Laboratoire ICAR » Plateforme TXM

package org.txm.macro.table

import org.kohsuke.args4j.*

import groovy.transform.Field

import java.nio.charset.Charset

import org.txm.rcpapplication.swt.widget.parameters.*

import org.txm.utils.*

import javax.xml.stream.*

import java.net.URL

import java.text.SimpleDateFormat

import org.apache.poi.ss.usermodel.*

import org.apache.poi.hssf.usermodel.*

import org.apache.poi.xssf.usermodel.*

import org.apache.poi.ss.usermodel.WorkbookFactory

import org.apache.poi.ss.util.*

def stringToIndent = { str -> org.txm.utils.AsciiUtils.buildAttributeId(org.txm.utils.AsciiUtils.convertNonAscii(str)).toLowerCase() }

// from http://www.java-connect.com/apache-poi-tutorials/read-all-type-of-excel-cell-value-as-string-using-poi

def getCellValueAsString = { cell ->

	strCellValue = null

	if (cell != null) {

		switch (cell.getCellType()) {

			case CellType.STRING:

				strCellValue = cell.toString()

				break

			case CellType.NUMERIC:

				if (DateUtil.isCellDateFormatted(cell)) {

					SimpleDateFormat dateFormat = new SimpleDateFormat("dd/MM/yyyy")

					strCellValue = dateFormat.format(cell.getDateCellValue())

				} else {

					value = cell.getNumericCellValue()

					longValue = value.longValue()

					strCellValue = new String(longValue.toString())

				}

				break

			case CellType.BOOLEAN:

				strCellValue = new String(new Boolean(cell.getBooleanCellValue()).toString())

				break

			case CellType.BLANK:

				strCellValue = ""

				break

		}

	}

	if (strCellValue == null) strCellValue = ""

	return strCellValue

}

@Field @Option(name="inputFile", usage="fichier Excel à traiter", widget="File", required=true, def="")

		File inputFile

@Field @Option(name="sheetName", usage="sheet name (if no name is given the first sheet will be used)", widget="String", required=false, def="")

		def sheetName

@Field @Option(name="rootTag", usage="root tag name", widget="String", required=false, def="root")

		def rootTag

@Field @Option(name="textTag", usage="line unit tag name", widget="String", required=false, def="unit")

		def textTag

@Field @Option(name="metadataColumnList", usage="metadata columns list separated by comma", widget="String", required=false, def="meta1,meta2")

		def metadataColumnList

@Field @Option(name="dateColumnList", usage="date columns list separated by comma", widget="String", required=false, def="meta1,meta2")

		def dateColumnList

@Field @Option(name="textColumnList", usage="text columns list separated by comma", widget="String", required=false, def="textColumnList1,textColumnList2")

		def textColumnList

@Field @Option(name="EmbedInTEI", usage="text columns list separated by comma", widget="Boolean", required=false, def="false")

		def EmbedInTEI

if (!ParametersDialog.open(this)) return

	if (!inputFile.exists()) {

		println "** Excel2XML: no '"+inputFile.name+"' file found. Aborting."

		return false

	}

if (!inputFile.canRead()) {

	println "** Excel2XML: '"+inputFile.name+"' file not readable. Aborting."

	return false

}

try {

	metadataColumnList = metadataColumnList.split(",").collect { it.trim() }

	dateColumnList = dateColumnList.split(",").collect { it.trim() }

	textColumnList = textColumnList.split(",").collect { it.trim() }

	textTag = textTag.trim()

	rootTag = rootTag.trim()

	wb = WorkbookFactory.create(inputFile)

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

		ws = wb.getSheetAt(0)

	} else {

		ws = wb.getSheet(sheetName)

		if (ws == null) {

			println "** Excel2XML: no '"+sheetName+" found. Aborting."

			return false

		}

	}

	if (ws == null) {

		println "** Excel2XML: no sheet found. Aborting."

		return false

	}

	nRows = ws.getPhysicalNumberOfRows()

	println nRows+" rows."

