src/html/helpdata.cpp

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        helpdata.cpp
   3 // Purpose:     wxHtmlHelpData
   4 // Notes:       Based on htmlhelp.cpp, implementing a monolithic
   5 //              HTML Help controller class,  by Vaclav Slavik
   6 // Author:      Harm van der Heijden and Vaclav Slavik
   7 // RCS-ID:      $Id$
   8 // Copyright:   (c) Harm van der Heijden and Vaclav Slavik
   9 // Licence:     wxWindows licence
  10 /////////////////////////////////////////////////////////////////////////////
  11
  12 #if defined(__GNUG__) && !defined(NO_GCC_PRAGMA)
  13 #pragma implementation "helpdata.h"
  14 #endif
  15
  16 // For compilers that support precompilation, includes "wx.h".
  17 #include "wx/wxprec.h"
  18
  19 #ifdef __BORLANDC__
  20 #pragma hdrstop
  21 #endif
  22
  23 #include "wx/defs.h"
  24
  25 #if wxUSE_HTML && wxUSE_STREAMS
  26
  27 #ifndef WXPRECOMP
  28     #include "wx/intl.h"
  29     #include "wx/log.h"
  30 #endif
  31
  32 #include <ctype.h>
  33
  34 #include "wx/html/helpdata.h"
  35 #include "wx/tokenzr.h"
  36 #include "wx/wfstream.h"
  37 #include "wx/busyinfo.h"
  38 #include "wx/encconv.h"
  39 #include "wx/fontmap.h"
  40 #include "wx/log.h"
  41 #include "wx/html/htmlpars.h"
  42 #include "wx/html/htmldefs.h"
  43 #include "wx/html/htmlfilt.h"
  44 #include "wx/filename.h"
  45
  46 #include "wx/arrimpl.cpp"
  47 WX_DEFINE_OBJARRAY(wxHtmlBookRecArray)
  48 WX_DEFINE_OBJARRAY(wxHtmlHelpDataItems)
  49
  50 //-----------------------------------------------------------------------------
  51 // static helper functions
  52 //-----------------------------------------------------------------------------
  53
  54 // Reads one line, stores it into buf and returns pointer to new line or NULL.
  55 static const wxChar* ReadLine(const wxChar *line, wxChar *buf, size_t bufsize)
  56 {
  57     wxChar *writeptr = buf;
  58     wxChar *endptr = buf + bufsize - 1;
  59     const wxChar *readptr = line;
  60
  61     while (*readptr != 0 && *readptr != _T('\r') && *readptr != _T('\n') &&
  62            writeptr != endptr)
  63         *(writeptr++) = *(readptr++);
  64     *writeptr = 0;
  65     while (*readptr == _T('\r') || *readptr == _T('\n'))
  66         readptr++;
  67     if (*readptr == 0)
  68         return NULL;
  69     else
  70         return readptr;
  71 }
  72
  73
  74
  75 static int
  76 wxHtmlHelpIndexCompareFunc(wxHtmlHelpDataItem **a, wxHtmlHelpDataItem **b)
  77 {
  78     wxHtmlHelpDataItem *ia = *a;
  79     wxHtmlHelpDataItem *ib = *b;
  80
  81     if (ia == NULL)
  82         return -1;
  83     if (ib == NULL)
  84         return 1;
  85
  86     if (ia->parent == ib->parent)
  87     {
  88         return ia->name.CmpNoCase(ib->name);
  89     }
  90     else if (ia->level == ib->level)
  91     {
  92         return wxHtmlHelpIndexCompareFunc(&ia->parent, &ib->parent);
  93     }
  94     else
  95     {
  96         wxHtmlHelpDataItem *ia2 = ia;
  97         wxHtmlHelpDataItem *ib2 = ib;
  98
  99         while (ia2->level > ib2->level)
 100         {
 101             ia2 = ia2->parent;
 102         }
 103         while (ib2->level > ia2->level)
 104         {
 105             ib2 = ib2->parent;
 106         }
 107
 108         wxASSERT(ia2);
 109         wxASSERT(ib2);
 110         int res = wxHtmlHelpIndexCompareFunc(&ia2, &ib2);
 111         if (res != 0)
 112             return res;
 113         else if (ia->level > ib->level)
 114             return 1;
 115         else
 116             return -1;
 117     }
 118 }
 119
 120 //-----------------------------------------------------------------------------
 121 // HP_Parser
 122 //-----------------------------------------------------------------------------
 123
 124 class HP_Parser : public wxHtmlParser
 125 {
 126 public:
 127     HP_Parser()
 128     {
