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             wxString::const_iterator iter = m_Source.begin() + i;
 175             if ( SkipCommentTag(iter, m_Source.end()) )
 176             {
 177                 textBeginning =
 178                 i = iter - m_Source.begin() + 1; // skip closing '>' too
 179             }
 180
 181             // add another tag to the tree:
 182             else if (i < end_pos-1 && m_Source.GetChar(i+1) != wxT('/'))
 183             {
 184                 wxHtmlTag *chd;
 185                 if (cur)
 186                     chd = new wxHtmlTag(cur, m_Source,
 187                                         i, end_pos, cache, m_entitiesParser);
 188                 else
 189                 {
 190                     chd = new wxHtmlTag(NULL, m_Source,
 191                                         i, end_pos, cache, m_entitiesParser);
 192                     if (!m_Tags)
 193                     {
 194                         // if this is the first tag to be created make the root
 195                         // m_Tags point to it:
 196                         m_Tags = chd;
 197                     }
 198                     else
 199                     {
 200                         // if there is already a root tag add this tag as
 201                         // the last sibling:
 202                         chd->m_Prev = m_Tags->GetLastSibling();
 203                         chd->m_Prev->m_Next = chd;
 204                     }
 205                 }
 206
 207                 if (chd->HasEnding())
 208                 {
 209                     CreateDOMSubTree(chd,
 210                                      chd->GetBeginPos(), chd->GetEndPos1(),
 211                                      cache);
 212                     i = chd->GetEndPos2();
 213                 }
 214                 else
 215                     i = chd->GetBeginPos();
 216
 217                 textBeginning = i;
 218             }
 219
 220             // ... or skip ending tag:
 221             else
 222             {
 223                 while (i < end_pos && m_Source.GetChar(i) != wxT('>')) i++;
 224                 textBeginning = i+1;
 225             }
 226         }
 227         else i++;
 228     }
 229
 230     // add remaining text to m_TextPieces:
 231     if (end_pos - textBeginning > 0)
 232         m_TextPieces->Add(
 233             wxHtmlTextPiece(textBeginning, end_pos - textBeginning));
 234 }
 235
 236 void wxHtmlParser::DestroyDOMTree()
 237 {
 238     wxHtmlTag *t1, *t2;
 239     t1 = m_Tags;
 240     while (t1)
 241     {
 242         t2 = t1->GetNextSibling();
 243         delete t1;
 244         t1 = t2;
 245     }
 246     m_Tags = m_CurTag = NULL;
 247
 248     delete m_TextPieces;
 249     m_TextPieces = NULL;
 250 }
 251
 252 void wxHtmlParser::DoParsing()
 253 {
 254     m_CurTag = m_Tags;
 255     m_CurTextPiece = 0;
 256     DoParsing(0, m_Source.length());
 257 }
 258
 259 void wxHtmlParser::DoParsing(int begin_pos, int end_pos)
 260 {
 261     if (end_pos <= begin_pos) return;
 262
 263     wxHtmlTextPieces& pieces = *m_TextPieces;
 264     size_t piecesCnt = pieces.GetCount();
 265
 266     while (begin_pos < end_pos)
 267     {
 268         while (m_CurTag && m_CurTag->GetBeginPos() < begin_pos)
 269             m_CurTag = m_CurTag->GetNextTag();
 270         while (m_CurTextPiece < piecesCnt &&
 271                pieces[m_CurTextPiece].m_pos < begin_pos)
 272             m_CurTextPiece++;
 273
 274         if (m_CurTextPiece < piecesCnt &&
 275             (!m_CurTag ||
 276              pieces[m_CurTextPiece].m_pos < m_CurTag->GetBeginPos()))
 277         {
 278             // Add text:
 279             AddText(GetEntitiesParser()->Parse(
 280                        m_Source.Mid(pieces[m_CurTextPiece].m_pos,
 281                                     pieces[m_CurTextPiece].m_lng)));
