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) || page.Find(wxT("file:")) == 0)
 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 #if wxUSE_WCHAR_T
 504         #if wxUSE_UNICODE
 505             #define CORRECT_STR(str, conv) \
 506                 str = wxString((str).mb_str(wxConvISO8859_1), conv)
 507         #else
 508             #define CORRECT_STR(str, conv) \
 509                 str = wxString((str).wc_str(conv), wxConvLocal)
 510         #endif
 511 #else
 512     #define CORRECT_STR(str, conv)
 513 #endif
 514
 515     wxFileSystem fsys;
 516     wxFSFile *fi;
 517     wxHtmlBookRecord *bookr;
 518
 519     int IndexOld = m_index.size(),
 520         ContentsOld = m_contents.size();
 521
 522     if (!path.empty())
 523         fsys.ChangePathTo(path, true);
 524
 525     size_t booksCnt = m_bookRecords.GetCount();
 526     for (size_t i = 0; i < booksCnt; i++)
 527     {
 528         if ( m_bookRecords[i].GetBookFile() == bookfile.GetLocation() )
 529             return true; // book is (was) loaded
 530     }
 531
 532     wxString title1 = title;
 533     if (encoding != wxFONTENCODING_SYSTEM)
 534     {
 535         wxCSConv conv(encoding);
 536         CORRECT_STR(title1, conv);
 537         if (title1.IsEmpty() && !title.IsEmpty())
 538             title1 = title;
 539     }
 540
 541     bookr = new wxHtmlBookRecord(bookfile.GetLocation(), fsys.GetPath(), title1, deftopic);
 542
 543     wxHtmlHelpDataItem *bookitem = new wxHtmlHelpDataItem;
 544     bookitem->level = 0;
 545     bookitem->id = 0;
 546     bookitem->page = deftopic;
 547     bookitem->name = title;
 548     bookitem->book = bookr;
 549
 550     // store the contents index for later
 551     int cont_start = m_contents.size();
 552
 553     m_contents.Add(bookitem);
 554
 555     // Try to find cached binary versions:
 556     // 1. save file as book, but with .hhp.cached extension
 557     // 2. same as 1. but in temp path
 558     // 3. otherwise or if cache load failed, load it from MS.
 559
 560     fi = fsys.OpenFile(bookfile.GetLocation() + wxT(".cached"));
 561
 562     if (fi == NULL ||
 563 #if wxUSE_DATETIME
 564           fi->GetModificationTime() < bookfile.GetModificationTime() ||
 565 #endif // wxUSE_DATETIME
 566           !LoadCachedBook(bookr, fi->GetStream()))
 567     {
 568         if (fi != NULL) delete fi;
 569         fi = fsys.OpenFile(m_tempPath + wxFileNameFromPath(bookfile.GetLocation()) + wxT(".cached"));
 570         if (m_tempPath.empty() || fi == NULL ||
 571 #if wxUSE_DATETIME
 572             fi->GetModificationTime() < bookfile.GetModificationTime() ||
 573 #endif // wxUSE_DATETIME
 574             !LoadCachedBook(bookr, fi->GetStream()))
 575         {
 576             LoadMSProject(bookr, fsys, indexfile, contfile);
 577             if (!m_tempPath.empty())
 578             {
 579                 wxFileOutputStream *outs = new wxFileOutputStream(m_tempPath +
 580                                                   SafeFileName(wxFileNameFromPath(bookfile.GetLocation())) + wxT(".cached"));
 581                 SaveCachedBook(bookr, outs);
 582                 delete outs;
 583             }
 584         }
 585     }
 586
 587     if (fi != NULL) delete fi;
 588
 589     // Now store the contents range
 590     bookr->SetContentsRange(cont_start, m_contents.size());
 591
 592     // MS HTML Help files [written by MS HTML Help Workshop] are broken
 593     // in that the data are iso-8859-1 (including HTML entities), but must
 594     // be interpreted as being in language's windows charset. Correct the
 595     // differences here and also convert to wxConvLocal in ANSI build
 596     if (encoding != wxFONTENCODING_SYSTEM)
 597     {
 598         wxCSConv conv(encoding);
 599         size_t IndexCnt = m_index.size();
 600         size_t ContentsCnt = m_contents.size();
 601         size_t i;
 602         for (i = IndexOld; i < IndexCnt; i++)
 603         {
 604             CORRECT_STR(m_index[i].name, conv);
 605         }
 606         for (i = ContentsOld; i < ContentsCnt; i++)
 607         {
 608             CORRECT_STR(m_contents[i].name, conv);
 609         }
