src/html/helpdata.cpp

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