 282             begin_pos = pieces[m_CurTextPiece].m_pos +
 283                         pieces[m_CurTextPiece].m_lng;
 284             m_CurTextPiece++;
 285         }
 286         else if (m_CurTag)
 287         {
 288             if (m_CurTag->HasEnding())
 289                 begin_pos = m_CurTag->GetEndPos2();
 290             else
 291                 begin_pos = m_CurTag->GetBeginPos();
 292             wxHtmlTag *t = m_CurTag;
 293             m_CurTag = m_CurTag->GetNextTag();
 294             AddTag(*t);
 295             if (m_stopParsing)
 296                 return;
 297         }
 298         else break;
 299     }
 300 }
 301
 302 void wxHtmlParser::AddTag(const wxHtmlTag& tag)
 303 {
 304     wxHtmlTagHandler *h;
 305     bool inner = false;
 306
 307     h = (wxHtmlTagHandler*) m_HandlersHash.Get(tag.GetName());
 308     if (h)
 309     {
 310         inner = h->HandleTag(tag);
 311         if (m_stopParsing)
 312             return;
 313     }
 314     if (!inner)
 315     {
 316         if (tag.HasEnding())
 317             DoParsing(tag.GetBeginPos(), tag.GetEndPos1());
 318     }
 319 }
 320
 321 void wxHtmlParser::AddTagHandler(wxHtmlTagHandler *handler)
 322 {
 323     wxString s(handler->GetSupportedTags());
 324     wxStringTokenizer tokenizer(s, wxT(", "));
 325
 326     while (tokenizer.HasMoreTokens())
 327         m_HandlersHash.Put(tokenizer.GetNextToken(), handler);
 328
 329     if (m_HandlersList.IndexOf(handler) == wxNOT_FOUND)
 330         m_HandlersList.Append(handler);
 331
 332     handler->SetParser(this);
 333 }
 334
 335 void wxHtmlParser::PushTagHandler(wxHtmlTagHandler *handler, const wxString& tags)
 336 {
 337     wxStringTokenizer tokenizer(tags, wxT(", "));
 338     wxString key;
 339
 340     if (m_HandlersStack == NULL)
 341     {
 342         m_HandlersStack = new wxList;
 343     }
 344
 345     m_HandlersStack->Insert((wxObject*)new wxHashTable(m_HandlersHash));
 346
 347     while (tokenizer.HasMoreTokens())
 348     {
 349         key = tokenizer.GetNextToken();
 350         m_HandlersHash.Delete(key);
 351         m_HandlersHash.Put(key, handler);
 352     }
 353 }
 354
 355 void wxHtmlParser::PopTagHandler()
 356 {
 357     wxList::compatibility_iterator first;
 358
 359     if ( !m_HandlersStack ||
 360 #if wxUSE_STL
 361          !(first = m_HandlersStack->GetFirst())
 362 #else // !wxUSE_STL
 363          ((first = m_HandlersStack->GetFirst()) == NULL)
 364 #endif // wxUSE_STL/!wxUSE_STL
 365         )
 366     {
 367         wxLogWarning(_("Warning: attempt to remove HTML tag handler from empty stack."));
 368         return;
 369     }
 370     m_HandlersHash = *((wxHashTable*) first->GetData());
 371     delete (wxHashTable*) first->GetData();
 372     m_HandlersStack->Erase(first);
 373 }
 374
 375 void wxHtmlParser::SetSourceAndSaveState(const wxString& src)
 376 {
 377     wxHtmlParserState *s = new wxHtmlParserState;
 378
 379     s->m_curTag = m_CurTag;
 380     s->m_tags = m_Tags;
 381     s->m_textPieces = m_TextPieces;
 382     s->m_curTextPiece = m_CurTextPiece;
 383     s->m_source = m_Source;
 384
 385     s->m_nextState = m_SavedStates;
 386     m_SavedStates = s;
 387
 388     m_CurTag = NULL;
 389     m_Tags = NULL;
 390     m_TextPieces = NULL;
 391     m_CurTextPiece = 0;
 392     m_Source = wxEmptyString;
 393
 394     SetSource(src);
 395 }
 396
 397 bool wxHtmlParser::RestoreState()
 398 {
 399     if (!m_SavedStates) return false;
 400
 401     DestroyDOMTree();
 402
 403     wxHtmlParserState *s = m_SavedStates;
 404     m_SavedStates = s->m_nextState;
 405
 406     m_CurTag = s->m_curTag;
 407     m_Tags = s->m_tags;
 408     m_TextPieces = s->m_textPieces;
 409     m_CurTextPiece = s->m_curTextPiece;
 410     m_Source = s->m_source;
 411
 412     delete s;
 413     return true;
 414 }
 415
 416 wxString wxHtmlParser::GetInnerSource(const wxHtmlTag& tag)
 417 {
