SpringBoot实现文件下载的几种方式比较

目前，网上一搜SpringBoot环境下载文件。有多种实现方式，大概率出来的会是基础版的，基础版的有几个坑，我这里分别将基础版以及基础版会出现的问题，从而引申各种解决方法。

基础版

话不多说，直接上源码：

@RestController
public class TestDownload {
    @Value("${gen.base.path}")
    private String baseFilePath;
    @ApiOperation(value = "SpringBoot实现文件下载", notes = "基础版")
    @RequestMapping(value = "/baseDownload", method = RequestMethod.GET)
    @ResponseBody
   public ResponseEntity<InputStreamResource> baseDownload(@RequestParam("fileId") String fileId) throws Exception {
        FileSystemResource file = new FileSystemResource(baseFilePath+fileId+".pdf");
        if (!file.exists()) {
        throw new HaBizException(1,"请您输入正确的文件ID");
    }
        HttpHeaders headers = new HttpHeaders();
        headers.add("Cache-Control", "no-cache, no-store, must-revalidate");
        headers.add("Content-Disposition", String.format("attachment; filename=%s", file.getFilename()));
        headers.add("Pragma", "no-cache");
        headers.add("Expires", "0");
        return ResponseEntity
                .ok()
                .headers(headers)
                .contentLength(file.contentLength())
                .contentType(MediaType.parseMediaType("application/octet-stream"))
                .body(new InputStreamResource(file.getInputStream()));
    }

}

这个基础版是有两个问题：1、不能手动（主动）关闭文件流
2、下载文件较大时，jvm会内存溢出。

然后这里先引申出进阶版：主动关闭文件流。

@ApiOperation(value = "SpringBoot实现文件下载", notes = "进阶版")
@RequestMapping(value = "/advancedDownload", method = RequestMethod.GET)
@ResponseBody
public ResponseEntity<StreamingResponseBody> advancedDownload(@RequestParam("fileId") String fileId) throws Exception {
    FileSystemResource file = new FileSystemResource(baseFilePath+fileId);
    if (!file.exists()) {
        throw  new HaBizException(1,"请您输入正确的文件ID");
    }
    InputStream inputStream = file.getInputStream();
    StreamingResponseBody responseBody = outputStream -> {
        int numberOfBytesToWrite;
        byte[] data = new byte[1024];
        while ((numberOfBytesToWrite = inputStream.read(data, 0, data.length)) != -1) {
            outputStream.write(data, 0, numberOfBytesToWrite);
        }
        inputStream.close();
    };
    HttpHeaders headers = new HttpHeaders();
    headers.add("Cache-Control", "no-cache, no-store, must-revalidate");
    headers.add("Content-Disposition", String.format("attachment; filename=%s", file.getFilename()));
    headers.add("Pragma", "no-cache");
    headers.add("Expires", "0");
    return ResponseEntity.ok()
            .headers(headers)
            .contentType(MediaType.APPLICATION_OCTET_STREAM)
            .body(responseBody);
}

这个版本普遍适用，但是仍然不能解决下载文件较大导致内存溢出的问题，所以这里就来个功能更全一点的，使用缓存流，边读边写

使用缓存流，边读边写

@ApiOperation(value = "SpringBoot实现文件下载", notes = "使用缓存流，边读边写")
@RequestMapping(value = "/cacheDownload", method = RequestMethod.GET)
@ResponseBody
public void cacheDownload(@RequestParam("fileId") String fileId, HttpServletResponse response) throws Exception {
    FileSystemResource file = new FileSystemResource(baseFilePath+fileId);
    if (!file.exists()) {
        throw  new HaBizException(1,"请您输入正确的文件ID");
    }
    String filename = file.getFilename();
    InputStream inputStream = null;
    BufferedInputStream bufferedInputStream = null;
    BufferedOutputStream bufferedOutputStream = null;
    response.setHeader("Content-Disposition", "attachment;filename=" + URLEncoder.encode(filename, "UTF-8"));
    try {
        inputStream = file.getInputStream();
        bufferedInputStream = new BufferedInputStream(inputStream);
        bufferedOutputStream = new BufferedOutputStream(response.getOutputStream());
        FileCopyUtils.copy(bufferedInputStream, bufferedOutputStream);
        }catch(Exception e){
            throw  new HaBizException(1,"下载文件异常"+e);
        }finally {
            if(null!=inputStream){
                inputStream.close();
            }
            if(null!=bufferedInputStream){
                bufferedInputStream.close();
            }
            if(null!=bufferedOutputStream){
            	bufferedOutputStream.flush();
                bufferedOutputStream.close();
            }
    }
}

当然，如果不想使用java自带的BufferedInputStream当缓冲区的话，也可以自己写，例子如下

@RequestMapping("/baseCacheDownload")
public void baseCacheDownload(@RequestParam("fileId") String fileId, HttpServletResponse response) throws Exception {
    FileSystemResource file = new FileSystemResource(baseFilePath+fileId);
    if (!file.exists()) {
        throw  new HaBizException(1,"请您输入正确的文件ID");
    }
    InputStream inputStream = file.getInputStream();// 文件的存放路径
    response.reset();
    response.setContentType("application/octet-stream");
    String filename = file.getFilename();
    response.addHeader("Content-Disposition", "attachment; filename=" + URLEncoder.encode(filename, "UTF-8"));//①tips
    ServletOutputStream outputStream = response.getOutputStream();
    byte[] b = new byte[1024];
    int len;
    //缓冲区（自己可以设置大小）：从输入流中读取一定数量的字节，并将其存储在缓冲区字节数组中，读到末尾返回-1
    while ((len = inputStream.read(b)) > 0) {
        outputStream.write(b, 0, len);
    }
    inputStream.close();
}

这里有一个技巧，①tips处的代码，Content-disposition 为属性名。 attachment 表示以附件方式下载。如果要在页面中打开，则改为 inline。 filename 如果为中文，则会出现乱码。解决办法有两种：1、 使用 fileName = new String(fileName.getBytes(), “ISO8859-1”) 语句 2、使用 fileName = HttpUtility.UrlEncode(filename, System.Text.Encoding.UTF8) 语句