src/html/htmlpars.cpp

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        src/html/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 #include "wx/wxprec.h"
  11
  12 #ifdef __BORLANDC__
  13     #pragma hdrstop
  14 #endif
  15
  16 #if wxUSE_HTML && wxUSE_STREAMS
  17
  18 #ifndef WXPRECOMP
  19     #include "wx/dynarray.h"
  20     #include "wx/log.h"
  21     #include "wx/intl.h"
  22     #include "wx/app.h"
  23 #endif
  24
  25 #include "wx/tokenzr.h"
  26 #include "wx/wfstream.h"
  27 #include "wx/url.h"
  28 #include "wx/fontmap.h"
  29 #include "wx/html/htmldefs.h"
  30 #include "wx/html/htmlpars.h"
  31 #include "wx/arrimpl.cpp"
  32
  33 #ifdef __WXWINCE__
  34     #include "wx/msw/wince/missing.h"       // for bsearch()
  35 #endif
  36
  37 // DLL options compatibility check:
  38 WX_CHECK_BUILD_OPTIONS("wxHTML")
  39
  40 const wxChar *wxTRACE_HTML_DEBUG = _T("htmldebug");
  41
  42 //-----------------------------------------------------------------------------
  43 // wxHtmlParser helpers
  44 //-----------------------------------------------------------------------------
  45
  46 class wxHtmlTextPiece
  47 {
  48 public:
  49     wxHtmlTextPiece(int pos, int lng) : m_pos(pos), m_lng(lng) {}
  50     int m_pos, m_lng;
  51 };
  52
  53 WX_DECLARE_OBJARRAY(wxHtmlTextPiece, wxHtmlTextPieces);
  54 WX_DEFINE_OBJARRAY(wxHtmlTextPieces)
  55
  56 class wxHtmlParserState
  57 {
  58 public:
  59     wxHtmlTag         *m_curTag;
  60     wxHtmlTag         *m_tags;
  61     wxHtmlTextPieces  *m_textPieces;
  62     int                m_curTextPiece;
  63     wxString           m_source;
  64     wxHtmlParserState *m_nextState;
  65 };
  66
  67 //-----------------------------------------------------------------------------
  68 // wxHtmlParser
  69 //-----------------------------------------------------------------------------
  70
  71 IMPLEMENT_ABSTRACT_CLASS(wxHtmlParser,wxObject)
  72
  73 wxHtmlParser::wxHtmlParser()
  74     : wxObject(), m_HandlersHash(wxKEY_STRING),
  75       m_FS(NULL), m_HandlersStack(NULL)
  76 {
  77     m_entitiesParser = new wxHtmlEntitiesParser;
  78     m_Tags = NULL;
  79     m_CurTag = NULL;
  80     m_TextPieces = NULL;
  81     m_CurTextPiece = 0;
  82     m_SavedStates = NULL;
  83 }
  84
  85 wxHtmlParser::~wxHtmlParser()
  86 {
  87     while (RestoreState()) {}
  88     DestroyDOMTree();
  89
  90     if (m_HandlersStack)
  91     {
  92         wxList& tmp = *m_HandlersStack;
  93         wxList::iterator it, en;
  94         for( it = tmp.begin(), en = tmp.end(); it != en; ++it )
  95             delete (wxHashTable*)*it;
  96         tmp.clear();
  97     }
  98     delete m_HandlersStack;
  99     m_HandlersHash.Clear();
 100     WX_CLEAR_LIST(wxList, m_HandlersList);
 101     delete m_entitiesParser;
 102 }
 103
 104 wxObject* wxHtmlParser::Parse(const wxString& source)
 105 {
 106     InitParser(source);
 107     DoParsing();
 108     wxObject *result = GetProduct();
 109     DoneParser();
 110     return result;
 111 }
 112
 113 void wxHtmlParser::InitParser(const wxString& source)
 114 {
 115     SetSource(source);
 116     m_stopParsing = false;
 117 }
 118
 119 void wxHtmlParser::DoneParser()
 120 {
 121     DestroyDOMTree();
 122 }
 123
 124 void wxHtmlParser::SetSource(const wxString& src)
 125 {
 126     DestroyDOMTree();
 127     m_Source = src;
 128     CreateDOMTree();
 129     m_CurTag = NULL;
 130     m_CurTextPiece = 0;
 131 }
 132
 133 void wxHtmlParser::CreateDOMTree()
 134 {
 135     wxHtmlTagsCache cache(m_Source);
 136     m_TextPieces = new wxHtmlTextPieces;
 137     CreateDOMSubTree(NULL, 0, m_Source.length(), &cache);
 138     m_CurTextPiece = 0;
 139 }
 140
 141 extern bool wxIsCDATAElement(const wxChar *tag);
 142
 143 void wxHtmlParser::CreateDOMSubTree(wxHtmlTag *cur,
 144                                     int begin_pos, int end_pos,
 145                                     wxHtmlTagsCache *cache)
