问题描述
我有以下代码。当我尝试使用少量记录（大约 200 条记录）将记录列表导出到 excel 时，这很好用：
@Path("/toExcel")
@GET
public Response excelCustomersReport(@QueryParam("page") Integer page,@QueryParam("start") Integer start,@QueryParam("limit") Integer limit,@QueryParam("filter") FilterWrapper filter,@QueryParam("sort") SortWrapper sort) throws Exception {
response.setContentType("application/octet-stream");
DateFormat dateFormatter = new SimpleDateFormat("yyyy-MM-dd_HH:mm:ss");
String currentDateTime = dateFormatter.format(new Date());
String headerKey = "Content-disposition";
String headerValue = "attachment; filename=report_" + currentDateTime + ".xlsx";
response.setHeader(headerKey,headerValue);
hssfWorkbook workbook=userService.export(page,start,limit,filter,sort);
ServletoutputStream outputStream=response.getoutputStream();
workbook.write(outputStream);
workbook.close();
outputStream.close();
return ResponseHelper.ok(response);
}
当列表大小变大（例如超过 200 条记录）时，悲剧就开始了。它给出了以下错误：
<An I/O error has occurred while writing a response message entity to the container output stream.
org.glassfish.jersey.server.internal.process.MappableException: java.io.IOException: CLOSED
public Response excelCustomersReport(@QueryParam("page") Integer page,sort);
response.setHeader("Content-Length",String.valueOf(workbook.getBytes().length));
InputStream is=new ByteArrayInputStream(workbook.getBytes());
ServletoutputStream sos = response.getoutputStream();
long byteRead = 0;
try {
byte[] buf = new byte[1000];
while (true) {
int r = is.read(buf);
if (r == -1)
break;
sos.write(buf,r);
byteRead +=r;
if(byteRead > 1024*1024){ //flushes after 1mb
byteRead = 0;
sos.flush();
}
}
}catch (Exception e) {
e.printstacktrace();
}finally{
if(sos != null){
sos.flush();
}
try{is.close();}catch(Exception e){}
try{sos.close();}catch(Exception e){}
}
return ResponseHelper.ok(response);
}
冲洗流是大多数人指向的唯一方法。现在这个解决方案不仅适用于大数据，也适用于较小的数据（不适用于这两种情况）。它给出了这个错误：
An I/O error has occurred while writing a response message entity to the container output stream.
org.glassfish.jersey.server.internal.process.MappableException: com.fasterxml.jackson.databind.JsonMappingException: Direct self-reference leading to cycle
我尝试了 XSSFWorkbook 而不是 hssfWorkbook 。但 weblogic 甚至没有构建该项目并引发了与 Meta-model 相关的错误。
无论如何，这段代码有什么问题。如何将较大的文件写入流？我的 apache poi 版本是 3.11。
解决方法
正如 Paul Samsotha 评论的 S.O 链接所说，我应该使用 StreamingOutput 而不是 ServletOutputStream。所以该方法应该被重写为：
@Path("/toExcel")
@GET
public Response excelCustomersReport(@QueryParam("page") Integer page,@QueryParam("start") Integer start,@QueryParam("limit") Integer limit,@QueryParam("filter") FilterWrapper filter,@QueryParam("sort") SortWrapper sort) throws Exception {
response.setContentType("application/octet-stream");
DateFormat dateFormatter = new SimpleDateFormat("yyyy-MM-dd_HH:mm:ss");
String currentDateTime = dateFormatter.format(new Date());
String headerKey = "Content-Disposition";
String headerValue = "attachment; filename=report_" + currentDateTime + ".xlsx";
response.setHeader(headerKey,headerValue);
HSSFWorkbook workbook=userService.export(page,start,limit,filter,sort);
StreamingOutput output = new StreamingOutput() {
@Override
public void write(OutputStream out)
throws IOException,WebApplicationException {
workbook.write(out);
out.flush();
}
};
return Response.ok(output)
.build();
}