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 WXPRECOMP
  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 != _T('\r') && *readptr != _T('\n') &&
  55            writeptr != endptr)
  56         *(writeptr++) = *(readptr++);
  57     *writeptr = 0;
  58     while (*readptr == _T('\r') || *readptr == _T('\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 wxChar* WXUNUSED(txt)) {}
 129
 130     DECLARE_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     DECLARE_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 << _T("   ");
 259     s << name;
 260     return s;
 261 }
 262
 263
 264 IMPLEMENT_DYNAMIC_CLASS(wxHtmlHelpData, wxObject)
 265
 266 wxHtmlHelpData::wxHtmlHelpData()
 267 {
 268 #if WXWIN_COMPATIBILITY_2_4
 269     m_cacheContents = NULL;
 270     m_cacheIndex = NULL;
 271 #endif
 272 }
 273
 274 wxHtmlHelpData::~wxHtmlHelpData()
 275 {
 276 #if WXWIN_COMPATIBILITY_2_4
 277     CleanCompatibilityData();
 278 #endif
 279 }
 280
 281 bool wxHtmlHelpData::LoadMSProject(wxHtmlBookRecord *book, wxFileSystem& fsys,
 282                                    const wxString& indexfile,
 283                                    const wxString& contentsfile)
 284 {
 285     wxFSFile *f;
 286     wxHtmlFilterHTML filter;
 287     wxString buf;
 288     wxString string;
 289
 290     HP_Parser parser;
 291     HP_TagHandler *handler = new HP_TagHandler(book);
 292     parser.AddTagHandler(handler);
 293
 294     f = ( contentsfile.empty() ? (wxFSFile*) NULL : fsys.OpenFile(contentsfile) );
 295     if (f)
 296     {
 297         buf.clear();
 298         buf = filter.ReadFile(*f);
 299         delete f;
 300         handler->Reset(m_contents);
 301         parser.Parse(buf);
 302     }
 303     else
 304     {
 305         wxLogError(_("Cannot open contents file: %s"), contentsfile.c_str());
 306     }
 307
 308     f = ( indexfile.empty() ? (wxFSFile*) NULL : fsys.OpenFile(indexfile) );
 309     if (f)
 310     {
 311         buf.clear();
 312         buf = filter.ReadFile(*f);
 313         delete f;
 314         handler->Reset(m_index);
 315         parser.Parse(buf);
 316     }
 317     else if (!indexfile.empty())
 318     {
 319         wxLogError(_("Cannot open index file: %s"), indexfile.c_str());
 320     }
 321     return true;
 322 }
 323
 324 inline static void CacheWriteInt32(wxOutputStream *f, wxInt32 value)
 325 {
 326     wxInt32 x = wxINT32_SWAP_ON_BE(value);
 327     f->Write(&x, sizeof(x));
 328 }
 329
 330 inline static wxInt32 CacheReadInt32(wxInputStream *f)
 331 {
 332     wxInt32 x;
 333     f->Read(&x, sizeof(x));
 334     return wxINT32_SWAP_ON_BE(x);
 335 }
 336
 337 inline static void CacheWriteString(wxOutputStream *f, const wxString& str)
 338 {
 339     const wxWX2MBbuf mbstr = str.mb_str(wxConvUTF8);
 340     size_t len = strlen((const char*)mbstr)+1;
 341     CacheWriteInt32(f, len);
 342     f->Write((const char*)mbstr, len);
 343 }
 344
 345 inline static wxString CacheReadString(wxInputStream *f)
 346 {
 347     size_t len = (size_t)CacheReadInt32(f);
 348     wxCharBuffer str(len-1);
 349     f->Read(str.data(), len);
 350     return wxString(str, wxConvUTF8);
 351 }
 352
 353 #define CURRENT_CACHED_BOOK_VERSION     5
 354
 355 // Additional flags to detect incompatibilities of the runtime environment:
 356 #define CACHED_BOOK_FORMAT_FLAGS \
 357                      (wxUSE_UNICODE << 0)
 358
 359
 360 bool wxHtmlHelpData::LoadCachedBook(wxHtmlBookRecord *book, wxInputStream *f)
 361 {
 362     int i, st, newsize;
 363     wxInt32 version;
 364
 365     /* load header - version info : */
 366     version = CacheReadInt32(f);
 367
 368     if (version != CURRENT_CACHED_BOOK_VERSION)
 369     {
 370         // NB: We can just silently return false here and don't worry about
 371         //     it anymore, because AddBookParam will load the MS project in
 372         //     absence of (properly versioned) .cached file and automatically
 373         //     create new .cached file immediately afterward.
