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 <o3tl/string_view.hxx>
  23
  24 #include <cstddef>
  25 #include <iostream>
  26 #include <memory>
  27 #include <string>
  28
  29 #include <common.hxx>
  30 #include <po.hxx>
  31 #include <lngmerge.hxx>
  32 #include <utility>
  33
  34 namespace {
  35
  36 bool lcl_isNextGroup(OString &sGroup_out, std::string_view sLineTrim)
  37 {
  38     if (o3tl::starts_with(sLineTrim, "[") && o3tl::ends_with(sLineTrim, "]"))
  39     {
  40         sLineTrim = o3tl::getToken(sLineTrim, 1, '[');
  41         sLineTrim = o3tl::getToken(sLineTrim, 0, ']');
  42         sGroup_out = OString(o3tl::trim(sLineTrim));
  43         return true;
  44     }
  45     return false;
  46 }
  47
  48 void lcl_RemoveUTF8ByteOrderMarker( OString &rString )
  49 {
  50     if( rString.getLength() >= 3 && rString[0] == '\xEF' &&
  51         rString[1] == '\xBB' && rString[2] == '\xBF' )
  52     {
  53         rString = rString.copy(3);
  54     }
  55 }
  56
  57 }
  58
  59
  60
  61 LngParser::LngParser(OString sLngFile)
  62     : sSource(std::move( sLngFile ))
  63 {
  64     std::ifstream aStream(sSource.getStr());
  65     if (!aStream.is_open())
  66         return;
  67
  68     bool bFirstLine = true;
  69     std::string s;
  70     while (std::getline(aStream, s))
  71     {
  72         OString sLine(s.data(), s.length());
  73
  74         if( bFirstLine )
  75         {
  76             // Always remove UTF8 BOM from the first line
  77             lcl_RemoveUTF8ByteOrderMarker( sLine );
  78             bFirstLine = false;
  79         }
  80
  81         mvLines.push_back( sLine );
  82     }
  83     mvLines.push_back( OString() );
  84 }
  85
  86 LngParser::~LngParser()
  87 {
  88 }
  89
  90 void 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 << "\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 < mvLines.size() ) {
 104         sLine = mvLines[ nPos++ ];
 105         while( nPos < mvLines.size() && !isNextGroup( sGroup , sLine ) ) {
 106             ReadLine( sLine , Text );
 107             sID = sGroup;
 108             sLine = mvLines[ 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"_ostr);
 118     }
 119     aPOStream.close();
 120 }
 121
 122 void LngParser::WritePO(PoOfstream &aPOStream,
 123     OStringHashMap &rText_inout, const OString &rActFileName,
 124     const OString &rID)
 125 {
 126     common::writePoEntry(
 127         "Ulfex"_ostr, aPOStream, rActFileName, "LngText",
 128         rID, OString(), rText_inout.count("x-comment"_ostr) ? rText_inout["x-comment"_ostr] : OString(), rText_inout["en-US"_ostr]);
 129 }
 130
 131 bool LngParser::isNextGroup(OString &sGroup_out, std::string_view sLine_in)
 132 {
 133     return lcl_isNextGroup(sGroup_out, o3tl::trim(sLine_in));
 134 }
 135
 136 void LngParser::ReadLine(std::string_view rLine_in,
 137         OStringHashMap &rText_inout)
 138 {
 139     if (!o3tl::starts_with(rLine_in, " *") && !o3tl::starts_with(rLine_in, "/*"))
 140     {
 141         OString sLang(o3tl::trim(o3tl::getToken(rLine_in, 0, '=')));
 142         if (!sLang.isEmpty()) {
 143             rText_inout[sLang] = OString(o3tl::getToken(rLine_in, 1, '"'));
 144         }
 145     }
 146 }
 147
 148 void LngParser::Merge(
 149     const OString &rPOFile,
 150     const OString &rDestinationFile,
 151     std::string_view rLanguage )
 152 {
 153     std::ofstream aDestination(
 154         rDestinationFile.getStr(), std::ios_base::out | std::ios_base::trunc);
 155
 156     MergeDataFile aMergeDataFile( rPOFile, sSource, false, true );
 157     if( o3tl::equalsIgnoreAsciiCase(rLanguage, "ALL") )
 158         aLanguages = aMergeDataFile.GetLanguages();
 159
 160     size_t nPos = 0;
 161     bool bGroup = false;
 162     OString sGroup;
 163
 164     // seek to next group
 165     while ( nPos < mvLines.size() && !bGroup )
 166         bGroup = lcl_isNextGroup(sGroup, o3tl::trim(mvLines[nPos++]));
 167
 168     while ( nPos < mvLines.size()) {
 169         OStringHashMap Text;
 170         OString sID( sGroup );
 171         std::size_t nLastLangPos = 0;
 172
 173         ResData aResData( sID, sSource );
 174         aResData.sResTyp = "LngText"_ostr;