 146 {
 147     if (end_pos <= begin_pos) return;
 148
 149     wxChar c;
 150     int i = begin_pos;
 151     int textBeginning = begin_pos;
 152
 153     // If the tag contains CDATA text, we include the text between beginning
 154     // and ending tag verbosely. Setting i=end_pos will skip to the very
 155     // end of this function where text piece is added, bypassing any child
 156     // tags parsing (CDATA element can't have child elements by definition):
 157     if (cur != NULL && wxIsCDATAElement(cur->GetName().c_str()))
 158     {
 159         i = end_pos;
 160     }
 161
 162     while (i < end_pos)
 163     {
 164         c = m_Source.GetChar(i);
 165
 166         if (c == wxT('<'))
 167         {
 168             // add text to m_TextPieces:
 169             if (i - textBeginning > 0)
 170                 m_TextPieces->Add(
 171                     wxHtmlTextPiece(textBeginning, i - textBeginning));
 172
 173             // if it is a comment, skip it:
 174             if (i < end_pos-6 && m_Source.GetChar(i+1) == wxT('!') &&
 175                                  m_Source.GetChar(i+2) == wxT('-') &&
 176                                  m_Source.GetChar(i+3) == wxT('-'))
 177             {
 178                 // Comments begin with "<!--" and end with "--[ \t\r\n]*>"
 179                 // according to HTML 4.0
 180                 int dashes = 0;
 181                 i += 4;
 182                 while (i < end_pos)
 183                 {
 184                     c = m_Source.GetChar(i++);
 185                     if ((c == wxT(' ') || c == wxT('\n') ||
 186                         c == wxT('\r') || c == wxT('\t')) && dashes >= 2) {}
 187                     else if (c == wxT('>') && dashes >= 2)
 188                     {
 189                         textBeginning = i;
 190                         break;
 191                     }
 192                     else if (c == wxT('-'))
 193                         dashes++;
 194                     else
 195                         dashes = 0;
 196                 }
 197             }
 198
 199             // add another tag to the tree:
 200             else if (i < end_pos-1 && m_Source.GetChar(i+1) != wxT('/'))
 201             {
 202                 wxHtmlTag *chd;
 203                 if (cur)
 204                     chd = new wxHtmlTag(cur, m_Source,
 205                                         i, end_pos, cache, m_entitiesParser);
 206                 else
 207                 {
 208                     chd = new wxHtmlTag(NULL, m_Source,
 209                                         i, end_pos, cache, m_entitiesParser);
 210                     if (!m_Tags)
 211                     {
 212                         // if this is the first tag to be created make the root
 213                         // m_Tags point to it:
 214                         m_Tags = chd;
 215                     }
 216                     else
 217                     {
 218                         // if there is already a root tag add this tag as
 219                         // the last sibling:
 220                         chd->m_Prev = m_Tags->GetLastSibling();
 221                         chd->m_Prev->m_Next = chd;
 222                     }
 223                 }
 224
 225                 if (chd->HasEnding())
 226                 {
 227                     CreateDOMSubTree(chd,
 228                                      chd->GetBeginPos(), chd->GetEndPos1(),
 229                                      cache);
 230                     i = chd->GetEndPos2();
 231                 }
 232                 else
 233                     i = chd->GetBeginPos();
 234
 235                 textBeginning = i;
 236             }
 237
 238             // ... or skip ending tag:
 239             else
 240             {
 241                 while (i < end_pos && m_Source.GetChar(i) != wxT('>')) i++;
 242                 textBeginning = i+1;