 418     return GetSource()->Mid(tag.GetBeginPos(),
 419                             tag.GetEndPos1() - tag.GetBeginPos());
 420 }
 421
 422 //-----------------------------------------------------------------------------
 423 // wxHtmlTagHandler
 424 //-----------------------------------------------------------------------------
 425
 426 IMPLEMENT_ABSTRACT_CLASS(wxHtmlTagHandler,wxObject)
 427
 428 void wxHtmlTagHandler::ParseInnerSource(const wxString& source)
 429 {
 430     // It is safe to temporarily change the source being parsed,
 431     // provided we restore the state back after parsing
 432     m_Parser->SetSourceAndSaveState(source);
 433     m_Parser->DoParsing();
 434     m_Parser->RestoreState();
 435 }
 436
 437
 438 //-----------------------------------------------------------------------------
 439 // wxHtmlEntitiesParser
 440 //-----------------------------------------------------------------------------
 441
 442 IMPLEMENT_DYNAMIC_CLASS(wxHtmlEntitiesParser,wxObject)
 443
 444 wxHtmlEntitiesParser::wxHtmlEntitiesParser()
 445 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 446     : m_conv(NULL), m_encoding(wxFONTENCODING_SYSTEM)
 447 #endif
 448 {
 449 }
 450
 451 wxHtmlEntitiesParser::~wxHtmlEntitiesParser()
 452 {
 453 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 454     delete m_conv;
 455 #endif
 456 }
 457
 458 void wxHtmlEntitiesParser::SetEncoding(wxFontEncoding encoding)
 459 {
 460 #if wxUSE_WCHAR_T && !wxUSE_UNICODE
 461     if (encoding == m_encoding)
 462         return;
 463
 464     delete m_conv;
 465
 466     m_encoding = encoding;
 467     if (m_encoding == wxFONTENCODING_SYSTEM)
 468         m_conv = NULL;
 469     else
 470         m_conv = new wxCSConv(wxFontMapper::GetEncodingName(m_encoding));
 471 #else
 472     (void) encoding;
 473 #endif
 474 }
 475
 476 wxString wxHtmlEntitiesParser::Parse(const wxString& input)
 477 {
 478     const wxChar *c, *last;
 479     const wxChar *in_str = input.c_str();
 480     wxString output;
 481
 482     output.reserve(input.length());
 483
 484     for (c = in_str, last = in_str; *c != wxT('\0'); c++)
 485     {
 486         if (*c == wxT('&'))
 487         {
 488             if (c - last > 0)
 489                 output.append(last, c - last);
 490             if ( *++c == wxT('\0') )
 491                 break;
 492
 493             wxString entity;
 494             const wxChar *ent_s = c;
 495             wxChar entity_char;
 496
 497             for (; (*c >= wxT('a') && *c <= wxT('z')) ||
 498                    (*c >= wxT('A') && *c <= wxT('Z')) ||
 499                    (*c >= wxT('0') && *c <= wxT('9')) ||
 500                    *c == wxT('_') || *c == wxT('#'); c++) {}
 501             entity.append(ent_s, c - ent_s);
 502             if (*c != wxT(';')) c--;
 503             last = c+1;
 504             entity_char = GetEntityChar(entity);
 505             if (entity_char)
 506                 output << entity_char;
 507             else
 508             {
 509                 output.append(ent_s-1, c-ent_s+2);
 510                 wxLogTrace(wxTRACE_HTML_DEBUG,
 511                            wxT("Unrecognized HTML entity: '%s'"),
 512                            entity.c_str());
 513             }
 514         }
 515     }
 516     if (*last != wxT('\0'))
 517         output.append(last);
 518     return output;
 519 }
 520
 521 struct wxHtmlEntityInfo
 522 {
 523     const wxChar *name;
 524     unsigned code;
 525 };
 526
 527 extern "C" int LINKAGEMODE wxHtmlEntityCompare(const void *key, const void *item)
 528 {
 529     return wxStrcmp((wxChar*)key, ((wxHtmlEntityInfo*)item)->name);
 530 }
 531
 532 #if !wxUSE_UNICODE
 533 wxChar wxHtmlEntitiesParser::GetCharForCode(unsigned code)
 534 {
 535 #if wxUSE_WCHAR_T
 536     char buf[2];
 537     wchar_t wbuf[2];
 538     wbuf[0] = (wchar_t)code;
 539     wbuf[1] = 0;
