/*************************************************************************** * Copyright (C) 2004-2007 by Georgy Yunaev, gyunaev@ulduzsoft.com * * Portions Copyright (C) 2003 Razvan Cojocaru * * Please do not use email address above for bug reports; see * * the README file * * * * This program is free software; you can redistribute it and/or modify * * it under the terms of the GNU General Public License as published by * * the Free Software Foundation; either version 2 of the License, or * * (at your option) any later version. * * * * This program is distributed in the hope that it will be useful, * * but WITHOUT ANY WARRANTY; without even the implied warranty of * * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * * GNU General Public License for more details. * * * * You should have received a copy of the GNU General Public License * * along with this program; if not, write to the * * Free Software Foundation, Inc., * * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. * ***************************************************************************/ #include #include #include #include #include "config.h" #include "chm_lib.h" #include "bitfiddle.h" #include "libchmfile.h" #include "libchmurlfactory.h" #include "libchmfileimpl.h" // Big-enough buffer size for use with various routines. #define BUF_SIZE 4096 #define COMMON_BUF_LEN 1025 #define TOPICS_ENTRY_LEN 16 #define URLTBL_ENTRY_LEN 12 //#define DEBUGPARSER(A) tqDebug A #define DEBUGPARSER(A) ; class KCHMShowWaitCursor { public: KCHMShowWaitCursor () { TQApplication::setOverrideCursor( TQCursor(TQt::WaitCursor) ); } ~KCHMShowWaitCursor () { TQApplication::restoreOverrideCursor(); } }; LCHMFileImpl::LCHMFileImpl( ) { m_chmFile = NULL; m_home = m_filename = m_home = m_topicsFile = m_indexFile = m_font = TQString(); m_entityDecodeMap.clear(); m_textCodec = 0; m_textCodecForSpecialFiles = 0; m_detectedLCID = 0; m_currentEncoding = 0; } LCHMFileImpl::~ LCHMFileImpl( ) { closeAll(); } bool LCHMFileImpl::loadFile( const TQString & archiveName ) { TQString filename; // If the file has a file:// prefix, remove it if ( archiveName.startsWith( "file://" ) ) filename = archiveName.mid( 7 ); // strip it else filename = archiveName; if( m_chmFile ) closeAll(); m_chmFile = chm_open( TQFile::encodeName(filename) ); if ( m_chmFile == NULL ) return false; m_filename = filename; // Reset encoding m_textCodec = 0; m_textCodecForSpecialFiles = 0; m_currentEncoding = 0; // Get information from /#WINDOWS and /#SYSTEM files (encoding, title, context file and so) // and guess the encoding getInfoFromWindows(); getInfoFromSystem(); guessTextEncoding(); // Check whether the search tables are present if ( ResolveObject("/#TOPICS", &m_chmTOPICS) && ResolveObject("/#STRINGS", &m_chmSTRINGS) && ResolveObject("/#URLTBL", &m_chmURLTBL) && ResolveObject("/#URLSTR", &m_chmURLSTR) ) { m_lookupTablesValid = true; fillTopicsUrlMap(); } else m_lookupTablesValid = false; if ( m_lookupTablesValid && ResolveObject ("/$FIftiMain", &m_chmFIftiMain) ) m_searchAvailable = true; else m_searchAvailable = false; return true; } void LCHMFileImpl::closeAll( ) { if ( m_chmFile == NULL ) return; chm_close( m_chmFile ); m_chmFile = NULL; m_home = m_filename = m_home = m_topicsFile = m_indexFile = m_font = TQString(); m_entityDecodeMap.clear(); m_textCodec = 0; m_textCodecForSpecialFiles = 0; m_detectedLCID = 0; m_currentEncoding = 0; } TQString LCHMFileImpl::decodeEntity( const TQString & entity ) { // Set up m_entityDecodeMap characters according to current textCodec if ( m_entityDecodeMap.isEmpty() ) { m_entityDecodeMap["AElig"] = encodeWithCurrentCodec ("\306"); // capital AE diphthong (ligature) m_entityDecodeMap["Aacute"] = encodeWithCurrentCodec ("\301"); // capital A, acute accent m_entityDecodeMap["Acirc"] = encodeWithCurrentCodec ("\302"); // capital A, circumflex accent m_entityDecodeMap["Agrave"] = encodeWithCurrentCodec ("\300"); // capital A, grave accent m_entityDecodeMap["Aring"] = encodeWithCurrentCodec ("\305"); // capital A, ring m_entityDecodeMap["Atilde"] = encodeWithCurrentCodec ("\303"); // capital A, tilde m_entityDecodeMap["Auml"] = encodeWithCurrentCodec ("\304"); // capital A, dieresis or umlaut mark m_entityDecodeMap["Ccedil"] = encodeWithCurrentCodec ("\307"); // capital C, cedilla m_entityDecodeMap["Dstrok"] = encodeWithCurrentCodec ("\320"); // whatever m_entityDecodeMap["ETH"] = encodeWithCurrentCodec ("\320"); // capital Eth, Icelandic m_entityDecodeMap["Eacute"] = encodeWithCurrentCodec ("\311"); // capital E, acute accent m_entityDecodeMap["Ecirc"] = encodeWithCurrentCodec ("\312"); // capital E, circumflex accent m_entityDecodeMap["Egrave"] = encodeWithCurrentCodec ("\310"); // capital E, grave accent m_entityDecodeMap["Euml"] = encodeWithCurrentCodec ("\313"); // capital E, dieresis or umlaut mark m_entityDecodeMap["Iacute"] = encodeWithCurrentCodec ("\315"); // capital I, acute accent m_entityDecodeMap["Icirc"] = encodeWithCurrentCodec ("\316"); // capital I, circumflex accent m_entityDecodeMap["Igrave"] = encodeWithCurrentCodec ("\314"); // capital I, grave accent m_entityDecodeMap["Iuml"] = encodeWithCurrentCodec ("\317"); // capital I, dieresis or umlaut mark m_entityDecodeMap["Ntilde"] = encodeWithCurrentCodec ("\321"); // capital N, tilde m_entityDecodeMap["Oacute"] = encodeWithCurrentCodec ("\323"); // capital O, acute accent m_entityDecodeMap["Ocirc"] = encodeWithCurrentCodec ("\324"); // capital O, circumflex accent m_entityDecodeMap["Ograve"] = encodeWithCurrentCodec ("\322"); // capital O, grave accent m_entityDecodeMap["Oslash"] = encodeWithCurrentCodec ("\330"); // capital O, slash m_entityDecodeMap["Otilde"] = encodeWithCurrentCodec ("\325"); // capital O, tilde m_entityDecodeMap["Ouml"] = encodeWithCurrentCodec ("\326"); // capital O, dieresis or umlaut mark m_entityDecodeMap["THORN"] = encodeWithCurrentCodec ("\336"); // capital THORN, Icelandic m_entityDecodeMap["Uacute"] = encodeWithCurrentCodec ("\332"); // capital U, acute accent m_entityDecodeMap["Ucirc"] = encodeWithCurrentCodec ("\333"); // capital U, circumflex accent m_entityDecodeMap["Ugrave"] = encodeWithCurrentCodec ("\331"); // capital U, grave accent m_entityDecodeMap["Uuml"] = encodeWithCurrentCodec ("\334"); // capital U, dieresis or umlaut mark m_entityDecodeMap["Yacute"] = encodeWithCurrentCodec ("\335"); // capital Y, acute accent m_entityDecodeMap["OElig"] = encodeWithCurrentCodec ("\338"); // capital Y, acute accent m_entityDecodeMap["oelig"] = encodeWithCurrentCodec ("\339"); // capital Y, acute accent m_entityDecodeMap["aacute"] = encodeWithCurrentCodec ("\341"); // small a, acute accent m_entityDecodeMap["acirc"] = encodeWithCurrentCodec ("\342"); // small a, circumflex accent m_entityDecodeMap["aelig"] = encodeWithCurrentCodec ("\346"); // small ae diphthong (ligature) m_entityDecodeMap["agrave"] = encodeWithCurrentCodec ("\340"); // small a, grave accent m_entityDecodeMap["aring"] = encodeWithCurrentCodec ("\345"); // small a, ring m_entityDecodeMap["atilde"] = encodeWithCurrentCodec ("\343"); // small a, tilde m_entityDecodeMap["auml"] = encodeWithCurrentCodec ("\344"); // small a, dieresis or umlaut mark m_entityDecodeMap["ccedil"] = encodeWithCurrentCodec ("\347"); // small c, cedilla m_entityDecodeMap["eacute"] = encodeWithCurrentCodec ("\351"); // small e, acute accent m_entityDecodeMap["ecirc"] = encodeWithCurrentCodec ("\352"); // small e, circumflex accent m_entityDecodeMap["Scaron"] = encodeWithCurrentCodec ("\352"); // small e, circumflex accent m_entityDecodeMap["egrave"] = encodeWithCurrentCodec ("\350"); // small e, grave accent m_entityDecodeMap["eth"] = encodeWithCurrentCodec ("\360"); // small eth, Icelandic m_entityDecodeMap["euml"] = encodeWithCurrentCodec ("\353"); // small e, dieresis or umlaut mark m_entityDecodeMap["iacute"] = encodeWithCurrentCodec ("\355"); // small i, acute accent m_entityDecodeMap["icirc"] = encodeWithCurrentCodec ("\356"); // small i, circumflex accent m_entityDecodeMap["igrave"] = encodeWithCurrentCodec ("\354"); // small i, grave accent m_entityDecodeMap["iuml"] = encodeWithCurrentCodec ("\357"); // small i, dieresis or umlaut mark m_entityDecodeMap["ntilde"] = encodeWithCurrentCodec ("\361"); // small n, tilde m_entityDecodeMap["oacute"] = encodeWithCurrentCodec ("\363"); // small o, acute accent m_entityDecodeMap["ocirc"] = encodeWithCurrentCodec ("\364"); // small o, circumflex accent m_entityDecodeMap["ograve"] = encodeWithCurrentCodec ("\362"); // small o, grave accent m_entityDecodeMap["oslash"] = encodeWithCurrentCodec ("\370"); // small o, slash m_entityDecodeMap["otilde"] = encodeWithCurrentCodec ("\365"); // small o, tilde m_entityDecodeMap["ouml"] = encodeWithCurrentCodec ("\366"); // small o, dieresis or umlaut mark m_entityDecodeMap["szlig"] = encodeWithCurrentCodec ("\337"); // small sharp s, German (sz ligature) m_entityDecodeMap["thorn"] = encodeWithCurrentCodec ("\376"); // small thorn, Icelandic m_entityDecodeMap["uacute"] = encodeWithCurrentCodec ("\372"); // small u, acute accent m_entityDecodeMap["ucirc"] = encodeWithCurrentCodec ("\373"); // small u, circumflex accent m_entityDecodeMap["ugrave"] = encodeWithCurrentCodec ("\371"); // small u, grave accent m_entityDecodeMap["uuml"] = encodeWithCurrentCodec ("\374"); // small u, dieresis or umlaut mark m_entityDecodeMap["yacute"] = encodeWithCurrentCodec ("\375"); // small y, acute accent m_entityDecodeMap["yuml"] = encodeWithCurrentCodec ("\377"); // small y, dieresis or umlaut mark m_entityDecodeMap["iexcl"] = encodeWithCurrentCodec ("\241"); m_entityDecodeMap["cent"] = encodeWithCurrentCodec ("\242"); m_entityDecodeMap["pound"] = encodeWithCurrentCodec ("\243"); m_entityDecodeMap["curren"] = encodeWithCurrentCodec ("\244"); m_entityDecodeMap["yen"] = encodeWithCurrentCodec ("\245"); m_entityDecodeMap["brvbar"] = encodeWithCurrentCodec ("\246"); m_entityDecodeMap["sect"] = encodeWithCurrentCodec ("\247"); m_entityDecodeMap["uml"] = encodeWithCurrentCodec ("\250"); m_entityDecodeMap["ordf"] = encodeWithCurrentCodec ("\252"); m_entityDecodeMap["laquo"] = encodeWithCurrentCodec ("\253"); m_entityDecodeMap["not"] = encodeWithCurrentCodec ("\254"); m_entityDecodeMap["shy"] = encodeWithCurrentCodec ("\255"); m_entityDecodeMap["macr"] = encodeWithCurrentCodec ("\257"); m_entityDecodeMap["deg"] = encodeWithCurrentCodec ("\260"); m_entityDecodeMap["plusmn"] = encodeWithCurrentCodec ("\261"); m_entityDecodeMap["sup1"] = encodeWithCurrentCodec ("\271"); m_entityDecodeMap["sup2"] = encodeWithCurrentCodec ("\262"); m_entityDecodeMap["sup3"] = encodeWithCurrentCodec ("\263"); m_entityDecodeMap["acute"] = encodeWithCurrentCodec ("\264"); m_entityDecodeMap["micro"] = encodeWithCurrentCodec ("\265"); m_entityDecodeMap["para"] = encodeWithCurrentCodec ("\266"); m_entityDecodeMap["middot"] = encodeWithCurrentCodec ("\267"); m_entityDecodeMap["cedil"] = encodeWithCurrentCodec ("\270"); m_entityDecodeMap["ordm"] = encodeWithCurrentCodec ("\272"); m_entityDecodeMap["raquo"] = encodeWithCurrentCodec ("\273"); m_entityDecodeMap["frac14"] = encodeWithCurrentCodec ("\274"); m_entityDecodeMap["frac12"] = encodeWithCurrentCodec ("\275"); m_entityDecodeMap["frac34"] = encodeWithCurrentCodec ("\276"); m_entityDecodeMap["iquest"] = encodeWithCurrentCodec ("\277"); m_entityDecodeMap["times"] = encodeWithCurrentCodec ("\327"); m_entityDecodeMap["divide"] = encodeWithCurrentCodec ("\367"); m_entityDecodeMap["copy"] = encodeWithCurrentCodec ("\251"); // copyright sign m_entityDecodeMap["reg"] = encodeWithCurrentCodec ("\256"); // registered sign m_entityDecodeMap["nbsp"] = encodeWithCurrentCodec ("\240"); // non breaking space m_entityDecodeMap["fnof"] = TQChar((unsigned short) 402); m_entityDecodeMap["Delta"] = TQChar((unsigned short) 916); m_entityDecodeMap["Pi"] = TQChar((unsigned short) 928); m_entityDecodeMap["Sigma"] = TQChar((unsigned short) 931); m_entityDecodeMap["beta"] = TQChar((unsigned short) 946); m_entityDecodeMap["gamma"] = TQChar((unsigned short) 947); m_entityDecodeMap["delta"] = TQChar((unsigned short) 948); m_entityDecodeMap["eta"] = TQChar((unsigned short) 951); m_entityDecodeMap["theta"] = TQChar((unsigned short) 952); m_entityDecodeMap["lambda"] = TQChar((unsigned short) 955); m_entityDecodeMap["mu"] = TQChar((unsigned short) 956); m_entityDecodeMap["nu"] = TQChar((unsigned short) 957); m_entityDecodeMap["pi"] = TQChar((unsigned short) 960); m_entityDecodeMap["rho"] = TQChar((unsigned short) 961); m_entityDecodeMap["lsquo"] = TQChar((unsigned short) 8216); m_entityDecodeMap["rsquo"] = TQChar((unsigned short) 8217); m_entityDecodeMap["rdquo"] = TQChar((unsigned short) 8221); m_entityDecodeMap["bdquo"] = TQChar((unsigned short) 8222); m_entityDecodeMap["trade"] = TQChar((unsigned short) 8482); m_entityDecodeMap["ldquo"] = TQChar((unsigned short) 8220); m_entityDecodeMap["ndash"] = TQChar((unsigned short) 8211); m_entityDecodeMap["mdash"] = TQChar((unsigned short) 8212); m_entityDecodeMap["bull"] = TQChar((unsigned short) 8226); m_entityDecodeMap["hellip"] = TQChar((unsigned short) 8230); m_entityDecodeMap["emsp"] = TQChar((unsigned short) 8195); m_entityDecodeMap["rarr"] = TQChar((unsigned short) 8594); m_entityDecodeMap["rArr"] = TQChar((unsigned short) 8658); m_entityDecodeMap["crarr"] = TQChar((unsigned short) 8629); m_entityDecodeMap["le"] = TQChar((unsigned short) 8804); m_entityDecodeMap["ge"] = TQChar((unsigned short) 8805); m_entityDecodeMap["lte"] = TQChar((unsigned short) 8804); // wrong, but used somewhere m_entityDecodeMap["gte"] = TQChar((unsigned short) 8805); // wrong, but used somewhere m_entityDecodeMap["dagger"] = TQChar((unsigned short) 8224); m_entityDecodeMap["Dagger"] = TQChar((unsigned short) 8225); m_entityDecodeMap["euro"] = TQChar((unsigned short) 