 243             }
 244         }
 245         else i++;
 246     }
 247
 248     // add remaining text to m_TextPieces:
 249     if (end_pos - textBeginning > 0)
 250         m_TextPieces->Add(
 251             wxHtmlTextPiece(textBeginning, end_pos - textBeginning));
 252 }
 253
 254 void wxHtmlParser::DestroyDOMTree()
 255 {
 256     wxHtmlTag *t1, *t2;
 257     t1 = m_Tags;
 258     while (t1)
 259     {
 260         t2 = t1->GetNextSibling();
 261         delete t1;
 262         t1 = t2;
 263     }
 264     m_Tags = m_CurTag = NULL;
 265
 266     delete m_TextPieces;
 267     m_TextPieces = NULL;
 268 }
 269
 270 void wxHtmlParser::DoParsing()
 271 {
 272     m_CurTag = m_Tags;
 273     m_CurTextPiece = 0;
 274     DoParsing(0, m_Source.length());
 275 }
 276
 277 void wxHtmlParser::DoParsing(int begin_pos, int end_pos)
 278 {
 279     if (end_pos <= begin_pos) return;
 280
 281     wxHtmlTextPieces& pieces = *m_TextPieces;
 282     size_t piecesCnt = pieces.GetCount();
 283
 284     while (begin_pos < end_pos)
 285     {
 286         while (m_CurTag && m_CurTag->GetBeginPos() < begin_pos)
 287             m_CurTag = m_CurTag->GetNextTag();
 288         while (m_CurTextPiece < piecesCnt &&
 289                pieces[m_CurTextPiece].m_pos < begin_pos)
 290             m_CurTextPiece++;
 291
 292         if (m_CurTextPiece < piecesCnt &&
 293             (!m_CurTag ||
 294              pieces[m_CurTextPiece].m_pos < m_CurTag->GetBeginPos()))
 295         {
 296             // Add text:
 297             AddText(GetEntitiesParser()->Parse(
 298                        m_Source.Mid(pieces[m_CurTextPiece].m_pos,
 299                                     pieces[m_CurTextPiece].m_lng)));
 300             begin_pos = pieces[m_CurTextPiece].m_pos +
 301                         pieces[m_CurTextPiece].m_lng;
 302             m_CurTextPiece++;
 303         }
 304         else if (m_CurTag)
 305         {
 306             if (m_CurTag->HasEnding())
 307                 begin_pos = m_CurTag->GetEndPos2();
 308             else
 309                 begin_pos = m_CurTag->GetBeginPos();
 310             wxHtmlTag *t = m_CurTag;
 311             m_CurTag = m_CurTag->GetNextTag();
 312             AddTag(*t);
 313             if (m_stopParsing)
 314                 return;
 315         }
 316         else break;
 317     }
 318 }
 319
 320 void wxHtmlParser::AddTag(const wxHtmlTag& tag)
 321 {
 322     wxHtmlTagHandler *h;
 323     bool inner = false;
 324
 325     h = (wxHtmlTagHandler*) m_HandlersHash.Get(tag.GetName());
 326     if (h)
 327     {
 328         inner = h->HandleTag(tag);
 329         if (m_stopParsing)
 330             return;
 331     }
 332     if (!inner)
 333     {
 334         if (tag.HasEnding())
 335             DoParsing(tag.GetBeginPos(), tag.GetEndPos1());
 336     }
 337 }
 338
 339 void wxHtmlParser::AddTagHandler(wxHtmlTagHandler *handler)
 340 {
 341     wxString s(handler->GetSupportedTags());
 342     wxStringTokenizer tokenizer(s, wxT(", "));
 343
 344     while (tokenizer.HasMoreTokens())
 345         m_HandlersHash.Put(tokenizer.GetNextToken(), handler);
 346
 347     if (m_HandlersList.IndexOf(handler) == wxNOT_FOUND)
 348         m_HandlersList.Append(handler);
 349
 350     handler->SetParser(this);
 351 }
 352
 353 void wxHtmlParser::PushTagHandler(wxHtmlTagHandler *handler, const wxString& tags)
 354 {
 355     wxStringTokenizer tokenizer(tags, wxT(", "));
 356     wxString key;
 357
 358     if (m_HandlersStack == NULL)
 359     {
 360         m_HandlersStack = new wxList;
 361     }
 362
 363     m_HandlersStack->Insert((wxObject*)new wxHashTable(m_HandlersHash));
 364
 365     while (tokenizer.HasMoreTokens())
 366     {
 367         key = tokenizer.GetNextToken();
 368         m_HandlersHash.Delete(key);
 369         m_HandlersHash.Put(key, handler);
 370     }
 371 }
 372
 373 void wxHtmlParser::PopTagHandler()
 374 {
