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