 540     wxMBConv *conv = m_conv ? m_conv : &wxConvLocal;
 541     if (conv->WC2MB(buf, wbuf, 2) == (size_t)-1)
 542         return '?';
 543     return buf[0];
 544 #else
 545     return (code < 256) ? (wxChar)code : '?';
 546 #endif
 547 }
 548 #endif
 549
 550 wxChar wxHtmlEntitiesParser::GetEntityChar(const wxString& entity)
 551 {
 552     unsigned code = 0;
 553
 554     if (entity[0] == wxT('#'))
 555     {
 556         const wxChar *ent_s = entity.c_str();
 557         const wxChar *format;
 558
 559         if (ent_s[1] == wxT('x') || ent_s[1] == wxT('X'))
 560         {
 561             format = wxT("%x");
 562             ent_s++;
 563         }
 564         else
 565             format = wxT("%u");
 566         ent_s++;
 567
 568         if (wxSscanf(ent_s, format, &code) != 1)
 569             code = 0;
 570     }
 571     else
 572     {
 573         static wxHtmlEntityInfo substitutions[] = {
 574             { wxT("AElig"),198 },
 575             { wxT("Aacute"),193 },
 576             { wxT("Acirc"),194 },
 577             { wxT("Agrave"),192 },
 578             { wxT("Alpha"),913 },
 579             { wxT("Aring"),197 },
 580             { wxT("Atilde"),195 },
 581             { wxT("Auml"),196 },
 582             { wxT("Beta"),914 },
 583             { wxT("Ccedil"),199 },
 584             { wxT("Chi"),935 },
 585             { wxT("Dagger"),8225 },
 586             { wxT("Delta"),916 },
 587             { wxT("ETH"),208 },
 588             { wxT("Eacute"),201 },
 589             { wxT("Ecirc"),202 },
 590             { wxT("Egrave"),200 },
 591             { wxT("Epsilon"),917 },
 592             { wxT("Eta"),919 },
 593             { wxT("Euml"),203 },
 594             { wxT("Gamma"),915 },
 595             { wxT("Iacute"),205 },
 596             { wxT("Icirc"),206 },
 597             { wxT("Igrave"),204 },
 598             { wxT("Iota"),921 },
 599             { wxT("Iuml"),207 },
 600             { wxT("Kappa"),922 },
 601             { wxT("Lambda"),923 },
 602             { wxT("Mu"),924 },
 603             { wxT("Ntilde"),209 },
 604             { wxT("Nu"),925 },
 605             { wxT("OElig"),338 },
 606             { wxT("Oacute"),211 },
 607             { wxT("Ocirc"),212 },
 608             { wxT("Ograve"),210 },
 609             { wxT("Omega"),937 },
 610             { wxT("Omicron"),927 },
 611             { wxT("Oslash"),216 },
 612             { wxT("Otilde"),213 },
 613             { wxT("Ouml"),214 },
 614             { wxT("Phi"),934 },
 615             { wxT("Pi"),928 },
 616             { wxT("Prime"),8243 },
 617             { wxT("Psi"),936 },
 618             { wxT("Rho"),929 },
 619             { wxT("Scaron"),352 },
 620             { wxT("Sigma"),931 },
 621             { wxT("THORN"),222 },
 622             { wxT("Tau"),932 },
 623             { wxT("Theta"),920 },
 624             { wxT("Uacute"),218 },
 625             { wxT("Ucirc"),219 },
 626             { wxT("Ugrave"),217 },
 627             { wxT("Upsilon"),933 },
 628             { wxT("Uuml"),220 },
 629             { wxT("Xi"),926 },
 630             { wxT("Yacute"),221 },
 631             { wxT("Yuml"),376 },
 632             { wxT("Zeta"),918 },
 633             { wxT("aacute"),225 },
 634             { wxT("acirc"),226 },
 635             { wxT("acute"),180 },
 636             { wxT("aelig"),230 },
 637             { wxT("agrave"),224 },
 638             { wxT("alefsym"),8501 },
 639             { wxT("alpha"),945 },
 640             { wxT("amp"),38 },
 641             { wxT("and"),8743 },
 642             { wxT("ang"),8736 },
 643             { wxT("aring"),229 },
 644             { wxT("asymp"),8776 },
 645             { wxT("atilde"),227 },
 646             { wxT("auml"),228 },
 647             { wxT("bdquo"),8222 },
 648             { wxT("beta"),946 },
 649             { wxT("brvbar"),166 },
 650             { wxT("bull"),8226 },
 651             { wxT("cap"),8745 },
