我有以下方法在服务器上发出HTTP Get请求,将响应解析为Java源代码。我在同一时间为多个文件执行此操作。这对于前几个文件非常有效,但在一段时间后我得到一个例外:
Exception in thread "main" java.lang.OutOfMemoryError: Java heap space
行中出现异常:jp.parse(new StringReader(responseString));
我认为问题是内存泄漏,因为我试图解析的文件并不是很大。只有几十行代码。但我找不到这个例外的原因。任何提示?
public void retrieveSourceCode() {
try {
System.out.println("Try to get: " + getSourceCodeURI());
String responseString = RestServices.getInstance().sendGetRequestJsonTextToString(getSourceCodeURI());
JavaSourceFactory jsf = new JavaSourceFactory();
JavaParser jp = new JavaParser(jsf);
jp.parse(new StringReader(responseString));
Iterator<?> iterator = jsf.getJavaSources();
while(iterator.hasNext()) {
JavaSource source = ((JavaSource) iterator.next());
fileName = source.getQName().toString();
sourceCode = source.toString();
}
} catch (ClientProtocolException e) {
fileName = "no file name";
sourceCode = "no sourcecode available";
e.printStackTrace();
} catch (IOException e) {
fileName = "no file name";
sourceCode = "no sourcecode available";
e.printStackTrace();
} catch (RestServicesException e) {
fileName = "no file name";
sourceCode = "no sourcecode available";
e.printStackTrace();
} catch (RecognitionException e) {
fileName = "no file name";
sourceCode = "no sourcecode available";
e.printStackTrace();
} catch (TokenStreamException e) {
fileName = "no file name";
sourceCode = "no sourcecode available";
e.printStackTrace();
}
if (before == null) {
beforeSourceCode = "no before sourcecode available";
} else {
try {
String responseString = RestServices.getInstance().sendGetRequestJsonTextToString(getBeforeVersionURI());
JavaSourceFactory jsf = new JavaSourceFactory();
JavaParser jp = new JavaParser(jsf);
jp.parse(new StringReader(responseString));
Iterator<?> iterator = jsf.getJavaSources();
while(iterator.hasNext()) {
JavaSource source = (JavaSource) iterator.next();
beforeSourceCode = source.toString();
}
} catch (ClientProtocolException e) {
beforeSourceCode = "no before sourcecode available";
} catch (RecognitionException e) {
beforeSourceCode = "no before sourcecode available";
e.printStackTrace();
} catch (TokenStreamException e) {
beforeSourceCode = "no before sourcecode available";
e.printStackTrace();
} catch (IOException e) {
beforeSourceCode = "no before sourcecode available";
e.printStackTrace();
} catch (RestServicesException e) {
beforeSourceCode = "no before sourcecode available";
e.printStackTrace();
}
}
if (after == null) {
afterSourceCode = "no after sourcecode available";
} else {
try {
String responseString = RestServices.getInstance().sendGetRequestJsonTextToString(getAfterVersionURI());
JavaSourceFactory jsf = new JavaSourceFactory();
JavaParser jp = new JavaParser(jsf);
jp.parse(new StringReader(responseString));
Iterator<?> iterator = jsf.getJavaSources();
while(iterator.hasNext()) {
JavaSource source = (JavaSource) iterator.next();
afterSourceCode = source.toString();
}
} catch (ClientProtocolException e) {
afterSourceCode = "no after sourcecode available";
} catch (RecognitionException e) {
afterSourceCode = "no after sourcecode available";
e.printStackTrace();
} catch (TokenStreamException e) {
afterSourceCode = "no after sourcecode available";
e.printStackTrace();
} catch (IOException e) {
afterSourceCode = "no after sourcecode available";
e.printStackTrace();
} catch (RestServicesException e) {
afterSourceCode = "no after sourcecode available";
e.printStackTrace();
}
}
getChangeSet().addAffectedFile(getFileName());
}
答案 0 :(得分:2)
getChangeSet().addAffectedFile()
做什么?您可能想要使用已建议的探查器?
坚持下去吗?此外,您可能希望将该方法拆分为多个部分。
答案 1 :(得分:1)
为什么要继续使用相同的处理代码重复这些catch块?我会抓住Exception一次并完成它。其余的人什么都不做,只会增加视觉上的混乱。
但这不是你问题的答案......
这些课程是您的,还是来自图书馆?我猜它是Apache JAXMe库。
如果您使用的是Sun JVM,可以下载Visual VM 1.3.3,安装所有插件,打开它,然后开始您的流程以了解正在发生的事情。它将通过生成,CPU,线程等向您展示内存。这是一个很棒的工具。
为什么不尝试添加finally块并清除阵列?你没有在电话之间保留这些数据,对吗?如果是,其他呼叫者如何查找并使用结果?也许WeakHashMap可能是你的解决方案。如果需要,您将为JVM提供清理的机会。
你应该小心线程安全。 Servlet是共享的,所以如果你挂在那个可变对象上,你需要担心同步。