Laboratoire ICAR » CCC

eclipse.preferences.version=1

org.eclipse.jdt.core.compiler.codegen.inlineJsrBytecode=enabled

org.eclipse.jdt.core.compiler.codegen.targetPlatform=1.6

org.eclipse.jdt.core.compiler.codegen.unusedLocal=preserve

org.eclipse.jdt.core.compiler.compliance=1.6

org.eclipse.jdt.core.compiler.debug.lineNumber=generate

org.eclipse.jdt.core.compiler.debug.localVariable=generate

org.eclipse.jdt.core.compiler.debug.sourceFile=generate

org.eclipse.jdt.core.compiler.problem.assertIdentifier=error

org.eclipse.jdt.core.compiler.problem.enumIdentifier=error

org.eclipse.jdt.core.compiler.source=1.6

<?xml version="1.0" encoding="UTF-8"?>

<classpath>

	<classpathentry kind="src" path="src"/>

	<classpathentry kind="con" path="org.eclipse.jdt.launching.JRE_CONTAINER/org.eclipse.jdt.internal.debug.ui.launcher.StandardVMType/JavaSE-1.6"/>

	<classpathentry exported="true" kind="con" path="GROOVY_SUPPORT"/>

	<classpathentry exported="true" kind="con" path="GROOVY_DSL_SUPPORT"/>

	<classpathentry kind="lib" path="saxon9he.jar"/>

	<classpathentry kind="output" path="bin"/>

</classpath>

<?xml version="1.0" encoding="UTF-8"?>

<projectDescription>

	<name>CMC2ELAN</name>

	<comment></comment>

	<projects>

	</projects>

	<buildSpec>

		<buildCommand>

			<name>org.eclipse.jdt.core.javabuilder</name>

			<arguments>

			</arguments>

		</buildCommand>

	</buildSpec>

	<natures>

		<nature>org.eclipse.jdt.groovy.core.groovyNature</nature>

		<nature>org.eclipse.jdt.core.javanature</nature>

	</natures>

</projectDescription>

// 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.

import java.text.SimpleDateFormat;

import java.util.Date;

class FixMissingTimings {

	File trsInFile = new File("/home/mdecorde/xml/transana/out.xml")

	File trsOutFile = new File("/home/mdecorde/xml/transana/TrP1S8_08102010.trs")

	public FixMissingTimings(File trsInFile, File trsOutFile) {

		this.trsInFile = trsInFile;

		this.trsOutFile = trsOutFile;

	}

	public boolean process() {

		def slurper = new XmlParser();

		def trs = slurper.parse(trsInFile)

		String currentEnd = null;

		for (def turn : trs.Episode.Section.Turn) {

			def children = turn.children()

			if (children.size() > 0) {

				//println "fixing turn start="+turn.@startTime+" end="+turn.@endTime+" previous end="+currentEnd

				if (currentEnd != null && currentEnd.length() > 0 && turn.@startTime.length() == 0) { // using last Sync

					//println "fixing with previous Turn $currentEnd"

					turn.@startTime = currentEnd

					setStartTime(children[0], currentEnd)

				} else if (!(children[0] instanceof String)) { // no previous Sync, using next Sync

					def start = getStartTime(children[0])

					if (start != null && start.length() > 0)	turn.@startTime = start

				}

				currentEnd = null

				if (!(children[children.size() - 1] instanceof String)) { // this is the last Sync

					def end = getStartTime(children[children.size() - 1])

					if (end != null && end.length() > 0)  {

						turn.@endTime = end

						currentEnd = end;

					}

				}

			}

		}

		def nodes = []

		// get Sync nodes to fix

		for (def section : trs.Episode.Section) { // all Section

			for (def child : section.children()) { // all Section children

				if (!(child instanceof String)) { // is a Tag

					switch(child.name()) {

						case "Turn": // get Sync in Turn

							//nodes << child;

							for (def sync : child.Sync)

								nodes << sync

							break;

						case "Sync":

							nodes << child;

							break;

						default: break;

					}

				}

			}

		}

		//Fixing Sync@time

		for (int i = 0 ; i < nodes.size() ; i++) {

			def node = nodes[i]

			String time = getStartTime(node)	

			if (time.length() == 0) {				

				def list = [];

				String previous;

				if (i > 0) { previous = getStartTime(nodes[i-1]);

				} else { previous = "0.0" }

				String next = null;

				while ((next == null || next.length() == 0) && i <= nodes.size()) {

					list << node

					node = nodes[++i]

					next = getStartTime(node)

				}

				if (next != null) {

					if (node != null) list << node

					fixSyncTimes(list, previous, next)

				} else {

					println "ERROR: no end time in the transcription"

				}

			}

		}

		// fixing startTime of Turn using next Sync.@time

		def turns = trs.Episode.Section.Turn

		for (int i = 0 ; i < turns.size() ; i++) {

			def turn = turns[i]

			def syncs = turn.Sync

			if (syncs.size() > 0) {

				turn.@startTime = syncs[0].@time

			} else {

				println "Error: Turn with no Sync milestone"

			}

		}

		// fixing endTimes of Turn using next Turn.@startTime

		for (int i = 0 ; i < turns.size() ; i++) {

			def turn = turns[i]