 652             { wxT("ccedil"),231 },
 653             { wxT("cedil"),184 },
 654             { wxT("cent"),162 },
 655             { wxT("chi"),967 },
 656             { wxT("circ"),710 },
 657             { wxT("clubs"),9827 },
 658             { wxT("cong"),8773 },
 659             { wxT("copy"),169 },
 660             { wxT("crarr"),8629 },
 661             { wxT("cup"),8746 },
 662             { wxT("curren"),164 },
 663             { wxT("dArr"),8659 },
 664             { wxT("dagger"),8224 },
 665             { wxT("darr"),8595 },
 666             { wxT("deg"),176 },
 667             { wxT("delta"),948 },
 668             { wxT("diams"),9830 },
 669             { wxT("divide"),247 },
 670             { wxT("eacute"),233 },
 671             { wxT("ecirc"),234 },
 672             { wxT("egrave"),232 },
 673             { wxT("empty"),8709 },
 674             { wxT("emsp"),8195 },
 675             { wxT("ensp"),8194 },
 676             { wxT("epsilon"),949 },
 677             { wxT("equiv"),8801 },
 678             { wxT("eta"),951 },
 679             { wxT("eth"),240 },
 680             { wxT("euml"),235 },
 681             { wxT("euro"),8364 },
 682             { wxT("exist"),8707 },
 683             { wxT("fnof"),402 },
 684             { wxT("forall"),8704 },
 685             { wxT("frac12"),189 },
 686             { wxT("frac14"),188 },
 687             { wxT("frac34"),190 },
 688             { wxT("frasl"),8260 },
 689             { wxT("gamma"),947 },
 690             { wxT("ge"),8805 },
 691             { wxT("gt"),62 },
 692             { wxT("hArr"),8660 },
 693             { wxT("harr"),8596 },
 694             { wxT("hearts"),9829 },
 695             { wxT("hellip"),8230 },
 696             { wxT("iacute"),237 },
 697             { wxT("icirc"),238 },
 698             { wxT("iexcl"),161 },
 699             { wxT("igrave"),236 },
 700             { wxT("image"),8465 },
 701             { wxT("infin"),8734 },
 702             { wxT("int"),8747 },
 703             { wxT("iota"),953 },
 704             { wxT("iquest"),191 },
 705             { wxT("isin"),8712 },
 706             { wxT("iuml"),239 },
 707             { wxT("kappa"),954 },
 708             { wxT("lArr"),8656 },
 709             { wxT("lambda"),955 },
 710             { wxT("lang"),9001 },
 711             { wxT("laquo"),171 },
 712             { wxT("larr"),8592 },
 713             { wxT("lceil"),8968 },
 714             { wxT("ldquo"),8220 },
 715             { wxT("le"),8804 },
 716             { wxT("lfloor"),8970 },
 717             { wxT("lowast"),8727 },
 718             { wxT("loz"),9674 },
 719             { wxT("lrm"),8206 },
 720             { wxT("lsaquo"),8249 },
 721             { wxT("lsquo"),8216 },
 722             { wxT("lt"),60 },
 723             { wxT("macr"),175 },
 724             { wxT("mdash"),8212 },
 725             { wxT("micro"),181 },
 726             { wxT("middot"),183 },
 727             { wxT("minus"),8722 },
 728             { wxT("mu"),956 },
 729             { wxT("nabla"),8711 },
 730             { wxT("nbsp"),160 },
 731             { wxT("ndash"),8211 },
 732             { wxT("ne"),8800 },
 733             { wxT("ni"),8715 },
 734             { wxT("not"),172 },
 735             { wxT("notin"),8713 },
 736             { wxT("nsub"),8836 },
 737             { wxT("ntilde"),241 },
 738             { wxT("nu"),957 },
 739             { wxT("oacute"),243 },
 740             { wxT("ocirc"),244 },
 741             { wxT("oelig"),339 },
 742             { wxT("ograve"),242 },
 743             { wxT("oline"),8254 },
 744             { wxT("omega"),969 },
 745             { wxT("omicron"),959 },
 746             { wxT("oplus"),8853 },
 747             { wxT("or"),8744 },
 748             { wxT("ordf"),170 },
 749             { wxT("ordm"),186 },
 750             { wxT("oslash"),248 },
 751             { wxT("otilde"),245 },
