src/html/helpdata.cpp

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        src/html/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 // For compilers that support precompilation, includes "wx.h".
  13 #include "wx/wxprec.h"
  14
  15 #ifdef __BORLANDC__
  16     #pragma hdrstop
  17 #endif
  18
  19 #if wxUSE_HTML && wxUSE_STREAMS
  20
  21 #ifndef WX_PRECOMP
  22     #include "wx/intl.h"
  23     #include "wx/log.h"
  24 #endif
  25
  26 #include <ctype.h>
  27
  28 #include "wx/html/helpdata.h"
  29 #include "wx/tokenzr.h"
  30 #include "wx/wfstream.h"
  31 #include "wx/busyinfo.h"
  32 #include "wx/encconv.h"
  33 #include "wx/fontmap.h"
  34 #include "wx/html/htmlpars.h"
  35 #include "wx/html/htmldefs.h"
  36 #include "wx/html/htmlfilt.h"
  37 #include "wx/filename.h"
  38
  39 #include "wx/arrimpl.cpp"
  40 WX_DEFINE_OBJARRAY(wxHtmlBookRecArray)
  41 WX_DEFINE_OBJARRAY(wxHtmlHelpDataItems)
  42
  43 //-----------------------------------------------------------------------------
  44 // static helper functions
  45 //-----------------------------------------------------------------------------
  46
  47 // Reads one line, stores it into buf and returns pointer to new line or NULL.
  48 static const wxChar* ReadLine(const wxChar *line, wxChar *buf, size_t bufsize)
  49 {
  50     wxChar *writeptr = buf;
  51     wxChar *endptr = buf + bufsize - 1;
  52     const wxChar *readptr = line;
  53
  54     while (*readptr != 0 && *readptr != wxT('\r') && *readptr != wxT('\n') &&
  55            writeptr != endptr)
  56         *(writeptr++) = *(readptr++);
  57     *writeptr = 0;
  58     while (*readptr == wxT('\r') || *readptr == wxT('\n'))
  59         readptr++;
  60     if (*readptr == 0)
  61         return NULL;
  62     else
  63         return readptr;
  64 }
  65
  66
  67
  68 static int
  69 wxHtmlHelpIndexCompareFunc(wxHtmlHelpDataItem **a, wxHtmlHelpDataItem **b)
  70 {
  71     wxHtmlHelpDataItem *ia = *a;
  72     wxHtmlHelpDataItem *ib = *b;
  73
  74     if (ia == NULL)
  75         return -1;
  76     if (ib == NULL)
  77         return 1;
  78
  79     if (ia->parent == ib->parent)
  80     {
  81         return ia->name.CmpNoCase(ib->name);
  82     }
  83     else if (ia->level == ib->level)
  84     {
  85         return wxHtmlHelpIndexCompareFunc(&ia->parent, &ib->parent);
  86     }
  87     else
  88     {
  89         wxHtmlHelpDataItem *ia2 = ia;
  90         wxHtmlHelpDataItem *ib2 = ib;
  91
  92         while (ia2->level > ib2->level)
  93         {
  94             ia2 = ia2->parent;
  95         }
  96         while (ib2->level > ia2->level)
  97         {
  98             ib2 = ib2->parent;
  99         }
 100
 101         wxASSERT(ia2);
 102         wxASSERT(ib2);
 103         int res = wxHtmlHelpIndexCompareFunc(&ia2, &ib2);
 104         if (res != 0)
 105             return res;
 106         else if (ia->level > ib->level)
 107             return 1;
 108         else
 109             return -1;
 110     }
 111 }
 112
 113 //-----------------------------------------------------------------------------
 114 // HP_Parser
 115 //-----------------------------------------------------------------------------
 116
 117 class HP_Parser : public wxHtmlParser
 118 {
 119 public:
 120     HP_Parser()
 121     {
 122         GetEntitiesParser()->SetEncoding(wxFONTENCODING_ISO8859_1);
 123     }
 124
 125     wxObject* GetProduct() { return NULL; }
 126
 127 protected:
 128     virtual void AddText(const wxString& WXUNUSED(txt)) {}
 129
 130     wxDECLARE_NO_COPY_CLASS(HP_Parser);
 131 };
 132
 133
 134 //-----------------------------------------------------------------------------