	firstRow = ws.getRow(0)

	colMax = firstRow.getLastCellNum()

	headers = (0..colMax-1).collect { getCellValueAsString(firstRow.getCell(it)) }

	println "Headers: $headers"

	normalizedHeaders = headers.collect { stringToIndent(it) }

	ok = true

	metadataColumnList.each { m ->

		if (!headers.contains(m)) {

			println "** Excel2XML: missing metadataColumnList column: $m"

			ok = false

		}

	}

	textColumnList.each { t ->

		if (!headers.contains(t)) {

			println "** Excel2XML: missing textColumnList column: $t"

			ok = false

		}

	}

	if (!ok) { return false }

	metadataColumnIndex = metadataColumnList.collect { headers.indexOf(it) }

	dateColumnsIndex = dateColumnList.collect { headers.indexOf(it) }

	textColumnIndex = textColumnList.collect { headers.indexOf(it) }

	println "metadataColumnList = "+metadataColumnList

	println "metadataColumnIndex = "+metadataColumnIndex

	println "dateColumnList = "+dateColumnList

	println "dateColumnsIndex = "+dateColumnsIndex

	println "textColumnList = "+textColumnList

	println "textColumnIndex = "+textColumnIndex

	name = inputFile.getName()

	idx = name.lastIndexOf(".")

	if (idx > 0) name = name.substring(0, idx)

	outputFile = new File(inputFile.getParentFile(), name+".xml")

	factory = XMLOutputFactory.newInstance()

	output = new FileOutputStream(outputFile)

	writer = factory.createXMLStreamWriter(output, "UTF-8")

	writer.writeStartDocument("UTF-8","1.0")

	writer.writeCharacters("\n") // simple XML formating

	if (EmbedInTEI) {

		writer.writeStartElement("TEI")

		writer.writeStartElement("teiHeader")

		writer.writeEndElement() // teiHeader

		writer.writeStartElement("text")

		writer.writeCharacters("\n")

	}

	writer.writeStartElement(rootTag)

	writer.writeCharacters("\n")

	pb_n = 1

	(1..nRows-1).each { rowIndex ->

		writer.writeCharacters("  ")

		writer.writeEmptyElement("pb") // <pb/> to get one page per input line (don't forget high number of words per page in import module)

		writer.writeAttribute("n", ""+pb_n++)

		writer.writeCharacters("\n") // simple XML formating

		writer.writeCharacters("  ")

		writer.writeStartElement(textTag)

		metadataColumnIndex.each { colIndex -> // build an attribute for each metadata

			def row = ws.getRow(rowIndex)

			if (row != null) {

				String s = getCellValueAsString(row.getCell(colIndex));

				if (s == null) s ="";

				value = s.replaceAll("\n", ";").trim()

				writer.writeAttribute(normalizedHeaders[colIndex], value)

				if (colIndex in dateColumnsIndex) { // also split date attributes in day+month+year attributes

					matches = (value =~ /([0-9]{2})\/([0-9]{2})\/([0-9]{4})/)

					writer.writeAttribute(normalizedHeaders[colIndex]+"jour", matches[0][1])

					writer.writeAttribute(normalizedHeaders[colIndex]+"joursemaine", new java.text.SimpleDateFormat('EEEE').format(Date.parse("dd/MM/yyyy", value)))

					writer.writeAttribute(normalizedHeaders[colIndex]+"mois", matches[0][2])

					writer.writeAttribute(normalizedHeaders[colIndex]+"annee", matches[0][3])

				}

			}

		}

		writer.writeCharacters("\n")

		writer.writeCharacters("    ")

		writer.writeStartElement("metadata")

		writer.writeStartElement("list")

		writer.writeAttribute("type", "unordered")

		writer.writeCharacters("\n")

		metadataColumnIndex.each { colIndex ->

			def row = ws.getRow(rowIndex)

			if (row != null) {

				writer.writeStartElement("item")

				writer.writeCharacters(headers[colIndex]+" : "+getCellValueAsString(row.getCell(colIndex)).replaceAll("\n", ";"))

				writer.writeEndElement() // item

				writer.writeCharacters("\n")

			}

		}

		writer.writeCharacters("    ")

		writer.writeEndElement() // list

		writer.writeEndElement() // head

		writer.writeCharacters("\n")

		textColumnIndex.each { colIndex ->

			writer.writeCharacters("    ")

			writer.writeStartElement(normalizedHeaders[colIndex])

			writer.writeStartElement("p")

			writer.writeStartElement("head")

			writer.writeStartElement("hi")

			writer.writeCharacters(headers[colIndex]+" : ")

			writer.writeEndElement() // hi

			writer.writeEndElement() // head

			def row = ws.getRow(rowIndex)

			if (row != null) {

				value = getCellValueAsString(row.getCell(colIndex))

				if (value ==~ /(?s)^[A-Z]{3}:  [^;\n]+? +[;\n].*/) {

					value.findAll( /(?s)[A-Z]{3}:  ([^;\n]+?) +[;\n]/ ).each { desc ->

						writer.writeStartElement("descripteur")

						matches = (desc =~ /(?s)([A-Z]{3}):  ([^;\n]+?) +[;\n]/)

						writer.writeAttribute("type", matches[0][1])

						writer.writeCharacters(matches[0][2])

						writer.writeEndElement() // descripteur

					}

				} else {

					writer.writeCharacters(value)

				}

				writer.writeEndElement() // p

				writer.writeEndElement() // textColumn

				writer.writeCharacters("\n")

			}

		}

		writer.writeCharacters("  ")

		writer.writeEndElement() // textTag

		writer.writeCharacters("\n")

	}

	writer.writeEndElement() // rootTag

	writer.writeCharacters("\n")

	if (EmbedInTEI) {

		writer.writeEndElement() // text

		writer.writeCharacters("\n")

		writer.writeEndElement() // TEI

		writer.writeCharacters("\n")

	}

	writer.close()

	output.close()

	println "Result file: $outputFile"

} catch (Exception e) {

	println "** Excel2XML: unable to read input file. Aborting."

	println e.getLocalizedMessage()

	println e.printStackTrace()

	return false

}

return true

package org.txm.macro.table

import org.kohsuke.args4j.*

import groovy.transform.Field

import java.nio.charset.Charset

import org.txm.libs.msoffice.ReadExcel

import org.txm.rcpapplication.swt.widget.parameters.*

import org.txm.utils.*

import javax.xml.stream.*

import java.net.URL

import java.text.SimpleDateFormat

import org.apache.poi.ss.usermodel.*

import org.apache.poi.hssf.usermodel.*

import org.apache.poi.xssf.usermodel.*

import org.apache.poi.ss.usermodel.WorkbookFactory

import org.apache.poi.ss.util.*

def stringToIndent = { str -> org.txm.utils.AsciiUtils.buildAttributeId(org.txm.utils.AsciiUtils.convertNonAscii(str)).toLowerCase() }

@Field @Option(name="inputFile", usage="fichier Excel à traiter", widget="File", required=true, def="")

		File inputFile

@Field @Option(name="sheetName", usage="sheet name (if no name is given the first sheet will be used)", widget="String", required=false, def="")

		def sheetName

@Field @Option(name="EmbedInTEI", usage="embed xml content in  TEI text element", widget="Boolean", required=false, def="false")

		def EmbedInTEI

@Field @Option(name="rootTag", usage="root tag name", widget="String", required=false, def="root")

		def rootTag

@Field @Option(name="textTag", usage="line unit tag name", widget="String", required=false, def="unit")

		def textTag

@Field @Option(name="metadataColumnList", usage="metadata columns list separated by comma", widget="String", required=false, def="meta1,meta2")

		def metadataColumnList

@Field @Option(name="dateColumnList", usage="date columns list separated by comma", widget="String", required=false, def="meta1,meta2")

		def dateColumnList

@Field @Option(name="textColumnList", usage="text columns list separated by comma", widget="String", required=false, def="textColumnList1,textColumnList2")

		def textColumnList

@Field @Option(name="prefixContentColumnList", usage="columns which content is a semi-colon separated prefixes list", widget="String", required=false, def="")

		def prefixContentColumnList

@Field @Option(name="listContentColumnList", usage="columns which content is a semi-colon separated list", widget="String", required=false, def="")

		def listContentColumnList

if (!ParametersDialog.open(this)) return

	if (!inputFile.exists()) {

		println "** Excel2XML: no '"+inputFile.name+"' file found. Aborting."

		return false

	}

if (!inputFile.canRead()) {

	println "** Excel2XML: '"+inputFile.name+"' file not readable. Aborting."

	return false

}

try {

	metadataColumnList = metadataColumnList.split(",").collect { it.trim() }

	dateColumnList = dateColumnList.split(",").collect { it.trim() }

	textColumnList = textColumnList.split(",").collect { it.trim() }

	prefixContentColumnList = prefixContentColumnList.split(",").collect { it.trim() }

	listContentColumnList = listContentColumnList.split(",").collect { it.trim() }

	textTag = textTag.trim()

	rootTag = rootTag.trim()

