在使用POI进行excel操作时,当数据量较大时经常会产生内存溢出异常。下面我们通过分析如何解决该问题
一、POI结构图
二、内存溢出问题
在项目中遇到二十万行数据要写入到excel中时会内存溢出,一般方法是调大tomcat的内存,但是调到2048M还是会内存溢出报错。因此我们分析其原因。我们通过分析其源码,得出其实现步骤为通过InputStream一行行读取到TreeMap类型的HSSFRow结构体中,因此当数据量大时就会造成内存溢出。
public HSSFWorkbook(DirectoryNode directory, boolean preserveNodes) throws IOException { super(directory); String workbookName = getWorkbookDirEntryName(directory); this.preserveNodes = preserveNodes; // If we're not preserving nodes, don't track the // POIFS any more if(! preserveNodes) { clearDirectory(); } _sheets = new ArrayList<HSSFSheet>(INITIAL_CAPACITY); names = new ArrayList<HSSFName>(INITIAL_CAPACITY); // Grab the data from the workbook stream, however // it happens to be spelled. InputStream stream = directory.createDocumentInputStream(workbookName); List<Record> records = RecordFactory.createRecords(stream); workbook = InternalWorkbook.createWorkbook(records); setPropertiesFromWorkbook(workbook); int recOffset = workbook.getNumRecords(); // convert all LabelRecord records to LabelSSTRecord convertLabelRecords(records, recOffset); RecordStream rs = new RecordStream(records, recOffset); while (rs.hasNext()) { try { InternalSheet sheet = InternalSheet.createSheet(rs); _sheets.add(new HSSFSheet(this, sheet)); } catch (UnsupportedBOFType eb) { // Hopefully there's a supported one after this! log.log(POILogger.WARN, "Unsupported BOF found of type " + eb.getType()); } } for (int i = 0 ; i < workbook.getNumNames() ; ++i){ NameRecord nameRecord = workbook.getNameRecord(i); HSSFName name = new HSSFName(this, nameRecord, workbook.getNameCommentRecord(nameRecord)); names.add(name); } }
/** * add a row to the sheet * * @param addLow whether to add the row to the low level model - false if its already there */ private void addRow(HSSFRow row, boolean addLow) { _rows.put(Integer.valueOf(row.getRowNum()), row); if (addLow) { _sheet.addRow(row.getRowRecord()); } boolean firstRow = _rows.size() == 1; if (row.getRowNum() > getLastRowNum() || firstRow) { _lastrow = row.getRowNum(); } if (row.getRowNum() < getFirstRowNum() || firstRow) { _firstrow = row.getRowNum(); } }
excel数据行读取到内存的存储结构如下:
三、解决方案
poi官网给了一种大批量数据写入的方法,使用SXXFWorkbook类进行大批量写入操作解决了这个问题,可以监控该样例,我们会发现整体内存呈现锯齿状,能够及时回收,内存相对比较平稳。
package org.bird.poi; import java.io.File; import java.io.FileOutputStream; import java.io.IOException; import java.net.URL; import org.apache.poi.ss.usermodel.Cell; import org.apache.poi.ss.usermodel.Row; import org.apache.poi.ss.usermodel.Sheet; import org.apache.poi.ss.util.CellReference; import org.apache.poi.xssf.streaming.SXSSFWorkbook; import org.junit.Assert; public class XSSFWriter { private static SXSSFWorkbook wb; public static void main(String[] args) throws IOException { wb = new SXSSFWorkbook(10000); Sheet sh = wb.createSheet(); for(int rownum = 0; rownum < 100000; rownum++){ Row row = sh.createRow(rownum); for(int cellnum = 0; cellnum < 10; cellnum++){ Cell cell = row.createCell(cellnum); String address = new CellReference(cell).formatAsString(); cell.setCellValue(address); } } // Rows with rownum < 900 are flushed and not accessible for(int rownum = 0; rownum < 90000; rownum++){ Assert.assertNull(sh.getRow(rownum)); } // ther last 100 rows are still in memory for(int rownum = 90000; rownum < 100000; rownum++){ Assert.assertNotNull(sh.getRow(rownum)); } URL url = XSSFWriter.class.getClassLoader().getResource(""); FileOutputStream out = new FileOutputStream(url.getPath() + File.separator + "wirter.xlsx"); wb.write(out); out.close(); // dispose of temporary files backing this workbook on disk wb.dispose(); } }