			//println "turn start="+turn.@startTime+" end="+turn.@endTime

			if (i < turns.size() - 1) {

				if (turn.@endTime.length() == 0 && turns[i+1].@startTime.length() > 0) {

					//println "fixing turn.@endTime "+turn.@endTime+" with turns[i+1].@startTime "+(turns[i+1]);

					turn.@endTime = turns[i+1].@startTime

				} 

			} else {

				Float e = Float.parseFloat(turn.@startTime)

				e += turn.text().length()*1.0f 

				turn.@endTime = e.toString()

			}

		}

		// fixing endTimes of Section using last Turn.@endTime

		// fixing startTimes of Section using first Turn.@startTime

		def sections = trs.Episode.Section

		for (int i = 0 ; i < sections.size() ; i++) {

			def t = sections[i].Turn

			if (t.size() == 0) continue;

			sections[i].@startTime = t[0].@startTime

			sections[i].@endTime = t[-1].@endTime

		}

		trsOutFile.withWriter("UTF-8") { writer ->

			writer.write('<?xml version="1.0" encoding="UTF-8"?>\n<!DOCTYPE Trans SYSTEM "trans-14.dtd">\n')

			new groovy.util.XmlNodePrinter(new PrintWriter(writer)).print(trs) }

		return true;

	}

	def getStartTime(def node) {

		def ret = null;

		if (node == null) return "0.0"

		if (node.name() == null) return "0.0"

		switch(node.name()) {

			case "Turn":

			case "Section":

			//			println "Turn"

				ret = node.@startTime

				break

			case "Sync":

			//		println "Sync"

				ret = node.@time

				break

			default: break;

		}

		//println "getStartTime "+node.name()+" $ret"

		return ret

	}

	def setStartTime(def node, def value) {

		switch(node.name()) {

			case "Turn":

			case "Section":

				node.@startTime = ""+value

				break;

			case "Sync":

				return node.@time = ""+value

			default: break;

		}

	}

	def fixSyncTimes(def list, def start, def end) {

		//println "Nfix: "+list.size()+" "+list

		//println "$start $end"

		def startf = Double.parseDouble(start)

		def endf = Double.parseDouble(end)

		def delta = (endf-startf)/list.size()

		//println delta

		float c = startf;

		for (int i = 0 ; i < list.size() ; i++) {

			c += delta;

			list[i].@time = ""+c

		}

	}

}

File splitDir = new File("/home/mdecorde/xml/comere/split")

for (File f : splitDir.listFiles()) {

	XPathResult xpathProcessor = new XPathResult(f);

	def files = xpathProcessor.getXpathResponses("//tei:dateline/tei:name/text()")

	files = new HashSet(files)

	if (files.size() > 1) println "$f	$files"

}

// 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.

//

// This file is part of the TXM platform.

//

// The TXM platform is free software: you can redistribute it and/or modif y

// it under the terms of the GNU General Public License as published by

// the Free Software Foundation, either version 3 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: 2011-11-03 17:59:23 +0100 (jeu., 03 nov. 2011) $

// $LastChangedRevision: 2051 $

// $LastChangedBy: mdecorde $ 

//

/**

 * @author mdecorde

 */

import java.io.File;

import java.io.FileNotFoundException;

import java.io.FileOutputStream;

import java.util.HashMap;

import javax.xml.transform.Transformer;

import javax.xml.transform.TransformerConfigurationException;

import javax.xml.transform.TransformerException;

import javax.xml.transform.stream.StreamResult;

import javax.xml.transform.stream.StreamSource;

import net.sf.saxon.TransformerFactoryImpl;

// TODO: Auto-generated Javadoc

/**

 * apply a xsl file to a file : use process(File infile, File outfile, String[]

 * args) to apply the xslt file loaded by the constructor args =

 * ["name1","value1","name2",value2, ...]

 * 

 * @author mdecorde

 */

public class ApplyXsl2 {

	//	/** The proc. */

	//	private Processor proc;

	//	

	//	/** The comp. */

	//	private XsltCompiler comp;

	//	

	//	/** The exp. */

	//	private XsltExecutable exp;

	//

	//	/** The source. */

	//	private XdmNode source;

	//	

	//	/** The out. */

	//	private Serializer out;

	//	

	//	/** The trans. */

	//	private XsltTransformer trans;

	net.sf.saxon.TransformerFactoryImpl tFactory;

	Transformer transformer;

	private HashMap<String, String> params = new HashMap<String, String>();

	private File xsltfile;

	/**

	 * initialize with the xslt file.

	 *

	 * @param xsltfile the xsltfile

	 * @throws TransformerConfigurationException 

	 */

	public ApplyXsl2(String xsltfile) throws TransformerConfigurationException {

		System.setProperty("javax.xml.transform.TransformerFactory", "net.sf.saxon.TransformerFactoryImpl");

		this.xsltfile = new File(xsltfile);

		tFactory = new TransformerFactoryImpl();

		transformer = tFactory.newTransformer(new StreamSource(this.xsltfile));

		//println("new tFactory: "+tFactory);

		//println("new transformer: "+transformer);

	}

	/**

	 * initialize with the xslt file.