	ReadExcel excel = new ReadExcel(inputFile, sheetName);

	excel.readHeaders();

	headers = excel.getHeaders()

	println "Headers: $headers"

	ok = true

	metadataColumnList.each { m ->

		if (!headers.contains(m)) {

			println "** Excel2XML: missing metadataColumnList column: $m"

			ok = false

		}

	}

	textColumnList.each { t ->

		if (!headers.contains(t)) {

			println "** Excel2XML: missing textColumnList column: $t"

			ok = false

		}

	}

	if (!ok) { return false }

	println "metadataColumnList = "+metadataColumnList

	println "dateColumnList = "+dateColumnList

	println "textColumnList = "+textColumnList

	name = inputFile.getName()

	idx = name.lastIndexOf(".")

	if (idx > 0) name = name.substring(0, idx)

	outputFile = new File(inputFile.getParentFile(), name+".xml")

	factory = XMLOutputFactory.newInstance()

	output = new FileOutputStream(outputFile)

	writer = factory.createXMLStreamWriter(output, "UTF-8")

	writer.writeStartDocument("UTF-8","1.0")

	writer.writeCharacters("\n") // simple XML formating

	if (EmbedInTEI) {

		writer.writeStartElement("TEI")

		writer.writeStartElement("teiHeader")

		writer.writeEndElement() // teiHeader

		writer.writeStartElement("text")

		writer.writeCharacters("\n")

	}

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

		writer.writeStartElement(rootTag)

		writer.writeCharacters("\n")

	}

	pb_n = 1

	//(1..nRows-1).each { rowIndex ->

	def record = null

	while (excel.readRecord()) {

		record = excel.getRecord()

		if (record.isEmpty()) continue;

		//println "record=$record"

		writer.writeCharacters("  ")

		writer.writeEmptyElement("pb") // <pb/> to get one page per input line (don't forget high number of words per page in import module)

		writer.writeAttribute("n", ""+pb_n++)

		writer.writeCharacters("\n") // simple XML formating

		writer.writeCharacters("  ")

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

			writer.writeStartElement(textTag)

			//metadataColumnIndex.each { colIndex -> // build an attribute for each metadata

			for (def metadataName : metadataColumnList) {

				String s = record.get(metadataName);

				if (s == null) s = "";

				value = s.replaceAll("\n", ";").trim()

				writer.writeAttribute(stringToIndent(metadataName), value)

				if (metadataName in dateColumnList) { // also split date attributes in day+month+year attributes

					matches = (value =~ /([0-9]{2})\/([0-9]{2})\/([0-9]{4})/)

					writer.writeAttribute(metadataName+"-jour", matches[0][1])

					writer.writeAttribute(metadataName+"-joursemaine", new java.text.SimpleDateFormat('EEEE').format(Date.parse("dd/MM/yyyy", value)))

					writer.writeAttribute(metadataName+"-mois", matches[0][2])

					writer.writeAttribute(metadataName+"-annee", matches[0][3])

				}

			}

			writer.writeCharacters("\n")

		}

		writer.writeCharacters("    ")

		writer.writeStartElement("metadata")

		writer.writeStartElement("list")

		writer.writeAttribute("type", "unordered")

		writer.writeCharacters("\n")

		//metadataColumnIndex.each { colIndex ->

		for (def metadataName : metadataColumnList) {

			writer.writeStartElement("item")

			writer.writeCharacters(metadataName+" : "+record.get(metadataName).replaceAll("\n", ";"))

			writer.writeEndElement() // item

			writer.writeCharacters("\n")

		}

		writer.writeCharacters("    ")

		writer.writeEndElement() // list

		writer.writeEndElement() // head

		writer.writeCharacters("\n")

