Laboratoire ICAR » Plateforme TXM

// Copyright © 2010-2013 ENS de Lyon.

// Copyright © 2007-2010 ENS de Lyon, CNRS, INRP, University of

// Lyon 2, University of Franche-Comté, University of Nice

// Sophia Antipolis, University of Paris 3.

// 

// The TXM platform 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.

// 

// The TXM platform 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 the TXM platform. If not, see

// http://www.gnu.org/licenses.

// 

// 

// 

// $LastChangedDate: 2013-11-08 13:38:06 +0100 (ven. 08 nov. 2013) $

// $LastChangedRevision: 2569 $

// $LastChangedBy: mdecorde $ 

//

package org.txm.scripts.importer.corptef;

import org.txm.importer.cwb.*

import org.txm.importer.scripts.xmltxm.*;

import org.txm.scripts.importer.*;

import org.txm.utils.treetagger.TreeTagger;

import org.txm.Toolbox;

// TODO: Auto-generated Javadoc

/**

 * The Class annotate.

 */

class annotate {

        /**

         * Run.

         *

         * @param dir the dir

         */

        public void run(File dir)

        {

                String rootDir = dir.getAbsolutePath()+"/";//"~/xml/discours/src";

                String txmDir = dir.getAbsolutePath()+"/txm/"

                //cleaning

                new File(rootDir,"annotations").deleteDir();

                new File(rootDir,"annotations").mkdir();

                new File(rootDir,"treetagger").deleteDir();

                new File(rootDir,"treetagger").mkdir();

                println txmDir

                List<File> listfiles = new File(txmDir).listFiles();

                for(File teifile : listfiles)

                {

                        println("annotate "+teifile)

                        File modelfile = new File(Toolbox.getPreference(Toolbox.TREETAGGER_MODELS_PATH),"/rgaqcj.par");

                        if(!modelfile.exists())

                        {

                                println "Skipping ANNOTATE: Incorrect modelfile path: "+modelfile;

                                return;

                        }

                        File annotfile = new File(rootDir+"annotations",teifile.getName()+"-STDOFF.xml");

                        File ttsrcfile = new File(rootDir+"treetagger",teifile.getName()+"-src.tt");

                        File ttrezfile = new File(rootDir+"treetagger",teifile.getName()+"-out.tt");

                        //prepare file to be tagged

                        def builder = new BuildTTSrc(teifile.toURL());

                        builder.process(ttsrcfile);

                        //Apply TT

                        if(!new File(Toolbox.getPreference(Toolbox.TREETAGGER_INSTALL_PATH)+"/bin/").exists() || Toolbox.getPreference(Toolbox.TREETAGGER_INSTALL_PATH).length() == 0 )

                        {

                                println("Path to TreeTagger is wrong: "+Toolbox.getPreference(Toolbox.TREETAGGER_INSTALL_PATH)+"/bin/")

                                return;

                        }

                        TreeTagger tt = new TreeTagger(Toolbox.getPreference(Toolbox.TREETAGGER_INSTALL_PATH)+"/bin/");

                        tt.settoken();

                        tt.setlemma();

                        tt.setquiet();

                        tt.setnounknown();

                        tt.setsgml();

                        tt.seteostag("<s>");

                        tt.treetagger( modelfile.getAbsolutePath(), ttsrcfile.getAbsolutePath(), ttrezfile.getAbsolutePath())

                        //create stand-off annotation file

                        //targeted file for annotations

                        String target = teifile.getName()

                        //contains txm:application/txm:commandLine

                        File reportFile = new File(rootDir,"NLPToolsParameters.xml");

                        String respPerson = System.getProperty("user.name");

                        String respId = "txm";

                        String respDesc = "NLP annotation tool";

                        String respDate = "Tue Mar  11 1:02:55 Paris, Madrid 2010";

                        String respWhen = ""

                        String appIdent = "TreeTagger";

                        String appVersion = "3.2";

                        String distributor = "";

                        String publiStmt = """""";

                        String sourceStmt = """""";

                        def types = ["ttpos","ttlemma"];

                        def typesTITLE = ["ttpos","ttlemma"];

                        def typesDesc = ["fr pos","fr lemma"]

                        def typesTAGSET = ["",""]

                        def typesWEB = ["",""]

                        String idform ="w_";

                        String encoding ="UTF-8";

                        def transfo = new CSV2W_ANA();

                        transfo.setAnnotationTypes( types, typesDesc, typesTAGSET, typesWEB, idform);

                        transfo.setResp(respId, respDesc,respDate, respPerson, respWhen);

                        transfo.setApp(appIdent, appVersion);

                        transfo.setTarget(target, reportFile);

                        transfo.setInfos(distributor,  publiStmt, sourceStmt);

                        transfo.process( ttrezfile, annotfile, encoding );

                        //merge into the tei file

                        builder = new AnnotationInjection(teifile.toURL(), annotfile.toURL(), new ArrayList<String>());

                        builder.transfomFile(rootDir,"temp");

                        if (!(teifile.delete() && new File(rootDir,"temp").renameTo(teifile))) println "Warning can't rename file "+new File(rootDir,"temp")+" to "+teifile

                }

        }

        /**

         * The main method.

         *

         * @param args the arguments

         */

        public static void main(String[] args)

        {

                File dir = new File("C:/Documents and Settings/alavrent/TXM/corpora/corptef/")

                new annotate().run(dir);

        }

}