Java OkHttp：避免泄漏连接警告
声明：本页面是StackOverFlow热门问题的中英对照翻译，遵循CC BY-SA 4.0协议，如果您需要使用它，必须同样遵循CC BY-SA许可，注明原文地址和作者信息，同时你必须将它归于原作者（不是我）：StackOverFlow
原文地址: http://stackoverflow.com/questions/38380371/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me):
StackOverFlow
OkHttp: avoid leaked connection warning
提问by Alphaaa
I am using OkHttp 3, and I keep getting leaked connection warnings:
我正在使用 OkHttp 3，并且不断收到泄漏的连接警告：
WARNING: A connection to https://help.helpling.com/ was leaked. Did you forget to close a response body?
Jul 14, 2016 6:57:09 PM okhttp3.ConnectionPool pruneAndGetAllocationCount
Everytime I get a ResponseBody, I either call .string()which supposedly closes the stream for me, or I explicitly close it in a finallyblock, in the following way:
每次我得到 a 时ResponseBody，我要么调用.string()它为我关闭流，要么finally通过以下方式在块中显式关闭它：
ResponseBody responseBody = response.body();
try (Reader responseReader = responseBody.charStream()) {
...
}
finally {
responseBody.close();
}
My application makes intense use of the network, and yet that warning appears frequently. I never observed any problem caused by this presumed leak, but I would still like to understand ifand whatI am doing wrong.
我的应用程序大量使用网络，但该警告经常出现。我从未观察到由这种假定的泄漏引起的任何问题，但我仍然想了解我是否做错了以及我做错了什么。
Could anyone shed some light on this?
任何人都可以对此有所了解吗？
采纳答案by Alphaaa
By upgrading to OkHttp 3.7, Eclipse started warning me of potential resource leaks. I found my problem to be in this method I wrote:
通过升级到 OkHttp 3.7，Eclipse 开始警告我潜在的资源泄漏。我发现我的问题出在我写的这个方法中：
public static Response getResponse(HttpUrl url, OkHttpClient client) throws IOException {
Builder request = new Request.Builder().url(url);
Response response = client.newCall(request.build()).execute();
if (!response.isSuccessful()) {
boolean repeatRequest = handleHttpError(response);
if (repeatRequest)
return getResponse(url, client, etag);
else
throw new IOException(String.format("Cannot get successful response for url %s", url));
}
return response;
}
I assumed that by always calling getResponse(url, client).body().string()the stream would close automatically. But, whenever a response was unsuccessful, an exception would raise before the execution of .string(), thus the stream would remain open.
我认为通过始终调用getResponse(url, client).body().string()流会自动关闭。但是，只要响应不成功，就会在执行之前引发异常.string()，因此流将保持打开状态。
Adding an explicit close in case of unsuccessful response solved the problem.
在响应不成功的情况下添加显式关闭解决了问题。
if (!response.isSuccessful()) {
boolean repeatRequest = handleHttpError(response);
response.close();
}
回答by Corrigan Johnson
As mentioned in the other answers, you have to close the response. A slightly cleaner approach would be to declare the ResponseBodyin the try block, so that it will be automatically closed.
正如其他答案中所述，您必须关闭响应。一种稍微简洁的方法是ResponseBody在 try 块中声明，以便它会自动关闭。
try(ResponseBody body = ....){
....
}
