001/*
002 * (C) Copyright 2006-2007 Nuxeo SA (http://nuxeo.com/) and others.
003 *
004 * Licensed under the Apache License, Version 2.0 (the "License");
005 * you may not use this file except in compliance with the License.
006 * You may obtain a copy of the License at
007 *
008 *     http://www.apache.org/licenses/LICENSE-2.0
009 *
010 * Unless required by applicable law or agreed to in writing, software
011 * distributed under the License is distributed on an "AS IS" BASIS,
012 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
013 * See the License for the specific language governing permissions and
014 * limitations under the License.
015 *
016 * Contributors:
017 *     Nuxeo - initial API and implementation
018 *
019 */
020package org.nuxeo.ecm.core.convert.plugins.text.extractors;
021
022import java.io.IOException;
023import java.io.InputStream;
024import java.io.Serializable;
025import java.util.Iterator;
026import java.util.Map;
027
028import org.apache.commons.logging.Log;
029import org.apache.commons.logging.LogFactory;
030import org.apache.poi.openxml4j.exceptions.OpenXML4JException;
031import org.apache.poi.openxml4j.opc.OPCPackage;
032import org.apache.poi.ss.usermodel.Cell;
033import org.apache.poi.ss.usermodel.Row;
034import org.apache.poi.xssf.usermodel.XSSFCell;
035import org.apache.poi.xssf.usermodel.XSSFRow;
036import org.apache.poi.xssf.usermodel.XSSFSheet;
037import org.apache.poi.xssf.usermodel.XSSFWorkbook;
038import org.nuxeo.ecm.core.api.Blob;
039import org.nuxeo.ecm.core.api.Blobs;
040import org.nuxeo.ecm.core.api.blobholder.BlobHolder;
041import org.nuxeo.ecm.core.convert.api.ConversionException;
042import org.nuxeo.ecm.core.convert.cache.SimpleCachableBlobHolder;
043import org.nuxeo.ecm.core.convert.extension.Converter;
044
045public class XLX2TextConverter extends BaseOfficeXMLTextConverter implements Converter {
046
047    private static final Log log = LogFactory.getLog(XLX2TextConverter.class);
048
049    private static final String CELL_SEP = " ";
050
051    private static final String ROW_SEP = "\n";
052
053    @Override
054    public BlobHolder convert(BlobHolder blobHolder, Map<String, Serializable> parameters) throws ConversionException {
055
056        StringBuffer sb = new StringBuffer();
057
058        try {
059            Blob blob = blobHolder.getBlob();
060
061            if (blob.getLength() > maxSize4POI) {
062                return runFallBackConverter(blobHolder, "xl/");
063            }
064
065            try (InputStream stream = blob.getStream(); //
066                    OPCPackage p = OPCPackage.open(stream); //
067                    XSSFWorkbook workbook = new XSSFWorkbook(p)) {
068                for (int i = 0; i < workbook.getNumberOfSheets(); i++) {
069                    XSSFSheet sheet = workbook.getSheetAt(i);
070                    Iterator<Row> rows = sheet.rowIterator();
071                    while (rows.hasNext()) {
072                        XSSFRow row = (XSSFRow) rows.next();
073                        Iterator<Cell> cells = row.cellIterator();
074                        while (cells.hasNext()) {
075                            XSSFCell cell = (XSSFCell) cells.next();
076                            appendTextFromCell(cell, sb);
077                        }
078                        sb.append(ROW_SEP);
079                    }
080                }
081            }
082            return new SimpleCachableBlobHolder(Blobs.createBlob(sb.toString()));
083        } catch (IOException | OpenXML4JException e) {
084            throw new ConversionException("Error during XLX2Text conversion", e);
085        }
086    }
087
088    protected void appendTextFromCell(XSSFCell cell, StringBuffer sb) {
089        String cellValue = null;
090        switch (cell.getCellType()) {
091        case XSSFCell.CELL_TYPE_NUMERIC:
092            cellValue = Double.toString(cell.getNumericCellValue()).trim();
093            break;
094        case XSSFCell.CELL_TYPE_STRING:
095            cellValue = cell.getStringCellValue().trim();
096            break;
097        }
098
099        if (cellValue != null && cellValue.length() > 0) {
100            sb.append(cellValue).append(CELL_SEP);
101        }
102    }
103
104}