 129         GetEntitiesParser()->SetEncoding(wxFONTENCODING_ISO8859_1);
 130     }
 131
 132     wxObject* GetProduct() { return NULL; }
 133
 134 protected:
 135     virtual void AddText(const wxChar* WXUNUSED(txt)) {}
 136
 137     DECLARE_NO_COPY_CLASS(HP_Parser)
 138 };
 139
 140
 141 //-----------------------------------------------------------------------------
 142 // HP_TagHandler
 143 //-----------------------------------------------------------------------------
 144
 145 class HP_TagHandler : public wxHtmlTagHandler
 146 {
 147     private:
 148         wxString m_name, m_page;
 149         int m_level;
 150         int m_id;
 151         int m_index;
 152         int m_count;
 153         wxHtmlHelpDataItem *m_parentItem;
 154         wxHtmlBookRecord *m_book;
 155
 156         wxHtmlHelpDataItems *m_data;
 157
 158     public:
 159         HP_TagHandler(wxHtmlBookRecord *b) : wxHtmlTagHandler()
 160         {
 161             m_data = NULL;
 162             m_book = b;
 163             m_name = m_page = wxEmptyString;
 164             m_level = 0;
 165             m_id = wxID_ANY;
 166             m_count = 0;
 167             m_parentItem = NULL;
 168         }
 169         wxString GetSupportedTags() { return wxT("UL,OBJECT,PARAM"); }
 170         bool HandleTag(const wxHtmlTag& tag);
 171
 172         void Reset(wxHtmlHelpDataItems& data)
 173         {
 174             m_data = &data;
 175             m_count = 0;
 176             m_level = 0;
 177             m_parentItem = NULL;
 178         }
 179
 180     DECLARE_NO_COPY_CLASS(HP_TagHandler)
 181 };
 182
 183
 184 bool HP_TagHandler::HandleTag(const wxHtmlTag& tag)
 185 {
 186     if (tag.GetName() == wxT("UL"))
 187     {
 188         wxHtmlHelpDataItem *oldparent = m_parentItem;
 189         m_level++;
 190         m_parentItem = (m_count > 0) ? &(*m_data)[m_data->size()-1] : NULL;
 191         ParseInner(tag);
 192         m_level--;
 193         m_parentItem = oldparent;
 194         return true;
 195     }
 196     else if (tag.GetName() == wxT("OBJECT"))
 197     {
 198         m_name = m_page = wxEmptyString;
 199         ParseInner(tag);
 200
 201 #if 0
 202          if (!page.IsEmpty())
 203         /* Valid HHW's file may contain only two object tags:
 204
 205            <OBJECT type="text/site properties">
 206                <param name="ImageType" value="Folder">
 207            </OBJECT>
 208
 209            or
 210
 211            <OBJECT type="text/sitemap">
 212                <param name="Name" value="main page">
 213                <param name="Local" value="another.htm">
 214            </OBJECT>
 215
 216            We're interested in the latter. !page.IsEmpty() is valid
 217            condition because text/site properties does not contain Local param
 218         */
 219 #endif
 220         if (tag.GetParam(wxT("TYPE")) == wxT("text/sitemap"))
 221         {
 222             wxHtmlHelpDataItem *item = new wxHtmlHelpDataItem();
 223             item->parent = m_parentItem;
 224             item->level = m_level;
 225             item->id = m_id;
 226             item->page = m_page;
 227             item->name = m_name;
 228
 229             item->book = m_book;
 230             m_data->Add(item);
 231             m_count++;
 232         }
 233
 234         return true;
 235     }
 236     else
 237     { // "PARAM"
 238         if (m_name.empty() && tag.GetParam(wxT("NAME")) == wxT("Name"))
 239             m_name = tag.GetParam(wxT("VALUE"));
 240         if (tag.GetParam(wxT("NAME")) == wxT("Local"))
 241             m_page = tag.GetParam(wxT("VALUE"));
 242         if (tag.GetParam(wxT("NAME")) == wxT("ID"))
 243             tag.GetParamAsInt(wxT("VALUE"), &m_id);
 244         return false;
 245     }
 246 }
 247
 248
 249 //-----------------------------------------------------------------------------
 250 // wxHtmlHelpData