 375     wxList::compatibility_iterator first;
 376
 377     if ( !m_HandlersStack ||
 378 #if wxUSE_STL
 379          !(first = m_HandlersStack->GetFirst())
 380 #else // !wxUSE_STL
 381          ((first = m_HandlersStack->GetFirst()) == NULL)
 382 #endif // wxUSE_STL/!wxUSE_STL
 383         )
 384     {
 385         wxLogWarning(_("Warning: attempt to remove HTML tag handler from empty stack."));
 386         return;
 387     }
 388     m_HandlersHash = *((wxHashTable*) first->GetData());
 389     delete (wxHashTable*) first->GetData();
 390     m_HandlersStack->Erase(first);
 391 }
 392
 393 void wxHtmlParser::SetSourceAndSaveState(const wxString& src)
 394 {
 395     wxHtmlParserState *s = new wxHtmlParserState;
 396
 397     s->m_curTag = m_CurTag;
 398     s->m_tags = m_Tags;
 399     s->m_textPieces = m_TextPieces;
 400     s->m_curTextPiece = m_CurTextPiece;
 401     s->m_source = m_Source;
 402
 403     s->m_nextState = m_SavedStates;
 404     m_SavedStates = s;
 405
 406     m_CurTag = NULL;
 407     m_Tags = NULL;
 408     m_TextPieces = NULL;
 409     m_CurTextPiece = 0;
 410     m_Source = wxEmptyString;
 411
 412     SetSource(src);
 413 }
 414
 415 bool wxHtmlParser::RestoreState()
 416 {
 417     if (!m_SavedStates) return false;
 418
 419     DestroyDOMTree();
 420
 421     wxHtmlParserState *s = m_SavedStates;
 422     m_SavedStates = s->m_nextState;
 423
 424     m_CurTag = s->m_curTag;
 425     m_Tags = s->m_tags;
 426     m_TextPieces = s->m_textPieces;
 427     m_CurTextPiece = s->m_curTextPiece;
 428     m_Source = s->m_source;
 429
 430     delete s;
 431     return true;
 432 }
 433
 434 wxString wxHtmlParser::GetInnerSource(const wxHtmlTag& tag)
 435 {
 436     return GetSource()->Mid(tag.GetBeginPos(),
 437                             tag.GetEndPos1() - tag.GetBeginPos());
 438 }
 439
 440 //-----------------------------------------------------------------------------
 441 // wxHtmlTagHandler
 442 //-----------------------------------------------------------------------------
 443
 444 IMPLEMENT_ABSTRACT_CLASS(wxHtmlTagHandler,wxObject)
 445
 446 void wxHtmlTagHandler::ParseInnerSource(const wxString& source)
 447 {
 448     // It is safe to temporarily change the source being parsed,
 449     // provided we restore the state back after parsing
 450     m_Parser->SetSourceAndSaveState(source);
 451     m_Parser->DoParsing();
 452     m_Parser->RestoreState();
 453 }
 454
 455
 456 //-----------------------------------------------------------------------------
 457 // wxHtmlEntitiesParser
 458 //-----------------------------------------------------------------------------
 459
 460 IMPLEMENT_DYNAMIC_CLASS(wxHtmlEntitiesParser,wxObject)
 461
 462 wxHtmlEntitiesParser::wxHtmlEntitiesParser()
 463 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 464     : m_conv(NULL), m_encoding(wxFONTENCODING_SYSTEM)
 465 #endif
 466 {
 467 }
 468
 469 wxHtmlEntitiesParser::~wxHtmlEntitiesParser()
 470 {
 471 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 472     delete m_conv;
 473 #endif
 474 }
 475
 476 void wxHtmlEntitiesParser::SetEncoding(wxFontEncoding encoding)
 477 {
 478 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 479     if (encoding == m_encoding)
 480         return;
 481
 482     delete m_conv;
 483
 484     m_encoding = encoding;
 485     if (m_encoding == wxFONTENCODING_SYSTEM)
 486         m_conv = NULL;
 487     else
 488         m_conv = new wxCSConv(wxFontMapper::GetEncodingName(m_encoding));
 489 #else
 490     (void) encoding;
 491 #endif
 492 }
 493
 494 wxString wxHtmlEntitiesParser::Parse(const wxString& input)
 495 {
 496     const wxChar *c, *last;
 497     const wxChar *in_str = input.c_str();
 498     wxString output;
 499
 500     output.reserve(input.length());
 501
 502     for (c = in_str, last = in_str; *c != wxT('\0'); c++)
 503     {
 504         if (*c == wxT('&'))
 505         {
 506             if (c - last > 0)
