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.

//

// 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: 2017-04-11 15:30:35 +0200 (mar. 11 avril 2017) $

// $LastChangedRevision: 3426 $

// $LastChangedBy: mdecorde $ 

//

package org.txm.scripts.importer;

import org.txm.utils.i18n.DetectBOM;

import org.w3c.dom.Document;

import org.w3c.dom.Element;

import org.w3c.dom.NodeList;

import org.xml.sax.SAXException;

import java.io.File;

import java.io.FileInputStream;

import java.io.InputStreamReader;

import javax.xml.stream.*;

import java.net.URL;

// TODO: Auto-generated Javadoc

/**

 * The Class FortranColumns2XML.

 */

class FortranColumns2XML {

        /** The infile. */

        File infile;

        /** The outfile. */

        File outfile;

        /** The encoding. */

        String encoding;

        /** The column names. */

        List<String> columnNames

        /** The word tag. */

        String wordTag;

        /** The wordcolumn. */

        int wordcolumn;

        /** The linecounter. */

        int linecounter = 0;

        /** The check line number. */

        int checkLineNumber;

        /** The check line number tag. */

        String checkLineNumberTag;

        /** The columnindex. */

        private List<Integer> columnindex;

        /** The endcolumnindex. */

        private List<Integer> endcolumnindex;

        /** The line. */

        private String line;

        /** The writer. */

        private XMLStreamWriter writer;

        /** The reader. */

        private Reader reader;

        /** The output. */

        private FileOutputStream output;

        String textname;

        /**

         * Instantiates a new fortran columns2 xml.

         *

         * @param infile the infile

         * @param outfile the outfile

         * @param encoding the encoding

         */

        FortranColumns2XML(File infile, File outfile, String encoding)

        {

                this.infile = infile;

                this.outfile = outfile;

                textname = infile.getName();

                int idx = -1;

                idx = textname.lastIndexOf(".");

                if (idx > 0)

                        textname = textname.substring(0, idx);

                def input = new FileInputStream(infile)                 

                reader = new InputStreamReader(input , encoding);

                DetectBOM bomdetector = new DetectBOM(infile);

                for (int ibom = 0 ; ibom < bomdetector.getBOMSize() ; ibom++) input.read()

                XMLOutputFactory factory = XMLOutputFactory.newInstance();

                output = new FileOutputStream(outfile)

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

        }

        /**

         * Process.

         *

         * @return true, if successful

         */

        public boolean process()

        {

                line = reader.readLine(); // use the first line to find columns 

                if(columnindex == null || endcolumnindex == null)

                {

                        searchNumberOfColumns();

                }

                if(columnindex.size() != columnNames.size())

                {

                        System.err.println("Number of column("+columnindex.size()+") different of number of column name("+columnNames.size()+")");

                        return false;

                }

                //write start document

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

                writer.writeStartElement("text")

                writer.writeAttribute("id",infile.getName())

                boolean gogogo = true;

                while(gogogo)

                {        

                        //println "line : "+line

                        processLine();//split line accordinly to columnindexes

                        linecounter++;

                        //continue to read the document

                        line = reader.readLine();

                        if(line == null)

                                gogogo= false;

                }

                writer.writeEndElement();

                writer.writeEndDocument();

                writer.close();

                output.close();

                return true;

        }

        /**

         * Process line.

         */

        int wordcount = 1;

        private void processLine()

        {

                String pcdata;

                if(line.length() > 0)

                {

                        writer.writeStartElement(wordTag)

                        if(setIds)

                                writer.writeAttribute("id", "w_"+textname+"_"+(wordcount++));

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

                        {

                                if(line.length() > endcolumnindex.get(i))

                                {

                                        String substring = line.substring(columnindex.get(i), endcolumnindex.get(i))

                                        if(i != wordcolumn)

                                        {

                                                writer.writeAttribute(columnNames.get(i), substring.trim())

                                        }

                                        else

                                        {

                                                pcdata = substring.trim()

                                        }

                                }

                                else

                                {

                                        println "Error: line $linecounter: len="+line.length()+" maxidx="+endcolumnindex.get(i)

                                        println line;

                                }

                        }

                        writer.writeCharacters(pcdata);

                        writer.writeEndElement();

                }

        }

        /**

         * Sets the column indexes.

         *

         * @param starts the starts

         * @param ends the ends

         */

        private void setColumnIndexes(List<Integer> starts, List<Integer> ends)

        {

                columnindex = starts;

                endcolumnindex = ends;

        }

        boolean setIds = false;

        public void setAddIds(boolean setIds)

        {

                this.setIds = setIds;

        }

        /**

         * Search number of columns.

         */

        private void searchNumberOfColumns()

        {

                columnindex = new ArrayList<Integer>();

                endcolumnindex = new ArrayList<Integer>();

                columnindex.add(0);

                int ic = 0

                char c;

                char[] cline = line.toCharArray();

                int state = 0; // 0 lettre, 1 espace

                for(ic = 0 ; ic < cline.length ; ic++)

                {

                        c = cline[ic];

                        if(state == 0)// zone lettre

                        {

                                if(c == " ")

                                {

                                        state = 1;

                                }

                        }

                        else // zone espace

                        {

                                if(c != " ")

                                {

                                        columnindex.add(ic);

                                        endcolumnindex.add(ic-1);

                                        state = 0;

                                }

                        }

                }

                endcolumnindex.add(cline.length);

                println columnindex;

                println endcolumnindex;

        }

        /**

         * Sets the columns lines names.

         *

         * @param columnNames the column names

         * @param unitTag the unit tag

         * @param wordcolumn the wordcolumn

         */

        public void setColumnsLinesNames(List<String> columnNames, String unitTag, int wordcolumn)

        {

                this.columnNames = columnNames;

                this.wordTag = unitTag;

                this.wordcolumn = wordcolumn;

        }

        /**

         * Sets the line check.

         *

         * @param colindex the colindex

         * @param tagname the tagname

         */

        public void setLineCheck(int colindex, String tagname)

        {

                this.checkLineNumber = colindex;

                this.checkLineNumberTag = tagname;

        }

        /**

         * The main method.

         *

         * @param args the arguments

         */

        public static void main(String[] args)

        {

                File fortranfile = new File(System.getProperty("user.home"),"\\xml\\lasla\\lasla\\L01PlaAmphi.APN");

                File outfile = new File(System.getProperty("user.home"),"\\xml\\lasla\\out.xml");

                FortranColumns2XML builder = new FortranColumns2XML(fortranfile, outfile, "iso-8859-1");

                builder.setColumnIndexes([0, 3, 4, 8, 29, 55, 67],        [3, 4, 8, 28, 54, 66, 79]);

                builder.setColumnsLinesNames(["ref","s","sent","word","lemme","line","pos"], "w", 3);

                builder.setLineCheck(3, "l");

                builder.process();

        }

}