 374         return false;
 375     }
 376
 377     if (CacheReadInt32(f) != CACHED_BOOK_FORMAT_FLAGS)
 378         return false;
 379
 380     /* load contents : */
 381     st = m_contents.size();
 382     newsize = st + CacheReadInt32(f);
 383     m_contents.Alloc(newsize);
 384     for (i = st; i < newsize; i++)
 385     {
 386         wxHtmlHelpDataItem *item = new wxHtmlHelpDataItem;
 387         item->level = CacheReadInt32(f);
 388         item->id = CacheReadInt32(f);
 389         item->name = CacheReadString(f);
 390         item->page = CacheReadString(f);
 391         item->book = book;
 392         m_contents.Add(item);
 393     }
 394
 395     /* load index : */
 396     st = m_index.size();
 397     newsize = st + CacheReadInt32(f);
 398     m_index.Alloc(newsize);
 399     for (i = st; i < newsize; i++)
 400     {
 401         wxHtmlHelpDataItem *item = new wxHtmlHelpDataItem;
 402         item->name = CacheReadString(f);
 403         item->page = CacheReadString(f);
 404         item->level = CacheReadInt32(f);
 405         item->book = book;
 406         int parentShift = CacheReadInt32(f);
 407         if (parentShift != 0)
 408             item->parent = &m_index[m_index.size() - parentShift];
 409         m_index.Add(item);
 410     }
 411     return true;
 412 }
 413
 414
 415 bool wxHtmlHelpData::SaveCachedBook(wxHtmlBookRecord *book, wxOutputStream *f)
 416 {
 417     int i;
 418     wxInt32 cnt;
 419
 420     /* save header - version info : */
 421     CacheWriteInt32(f, CURRENT_CACHED_BOOK_VERSION);
 422     CacheWriteInt32(f, CACHED_BOOK_FORMAT_FLAGS);
 423
 424     /* save contents : */
 425     int len = m_contents.size();
 426     for (cnt = 0, i = 0; i < len; i++)
 427         if (m_contents[i].book == book && m_contents[i].level > 0)
 428             cnt++;
 429     CacheWriteInt32(f, cnt);
 430
 431     for (i = 0; i < len; i++)
 432     {
 433         if (m_contents[i].book != book || m_contents[i].level == 0)
 434             continue;
 435         CacheWriteInt32(f, m_contents[i].level);
 436         CacheWriteInt32(f, m_contents[i].id);
 437         CacheWriteString(f, m_contents[i].name);
 438         CacheWriteString(f, m_contents[i].page);
 439     }
 440
 441     /* save index : */
 442     len = m_index.size();
 443     for (cnt = 0, i = 0; i < len; i++)
 444         if (m_index[i].book == book && m_index[i].level > 0)
 445             cnt++;
 446     CacheWriteInt32(f, cnt);
 447
 448     for (i = 0; i < len; i++)
 449     {
 450         if (m_index[i].book != book || m_index[i].level == 0)
 451             continue;
 452         CacheWriteString(f, m_index[i].name);
 453         CacheWriteString(f, m_index[i].page);
 454         CacheWriteInt32(f, m_index[i].level);
 455         // save distance to parent item, if any:
 456         if (m_index[i].parent == NULL)
 457         {
 458             CacheWriteInt32(f, 0);
 459         }
 460         else
 461         {
 462             int cnt2 = 0;
 463             wxHtmlHelpDataItem *parent = m_index[i].parent;
 464             for (int j = i-1; j >= 0; j--)
 465             {
 466                 if (m_index[j].book == book && m_index[j].level > 0)
 467                     cnt2++;
 468                 if (&m_index[j] == parent)
 469                     break;
 470             }
 471             wxASSERT(cnt2 > 0);
 472             CacheWriteInt32(f, cnt2);
 473         }
 474     }
 475     return true;
 476 }
 477
 478
 479 void wxHtmlHelpData::SetTempDir(const wxString& path)
 480 {
 481     if (path.empty())
 482         m_tempPath = path;
 483     else
 484     {
 485         if (wxIsAbsolutePath(path)) m_tempPath = path;
