java IO流下载.docx文件时报文件损坏的错误

String fileId = request.getParameter("fileId");
            AttachFileVO attachFileVO = new AttachFileVO();
            attachFileVO.setFileId(fileId);       
            AttachFileVO attachFile = applicationService.getAttachFile(attachFileVO);
            String fileName = attachFile.getFileName();
            String formatFileName = CommonUtil.encodingFileName(fileName);  
            if(attachFile!=null&&attachFile.getFileContent()!=null){
                InputStream in = null;
                OutputStream os = null;
                try{               
                    response.reset();
                    response.setContentType(attachFile.getFileType());
                    response.setHeader("Content-Disposition", "attachment;filename="+formatFileName);
 
                    in = new ByteArrayInputStream(attachFile.getFileContent());
                    byte[] buffer = new byte[1024];
                    os = response.getOutputStream();
                    while (in.read(buffer) > 0) {
                        os.write(buffer);
                    }               
                }catch(Exception ex){
                    throw ex;
                }finally{
                    if(in!=null){
                        in.close();
                    }
                    if(os!=null){
                        os.close();
                    }                   
                }               
                return null;
            }

上面写法下载docx报错,更改:

 in = new ByteArrayInputStream(attachFile.getFileContent());
                byte[] buffer = new byte[1024];
                os = response.getOutputStream();
                int len = 0;
                while ((len = in.read(buffer)) > 0) {
                    os.write(buffer, 0, len);
                }  

  意思是将buffer里面0~len的字节写到输出流里面,下载的文件转换成字节流的字节数一般是不能被1024整除的,当输出最后一部分字节的时候,可能没有1024个字节,那么buffer字节数组里多出来的位置就是初始值0,这些也写到输出流里面去了,也就是多写了很多个值为0的字节进去,所以格式不对。

猜你喜欢

转载自paynexss.iteye.com/blog/2364051