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.hssf.usermodel.HSSFCell;
031import org.apache.poi.hssf.usermodel.HSSFRow;
032import org.apache.poi.hssf.usermodel.HSSFSheet;
033import org.apache.poi.hssf.usermodel.HSSFWorkbook;
034import org.apache.poi.poifs.filesystem.POIFSFileSystem;
035import org.apache.poi.ss.usermodel.Row;
036import org.nuxeo.ecm.core.api.Blobs;
037import org.nuxeo.ecm.core.api.blobholder.BlobHolder;
038import org.nuxeo.ecm.core.convert.api.ConversionException;
039import org.nuxeo.ecm.core.convert.cache.SimpleCachableBlobHolder;
040import org.nuxeo.ecm.core.convert.extension.Converter;
041import org.nuxeo.ecm.core.convert.extension.ConverterDescriptor;
042
043public class XL2TextConverter implements Converter {
044
045    private static final Log log = LogFactory.getLog(XL2TextConverter.class);
046
047    private static final String CELL_SEP = " ";
048
049    private static final String ROW_SEP = "\n\n";
050
051    @Override
052    public BlobHolder convert(BlobHolder blobHolder, Map<String, Serializable> parameters) throws ConversionException {
053
054        StringBuilder sb = new StringBuilder();
055        try (InputStream stream = blobHolder.getBlob().getStream(); //
056                POIFSFileSystem fs = new POIFSFileSystem(stream); //
057                HSSFWorkbook workbook = new HSSFWorkbook(fs)) {
058            for (int i = 0; i < workbook.getNumberOfSheets(); i++) {
059                HSSFSheet sheet = workbook.getSheetAt(i);
060                Iterator<Row> rows = sheet.rowIterator();
061                while (rows.hasNext()) {
062                    HSSFRow row = (HSSFRow) rows.next();
063                    Iterator<?> cells = row.cellIterator();
064                    while (cells.hasNext()) {
065                        HSSFCell cell = (HSSFCell) cells.next();
066                        appendTextFromCell(cell, sb);
067                        sb.append(CELL_SEP);
068                    }
069                    sb.append(ROW_SEP);
070                }
071            }
072            return new SimpleCachableBlobHolder(Blobs.createBlob(sb.toString()));
073        } catch (IOException e) {
074            throw new ConversionException("Error during XL2Text conversion", blobHolder, e);
075        }
076    }
077
078    protected void appendTextFromCell(HSSFCell cell, StringBuilder sb) {
079        String cellValue = null;
080        switch (cell.getCellType()) {
081        case NUMERIC:
082            cellValue = Double.toString(cell.getNumericCellValue()).trim();
083            break;
084        case STRING:
085            cellValue = cell.getStringCellValue().trim().replaceAll("\n", " ");
086            break;
087        }
088
089        if (cellValue != null && cellValue.length() > 0) {
090            sb.append(cellValue);
091        }
092    }
093
094    @Override
095    public void init(ConverterDescriptor descriptor) {
096    }
097
098}