In message <20100714103700.1592f2388...@eris.apache.org>, odea...@apache.org
writes:
>
> Author: odeakin
> Date: Wed Jul 14 10:36:59 2010
> New Revision: 964001
>
> URL: http://svn.apache.org/viewvc?rev=964001&view=rev
> Log:
> To match the RI's behaviour, resolve and store the hostname when getHostName
> is called. Add a flag so we do not resolve the hostname for every call to get
> HostName. Also add a regression test for the behaviour.
>
> [SNIP]
>
> --- harmony/enhanced/java/trunk/classlib/modules/luni/src/test/api/common/org
> /apache/harmony/luni/tests/java/net/InetSocketAddressTest.java (original)
> +++ harmony/enhanced/java/trunk/classlib/modules/luni/src/test/api/common/org
> /apache/harmony/luni/tests/java/net/InetSocketAddressTest.java Wed Jul 14 10:
> 36:59 2010
> @@ -31,7 +31,9 @@ public class InetSocketAddressTest exten
> public void test_ConstructorLjava_lang_StringI() throws Exception {
> // regression test for Harmony-1042
> InetSocketAddress address = new InetSocketAddress("127.0.0.1", 0);
> - assertNotNull(address.getHostName());
> + assertEquals("/127.0.0.1:0", address.toString());
> + assertEquals("localhost", address.getHostName());
> + assertEquals("localhost/127.0.0.1:0", address.toString());
> }
>
> /**
I would have assumed that the removed:
assertNotNull(address.getHostName());
was not:
assertEquals("localhost", address.getHostName());
for good reasons[0].
Therefore it might be better to make the new code:
assertEquals("/127.0.0.1:0", address.toString());
String localhostName = address.getHostName();
assertNotNull(localhostName);
assertEquals(localhostName+"/127.0.0.1:0", address.toString());
or something like that?
Regards,
Mark.
[0] Such as "localhost" being a rather arbitrary string that may be
different on different machines? Some older Debian machines
have "localhost.localdomain" first in their /etc/hosts entries IIRC.