 135 // HP_TagHandler
 136 //-----------------------------------------------------------------------------
 137
 138 class HP_TagHandler : public wxHtmlTagHandler
 139 {
 140     private:
 141         wxString m_name, m_page;
 142         int m_level;
 143         int m_id;
 144         int m_index;
 145         int m_count;
 146         wxHtmlHelpDataItem *m_parentItem;
 147         wxHtmlBookRecord *m_book;
 148
 149         wxHtmlHelpDataItems *m_data;
 150
 151     public:
 152         HP_TagHandler(wxHtmlBookRecord *b) : wxHtmlTagHandler()
 153         {
 154             m_data = NULL;
 155             m_book = b;
 156             m_name = m_page = wxEmptyString;
 157             m_level = 0;
 158             m_id = wxID_ANY;
 159             m_count = 0;
 160             m_parentItem = NULL;
 161         }
 162         wxString GetSupportedTags() { return wxT("UL,OBJECT,PARAM"); }
 163         bool HandleTag(const wxHtmlTag& tag);
 164
 165         void Reset(wxHtmlHelpDataItems& data)
 166         {
 167             m_data = &data;
 168             m_count = 0;
 169             m_level = 0;
 170             m_parentItem = NULL;
 171         }
 172
 173     wxDECLARE_NO_COPY_CLASS(HP_TagHandler);
 174 };
 175
 176
 177 bool HP_TagHandler::HandleTag(const wxHtmlTag& tag)
 178 {
 179     if (tag.GetName() == wxT("UL"))
 180     {
 181         wxHtmlHelpDataItem *oldparent = m_parentItem;
 182         m_level++;
 183         m_parentItem = (m_count > 0) ? &(*m_data)[m_data->size()-1] : NULL;
 184         ParseInner(tag);
 185         m_level--;
 186         m_parentItem = oldparent;
 187         return true;
 188     }
 189     else if (tag.GetName() == wxT("OBJECT"))
 190     {
 191         m_name = m_page = wxEmptyString;
 192         ParseInner(tag);
 193
 194 #if 0
 195          if (!page.IsEmpty())
 196         /* Valid HHW's file may contain only two object tags:
 197
 198            <OBJECT type="text/site properties">
 199                <param name="ImageType" value="Folder">
 200            </OBJECT>
 201
 202            or
 203
 204            <OBJECT type="text/sitemap">
 205                <param name="Name" value="main page">
 206                <param name="Local" value="another.htm">
 207            </OBJECT>
 208
 209            We're interested in the latter. !page.IsEmpty() is valid
 210            condition because text/site properties does not contain Local param
 211         */
 212 #endif
 213         if (tag.GetParam(wxT("TYPE")) == wxT("text/sitemap"))
 214         {
 215             wxHtmlHelpDataItem *item = new wxHtmlHelpDataItem();
 216             item->parent = m_parentItem;
 217             item->level = m_level;
 218             item->id = m_id;
 219             item->page = m_page;
 220             item->name = m_name;
 221
 222             item->book = m_book;
 223             m_data->Add(item);
 224             m_count++;
 225         }
 226
 227         return true;
 228     }
 229     else
 230     { // "PARAM"
 231         if (m_name.empty() && tag.GetParam(wxT("NAME")) == wxT("Name"))
 232             m_name = tag.GetParam(wxT("VALUE"));
 233         if (tag.GetParam(wxT("NAME")) == wxT("Local"))
 234             m_page = tag.GetParam(wxT("VALUE"));
 235         if (tag.GetParam(wxT("NAME")) == wxT("ID"))
 236             tag.GetParamAsInt(wxT("VALUE"), &m_id);
 237         return false;
 238     }
 239 }
 240
 241
 242 //-----------------------------------------------------------------------------
 243 // wxHtmlHelpData
 244 //-----------------------------------------------------------------------------