 251 //-----------------------------------------------------------------------------
 252
 253 wxString wxHtmlBookRecord::GetFullPath(const wxString &page) const
 254 {
 255     if (wxIsAbsolutePath(page))
 256         return page;
 257     else
 258         return m_BasePath + page;
 259 }
 260
 261 wxString wxHtmlHelpDataItem::GetIndentedName() const
 262 {
 263     wxString s;
 264     for (int i = 1; i < level; i++)
 265         s << _T("   ");
 266     s << name;
 267     return s;
 268 }
 269
 270
 271 IMPLEMENT_DYNAMIC_CLASS(wxHtmlHelpData, wxObject)
 272
 273 wxHtmlHelpData::wxHtmlHelpData()
 274 {
 275 #if WXWIN_COMPATIBILITY_2_4
 276     m_cacheContents = NULL;
 277     m_cacheIndex = NULL;
 278 #endif
 279 }
 280
 281 wxHtmlHelpData::~wxHtmlHelpData()
 282 {
 283 #if WXWIN_COMPATIBILITY_2_4
 284     CleanCompatibilityData();
 285 #endif
 286 }
 287
 288 bool wxHtmlHelpData::LoadMSProject(wxHtmlBookRecord *book, wxFileSystem& fsys,
 289                                    const wxString& indexfile,
 290                                    const wxString& contentsfile)
 291 {
 292     wxFSFile *f;
 293     wxHtmlFilterHTML filter;
 294     wxString buf;
 295     wxString string;
 296
 297     HP_Parser parser;
 298     HP_TagHandler *handler = new HP_TagHandler(book);
 299     parser.AddTagHandler(handler);
 300
 301     f = ( contentsfile.IsEmpty() ? (wxFSFile*) NULL : fsys.OpenFile(contentsfile) );
 302     if (f)
 303     {
 304         buf.clear();
 305         buf = filter.ReadFile(*f);
 306         delete f;
 307         handler->Reset(m_contents);
 308         parser.Parse(buf);
 309     }
 310     else
 311     {
 312         wxLogError(_("Cannot open contents file: %s"), contentsfile.c_str());
 313     }
 314
 315     f = ( indexfile.IsEmpty() ? (wxFSFile*) NULL : fsys.OpenFile(indexfile) );
 316     if (f)
 317     {
 318         buf.clear();
 319         buf = filter.ReadFile(*f);
 320         delete f;
 321         handler->Reset(m_index);
 322         parser.Parse(buf);
 323     }
 324     else if (!indexfile.IsEmpty())
 325     {
 326         wxLogError(_("Cannot open index file: %s"), indexfile.c_str());
 327     }
 328     return true;
 329 }
 330
 331 inline static void CacheWriteInt32(wxOutputStream *f, wxInt32 value)
 332 {
 333     wxInt32 x = wxINT32_SWAP_ON_BE(value);
 334     f->Write(&x, sizeof(x));
 335 }
 336
 337 inline static wxInt32 CacheReadInt32(wxInputStream *f)
 338 {
 339     wxInt32 x;
 340     f->Read(&x, sizeof(x));
 341     return wxINT32_SWAP_ON_BE(x);
 342 }
 343
 344 inline static void CacheWriteString(wxOutputStream *f, const wxString& str)
 345 {
 346     const wxWX2MBbuf mbstr = str.mb_str(wxConvUTF8);
 347     size_t len = strlen((const char*)mbstr)+1;
 348     CacheWriteInt32(f, len);
 349     f->Write((const char*)mbstr, len);
 350 }
 351
 352 inline static wxString CacheReadString(wxInputStream *f)
 353 {
 354     size_t len = (size_t)CacheReadInt32(f);
 355     wxCharBuffer str(len-1);
 356     f->Read(str.data(), len);
 357     return wxString(str, wxConvUTF8);
 358 }
 359
 360 #define CURRENT_CACHED_BOOK_VERSION     5
 361
 362 // Additional flags to detect incompatibilities of the runtime environment:
 363 #define CACHED_BOOK_FORMAT_FLAGS \
 364                      (wxUSE_UNICODE << 0)
 365
 366
 367 bool wxHtmlHelpData::LoadCachedBook(wxHtmlBookRecord *book, wxInputStream *f)
 368 {
 369     int i, st, newsize;
 370     wxInt32 version;
 371
 372     /* load header - version info : */
 373     version = CacheReadInt32(f);
 374
 375     if (version != CURRENT_CACHED_BOOK_VERSION)
 376     {
 377         // NB: We can just silently return false here and don't worry about
 378         //     it anymore, because AddBookParam will load the MS project in
 379         //     absence of (properly versioned) .cached file and automatically
 380         //     create new .cached file immediately afterward.
