This is the mail archive of the
java-patches@gcc.gnu.org
mailing list for the Java project.
Re: [patch] minimal test case for sha-1 and md5
- From: Casey Marshall <csm at gnu dot org>
- To: tromey at redhat dot com
- Cc: Andreas Tobler <toa at pop dot agri dot ch>,Java Patches <java-patches at gcc dot gnu dot org>
- Date: Mon, 11 Oct 2004 16:16:20 -0700
- Subject: Re: [patch] minimal test case for sha-1 and md5
- References: <416AD974.1010600@pop.agri.ch><m3is9hc8l5.fsf@localhost.localdomain>
>>>>> "Tom" == Tom Tromey <tromey@redhat.com> writes:
>>>>> "Andreas" == Andreas Tobler <toa@pop.agri.ch> writes:
Andreas> xxx added a test case for 17903, I asked him to add one for
Andreas> md5 too.
Andreas> Here they are. Where should we place them? Is libjava.lang
Andreas> ok? It gives there 36 new results if everything works fine.
Tom> There wasn't an attachment or anything, so I didn't read the test
Tom> cases.
Tom> Is this something we can add to Mauve? For instance testing via
Tom> java.security.MessageDigest?
FYI GNU Crypto has a handful of tests for both MD5 and SHA-1, and we
already use Mauve as the test harness. The difference is we write to
our private API, while a Mauve test would write to MessageDigest.
I see now that the tests are based on GNU Crypto's, which in turn were
based off of published test vectors.
Maybe it would be better to find more test vectors (NIST republished
SHA-1 as SHA-160, so they may have more tests than we currently
implement) and to add those to Mauve. Because right now GNU Crypto's
SHA-1 test suite is one test vector, so we could really use more.
--
Casey Marshall || csm@gnu.org