 507                 output.append(last, c - last);
 508             if ( *++c == wxT('\0') )
 509                 break;
 510
 511             wxString entity;
 512             const wxChar *ent_s = c;
 513             wxChar entity_char;
 514
 515             for (; (*c >= wxT('a') && *c <= wxT('z')) ||
 516                    (*c >= wxT('A') && *c <= wxT('Z')) ||
 517                    (*c >= wxT('0') && *c <= wxT('9')) ||
 518                    *c == wxT('_') || *c == wxT('#'); c++) {}
 519             entity.append(ent_s, c - ent_s);
 520             if (*c != wxT(';')) c--;
 521             last = c+1;
 522             entity_char = GetEntityChar(entity);
 523             if (entity_char)
 524                 output << entity_char;
 525             else
 526             {
 527                 output.append(ent_s-1, c-ent_s+2);
 528                 wxLogTrace(wxTRACE_HTML_DEBUG,
 529                            wxT("Unrecognized HTML entity: '%s'"),
 530                            entity.c_str());
 531             }
 532         }
 533     }
 534     if (*last != wxT('\0'))
 535         output.append(last);
 536     return output;
 537 }
 538
 539 struct wxHtmlEntityInfo
 540 {
 541     const wxChar *name;
 542     unsigned code;
 543 };
 544
 545 extern "C" int LINKAGEMODE wxHtmlEntityCompare(const void *key, const void *item)
 546 {
 547     return wxStrcmp((wxChar*)key, ((wxHtmlEntityInfo*)item)->name);
 548 }
 549
 550 #if !wxUSE_UNICODE
 551 wxChar wxHtmlEntitiesParser::GetCharForCode(unsigned code)
 552 {
 553 #if wxUSE_WCHAR_T
 554     char buf[2];
 555     wchar_t wbuf[2];
 556     wbuf[0] = (wchar_t)code;
 557     wbuf[1] = 0;
 558     wxMBConv *conv = m_conv ? m_conv : &wxConvLocal;
 559     if (conv->WC2MB(buf, wbuf, 2) == (size_t)-1)
 560         return '?';
 561     return buf[0];
 562 #else
 563     return (code < 256) ? (wxChar)code : '?';
 564 #endif
 565 }
 566 #endif
 567
 568 wxChar wxHtmlEntitiesParser::GetEntityChar(const wxString& entity)
 569 {
 570     unsigned code = 0;
 571
 572     if (entity[0] == wxT('#'))
 573     {
 574         const wxChar *ent_s = entity.c_str();
 575         const wxChar *format;
 576
 577         if (ent_s[1] == wxT('x') || ent_s[1] == wxT('X'))
 578         {
 579             format = wxT("%x");
 580             ent_s++;
 581         }
 582         else
 583             format = wxT("%u");
 584         ent_s++;
 585
 586         if (wxSscanf(ent_s, format, &code) != 1)
 587             code = 0;
 588     }
 589     else
 590     {
 591         static wxHtmlEntityInfo substitutions[] = {
 592             { wxT("AElig"),198 },
 593             { wxT("Aacute"),193 },
 594             { wxT("Acirc"),194 },
 595             { wxT("Agrave"),192 },
 596             { wxT("Alpha"),913 },
 597             { wxT("Aring"),197 },
 598             { wxT("Atilde"),195 },
 599             { wxT("Auml"),196 },
 600             { wxT("Beta"),914 },
 601             { wxT("Ccedil"),199 },
 602             { wxT("Chi"),935 },
 603             { wxT("Dagger"),8225 },
 604             { wxT("Delta"),916 },
 605             { wxT("ETH"),208 },
 606             { wxT("Eacute"),201 },
 607             { wxT("Ecirc"),202 },
 608             { wxT("Egrave"),200 },
 609             { wxT("Epsilon"),917 },
 610             { wxT("Eta"),919 },
 611             { wxT("Euml"),203 },
 612             { wxT("Gamma"),915 },
 613             { wxT("Iacute"),205 },
 614             { wxT("Icirc"),206 },
 615             { wxT("Igrave"),204 },
 616             { wxT("Iota"),921 },
 617             { wxT("Iuml"),207 },
 618             { wxT("Kappa"),922 },
 619             { wxT("Lambda"),923 },
 620             { wxT("Mu"),924 },
 621             { wxT("Ntilde"),209 },
 622             { wxT("Nu"),925 },
 623             { wxT("OElig"),338 },
 624             { wxT("Oacute"),211 },
 625             { wxT("Ocirc"),212 },
 626             { wxT("Ograve"),210 },