 245
 246 wxString wxHtmlBookRecord::GetFullPath(const wxString &page) const
 247 {
 248     if (wxIsAbsolutePath(page))
 249         return page;
 250     else
 251         return m_BasePath + page;
 252 }
 253
 254 wxString wxHtmlHelpDataItem::GetIndentedName() const
 255 {
 256     wxString s;
 257     for (int i = 1; i < level; i++)
 258         s << wxT("   ");
 259     s << name;
 260     return s;
 261 }
 262
 263
 264 IMPLEMENT_DYNAMIC_CLASS(wxHtmlHelpData, wxObject)
 265
 266 wxHtmlHelpData::wxHtmlHelpData()
 267 {
 268 }
 269
 270 wxHtmlHelpData::~wxHtmlHelpData()
 271 {
 272 }
 273
 274 bool wxHtmlHelpData::LoadMSProject(wxHtmlBookRecord *book, wxFileSystem& fsys,
 275                                    const wxString& indexfile,
 276                                    const wxString& contentsfile)
 277 {
 278     wxFSFile *f;
 279     wxHtmlFilterHTML filter;
 280     wxString buf;
 281     wxString string;
 282
 283     HP_Parser parser;
 284     HP_TagHandler *handler = new HP_TagHandler(book);
 285     parser.AddTagHandler(handler);
 286
 287     f = ( contentsfile.empty() ? NULL : fsys.OpenFile(contentsfile) );
 288     if (f)
 289     {
 290         buf.clear();
 291         buf = filter.ReadFile(*f);
 292         delete f;
 293         handler->Reset(m_contents);
 294         parser.Parse(buf);
 295     }
 296     else
 297     {
 298         wxLogError(_("Cannot open contents file: %s"), contentsfile.c_str());
 299     }
 300
 301     f = ( indexfile.empty() ? NULL : fsys.OpenFile(indexfile) );
 302     if (f)
 303     {
 304         buf.clear();
 305         buf = filter.ReadFile(*f);
 306         delete f;
 307         handler->Reset(m_index);
 308         parser.Parse(buf);
 309     }
 310     else if (!indexfile.empty())
 311     {
 312         wxLogError(_("Cannot open index file: %s"), indexfile.c_str());
 313     }
 314     return true;
 315 }
 316
 317 inline static void CacheWriteInt32(wxOutputStream *f, wxInt32 value)
 318 {
 319     wxInt32 x = wxINT32_SWAP_ON_BE(value);
 320     f->Write(&x, sizeof(x));
 321 }
 322
 323 inline static wxInt32 CacheReadInt32(wxInputStream *f)
 324 {
 325     wxInt32 x;
 326     f->Read(&x, sizeof(x));
 327     return wxINT32_SWAP_ON_BE(x);
 328 }
 329
 330 inline static void CacheWriteString(wxOutputStream *f, const wxString& str)
 331 {
 332     const wxWX2MBbuf mbstr = str.mb_str(wxConvUTF8);
 333     size_t len = strlen((const char*)mbstr)+1;
 334     CacheWriteInt32(f, len);
 335     f->Write((const char*)mbstr, len);
 336 }
 337
 338 inline static wxString CacheReadString(wxInputStream *f)
 339 {
 340     size_t len = (size_t)CacheReadInt32(f);
 341     wxCharBuffer str(len-1);
 342     f->Read(str.data(), len);
 343     return wxString(str, wxConvUTF8);
 344 }
 345
 346 #define CURRENT_CACHED_BOOK_VERSION     5
 347
 348 // Additional flags to detect incompatibilities of the runtime environment:
 349 #define CACHED_BOOK_FORMAT_FLAGS \
 350                      (wxUSE_UNICODE << 0)
 351
 352
 353 bool wxHtmlHelpData::LoadCachedBook(wxHtmlBookRecord *book, wxInputStream *f)
 354 {
 355     int i, st, newsize;
 356     wxInt32 version;
 357
 358     /* load header - version info : */
 359     version = CacheReadInt32(f);
 360
 361     if (version != CURRENT_CACHED_BOOK_VERSION)
 362     {
 363         // NB: We can just silently return false here and don't worry about
 364         //     it anymore, because AddBookParam will load the MS project in
 365         //     absence of (properly versioned) .cached file and automatically
 366         //     create new .cached file immediately afterward.