 610         #undef CORRECT_STR
 611     }
 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, wxT("title=")) == linebuf)
 685             title = linebuf + wxStrlen(wxT("title="));
 686         if (wxStrstr(linebuf, wxT("default topic=")) == linebuf)
 687             start = linebuf + wxStrlen(wxT("default topic="));
 688         if (wxStrstr(linebuf, wxT("index file=")) == linebuf)
 689             index = linebuf + wxStrlen(wxT("index file="));
 690         if (wxStrstr(linebuf, wxT("contents file=")) == linebuf)
 691             contents = linebuf + wxStrlen(wxT("contents file="));
 692         if (wxStrstr(linebuf, wxT("charset=")) == linebuf)
 693             charset = linebuf + wxStrlen(wxT("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     return rtval;
 707 }
 708
 709 wxString wxHtmlHelpData::FindPageByName(const wxString& x)
 710 {
 711     int i;
 712
 713     bool has_non_ascii = false;
 714     wxString::const_iterator it;
 715     for (it = x.begin(); it != x.end(); ++it)
 716     {
 717         wxUniChar ch = *it;
 718         if (!ch.IsAscii())
 719         {
 720             has_non_ascii = true;
 721             break;
 722         }
 723     }
 724
 725     int cnt = m_bookRecords.GetCount();
 726
 727     if (!has_non_ascii)
 728     {
 729       wxFileSystem fsys;
 730       wxFSFile *f;
 731       // 1. try to open given file:
 732       for (i = 0; i < cnt; i++)
 733       {
 734         f = fsys.OpenFile(m_bookRecords[i].GetFullPath(x));
 735         if (f)
 736         {
 737             wxString url = m_bookRecords[i].GetFullPath(x);
 738             delete f;
 739             return url;
 740         }
 741       }
 742     }
 743
 744
 745     // 2. try to find a book:
 746     for (i = 0; i < cnt; i++)
 747     {
 748         if (m_bookRecords[i].GetTitle() == x)
 749             return m_bookRecords[i].GetFullPath(m_bookRecords[i].GetStart());
 750     }
 751
 752     // 3. try to find in contents:
 753     cnt = m_contents.size();
 754     for (i = 0; i < cnt; i++)
 755     {
 756         if (m_contents[i].name == x)
 757             return m_contents[i].GetFullPath();
 758     }
 759
 760
 761     // 4. try to find in index:
 762     cnt = m_index.size();
 763     for (i = 0; i < cnt; i++)
 764     {
 765         if (m_index[i].name == x)
 766             return m_index[i].GetFullPath();
 767     }
 768
 769     // 4b. if still not found, try case-insensitive comparison
 770     for (i = 0; i < cnt; i++)
 771     {
 772         if (m_index[i].name.CmpNoCase(x) == 0)
 773             return m_index[i].GetFullPath();
 774     }
 775
 776     return wxEmptyString;
 777 }
 778
 779 wxString wxHtmlHelpData::FindPageById(int id)
 780 {
 781     size_t cnt = m_contents.size();
 782     for (size_t i = 0; i < cnt; i++)
 783     {
 784         if (m_contents[i].id == id)
 785         {
 786             return m_contents[i].GetFullPath();
 787         }
 788     }
 789
 790     return wxEmptyString;
 791 }
 792
 793
 794 //----------------------------------------------------------------------------------
 795 // wxHtmlSearchStatus functions
 796 //----------------------------------------------------------------------------------
 797
 798 wxHtmlSearchStatus::wxHtmlSearchStatus(wxHtmlHelpData* data, const wxString& keyword,
 799                                        bool case_sensitive, bool whole_words_only,
 800                                        const wxString& book)
 801 {
 802     m_Data = data;
 803     m_Keyword = keyword;
 804     wxHtmlBookRecord* bookr = NULL;
 805     if (book != wxEmptyString)
 806     {
 807         // we have to search in a specific book. Find it first
 808         int i, cnt = data->m_bookRecords.GetCount();
 809         for (i = 0; i < cnt; i++)
 810             if (data->m_bookRecords[i].GetTitle() == book)
 811             {
 812                 bookr = &(data->m_bookRecords[i]);
 813                 m_CurIndex = bookr->GetContentsStart();
 814                 m_MaxIndex = bookr->GetContentsEnd();
 815                 break;
 816             }
 817         // check; we won't crash if the book doesn't exist, but it's Bad Anyway.
