Laboratoire ICAR » Plateforme TXM

import org.txm.importer.xmltxm.AnnotateCQP;

import org.txm.utils.FileUtils;

		boolean fixExistingValues = false; // default behavior is to replace existing values

		if (parameters.get("fix_existing_values") != null) {

			fixExistingValues = "true".equals(parameters.get("fix_existing_values"));

		}

		if (FileUtils.isExtension(xmlFile, "cqp")) {

			AnnotateCQP annotate = new AnnotateCQP();

			return annotate.run(xmlFile, lang, binaryCorpusDirectory, xmlFile.getParentFile());

		} else {

			Annotate annotate = new Annotate();

			return annotate.run(xmlFile, lang, fixExistingValues, binaryCorpusDirectory, xmlFile.getParentFile());

		}

	public boolean run(File f, String lang, boolean fixExistingValues, File binDir, File txmDir) {

		return run(f, lang, binDir, txmDir, fixExistingValues, new String[0], new String[0], annotDir, ptreetaggerDir, treetaggerDir)

	public boolean run(File f, String lang, File binDir, File txmDir, boolean fixExistingValues, String[] properties, String[] options, File annotDir, File ptreetaggerDir, File treetaggerDir) {

		builder = new AnnotationInjection(f.toURL(), annotfile.toURL(), fixExistingValues);

// 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: 2013-12-06 16:57:35 +0100 (Fri, 06 Dec 2013) $

// $LastChangedRevision: 2583 $

// $LastChangedBy: mdecorde $

//

package org.txm.importer.xmltxm

import java.io.File

import java.text.DateFormat

import java.util.Date

import org.txm.Toolbox

import org.txm.importer.*

import org.txm.importer.cwb.*

import org.txm.objects.*

import org.txm.treetagger.core.preferences.TreeTaggerPreferences

import org.txm.utils.ConsoleProgressBar

import org.txm.utils.DeleteDir;

import org.txm.utils.LangDetector;

import org.txm.utils.io.IOUtils

import org.txm.utils.treetagger.TreeTagger

// TODO: Auto-generated Javadoc

/**

 * Annotate and replace the TEI-TXM files of the folder $rootDirFile/txm with TreeTagger.

 * creates $rootDirFile/interp and $rootDirFile/treetagger

 *

 */

class AnnotateCQP {

	/** The debug. */

	boolean debug = true;

	/**

	 * Sets the debug.

	 */

	public void setDebug() { debug=true; }

	String id;

	/**

	 * Apply tt.

	 *

	 * @param ttsrcfile the ttsrcfile

	 * @param ttoutfile the ttoutfile

	 * @param modelfile the modelfile

	 * @return true, if successful

	 */

	public boolean applyTT(File ttsrcfile, File ttoutfile, File modelfile)

	{

		try {

			File infile = ttsrcfile;

			File outfile = ttoutfile;

			// TODO: why org.txm.utils.treetagger.TreeTagger tt = ..., throw IllegalAccessError ???

			String opt = Toolbox.getPreference(Toolbox.TREETAGGER_OPTIONS);

			if (opt == null) opt = "";

			String[] options = opt.split("  ");

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

			tt.settoken();

			tt.setlemma();

			tt.setquiet();

			tt.setsgml();

			tt.setnounknown();

			tt.seteostag("<s>");

			tt.treetagger(modelfile.getAbsolutePath(), infile.getAbsolutePath(), outfile.getAbsolutePath())

			//infile.delete();

		}

		catch(Exception e)

		{

			e.printStackTrace();

			return false;

		}

		return true;

	}

	/**

	 * Run.

	 *

	 * @param rootDirFile the root dir file

	 * @param modelfilename the modelfilename

	 * @return true, if successful

	 */

	public boolean run(File binDir, File txmDir,  String modelfilename)

	{

		//test if modelfile exists

		if(debug) {

			println "rootDirFile "+binDir

			println "txmDir "+txmDir

			println "TREETAGGER INSTALL PATH : "+TreeTaggerPreferences.getInstance().getBoolean(TreeTaggerPreferences.INSTALL_PATH)

			println "TREETAGGER MODELS PATH : "+TreeTaggerPreferences.getInstance().getBoolean(TreeTaggerPreferences.MODELS_PATH)

		}

		//test if the Toolbox know TreeTagger

		if (!new File(Toolbox.getPreference(Toolbox.TREETAGGER_INSTALL_PATH)+"/bin/").exists())

		{

			println("Could not find TreeTagger binaries in "+Toolbox.getPreference(Toolbox.TREETAGGER_INSTALL_PATH)+"/bin/")

			return false;

		}

		//cleaning

		File annotDir = new File(binDir, "annotations")

		annotDir.mkdir();

		//BUILD TT FILE READY TO BE TAGGED

		List<File> files = txmDir.listFiles(IOUtils.HIDDENFILE_FILTER)

		// get model file and check it

		File modelfile = new File(TreeTaggerPreferences.getInstance().getBoolean(TreeTaggerPreferences.MODELS_PATH), modelfilename);

		if (debug)

			println "model file : "+modelfile;

		if (!modelfile.exists()) {

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

			if(System.getProperty("os.name").startsWith("Windows"))

				println "Windows users: Windows might be hiding files extension. To see them, in the explorer parameters."

			return false;

		}

		//APPLY TREETAGGER

		println("Applying $modelfilename TreeTagger model on dir: "+txmDir+ " ("+files.size()+" files)")

		if (files == null || files.size() == 0)

			return false;

		ConsoleProgressBar cpb = new ConsoleProgressBar(files.size());

		for (File f : files) {

			cpb.tick();

			File infile = f;

			File outfile = new File(new File(binDir, "annotations"), f.getName());

			if (outfile.exists() && // outfile exists

				outfile.lastModified() >= infile.lastModified() && // outfile is more recent 

				outfile.length() > infile.length()) { // outfile is bigger

				// skip

			} else {

				if (!applyTT(infile, outfile, modelfile)) {

					System.out.println("Failed to apply treetagger on file "+f);

					return false;

				}

			}

		}

		println("")

//		if (DeleteDir.deleteDirectory(txmDir)) {

//			if (annotDir.renameTo(txmDir)) {

//

//			} else {

//				println "Could not rename  'annotations' directory to 'txm' directory"

//				return false

//			}

//		} else {

//			println "Could not delete txmDir: $txmDir (and could not copy 'annotations' directory"

//		}

		return true;

	}

	public boolean run(File xmlFile, String lang, File binaryCorpusDirectory, File parentFile) {

		System.out.println("AnnotateCQP called from the TreeTaggerEngine.processFile() is not yet implemented.")

		return false;

	}

}

	/**

	 * if set, existing values are replaced

	 */

	boolean fixExistingValues

	public AnnotationInjection(URL url, URL anaurl, boolean fixExistingValues) {

			this.fixExistingValues = fixExistingValues

			String newValue = getNextAnaValue(link, wordId)

			if (fixExistingValues || anaValues.containsKey(link) == null) {

				anaValues.put(link, newValue)

			} else {

				anaValues.put(link, "")

			writer.writeCharacters(anaValues.get(link));

		} else if (flagW && localname.equals("ana")) {