 367         return false;
 368     }
 369
 370     if (CacheReadInt32(f) != CACHED_BOOK_FORMAT_FLAGS)
 371         return false;
 372
 373     /* load contents : */
 374     st = m_contents.size();
 375     newsize = st + CacheReadInt32(f);
 376     m_contents.Alloc(newsize);
 377     for (i = st; i < newsize; i++)
 378     {
 379         wxHtmlHelpDataItem *item = new wxHtmlHelpDataItem;
 380         item->level = CacheReadInt32(f);
 381         item->id = CacheReadInt32(f);
 382         item->name = CacheReadString(f);
 383         item->page = CacheReadString(f);
 384         item->book = book;
 385         m_contents.Add(item);
 386     }
 387
 388     /* load index : */
 389     st = m_index.size();
 390     newsize = st + CacheReadInt32(f);
 391     m_index.Alloc(newsize);
 392     for (i = st; i < newsize; i++)
 393     {
 394         wxHtmlHelpDataItem *item = new wxHtmlHelpDataItem;
 395         item->name = CacheReadString(f);
 396         item->page = CacheReadString(f);
 397         item->level = CacheReadInt32(f);
 398         item->book = book;
 399         int parentShift = CacheReadInt32(f);
 400         if (parentShift != 0)
 401             item->parent = &m_index[m_index.size() - parentShift];
 402         m_index.Add(item);
 403     }
 404     return true;
 405 }
 406
 407
 408 bool wxHtmlHelpData::SaveCachedBook(wxHtmlBookRecord *book, wxOutputStream *f)
 409 {
 410     int i;
 411     wxInt32 cnt;
 412
 413     /* save header - version info : */
 414     CacheWriteInt32(f, CURRENT_CACHED_BOOK_VERSION);
 415     CacheWriteInt32(f, CACHED_BOOK_FORMAT_FLAGS);
 416
 417     /* save contents : */
 418     int len = m_contents.size();
 419     for (cnt = 0, i = 0; i < len; i++)
 420         if (m_contents[i].book == book && m_contents[i].level > 0)
 421             cnt++;
 422     CacheWriteInt32(f, cnt);
 423
 424     for (i = 0; i < len; i++)
 425     {
 426         if (m_contents[i].book != book || m_contents[i].level == 0)
 427             continue;
 428         CacheWriteInt32(f, m_contents[i].level);
 429         CacheWriteInt32(f, m_contents[i].id);
 430         CacheWriteString(f, m_contents[i].name);
 431         CacheWriteString(f, m_contents[i].page);
 432     }
 433
 434     /* save index : */
 435     len = m_index.size();
 436     for (cnt = 0, i = 0; i < len; i++)
 437         if (m_index[i].book == book && m_index[i].level > 0)
 438             cnt++;
 439     CacheWriteInt32(f, cnt);
 440
 441     for (i = 0; i < len; i++)
 442     {
 443         if (m_index[i].book != book || m_index[i].level == 0)
 444             continue;
 445         CacheWriteString(f, m_index[i].name);
 446         CacheWriteString(f, m_index[i].page);
 447         CacheWriteInt32(f, m_index[i].level);
 448         // save distance to parent item, if any:
 449         if (m_index[i].parent == NULL)
 450         {
 451             CacheWriteInt32(f, 0);
 452         }
 453         else
 454         {
 455             int cnt2 = 0;
 456             wxHtmlHelpDataItem *parent = m_index[i].parent;
 457             for (int j = i-1; j >= 0; j--)
 458             {
 459                 if (m_index[j].book == book && m_index[j].level > 0)
 460                     cnt2++;
 461                 if (&m_index[j] == parent)
 462                     break;
 463             }
 464             wxASSERT(cnt2 > 0);
 465             CacheWriteInt32(f, cnt2);
 466         }
 467     }
 468     return true;
 469 }
 470
 471
 472 void wxHtmlHelpData::SetTempDir(const wxString& path)
 473 {
 474     if (path.empty())
 475         m_tempPath = path;
 476     else
 477     {
 478         wxFileName fn(path);
 479         fn.MakeAbsolute();
 480