8364); m_entityDecodeMap["asymp"] = TQChar((unsigned short) 8776); m_entityDecodeMap["isin"] = TQChar((unsigned short) 8712); m_entityDecodeMap["notin"] = TQChar((unsigned short) 8713); m_entityDecodeMap["prod"] = TQChar((unsigned short) 8719); m_entityDecodeMap["ne"] = TQChar((unsigned short) 8800); m_entityDecodeMap["amp"] = "&"; // ampersand m_entityDecodeMap["gt"] = ">"; // greater than m_entityDecodeMap["lt"] = "<"; // less than m_entityDecodeMap["quot"] = "\""; // double quote m_entityDecodeMap["apos"] = "'"; // single quote m_entityDecodeMap["frasl"] = "/"; m_entityDecodeMap["minus"] = "-"; m_entityDecodeMap["oplus"] = "+"; m_entityDecodeMap["Prime"] = "\""; } // If entity is an ASCII code like 〽 - just decode it if ( entity[0] == '#' ) { bool valid; unsigned int ascode = entity.mid(1).toUInt( &valid ); if ( !valid ) { tqWarning ( "LCHMFileImpl::decodeEntity: could not decode HTML entity '%s'", entity.ascii() ); return TQString(); } return (TQString) (TQChar( ascode )); } else { TQMap::const_iterator it = m_entityDecodeMap.find( entity ); if ( it == m_entityDecodeMap.end() ) { tqWarning ("LCHMFileImpl::decodeEntity: could not decode HTML entity '%s'", entity.ascii()); return TQString(); } return *it; } } inline int LCHMFileImpl::findStringInQuotes (const TQString& tag, int offset, TQString& value, bool firstquote, bool decodeentities) { int qbegin = tag.find ('"', offset); if ( qbegin == -1 ) tqFatal ("LCHMFileImpl::findStringInQuotes: cannot find first quote in tag: '%s'", tag.ascii()); int qend = firstquote ? tag.find ('"', qbegin + 1) : tag.findRev ('"'); if ( qend == -1 || qend <= qbegin ) tqFatal ("LCHMFileImpl::findStringInQuotes: cannot find last quote in tag: '%s'", tag.ascii()); // If we do not need to decode HTML entities, just return. if ( decodeentities ) { TQString htmlentity = TQString(); bool fill_entity = false; value.reserve (qend - qbegin); // to avoid multiple memory allocations for ( int i = qbegin + 1; i < qend; i++ ) { if ( !fill_entity ) { if ( tag[i] == '&' ) // HTML entity starts fill_entity = true; else value.append (tag[i]); } else { if ( tag[i] == ';' ) // HTML entity ends { // If entity is an ASCII code, just decode it TQString decode = decodeEntity( htmlentity ); if ( decode.isNull() ) break; value.append ( decode ); htmlentity = TQString(); fill_entity = false; } else htmlentity.append (tag[i]); } } } else value = tag.mid (qbegin + 1, qend - qbegin - 1); return qend + 1; } bool LCHMFileImpl::searchWord (const TQString& text, bool wholeWords, bool titlesOnly, LCHMSearchProgressResults& results, bool phrase_search) { bool partial = false; if ( text.isEmpty() || !m_searchAvailable ) return false; TQString searchword = (TQString) convertSearchWord (text); #define FTS_HEADER_LEN 0x32 unsigned char header[FTS_HEADER_LEN]; if ( RetrieveObject (&m_chmFIftiMain, header, 0, FTS_HEADER_LEN) == 0 ) return false; unsigned char doc_index_s = header[0x1E], doc_index_r = header[0x1F]; unsigned char code_count_s = header[0x20], code_count_r = header[0x21]; unsigned char loc_codes_s = header[0x22], loc_codes_r = header[0x23]; if(doc_index_s != 2 || code_count_s != 2 || loc_codes_s != 2) { // Don't know how to use values other than 2 yet. Maybe next chmspec. return false; } unsigned char* cursor32 = header + 0x14; u_int32_t node_offset = UINT32ARRAY(cursor32); cursor32 = header + 0x2e; u_int32_t node_len = UINT32ARRAY(cursor32); unsigned char* cursor16 = header + 0x18; u_int16_t tree_depth = UINT16ARRAY(cursor16); unsigned char word_len, pos; TQString word; u_int32_t i = sizeof(u_int16_t); u_int16_t free_space; TQMemArray buffer(node_len); node_offset = GetLeafNodeOffset (searchword, node_offset, node_len, tree_depth); if ( !node_offset ) return false; do { // got a leaf node here. if ( RetrieveObject (&m_chmFIftiMain, buffer.data(), node_offset, node_len) == 0 ) return false; cursor16 = buffer.data() + 6; free_space = UINT16ARRAY(cursor16); i = sizeof(u_int32_t) + sizeof(u_int16_t) + sizeof(u_int16_t); u_int64_t wlc_count, wlc_size; u_int32_t wlc_offset; while (i < node_len - free_space) { word_len = *(buffer.data() + i); pos = *(buffer.data() + i + 1); char *wrd_buf = new char[word_len]; memcpy (wrd_buf, buffer.data() + i + 2, word_len - 1); wrd_buf[word_len - 1] = 0; if ( pos == 0 ) word = wrd_buf; else word = word.mid (0, pos) + wrd_buf; delete[] wrd_buf; i += 2 + word_len; unsigned char title = *(buffer.data() + i - 1); size_t encsz; wlc_count = be_encint (buffer.data() + i, encsz); i += encsz; cursor32 = buffer.data() + i; wlc_offset = UINT32ARRAY(cursor32); i += sizeof(u_int32_t) + sizeof(u_int16_t); wlc_size = be_encint (buffer.data() + i, encsz); i += encsz; cursor32 = buffer.data(); node_offset = UINT32ARRAY(cursor32); if ( !title && titlesOnly ) continue; if ( wholeWords && searchword == word ) return ProcessWLC(wlc_count, wlc_size, wlc_offset, doc_index_s, doc_index_r,code_count_s, code_count_r, loc_codes_s, loc_codes_r, results, phrase_search); if ( !wholeWords ) { if ( word.startsWith (searchword)) { partial = true; ProcessWLC(wlc_count, wlc_size, wlc_offset, doc_index_s, doc_index_r,code_count_s, code_count_r, loc_codes_s, loc_codes_r, results, phrase_search); } else if ( TQString::compare (searchword, word.mid(0, searchword.length())) < -1 ) break; } } } while ( !wholeWords && word.startsWith (searchword) && node_offset ); return partial; } bool LCHMFileImpl::ResolveObject(const TQString& fileName, chmUnitInfo *ui) const { return m_chmFile != NULL && ::chm_resolve_object(m_chmFile, fileName.ascii(), ui) == CHM_RESOLVE_SUCCESS; } size_t LCHMFileImpl::RetrieveObject(const chmUnitInfo *ui, unsigned char *buffer, LONGUINT64 fileOffset, LONGINT64 bufferSize) const { #if USE_BUILTIN_CHMLIB return ::chm_retrieve_object(m_chmFile, ui, buffer, fileOffset, bufferSize); #else return ::chm_retrieve_object(m_chmFile, const_cast(ui), buffer, fileOffset, bufferSize); #endif } inline u_int32_t LCHMFileImpl::GetLeafNodeOffset(const TQString& text, u_int32_t initialOffset, u_int32_t buffSize, u_int16_t treeDepth) { u_int32_t test_offset = 0; unsigned char* cursor16, *cursor32; unsigned char word_len, pos; u_int32_t i = sizeof(u_int16_t); TQMemArray buffer(buffSize); TQString word; while(--treeDepth) { if ( initialOffset == test_offset ) return 0; test_offset = initialOffset; if ( RetrieveObject (&m_chmFIftiMain, buffer.data(), initialOffset, buffSize) == 0 ) return 0; cursor16 = buffer.data(); u_int16_t free_space = UINT16ARRAY(cursor16); while (i < buffSize - free_space ) { word_len = *(buffer.data() + i); pos = *(buffer.data() + i + 1); char *wrd_buf = new char[word_len]; memcpy ( wrd_buf, buffer.data() + i + 2, word_len - 1 ); wrd_buf[word_len - 1] = 0; if ( pos == 0 ) word = wrd_buf; else word = word.mid(0, pos) + wrd_buf; delete[] wrd_buf; if ( text <= word ) { cursor32 = buffer.data() + i + word_len + 1; initialOffset = UINT32ARRAY(cursor32); break; } i += word_len + sizeof(unsigned char) + sizeof(u_int32_t) + sizeof(u_int16_t); } } if ( initialOffset == test_offset ) return 0; return initialOffset; } inline bool LCHMFileImpl::ProcessWLC (u_int64_t wlc_count, u_int64_t wlc_size, u_int32_t wlc_offset, unsigned char ds, unsigned char dr, unsigned char cs, unsigned char cr, unsigned char ls, unsigned char lr, LCHMSearchProgressResults& results, bool phrase_search) { int wlc_bit = 7; u_int64_t index = 0, count; size_t length, off = 0; TQMemArray buffer (wlc_size); unsigned char *cursor32; unsigned char entry[TOPICS_ENTRY_LEN]; unsigned char combuf[13]; if ( RetrieveObject (&m_chmFIftiMain, buffer.data(), wlc_offset, wlc_size) == 0 ) return false; for ( u_int64_t i = 0; i < wlc_count; ++i ) { if ( wlc_bit != 7 ) { ++off; wlc_bit = 7; } index += sr_int (buffer.data() + off, &wlc_bit, ds, dr, length); off += length; if ( RetrieveObject (&m_chmTOPICS, entry, index * 16, TOPICS_ENTRY_LEN) == 0 ) return false; LCHMSearchProgressResult progres; cursor32 = entry + 4; progres.titleoff = UINT32ARRAY(cursor32); cursor32 = entry + 8; progres.urloff = UINT32ARRAY(cursor32); if ( RetrieveObject (&m_chmURLTBL, combuf, progres.urloff, 12) == 0 ) return false; cursor32 = combuf + 8; progres.urloff = UINT32ARRAY (cursor32); count = sr_int (buffer.data() + off, &wlc_bit, cs, cr, length); off += length; if ( phrase_search ) progres.offsets.reserve (count); for (u_int64_t j = 0; j < count; ++j) { u_int64_t lcode = sr_int (buffer.data() + off, &wlc_bit, ls, lr, length); off += length; if ( phrase_search ) progres.offsets.push_back (lcode); } results.push_back (progres); } return true; } bool LCHMFileImpl::getInfoFromWindows() { #define WIN_HEADER_LEN 0x08 unsigned char buffer[BUF_SIZE]; unsigned int factor; chmUnitInfo ui; long size = 0; if ( ResolveObject("/#WINDOWS", &ui) ) { if ( !RetrieveObject(&ui, buffer, 0, WIN_HEADER_LEN) ) return false; u_int32_t entries = get_int32_le( (u_int32_t *)(buffer) ); u_int32_t entry_size = get_int32_le( (u_int32_t *)(buffer + 0x04) ); TQByteArray uptr(entries * entry_size); unsigned char* raw = (unsigned char*) uptr.data(); if ( !RetrieveObject (&ui, raw, 8, entries * entry_size) ) return false; if( !ResolveObject ("/#STRINGS", &ui) ) return false; for ( u_int32_t i = 0; i < entries; ++i ) { u_int32_t offset = i * entry_size; u_int32_t off_title = get_int32_le( (u_int32_t *)(raw + offset + 0x14) ); u_int32_t off_home = get_int32_le( (u_int32_t *)(raw + offset + 0x68) ); u_int32_t off_hhc = get_int32_le( (u_int32_t *)(raw + offset + 0x60) ); u_int32_t off_hhk = get_int32_le( (u_int32_t *)(raw + offset + 0x64) ); factor = off_title / 4096; if ( size == 0 ) size = RetrieveObject(&ui, buffer, factor * 4096, BUF_SIZE); if ( size && off_title ) m_title = TQString ((const char*) (buffer + off_title % 4096)); if ( factor != off_home / 4096) { factor = off_home / 4096; size = RetrieveObject (&ui, buffer, factor * 4096, BUF_SIZE); } if ( size && off_home ) m_home = TQString("/") + TQString( (const char*) buffer + off_home % 4096); if ( factor != off_hhc / 4096) { factor = off_hhc / 4096; size = RetrieveObject(&ui, buffer, factor * 4096, BUF_SIZE); } if ( size && off_hhc ) m_topicsFile = TQString("/") + TQString ((const char*) buffer + off_hhc % 4096); if ( factor != off_hhk / 4096) { factor = off_hhk / 4096; size = RetrieveObject (&ui, buffer, factor * 4096, BUF_SIZE); } if ( size && off_hhk ) m_indexFile = TQString("/") + TQString((const char*) buffer + off_hhk % 4096); } } return true; } bool LCHMFileImpl::getInfoFromSystem() { unsigned char buffer[BUF_SIZE]; chmUnitInfo ui; int index = 0; unsigned char* cursor = NULL, *p; u_int16_t value = 0; long size = 0; // Run the first loop to detect the encoding. We need this, because title could be // already encoded in user encoding. Same for file names if ( !ResolveObject ("/#SYSTEM", &ui) ) return false; // Can we pull BUFF_SIZE bytes of the #SYSTEM file? if ( (size = RetrieveObject (&ui, buffer, 4, BUF_SIZE)) == 0 ) return false; buffer[size - 1] = 0; // First loop to detect the encoding for ( index = 0; index < (size - 1 - (long)sizeof(u_int16_t)) ;) { cursor = buffer + index; value = UINT16ARRAY(cursor); switch(value) { case 0: index += 2; cursor = buffer + index; if(m_topicsFile.isEmpty()) m_topicsFile = TQString("/") + TQString((const char*) buffer + index + 2); break; case 1: index += 2; cursor = buffer + index; if(m_indexFile.isEmpty()) m_indexFile = TQString("/") + TQString ((const char*)buffer + index + 2); break; case 2: index += 2; cursor = buffer + index; if(m_home.isEmpty() || m_home == "/") m_home = TQString("/") + TQString ((const char*) buffer + index + 2); break; case 3: index += 2; cursor = buffer + index; m_title = TQString((const char*) (buffer + index + 2)); break; case 4: index += 2; cursor = buffer + index; p = buffer + index + 2; m_detectedLCID = (short) (p[0] | (p[1]<<8)); break; case 6: index += 2; cursor = buffer + index; if(m_topicsFile.isEmpty()) { TQString topicAttempt = "/", tmp; topicAttempt += TQString ((const char*) buffer +index +2); tmp = topicAttempt + ".hhc"; if ( ResolveObject( tmp, &ui) ) m_topicsFile = tmp; tmp = topicAttempt + ".hhk"; if ( ResolveObject( tmp, &ui) ) m_indexFile = tmp; } break; case 16: index += 2; cursor = buffer + index; m_font = TQString ((const char*) buffer + index + 2); break; default: index += 2; cursor = buffer + index; } value = UINT16ARRAY(cursor); index += value + 2; } return true; } TQCString LCHMFileImpl::convertSearchWord( const TQString & src ) { static const char * searchwordtable[128] = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, "s", 0, "oe", 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, "s", 0, "oe", 0, 0, "y", 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, "a", "a", "a", "a", "a", "a", "ae", "c", "e", "e", "e", "e", "i", "i", "i", "i", "d", "n", "o", "o", "o", "o", "o", 0, "o", "u", "u", "u", "u", "y", "\xDE", "ss", "a", "a", "a", "a", "a", "a", "ae", "c", "e", "e", "e", "e", "i", "i", "i", "i", "o", "n", "o", "o", "o", "o", "o", 0, "o", "u", "u", "u", "u", "y", "\xFE", "y" }; if ( !m_textCodec ) return (TQCString) src.lower(); TQCString dest = m_textCodec->fromUnicode (src); for ( unsigned int i = 0; i < dest.size(); i++ ) { if ( dest[i] & 0x80 ) { int index = dest[i] & 0x7F; if ( searchwordtable[index] ) dest.replace (i, 1, searchwordtable[index]); else dest.remove (i, 1); } } return dest.lower(); } void LCHMFileImpl::getSearchResults( const LCHMSearchProgressResults& tempres, TQStringList * results, unsigned int limit_results ) { unsigned char combuf [COMMON_BUF_LEN]; TQMap urlsmap; // used to prevent duplicated urls for ( unsigned int i = 0; i < tempres.size(); i++ ) { if ( urlsmap.find (tempres[i].urloff) != urlsmap.end() ) continue; urlsmap[tempres[i].urloff] = 1; if ( RetrieveObject (&m_chmURLSTR, combuf, tempres[i].urloff + 8, COMMON_BUF_LEN - 1) == 