		//textColumnIndex.each { colIndex ->

		for (def textColumnName : textColumnList) {

			writer.writeCharacters("    ")

			writer.writeStartElement(stringToIndent(textColumnName))

			writer.writeStartElement("p")

			writer.writeStartElement("head")

			writer.writeStartElement("hi")

			writer.writeCharacters(textColumnName+" : ")

			writer.writeEndElement() // hi

			writer.writeEndElement() // head

			value = record.get(textColumnName)

			if (value ==~ /(?s)^[A-Z]{3}:  [^;\n]+? +[;\n].*/) {

				value.findAll( /(?s)[A-Z]{3}:  ([^;\n]+?) +[;\n]/ ).each { desc ->

					writer.writeStartElement("descripteur")

					matches = (desc =~ /(?s)([A-Z]{3}):  ([^;\n]+?) +[;\n]/)

					writer.writeAttribute("type", matches[0][1])

					writer.writeCharacters(matches[0][2])

					writer.writeEndElement() // descripteur

				}

			} else {

				writer.writeCharacters(value)

			}

			writer.writeEndElement() // p

			writer.writeEndElement() // textColumn

			writer.writeCharacters("\n")

		}

		writer.writeCharacters("  ")

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

			writer.writeEndElement() // textTag

			writer.writeCharacters("\n")

		}

	}

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

		writer.writeEndElement() // rootTag

		writer.writeCharacters("\n")

	}

	if (EmbedInTEI) {

		writer.writeEndElement() // text

		writer.writeCharacters("\n")

		writer.writeEndElement() // TEI

		writer.writeCharacters("\n")

	}

	writer.close()

	output.close()

	println "Result file: $outputFile"

} catch (Exception e) {

	println "** Excel2XML: unable to read input file. Aborting."

	println e.getLocalizedMessage()

	println e.printStackTrace()

	return false

}

return true

package org.txm.macro.csv

import org.kohsuke.args4j.*

import groovy.transform.Field

import java.nio.charset.Charset

import org.txm.rcp.swt.widget.parameters.*

import org.txm.utils.*

import javax.xml.stream.*

import java.net.URL

import org.txm.libs.msoffice.ReadExcel

@Field @Option(name="inputFile", usage="CSV File", widget="File", required=false, def="file.xlsx")

		File inputFile;

@Field @Option(name="outputDirectory", usage="output directory", widget="Folder", required=false, def="directory")

		File outputDirectory;

@Field @Option(name="textsSeparator", usage="Texts", widget="Separator", required=false, def="Texts")

		def textsSeparator

@Field @Option(name="textIDColumn", usage="<text> id column", widget="String", required=false, def="Identifiant de la notice")

		def textIDColumn;

@Field @Option(name="textSelector", usage="<text> lines selector format is column=value", widget="String", required=false, def="Type de notice=Notice sommaire")

		def textSelector;

@Field @Option(name="textMetadataColumnList", usage="<text> metadata columns", widget="String", required=false, def="Identifiant de la notice,Date de diffusion,Durée,Genre,Identifiant Matériels (info.)")

		def textMetadataColumnList;

@Field @Option(name="textContentColumnList", usage="<text> textual content columns", widget="String", required=false, def="Titre propre,Résumé,Descripteurs (Aff. Lig.),Descripteurs (Aff. Col.),Séquences")

		def textContentColumnList;

@Field @Option(name="structuresSeparator", usage="Structures", widget="Separator", required=false, def="Structures")

		def structuresSeparator

@Field @Option(name="structureTag", usage="structure element to create", widget="String", required=false, def="div")

		def structureTag;

@Field @Option(name="structureSelector", usage="column_to_test=regexp", widget="String", required=false, def="Type de notice=Notice sujet")

		def structureSelector;

@Field @Option(name="textJoinColumn", usage="jointure column, values should point to the textIDColumn values", widget="String", required=false, def="Lien notice principale")

		def textJoinColumn;

@Field @Option(name="structureMetadataColumnList", usage="structure metadata columns", widget="String", required=false, def="Identifiant de la notice,Date de diffusion,Durée,Genre,Identifiant Matériels (info.)")

		def structureMetadataColumnList;

@Field @Option(name="structureContentColumnList", usage="structure textual content columns", widget="String", required=false, def="Titre propre,Résumé,Descripteurs (Aff. Lig.),Descripteurs (Aff. Col.),Séquences")

		def structureContentColumnList;