 481         m_tempPath = fn.GetPath(wxPATH_GET_VOLUME | wxPATH_GET_SEPARATOR);
 482     }
 483 }
 484
 485
 486
 487 static wxString SafeFileName(const wxString& s)
 488 {
 489     wxString res(s);
 490     res.Replace(wxT("#"), wxT("_"));
 491     res.Replace(wxT(":"), wxT("_"));
 492     res.Replace(wxT("\\"), wxT("_"));
 493     res.Replace(wxT("/"), wxT("_"));
 494     return res;
 495 }
 496
 497 bool wxHtmlHelpData::AddBookParam(const wxFSFile& bookfile,
 498                                   wxFontEncoding encoding,
 499                                   const wxString& title, const wxString& contfile,
 500                                   const wxString& indexfile, const wxString& deftopic,
 501                                   const wxString& path)
 502 {
 503     wxFileSystem fsys;
 504     wxFSFile *fi;
 505     wxHtmlBookRecord *bookr;
 506
 507     int IndexOld = m_index.size(),
 508         ContentsOld = m_contents.size();
 509
 510     if (!path.empty())
 511         fsys.ChangePathTo(path, true);
 512
 513     size_t booksCnt = m_bookRecords.GetCount();
 514     for (size_t i = 0; i < booksCnt; i++)
 515     {
 516         if ( m_bookRecords[i].GetBookFile() == bookfile.GetLocation() )
 517             return true; // book is (was) loaded
 518     }
 519
 520     bookr = new wxHtmlBookRecord(bookfile.GetLocation(), fsys.GetPath(), title, deftopic);
 521
 522     wxHtmlHelpDataItem *bookitem = new wxHtmlHelpDataItem;
 523     bookitem->level = 0;
 524     bookitem->id = 0;
 525     bookitem->page = deftopic;
 526     bookitem->name = title;
 527     bookitem->book = bookr;
 528
 529     // store the contents index for later
 530     int cont_start = m_contents.size();
 531
 532     m_contents.Add(bookitem);
 533
 534     // Try to find cached binary versions:
 535     // 1. save file as book, but with .hhp.cached extension
 536     // 2. same as 1. but in temp path
 537     // 3. otherwise or if cache load failed, load it from MS.
 538
 539     fi = fsys.OpenFile(bookfile.GetLocation() + wxT(".cached"));
 540
 541     if (fi == NULL ||
 542 #if wxUSE_DATETIME
 543           fi->GetModificationTime() < bookfile.GetModificationTime() ||
 544 #endif // wxUSE_DATETIME
 545           !LoadCachedBook(bookr, fi->GetStream()))
 546     {
 547         if (fi != NULL) delete fi;
 548         fi = fsys.OpenFile(m_tempPath + wxFileNameFromPath(bookfile.GetLocation()) + wxT(".cached"));
 549         if (m_tempPath.empty() || fi == NULL ||
 550 #if wxUSE_DATETIME
 551             fi->GetModificationTime() < bookfile.GetModificationTime() ||
 552 #endif // wxUSE_DATETIME
 553             !LoadCachedBook(bookr, fi->GetStream()))
 554         {
 555             LoadMSProject(bookr, fsys, indexfile, contfile);
 556             if (!m_tempPath.empty())
 557             {
 558                 wxFileOutputStream *outs = new wxFileOutputStream(m_tempPath +
 559                                                   SafeFileName(wxFileNameFromPath(bookfile.GetLocation())) + wxT(".cached"));
 560                 SaveCachedBook(bookr, outs);
 561                 delete outs;
 562             }
 563         }
 564     }
 565
 566     if (fi != NULL) delete fi;
 567
 568     // Now store the contents range
 569     bookr->SetContentsRange(cont_start, m_contents.size());
 570
 571     // MS HTML Help files [written by MS HTML Help Workshop] are broken
 572     // in that the data are iso-8859-1 (including HTML entities), but must
 573     // be interpreted as being in language's windows charset. Correct the
 574     // differences here and also convert to wxConvLocal in ANSI build
 575     if (encoding != wxFONTENCODING_SYSTEM)
 576     {
 577         #if wxUSE_UNICODE