 381         return false;
 382     }
 383
 384     if (CacheReadInt32(f) != CACHED_BOOK_FORMAT_FLAGS)
 385         return false;
 386
 387     /* load contents : */
 388     st = m_contents.size();
 389     newsize = st + CacheReadInt32(f);
 390     m_contents.Alloc(newsize);
 391     for (i = st; i < newsize; i++)
 392     {
 393         wxHtmlHelpDataItem *item = new wxHtmlHelpDataItem;
 394         item->level = CacheReadInt32(f);
 395         item->id = CacheReadInt32(f);
 396         item->name = CacheReadString(f);
 397         item->page = CacheReadString(f);
 398         item->book = book;
 399         m_contents.Add(item);
 400     }
 401
 402     /* load index : */
 403     st = m_index.size();
 404     newsize = st + CacheReadInt32(f);
 405     m_index.Alloc(newsize);
 406     for (i = st; i < newsize; i++)
 407     {
 408         wxHtmlHelpDataItem *item = new wxHtmlHelpDataItem;
 409         item->name = CacheReadString(f);
 410         item->page = CacheReadString(f);
 411         item->level = CacheReadInt32(f);
 412         item->book = book;
 413         int parentShift = CacheReadInt32(f);
 414         if (parentShift != 0)
 415             item->parent = &m_index[m_index.size() - parentShift];
 416         m_index.Add(item);
 417     }
 418     return true;
 419 }
 420
 421
 422 bool wxHtmlHelpData::SaveCachedBook(wxHtmlBookRecord *book, wxOutputStream *f)
 423 {
 424     int i;
 425     wxInt32 cnt;
 426
 427     /* save header - version info : */
 428     CacheWriteInt32(f, CURRENT_CACHED_BOOK_VERSION);
 429     CacheWriteInt32(f, CACHED_BOOK_FORMAT_FLAGS);
 430
 431     /* save contents : */
 432     int len = m_contents.size();
 433     for (cnt = 0, i = 0; i < len; i++)
 434         if (m_contents[i].book == book && m_contents[i].level > 0)
 435             cnt++;
 436     CacheWriteInt32(f, cnt);
 437
 438     for (i = 0; i < len; i++)
 439     {
 440         if (m_contents[i].book != book || m_contents[i].level == 0)
 441             continue;
 442         CacheWriteInt32(f, m_contents[i].level);
 443         CacheWriteInt32(f, m_contents[i].id);
 444         CacheWriteString(f, m_contents[i].name);
 445         CacheWriteString(f, m_contents[i].page);
 446     }
 447
 448     /* save index : */
 449     len = m_index.size();
 450     for (cnt = 0, i = 0; i < len; i++)
 451         if (m_index[i].book == book && m_index[i].level > 0)
 452             cnt++;
 453     CacheWriteInt32(f, cnt);
 454
 455     for (i = 0; i < len; i++)
 456     {
 457         if (m_index[i].book != book || m_index[i].level == 0)
 458             continue;
 459         CacheWriteString(f, m_index[i].name);
 460         CacheWriteString(f, m_index[i].page);
 461         CacheWriteInt32(f, m_index[i].level);
 462         // save distance to parent item, if any:
 463         if (m_index[i].parent == NULL)
 464         {
 465             CacheWriteInt32(f, 0);
 466         }
 467         else
 468         {
 469             int cnt = 0;
 470             wxHtmlHelpDataItem *parent = m_index[i].parent;
 471             for (int j = i-1; j >= 0; j--)
 472             {
 473                 if (m_index[j].book == book && m_index[j].level > 0)
 474                     cnt++;
 475                 if (&m_index[j] == parent)
 476                     break;
 477             }
 478             wxASSERT(cnt > 0);
 479             CacheWriteInt32(f, cnt);
 480         }
 481     }
 482     return true;
 483 }
 484
 485
 486 void wxHtmlHelpData::SetTempDir(const wxString& path)
 487 {
 488     if (path.empty())
 489         m_tempPath = path;
 490     else
 491     {
 492         if (wxIsAbsolutePath(path)) m_tempPath = path;
 493         else m_tempPath = wxGetCwd() + _T("/") + path;
 494
 495         if (m_tempPath[m_tempPath.Length() - 1] != _T('/'))
 496             m_tempPath << _T('/');
 497     }
 498 }
 499
 500
 501
 502 static wxString SafeFileName(const wxString& s)
 503 {
 504     wxString res(s);
 505     res.Replace(wxT("#"), wxT("_"));
 506     res.Replace(wxT(":"), wxT("_"));
 507     res.Replace(wxT("\\"), wxT("_"));
 508     res.Replace(wxT("/"), wxT("_"));
 509     return res;
 510 }
 511
 512 bool wxHtmlHelpData::AddBookParam(const wxFSFile& bookfile,
 513                                   wxFontEncoding encoding,
 514                                   const wxString& title, const wxString& contfile,
 515                                   const wxString& indexfile, const wxString& deftopic,
 516                                   const wxString& path)
 517 {
 518     wxFileSystem fsys;
 519     wxFSFile *fi;
 520     wxHtmlBookRecord *bookr;
 521
 522     int IndexOld = m_index.size(),
 523         ContentsOld = m_contents.size();
 524
 525     if (!path.IsEmpty())
 526         fsys.ChangePathTo(path, true);
 527
 528     size_t booksCnt = m_bookRecords.GetCount();
 529     for (size_t i = 0; i < booksCnt; i++)
 530     {
 531         if ( m_bookRecords[i].GetBookFile() == bookfile.GetLocation() )
 532             return true; // book is (was) loaded
 533     }
 534
 535     bookr = new wxHtmlBookRecord(bookfile.GetLocation(), fsys.GetPath(), title, deftopic);
 536
 537     wxHtmlHelpDataItem *bookitem = new wxHtmlHelpDataItem;
 538     bookitem->level = 0;
 539     bookitem->id = 0;
 540     bookitem->page = deftopic;
 541     bookitem->name = title;
 542     bookitem->book = bookr;
 543
 544     // store the contents index for later
 545     int cont_start = m_contents.size();
 546
 547     m_contents.Add(bookitem);
 548
 549     // Try to find cached binary versions:
 550     // 1. save file as book, but with .hhp.cached extension
 551     // 2. same as 1. but in temp path
 552     // 3. otherwise or if cache load failed, load it from MS.