 486         else m_tempPath = wxGetCwd() + _T("/") + path;
 487
 488         if (m_tempPath[m_tempPath.length() - 1] != _T('/'))
 489             m_tempPath << _T('/');
 490     }
 491 }
 492
 493
 494
 495 static wxString SafeFileName(const wxString& s)
 496 {
 497     wxString res(s);
 498     res.Replace(wxT("#"), wxT("_"));
 499     res.Replace(wxT(":"), wxT("_"));
 500     res.Replace(wxT("\\"), wxT("_"));
 501     res.Replace(wxT("/"), wxT("_"));
 502     return res;
 503 }
 504
 505 bool wxHtmlHelpData::AddBookParam(const wxFSFile& bookfile,
 506                                   wxFontEncoding encoding,
 507                                   const wxString& title, const wxString& contfile,
 508                                   const wxString& indexfile, const wxString& deftopic,
 509                                   const wxString& path)
 510 {
 511     wxFileSystem fsys;
 512     wxFSFile *fi;
 513     wxHtmlBookRecord *bookr;
 514
 515     int IndexOld = m_index.size(),
 516         ContentsOld = m_contents.size();
 517
 518     if (!path.empty())
 519         fsys.ChangePathTo(path, true);
 520
 521     size_t booksCnt = m_bookRecords.GetCount();
 522     for (size_t i = 0; i < booksCnt; i++)
 523     {
 524         if ( m_bookRecords[i].GetBookFile() == bookfile.GetLocation() )
 525             return true; // book is (was) loaded
 526     }
 527
 528     bookr = new wxHtmlBookRecord(bookfile.GetLocation(), fsys.GetPath(), title, deftopic);
 529
 530     wxHtmlHelpDataItem *bookitem = new wxHtmlHelpDataItem;
 531     bookitem->level = 0;
 532     bookitem->id = 0;
 533     bookitem->page = deftopic;
 534     bookitem->name = title;
 535     bookitem->book = bookr;
 536
 537     // store the contents index for later
 538     int cont_start = m_contents.size();
 539
 540     m_contents.Add(bookitem);
 541
 542     // Try to find cached binary versions:
 543     // 1. save file as book, but with .hhp.cached extension
 544     // 2. same as 1. but in temp path
 545     // 3. otherwise or if cache load failed, load it from MS.
 546
 547     fi = fsys.OpenFile(bookfile.GetLocation() + wxT(".cached"));
 548
 549     if (fi == NULL ||
 550 #if wxUSE_DATETIME
 551           fi->GetModificationTime() < bookfile.GetModificationTime() ||
 552 #endif // wxUSE_DATETIME
 553           !LoadCachedBook(bookr, fi->GetStream()))
 554     {
 555         if (fi != NULL) delete fi;
 556         fi = fsys.OpenFile(m_tempPath + wxFileNameFromPath(bookfile.GetLocation()) + wxT(".cached"));
 557         if (m_tempPath.empty() || fi == NULL ||
 558 #if wxUSE_DATETIME
 559             fi->GetModificationTime() < bookfile.GetModificationTime() ||
 560 #endif // wxUSE_DATETIME
 561             !LoadCachedBook(bookr, fi->GetStream()))
 562         {
 563             LoadMSProject(bookr, fsys, indexfile, contfile);
 564             if (!m_tempPath.empty())
 565             {
 566                 wxFileOutputStream *outs = new wxFileOutputStream(m_tempPath +
 567                                                   SafeFileName(wxFileNameFromPath(bookfile.GetLocation())) + wxT(".cached"));
 568                 SaveCachedBook(bookr, outs);
 569                 delete outs;
 570             }
 571         }
 572     }
 573
 574     if (fi != NULL) delete fi;
 575
 576     // Now store the contents range
 577     bookr->SetContentsRange(cont_start, m_contents.size());
 578
 579 #if wxUSE_WCHAR_T
 580     // MS HTML Help files [written by MS HTML Help Workshop] are broken
 581     // in that the data are iso-8859-1 (including HTML entities), but must
 582     // be interpreted as being in language's windows charset. Correct the
 583     // differences here and also convert to wxConvLocal in ANSI build