	 *

	 * @param xsltfile the xsltfile

	 */

	public ApplyXsl2(File xsltfile) {

		this.xsltfile = xsltfile;

		System.setProperty("javax.xml.transform.TransformerFactory", "net.sf.saxon.TransformerFactoryImpl");

		tFactory = new net.sf.saxon.TransformerFactoryImpl();

		println("new tFactory: "+tFactory);

		println("ApplyXsl2 from file: "+xsltfile);

		reload();

		println("new transformer: "+transformer);

	}

	/**

	 * Set in and out file.

	 *

	 * @param xmlinfile the xmlinfile

	 * @param xmloutfile the xmloutfile

	 * @return true, if successful

	 */

	private boolean SetInOutSource(File xmlinfile, File xmloutfile) {

		//		if (proc == null || exp == null || comp == null)

		//			return false;

		//

		//		source = proc.newDocumentBuilder().build(

		//				new StreamSource(new File(xmlinfile)));

		//		out = new Serializer();

		//		out.setOutputFile(new File(xmloutfile));

		//

		//		trans = exp.load();

		//		transformer.set.setInitialContextNode(source);

		//		transformer.setDestination(out);

		return true;

	}

	/**

	 * Set a xslt param.

	 *

	 * @param name the name

	 * @param value the value

	 * @return true, if successful

	 */

	public boolean SetParam(String name, Object value) {

		if (transformer != null) {

			transformer.setParameter(name, value);

			params.put(name, value.toString());

			return true;

		} else

			return false;

	}

	public boolean resetParams() {

		if (transformer != null) {

			transformer.reset();

			params.clear();

		} else

			return false;

		return true;

	}

	/**

	 * Force java to clean memory after processing a xslt. If not we might catch

	 * a JavaHeapspace Exception

	 * 

	 */

	private void cleanMemory() {

		reload();

		for (String name : params.keySet()) { // reload parameters

			transformer.setParameter(name, params.get(name));

		}

	}

	private boolean reload() {

		try {

			transformer = null;

			transformer = tFactory.newTransformer(new StreamSource(xsltfile));

		} catch (TransformerConfigurationException e) {

			e.printStackTrace();

			println("Error while reloading transformer: "+e);

			return false;

		}

		return true;

	}

	/**

	 * Process files with xslt args.

	 *

	 * @param xmlinfile file to be processed

	 * @param xmloutfile output file

	 * @param args xslt args ["arg1","arg1value","arg2","arg2value"...]

	 * @return true, if successful

	 * @throws TransformerException 

	 * @throws FileNotFoundException 

	 */

	public boolean process(String xmlinfile, String xmloutfile, String[] args) throws Exception {

		File infile = new File(xmlinfile);

		File outfile = new File(xmloutfile);

		if (!this.SetInOutSource(infile, outfile))

			return false;

		//System.out.println(xmlinfile+" Params: "+Arrays.toString(args));

		for (int i = 0; i < args.length; i = i + 2) {

			if (!this.SetParam(args[i], args[i + 1]))

				return false;

		}

		transformer.transform(new StreamSource(xmlinfile), new StreamResult(new FileOutputStream(xmloutfile)));

		cleanMemory(); // save memory

		// System.out.println("Done");

		return true;

	}

	/**

	 * Process files without xslt args.

	 *

	 * @param xmlinfile the xmlinfile

	 * @param xmloutfile the xmloutfile

	 * @return true, if successful

	 * @throws TransformerException 

	 * @throws FileNotFoundException 

	 */

	public boolean process(File xmlinfile, File xmloutfile) throws FileNotFoundException, TransformerException {

		if (!this.SetInOutSource(xmlinfile, xmloutfile))

			return false;

		if (transformer == null) return false;

		transformer.transform(new StreamSource(xmlinfile), new StreamResult(new FileOutputStream(xmloutfile)));

		cleanMemory();

		return true;

	}

	public static boolean processImportSources(File xslFile, File srcdir, File outdir) throws FileNotFoundException, TransformerException

	{

		HashMap<String, String> params = new HashMap<String, String>();

		return processImportSources(xslFile, srcdir, outdir, params);

	}

	public static boolean processImportSources(File xslFile, File srcdir, File outdir, HashMap<String, String> params) throws FileNotFoundException, TransformerException

	{

		outdir.mkdir();

		if (xslFile.exists()) {

			ApplyXsl2 builder = new ApplyXsl2(xslFile);

			for(String name : params.keySet())

				builder.SetParam(name, params.get(name));

			System.out.println("-- Apply xsl "+xslFile+" with parameters: "+params);

			File[] files = srcdir.listFiles();

			if (files != null)

			for (File f : files) {

				if (f.isHidden() || f.isDirectory())

					continue;

				if (f.getName().endsWith(".properties"))

					continue;

				if (f.getName().equals("import.xml"))

					continue;

				if (f.getName().endsWith(".csv"))

					continue;

				if (f.getName().endsWith(".dtd"))

					continue;

				if (f.getName().endsWith(".xsl"))

					continue;

				System.out.print(".");

				File outfile = new File(outdir, f.getName());

				if (!builder.process(f, outfile)) {

					System.out.println("Failed with file "+ f);

					outfile.delete();