 553
 554     fi = fsys.OpenFile(bookfile.GetLocation() + wxT(".cached"));
 555
 556     if (fi == NULL ||
 557 #if wxUSE_DATETIME
 558           fi->GetModificationTime() < bookfile.GetModificationTime() ||
 559 #endif // wxUSE_DATETIME
 560           !LoadCachedBook(bookr, fi->GetStream()))
 561     {
 562         if (fi != NULL) delete fi;
 563         fi = fsys.OpenFile(m_tempPath + wxFileNameFromPath(bookfile.GetLocation()) + wxT(".cached"));
 564         if (m_tempPath.empty() || fi == NULL ||
 565 #if wxUSE_DATETIME
 566             fi->GetModificationTime() < bookfile.GetModificationTime() ||
 567 #endif // wxUSE_DATETIME
 568             !LoadCachedBook(bookr, fi->GetStream()))
 569         {
 570             LoadMSProject(bookr, fsys, indexfile, contfile);
 571             if (!m_tempPath.empty())
 572             {
 573                 wxFileOutputStream *outs = new wxFileOutputStream(m_tempPath +
 574                                                   SafeFileName(wxFileNameFromPath(bookfile.GetLocation())) + wxT(".cached"));
 575                 SaveCachedBook(bookr, outs);
 576                 delete outs;
 577             }
 578         }
 579     }
 580
 581     if (fi != NULL) delete fi;
 582
 583     // Now store the contents range
 584     bookr->SetContentsRange(cont_start, m_contents.size());
 585
 586 #if wxUSE_WCHAR_T
 587     // MS HTML Help files [written by MS HTML Help Workshop] are broken
 588     // in that the data are iso-8859-1 (including HTML entities), but must
 589     // be interpreted as being in language's windows charset. Correct the
 590     // differences here and also convert to wxConvLocal in ANSI build
 591     if (encoding != wxFONTENCODING_SYSTEM)
 592     {
 593         #if wxUSE_UNICODE
 594             #define CORRECT_STR(str, conv) \
 595                 str = wxString((str).mb_str(wxConvISO8859_1), conv)
 596         #else
 597             #define CORRECT_STR(str, conv) \
 598                 str = wxString((str).wc_str(conv), wxConvLocal)
 599         #endif
 600         wxCSConv conv(encoding);
 601         size_t IndexCnt = m_index.size();
 602         size_t ContentsCnt = m_contents.size();
 603         size_t i;
 604         for (i = IndexOld; i < IndexCnt; i++)
 605         {
 606             CORRECT_STR(m_index[i].name, conv);
 607         }
 608         for (i = ContentsOld; i < ContentsCnt; i++)
 609         {
 610             CORRECT_STR(m_contents[i].name, conv);
 611         }
 612         #undef CORRECT_STR
 613     }
 614 #else
 615     wxUnusedVar(IndexOld);
 616     wxUnusedVar(ContentsOld);
 617     wxASSERT_MSG(encoding == wxFONTENCODING_SYSTEM, wxT("Help files need charset conversion, but wxUSE_WCHAR_T is 0"));
 618 #endif // wxUSE_WCHAR_T/!wxUSE_WCHAR_T
 619
 620     m_bookRecords.Add(bookr);
 621     if (!m_index.empty())
 622     {
 623         m_index.Sort(wxHtmlHelpIndexCompareFunc);
 624     }
 625
 626     return true;
 627 }
 628
 629
 630 bool wxHtmlHelpData::AddBook(const wxString& book)
 631 {
 632     wxString extension(book.Right(4).Lower());
 633     if (extension == wxT(".zip") ||
 634 #if wxUSE_LIBMSPACK
 635         extension == wxT(".chm") /*compressed html help book*/ ||
 636 #endif
 637         extension == wxT(".htb") /*html book*/)
 638     {
 639         wxFileSystem fsys;
 640         wxString s;
 641         bool rt = false;
 642
 643 #if wxUSE_LIBMSPACK
 644         if (extension == wxT(".chm"))
 645             s = fsys.FindFirst(book + wxT("#chm:*.hhp"), wxFILE);
 646         else
 647 #endif
 648             s = fsys.FindFirst(book + wxT("#zip:*.hhp"), wxFILE);
 649
 650         while (!s.IsEmpty())