 578             #define CORRECT_STR(str, conv) \
 579                 str = wxString((str).mb_str(wxConvISO8859_1), conv)
 580         #else
 581             #define CORRECT_STR(str, conv) \
 582                 str = wxString((str).wc_str(conv), wxConvLocal)
 583         #endif
 584         wxCSConv conv(encoding);
 585         size_t IndexCnt = m_index.size();
 586         size_t ContentsCnt = m_contents.size();
 587         size_t i;
 588         for (i = IndexOld; i < IndexCnt; i++)
 589         {
 590             CORRECT_STR(m_index[i].name, conv);
 591         }
 592         for (i = ContentsOld; i < ContentsCnt; i++)
 593         {
 594             CORRECT_STR(m_contents[i].name, conv);
 595         }
 596         #undef CORRECT_STR
 597     }
 598
 599     m_bookRecords.Add(bookr);
 600     if (!m_index.empty())
 601     {
 602         m_index.Sort(wxHtmlHelpIndexCompareFunc);
 603     }
 604
 605     return true;
 606 }
 607
 608
 609 bool wxHtmlHelpData::AddBook(const wxString& book)
 610 {
 611     wxString extension(book.Right(4).Lower());
 612     if (extension == wxT(".zip") ||
 613 #if wxUSE_LIBMSPACK
 614         extension == wxT(".chm") /*compressed html help book*/ ||
 615 #endif
 616         extension == wxT(".htb") /*html book*/)
 617     {
 618         wxFileSystem fsys;
 619         wxString s;
 620         bool rt = false;
 621
 622 #if wxUSE_LIBMSPACK
 623         if (extension == wxT(".chm"))
 624             s = fsys.FindFirst(book + wxT("#chm:*.hhp"), wxFILE);
 625         else
 626 #endif
 627             s = fsys.FindFirst(book + wxT("#zip:*.hhp"), wxFILE);
 628
 629         while (!s.empty())
 630         {
 631             if (AddBook(s)) rt = true;
 632             s = fsys.FindNext();
 633         }
 634
 635         return rt;
 636     }
 637
 638     wxFSFile *fi;
 639     wxFileSystem fsys;
 640
 641     wxString title = _("noname"),
 642              safetitle,
 643              start = wxEmptyString,
 644              contents = wxEmptyString,
 645              index = wxEmptyString,
 646              charset = wxEmptyString;
 647
 648     fi = fsys.OpenFile(book);
 649     if (fi == NULL)
 650     {
 651         wxLogError(_("Cannot open HTML help book: %s"), book.c_str());
 652         return false;
 653     }
 654     fsys.ChangePathTo(book);
 655
 656     const wxChar *lineptr;
 657     wxChar linebuf[300];
 658     wxString tmp;
 659     wxHtmlFilterPlainText filter;
 660     tmp = filter.ReadFile(*fi);
 661     lineptr = tmp.c_str();
 662
 663     do
 664     {
 665         lineptr = ReadLine(lineptr, linebuf, 300);
 666
 667         for (wxChar *ch = linebuf; *ch != wxT('\0') && *ch != wxT('='); ch++)
 668            *ch = (wxChar)wxTolower(*ch);
 669
 670         if (wxStrstr(linebuf, wxT("title=")) == linebuf)
 671             title = linebuf + wxStrlen(wxT("title="));
 672         if (wxStrstr(linebuf, wxT("default topic=")) == linebuf)
 673             start = linebuf + wxStrlen(wxT("default topic="));
 674         if (wxStrstr(linebuf, wxT("index file=")) == linebuf)
 675             index = linebuf + wxStrlen(wxT("index file="));
 676         if (wxStrstr(linebuf, wxT("contents file=")) == linebuf)
 677             contents = linebuf + wxStrlen(wxT("contents file="));
 678         if (wxStrstr(linebuf, wxT("charset=")) == linebuf)
 679             charset = linebuf + wxStrlen(wxT("charset="));
 680     } while (lineptr != NULL);
 681
 682     wxFontEncoding enc = wxFONTENCODING_SYSTEM;
 683 #if wxUSE_FONTMAP
 684     if (charset != wxEmptyString)
 685         enc = wxFontMapper::Get()->CharsetToEncoding(charset);
 686 #endif
 687
 688     // No conversion was done on the title yet; at least
 689     // test for a common case.
