l10ntools/source/lngmerge.cxx

   1 /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
   2 /*
   3  * This file is part of the LibreOffice project.
   4  *
   5  * This Source Code Form is subject to the terms of the Mozilla Public
   6  * License, v. 2.0. If a copy of the MPL was not distributed with this
   7  * file, You can obtain one at http://mozilla.org/MPL/2.0/.
   8  *
   9  * This file incorporates work covered by the following license notice:
  10  *
  11  *   Licensed to the Apache Software Foundation (ASF) under one or more
  12  *   contributor license agreements. See the NOTICE file distributed
  13  *   with this work for additional information regarding copyright
  14  *   ownership. The ASF licenses this file to you under the Apache
  15  *   License, Version 2.0 (the "License"); you may not use this file
  16  *   except in compliance with the License. You may obtain a copy of
  17  *   the License at http://www.apache.org/licenses/LICENSE-2.0 .
  18  */
  19
  20 #include "sal/config.h"
  21
  22 #include <cstddef>
  23 #include <fstream>
  24 #include <iterator>
  25 #include <string>
  26
  27 #include "po.hxx"
  28 #include "lngmerge.hxx"
  29
  30 namespace {
  31
  32 OString getBracketedContent(const OString& text) {
  33     return text.getToken(1, '[').getToken(0, ']');
  34 }
  35
  36 static void lcl_RemoveUTF8ByteOrderMarker( OString &rString )
  37 {
  38     if( rString.getLength() >= 3 && rString[0] == '\xEF' &&
  39         rString[1] == '\xBB' && rString[2] == '\xBF' )
  40     {
  41         rString = rString.copy(3);
  42     }
  43 }
  44
  45 }
  46
  47
  48 // class LngParser
  49
  50 LngParser::LngParser(const OString &rLngFile)
  51     : nError( LNG_OK )
  52     , pLines( NULL )
  53     , sSource( rLngFile )
  54 {
  55     pLines = new LngLineList();
  56     std::ifstream aStream(sSource.getStr());
  57     if (aStream.is_open())
  58     {
  59         bool bFirstLine = true;
  60         std::string s;
  61         std::getline(aStream, s);
  62         while (!aStream.eof())
  63         {
  64             OString sLine(s.data(), s.length());
  65
  66             if( bFirstLine )
  67             {
  68                 // Always remove UTF8 BOM from the first line
  69                 lcl_RemoveUTF8ByteOrderMarker( sLine );
  70                 bFirstLine = false;
  71             }
  72
  73             pLines->push_back( new OString(sLine) );
  74             std::getline(aStream, s);
  75         }
  76         pLines->push_back( new OString() );
  77     }
  78     else
  79         nError = LNG_COULD_NOT_OPEN;
  80 }
  81
  82 LngParser::~LngParser()
  83 {
  84     for ( size_t i = 0, n = pLines->size(); i < n; ++i )
  85         delete (*pLines)[ i ];
  86     pLines->clear();
  87     delete pLines;
  88 }
  89
  90 bool LngParser::CreatePO( const OString &rPOFile )
  91 {
  92     PoOfstream aPOStream( rPOFile, PoOfstream::APP );
  93     if (!aPOStream.isOpen()) {
  94         std::cerr << "Ulfex error: Can't open po file:" << rPOFile.getStr() << "\n";
  95     }
  96
  97     size_t nPos  = 0;
  98     bool bStart = true;
  99     OString sGroup, sLine;
 100     OStringHashMap Text;
 101     OString sID;
 102
 103     while( nPos < pLines->size() ) {
 104         sLine = *(*pLines)[ nPos++ ];
 105         while( nPos < pLines->size() && !isNextGroup( sGroup , sLine ) ) {
 106             ReadLine( sLine , Text );
 107             sID = sGroup;
 108             sLine = *(*pLines)[ nPos++ ];
 109         }
 110         if( bStart ) {
 111             bStart = false;
 112             sID = sGroup;
 113         }
 114         else {
 115             WritePO( aPOStream , Text , sSource , sID );
 116         }
 117         Text.erase("x-comment");
 118     }
 119     aPOStream.close();
 120     return true;
 121 }
 122
 123 void LngParser::WritePO(PoOfstream &aPOStream,
 124     OStringHashMap &rText_inout, const OString &rActFileName,
 125     const OString &rID)
 126 {
 127
 128    bool bExport = true;
 129    if ( bExport )
 130     {
 131         common::writePoEntry(
 132             "Ulfex", aPOStream, rActFileName, "LngText",
 133             rID, OString(), rText_inout.count("x-comment") ? rText_inout["x-comment"] : OString(), rText_inout["en-US"]);
 134    }
 135 }
 136
 137 bool LngParser::isNextGroup(OString &sGroup_out, const OString &sLine_in)
 138 {
 139     const OString sLineTrim = sLine_in.trim();
 140     if (sLineTrim.startsWith("[") && sLineTrim.endsWith("]"))
 141     {
 142         sGroup_out = getBracketedContent(sLineTrim).trim();
 143         return true;
 144     }
 145     return false;
 146 }
 147
 148 void LngParser::ReadLine(const OString &rLine_in,
 149         OStringHashMap &rText_inout)
 150 {
 151     if (!rLine_in.match(" *") && !rLine_in.match("/*"))
 152     {
 153         OString sLang(rLine_in.getToken(0, '=').trim());
 154         if (!sLang.isEmpty()) {
 155             OString sText(rLine_in.getToken(1, '"'));
 156             rText_inout[sLang] = sText;
 157         }
 158     }
 159 }
 160
 161 bool LngParser::Merge(
 162     const OString &rPOFile,
 163     const OString &rDestinationFile,
 164     const OString &rLanguage )
 165 {
 166     std::ofstream aDestination(
 167         rDestinationFile.getStr(), std::ios_base::out | std::ios_base::trunc);
 168     if (!aDestination.is_open()) {
 169         nError = LNG_COULD_NOT_OPEN;
 170     }
 171     nError = LNG_OK;
 172
 173     MergeDataFile aMergeDataFile( rPOFile, sSource, false, true );
 174     if( rLanguage.equalsIgnoreAsciiCase("ALL") )
 175         aLanguages = aMergeDataFile.GetLanguages();