				}

			}

			System.out.println("");

			return true;

		} else {

			System.out.println("XSL file does not exists: "+ xslFile);

		}

		return false;

	}

}

import java.io.File;

import java.net.URL;

import javax.xml.stream.*;

public class SplitTEICMRPerLocutor extends StaxIdentityParser {

	File outputDirectory;

	def writers = [:];

	public SplitTEICMRPerLocutor(File inputFile, File outputDirectory) {

		super(inputFile);

		this.outputDirectory = outputDirectory;

		outputDirectory.mkdir()

	}

	protected void processStartElement() {

		if (parser.getLocalName() == "posting") {

			for (int i = 0 ; i < parser.getAttributeCount() ; i++) {

				if (parser.getAttributeLocalName(i) == "who") {

					String loc = parser.getAttributeValue(i)

					writer = writers.get(loc) // switch writer on locutor

					break;

				}

			}

		}

		if (writer != null) {

			processStartElement(writer);

		} else {

			for (def swriter : writers.values()) processStartElement(swriter);

		}

	}

	protected void processStartElement(def swriter)

	{

		String prefix = parser.getPrefix();

		if (prefix != null && prefix.length() > 0)

			swriter.writeStartElement(Nscontext.getNamespaceURI(prefix), localname)

		else

			swriter.writeStartElement(localname);

		for (int i = 0 ; i < parser.getNamespaceCount() ; i++) {

			swriter.writeNamespace(parser.getNamespacePrefix(i), parser.getNamespaceURI(i));

		}

		writeAttributes(swriter);

	}

	private void _processStartElement() {

		String prefix = parser.getPrefix();

		if (prefix != null && prefix.length() > 0)

			writer.writeStartElement(Nscontext.getNamespaceURI(prefix), localname)

		else

			writer.writeStartElement(localname);

		for (int i = 0 ; i < parser.getNamespaceCount() ; i++) {

			writer.writeNamespace(parser.getNamespacePrefix(i), parser.getNamespaceURI(i));

		}

		writeAttributes();

	}

	protected void processNamespace() {

		if (writer != null) {

			writer.writeNamespace(parser.getPrefix(), parser.getNamespaceURI());

		} else {

			for (def swriter : writers.values()) swriter.writeNamespace(parser.getPrefix(), parser.getNamespaceURI());

		}

	}

	protected void writeAttributes() {

		if (writer != null) {

			writeAttributes(writer);

		} else {

			for (def swriter : writers.values()) writeAttributes(swriter);

		}

	}

	protected void writeAttributes(def swriter) {

		for (int i = 0 ; i < parser.getAttributeCount() ; i++) {

			String attrPrefix = parser.getAttributePrefix(i);

			if (attrPrefix != null && attrPrefix.length() > 0)

				swriter.writeAttribute(attrPrefix+":"+parser.getAttributeLocalName(i), parser.getAttributeValue(i));

			else

				swriter.writeAttribute(parser.getAttributeLocalName(i), parser.getAttributeValue(i));

		}

	}

	protected void processCharacters()

	{

		if (writer != null) {

			writer.writeCharacters(parser.getText());

		} else {

			for (def swriter : writers.values()) swriter.writeCharacters(parser.getText());

		}

	}

	protected void processProcessingInstruction()

	{

		if (writer != null) {

			writer.writeProcessingInstruction(parser.getPITarget(), parser.getPIData());

		} else {

			for (def swriter : writers.values()) swriter.writeProcessingInstruction(parser.getPITarget(), parser.getPIData());

		}

	}

	protected void processDTD()

	{

		if (writer != null) {

			writer.writeDTD(parser.getText());

		} else {

			for (def swriter : writers.values()) swriter.writeDTD(parser.getText());

		}

	}

	protected void processCDATA()

	{

		if (writer != null) {

			writer.writeCData(parser.getText())

		} else {

			for (def swriter : writers.values()) swriter.writeCData(parser.getText())

		}

	}

	protected void processComment()

	{

		if (writer != null) {

			writer.writeComment(parser.getText());

		} else {

			for (def swriter : writers.values()) swriter.writeComment(parser.getText());

		}

	}

	protected void processEndElement()

	{

		if (writer != null) {

			writer.writeEndElement();

		} else {

			for (def swriter : writers.values()) swriter.writeEndElement();

		}

		if (parser.getLocalName() == "posting") writer = null;

	}

	protected void processEndDocument() {

		if (writer != null) {

			writer.writeEndDocument();

		} else {

			for (def swriter : writers.values()) swriter.writeEndDocument();

		}

	}

	protected void processEntityReference() {

		if (writer != null) {

			writer.writeEntityRef(parser.getLocalName());

		} else {

			for (def swriter : writers.values()) swriter.writeEntityRef(parser.getLocalName());

		}

	}

	public boolean processLocutors(def locutors) {

		for (String loc : locutors) {

			println "Create writer for $loc"

			writers.put(loc, new StaxStackWriter(new File(outputDirectory, loc+"-teicmr.xml").newWriter("UTF-8")))

		}

		println "writers: $writers"

		for (def swriter : writers.values()) {

			swriter.writeStartDocument("UTF-8", "1.0");

			swriter.writeCharacters("\n");