 651         {
 652             if (AddBook(s)) rt = true;
 653             s = fsys.FindNext();
 654         }
 655
 656         return rt;
 657     }
 658
 659     wxFSFile *fi;
 660     wxFileSystem fsys;
 661
 662     wxString title = _("noname"),
 663              safetitle,
 664              start = wxEmptyString,
 665              contents = wxEmptyString,
 666              index = wxEmptyString,
 667              charset = wxEmptyString;
 668
 669     fi = fsys.OpenFile(book);
 670     if (fi == NULL)
 671     {
 672         wxLogError(_("Cannot open HTML help book: %s"), book.c_str());
 673         return false;
 674     }
 675     fsys.ChangePathTo(book);
 676
 677     const wxChar *lineptr;
 678     wxChar linebuf[300];
 679     wxString tmp;
 680     wxHtmlFilterPlainText filter;
 681     tmp = filter.ReadFile(*fi);
 682     lineptr = tmp.c_str();
 683
 684     do
 685     {
 686         lineptr = ReadLine(lineptr, linebuf, 300);
 687
 688         for (wxChar *ch = linebuf; *ch != wxT('\0') && *ch != wxT('='); ch++)
 689            *ch = tolower(*ch);
 690
 691         if (wxStrstr(linebuf, _T("title=")) == linebuf)
 692             title = linebuf + wxStrlen(_T("title="));
 693         if (wxStrstr(linebuf, _T("default topic=")) == linebuf)
 694             start = linebuf + wxStrlen(_T("default topic="));
 695         if (wxStrstr(linebuf, _T("index file=")) == linebuf)
 696             index = linebuf + wxStrlen(_T("index file="));
 697         if (wxStrstr(linebuf, _T("contents file=")) == linebuf)
 698             contents = linebuf + wxStrlen(_T("contents file="));
 699         if (wxStrstr(linebuf, _T("charset=")) == linebuf)
 700             charset = linebuf + wxStrlen(_T("charset="));
 701     } while (lineptr != NULL);
 702
 703     wxFontEncoding enc;
 704     if (charset == wxEmptyString) enc = wxFONTENCODING_SYSTEM;
 705     else enc = wxFontMapper::Get()->CharsetToEncoding(charset);
 706
 707     bool rtval = AddBookParam(*fi, enc,
 708                               title, contents, index, start, fsys.GetPath());
 709     delete fi;
 710
 711 #if WXWIN_COMPATIBILITY_2_4
 712     CleanCompatibilityData();
 713 #endif
 714
 715     return rtval;
 716 }
 717
 718 wxString wxHtmlHelpData::FindPageByName(const wxString& x)
 719 {
 720     int cnt;
 721     int i;
 722     wxFileSystem fsys;
 723     wxFSFile *f;
 724
 725     /* 1. try to open given file: */
 726
 727     cnt = m_bookRecords.GetCount();
 728     for (i = 0; i < cnt; i++)
 729     {
 730         f = fsys.OpenFile(m_bookRecords[i].GetFullPath(x));
 731         if (f)
 732         {
 733             wxString url = m_bookRecords[i].GetFullPath(x);
 734             delete f;
 735             return url;
 736         }
 737     }
 738
 739
 740     /* 2. try to find a book: */
 741
 742     for (i = 0; i < cnt; i++)
 743     {
 744         if (m_bookRecords[i].GetTitle() == x)
 745             return m_bookRecords[i].GetFullPath(m_bookRecords[i].GetStart());
 746     }
 747
 748     /* 3. try to find in contents: */
 749
 750     cnt = m_contents.size();
 751     for (i = 0; i < cnt; i++)
 752     {
 753         if (m_contents[i].name == x)
 754             return m_contents[i].GetFullPath();
 755     }
 756
 757
 758     /* 4. try to find in index: */
 759
 760     cnt = m_index.size();
 761     for (i = 0; i < cnt; i++)
 762     {
 763         if (m_index[i].name == x)
 764             return m_index[i].GetFullPath();
 765     }
 766
 767     return wxEmptyString;
 768 }
 769
 770 wxString wxHtmlHelpData::FindPageById(int id)
 771 {
 772     size_t cnt = m_contents.size();
 773     for (size_t i = 0; i < cnt; i++)