 752             { wxT("otimes"),8855 },
 753             { wxT("ouml"),246 },
 754             { wxT("para"),182 },
 755             { wxT("part"),8706 },
 756             { wxT("permil"),8240 },
 757             { wxT("perp"),8869 },
 758             { wxT("phi"),966 },
 759             { wxT("pi"),960 },
 760             { wxT("piv"),982 },
 761             { wxT("plusmn"),177 },
 762             { wxT("pound"),163 },
 763             { wxT("prime"),8242 },
 764             { wxT("prod"),8719 },
 765             { wxT("prop"),8733 },
 766             { wxT("psi"),968 },
 767             { wxT("quot"),34 },
 768             { wxT("rArr"),8658 },
 769             { wxT("radic"),8730 },
 770             { wxT("rang"),9002 },
 771             { wxT("raquo"),187 },
 772             { wxT("rarr"),8594 },
 773             { wxT("rceil"),8969 },
 774             { wxT("rdquo"),8221 },
 775             { wxT("real"),8476 },
 776             { wxT("reg"),174 },
 777             { wxT("rfloor"),8971 },
 778             { wxT("rho"),961 },
 779             { wxT("rlm"),8207 },
 780             { wxT("rsaquo"),8250 },
 781             { wxT("rsquo"),8217 },
 782             { wxT("sbquo"),8218 },
 783             { wxT("scaron"),353 },
 784             { wxT("sdot"),8901 },
 785             { wxT("sect"),167 },
 786             { wxT("shy"),173 },
 787             { wxT("sigma"),963 },
 788             { wxT("sigmaf"),962 },
 789             { wxT("sim"),8764 },
 790             { wxT("spades"),9824 },
 791             { wxT("sub"),8834 },
 792             { wxT("sube"),8838 },
 793             { wxT("sum"),8721 },
 794             { wxT("sup"),8835 },
 795             { wxT("sup1"),185 },
 796             { wxT("sup2"),178 },
 797             { wxT("sup3"),179 },
 798             { wxT("supe"),8839 },
 799             { wxT("szlig"),223 },
 800             { wxT("tau"),964 },
 801             { wxT("there4"),8756 },
 802             { wxT("theta"),952 },
 803             { wxT("thetasym"),977 },
 804             { wxT("thinsp"),8201 },
 805             { wxT("thorn"),254 },
 806             { wxT("tilde"),732 },
 807             { wxT("times"),215 },
 808             { wxT("trade"),8482 },
 809             { wxT("uArr"),8657 },
 810             { wxT("uacute"),250 },
 811             { wxT("uarr"),8593 },
 812             { wxT("ucirc"),251 },
 813             { wxT("ugrave"),249 },
 814             { wxT("uml"),168 },
 815             { wxT("upsih"),978 },
 816             { wxT("upsilon"),965 },
 817             { wxT("uuml"),252 },
 818             { wxT("weierp"),8472 },
 819             { wxT("xi"),958 },
 820             { wxT("yacute"),253 },
 821             { wxT("yen"),165 },
 822             { wxT("yuml"),255 },
 823             { wxT("zeta"),950 },
 824             { wxT("zwj"),8205 },
 825             { wxT("zwnj"),8204 },
 826             {NULL, 0}};
 827         static size_t substitutions_cnt = 0;
 828
 829         if (substitutions_cnt == 0)
 830             while (substitutions[substitutions_cnt].code != 0)
 831                 substitutions_cnt++;
 832
 833         wxHtmlEntityInfo *info = NULL;
 834 #ifdef __WXWINCE__
 835         // bsearch crashes under WinCE for some reason
 836         size_t i;
 837         for (i = 0; i < substitutions_cnt; i++)
 838         {
 839             if (entity == substitutions[i].name)
 840             {
 841                 info = & substitutions[i];
 842                 break;
 843             }
 844         }
 845 #else
 846         info = (wxHtmlEntityInfo*) bsearch(entity.c_str(), substitutions,
 847                                            substitutions_cnt,
 848                                            sizeof(wxHtmlEntityInfo),
 849                                            wxHtmlEntityCompare);
 850 #endif
 851         if (info)
 852             code = info->code;
 853     }
 854
