解决POI读取Excel百万级内存溢出问题

使用传统poi来操作大数据量的excel会出现内存溢出的问题，根据各种资源，亲试了一个可用工具类，附代码如下：

一、基于eventusermodel的excel解析工具类

package com.taikang.task.service.excel;

import org.apache.poi.openxml4j.exceptions.InvalidFormatException;
import org.apache.poi.openxml4j.exceptions.OpenXML4JException;
import org.apache.poi.openxml4j.opc.OPCPackage;
import org.apache.poi.xssf.eventusermodel.ReadOnlySharedStringsTable;
import org.apache.poi.xssf.eventusermodel.XSSFReader;
import org.apache.poi.xssf.eventusermodel.XSSFSheetXMLHandler;
import org.apache.poi.xssf.eventusermodel.XSSFSheetXMLHandler.SheetContentsHandler;
import org.apache.poi.xssf.model.StylesTable;
import org.apache.poi.xssf.usermodel.XSSFComment;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.stereotype.Component;
import org.xml.sax.InputSource;
import org.xml.sax.SAXException;
import org.xml.sax.XMLReader;
import org.xml.sax.helpers.XMLReaderFactory;

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

/**
* 解析大数据量Excel工具类
* @author
*
*/
@Component
public class ExcelParser {
private static final Logger logger = LoggerFactory.getLogger(ExcelParser.class);
/**
* 表格默认处理器
*/
private ISheetContentHandler contentHandler = new DefaultSheetHandler();
/**
* 读取数据
*/
private List<String[]> datas = new ArrayList<String[]>();

/**
* 转换表格,默认为转换第一个表格
* @param stream
* @return
* @throws InvalidFormatException
* @throws IOException
* @throws ParseException
*/
public ExcelParser parse(InputStream stream)
throws InvalidFormatException, IOException, ParseException {
return parse(stream, 1);
}

/**
*
* @param stream
* @param sheetId:为要遍历的sheet索引，从1开始
* @return
* @throws InvalidFormatException
* @throws IOException
* @throws ParseException
*/
public synchronized ExcelParser parse(InputStream stream, int sheetId)
throws InvalidFormatException, IOException, ParseException {
// 每次转换前都清空数据
datas.clear();
// 打开表格文件输入流
OPCPackage pkg = OPCPackage.open(stream);
try {
// 创建表阅读器
XSSFReader reader;
try {
reader = new XSSFReader(pkg);
} catch (OpenXML4JException e) {
logger.error("读取表格出错");
throw new ParseException(e.fillInStackTrace());
}

// 转换指定单元表
InputStream shellStream = reader.getSheet("rId" + sheetId);
try {
InputSource sheetSource = new InputSource(shellStream);
StylesTable styles = reader.getStylesTable();
ReadOnlySharedStringsTable strings = new ReadOnlySharedStringsTable(pkg);
getContentHandler().init(datas);// 设置读取出的数据
// 获取转换器
XMLReader parser = getSheetParser(styles, strings);
parser.parse(sheetSource);
} catch (SAXException e) {
logger.error("读取表格出错");
throw new ParseException(e.fillInStackTrace());
} finally {
shellStream.close();
}
} finally {
pkg.close();

}
return this;

}

/**
* 获取表格读取数据,获取数据前，需要先转换数据<br>
* 此方法不会获取第一行数据
*
* @return 表格读取数据
*/
public List<String[]> getDatas() {
return getDatas(true);

}

/**
* 获取表格读取数据,获取数据前，需要先转换数据
*
* @param dropFirstRow
* 删除第一行表头记录
* @return 表格读取数据
*/
public List<String[]> getDatas(boolean dropFirstRow) {
if (dropFirstRow && datas.size() > 0) {
datas.remove(0);// 删除表头
}
return datas;

}

/**
* 获取读取表格的转换器
*
* @return 读取表格的转换器
* @throws SAXException
* SAX错误
*/
protected XMLReader getSheetParser(StylesTable styles, ReadOnlySharedStringsTable strings) throws SAXException {
XMLReader parser = XMLReaderFactory.createXMLReader();
parser.setContentHandler(new XSSFSheetXMLHandler(styles, strings, getContentHandler(), false));
return parser;
}

public ISheetContentHandler getContentHandler() {
return contentHandler;
}

public void setContentHandler(ISheetContentHandler contentHandler) {
this.contentHandler = contentHandler;
}