 176
 177     size_t nPos = 0;
 178     bool bGroup = false;
 179     OString sGroup;
 180
 181     // seek to next group
 182     while ( nPos < pLines->size() && !bGroup )
 183     {
 184         OString sLine( *(*pLines)[ nPos ] );
 185         sLine = sLine.trim();
 186         if ( sLine.startsWith("[") && sLine.endsWith("]") )
 187         {
 188             sGroup = getBracketedContent(sLine).trim();
 189             bGroup = true;
 190         }
 191         nPos ++;
 192     }
 193
 194     while ( nPos < pLines->size()) {
 195         OStringHashMap Text;
 196         OString sID( sGroup );
 197         std::size_t nLastLangPos = 0;
 198
 199         ResData  *pResData = new ResData( sID, sSource );
 200         pResData->sResTyp = "LngText";
 201         MergeEntrys *pEntrys = aMergeDataFile.GetMergeEntrys( pResData );
 202         // read languages
 203         bGroup = false;
 204
 205         OString sLanguagesDone;
 206
 207         while ( nPos < pLines->size() && !bGroup )
 208         {
 209             OString sLine( *(*pLines)[ nPos ] );
 210             sLine = sLine.trim();
 211             if ( sLine.startsWith("[") && sLine.endsWith("]") )
 212             {
 213                 sGroup = getBracketedContent(sLine).trim();
 214                 bGroup = true;
 215                 nPos ++;
 216                 sLanguagesDone = "";
 217             }
 218             else
 219             {
 220                 sal_Int32 n = 0;
 221                 OString sLang(sLine.getToken(0, '=', n));
 222                 if (n == -1 || static_cast<bool>(sLine.match("/*")))
 223                 {
 224                     ++nPos;
 225                 }
 226                 else
 227                 {
 228                     sLang = sLang.trim();
 229
 230                     OString sSearch( ";" );
 231                     sSearch += sLang;
 232                     sSearch += ";";
 233
 234                     if (( sLanguagesDone.indexOf( sSearch ) != -1 )) {
 235                         LngLineList::iterator it = pLines->begin();
 236                         std::advance( it, nPos );
 237                         pLines->erase( it );
 238                     }
 239                     if( pEntrys )
 240                     {
 241                         if( !sLang.isEmpty() )
 242                         {
 243                             OString sNewText;
 244                             pEntrys->GetText( sNewText, STRING_TYP_TEXT, sLang, true );
 245                             if( sLang == "qtz" )
 246                                 continue;
 247
 248                             if ( !sNewText.isEmpty()) {
 249                                 OString *pLine = (*pLines)[ nPos ];
 250
 251                                 OString sText1( sLang );
 252                                 sText1 += " = \"";
 253                                 // escape quotes, unescape double escaped quotes fdo#56648
 254                                 sText1 += sNewText.replaceAll("\"","\\\"").replaceAll("\\\\\"","\\\"");
 255                                 sText1 += "\"";
 256                                 *pLine = sText1;
 257                                 Text[ sLang ] = sNewText;
 258                             }
 259                         }
 260                         nLastLangPos = nPos;
 261                         nPos ++;
 262                         sLanguagesDone += sSearch;
 263                     }
 264                     else {
 265                         nLastLangPos = nPos;
 266                         nPos ++;
 267                         sLanguagesDone += sSearch;
 268                     }
 269                 }
 270             }
 271         }
 272         OString sCur;
 273         if ( nLastLangPos )
 274         {
 275             for(size_t n = 0; n < aLanguages.size(); ++n)
 276             {
 277                 sCur = aLanguages[ n ];
 278                 if( !sCur.equalsIgnoreAsciiCase("en-US") && Text[sCur].isEmpty() && pEntrys )
 279                 {
 280
 281                     OString sNewText;
 282                     pEntrys->GetText( sNewText, STRING_TYP_TEXT, sCur, true );
 283                     if( sCur == "qtz" )
 284                         continue;
 285                     if ( !sNewText.isEmpty() && sCur != "x-comment")
 286                     {
 287                         OString sLine;
 288                         sLine += sCur;
 289                         sLine += " = \"";
 290                         // escape quotes, unescape double escaped quotes fdo#56648
 291                         sLine += sNewText.replaceAll("\"","\\\"").replaceAll("\\\\\"","\\\"");
 292                         sLine += "\"";
 293
 294                         nLastLangPos++;
 295                         nPos++;
 296
 297                         if ( nLastLangPos < pLines->size() ) {
 298                             LngLineList::iterator it = pLines->begin();
 299                             std::advance( it, nLastLangPos );
 300                             pLines->insert( it, new OString(sLine) );
 301                         } else {
 302                             pLines->push_back( new OString(sLine) );
 303                         }
 304                     }
 305                 }
 306             }
 307         }
 308
 309         delete pResData;
 310     }
 311
 312     for ( size_t i = 0; i < pLines->size(); ++i )
 313         aDestination << (*pLines)[i]->getStr() << '\n';
 314
 315     aDestination.close();
 316     return true;
 317 }
 318
 319 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */