使用POI和dom4j来解析Excel生成Xml

工作中有时候我们需要使用Apache的POI来读写Excel文件，而dom4j则用来解析或者生成Xml文件，由于涉及到IO流，所以又使用了commons-io提供的IOUtils来关闭流。

需要的Jar包

POI和dom4j需要的jar包

一个应用POI和dom4j的简单例子

import java.io.File;
import java.io.FileInputStream;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.util.ArrayList;
import java.util.List;

import org.apache.commons.io.IOUtils;
import org.apache.poi.ss.usermodel.Cell;
import org.apache.poi.xssf.usermodel.XSSFCell;
import org.apache.poi.xssf.usermodel.XSSFRow;
import org.apache.poi.xssf.usermodel.XSSFSheet;
import org.apache.poi.xssf.usermodel.XSSFWorkbook;
import org.dom4j.Document;
import org.dom4j.DocumentHelper;
import org.dom4j.Element;
import org.dom4j.io.OutputFormat;
import org.dom4j.io.XMLWriter;

public class MyPOI {

private boolean elemStarted = false;
private boolean attrStarted = false;
Element tabNode;
Element elemsNode;
Element elemNode;
Element subNode;
private final List<String> list = new ArrayList<String>();

public void parseExcel(final String src, final String target) throws Exception {

XSSFWorkbook workbook = null;
XSSFSheet sheet = null;
XMLWriter writer = null;
InputStream fin = null;
OutputStream fos = null;
final File srcFile = new File(src);
final String fileName = srcFile.getName();
final Document doc = DocumentHelper.createDocument();
// create the root element
final Element root = doc.addElement(fileName.substring(0, fileName.indexOf('.')));
root.addAttribute("position", fileName);
try {
fin = new FileInputStream(srcFile);
fos = new FileOutputStream(target);
workbook = new XSSFWorkbook(fin);
final int sheetNum = workbook.getNumberOfSheets();
for (int i = 0; i < sheetNum; i++) {
sheet = workbook.getSheetAt(i);
if (sheet == null) {
continue;
}
// create the sheet element
final Element sheetNode = root.addElement("sheet");
sheetNode.addAttribute("id", sheet.getSheetName());
sheetNode.addAttribute("position", fileName + "," + sheet.getSheetName());
final int rowNum = sheet.getLastRowNum();
for (int j = 0; j <= rowNum; j++) {
final XSSFRow row = sheet.getRow(j);
if (row == null) {
continue;
}
final XSSFCell firstCell = row.getCell(0);
final String firstCellValue = getCellValue(firstCell);
if (firstCell == null || "".equals(firstCellValue)) {
continue;
}
if (firstCellValue.startsWith("#end_attr")) {
attrStarted = false;
continue;
}
if (firstCellValue.startsWith("#end_elem")) {
elemStarted = false;
continue;
}
if (firstCellValue.startsWith("##")) {
// create the tab element
c6a8

tabNode = sheetNode.addElement(firstCellValue.substring(2));
tabNode.addAttribute("position", fileName + "," + sheet.getSheetName() + "," + (j+1));
} else if (firstCellValue.startsWith("#begin")) {
j++;
final XSSFRow nextRow = sheet.getRow(j);
final int nextRowCellNum = nextRow.getLastCellNum();
if (firstCellValue.startsWith("#begin_attr")) {
attrStarted = true;
} else if (firstCellValue.startsWith("#begin_elem")) {
elemStarted = true;
// create the elements element
elemsNode = tabNode.addElement("elements");
if (firstCellValue.indexOf(":") == -1) {
elemsNode.addAttribute("id", "default");
} else {
elemsNode.addAttribute("id", firstCellValue.split(":")[1]);
}
}
if (!list.isEmpty()) {
list.clear();
}
for (int k = 0; k < nextRowCellNum; k++) {
final XSSFCell nextRowCell = nextRow.getCell(k);
String cellValue = getCellValue(nextRowCell);
if (nextRowCell != null && !"".equals(cellValue)) {
if (cellValue.endsWith("*")) {
cellValue = cellValue.substring(0, cellValue.length() - 1);
}
list.add(cellValue);
}
}
} else {
processRow(row);
}
}
}
writer = new XMLWriter(fos, OutputFormat.createPrettyPrint());
writer.write(doc);
} finally {
if (workbook != null) {
try {
workbook.close();
} catch (final IOException e) {
e.printStackTrace();
}
}
IOUtils.closeQuietly(fin);
IOUtils.closeQuietly(fos);
}
}

private String getCellValue(final Cell cell) {
String value = "";
if (cell == null) {
return value;
}
if(cell.getCellType() == Cell.CELL_TYPE_NUMERIC){
cell.setCellType(Cell.CELL_TYPE_STRING);
}
switch (cell.getCellType()) {
case Cell.CELL_TYPE_STRING:
if (cell.getStringCellValue().startsWith("( *")) {
value = "";
} else {
value = cell.getStringCellValue();
}
break;
case Cell.CELL_TYPE_NUMERIC:
value = String.valueOf(cell.getNumericCellValue());
break;
case Cell.CELL_TYPE_BOOLEAN:
value = String.valueOf(cell.getBooleanCellValue());
break;
case Cell.CELL_TYPE_FORMULA:
value = String.valueOf(cell.getCellFormula());
break;
case Cell.CELL_TYPE_BLANK:
value = "";
break;
default:
break;
}

return value;
}

private void processRow(final XSSFRow row) {
if (attrStarted) {
for (int i = 0; i < list.size(); i++) {
final XSSFCell cell = row.getCell(i);
tabNode.addAttribute(list.get(i), getCellValue(cell));
}
}
if (elemStarted) {
elemNode = elemsNode.addElement("element");
final String tabPosition = tabNode.attributeValue("position");
final String positionPrefix = tabPosition.substring(0, tabPosition.lastIndexOf(","));
elemNode.addAttribute("position", positionPrefix + "," + (row.getRowNum() + 1));
for (int i = 0; i < list.size(); i++) {
final XSSFCell cell = row.getCell(i);
subNode = elemNode.addElement(list.get(i));
if (!"".equals(getCellValue(cell))) {
subNode.setText(getCellValue(cell));
}
}
}
}

}

在上边的类中，要注意流的关闭，不要直接在try块里边关闭，应该在finally里边使用IOUtils的closeQuietly方法来关闭，这样就可以不用自己去判断流是否为null，也不用自己再去try-catch流的close方法。接着是一个测试类。

public class POITest {

public static void main(final String[] args) {
final ResourceBundle rb = ResourceBundle.getBundle("xml-generator");
final String src = rb.getString("gen.excel.path");
final String target = rb.getString("gen.target.path");
try {
new LewisPOI().parseExcel(src, target);
} catch (final Exception e) {
e.printStackTrace();
}
}

}

在POITest类，为了避免路径字符串的硬编码，我将路径存放到了一个名为xml-generator的properties文件中，该配置文件存放在项目的src目录下。该配置文件存放的是要读取的excel的路径，以及要生成的xml文件的路径，这里需要注意两个文件的目录是否正确，否则会无法正确的读取excel或者生成xml（可以自己先在代码中判断目录是否存在，若不存在则创建该目录）

gen.excel.path=D:/**/**.xlsx
gen.target.path=D:/**/**.xml