHandler(&charset));
 919     parser.Parse(markup);
 920     return charset;
 921 }
 922
 923 #endif