		}

		boolean ret = process(writer);

		for (def swriter : writers.values()) {

			try {swriter.close();} catch(Exception e){println "close writer exep: "+e}

		}

		if (parser != null)

			try {parser.close()} catch(Exception e){println "parser exep: "+e}

		return ret;

	}

	public static void main(String[] args) {

		File inputFile = new File("/home/mdecorde/xml/comere/ismael-textchat.xml")

		File outputDirectory = new File("/home/mdecorde/xml/comere/split_out")

		def p = new SplitTEICMRPerLocutor(inputFile, outputDirectory)

		println p.processLocutors(["s_1", "s_10", "s_11", "s_12", "s_13", "s_14", "s_3", "s_4", "s_5", "s_6", "s_7", "s_8", "s_9", "tt_1", "tt_12", "tt_2", "tt_3", "tt_4", "tt_5", "tt_6", "tt_7", "tt_8"])

	}

}

import java.io.File;

import java.net.URL;

import javax.xml.stream.*;

public class TEICMCPerLocPerDate extends StaxIdentityParser {

	File outputDirectory;

	def writers = [:];

	def counts = [:];

	public TEICMCPerLocPerDate(File inputFile, File outputDirectory) {

		super(inputFile);

		this.outputDirectory = outputDirectory;

		outputDirectory.mkdir()

	}

	boolean inName = false;

	String filenameValue = "";

	boolean inDate = false;

	String dateValue = "";

	boolean inTime = false;

	String timeValue = "";

	boolean inPosting = false;

	String postingID = ""

	String postingWho = ""

	boolean inP;

	String content = ""

	protected void processStartElement() {

		if (parser.getLocalName() == "posting") {

			inPosting = true

			for (int i = 0 ; i < parser.getAttributeCount() ; i++) {

				if (parser.getAttributeLocalName(i) == "who") {

					postingWho = parser.getAttributeValue(i)

				} else if (parser.getAttributeLocalName(i) == "id") {

					postingID = parser.getAttributeValue(i)

				}

			}

			return;

		} else if(inPosting) {

			if ( parser.getLocalName() == "dateline") {

			} else if (parser.getLocalName() == "name") {

				inName = true

				filenameValue = ""

			} else if (parser.getLocalName() == "date") {

				inDate = true;

				dateValue = ""

			} else if (parser.getLocalName() == "time") {

				inTime = true;

				timeValue = ""

			} else if (parser.getLocalName() == "p") {

				inP = true;

				content = ""

			}

			return;

		}

		if (writer != null) {

			processStartElement(writer);

		} else {

			for (def swriter : writers.values()) processStartElement(swriter);

		}

	}

	protected void processCharacters() {

		if (inPosting) {

			if (inName) {

				filenameValue += parser.getText();

				return;

			} else if (inDate) {

				dateValue += parser.getText();

				return;

			} else if (inTime) {

				timeValue += parser.getText();

				return;

			} else if (inP) {

				content += parser.getText();

				return;

			}

		}

		if (writer != null) {

			writer.writeCharacters(parser.getText());

		} else {

			for (def swriter : writers.values()) swriter.writeCharacters(parser.getText());

		}

	}

	protected void processEndElement()

	{

		if (inPosting) {

			if (inPosting && parser.getLocalName() == "name") {

				inName = false;

			} else if (parser.getLocalName() == "date") {

				inDate = false;

			} else if (parser.getLocalName() == "p") {

				inP = false;

			} else if (parser.getLocalName() == "time") {

				inTime = false;

				writer = getWriter(dateValue+postingWho)

			} else if (parser.getLocalName() == "posting") {

				if (writer == null) {

					println "Error Houston !!! with loc=$postingWho date=$dateValue at="parser.getLocation()

					return;

				}

				// write Turn

				//println "write Turn"

				writer.writeStartElement("posting");

				writer.writeAttribute("xml:id",postingID);

				writer.writeAttribute("who",postingWho);

				writer.writeStartElement("dateline");

				writer.writeStartElement("name");

				writer.writeAttribute("type","file");

				writer.writeCharacters(filenameValue)

				writer.writeEndElement(); //name

				writer.writeStartElement("date");

				writer.writeCharacters(dateValue)

				writer.writeEndElement(); //date

				writer.writeStartElement("time");

				writer.writeCharacters(timeValue)

				writer.writeEndElement(); //time

				writer.writeEndElement(); //dateline

				writer.writeStartElement("p");

				writer.writeCharacters(content)

				writer.writeEndElement(); //p

				writer.writeEndElement(); //posting // closed in the next lines

				//println "end write Turn"

				inPosting = false

				writer = null;

			}

			return;

		}

		if (writer != null) {

			writer.writeEndElement();

		} else {

			for (def swriter : writers.values()) swriter.writeEndElement();

		}

	}

	protected void processStartElement(def swriter)

	{

		String prefix = parser.getPrefix();

		if (prefix != null && prefix.length() > 0)

			swriter.writeStartElement(Nscontext.getNamespaceURI(prefix), localname)

		else

			swriter.writeStartElement(localname);

		for (int i = 0 ; i < parser.getNamespaceCount() ; i++) {

			swriter.writeNamespace(parser.getNamespacePrefix(i), parser.getNamespaceURI(i));

