javaEE杂项 --下载文件乱码的不同解决方案

在下载文件时,常见的方法:

 public void getuseopdoc(HttpServletResponse response) throws Exception {
        try {
            String filename = "某某文件";
            InputStream inputStream = getClass().getResourceAsStream("/doc/"+filename+".doc");
            response.reset();
            response.setCharacterEncoding("utf-8");
            response.setContentType("application/x-msdownload");
            response.setHeader("Content-Disposition","attachment; filename="+filename+".doc");
            IOUtils.copy(inputStream, response.getOutputStream());
            response.flushBuffer();
            inputStream.close();
        } catch (Exception e){
            log.debug("Request could not be completed at this moment. Please try again.");
            e.printStackTrace();
        }

    }

• 客户端得到的几乎都会是乱码, 因为filename是UTF8编码,它被设置为attachment属性值,然后通过HTTP传输到请求端浏览器,但是,HTTP的传输编码是ISO 8859-1(java的网络传输标准编码). 用ISO8859-1传输utf8字符串肯定不兼容导致乱码或者某些错误字符.
• 所以这时候我们想到把utf8字符串转换成iso8859-1的编码方法,传输到客户端后再用UTF8解码:

 public void getuseopdoc(HttpServletResponse response) throws Exception {
        try {
            String filename = "重庆水库大坝安全监测信息管理平台操作手册（业务版）";
            InputStream inputStream = getClass().getResourceAsStream("/doc/"+filename+".doc");
            response.reset();
            response.setCharacterEncoding("utf-8");
            response.setContentType("application/x-msdownload");   
            response.setHeader("Content-Disposition","attachment; filename="+new String(filename.getbytes("utf-8"),"iso8859-1")+".doc");
            IOUtils.copy(inputStream, response.getOutputStream());
            response.flushBuffer();
            inputStream.close();
        } catch (Exception e){
            log.debug("Request could not be completed at this moment. Please try again.");
            e.printStackTrace();
        }

    }

字符编码的更多信息可以查看http://blog.csdn.net/robertcpp/article/details/7837712.

utf8是Unicode编码的一种优化版,因为unicode对所有字符都编码为2个字节,这导致了标准unicode编码不和任何其他编码兼容. utf-8在Unicode基础上做了一定的修改,对属于ascii的字符依然采用1个字节编码,而其他的字符则采用更多的字节. 这也被称为变长字节编码. utf8这样处理就使得它与iso8859-1兼容了. 然后说说gb2312/gbk. 二者差别在于后者可以表示繁体字,可以看作前者超集. gb2312是中文编码的国标,对所有中文采用2字节编码,相对utf-8来说,他可能还更加的节省空间.

笔者在测试上面代码的时候,IE下文件名依然乱码, 原因未知. 不过我在把转换的编码从utf-8改为gb2312后就好了.

• 在求助google后,找到下面这段代码(依赖了spring的包):

  
      protected String getFileName(String filename) {
          try{
              HttpServletRequest request = ((ServletRequestAttributes) RequestContextHolder.getRequestAttributes()).getRequest();
              if (request.getHeader("User-Agent").indexOf("MSIE") != -1) {
                  return ‘\"‘ + java.net.URLEncoder.encode(filename, "UTF-8") + ‘\"‘;
              }
              byte[] bytes = filename.getBytes("UTF-8");
              StringBuilder buff = new StringBuilder(bytes.length << 2);
              buff.append("=?UTF-8?Q?");
              for (byte b : bytes) {
                  int unsignedByte = b & 0xFF;
                  buff.append(‘=‘).append(HEX_CHARS[unsignedByte >> 4]).append(HEX_CHARS[unsignedByte & 0xF]);
              }
              return buff.append("?=").toString();
          }catch(UnsupportedEncodingException e){
              return filename;
          }
      }
  
  

  这段代码对文件名并区分ie和其他浏览器,然后设置不同的编码格式.

  测试在不同的环境下都可以用

null