 774     {
 775         if (m_contents[i].id == id)
 776         {
 777             return m_contents[i].GetFullPath();
 778         }
 779     }
 780
 781     return wxEmptyString;
 782 }
 783
 784 #if WXWIN_COMPATIBILITY_2_4
 785 wxHtmlContentsItem::wxHtmlContentsItem()
 786     : m_Level(0), m_ID(wxID_ANY), m_Name(NULL), m_Page(NULL), m_Book(NULL),
 787       m_autofree(false)
 788 {
 789 }
 790
 791 wxHtmlContentsItem::wxHtmlContentsItem(const wxHtmlHelpDataItem& d)
 792 {
 793     m_autofree = true;
 794     m_Level = d.level;
 795     m_ID = d.id;
 796     m_Name = wxStrdup(d.name.c_str());
 797     m_Page = wxStrdup(d.page.c_str());
 798     m_Book = d.book;
 799 }
 800
 801 wxHtmlContentsItem& wxHtmlContentsItem::operator=(const wxHtmlContentsItem& d)
 802 {
 803     if (m_autofree)
 804     {
 805         free(m_Name);
 806         free(m_Page);
 807     }
 808     m_autofree = true;
 809     m_Level = d.m_Level;
 810     m_ID = d.m_ID;
 811     m_Name = d.m_Name ? wxStrdup(d.m_Name) : NULL;
 812     m_Page = d.m_Page ? wxStrdup(d.m_Page) : NULL;
 813     m_Book = d.m_Book;
 814     return *this;
 815 }
 816
 817 wxHtmlContentsItem::~wxHtmlContentsItem()
 818 {
 819     if (m_autofree)
 820     {
 821         free(m_Name);
 822         free(m_Page);
 823     }
 824 }
 825
 826 wxHtmlContentsItem* wxHtmlHelpData::GetContents()
 827 {
 828     if (!m_cacheContents && !m_contents.empty())
 829     {
 830         size_t len = m_contents.size();
 831         m_cacheContents = new wxHtmlContentsItem[len];
 832         for (size_t i = 0; i < len; i++)
 833             m_cacheContents[i] = m_contents[i];
 834     }
 835     return m_cacheContents;
 836 }
 837
 838 int wxHtmlHelpData::GetContentsCnt()
 839 {
 840     return m_contents.size();
 841 }
 842
 843 wxHtmlContentsItem* wxHtmlHelpData::GetIndex()
 844 {
 845     if (!m_cacheContents && !m_index.empty())
 846     {
 847         size_t len = m_index.size();
 848         m_cacheContents = new wxHtmlContentsItem[len];
 849         for (size_t i = 0; i < len; i++)
 850             m_cacheContents[i] = m_index[i];
 851     }
 852     return m_cacheContents;
 853 }
 854
 855 int wxHtmlHelpData::GetIndexCnt()
 856 {
 857     return m_index.size();
 858 }
 859
 860 void wxHtmlHelpData::CleanCompatibilityData()
 861 {
 862     delete[] m_cacheContents;
 863     m_cacheContents = NULL;
 864     delete[] m_cacheIndex;
 865     m_cacheIndex = NULL;
 866 }
 867 #endif // WXWIN_COMPATIBILITY_2_4
 868
 869 //----------------------------------------------------------------------------------
 870 // wxHtmlSearchStatus functions
 871 //----------------------------------------------------------------------------------
 872
 873 wxHtmlSearchStatus::wxHtmlSearchStatus(wxHtmlHelpData* data, const wxString& keyword,
 874                                        bool case_sensitive, bool whole_words_only,
 875                                        const wxString& book)
 876 {
 877     m_Data = data;
 878     m_Keyword = keyword;
 879     wxHtmlBookRecord* bookr = NULL;
 880     if (book != wxEmptyString)
 881     {
 882         // we have to search in a specific book. Find it first
 883         int i, cnt = data->m_bookRecords.GetCount();
 884         for (i = 0; i < cnt; i++)
 885             if (data->m_bookRecords[i].GetTitle() == book)
 886             {
 887                 bookr = &(data->m_bookRecords[i]);
 888                 m_CurIndex = bookr->GetContentsStart();
 889                 m_MaxIndex = bookr->GetContentsEnd();
 890                 break;
 891             }
 892         // check; we won't crash if the book doesn't exist, but it's Bad Anyway.
