diff --git a/datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/server/plugins/standard/AbstractFileTableReportingPlugin.java b/datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/server/plugins/standard/AbstractFileTableReportingPlugin.java
index e8c6edc47d6112a210ed421f148a0b49fad6c81d..405d997d1953b6abcf92bc87eaa4fbfdb7c9851c 100644
--- a/datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/server/plugins/standard/AbstractFileTableReportingPlugin.java
+++ b/datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/server/plugins/standard/AbstractFileTableReportingPlugin.java
@@ -24,12 +24,7 @@ import java.util.ArrayList;
import java.util.List;
import java.util.Properties;
-import org.apache.commons.io.FilenameUtils;
-import org.apache.poi.hssf.usermodel.HSSFWorkbook;
-import org.apache.poi.poifs.filesystem.POIFSFileSystem;
-import org.apache.poi.ss.usermodel.Sheet;
import org.apache.poi.ss.usermodel.Workbook;
-import org.apache.poi.xssf.usermodel.XSSFWorkbook;
import com.csvreader.CsvReader;
@@ -40,7 +35,7 @@ import ch.systemsx.cisd.common.parser.ParsingException;
import ch.systemsx.cisd.common.utilities.PropertyUtils;
import ch.systemsx.cisd.openbis.dss.generic.shared.utils.CodeAndLabelUtil;
import ch.systemsx.cisd.openbis.dss.generic.shared.utils.DatasetFileLines;
-import ch.systemsx.cisd.openbis.dss.generic.shared.utils.ExcelFileReaderHelper;
+import ch.systemsx.cisd.openbis.dss.generic.shared.utils.ExcelFileReader;
import ch.systemsx.cisd.openbis.generic.shared.basic.TableCellUtil;
import ch.systemsx.cisd.openbis.generic.shared.basic.dto.CodeAndLabel;
import ch.systemsx.cisd.openbis.generic.shared.basic.dto.ISerializableComparable;
@@ -110,12 +105,13 @@ abstract public class AbstractFileTableReportingPlugin extends AbstractTableMode
assert file != null : "Given file must not be null";
assert file.isFile() : "Given file '" + file.getAbsolutePath() + "' is not a file.";
- if (isExcelFile(file))
+ if (ExcelFileReader.isExcelFile(file))
{
try
{
- Sheet sheet = getExcelSheet(file);
- return load(dataset, sheet, file);
+ Workbook workbook = ExcelFileReader.getExcelWorkbook(file);
+ ExcelFileReader reader = new ExcelFileReader(workbook, ignoreComments);
+ return load(dataset, reader, file);
} catch (final IOException ex)
{
throw new IOExceptionUnchecked(ex);
@@ -140,47 +136,24 @@ abstract public class AbstractFileTableReportingPlugin extends AbstractTableMode
}
}
- private static boolean isExcelFile(File file)
+ /**
+ * Loads data from the specified reader.
+ */
+ private DatasetFileLines load(DatasetDescription dataset, ExcelFileReader reader,
+ File file) throws IOException
{
- return FilenameUtils.isExtension(file.getName().toLowerCase(), new String[]
- { "xls", "xlsx" });
- }
+ assert reader != null : "Unspecified reader";
- private Sheet getExcelSheet(File file) throws IOException
- {
- Workbook wb = getExcelWorkbook(file);
+ List<String[]> lines = null;
try
{
int index = Integer.parseInt(excelSheet);
- return wb.getSheetAt(index); // will throw exception if index is out of range
+ lines = reader.readLines(index); // will throw exception if index is out of range
} catch (NumberFormatException ex)
{
- Sheet sheet = wb.getSheet(excelSheet);
- if (sheet == null)
- {
- throw new UserFailureException(file.getName() + " doesn't contain sheet named "
- + excelSheet);
- }
- return sheet;
- }
- }
-
- private Workbook getExcelWorkbook(File file) throws IOException
- {
- final String extension = FilenameUtils.getExtension(file.getName()).toLowerCase();
- final FileInputStream stream = new FileInputStream(file);
- if ("xls".equals(extension))
- {
- POIFSFileSystem poifsFileSystem = new POIFSFileSystem(stream);
- return new HSSFWorkbook(poifsFileSystem);
- } else if ("xlsx".equals(extension))
- {
- return new XSSFWorkbook(stream);
- } else
- {
- throw new IllegalArgumentException(
- "Expected an Excel file with 'xls' or 'xlsx' extension, got " + file.getName());
+ lines = reader.readLines(excelSheet);
}
+ return new DatasetFileLines(file, dataset.getDataSetCode(), lines, ignoreTrailingEmptyCells);
}
private static CsvReader readFile(File file, boolean ignoreComments, char separator)
@@ -203,20 +176,6 @@ abstract public class AbstractFileTableReportingPlugin extends AbstractTableMode
return csvReader;
}
- /**
- * Loads data from the specified sheet.
- *
- * @throws IOException
- */
- private DatasetFileLines load(DatasetDescription dataset, Sheet sheet, File file)
- throws IOException
- {
- assert sheet != null : "Unspecified sheet";
-
- List<String[]> lines = ExcelFileReaderHelper.loadLines(sheet, ignoreComments);
- return new DatasetFileLines(file, dataset.getDataSetCode(), lines, ignoreTrailingEmptyCells);
- }
-
/**
* Loads data from the specified reader.
*/
diff --git a/datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/shared/utils/ExcelFileReaderHelper.java b/datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/shared/utils/ExcelFileReader.java
similarity index 64%
rename from datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/shared/utils/ExcelFileReaderHelper.java
rename to datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/shared/utils/ExcelFileReader.java
index 0cb1069d27936c65dada22d235bd5e1c958afadd..3639f1dd844462e55ae3233a4c922f1ea3dcdeec 100644
--- a/datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/shared/utils/ExcelFileReaderHelper.java
+++ b/datastore_server/source/java/ch/systemsx/cisd/openbis/dss/generic/shared/utils/ExcelFileReader.java
@@ -24,7 +24,7 @@ import java.util.Arrays;
import java.util.List;
import org.apache.commons.io.FilenameUtils;
-import org.apache.poi.hssf.extractor.ExcelExtractor;
+import org.apache.log4j.Logger;
import org.apache.poi.hssf.usermodel.HSSFWorkbook;
import org.apache.poi.poifs.filesystem.POIFSFileSystem;
import org.apache.poi.ss.usermodel.Cell;
@@ -32,85 +32,102 @@ import org.apache.poi.ss.usermodel.Row;
import org.apache.poi.ss.usermodel.Sheet;
import org.apache.poi.ss.usermodel.Workbook;
import org.apache.poi.ss.util.CellReference;
-import org.apache.poi.xssf.extractor.XSSFExcelExtractor;
import org.apache.poi.xssf.usermodel.XSSFWorkbook;
import ch.systemsx.cisd.common.exceptions.UserFailureException;
+import ch.systemsx.cisd.common.logging.LogCategory;
+import ch.systemsx.cisd.common.logging.LogFactory;
/**
+ * Reader of tabular data from excel files.
+ *
* @author Piotr Buczek
*/
-public class ExcelFileReaderHelper
+public class ExcelFileReader
{
- public static void main(String[] args) throws Exception
+ private static final Logger operationLog = LogFactory.getLogger(LogCategory.OPERATION,
+ ExcelFileReader.class);
+
+ private final boolean ignoreComments;
+
+ private final Workbook workbook;
+
+ public ExcelFileReader(Workbook workbook, boolean ignoreComments)
{
- String filename = args[0];
- File file = new File(filename);
- List<String[]> lines = loadExcelFile(file, true);
- for (String[] line : lines)
+ this.ignoreComments = ignoreComments;
+ this.workbook = workbook;
+ }
+
+ public List<String[]> readLines() throws IOException
+ {
+ Sheet sheet = workbook.getSheetAt(0);
+ return loadLines(sheet, ignoreComments);
+ }
+
+ public List<String[]> readLines(int sheetIndex) throws IOException
+ {
+ Sheet sheet = workbook.getSheetAt(sheetIndex);
+ return loadLines(sheet, ignoreComments);
+ }
+
+ public List<String[]> readLines(String sheetName) throws IOException
+ {
+ Sheet sheet = workbook.getSheet(sheetName);
+ if (sheet == null)
{
- System.err.println(Arrays.toString(line));
+ throw new UserFailureException("Couldn't find sheet named " + sheetName);
}
+ return loadLines(sheet, ignoreComments);
+ }
+
+ // if the line starts with this character and comments should be ignored, the line is ignored
+ private static final String COMMENT = "#";
+
+ /**
+ * Uses file extension to figure out if given <var>file</var> is supported Excel file.
+ *
+ * @return <code>true</code> if the <var>file</var> is an XSL or XLSX Excel file
+ */
+ public static boolean isExcelFile(File file)
+ {
+ return FilenameUtils.isExtension(file.getName().toLowerCase(), new String[]
+ { "xls", "xlsx" });
}
- public static List<String[]> loadExcelFile(File file, boolean ignoreComments)
- throws IOException
+ /**
+ * @return {@link Workbook} of an Excel <var>file</var>
+ * @throws IOException if an I/O problem occurs
+ * @throws IllegalArgumentException if the <var>file</var> is a format that is not supported
+ */
+ public static Workbook getExcelWorkbook(File file) throws IOException, IllegalArgumentException
{
final String extension = FilenameUtils.getExtension(file.getName()).toLowerCase();
final FileInputStream stream = new FileInputStream(file);
- Workbook wb = null;
if ("xls".equals(extension))
{
POIFSFileSystem poifsFileSystem = new POIFSFileSystem(stream);
- HSSFWorkbook hssfWorkbook = new HSSFWorkbook(poifsFileSystem);
- wb = hssfWorkbook;
- // HSSFWorkbook wb = new HSSFWorkbook(poifsFileSystem);
-
- ExcelExtractor extractor = new ExcelExtractor(hssfWorkbook);
- // HSSFFormulaEvaluator.evaluateAllFormulaCells(hssfWorkbook);
- //
- extractor.setIncludeBlankCells(true);
- extractor.setIncludeSheetNames(false);
- String text = extractor.getText();
- System.out.println(text);// returns TSV file
-
+ return new HSSFWorkbook(poifsFileSystem);
} else if ("xlsx".equals(extension))
{
- // spaces are not ignored
- XSSFWorkbook xssfWorkbook = new XSSFWorkbook(stream);
- wb = xssfWorkbook;
-
- // ExcelExtractor extractor = new ExcelExtractor(xssfWorkbook);
- //
- XSSFExcelExtractor extractor = new XSSFExcelExtractor(xssfWorkbook);
- // XSSFFormulaEvaluator evaluator = new XSSFFormulaEvaluator(xssfWorkbook);
- // XSSFFormulaEvaluator.evaluateAllFormulaCells(xssfWorkbook);
-
- extractor.setIncludeSheetNames(false);
- String text = extractor.getText();
- System.out.println(text);// returns TSV file
+ return new XSSFWorkbook(stream);
} else
{
- System.err.println(extension);
throw new IllegalArgumentException(
"Expected an Excel file with 'xls' or 'xlsx' extension, got " + file.getName());
}
-
- final Sheet sheet = wb.getSheetAt(0);
-
- return loadLines(sheet, ignoreComments);
}
- // if the line starts with this character and comments should be ignored, the line is ignored
- private static final String COMMENT = "#";
+ //
+ // helper methods
+ //
- public static List<String[]> loadLines(Sheet sheet, boolean ignoreComments) throws IOException
+ private static List<String[]> loadLines(Sheet sheet, boolean ignoreComments) throws IOException
{
final List<String[]> lines = new ArrayList<String[]>();
boolean firstLine = true;
- // NOTE: the following code is pretty ugly - it is because poi API is very limited
+ // NOTE: the following code is pretty ugly because poi API is very limited
int headerSize = 0;
for (Row row : sheet)
{
@@ -126,7 +143,10 @@ public class ExcelFileReaderHelper
for (Cell cell : row)
{
String value = extractCellValue(cell);
- System.out.println("cell " + extractCellPosition(cell) + ":" + value);
+ if (operationLog.isDebugEnabled())
+ {
+ operationLog.debug(extractCellPosition(cell) + ": " + value);
+ }
header[cell.getColumnIndex()] = value;
}
lines.add(header);
@@ -137,7 +157,10 @@ public class ExcelFileReaderHelper
for (Cell cell : row)
{
String value = extractCellValue(cell);
- System.out.println("cell " + extractCellPosition(cell) + ":" + value);
+ if (operationLog.isDebugEnabled())
+ {
+ operationLog.debug(extractCellPosition(cell) + ": " + value);
+ }
if (cell.getColumnIndex() >= line.length)
{
continue; // ignore for now
@@ -148,10 +171,14 @@ public class ExcelFileReaderHelper
}
}
- for (String[] line : lines)
+ if (operationLog.isDebugEnabled())
{
- System.err.println(Arrays.toString(line));
+ for (String[] line : lines)
+ {
+ operationLog.debug(Arrays.toString(line));
+ }
}
+
return lines;
}
@@ -197,4 +224,21 @@ public class ExcelFileReaderHelper
String row = "" + (cell.getRowIndex() + 1);
return col + row;
}
+
+ //
+ // for testing
+ //
+
+ public static void main(String[] args) throws Exception
+ {
+ String filename = args[0];
+ File file = new File(filename);
+ Workbook wb = getExcelWorkbook(file);
+ ExcelFileReader helper = new ExcelFileReader(wb, true);
+ List<String[]> lines = helper.readLines();
+ for (String[] line : lines)
+ {
+ System.err.println(Arrays.toString(line));
+ }
+ }
}