/**
* 表格转换错误
*/
public class ParseException extends Exception {
private static final long serialVersionUID = -2451526411018517607L;

public ParseException(Throwable t) {
super("表格转换错误", t);
}

}

public interface ISheetContentHandler extends SheetContentsHandler {

/**
* 设置转换后的数据集，用于存放转换结果
*
* @param datas
* 转换结果
*/
void init(List<String[]> datas);
}

/**
* 默认表格解析handder
*/
class DefaultSheetHandler implements ISheetContentHandler {
/**
* 读取数据
*/
private List<String[]> datas;
private int columsLength;
// 读取行信息
private String[] readRow;
private ArrayList<String> fristRow = new ArrayList<String>();

@Override
public void init(List<String[]> datas) {
this.datas = datas;
// this.columsLength = columsLength;
}

@Override
public void startRow(int rowNum) {
if (rowNum != 0) {
readRow = new String[columsLength];
}
}

@Override
public void endRow(int rowNum) {
//将Excel第一行表头的列数当做数组的长度，要保证后续的行的列数不能超过这个长度，这是个约定。
if (rowNum == 0) {
columsLength = fristRow.size();
readRow = fristRow.toArray(new String[fristRow.size()]);
}else {
readRow = fristRow.toArray(new String[columsLength]);
}
datas.add(readRow.clone());
readRow = null;
fristRow.clear();
}

@Override
public void cell(String cellReference, String formattedValue, XSSFComment comment) {
int index = getCellIndex(cellReference);//转换A1,B1,C1等表格位置为真实索引位置
try {
fristRow.set(index, formattedValue);
} catch (IndexOutOfBoundsException e) {
int size = fristRow.size();
for (int i = index - size+1;i>0;i--){
fristRow.add(null);
}
fristRow.set(index,formattedValue);
}
}

@Override
public void headerFooter(String text, boolean isHeader, String tagName) {
}

/**
* 转换表格引用为列编号
*
* @param cellReference
* 列引用
* @return 表格列位置，从0开始算
*/
public int getCellIndex(String cellReference) {
String ref = cellReference.replaceAll("\\d+", "");
int num = 0;
int result = 0;
for (int i = 0; i < ref.length(); i++) {
char ch = cellReference.charAt(ref.length() - i - 1);
num = (int) (ch - 'A' + 1);
num *= Math.pow(26, i);
result += num;
}
return result - 1;
}
}

以下为读取excel测试main方法

public static void main(String[] args) throws IOException, ParseException, InvalidFormatException {
File file = new File("E:\\tmp\\weibaozj.xlsx");
FileInputStream inputStream = new FileInputStream(file);
ExcelParser excelParser = new ExcelParser();
ExcelParser parse = excelParser.parse(inputStream);
List<String[]> datas = parse.getDatas();
String[] array5 = datas.get(0);
for (int i=0;i<array5.length;i++){

System.out.println("下标："+i+"--对应value :"+array5[i]);
}
//System.out.println(datas);
}
}