 690     if (charset == wxT("utf-8") && !title.IsEmpty())
 691     {
 692         char *buf = new char[title.Length()+1];
 693         size_t i;
 694         for (i = 0; i < title.Length(); i++)
 695             buf[i] = (char) title[i];
 696         buf[i] = 0;
 697         title = wxString::FromUTF8(buf);
 698         delete[] buf;
 699     }
 700
 701     bool rtval = AddBookParam(*fi, enc,
 702                               title, contents, index, start, fsys.GetPath());
 703     delete fi;
 704
 705     return rtval;
 706 }
 707
 708 wxString wxHtmlHelpData::FindPageByName(const wxString& x)
 709 {
 710     int i;
 711
 712     bool has_non_ascii = false;
 713     wxString::const_iterator it;
 714     for (it = x.begin(); it != x.end(); ++it)
 715     {
 716         wxUniChar ch = *it;
 717         if (!ch.IsAscii())
 718         {
 719             has_non_ascii = true;
 720             break;
 721         }
 722     }
 723
 724     int cnt = m_bookRecords.GetCount();
 725
 726     if (!has_non_ascii)
 727     {
 728       wxFileSystem fsys;
 729       wxFSFile *f;
 730       // 1. try to open given file:
 731       for (i = 0; i < cnt; i++)
 732       {
 733         f = fsys.OpenFile(m_bookRecords[i].GetFullPath(x));
 734         if (f)
 735         {
 736             wxString url = m_bookRecords[i].GetFullPath(x);
 737             delete f;
 738             return url;
 739         }
 740       }
 741     }
 742
 743
 744     // 2. try to find a book:
 745     for (i = 0; i < cnt; i++)
 746     {
 747         if (m_bookRecords[i].GetTitle() == x)
 748             return m_bookRecords[i].GetFullPath(m_bookRecords[i].GetStart());
 749     }
 750
 751     // 3. try to find in contents:
 752     cnt = m_contents.size();
 753     for (i = 0; i < cnt; i++)
 754     {
 755         if (m_contents[i].name == x)
 756             return m_contents[i].GetFullPath();
 757     }
 758
 759
 760     // 4. try to find in index:
 761     cnt = m_index.size();
 762     for (i = 0; i < cnt; i++)
 763     {
 764         if (m_index[i].name == x)
 765             return m_index[i].GetFullPath();
 766     }
 767
 768     // 4b. if still not found, try case-insensitive comparison
 769     for (i = 0; i < cnt; i++)
 770     {
 771         if (m_index[i].name.CmpNoCase(x) == 0)
 772             return m_index[i].GetFullPath();
 773     }
 774
 775     return wxEmptyString;
 776 }
 777
 778 wxString wxHtmlHelpData::FindPageById(int id)
 779 {
 780     size_t cnt = m_contents.size();
 781     for (size_t i = 0; i < cnt; i++)
 782     {
 783         if (m_contents[i].id == id)
 784         {
 785             return m_contents[i].GetFullPath();
 786         }
 787     }
 788
 789     return wxEmptyString;
 790 }
 791
 792
 793 //----------------------------------------------------------------------------------
 794 // wxHtmlSearchStatus functions
 795 //----------------------------------------------------------------------------------
 796
 797 wxHtmlSearchStatus::wxHtmlSearchStatus(wxHtmlHelpData* data, const wxString& keyword,
 798                                        bool case_sensitive, bool whole_words_only,
 799                                        const wxString& book)
 800 {
 801     m_Data = data;
 802     m_Keyword = keyword;
 803     wxHtmlBookRecord* bookr = NULL;
 804     if (book != wxEmptyString)
 805     {
 806         // we have to search in a specific book. Find it first
 807         int i, cnt = data->m_bookRecords.GetCount();
 808         for (i = 0; i < cnt; i++)
 809             if (data->m_bookRecords[i].GetTitle() == book)
 810             {
 811                 bookr = &(data->m_bookRecords[i]);
 812                 m_CurIndex = bookr->GetContentsStart();
 813                 m_MaxIndex = bookr->GetContentsEnd();
 814                 break;
 815             }
 816         // check; we won't crash if the book doesn't exist, but it's Bad Anyway.