@Field @Option(name="typesSeparator", usage="Columns types", widget="Separator", required=false, def="Columns types")

		def typesSeparator

@Field @Option(name="dateColumnTypeList", usage="metadata columns of type=Date", widget="String", required=false, def="")

		def dateColumnTypeList

@Field @Option(name="prefixesColumnTypeList", usage="metadata columns of type=Prefixes", widget="String", required=false, def="")

		def prefixesColumnTypeList

@Field @Option(name="listColumnTypeList", usage="metadata columns of type=List semi-colon separated", widget="String", required=false, def="")

		def listColumnTypeList

		@Field @Option(name="debug", usage="Show devug messages", widget="Boolean", required=false, def="false")

		def debug

//@Field @Option(name="structureOrderColumn", usage="structure column coding structure order", widget="String", required=false, def="Titre propre,Résumé,Descripteurs (Aff. Lig.),Descripteurs (Aff. Col.),Séquences")

//		def structureOrderColumn;

if (!ParametersDialog.open(this)) return;

textMetadataColumnList = textMetadataColumnList.split(",").collect { it.trim() }

textContentColumnList = textContentColumnList.split(",").collect { it.trim() }

structureMetadataColumnList = structureMetadataColumnList.split(",").collect { it.trim() }

structureContentColumnList = structureContentColumnList.split(",").collect { it.trim() }

dateColumnTypeList = dateColumnTypeList.split(",").collect { it.trim() }

prefixesColumnTypeList = prefixesColumnTypeList.split(",").collect { it.trim() }

listColumnTypeList = listColumnTypeList.split(",").collect { it.trim() }

if (textSelector != null && textSelector.contains("=")) {

	textSelector = textSelector.trim()

	textSelector = [textSelector.substring(0, textSelector.indexOf("=")).trim(), textSelector.substring(textSelector.indexOf("=")+1).trim()]

} else {

	textSelector = [null, null]

}

if (structureSelector != null && structureSelector.contains("=")) {

	structureSelector = structureSelector.trim()

	structureSelector = [structureSelector.substring(0, structureSelector.indexOf("=")).trim(), structureSelector.substring(structureSelector.indexOf("=")+1).trim()]

} else {

	structureSelector = [null, null]

}

println "textIDColumn, textJoinColumn=$textIDColumn, $textJoinColumn"

println "textMetadataColumnList columns: $textMetadataColumnList"

println "textContentColumnList columns: $textContentColumnList"

println "structureMetadataColumnList columns: $structureMetadataColumnList"

println "structureContentColumnList columns: $structureContentColumnList"

println "text selector="+textSelector

println "structure selector="+structureSelector

println "structureTag="+structureTag

def reader = null

try {

	reader = new TableReader(inputFile);

	println "Reading $inputFile with TableReader..."

} catch(Exception e) {

	reader = new ReadExcel(inputFile, null);

	println "Reading $inputFile with ReadExcel..."

}

if (!reader.readHeaders()) {

	println "** Error: no header"

	return

}

def headers = Arrays.asList(reader.getHeaders())

println "$inputFile table column names: $headers"

def ok = true

def hash = ["selection":[textIDColumn, textJoinColumn], "textMetadataColumnList":textMetadataColumnList,

	"textContentColumnList":textContentColumnList, "structureMetadataColumnList":structureMetadataColumnList,

	"structureContentColumnList":structureContentColumnList]

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

	for (def m : hash[key]) {

		if (!headers.contains(m)) {

			println "** Error: missing $key column: $m"

			ok = false

		}

}

}

if (!ok) { return; }

// group records by text

def texts = new LinkedHashMap()

def nRecord = 0

def nRecordToWrite = 0

while (reader.readRecord()) {

nRecord++

//println "record="+reader.getRecord().get(textSelector[0])+" "+reader.getRecord().get(structureSelector[0])

String id = reader.get(textIDColumn).trim()

String join = reader.get(textJoinColumn).trim()

String textSelectorValue = reader.get(textSelector[0]).trim()

String structureSelectorValue = reader.get(structureSelector[0]).trim()

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

	if (textSelectorValue != null && structureSelectorValue != null) {

		if (textSelectorValue != null && textSelectorValue.matches(textSelector[1])) {

			if (!texts.containsKey(id)) texts[id] = []

			texts[id].add(0, reader.getRecord())

		} else if (structureSelectorValue != null && structureSelectorValue.matches(structureSelector[1])) {

			if (!texts.containsKey(join)) texts[join] = []

			texts[join].add(reader.getRecord())

			nRecordToWrite++

		} else {

			// ignore record

		}

	} else { // no text&structure selector set, take the record if 'id' or 'join' are set

		if (id.length() > 0) {

			if (!texts.containsKey(id)) texts[id] = []

			texts[id].add(0, reader.getRecord())