		}

		writeAttributes(swriter);

	}

	private void _processStartElement() {

		String prefix = parser.getPrefix();

		if (prefix != null && prefix.length() > 0)

			writer.writeStartElement(Nscontext.getNamespaceURI(prefix), localname)

		else

			writer.writeStartElement(localname);

		for (int i = 0 ; i < parser.getNamespaceCount() ; i++) {

			writer.writeNamespace(parser.getNamespacePrefix(i), parser.getNamespaceURI(i));

		}

		writeAttributes();

	}

	protected void processNamespace() {

		if (writer != null) {

			writer.writeNamespace(parser.getPrefix(), parser.getNamespaceURI());

		} else {

			for (def swriter : writers.values()) swriter.writeNamespace(parser.getPrefix(), parser.getNamespaceURI());

		}

	}

	protected void writeAttributes() {

		if (writer != null) {

			writeAttributes(writer);

		} else {

			for (def swriter : writers.values()) writeAttributes(swriter);

		}

	}

	protected void writeAttributes(def swriter) {

		for (int i = 0 ; i < parser.getAttributeCount() ; i++) {

			String attrPrefix = parser.getAttributePrefix(i);

			if (attrPrefix != null && attrPrefix.length() > 0)

				swriter.writeAttribute(attrPrefix+":"+parser.getAttributeLocalName(i), parser.getAttributeValue(i));

			else

				swriter.writeAttribute(parser.getAttributeLocalName(i), parser.getAttributeValue(i));

		}

	}

	protected void processProcessingInstruction() {

		if (writer != null) {

			writer.writeProcessingInstruction(parser.getPITarget(), parser.getPIData());

		} else {

			for (def swriter : writers.values()) swriter.writeProcessingInstruction(parser.getPITarget(), parser.getPIData());

		}

	}

	protected void processDTD()

	{

		if (writer != null) {

			writer.writeDTD(parser.getText());

		} else {

			for (def swriter : writers.values()) swriter.writeDTD(parser.getText());

		}

	}

	protected void processCDATA()

	{

		if (writer != null) {

			writer.writeCData(parser.getText())

		} else {

			for (def swriter : writers.values()) swriter.writeCData(parser.getText())

		}

	}

	protected void processComment()

	{

		if (writer != null) {

			writer.writeComment(parser.getText());

		} else {

			for (def swriter : writers.values()) swriter.writeComment(parser.getText());

		}

	}

	protected getWriter(String key) {

		def w = writers.get(key)

		counts.put(key, counts.get(key)+1)

		return w

	}

	protected void processEndDocument() {

		if (writer != null) {

			writer.writeEndDocument();

		} else {

			for (def swriter : writers.values()) swriter.writeEndDocument();

		}

	}

	protected void processEntityReference() {

		if (writer != null) {

			writer.writeEntityRef(parser.getLocalName());

		} else {

			for (def swriter : writers.values()) swriter.writeEntityRef(parser.getLocalName());

		}

	}

	public boolean processDatesAndLocs(def dates, def locutors) {

		for (String date : dates) {

			for (String loc : locutors) {

				println "Create writer for $loc"

				writers.put(date+loc, new StaxStackWriter(new File(outputDirectory, "${loc}-${date}-teicmr.xml")))

				counts.put(date+loc, 0)

			}

		}

		println "writers: $writers"

		for (def swriter : writers.values()) {

			swriter.writeStartDocument("UTF-8", "1.0");

			swriter.writeCharacters("\n");

		}

		boolean ret = process(writer);

		for (def swriter : writers.values()) {

			try {swriter.close();} catch(Exception e){println "close writer exep: "+e}

		}

		if (parser != null)

			try {parser.close()} catch(Exception e){println "parser exep: "+e}

		//delete file (date+loc) not used

		for (def key : counts.keySet()) {

			if (counts.get(key) == 0) { // never used for a couple date+loc

				(writers.get(key)).getInfile().delete()

			}

		}

		return ret;

	}

	public static void main(String[] args) {

		File inputFile = new File("/home/mdecorde/xml/comere/ismael-textchat.xml")

		File outputDirectory = new File("/home/mdecorde/xml/comere/split_out")

		outputDirectory.deleteDir()

		outputDirectory.mkdir()

		XPathResult xpathProcessor = new XPathResult(inputFile);

		def dates = xpathProcessor.getXpathResponses("//tei:dateline/tei:date/text()")

		def locs = xpathProcessor.getXpathResponses("//tei:listPerson/tei:person/@xml:id")

		dates = new HashSet(dates)

		dates = new ArrayList(dates)

		println dates.sort()

		println locs

		def p = new TEICMCPerLocPerDate(inputFile, outputDirectory)

		println p.processDatesAndLocs(dates, locs)

	}

}

import java.io.File;

import java.net.URL;

import javax.xml.stream.*;

import java.text.SimpleDateFormat;

import java.util.Date;

public class TEICMC2TRS extends StaxIdentityParser {

	File outputDirectory;

	String loc;

	public TEICMC2TRS(File inputFile, String loc) {

		super(inputFile);

		this.loc = loc;

	}

	boolean inBody = false

	boolean inName = false;