 817         wxASSERT(bookr);
 818     }
 819     if (! bookr)
 820     {
 821         // no book specified; search all books
 822         m_CurIndex = 0;
 823         m_MaxIndex = m_Data->m_contents.size();
 824     }
 825     m_Engine.LookFor(keyword, case_sensitive, whole_words_only);
 826     m_Active = (m_CurIndex < m_MaxIndex);
 827 }
 828
 829 bool wxHtmlSearchStatus::Search()
 830 {
 831     wxFSFile *file;
 832     int i = m_CurIndex;  // shortcut
 833     bool found = false;
 834     wxString thepage;
 835
 836     if (!m_Active)
 837     {
 838         // sanity check. Illegal use, but we'll try to prevent a crash anyway
 839         wxASSERT(m_Active);
 840         return false;
 841     }
 842
 843     m_Name = wxEmptyString;
 844     m_CurItem = NULL;
 845     thepage = m_Data->m_contents[i].page;
 846
 847     m_Active = (++m_CurIndex < m_MaxIndex);
 848     // check if it is same page with different anchor:
 849     if (!m_LastPage.empty())
 850     {
 851         const wxChar *p1, *p2;
 852         for (p1 = thepage.c_str(), p2 = m_LastPage.c_str();
 853              *p1 != 0 && *p1 != wxT('#') && *p1 == *p2; p1++, p2++) {}
 854
 855         m_LastPage = thepage;
 856
 857         if (*p1 == 0 || *p1 == wxT('#'))
 858             return false;
 859     }
 860     else m_LastPage = thepage;
 861
 862     wxFileSystem fsys;
 863     file = fsys.OpenFile(m_Data->m_contents[i].book->GetFullPath(thepage));
 864     if (file)
 865     {
 866         if (m_Engine.Scan(*file))
 867         {
 868             m_Name = m_Data->m_contents[i].name;
 869             m_CurItem = &m_Data->m_contents[i];
 870             found = true;
 871         }
 872         delete file;
 873     }
 874     return found;
 875 }
 876
 877
 878
 879
 880
 881
 882
 883
 884 //--------------------------------------------------------------------------------
 885 // wxHtmlSearchEngine
 886 //--------------------------------------------------------------------------------
 887
 888 void wxHtmlSearchEngine::LookFor(const wxString& keyword, bool case_sensitive, bool whole_words_only)
 889 {
 890     m_CaseSensitive = case_sensitive;
 891     m_WholeWords = whole_words_only;
 892     m_Keyword = keyword;
 893
 894     if (!m_CaseSensitive)
 895         m_Keyword.LowerCase();
 896 }
 897
 898
 899 static inline bool WHITESPACE(wxChar c)
 900 {
 901     return c == wxT(' ') || c == wxT('\n') || c == wxT('\r') || c == wxT('\t');
 902 }
 903
 904 // replace continuous spaces by one single space
 905 static inline wxString CompressSpaces(const wxString & str)
 906 {
 907     wxString buf;
 908     buf.reserve( str.size() );
 909
 910     bool space_counted = false;
 911     for( const wxChar * pstr = str.c_str(); *pstr; ++pstr )
 912     {
 913         wxChar ch = *pstr;
 914         if( WHITESPACE( ch ) )
 915         {
 916             if( space_counted )
 917             {
 918                 continue;
 919             }
 920             ch = wxT(' ');
 921             space_counted = true;
 922         }
 923         else
 924         {
 925             space_counted = false;
 926         }
 927         buf += ch;
 928     }
 929
 930     return buf;
 931 }
 932
 933 bool wxHtmlSearchEngine::Scan(const wxFSFile& file)
 934 {
 935     wxASSERT_MSG(!m_Keyword.empty(), wxT("wxHtmlSearchEngine::LookFor must be called before scanning!"));
 936
 937     wxHtmlFilterHTML filter;
 938     wxString bufStr = filter.ReadFile(file);
 939
 940     if (!m_CaseSensitive)
 941         bufStr.LowerCase();
 942
 943     {   // remove html tags
 944         wxString bufStrCopy;
 945         bufStrCopy.reserve( bufStr.size() );
 946         bool insideTag = false;
 947         for (const wxChar * pBufStr = bufStr.c_str(); *pBufStr; ++pBufStr)
 948         {
 949             wxChar c = *pBufStr;
 950             if (insideTag)
 951             {
 952                 if (c == wxT('>'))
 953                 {
 954                     insideTag = false;
 955                     // replace the tag by an empty space
 956                     c = wxT(' ');
 957                 }
 958                 else
 959                     continue;
 960             }
 961             else if (c == wxT('<'))
 962             {
 963                 wxChar nextCh = *(pBufStr + 1);
 964                 if (nextCh == wxT('/') || !WHITESPACE(nextCh))
 965                 {
 966                     insideTag = true;
 967                     continue;
 968                 }
 969             }
 970             bufStrCopy += c;
 971         }
 972         bufStr.swap( bufStrCopy );
 973     }
 974
 975     wxString keyword = m_Keyword;
 976
 977     if (m_WholeWords)
 978     {
 979         // insert ' ' at the beginning and at the end
 980         keyword.insert( 0, wxT(" ") );
 981         keyword.append( wxT(" ") );
 982         bufStr.insert( 0, wxT(" ") );
 983         bufStr.append( wxT(" ") );
 984     }
 985
 986     // remove continuous spaces
 987     keyword = CompressSpaces( keyword );
 988     bufStr = CompressSpaces( bufStr );
 989
 990     // finally do the search
 991     return bufStr.find( keyword ) != wxString::npos;
 992 }
 993
 994 #endif