		} else if (join.length() > 0) {

			if (!texts.containsKey(join)) texts[join] = []

			texts[join].add(reader.getRecord())

			nRecordToWrite++

		} else {

			// ignore record

		}

	}

} else { // no id & join parameter set -> regroup all record in one file

	if (!texts.containsKey("result.xml")) texts["result.xml"] = []

	texts["result.xml"].add(reader.getRecord())

	nRecordToWrite++

}

}

println "N records: "+nRecord

println "N records to write: "+nRecordToWrite

println "N texts to build: "+texts.size()

if (texts.size()  == 0) {

println "No text found. Aborting."

return

}

outputDirectory.mkdir()

for (def id : texts.keySet()) {

def toWrite = texts[id]

def text = toWrite[0]

String textSelectorValue = text.get(textSelector[0]).trim()

if (textSelectorValue == null || (textSelectorValue != null && textSelectorValue.matches(textSelector[1]))) {

	if (debug) println "Processing text: $id"

	File outputfile = new File(outputDirectory, id+".xml")

	XMLOutputFactory factory = XMLOutputFactory.newInstance()

	FileOutputStream output = new FileOutputStream(outputfile)

	XMLStreamWriter writer = factory.createXMLStreamWriter(output, "UTF-8")

	writer.writeStartDocument("UTF-8","1.0")

	writer.writeCharacters("\n") // simple XML formating

	writer.writeStartElement("TEI")

	writer.writeCharacters("\n") // simple XML formating

	writer.writeStartElement("teiHeader")

	writer.writeEndElement() // teiHeader

	writer.writeCharacters("\n") // simple XML formating

	writer.writeStartElement("text")

	writer.writeAttribute("id", id)

	for (String att : textMetadataColumnList) {

		if (att in dateColumnTypeList) {

			writeMetadataDate(text, att, writer)

		} else {

			writer.writeAttribute(AsciiUtils.buildAttributeId(att), text.get(att)) // struct

		}

	}

	writer.writeCharacters("\n") // simple XML formating

	writer.writeStartElement("metadata")

	writer.writeStartElement("list")

	writer.writeAttribute("type", "unordered")

	writer.writeCharacters("\n")

	textMetadataColumnList.each { att ->

		writer.writeStartElement("item")

		writer.writeCharacters(att+" : "+text.get(att).replaceAll("\n", ";"))

		writer.writeEndElement() // item

		writer.writeCharacters("\n")

	}

	writer.writeEndElement() // list

	writer.writeEndElement() // metadata

	writer.writeCharacters("\n")

	for (String att : textContentColumnList) {

		if (att in prefixesColumnTypeList) {

			writePrefixTextContent(text, att, writer)

		} else if (att in listColumnTypeList) {

			writeListTextContent(text, att, writer)

		} else {

			writer.writeStartElement("p");

			writer.writeAttribute("type", att.trim())

			writer.writeCharacters("\n") // simple XML formating

			writer.writeStartElement("head")

			writer.writeStartElement("hi")

			writer.writeCharacters(att+" : ")

			writer.writeEndElement() // hi

			writer.writeEndElement() // head

			writer.writeCharacters(text.get(att)) // get textColumnList content

			writer.writeEndElement() // t

			writer.writeCharacters("\n") // simple XML formating)

		}

	}

	int pb_n = 1;

	for (int i = 1 ; i < toWrite.size() ; i++) {

		def record = toWrite[i]

		writer.writeEmptyElement("pb") // <pb/>

		writer.writeAttribute("n", ""+pb_n++)

		writer.writeAttribute("type", "record")

		writer.writeStartElement(structureTag)

		for (String att : structureMetadataColumnList) {

			if (att in dateColumnTypeList) {

				writeMetadataDate(record, att, writer)

			} else {

				writer.writeAttribute(AsciiUtils.buildAttributeId(att), record.get(att)) // struct