 584     if (encoding != wxFONTENCODING_SYSTEM)
 585     {
 586         #if wxUSE_UNICODE
 587             #define CORRECT_STR(str, conv) \
 588                 str = wxString((str).mb_str(wxConvISO8859_1), conv)
 589         #else
 590             #define CORRECT_STR(str, conv) \
 591                 str = wxString((str).wc_str(conv), wxConvLocal)
 592         #endif
 593         wxCSConv conv(encoding);
 594         size_t IndexCnt = m_index.size();
 595         size_t ContentsCnt = m_contents.size();
 596         size_t i;
 597         for (i = IndexOld; i < IndexCnt; i++)
 598         {
 599             CORRECT_STR(m_index[i].name, conv);
 600         }
 601         for (i = ContentsOld; i < ContentsCnt; i++)
 602         {
 603             CORRECT_STR(m_contents[i].name, conv);
 604         }
 605         #undef CORRECT_STR
 606     }
 607 #else
 608     wxUnusedVar(IndexOld);
 609     wxUnusedVar(ContentsOld);
 610     wxASSERT_MSG(encoding == wxFONTENCODING_SYSTEM, wxT("Help files need charset conversion, but wxUSE_WCHAR_T is 0"));
 611 #endif // wxUSE_WCHAR_T/!wxUSE_WCHAR_T
 612
 613     m_bookRecords.Add(bookr);
 614     if (!m_index.empty())
 615     {
 616         m_index.Sort(wxHtmlHelpIndexCompareFunc);
 617     }
 618
 619     return true;
 620 }
 621
 622
 623 bool wxHtmlHelpData::AddBook(const wxString& book)
 624 {
 625     wxString extension(book.Right(4).Lower());
 626     if (extension == wxT(".zip") ||
 627 #if wxUSE_LIBMSPACK
 628         extension == wxT(".chm") /*compressed html help book*/ ||
 629 #endif
 630         extension == wxT(".htb") /*html book*/)
 631     {
 632         wxFileSystem fsys;
 633         wxString s;
 634         bool rt = false;
 635
 636 #if wxUSE_LIBMSPACK
 637         if (extension == wxT(".chm"))
 638             s = fsys.FindFirst(book + wxT("#chm:*.hhp"), wxFILE);
 639         else
 640 #endif
 641             s = fsys.FindFirst(book + wxT("#zip:*.hhp"), wxFILE);
 642
 643         while (!s.empty())
 644         {
 645             if (AddBook(s)) rt = true;
 646             s = fsys.FindNext();
 647         }
 648
 649         return rt;
 650     }
 651
 652     wxFSFile *fi;
 653     wxFileSystem fsys;
 654
 655     wxString title = _("noname"),
 656              safetitle,
 657              start = wxEmptyString,
 658              contents = wxEmptyString,
 659              index = wxEmptyString,
 660              charset = wxEmptyString;
 661
 662     fi = fsys.OpenFile(book);
 663     if (fi == NULL)
 664     {
 665         wxLogError(_("Cannot open HTML help book: %s"), book.c_str());
 666         return false;
 667     }
 668     fsys.ChangePathTo(book);
 669
 670     const wxChar *lineptr;
 671     wxChar linebuf[300];
 672     wxString tmp;
 673     wxHtmlFilterPlainText filter;
 674     tmp = filter.ReadFile(*fi);
 675     lineptr = tmp.c_str();
 676
 677     do
 678     {
 679         lineptr = ReadLine(lineptr, linebuf, 300);
 680
 681         for (wxChar *ch = linebuf; *ch != wxT('\0') && *ch != wxT('='); ch++)
 682            *ch = (wxChar)wxTolower(*ch);
 683
 684         if (wxStrstr(linebuf, _T("title=")) == linebuf)
 685             title = linebuf + wxStrlen(_T("title="));
 686         if (wxStrstr(linebuf, _T("default topic=")) == linebuf)
 687             start = linebuf + wxStrlen(_T("default topic="));
 688         if (wxStrstr(linebuf, _T("index file=")) == linebuf)
 689             index = linebuf + wxStrlen(_T("index file="));
 690         if (wxStrstr(linebuf, _T("contents file=")) == linebuf)
 691             contents = linebuf + wxStrlen(_T("contents file="));
 692         if (wxStrstr(linebuf, _T("charset=")) == linebuf)