 627             { wxT("Omega"),937 },
 628             { wxT("Omicron"),927 },
 629             { wxT("Oslash"),216 },
 630             { wxT("Otilde"),213 },
 631             { wxT("Ouml"),214 },
 632             { wxT("Phi"),934 },
 633             { wxT("Pi"),928 },
 634             { wxT("Prime"),8243 },
 635             { wxT("Psi"),936 },
 636             { wxT("Rho"),929 },
 637             { wxT("Scaron"),352 },
 638             { wxT("Sigma"),931 },
 639             { wxT("THORN"),222 },
 640             { wxT("Tau"),932 },
 641             { wxT("Theta"),920 },
 642             { wxT("Uacute"),218 },
 643             { wxT("Ucirc"),219 },
 644             { wxT("Ugrave"),217 },
 645             { wxT("Upsilon"),933 },
 646             { wxT("Uuml"),220 },
 647             { wxT("Xi"),926 },
 648             { wxT("Yacute"),221 },
 649             { wxT("Yuml"),376 },
 650             { wxT("Zeta"),918 },
 651             { wxT("aacute"),225 },
 652             { wxT("acirc"),226 },
 653             { wxT("acute"),180 },
 654             { wxT("aelig"),230 },
 655             { wxT("agrave"),224 },
 656             { wxT("alefsym"),8501 },
 657             { wxT("alpha"),945 },
 658             { wxT("amp"),38 },
 659             { wxT("and"),8743 },
 660             { wxT("ang"),8736 },
 661             { wxT("aring"),229 },
 662             { wxT("asymp"),8776 },
 663             { wxT("atilde"),227 },
 664             { wxT("auml"),228 },
 665             { wxT("bdquo"),8222 },
 666             { wxT("beta"),946 },
 667             { wxT("brvbar"),166 },
 668             { wxT("bull"),8226 },
 669             { wxT("cap"),8745 },
 670             { wxT("ccedil"),231 },
 671             { wxT("cedil"),184 },
 672             { wxT("cent"),162 },
 673             { wxT("chi"),967 },
 674             { wxT("circ"),710 },
 675             { wxT("clubs"),9827 },
 676             { wxT("cong"),8773 },
 677             { wxT("copy"),169 },
 678             { wxT("crarr"),8629 },
 679             { wxT("cup"),8746 },
 680             { wxT("curren"),164 },
 681             { wxT("dArr"),8659 },
 682             { wxT("dagger"),8224 },
 683             { wxT("darr"),8595 },
 684             { wxT("deg"),176 },
 685             { wxT("delta"),948 },
 686             { wxT("diams"),9830 },
 687             { wxT("divide"),247 },
 688             { wxT("eacute"),233 },
 689             { wxT("ecirc"),234 },
 690             { wxT("egrave"),232 },
 691             { wxT("empty"),8709 },
 692             { wxT("emsp"),8195 },
 693             { wxT("ensp"),8194 },
 694             { wxT("epsilon"),949 },
 695             { wxT("equiv"),8801 },
 696             { wxT("eta"),951 },
 697             { wxT("eth"),240 },
 698             { wxT("euml"),235 },
 699             { wxT("euro"),8364 },
 700             { wxT("exist"),8707 },
 701             { wxT("fnof"),402 },
 702             { wxT("forall"),8704 },
 703             { wxT("frac12"),189 },
 704             { wxT("frac14"),188 },
 705             { wxT("frac34"),190 },
 706             { wxT("frasl"),8260 },
 707             { wxT("gamma"),947 },
 708             { wxT("ge"),8805 },
 709             { wxT("gt"),62 },
 710             { wxT("hArr"),8660 },
 711             { wxT("harr"),8596 },
 712             { wxT("hearts"),9829 },
 713             { wxT("hellip"),8230 },
 714             { wxT("iacute"),237 },
 715             { wxT("icirc"),238 },
 716             { wxT("iexcl"),161 },
 717             { wxT("igrave"),236 },
 718             { wxT("image"),8465 },
 719             { wxT("infin"),8734 },
 720             { wxT("int"),8747 },
 721             { wxT("iota"),953 },
 722             { wxT("iquest"),191 },
 723             { wxT("isin"),8712 },
 724             { wxT("iuml"),239 },
 725             { wxT("kappa"),954 },
 726             { wxT("lArr"),8656 },
 727             { wxT("lambda"),955 },