			}

		}

		writer.writeCharacters("\n")

		writer.writeStartElement("metadata")

		writer.writeStartElement("list")

		writer.writeAttribute("type", "unordered")

		writer.writeCharacters("\n")

		structureMetadataColumnList.each { att ->

			writer.writeStartElement("item")

			writer.writeCharacters(att+" : "+record.get(att).replaceAll("\n", ";"))

			writer.writeEndElement() // item

			writer.writeCharacters("\n")

		}

		writer.writeEndElement() // list

		writer.writeEndElement() // metadata

		writer.writeCharacters("\n")

		for (String att : structureContentColumnList) {

			if (att in prefixesColumnTypeList) {

				writePrefixTextContent(text, att, writer)

			} else if (att in listColumnTypeList) {

				writeListTextContent(text, att, writer)

			} else {

				writer.writeStartElement("p");

				writer.writeAttribute("type", att.trim())

				writer.writeAttribute("id", record.get(textIDColumn))

				writer.writeCharacters("\n") // simple XML formating

				writer.writeStartElement("head")

				writer.writeStartElement("hi")

				writer.writeCharacters(att+" : ")

				writer.writeEndElement() // hi

				writer.writeEndElement() // head

				writer.writeCharacters(record.get(att)) // get textColumnList content

				writer.writeEndElement() // t

				writer.writeCharacters("\n") // simple XML formating

			}

		}

		writer.writeEndElement() // struct

		writer.writeCharacters("\n") // simple XML formating

	}

	writer.writeEndElement() // text

	writer.writeCharacters("\n") // simple XML formating

	writer.writeEndElement() // TEI

	writer.close()

	output.close()

	reader.close()

} else {

	// error

	println "ERROR: '$id' text group with  no text line"

}

}

def writeListTextContent(def record, def att, def writer) {

writer.writeCharacters("\n")

found = false

def value = record.get(att)

value.findAll( /(?s)([^\n]+?)[\n]/ ).each { desc ->

	found = true

	writer.writeCharacters("\t")

	writer.writeStartElement("p")

	writer.writeAttribute("rend", "list")

	matches = (desc =~ /(?s)([^\n]+?)[\n]/)

	writer.writeCharacters(matches[0][1])

	writer.writeEndElement() // p

	writer.writeCharacters("\n")

}

if (!found) {

	writer.writeCharacters("\t")

	writer.writeStartElement("p")

	writer.writeAttribute("rend", "no-list")

	writer.writeCharacters(value)

	writer.writeEndElement() // p

	writer.writeCharacters("\n")

}

}

def writePrefixTextContent(def record, def att, def writer) {

writer.writeCharacters("\n")

writer.writeStartElement("list")

writer.writeAttribute("rend", "prefixes")

writer.writeAttribute("type", "unordered")

writer.writeCharacters("\n")

found = false

def value = record.get(att)

value.findAll( /(?s)[A-Z]{3}:? *([^;\n]+?) +[;\n]/ ).each { desc ->

	found = true

	writer.writeCharacters("\t")

	writer.writeStartElement("item")

	matches = (desc =~ /(?s)([A-Z]{3}):? *([^;\n]+?) +[;\n]/)

	writer.writeAttribute("type", matches[0][1])

	writer.writeStartElement("span")

	writer.writeCharacters(matches[0][1]+" ")

	writer.writeEndElement() // span

	writer.writeCharacters(matches[0][2])

	writer.writeEndElement() // item

	writer.writeCharacters("\n")

}

if (!found) {

	writer.writeCharacters(value)

}

writer.writeEndElement() // list

writer.writeCharacters("\n")

}

def writeMetadataDate(def record, def att, def writer) {

String value = record.get(att)

String att_normalized = AsciiUtils.buildAttributeId(att)

matches = (value =~ /([0-9]{2})\/([0-9]{2})\/([0-9]{4})/)

writer.writeAttribute(att_normalized+"jour", matches[0][1])

writer.writeAttribute(att_normalized+"joursemaine", new java.text.SimpleDateFormat('EEEE').format(Date.parse("dd/MM/yyyy", value)))

writer.writeAttribute(att_normalized+"mois", matches[0][2])

writer.writeAttribute(att_normalized+"annee", matches[0][3])

writer.writeAttribute(att_normalized+"tri", matches[0][3]+"-"+matches[0][2]+"-"+matches[0][1])

}

/*

 String name = inputFile.getName()

 int idx = name.lastIndexOf(".")

 if (idx > 0) name = name.substring(0, idx)

 */