 693             charset = linebuf + wxStrlen(_T("charset="));
 694     } while (lineptr != NULL);
 695
 696     wxFontEncoding enc = wxFONTENCODING_SYSTEM;
 697 #if wxUSE_FONTMAP
 698     if (charset != wxEmptyString)
 699         enc = wxFontMapper::Get()->CharsetToEncoding(charset);
 700 #endif
 701
 702     bool rtval = AddBookParam(*fi, enc,
 703                               title, contents, index, start, fsys.GetPath());
 704     delete fi;
 705
 706 #if WXWIN_COMPATIBILITY_2_4
 707     CleanCompatibilityData();
 708 #endif
 709
 710     return rtval;
 711 }
 712
 713 wxString wxHtmlHelpData::FindPageByName(const wxString& x)
 714 {
 715     int cnt;
 716     int i;
 717     wxFileSystem fsys;
 718     wxFSFile *f;
 719
 720     /* 1. try to open given file: */
 721
 722     cnt = m_bookRecords.GetCount();
 723     for (i = 0; i < cnt; i++)
 724     {
 725         f = fsys.OpenFile(m_bookRecords[i].GetFullPath(x));
 726         if (f)
 727         {
 728             wxString url = m_bookRecords[i].GetFullPath(x);
 729             delete f;
 730             return url;
 731         }
 732     }
 733
 734
 735     /* 2. try to find a book: */
 736
 737     for (i = 0; i < cnt; i++)
 738     {
 739         if (m_bookRecords[i].GetTitle() == x)
 740             return m_bookRecords[i].GetFullPath(m_bookRecords[i].GetStart());
 741     }
 742
 743     /* 3. try to find in contents: */
 744
 745     cnt = m_contents.size();
 746     for (i = 0; i < cnt; i++)
 747     {
 748         if (m_contents[i].name == x)
 749             return m_contents[i].GetFullPath();
 750     }
 751
 752
 753     /* 4. try to find in index: */
 754
 755     cnt = m_index.size();
 756     for (i = 0; i < cnt; i++)
 757     {
 758         if (m_index[i].name == x)
 759             return m_index[i].GetFullPath();
 760     }
 761
 762     return wxEmptyString;
 763 }
 764
 765 wxString wxHtmlHelpData::FindPageById(int id)
 766 {
 767     size_t cnt = m_contents.size();
 768     for (size_t i = 0; i < cnt; i++)
 769     {
 770         if (m_contents[i].id == id)
 771         {
 772             return m_contents[i].GetFullPath();
 773         }
 774     }
 775
 776     return wxEmptyString;
 777 }
 778
 779 #if WXWIN_COMPATIBILITY_2_4
 780 wxHtmlContentsItem::wxHtmlContentsItem()
 781     : m_Level(0), m_ID(wxID_ANY), m_Name(NULL), m_Page(NULL), m_Book(NULL),
 782       m_autofree(false)
 783 {
 784 }
 785
 786 wxHtmlContentsItem::wxHtmlContentsItem(const wxHtmlHelpDataItem& d)
 787 {
 788     m_autofree = true;
 789     m_Level = d.level;
 790     m_ID = d.id;
 791     m_Name = wxStrdup(d.name.c_str());
 792     m_Page = wxStrdup(d.page.c_str());
 793     m_Book = d.book;
 794 }
 795
 796 wxHtmlContentsItem& wxHtmlContentsItem::operator=(const wxHtmlContentsItem& d)
 797 {
 798     if (m_autofree)
 799     {
 800         free(m_Name);
 801         free(m_Page);
 802     }
 803     m_autofree = true;
 804     m_Level = d.m_Level;
 805     m_ID = d.m_ID;
 806     m_Name = d.m_Name ? wxStrdup(d.m_Name) : NULL;
 807     m_Page = d.m_Page ? wxStrdup(d.m_Page) : NULL;
 808     m_Book = d.m_Book;
 809     return *this;
 810 }
 811
 812 wxHtmlContentsItem::~wxHtmlContentsItem()
 813 {
 814     if (m_autofree)
 815     {
 816         free(m_Name);
 817         free(m_Page);
 818     }
 819 }
 820
 821 wxHtmlContentsItem* wxHtmlHelpData::GetContents()
 822 {
 823     if (!m_cacheContents && !m_contents.empty())
 824     {
 825         size_t len = m_contents.size();
 826         m_cacheContents = new wxHtmlContentsItem[len];
 827         for (size_t i = 0; i < len; i++)
 828             m_cacheContents[i] = m_contents[i];
 829     }
 830     return m_cacheContents;