二、实现项目中根据excel的数据批量写入数据库中

/**
* 请求数据导入功能
*
* @return
*/
@Transactional
public ReturnDTO<String> importBatchRequest() {
log.info("导入轻松筹入参数据开始");
int rowCount = 0;
String userName = "";
try {
//读取服务器路径
//String filePath = "D:\\MyDocuments\\itw_denghj\\工作簿1.xlsx";
String filePath = dataPropertiesUtil.getBatchFilePath();
File file = new File(filePath);
log.info("需要读取的文件fileName:{}", file.getName());
FileInputStream inputStream = new FileInputStream(file);
log.info("------解析excel文件开始-----------");
ExcelParser parse = excelParser.parse(inputStream);
List<String[]> dataList = parse.getDatas();
log.info("------解析excel文件结束，获取数据size:{}-----------",dataList.size());
List<HdBatchInsuraRequest> batchInsuraRequestList = new ArrayList<>();
/* InputStream inputStream = new FileInputStream(file);
Workbook workBook = ExcelUtils.getWorkBook(inputStream, file.getName());
List<List<Object>> sheetContentList = ExcelUtils.getCurrentSheetContentList(workBook.getSheetAt(0), 13);*/
for (int i = 0; i < dataList.size(); i++) {//遍历每一行
rowCount = i;
String[] rowList = dataList.get(i);
if (rowList.length !=12){
log.info("微保重疾rowlist内容长度不符合要求:{}",JSONObject.toJSONString(rowList));
continue;
}
HdBatchInsuraRequest hdBatchInsuraRequest = new HdBatchInsuraRequest();
//姓名
userName = rowList[0];
//身份证唯一标识
String uniqueIdentify = rowList[1];
//身份证号
String userId = rowList[2];
//投保时间
String proposalTime = rowList[3];
//保单号
String policyNo = rowList[4];
//保额
String sumInsured = rowList[5];
//保单有效期
String policyValid = rowList[6];
//是否发生理赔
String isClaim = rowList[7];
//理赔时间
String claimTime = rowList[8];
//是否购买医疗险
String buyMedicalInsurance = rowList[9];
//首次购买医疗险时间
String firstBuyTime = rowList[10];
//地址
String location = rowList[11];
hdBatchInsuraRequest.setUserId(AesUtil.encrypt(userId, dataPropertiesUtil.getCidAesKey()));
hdBatchInsuraRequest.setUniqueIdentify(uniqueIdentify);
hdBatchInsuraRequest.setBuyMedicalInsurance(buyMedicalInsurance);
hdBatchInsuraRequest.setCallStatus(0);
hdBatchInsuraRequest.setClaimTime(claimTime);
hdBatchInsuraRequest.setCreateTime(new Date());
hdBatchInsuraRequest.setFirstBuyTime(firstBuyTime);
hdBatchInsuraRequest.setSumInsured(sumInsured);
hdBatchInsuraRequest.setIsClaim(isClaim);
hdBatchInsuraRequest.setLocation(location);
hdBatchInsuraRequest.setBatchCode("weibao");
hdBatchInsuraRequest.setPolicyNo(policyNo);
hdBatchInsuraRequest.setPolicyValid(Integer.valueOf(policyValid));
hdBatchInsuraRequest.setUserName(userName);
hdBatchInsuraRequest.setProposalTime(proposalTime);
batchInsuraRequestList.add(hdBatchInsuraRequest);
}
log.info("------转换数据结束，数据batchInsuraRequestList.size:{}-----------",batchInsuraRequestList.size());
List<List<HdBatchInsuraRequest>> lists = subList(batchInsuraRequestList, 5000);
log.info("---------分批保存数据的集合大小为：{}",lists.size());
for (List<HdBatchInsuraRequest> requestList : lists){
int insertInsuraRequestList = hdBatchInsuraRequestExtMapper.insertInsuraRequestList(requestList);
}
log.info("导入微保入参数据成功");
} catch (Exception e) {
log.info("批量文件导入DB异常,异常行数：{},用户名：{}", rowCount, userName);
TransactionAspectSupport.currentTransactionStatus().setRollbackOnly();
e.printStackTrace();
return ReturnDTO.FAIL();
}
return ReturnDTO.SUCCESS();
}

三、List集合拆分通用工具方法

/**
* 拆分集合
* @param tList
* @param subNum
* @return
*/
public static List<List<HdBatchInsuraRequest>> subList(List<HdBatchInsuraRequest> tList, Integer subNum) {
// 新的截取到的list集合
List<List<HdBatchInsuraRequest>> tNewList = new ArrayList<List<HdBatchInsuraRequest>>();
// 要截取的下标上限
Integer priIndex = 0;
// 要截取的下标下限
Integer lastIndex = 0;
// 每次插入list的数量
// Integer subNum = 500;
// 查询出来list的总数目
Integer totalNum = tList.size();
// 总共需要插入的次数
Integer insertTimes = totalNum / subNum;
List<HdBatchInsuraRequest> subNewList = new ArrayList<HdBatchInsuraRequest>();
for (int i = 0; i <= insertTimes; i++) {
// [0--20) [20 --40) [40---60) [60---80) [80---100)
priIndex = subNum * i;
lastIndex = priIndex + subNum;
// 判断是否是最后一次
if (i == insertTimes) {
log.info("最后一次截取："+priIndex + "," + lastIndex);
subNewList = tList.subList(priIndex, tList.size());
} else {
// 非最后一次
subNewList = tList.subList(priIndex, lastIndex);

}
if (subNewList.size() > 0) {
//logger.info("开始将截取的list放入新的list中");
tNewList.add(subNewList);
}

}

return tNewList;

}