	String filenameValue = "";

	boolean inDate = false;

	String dateValue = "";

	boolean inTime = false;

	String timeValue = "";

	boolean inPosting = false;

	String postingID = ""

	String postingWho = ""

	boolean inP;

	String content = ""

	protected void processStartElement() {

		if (parser.getLocalName() == "body") { // write TRS HEADER

			inBody = true

			writer.writeStartElement("Trans");

			writer.writeAttribute("version","10");

			writer.writeAttribute("audio_filename","");

			writer.writeAttribute("scribe","");

			writer.writeAttribute("version_date","");

			writer.writeStartElement("Topics");

			writer.writeEndElement(); //Topics

			writer.writeStartElement("Speakers");

			writer.writeStartElement("Speaker");

			writer.writeAttribute("id",loc);

			writer.writeAttribute("name",loc);

			writer.writeAttribute("check","");

			writer.writeAttribute("dialect","");

			writer.writeAttribute("scope","");

			writer.writeEndElement(); //Speaker

			writer.writeEndElement(); //Speakers

			writer.writeStartElement("Episode");

			writer.writeStartElement("Section");

			writer.writeAttribute("type","report");

			writer.writeAttribute("startTime","");

			writer.writeAttribute("endTime","");

		} else if (parser.getLocalName() == "posting") {

			inPosting = true

			for (int i = 0 ; i < parser.getAttributeCount() ; i++) {

				if (parser.getAttributeLocalName(i) == "who") {

					postingWho = parser.getAttributeValue(i)

				} else if (parser.getAttributeLocalName(i) == "type") {

					postingID = parser.getAttributeValue(i)

				}

			}

			return;

		} else if(inPosting) {

			if ( parser.getLocalName() == "dateline") {

			} else if (parser.getLocalName() == "name") {

				inName = true

				filenameValue = ""

			} else if (parser.getLocalName() == "date") {

				inDate = true;

				dateValue = ""

			} else if (parser.getLocalName() == "time") {

				inTime = true;

				timeValue = ""

			} else if (parser.getLocalName() == "p") {

				inP = true;

				content = ""

			}

			return;

		}

	}

	protected void processCharacters() {

		if (inPosting) {

			if (inName) {

				filenameValue += parser.getText();

				return;

			} else if (inDate) {

				dateValue += parser.getText();

				return;

			} else if (inTime) {

				timeValue += parser.getText();

				return;

			} else if (inP) {

				content += parser.getText();

				return;

			}

		}

	}

	SimpleDateFormat formatter = new SimpleDateFormat("HHmmss");

	Date from = null// use only the first from

	protected void processEndElement()

	{

		if (parser.getLocalName() == "body") { // write TRS HEADER

			writer.writeEndElement(); //Section

			writer.writeEndElement(); //Episode

			writer.writeEndElement(); //Trans

		} else if (inPosting) {

			if (inPosting && parser.getLocalName() == "name") {

				inName = false;

			} else if (parser.getLocalName() == "date") {

				inDate = false;

			} else if (parser.getLocalName() == "p") {

				inP = false;

			} else if (parser.getLocalName() == "time") {

				inTime = false;

			} else if (parser.getLocalName() == "posting") {

				if (writer == null) {

					println "Error Houston !!! with loc=$postingWho date=$dateValue at="parser.getLocation()

					return;

				}

				// write Turn

				//println "write Turn"

				if (from == null) {

					int i1 = filenameValue.indexOf(".flv")

					int i2 = i1-6

					String date1 = filenameValue.substring(i2, i1)

					from = formatter.parse(date1);

				}

				Date to = formatter.parse(timeValue.replace(":", ""));

				float diff = ((float)((to.getTime() - from.getTime()))) / 1000.0f // ms

				float diff2 = diff + 10

				writer.writeStartElement("Turn");

				writer.writeAttribute("startTime","$diff");

				writer.writeAttribute("endTime","$diff2");

				writer.writeAttribute("speaker",postingWho);

				writer.writeStartElement("Sync");

				writer.writeAttribute("time","$diff");

				writer.writeEndElement(); //Sync

				writer.writeCharacters(content)

				writer.writeEndElement(); //Turn

				writer.writeCharacters("\n")

				//println "end write Turn"

				inPosting = false

			}

			return;

		}

	}

	public static void main(String[] args) {

		File inputFile = new File("/home/mdecorde/xml/comere/split_out/tt_5-2013-11-12-teicmr.xml")

		File outputFile = new File("/home/mdecorde/xml/comere/split_out/tt_5-2013-11-12-teicmr-broken.trs")

		File outputFixedFile = new File("/home/mdecorde/xml/comere/split_out/tt_5-2013-11-12-teicmr.trs")

		XPathResult xpathProcessor = new XPathResult(inputFile);

		def loc = xpathProcessor.getXpathResponse("//tei:posting/@who")

		def p1 = new TEICMC2TRS(inputFile, loc)		

		def p2 = new FixMissingTimings(outputFile, outputFixedFile)

		if (p1 && p2) {

			outputFile.delete()

		} else {

			println "Failed to convert $inputFile to TRS"