 175         MergeEntrys *pEntrys = aMergeDataFile.GetMergeEntrys( &aResData );
 176         // read languages
 177         bGroup = false;
 178
 179         OString sLanguagesDone;
 180
 181         while ( nPos < mvLines.size() && !bGroup )
 182         {
 183             const OString sLine{ mvLines[nPos].trim() };
 184             if ( lcl_isNextGroup(sGroup, sLine) )
 185             {
 186                 bGroup = true;
 187                 nPos ++;
 188                 sLanguagesDone = ""_ostr;
 189             }
 190             else
 191             {
 192                 sal_Int32 n = 0;
 193                 OString sLang(sLine.getToken(0, '=', n));
 194                 if (n == -1 || static_cast<bool>(sLine.match("/*")))
 195                 {
 196                     ++nPos;
 197                 }
 198                 else
 199                 {
 200                     sLang = sLang.trim();
 201
 202                     OString sSearch{ ";" + sLang + ";" };
 203
 204                     if ( sLanguagesDone.indexOf( sSearch ) != -1 ) {
 205                         mvLines.erase( mvLines.begin() + nPos );
 206                     }
 207                     if( pEntrys )
 208                     {
 209                         if( !sLang.isEmpty() )
 210                         {
 211                             OString sNewText;
 212                             pEntrys->GetText( sNewText, sLang, true );
 213                             if( sLang == "qtz" )
 214                                 continue;
 215
 216                             if ( !sNewText.isEmpty()) {
 217                                 mvLines[ nPos ] = sLang
 218                                     + " = \""
 219                                     // escape quotes, unescape double escaped quotes fdo#56648
 220                                     + sNewText.replaceAll("\""_ostr,"\\\""_ostr).replaceAll("\\\\\""_ostr,"\\\""_ostr)
 221                                     + "\"";
 222                                 Text[ sLang ] = sNewText;
 223                             }
 224                         }
 225                         nLastLangPos = nPos;
 226                         nPos ++;
 227                         sLanguagesDone += sSearch;
 228                     }
 229                     else {
 230                         nLastLangPos = nPos;
 231                         nPos ++;
 232                         sLanguagesDone += sSearch;
 233                     }
 234                 }
 235             }
 236         }
 237         OString sCur;
 238         if ( nLastLangPos )
 239         {
 240             for(size_t n = 0; n < aLanguages.size(); ++n)
 241             {
 242                 sCur = aLanguages[ n ];
 243                 if( !sCur.equalsIgnoreAsciiCase("en-US") && Text[sCur].isEmpty() && pEntrys )
 244                 {
 245
 246                     OString sNewText;
 247                     pEntrys->GetText( sNewText, sCur, true );
 248                     if( sCur == "qtz" )
 249                         continue;
 250                     if ( !sNewText.isEmpty() && sCur != "x-comment")
 251                     {
 252                         const OString sLine { sCur
 253                             + " = \""
 254                             // escape quotes, unescape double escaped quotes fdo#56648
 255                             + sNewText.replaceAll("\""_ostr,"\\\""_ostr).replaceAll("\\\\\""_ostr,"\\\""_ostr)
 256                             + "\"" };
 257
 258                         nLastLangPos++;
 259                         nPos++;
 260
 261                         if ( nLastLangPos < mvLines.size() ) {
 262                             mvLines.insert( mvLines.begin() + nLastLangPos, sLine );
 263                         } else {
 264                             mvLines.push_back( sLine );
 265                         }
 266                     }
 267                 }
 268             }
 269         }
 270     }
 271
 272     for ( size_t i = 0; i < mvLines.size(); ++i )
 273         aDestination << mvLines[i] << '\n';
 274
 275     aDestination.close();
 276 }
 277
 278 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */