kttsd/filters/stringreplacer/stringreplacerproc.cpp

   1 /***************************************************** vim:set ts=4 sw=4 sts=4:
   2   Generic String Replacement Filter Processing class.
   3   -------------------
   4   Copyright:
   5   (C) 2005 by Gary Cramblitt <garycramblitt@comcast.net>
   6   -------------------
   7   Original author: Gary Cramblitt <garycramblitt@comcast.net>
   8
   9   This program is free software; you can redistribute it and/or modify
  10   it under the terms of the GNU General Public License as published by
  11   the Free Software Foundation; either version 2 of the License, or
  12   (at your option) any later version.
  13
  14   This program is distributed in the hope that it will be useful,
  15   but WITHOUT ANY WARRANTY; without even the implied warranty of
  16   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  17   GNU General Public License for more details.
  18
  19   You should have received a copy of the GNU General Public License
  20   along with this program; if not, write to the Free Software
  21   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  22  ******************************************************************************/
  23
  24 // StringReplacer includes.
  25 #include "stringreplacerproc.h"
  26 #include "stringreplacerproc.moc"
  27
  28 // Qt includes.
  29 #include <QtXml/QDomDocument>
  30 #include <QtCore/QFile>
  31
  32 // KDE includes.
  33 #include <kdebug.h>
  34 #include <klocale.h>
  35 #include <kconfig.h>
  36 #include <kconfiggroup.h>
  37 #include <kglobal.h>
  38 #include <kstandarddirs.h>
  39
  40 // KTTS includes.
  41 #include "filterproc.h"
  42 #include "talkercode.h"
  43
  44 /**
  45  * Constructor.
  46  */
  47 StringReplacerProc::StringReplacerProc( QObject *parent, QVariantList list) :
  48     KttsFilterProc(parent, list)
  49 {
  50 }
  51
  52 /**
  53  * Destructor.
  54  */
  55 /*virtual*/ StringReplacerProc::~StringReplacerProc()
  56 {
  57     m_matchList.clear();
  58     m_substList.clear();
  59 }
  60
  61 /**
  62  * Initialize the filter.
  63  * @param config          Settings object.
  64  * @param configGroup     Settings Group.
  65  * @return                False if filter is not ready to filter.
  66  *
  67  * Note: The parameters are for reading from kttsdrc file.  Plugins may wish to maintain
  68  * separate configuration files of their own.
  69  */
  70 bool StringReplacerProc::init(KConfig* c, const QString& configGroup){
  71     //kDebug() << "StringReplacerProc::init: Running";
  72     QString wordsFilename =
  73         KGlobal::dirs()->saveLocation( "data" ,"kttsd/stringreplacer/", false );
  74     if ( wordsFilename.isEmpty() ) return false;
  75     wordsFilename += configGroup;
  76     KConfigGroup config( c, configGroup );
  77     wordsFilename = config.readEntry( "WordListFile", wordsFilename );
  78
  79     // Open existing word list.
  80     QFile file( wordsFilename );
  81     if ( !file.open( QIODevice::ReadOnly ) )
  82     {
  83         //kDebug() << "StringReplacerProc::init: couldn't open file " << wordsFilename;
  84         return false;
  85     }
  86     QDomDocument doc( "" );
  87     if ( !doc.setContent( &file ) ) {
  88         //kDebug() << "StringReplacerProc::init: couldn't get xml from file " << wordsFilename;
  89         file.close();
  90         return false;
  91     }
  92     file.close();
  93
  94     // Clear list.
  95     m_matchList.clear();
  96     m_substList.clear();
  97
  98     // Name setting.
  99     // QDomNodeList nameList = doc.elementsByTagName( "name" );
 100     // QDomNode nameNode = nameList.item( 0 );
 101     // m_widget->nameLineEdit->setText( nameNode.toElement().text() );
 102
 103     // Language Codes setting.  List may be single element of comma-separated values,
 104     // or multiple elements.
 105     m_languageCodeList.clear();
 106     QDomNodeList languageList = doc.elementsByTagName( "language-code" );
 107     for ( int ndx=0; ndx < languageList.count(); ++ndx )
 108     {
 109         QDomNode languageNode = languageList.item( ndx );
 110         m_languageCodeList += languageNode.toElement().text().split( ',', QString::SkipEmptyParts);
 111     }
 112
 113     // AppId.  Apply this filter only if DCOP appId of application that queued
 114     // the text contains this string.  List may be single element of comma-separated values,
 115     // or multiple elements.
 116     m_appIdList.clear();
 117     QDomNodeList appIdList = doc.elementsByTagName( "appid" );
 118     for ( int ndx=0; ndx < appIdList.count(); ++ndx )
 119     {
 120         QDomNode appIdNode = appIdList.item( ndx );
 121         m_appIdList += appIdNode.toElement().text().split( ',', QString::SkipEmptyParts);
 122     }
 123
 124     // Word list.
 125     QDomNodeList wordList = doc.elementsByTagName("word");
 126     const int wordListCount = wordList.count();
 127     for (int wordIndex = 0; wordIndex < wordListCount; ++wordIndex)
 128     {
 129         QDomNode wordNode = wordList.item(wordIndex);
 130         QDomNodeList propList = wordNode.childNodes();
 131         QString wordType;
 132         QString matchCase = "No"; // Default for old (v<=3.5.3) config files with no <case/>.
 133         QString match;
 134         QString subst;
 135         const int propListCount = propList.count();
 136         for (int propIndex = 0; propIndex < propListCount; ++propIndex)
 137         {
 138             QDomNode propNode = propList.item(propIndex);
 139             QDomElement prop = propNode.toElement();
 140             if (prop.tagName() == "type") wordType = prop.text();
 141             if (prop.tagName() == "case") matchCase = prop.text();
 142             if (prop.tagName() == "match") match = prop.text();
 143             if (prop.tagName() == "subst") subst = prop.text();
 144         }
 145         // Build Regular Expression for each word's match string.
 146         QRegExp rx;
 147         rx.setCaseSensitivity(matchCase == "Yes"?Qt::CaseInsensitive:Qt::CaseSensitive);
 148         if ( wordType == "Word" )
 149         {
 150                 // TODO: Does \b honor strange non-Latin1 encodings?
 151             rx.setPattern( "\\b" + match + "\\b" );
 152         }
 153         else
 154         {
 155             rx.setPattern( match );
 156         }
 157             // Add Regular Expression to list (if valid).
 158         if ( rx.isValid() )
 159         {
 160             m_matchList.append( rx );
 161             m_substList.append( subst );
 162         }
 163     }
 164     return true;
 165 }
 166
 167 /**
 168  * Convert input, returning output.
 169  * @param inputText         Input text.
 170  * @param talkerCode        TalkerCode structure for the talker that KTTSD intends to
 171  *                          use for synthing the text.  Useful for extracting hints about
 172  *                          how to filter the text.  For example, languageCode.
 173  * @param appId             The DCOP appId of the application that queued the text.
 174  *                          Also useful for hints about how to do the filtering.
 175  */
 176 /*virtual*/ QString StringReplacerProc::convert(const QString& inputText, TalkerCode* talkerCode,
 177     const QString& appId)
 178 {
 179     m_wasModified = false;
 180     // If language doesn't match, return input unmolested.
 181     if ( !m_languageCodeList.isEmpty() )
 182     {
 183         QString languageCode = talkerCode->languageCode();
 184         //kDebug() << "StringReplacerProc::convert: converting " << inputText <<
 185         //    " if language code " << languageCode << " matches " << m_languageCodeList << endl;
 186         if ( !m_languageCodeList.contains( languageCode ) )
 187         {
 188             if ( !talkerCode->countryCode().isEmpty() )
 189             {
 190                 languageCode += '_' + talkerCode->countryCode();
 191                  //kDebug() << "StringReplacerProc::convert: converting " << inputText <<
 192                  //   " if language code " << languageCode << " matches " << m_languageCodeList << endl;
 193                 if ( !m_languageCodeList.contains( languageCode ) ) return inputText;
 194             } else return inputText;
 195         }
 196     }
 197     // If appId doesn't match, return input unmolested.
 198     if ( !m_appIdList.isEmpty() )
 199     {
 200          //kDebug() << "StringReplacerProc::convert: converting " << inputText << " if appId "
 201          //    << appId << " matches " << m_appIdList << endl;
 202         bool found = false;
 203         QString appIdStr = appId;
 204         for ( int ndx=0; ndx < m_appIdList.count(); ++ndx )
 205         {
 206             if ( appIdStr.contains(m_appIdList[ndx]) )
 207             {
 208                 found = true;
 209                 break;
 210             }
 211         }
 212         if ( !found )
 213         {
 214              //kDebug() << "StringReplacerProc::convert: appId not found";
 215             return inputText;
 216         }
 217     }
 218     QString newText = inputText;
 219     const int listCount = m_matchList.count();
 220     for ( int index = 0; index < listCount; ++index )
 221     {
 222         //kDebug() << "newtext = " << newText << " matching " << m_matchList[index].pattern() << " replacing with " << m_substList[index];
 223         newText.replace( m_matchList[index], m_substList[index] );
 224     }
 225     m_wasModified = true;
 226     return newText;
 227 }
 228
 229 /**
 230  * Did this filter do anything?  If the filter returns the input as output
 231  * unmolested, it should return False when this method is called.
 232  */
 233 /*virtual*/ bool StringReplacerProc::wasModified() { return m_wasModified; }
 234