场景描述：

    在 Spring Boot 自定义拦截器中，需要对请求 body 中的内容做签名验证。在日志切面中，需要打印请求 body 中的内容。

报错内容：

    java.lang.IllegalStateException: getReader() has already been called for this request
        at org.apache.catalina.connector.Request.getInputStream(Request.java:1069)
        at org.apache.catalina.connector.RequestFacade.getInputStream(RequestFacade.java:365)
        at com.igg.aggregate.server.aspect.LogAspect.before(LogAspect.java:80)

原因分析：

    HttpServletRequest 的 getInputStream() 和 getReader() 都只能读取一次，由于 Request Body 是流的形式读取，那么

流读了一次就没有了，所以只能被调用一次。

解决办法：

    先将 Request Body 保存，然后通过 Servlet 自带的 HttpServletRequestWrapper 类覆盖 getReader() 和

getInputStream() 方法，使流从保存的body读取。然后再Filter中将ServletRequest替换为AuthenticationRequestWrapper。

代码示例：

    public class MyRequestWrapper extends HttpServletRequestWrapper {
        private byte[] body;
     
        public MyRequestWrapper(HttpServletRequest request) throws IOException {
            super(request);
     
            StringBuilder sb = new StringBuilder();
            String line;
            BufferedReader reader = request.getReader();
            while ((line = reader.readLine()) != null) {
                sb.append(line);
            }
            String body = sb.toString();
            this.body = body.getBytes(StandardCharsets.UTF_8);
        }
     
     
        public String getBody() {
            return new String(body, StandardCharsets.UTF_8);
        }
    }

    // 获取请求body
    try {
         MyRequestWrapper myRequestWrapper = new MyRequestWrapper(request);
         return myRequestWrapper.getBody();
    } catch (IOException e) {
         log.error("get request body exception", e);
         throw new RuntimeException(e);
    }

 

解决：getReader() has already been called for this request

2020-12-28阅读 2.7K0

在 Filter 中对 request 中的 body 进行参数签名校验, 会报如下错误:

getReader() has already been called for this request

原因是 request.getReader() 和 request.getInputStream() 都是只能调用一次

并且 getReader() 方法底层也是调用 getInputStream() 来实现的.

所以我们要使用 HttpServletRequestWrapper 来实现自定义的 CustomHttpServletRequestWrapper, 把 body 保存在 CustomHttpServletRequestWrapper 中, 并且重写 getInputStream() 方法

public class CustomHttpServletRequestWrapper extends HttpServletRequestWrapper{

	private byte[] body;

	public CustomHttpServletRequestWrapper(HttpServletRequest request) throws IOException {
		super(request);
		
		BufferedReader reader = request.getReader();
		try (StringWriter writer = new StringWriter()) {
			int read;
            char[] buf = new char[1024 * 8];
    		while ((read = reader.read(buf)) != -1) {
    			writer.write(buf, 0, read);
    		}
    		this.body = writer.getBuffer().toString().getBytes();
        }
	}
	
	public byte[] getBody() {
		return body;
	}
	
	@Override
	public ServletInputStream getInputStream() throws IOException {
		ByteArrayInputStream byteArrayInputStream = new ByteArrayInputStream(body);
		return new ServletInputStream() {
			
			@Override
			public int read() throws IOException {
				return byteArrayInputStream.read();
			}
			
			@Override
			public void setReadListener(ReadListener listener) {
			}
			
			@Override
			public boolean isReady() {
				return false;
			}
			
			@Override
			public boolean isFinished() {
				return false;
			}
		};
	}
}

然后, 获取 body 就调用 CustomHttpServletRequestWrapper 中的 getBody() 方法, 后面就使用

filterChain.doFilter(customHttpServletRequestWrapper, response);

就 OK 了