 893         wxASSERT(bookr);
 894     }
 895     if (! bookr)
 896     {
 897         // no book specified; search all books
 898         m_CurIndex = 0;
 899         m_MaxIndex = m_Data->m_contents.size();
 900     }
 901     m_Engine.LookFor(keyword, case_sensitive, whole_words_only);
 902     m_Active = (m_CurIndex < m_MaxIndex);
 903 }
 904
 905 #if WXWIN_COMPATIBILITY_2_4
 906 wxHtmlContentsItem* wxHtmlSearchStatus::GetContentsItem()
 907 {
 908     static wxHtmlContentsItem it;
 909     it = wxHtmlContentsItem(*m_CurItem);
 910     return &it;
 911 }
 912 #endif
 913
 914 bool wxHtmlSearchStatus::Search()
 915 {
 916     wxFSFile *file;
 917     int i = m_CurIndex;  // shortcut
 918     bool found = false;
 919     wxString thepage;
 920
 921     if (!m_Active)
 922     {
 923         // sanity check. Illegal use, but we'll try to prevent a crash anyway
 924         wxASSERT(m_Active);
 925         return false;
 926     }
 927
 928     m_Name = wxEmptyString;
 929     m_CurItem = NULL;
 930     thepage = m_Data->m_contents[i].page;
 931
 932     m_Active = (++m_CurIndex < m_MaxIndex);
 933     // check if it is same page with different anchor:
 934     if (!m_LastPage.empty())
 935     {
 936         const wxChar *p1, *p2;
 937         for (p1 = thepage.c_str(), p2 = m_LastPage.c_str();
 938              *p1 != 0 && *p1 != _T('#') && *p1 == *p2; p1++, p2++) {}
 939
 940         m_LastPage = thepage;
 941
 942         if (*p1 == 0 || *p1 == _T('#'))
 943             return false;
 944     }
 945     else m_LastPage = thepage;
 946
 947     wxFileSystem fsys;
 948     file = fsys.OpenFile(m_Data->m_contents[i].book->GetFullPath(thepage));
 949     if (file)
 950     {
 951         if (m_Engine.Scan(*file))
 952         {
 953             m_Name = m_Data->m_contents[i].name;
 954             m_CurItem = &m_Data->m_contents[i];
 955             found = true;
 956         }
 957         delete file;
 958     }
 959     return found;
 960 }
 961
 962
 963
 964
 965
 966
 967
 968
 969 //--------------------------------------------------------------------------------
 970 // wxHtmlSearchEngine
 971 //--------------------------------------------------------------------------------
 972
 973 void wxHtmlSearchEngine::LookFor(const wxString& keyword, bool case_sensitive, bool whole_words_only)
 974 {
 975     m_CaseSensitive = case_sensitive;
 976     m_WholeWords = whole_words_only;
 977     m_Keyword = keyword;
 978
 979     if (!m_CaseSensitive)
 980         m_Keyword.LowerCase();
 981 }
 982
 983
 984 static inline bool WHITESPACE(wxChar c)
 985 {
 986     return c == _T(' ') || c == _T('\n') || c == _T('\r') || c == _T('\t');
 987 }
 988
 989 bool wxHtmlSearchEngine::Scan(const wxFSFile& file)
 990 {
 991     wxASSERT_MSG(!m_Keyword.empty(), wxT("wxHtmlSearchEngine::LookFor must be called before scanning!"));
 992
 993     int i, j;
 994     int wrd = m_Keyword.Length();
 995     bool found = false;
 996     wxHtmlFilterHTML filter;
 997     wxString tmp = filter.ReadFile(file);
 998     int lng = tmp.length();
 999     const wxChar *buf = tmp.c_str();
1000
1001     if (!m_CaseSensitive)
1002         tmp.LowerCase();
1003
1004     const wxChar *kwd = m_Keyword.c_str();
1005
1006     if (m_WholeWords)
1007     {
1008         for (i = 0; i < lng - wrd; i++)
1009         {
1010             if (WHITESPACE(buf[i])) continue;
1011             j = 0;
1012             while ((j < wrd) && (buf[i + j] == kwd[j])) j++;
1013             if (j == wrd && WHITESPACE(buf[i + j])) { found = true; break; }
1014         }
1015     }
1016
1017     else
1018     {
1019         for (i = 0; i < lng - wrd; i++)
1020         {
1021             j = 0;
1022             while ((j < wrd) && (buf[i + j] == kwd[j])) j++;
1023             if (j == wrd) { found = true; break; }
1024         }
1025     }
1026
1027     return found;
1028 }
1029
1030
1031
1032 #endif