 728             { wxT("lang"),9001 },
 729             { wxT("laquo"),171 },
 730             { wxT("larr"),8592 },
 731             { wxT("lceil"),8968 },
 732             { wxT("ldquo"),8220 },
 733             { wxT("le"),8804 },
 734             { wxT("lfloor"),8970 },
 735             { wxT("lowast"),8727 },
 736             { wxT("loz"),9674 },
 737             { wxT("lrm"),8206 },
 738             { wxT("lsaquo"),8249 },
 739             { wxT("lsquo"),8216 },
 740             { wxT("lt"),60 },
 741             { wxT("macr"),175 },
 742             { wxT("mdash"),8212 },
 743             { wxT("micro"),181 },
 744             { wxT("middot"),183 },
 745             { wxT("minus"),8722 },
 746             { wxT("mu"),956 },
 747             { wxT("nabla"),8711 },
 748             { wxT("nbsp"),160 },
 749             { wxT("ndash"),8211 },
 750             { wxT("ne"),8800 },
 751             { wxT("ni"),8715 },
 752             { wxT("not"),172 },
 753             { wxT("notin"),8713 },
 754             { wxT("nsub"),8836 },
 755             { wxT("ntilde"),241 },
 756             { wxT("nu"),957 },
 757             { wxT("oacute"),243 },
 758             { wxT("ocirc"),244 },
 759             { wxT("oelig"),339 },
 760             { wxT("ograve"),242 },
 761             { wxT("oline"),8254 },
 762             { wxT("omega"),969 },
 763             { wxT("omicron"),959 },
 764             { wxT("oplus"),8853 },
 765             { wxT("or"),8744 },
 766             { wxT("ordf"),170 },
 767             { wxT("ordm"),186 },
 768             { wxT("oslash"),248 },
 769             { wxT("otilde"),245 },
 770             { wxT("otimes"),8855 },
 771             { wxT("ouml"),246 },
 772             { wxT("para"),182 },
 773             { wxT("part"),8706 },
 774             { wxT("permil"),8240 },
 775             { wxT("perp"),8869 },
 776             { wxT("phi"),966 },
 777             { wxT("pi"),960 },
 778             { wxT("piv"),982 },
 779             { wxT("plusmn"),177 },
 780             { wxT("pound"),163 },
 781             { wxT("prime"),8242 },
 782             { wxT("prod"),8719 },
 783             { wxT("prop"),8733 },
 784             { wxT("psi"),968 },
 785             { wxT("quot"),34 },
 786             { wxT("rArr"),8658 },
 787             { wxT("radic"),8730 },
 788             { wxT("rang"),9002 },
 789             { wxT("raquo"),187 },
 790             { wxT("rarr"),8594 },
 791             { wxT("rceil"),8969 },
 792             { wxT("rdquo"),8221 },
 793             { wxT("real"),8476 },
 794             { wxT("reg"),174 },
 795             { wxT("rfloor"),8971 },
 796             { wxT("rho"),961 },
 797             { wxT("rlm"),8207 },
 798             { wxT("rsaquo"),8250 },
 799             { wxT("rsquo"),8217 },
 800             { wxT("sbquo"),8218 },
 801             { wxT("scaron"),353 },
 802             { wxT("sdot"),8901 },
 803             { wxT("sect"),167 },
 804             { wxT("shy"),173 },
 805             { wxT("sigma"),963 },
 806             { wxT("sigmaf"),962 },
 807             { wxT("sim"),8764 },
 808             { wxT("spades"),9824 },
 809             { wxT("sub"),8834 },
 810             { wxT("sube"),8838 },
 811             { wxT("sum"),8721 },
 812             { wxT("sup"),8835 },
 813             { wxT("sup1"),185 },
 814             { wxT("sup2"),178 },
 815             { wxT("sup3"),179 },
 816             { wxT("supe"),8839 },
 817             { wxT("szlig"),223 },
 818             { wxT("tau"),964 },
 819             { wxT("there4"),8756 },
 820             { wxT("theta"),952 },
 821             { wxT("thetasym"),977 },
 822             { wxT("thinsp"),8201 },
 823             { wxT("thorn"),254 },
 824             { wxT("tilde"),732 },
 825             { wxT("times"),215 },
 826             { wxT("trade"),8482 },
 827             { wxT("uArr"),8657 },
 828             { wxT("uacute"),250 },