0 ) continue; combuf[COMMON_BUF_LEN - 1] = 0; results->push_back( LCHMUrlFactory::makeURLabsoluteIfNeeded( (const char*) combuf ) ); if ( --limit_results == 0 ) break; } } TQString LCHMFileImpl::normalizeUrl( const TQString & path ) const { int pos = path.find ('#'); TQString fixedpath = pos == -1 ? path : path.left (pos); return LCHMUrlFactory::makeURLabsoluteIfNeeded( fixedpath ); } /* * FIXME: * (from htmlhelp.chm) */ bool LCHMFileImpl::parseFileAndFillArray( const TQString & file, QT34VECTOR< LCHMParsedEntry > * data, bool asIndex ) { TQString src; const int MAX_NEST_DEPTH = 256; if ( !getFileContentAsString( &src, file ) || src.isEmpty() ) return false; KCHMShowWaitCursor wc; /* // Save the index for debugging purposes TQFile outfile( "parsed.htm" ); if ( outfile.open( IO_WriteOnly ) ) { TQTextStream textstream( &outfile ); textstream << src; outfile.close(); } */ unsigned int defaultimagenum = asIndex ? LCHMBookIcons::IMAGE_INDEX : LCHMBookIcons::IMAGE_AUTO; int pos = 0, indent = 0, root_indent_offset = 0; bool in_object = false, root_indent_offset_set = false; LCHMParsedEntry entry; entry.imageid = defaultimagenum; // Split the HHC file by HTML tags int stringlen = src.length(); while ( pos < stringlen && (pos = src.find ('<', pos)) != -1 ) { int i, word_end = 0; for ( i = ++pos; i < stringlen; i++ ) { // If a " or ' is found, skip to the next one. if ( (src[i] == '"' || src[i] == '\'') ) { // find where quote ends, either by another quote, or by '>' symbol (some people don't know HTML) int nextpos = src.find (src[i], i+1); if ( nextpos == -1 && (nextpos = src.find ('>', i+1)) == -1 ) { tqWarning ("LCHMFileImpl::ParseHhcAndFillTree: corrupted TOC: %s", src.mid(i).ascii()); return false; } i = nextpos; } else if ( src[i] == '>' ) break; else if ( !src[i].isLetterOrNumber() && src[i] != '/' && !word_end ) word_end = i; } TQString tagword, tag = src.mid (pos, i - pos); if ( word_end ) tagword = src.mid (pos, word_end - pos).lower(); else tagword = tag.lower(); //tqDebug ("tag: '%s', tagword: '%s'\n", tag.ascii(), tagword.ascii()); // - a topic entry if ( tagword == "object" && tag.find ("text/sitemap", 0, false) != -1 ) in_object = true; else if ( tagword == "/object" && in_object ) { // a topic entry closed. Add a tree item if ( entry.name ) { if ( !root_indent_offset_set ) { root_indent_offset_set = true; root_indent_offset = indent; if ( root_indent_offset > 1 ) tqWarning("CHM has improper index; root indent offset is %d", root_indent_offset); } int real_indent = indent - root_indent_offset; entry.indent = real_indent; data->push_back( entry ); } else { if ( !entry.urls.isEmpty() ) tqDebug ("LCHMFileImpl::ParseAndFillTopicsTree: tag with url \"%s\" is parsed, but name is empty.", entry.urls[0].ascii()); else tqDebug ("LCHMFileImpl::ParseAndFillTopicsTree: tag is parsed, but both name and url are empty."); } entry.name = TQString(); entry.urls.clear(); entry.imageid = defaultimagenum; in_object = false; } else if ( tagword == "param" && in_object ) { // int offset; // strlen("param ") TQString name_pattern = "name=", value_pattern = "value="; TQString pname, pvalue; if ( (offset = tag.find (name_pattern, 0, FALSE)) == -1 ) tqFatal ("LCHMFileImpl::ParseAndFillTopicsTree: bad tag '%s': no name=\n", tag.ascii()); // offset+5 skips 'name=' offset = findStringInQuotes (tag, offset + name_pattern.length(), pname, TRUE, FALSE); pname = pname.lower(); if ( (offset = tag.find (value_pattern, offset, FALSE)) == -1 ) tqFatal ("LCHMFileImpl::ParseAndFillTopicsTree: bad tag '%s': no value=\n", tag.ascii()); // offset+6 skips 'value=' findStringInQuotes (tag, offset + value_pattern.length(), pvalue, FALSE, TRUE); //tqDebug (": name '%s', value '%s'", pname.ascii(), pvalue.ascii()); if ( pname == "name" ) { // Some help files contain duplicate names, where the second name is empty. Work it around by keeping the first one if ( !pvalue.isEmpty() ) entry.name = pvalue; } else if ( pname == "local" ) { // Check for URL duplication TQString url = LCHMUrlFactory::makeURLabsoluteIfNeeded( pvalue ); if ( entry.urls.find( url ) == entry.urls.end() ) entry.urls.push_back( url ); } else if ( pname == "see also" && asIndex && entry.name != pvalue ) entry.urls.push_back (":" + pvalue); else if ( pname == "imagenumber" ) { bool bok; int imgnum = pvalue.toInt (&bok); if ( bok && imgnum >= 0 && imgnum < LCHMBookIcons::MAX_BUILTIN_ICONS ) entry.imageid = imgnum; } } else if ( tagword == "ul" ) // increase indent level { // Fix for buggy help files if ( ++indent >= MAX_NEST_DEPTH ) tqFatal("LCHMFileImpl::ParseAndFillTopicsTree: max nest depth (%d) is reached, error in help file", MAX_NEST_DEPTH); // This intended to fix