 831 }
 832
 833 int wxHtmlHelpData::GetContentsCnt()
 834 {
 835     return m_contents.size();
 836 }
 837
 838 wxHtmlContentsItem* wxHtmlHelpData::GetIndex()
 839 {
 840     if (!m_cacheContents && !m_index.empty())
 841     {
 842         size_t len = m_index.size();
 843         m_cacheContents = new wxHtmlContentsItem[len];
 844         for (size_t i = 0; i < len; i++)
 845             m_cacheContents[i] = m_index[i];
 846     }
 847     return m_cacheContents;
 848 }
 849
 850 int wxHtmlHelpData::GetIndexCnt()
 851 {
 852     return m_index.size();
 853 }
 854
 855 void wxHtmlHelpData::CleanCompatibilityData()
 856 {
 857     delete[] m_cacheContents;
 858     m_cacheContents = NULL;
 859     delete[] m_cacheIndex;
 860     m_cacheIndex = NULL;
 861 }
 862 #endif // WXWIN_COMPATIBILITY_2_4
 863
 864 //----------------------------------------------------------------------------------
 865 // wxHtmlSearchStatus functions
 866 //----------------------------------------------------------------------------------
 867
 868 wxHtmlSearchStatus::wxHtmlSearchStatus(wxHtmlHelpData* data, const wxString& keyword,
 869                                        bool case_sensitive, bool whole_words_only,
 870                                        const wxString& book)
 871 {
 872     m_Data = data;
 873     m_Keyword = keyword;
 874     wxHtmlBookRecord* bookr = NULL;
 875     if (book != wxEmptyString)
 876     {
 877         // we have to search in a specific book. Find it first
 878         int i, cnt = data->m_bookRecords.GetCount();
 879         for (i = 0; i < cnt; i++)
 880             if (data->m_bookRecords[i].GetTitle() == book)
 881             {
 882                 bookr = &(data->m_bookRecords[i]);
 883                 m_CurIndex = bookr->GetContentsStart();
 884                 m_MaxIndex = bookr->GetContentsEnd();
 885                 break;
 886             }
 887         // check; we won't crash if the book doesn't exist, but it's Bad Anyway.
 888         wxASSERT(bookr);
 889     }
 890     if (! bookr)
 891     {
 892         // no book specified; search all books
 893         m_CurIndex = 0;
 894         m_MaxIndex = m_Data->m_contents.size();
 895     }
 896     m_Engine.LookFor(keyword, case_sensitive, whole_words_only);
 897     m_Active = (m_CurIndex < m_MaxIndex);
 898 }
 899
 900 #if WXWIN_COMPATIBILITY_2_4
 901 wxHtmlContentsItem* wxHtmlSearchStatus::GetContentsItem()
 902 {
 903     static wxHtmlContentsItem it;
 904     it = wxHtmlContentsItem(*m_CurItem);
 905     return &it;
 906 }
 907 #endif
 908
 909 bool wxHtmlSearchStatus::Search()
 910 {
 911     wxFSFile *file;
 912     int i = m_CurIndex;  // shortcut
 913     bool found = false;
 914     wxString thepage;
 915
 916     if (!m_Active)
 917     {
 918         // sanity check. Illegal use, but we'll try to prevent a crash anyway
 919         wxASSERT(m_Active);
 920         return false;
 921     }
 922
 923     m_Name = wxEmptyString;
 924     m_CurItem = NULL;
 925     thepage = m_Data->m_contents[i].page;
 926
 927     m_Active = (++m_CurIndex < m_MaxIndex);
 928     // check if it is same page with different anchor:
 929     if (!m_LastPage.empty())
 930     {
 931         const wxChar *p1, *p2;
 932         for (p1 = thepage.c_str(), p2 = m_LastPage.c_str();
 933              *p1 != 0 && *p1 != _T('#') && *p1 == *p2; p1++, p2++) {}
 934
 935         m_LastPage = thepage;
 936
 937         if (*p1 == 0 || *p1 == _T('#'))
 938             return false;
 939     }
 940     else m_LastPage = thepage;
 941
 942     wxFileSystem fsys;
 943     file = fsys.OpenFile(m_Data->m_contents[i].book->GetFullPath(thepage));
 944     if (file)
 945     {