 855     if (code == 0)
 856         return 0;
 857     else
 858         return GetCharForCode(code);
 859 }
 860
 861 wxFSFile *wxHtmlParser::OpenURL(wxHtmlURLType WXUNUSED(type),
 862                                 const wxString& url) const
 863 {
 864     return m_FS ? m_FS->OpenFile(url) : NULL;
 865
 866 }
 867
 868
 869 //-----------------------------------------------------------------------------
 870 // wxHtmlParser::ExtractCharsetInformation
 871 //-----------------------------------------------------------------------------
 872
 873 class wxMetaTagParser : public wxHtmlParser
 874 {
 875 public:
 876     wxMetaTagParser() { }
 877
 878     wxObject* GetProduct() { return NULL; }
 879
 880 protected:
 881     virtual void AddText(const wxChar* WXUNUSED(txt)) {}
 882
 883     DECLARE_NO_COPY_CLASS(wxMetaTagParser)
 884 };
 885
 886 class wxMetaTagHandler : public wxHtmlTagHandler
 887 {
 888 public:
 889     wxMetaTagHandler(wxString *retval) : wxHtmlTagHandler(), m_retval(retval) {}
 890     wxString GetSupportedTags() { return wxT("META,BODY"); }
 891     bool HandleTag(const wxHtmlTag& tag);
 892
 893 private:
 894     wxString *m_retval;
 895
 896     DECLARE_NO_COPY_CLASS(wxMetaTagHandler)
 897 };
 898
 899 bool wxMetaTagHandler::HandleTag(const wxHtmlTag& tag)
 900 {
 901     if (tag.GetName() == _T("BODY"))
 902     {
 903         m_Parser->StopParsing();
 904         return false;
 905     }
 906
 907     if (tag.HasParam(_T("HTTP-EQUIV")) &&
 908         tag.GetParam(_T("HTTP-EQUIV")).IsSameAs(_T("Content-Type"), false) &&
 909         tag.HasParam(_T("CONTENT")))
 910     {
 911         wxString content = tag.GetParam(_T("CONTENT")).Lower();
 912         if (content.Left(19) == _T("text/html; charset="))
 913         {
 914             *m_retval = content.Mid(19);
 915             m_Parser->StopParsing();
 916         }
 917     }
 918     return false;
 919 }
 920
 921
 922 /*static*/
 923 wxString wxHtmlParser::ExtractCharsetInformation(const wxString& markup)
 924 {
 925     wxString charset;
 926     wxMetaTagParser *parser = new wxMetaTagParser();
 927     if(parser)
 928     {
 929         parser->AddTagHandler(new wxMetaTagHandler(&charset));
 930         parser->Parse(markup);
 931         delete parser;
 932     }
 933     return charset;
 934 }
 935
 936 /* static */
 937 bool
 938 wxHtmlParser::SkipCommentTag(wxString::const_iterator& start,
 939                              wxString::const_iterator end)
 940 {
 941     wxASSERT_MSG( *start == '<', _T("should be called on the tag start") );
 942
 943     wxString::const_iterator p = start;
 944
 945     // comments begin with "<!--" in HTML 4.0
 946     if ( end - p < 3 || *++p != '!' || *++p != '-' || *++p != '-' )
 947     {
 948         // not a comment at all
 949         return false;
 950     }
 951
 952     // skip the start of the comment tag in any case, if we don't find the
 953     // closing tag we should ignore broken markup
 954     start = p;
 955
 956     // comments end with "--[ \t\r\n]*>", i.e. white space is allowed between
 957     // comment delimiter and the closing tag character (section 3.2.4 of
 958     // http://www.w3.org/TR/html401/)
 959     int dashes = 0;
 960     while ( ++p < end )
 961     {
 962         const wxChar c = *p;
 963
 964         if ( (c == wxT(' ') || c == wxT('\n') ||
 965               c == wxT('\r') || c == wxT('\t')) && dashes >= 2 )
 966         {
 967             // ignore white space before potential tag end
 968             continue;
 969         }
 970
 971         if ( c == wxT('>') && dashes >= 2 )
 972         {
 973             // found end of comment
 974             start = p;
 975             break;
 976         }
 977
 978         if ( c == wxT('-') )
 979             dashes++;
 980         else
 981             dashes = 0;
 982     }
 983
 984     return true;
 985 }
 986
 987 #endif // wxUSE_HTML