		}

	}

}

import javax.xml.namespace.NamespaceContext;

import javax.xml.stream.*;

public class StaxStackWriter implements XMLStreamWriter {

	public XMLStreamWriter writer;

	public Writer output;

	public File infile;

	public def events = [];

	public StaxStackWriter(def output) {

		XMLOutputFactory factory = XMLOutputFactory.newInstance();

		this.output = output

		this.writer = factory.createXMLStreamWriter(output);//create a new file

	}

	public StaxStackWriter(File infile) {

		this(infile.newWriter("UTF-8"))

		this.infile = infile

	}

	public File getInfile() {

		return infile;

	}

	public void write(String txt) {

		output.write(txt)

	}

	public void writeEndElements() {

		int size = events.size();

		for (int i = 0 ; i < size ; i++) {

			writeEndElement();

		}

		events = [];

	}

	public void writeStartElements(def tagsToWrite) {

		for (String tag : tagsToWrite)

			writeStartElement(tag);

	}

	public def getTagStack() {

		return events;

	}

	@Override

	public void close () throws XMLStreamException {

		writer.close();

		output.close();

	}

	@Override

	public void flush () throws XMLStreamException {

		writer.flush();

	}

	@Override

	public NamespaceContext getNamespaceContext() {

		return writer.getNamespaceContext();

	}

	@Override

	public String getPrefix (String uri) throws XMLStreamException {

		return writer.getPrefix(uri);

	}

	@Override

	public Object getProperty (String name) throws IllegalArgumentException {

		return writer.getProperty(name);

	}

	@Override

	public void setDefaultNamespace (String uri) throws XMLStreamException {

		writer.setDefaultNamespace(uri);

	}

	@Override

	public void setNamespaceContext(NamespaceContext context)

	throws XMLStreamException {

		writer.setNamespaceContext(context);

	}

	@Override

	public void setPrefix (String prefix, String uri) throws XMLStreamException {

		writer.setPrefix(prefix, uri);

	}

	@Override

	public void writeAttribute (String localName, String value) throws XMLStreamException {

		writer.writeAttribute(localName, value);

	}

	@Override

	public void writeAttribute (String namespaceURI, String localName, String value) throws XMLStreamException {

		writer.writeAttribute(namespaceURI, localName, value);

	}

	@Override

	public void writeAttribute (String prefix, String namespaceURI, String localName, String value) throws XMLStreamException {

		writer.writeAttribute(prefix, namespaceURI, localName, value);

	}

	@Override

	public void writeCData (String data) throws XMLStreamException {

		writer.writeCData(data);

	}

	@Override

	public void writeCharacters (String text) throws XMLStreamException {

		writer.writeCharacters(text);

	}

	@Override

	public void writeCharacters (char[] text, int start, int len) throws XMLStreamException {

		writer.writeCharacters(text, start, len);

	}

	@Override

	public void writeComment (String data) throws XMLStreamException {

		writer.writeComment(data);

	}

	@Override

	public void writeDTD (String dtd) throws XMLStreamException {

		writer.writeDTD(dtd);

	}

	@Override

	public void writeDefaultNamespace(String namespaceURI)

	throws XMLStreamException {

		writer.writeDefaultNamespace(namespaceURI);

	}

	public void writeElement (String localName, String text) throws XMLStreamException {

		writer.writeStartElement(localName);

		writer.writeCharacters(text);

		writer.writeEndElement();

	}

	public void writeElement (String localName, def map, String text) throws XMLStreamException {

		writer.writeStartElement(localName);

		for (def key : map.keySet()) writeAttribute(key.toString(), map[key].toString());

		writer.writeCharacters(text);

		writer.writeEndElement();

	}

	@Override

	public void writeEmptyElement (String localName) throws XMLStreamException {

		writer.writeEmptyElement(localName);

	}

	public void writeEmptyElement (String localName, def map) throws XMLStreamException {

		writer.writeEmptyElement(localName);

		for (def key : map.keySet())

			writeAttribute(key.toString(), map[key].toString());

	}

	@Override

	public void writeEmptyElement(String namespaceURI, String localName)

	throws XMLStreamException {

		writer.writeEmptyElement(namespaceURI, localName);

	}

	@Override

	public void writeEmptyElement (String prefix, String localName, String namespaceURI) throws XMLStreamException {

		writer.writeEmptyElement(prefix, localName, namespaceURI);

	}

	@Override

	public void writeEndDocument () throws XMLStreamException {

		writer.writeEndDocument();

	}

	@Override

	public void writeEndElement () throws XMLStreamException {

		writer.writeEndElement();

		events.pop()

	}

	@Override

	public void writeEntityRef (String name) throws XMLStreamException {

		writer.writeEntityRef(name);

	}

	@Override

	public void writeNamespace(String prefix, String namespaceURI)

	throws XMLStreamException {

		writer.writeNamespace(prefix, namespaceURI);

	}

	@Override

	public void writeProcessingInstruction(String target)

	throws XMLStreamException {

		writer.writeProcessingInstruction(target);

	}

	@Override

	public void writeProcessingInstruction(String target, String data)

	throws XMLStreamException {

		writer.writeProcessingInstruction(target, data);

	}

	@Override

	public void writeStartDocument () throws XMLStreamException {

		writer.writeStartDocument();

	}