 818         wxASSERT(bookr);
 819     }
 820     if (! bookr)
 821     {
 822         // no book specified; search all books
 823         m_CurIndex = 0;
 824         m_MaxIndex = m_Data->m_contents.size();
 825     }
 826     m_Engine.LookFor(keyword, case_sensitive, whole_words_only);
 827     m_Active = (m_CurIndex < m_MaxIndex);
 828 }
 829
 830 bool wxHtmlSearchStatus::Search()
 831 {
 832     wxFSFile *file;
 833     int i = m_CurIndex;  // shortcut
 834     bool found = false;
 835     wxString thepage;
 836
 837     if (!m_Active)
 838     {
 839         // sanity check. Illegal use, but we'll try to prevent a crash anyway
 840         wxASSERT(m_Active);
 841         return false;
 842     }
 843
 844     m_Name = wxEmptyString;
 845     m_CurItem = NULL;
 846     thepage = m_Data->m_contents[i].page;
 847
 848     m_Active = (++m_CurIndex < m_MaxIndex);
 849     // check if it is same page with different anchor:
 850     if (!m_LastPage.empty())
 851     {
 852         const wxChar *p1, *p2;
 853         for (p1 = thepage.c_str(), p2 = m_LastPage.c_str();
 854              *p1 != 0 && *p1 != wxT('#') && *p1 == *p2; p1++, p2++) {}
 855
 856         m_LastPage = thepage;
 857
 858         if (*p1 == 0 || *p1 == wxT('#'))
 859             return false;
 860     }
 861     else m_LastPage = thepage;
 862
 863     wxFileSystem fsys;
 864     file = fsys.OpenFile(m_Data->m_contents[i].book->GetFullPath(thepage));
 865     if (file)
 866     {
 867         if (m_Engine.Scan(*file))
 868         {
 869             m_Name = m_Data->m_contents[i].name;
 870             m_CurItem = &m_Data->m_contents[i];
 871             found = true;
 872         }
 873         delete file;
 874     }
 875     return found;
 876 }
 877
 878
 879
 880
 881
 882
 883
 884
 885 //--------------------------------------------------------------------------------
 886 // wxHtmlSearchEngine
 887 //--------------------------------------------------------------------------------
 888
 889 void wxHtmlSearchEngine::LookFor(const wxString& keyword, bool case_sensitive, bool whole_words_only)
 890 {
 891     m_CaseSensitive = case_sensitive;
 892     m_WholeWords = whole_words_only;
 893     m_Keyword = keyword;
 894
 895     if (!m_CaseSensitive)
 896         m_Keyword.LowerCase();
 897 }
 898
 899
 900 static inline bool WHITESPACE(wxChar c)
 901 {
 902     return c == wxT(' ') || c == wxT('\n') || c == wxT('\r') || c == wxT('\t');
 903 }
 904
 905 // replace continuous spaces by one single space
 906 static inline wxString CompressSpaces(const wxString & str)
 907 {
 908     wxString buf;
 909     buf.reserve( str.size() );
 910
 911     bool space_counted = false;
 912     for( const wxChar * pstr = str.c_str(); *pstr; ++pstr )
 913     {
 914         wxChar ch = *pstr;
 915         if( WHITESPACE( ch ) )
 916         {
 917             if( space_counted )
 918             {
 919                 continue;
 920             }
 921             ch = wxT(' ');
 922             space_counted = true;
 923         }
 924         else
 925         {
 926             space_counted = false;
 927         }
 928         buf += ch;
 929     }
 930
 931     return buf;
 932 }
 933
 934 bool wxHtmlSearchEngine::Scan(const wxFSFile& file)
 935 {
 936     wxASSERT_MSG(!m_Keyword.empty(), wxT("wxHtmlSearchEngine::LookFor must be called before scanning!"));
 937
 938     wxHtmlFilterHTML filter;
 939     wxString bufStr = filter.ReadFile(file);
 940
 941     if (!m_CaseSensitive)
 942         bufStr.LowerCase();
 943
 944     {   // remove html tags
 945         wxString bufStrCopy;
 946         bufStrCopy.reserve( bufStr.size() );
 947         bool insideTag = false;
 948         for (const wxChar * pBufStr = bufStr.c_str(); *pBufStr; ++pBufStr)
 949         {
 950             wxChar c = *pBufStr;
 951             if (insideTag)
 952             {
 953                 if (c == wxT('>'))
 954                 {
 955                     insideTag = false;
 956                     // replace the tag by an empty space
 957                     c = wxT(' ');
 958                 }
 959                 else
 960                     continue;
 961             }
 962             else if (c == wxT('<'))
 963             {
 964                 wxChar nextCh = *(pBufStr + 1);
 965                 if (nextCh == wxT('/') || !WHITESPACE(nextCh))
 966                 {
 967                     insideTag = true;
 968                     continue;
 969                 }
 970             }
 971             bufStrCopy += c;
 972         }
 973         bufStr.swap( bufStrCopy );
 974     }
 975
 976     wxString keyword = m_Keyword;
 977
 978     if (m_WholeWords)
 979     {
 980         // insert ' ' at the beginning and at the end
 981         keyword.insert( 0, wxT(" ") );
 982         keyword.append( wxT(" ") );
 983         bufStr.insert( 0, wxT(" ") );
 984         bufStr.append( wxT(" ") );
 985     }
 986
 987     // remove continuous spaces
 988     keyword = CompressSpaces( keyword );
 989     bufStr = CompressSpaces( bufStr );
 990
 991     // finally do the search
 992     return bufStr.find( keyword ) != wxString::npos;
 993 }
 994
 995 #endif