      , which was seen in some buggy chm files, // and brokes rootentry[indent-1] check } else if ( tagword == "/ul" ) // decrease indent level { if ( --indent < root_indent_offset ) indent = root_indent_offset; DEBUGPARSER(("
    : new intent is %d\n", indent - root_indent_offset)); } pos = i; } return true; } bool LCHMFileImpl::getFileContentAsBinary( TQByteArray * data, const TQString & url ) const { chmUnitInfo ui; if( !ResolveObject( url, &ui ) ) return false; data->resize( ui.length ); if ( RetrieveObject( &ui, (unsigned char*) data->data(), 0, ui.length ) ) return true; else return false; } bool LCHMFileImpl::getFileContentAsString( TQString * str, const TQString & url, bool internal_encoding ) { TQByteArray buf; if ( getFileContentAsBinary( &buf, url ) ) { unsigned int length = buf.size(); if ( length > 0 ) { buf.resize( length + 1 ); buf [length] = '\0'; *str = internal_encoding ? (TQString)((const char*) buf) : encodeWithCurrentCodec((const char*) buf); return true; } } return false; } TQString LCHMFileImpl::getTopicByUrl( const TQString & url ) const { TQMap< TQString, TQString >::const_iterator it = m_url2topics.find( url ); if ( it == m_url2topics.end() ) return TQString(); return it.data(); } static int chm_enumerator_callback( struct chmFile*, struct chmUnitInfo *ui, void *context ) { ((TQStringList*) context)->push_back( ui->path ); return CHM_ENUMERATOR_CONTINUE; } bool LCHMFileImpl::enumerateFiles( TQStringList * files ) { files->clear(); return chm_enumerate( m_chmFile, CHM_ENUMERATE_ALL, chm_enumerator_callback, files ); } const TQPixmap * LCHMFileImpl::getBookIconPixmap( unsigned int imagenum ) { return m_imagesKeeper.getImage( imagenum ); } bool LCHMFileImpl::setCurrentEncoding( const LCHMTextEncoding * encoding ) { m_currentEncoding = encoding; return changeFileEncoding( encoding->qtcodec ); } bool LCHMFileImpl::guessTextEncoding( ) { const LCHMTextEncoding * enc = 0; if ( !m_detectedLCID || (enc = lookupByLCID (m_detectedLCID)) == 0 ) tqFatal ("Could not detect text encoding by LCID"); if ( changeFileEncoding (enc->qtcodec) ) { m_currentEncoding = enc; return true; } return false; } bool LCHMFileImpl::changeFileEncoding( const char *qtencoding ) { // Encoding could be either simple TQt codepage, or set like CP1251/KOI8, which allows to // set up encodings separately for text (first) and internal files (second) const char * p = strchr( qtencoding, '/' ); if ( p ) { char buf[128]; // much bigger that any encoding possible. No DoS; all encodings are hardcoded. strcpy( buf, qtencoding ); buf[p - qtencoding] = '\0'; m_textCodec = TQTextCodec::codecForName( buf ); if ( !m_textCodec ) { tqWarning( "Could not set up Text Codec for encoding '%s'", buf ); return false; } m_textCodecForSpecialFiles = TQTextCodec::codecForName( p + 1 ); if ( !m_textCodecForSpecialFiles ) { tqWarning( "Could not set up Text Codec for encoding '%s'", p + 1 ); return false; } } else { m_textCodecForSpecialFiles = m_textCodec = TQTextCodec::codecForName (qtencoding); if ( !m_textCodec ) { tqWarning( "Could not set up Text Codec for encoding '%s'", qtencoding ); return false; } } m_entityDecodeMap.clear(); return true; } void LCHMFileImpl::fillTopicsUrlMap() { if ( !m_lookupTablesValid ) return; // Read those tables TQByteArray topics( m_chmTOPICS.length ), urltbl( m_chmURLTBL.length ), urlstr( m_chmURLSTR.length ), strings( m_chmSTRINGS.length ); if ( !RetrieveObject( &m_chmTOPICS, (unsigned char*) topics.data(), 0, m_chmTOPICS.length ) || !RetrieveObject( &m_chmURLTBL, (unsigned char*) urltbl.data(), 0, m_chmURLTBL.length ) || !RetrieveObject( &m_chmURLSTR, (unsigned char*) urlstr.data(), 0, m_chmURLSTR.length ) || !RetrieveObject( &m_chmSTRINGS, (unsigned char*) strings.data(), 0, m_chmSTRINGS.length ) ) return; for ( unsigned int i = 0; i < m_chmTOPICS.length; i += TOPICS_ENTRY_LEN ) { u_int32_t off_title = get_int32_le( (u_int32_t *)(topics.data() + i + 4) ); u_int32_t off_url = get_int32_le( (u_int32_t *)(topics.data() + i + 8) ); off_url = get_int32_le( (u_int32_t *)( urltbl.data() + off_url + 8) ) + 8; TQString url = LCHMUrlFactory::makeURLabsoluteIfNeeded( (const char*) urlstr.data() + off_url ); if ( off_title < strings.size() ) m_url2topics[url] = encodeWithCurrentCodec ( (const char*) strings.data() + off_title ); else m_url2topics[url] = "Untitled"; } } bool LCHMFileImpl::getFileSize(unsigned int * size, const TQString & url) { chmUnitInfo ui; if( !ResolveObject( url, &ui ) ) return false; *size = ui.length; return true; }