 829             { wxT("uarr"),8593 },
 830             { wxT("ucirc"),251 },
 831             { wxT("ugrave"),249 },
 832             { wxT("uml"),168 },
 833             { wxT("upsih"),978 },
 834             { wxT("upsilon"),965 },
 835             { wxT("uuml"),252 },
 836             { wxT("weierp"),8472 },
 837             { wxT("xi"),958 },
 838             { wxT("yacute"),253 },
 839             { wxT("yen"),165 },
 840             { wxT("yuml"),255 },
 841             { wxT("zeta"),950 },
 842             { wxT("zwj"),8205 },
 843             { wxT("zwnj"),8204 },
 844             {NULL, 0}};
 845         static size_t substitutions_cnt = 0;
 846
 847         if (substitutions_cnt == 0)
 848             while (substitutions[substitutions_cnt].code != 0)
 849                 substitutions_cnt++;
 850
 851         wxHtmlEntityInfo *info = NULL;
 852 #ifdef __WXWINCE__
 853         // bsearch crashes under WinCE for some reason
 854         size_t i;
 855         for (i = 0; i < substitutions_cnt; i++)
 856         {
 857             if (entity == substitutions[i].name)
 858             {
 859                 info = & substitutions[i];
 860                 break;
 861             }
 862         }
 863 #else
 864         info = (wxHtmlEntityInfo*) bsearch(entity.c_str(), substitutions,
 865                                            substitutions_cnt,
 866                                            sizeof(wxHtmlEntityInfo),
 867                                            wxHtmlEntityCompare);
 868 #endif
 869         if (info)
 870             code = info->code;
 871     }
 872
 873     if (code == 0)
 874         return 0;
 875     else
 876         return GetCharForCode(code);
 877 }
 878
 879 wxFSFile *wxHtmlParser::OpenURL(wxHtmlURLType WXUNUSED(type),
 880                                 const wxString& url) const
 881 {
 882     return m_FS ? m_FS->OpenFile(url) : NULL;
 883
 884 }
 885
 886
 887 //-----------------------------------------------------------------------------
 888 // wxHtmlParser::ExtractCharsetInformation
 889 //-----------------------------------------------------------------------------
 890
 891 class wxMetaTagParser : public wxHtmlParser
 892 {
 893 public:
 894     wxMetaTagParser() { }
 895
 896     wxObject* GetProduct() { return NULL; }
 897
 898 protected:
 899     virtual void AddText(const wxChar* WXUNUSED(txt)) {}
 900
 901     DECLARE_NO_COPY_CLASS(wxMetaTagParser)
 902 };
 903
 904 class wxMetaTagHandler : public wxHtmlTagHandler
 905 {
 906 public:
 907     wxMetaTagHandler(wxString *retval) : wxHtmlTagHandler(), m_retval(retval) {}
 908     wxString GetSupportedTags() { return wxT("META,BODY"); }
 909     bool HandleTag(const wxHtmlTag& tag);
 910
 911 private:
 912     wxString *m_retval;
 913
 914     DECLARE_NO_COPY_CLASS(wxMetaTagHandler)
 915 };
 916
 917 bool wxMetaTagHandler::HandleTag(const wxHtmlTag& tag)
 918 {
 919     if (tag.GetName() == _T("BODY"))
 920     {
 921         m_Parser->StopParsing();
 922         return false;
 923     }
 924
 925     if (tag.HasParam(_T("HTTP-EQUIV")) &&
 926         tag.GetParam(_T("HTTP-EQUIV")).IsSameAs(_T("Content-Type"), false) &&
 927         tag.HasParam(_T("CONTENT")))
 928     {
 929         wxString content = tag.GetParam(_T("CONTENT")).Lower();
 930         if (content.Left(19) == _T("text/html; charset="))
 931         {
 932             *m_retval = content.Mid(19);
 933             m_Parser->StopParsing();
 934         }
 935     }
 936     return false;
 937 }
 938
 939
 940 /*static*/
 941 wxString wxHtmlParser::ExtractCharsetInformation(const wxString& markup)
 942 {
 943     wxString charset;
 944     wxMetaTagParser *parser = new wxMetaTagParser();
 945     if(parser)
 946     {
 947         parser->AddTagHandler(new wxMetaTagHandler(&charset));
 948         parser->Parse(markup);
 949         delete parser;
 950     }
 951     return charset;
 952 }
 953
 954 #endif