 946         if (m_Engine.Scan(*file))
 947         {
 948             m_Name = m_Data->m_contents[i].name;
 949             m_CurItem = &m_Data->m_contents[i];
 950             found = true;
 951         }
 952         delete file;
 953     }
 954     return found;
 955 }
 956
 957
 958
 959
 960
 961
 962
 963
 964 //--------------------------------------------------------------------------------
 965 // wxHtmlSearchEngine
 966 //--------------------------------------------------------------------------------
 967
 968 void wxHtmlSearchEngine::LookFor(const wxString& keyword, bool case_sensitive, bool whole_words_only)
 969 {
 970     m_CaseSensitive = case_sensitive;
 971     m_WholeWords = whole_words_only;
 972     m_Keyword = keyword;
 973
 974     if (!m_CaseSensitive)
 975         m_Keyword.LowerCase();
 976 }
 977
 978
 979 static inline bool WHITESPACE(wxChar c)
 980 {
 981     return c == _T(' ') || c == _T('\n') || c == _T('\r') || c == _T('\t');
 982 }
 983
 984 // replace continuous spaces by one single space
 985 static inline wxString CompressSpaces(const wxString & str)
 986 {
 987     wxString buf;
 988     buf.reserve( str.size() );
 989
 990     bool space_counted = false;
 991     for( const wxChar * pstr = str.c_str(); *pstr; ++pstr )
 992     {
 993         wxChar ch = *pstr;
 994         if( WHITESPACE( ch ) )
 995         {
 996             if( space_counted )
 997             {
 998                 continue;
 999             }
1000             ch = _T(' ');
1001             space_counted = true;
1002         }
1003         else
1004         {
1005             space_counted = false;
1006         }
1007         buf += ch;
1008     }
1009
1010     return buf;
1011 }
1012
1013 bool wxHtmlSearchEngine::Scan(const wxFSFile& file)
1014 {
1015     wxASSERT_MSG(!m_Keyword.empty(), wxT("wxHtmlSearchEngine::LookFor must be called before scanning!"));
1016
1017     wxHtmlFilterHTML filter;
1018     wxString bufStr = filter.ReadFile(file);
1019
1020     if (!m_CaseSensitive)
1021         bufStr.LowerCase();
1022
1023     {   // remove html tags
1024         wxString bufStrCopy;
1025         bufStrCopy.reserve( bufStr.size() );
1026         bool insideTag = false;
1027         for (const wxChar * pBufStr = bufStr.c_str(); *pBufStr; ++pBufStr)
1028         {
1029             wxChar c = *pBufStr;
1030             if (insideTag)
1031             {
1032                 if (c == _T('>'))
1033                 {
1034                     insideTag = false;
1035                     // replace the tag by an empty space
1036                     c = _T(' ');
1037                 }
1038                 else
1039                     continue;
1040             }
1041             else if (c == _T('<'))
1042             {
1043                 wxChar nextCh = *(pBufStr + 1);
1044                 if (nextCh == _T('/') || !WHITESPACE(nextCh))
1045                 {
1046                     insideTag = true;
1047                     continue;
1048                 }
1049             }
1050             bufStrCopy += c;
1051         }
1052         bufStr.swap( bufStrCopy );
1053     }
1054
1055     wxString keyword = m_Keyword;
1056
1057     if (m_WholeWords)
1058     {
1059         // insert ' ' at the beginning and at the end
1060         keyword.insert( 0, _T(" ") );
1061         keyword.append( _T(" ") );
1062         bufStr.insert( 0, _T(" ") );
1063         bufStr.append( _T(" ") );
1064     }
1065
1066     // remove continuous spaces
1067     keyword = CompressSpaces( keyword );
1068     bufStr = CompressSpaces( bufStr );
1069
1070     // finally do the search
1071     return bufStr.find( keyword ) != wxString::npos;
1072 }
1073
1074 #endif