OkHttp如何记录请求体

Question

我正在使用一个拦截器,我想记录我正在制作的请求的正文,但我看不到这样做的任何方法.

可能吗 ？

public class LoggingInterceptor implements Interceptor {
    @Override
    public Response intercept(Chain chain) throws IOException {
        Request request = chain.request();

        long t1 = System.nanoTime();
        Response response = chain.proceed(request);
        long t2 = System.nanoTime();

        double time = (t2 - t1) / 1e6d;

        if (request.method().equals("GET")) {
            Logs.info(String.format("GET " + F_REQUEST_WITHOUT_BODY + F_RESPONSE_WITH_BODY, request.url(), time, request.headers(), response.code(), response.headers(), response.body().charStream()));
        } else if (request.method().equals("POST")) {
            Logs.info(String.format("POST " + F_REQUEST_WITH_BODY + F_RESPONSE_WITH_BODY, request.url(), time, request.headers(), request.body(), response.code(), response.headers(), response.body().charStream()));
        } else if (request.method().equals("PUT")) {
            Logs.info(String.format("PUT " + F_REQUEST_WITH_BODY + F_RESPONSE_WITH_BODY, request.url(), time, request.headers(), request.body().toString(), response.code(), response.headers(), response.body().charStream()));
        } else if (request.method().equals("DELETE")) {
            Logs.info(String.format("DELETE " + F_REQUEST_WITHOUT_BODY + F_RESPONSE_WITHOUT_BODY, request.url(), time, request.headers(), response.code(), response.headers()));
        }

        return response;
    }
}

结果:

POST  [some url] in 88,7ms
    ZoneName: touraine
    Source: Android
    body: retrofit.client.OkClient$1@1df53f05 <-request.body().toString() gives me this, but I would like the content string
    Response: 500
    Date: Tue, 24 Feb 2015 10:14:22 GMT
    body: [some content] 

Answer 1

尼古拉的回答对我不起作用.我的猜测是ByteString#toString()改变的实施.这个解决方案对我有用:

private static String bodyToString(final Request request){

    try {
        final Request copy = request.newBuilder().build();
        final Buffer buffer = new Buffer();
        copy.body().writeTo(buffer);
        return buffer.readUtf8();
    } catch (final IOException e) {
        return "did not work";
    }
}

来自以下文件readUtf8():

从中删除所有字节,将它们解码为UTF-8,并返回字符串.

这应该是你想要的.

Answer 2

我试图评论@msung的正确答案,但我的声誉不够高.

这是我在完成请求之前打印RequestBody所做的修改.它就像一个魅力.谢谢

private static String bodyToString(final RequestBody request){
        try {
            final RequestBody copy = request;
            final Buffer buffer = new Buffer();
            copy.writeTo(buffer);
            return buffer.readUtf8();
        } 
        catch (final IOException e) {
            return "did not work";
        }
}

Answer 3

编辑

因为我看到仍有一些人对这篇文章感兴趣,这里是我的日志拦截器的最终版本(直到下一次改进).我希望它可以节省你们一些人的时间.

请注意,此代码正在使用OkHttp 2.2.0(和Retrofit 1.9.0)

import com.squareup.okhttp.*;
import okio.Buffer;
import java.io.IOException;

public class LoggingInterceptor implements Interceptor {

private static final String F_BREAK = " %n";
private static final String F_URL = " %s";
private static final String F_TIME = " in %.1fms";
private static final String F_HEADERS = "%s";
private static final String F_RESPONSE = F_BREAK + "Response: %d";
private static final String F_BODY = "body: %s";

private static final String F_BREAKER = F_BREAK + "-------------------------------------------" + F_BREAK;
private static final String F_REQUEST_WITHOUT_BODY = F_URL + F_TIME + F_BREAK + F_HEADERS;
private static final String F_RESPONSE_WITHOUT_BODY = F_RESPONSE + F_BREAK + F_HEADERS + F_BREAKER;
private static final String F_REQUEST_WITH_BODY = F_URL + F_TIME + F_BREAK + F_HEADERS + F_BODY + F_BREAK;
private static final String F_RESPONSE_WITH_BODY = F_RESPONSE + F_BREAK + F_HEADERS + F_BODY + F_BREAK + F_BREAKER;

@Override
public Response intercept(Chain chain) throws IOException {
    Request request = chain.request();

    long t1 = System.nanoTime();
    Response response = chain.proceed(request);
    long t2 = System.nanoTime();

    MediaType contentType = null;
    String bodyString = null;
    if (response.body() != null) {
        contentType = response.body().contentType();
        bodyString = response.body().string();
    }

    double time = (t2 - t1) / 1e6d;

    if (request.method().equals("GET")) {
        System.out.println(String.format("GET " + F_REQUEST_WITHOUT_BODY + F_RESPONSE_WITH_BODY, request.url(), time, request.headers(), response.code(), response.headers(), stringifyResponseBody(bodyString)));
    } else if (request.method().equals("POST")) {
        System.out.println(String.format("POST " + F_REQUEST_WITH_BODY + F_RESPONSE_WITH_BODY, request.url(), time, request.headers(), stringifyRequestBody(request), response.code(), response.headers(), stringifyResponseBody(bodyString)));
    } else if (request.method().equals("PUT")) {
        System.out.println(String.format("PUT " + F_REQUEST_WITH_BODY + F_RESPONSE_WITH_BODY, request.url(), time, request.headers(), request.body().toString(), response.code(), response.headers(), stringifyResponseBody(bodyString)));
    } else if (request.method().equals("DELETE")) {
        System.out.println(String.format("DELETE " + F_REQUEST_WITHOUT_BODY + F_RESPONSE_WITHOUT_BODY, request.url(), time, request.headers(), response.code(), response.headers()));
    }

    if (response.body() != null) {
        ResponseBody body = ResponseBody.create(contentType, bodyString);
        return response.newBuilder().body(body).build();
    } else {
        return response;
    }
}


private static String stringifyRequestBody(Request request) {
    try {
        final Request copy = request.newBuilder().build();
        final Buffer buffer = new Buffer();
        copy.body().writeTo(buffer);
        return buffer.readUtf8();
    } catch (final IOException e) {
        return "did not work";
    }
}

public String stringifyResponseBody(String responseBody) {
    return responseBody;
}
}

Answer 4

处理带或不带正文的请求的版本：

private String stringifyRequestBody(Request request) {
  if (request.body() != null) {
      try {
          final Request copy = request.newBuilder().build();
          final Buffer buffer = new Buffer();
          copy.body().writeTo(buffer);
          return buffer.readUtf8();
      } catch (final IOException e) {
          Log.w(TAG, "Failed to stringify request body: " + e.getMessage());
      }
  }
  return "";
}

Answer 5

使用当前版本的 OkHttp，您可以使用HTTP 日志拦截器并将级别设置为BODY

HttpLoggingInterceptor logging = new HttpLoggingInterceptor();
logging.setLevel(Level.BODY);

有了这个，你不能为不同的 HTTP 方法细粒度地配置输出，但它也适用于可能有主体的其他方法。

这是一个显示PATCH请求输出的示例（最少编辑）：

--> PATCH https://hostname/api/something/123456 HTTP/1.1
Content-Type: application/json-patch+json; charset=utf-8
Content-Length: 49
Authorization: Basic YWRtaW46c2VjcmV0Cg==
Accept: application/json

[ { "op": "add", "path": "/path", "value": true }]
--> END PATCH (xx-byte body)

正如您所看到的，这也会打印出标题，并且正如文档所述，您应该非常小心：

此拦截器在使用HEADERS或BODY级别时生成的日志有可能泄漏敏感信息，例如“授权”或“Cookie”标头以及请求和响应正文的内容。此数据应仅以受控方式或在非生产环境中记录。

您可以通过调用 来编辑可能包含敏感信息的标头redactHeader()。

logging.redactHeader("Authorization");
logging.redactHeader("Cookie");