I found this on the web that seems to explain why the problem is occurring...
http://opensource2.atlassian.com/projects/spring/browse/SPR-1518 "Sean Dockery" <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED] >I have a TestCase as follows... > > package example; > > import java.util.Date; > > import junit.framework.TestCase; > > public class JavaBeanTest extends TestCase { > > public JavaBeanTest(String testName) { > super(testName); > } > > public void testThrowawayCopy() throws Exception { > // setup > final JavaBean bean = new JavaBean(); > bean.setId(1L); > bean.setName("hello"); > // exercise > final JavaBean copy = bean.throwawayCopy(); > bean.setExpired(new Date()); > copy.setId(2L); > copy.setName("world"); > // verify > assertEquals("id affected by copy.setId(long) call", 2L, bean.getId()); > assertEquals("name not affected by copy.setName(String) call", "hello", > bean.getName()); > assertNotNull("expired has changed", bean.getExpired()); > > assertEquals("id has changed", 2L, copy.getId()); > assertEquals("name has changed", "world", copy.getName()); > assertNull("expired not affected by bean.setExpired(Date) call", > copy.getExpired()); > } > > public static Test suite() { > TestSuite suite = new TestSuite(); > Test test = new JavaBeanTest("testThrowawayCopy"); > for (int i = 0; i < 1000; i++) { > suite.addTest(test); > } > return suite; > } > > } > > ...for which I have a JavaBean class as follows... > > package example; > > import java.lang.reflect.Method; > import java.util.Date; > > public class JavaBean { > > private long id; > private String name; > private Date expired; > > public JavaBean() { > // default constructor for JavaBean > } > > public Date getExpired() { > return expired; > } > > public void setExpired(Date expired) { > this.expired = expired; > } > > public long getId() { > return id; > } > > public void setId(long id) { > this.id = id; > } > > public String getName() { > return name; > } > > public void setName(String name) { > this.name = name; > } > > public JavaBean throwawayCopy() { > return new ThrowawayJavaBean(); > } > > private final class ThrowawayJavaBean extends JavaBean { > > private boolean propagateId; > > private ThrowawayJavaBean() { > copyProperties(); > propagateId = true; > } > > private void copyProperties() { > try { > Method[] methods = JavaBean.this.getClass().getDeclaredMethods(); > for (int i = 0; i < methods.length; i++) { > final Method source = methods[i]; > if (source.getName().startsWith("get")) { > Method target = getClass().getMethod("set" + > source.getName().substring(3), new Class[] { source.getReturnType() }); > target.invoke(this, new Object[] { source.invoke(JavaBean.this, new > Object[] {}) }); > } > } > } catch (Exception e) { > throw unchecked(e); > } > } > > public void setId(long id) { > super.setId(id); > > if (propagateId) { > JavaBean.this.setId(id); > } > } > > private RuntimeException unchecked(Exception e) { > if (e instanceof RuntimeException) { > return (RuntimeException) e; > } > return new RuntimeException(e); > } > > } > > } > > > This code has been adapted from some code in a production application at > my > work. The production code actually uses Spring's > BeanUtils.copyProperties(Object source, Object target) method in the > ThrowawayJavaBean constructor, but we should be accomplishing the same > thing > by using reflection directly. > > As you can see, the ThrowawayJavaBean class is a *private* nested class. > During load tests of the production application, the production code > (running in Tomcat) *sometimes* experiences an IllegalAccessException due > to > the fact that we're using reflection against a private nested class. The > problem has never occurred when the visibility of the nested class is > public. Additionally, I have *never* seen the exception occur in the > JUnit > TestRunner with neither the code I have presented above, nor the > production > code itself. > > My confusion lies with the question of why the IllegalAccessException > occurs > only infrequently in the production code running under Tomcat instead of > always occurring. That is, why doesn't (wouldn't) it happen all the > time--for example, in my TestCase? > > Thanks for your time. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
