On 1/25/08 6:05 PM, "Tore Halset" <[EMAIL PROTECTED]> wrote:
>> PostgreSQL 8.2: FAIL (74 failed tests)
>
> Here, only(?) one failed. It was
> merge.CreateTableToModelTest.testAddTable. I am looking into that one
> now.
> Did you do a "createlang plpgsql databasename"? I got fooled by that
> one before.
That did the trick. I had reformatted my machine after my botched Leopard
upgrade. I'll make a note in the developer's guide.
The one failed test I have seems to be related to merge stuff:
----------------------------------------------------------------------------
---
Test set: org.apache.cayenne.merge.CreateTableToModelTest
----------------------------------------------------------------------------
---
Tests run: 1, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 20.947 sec
<<< FAILURE!
testAddTable(org.apache.cayenne.merge.CreateTableToModelTest) Time elapsed:
20.944 sec <<< FAILURE!
junit.framework.AssertionFailedError: tokens to db expected:<1> but was:<0>
at junit.framework.Assert.fail(Assert.java:47)
at junit.framework.Assert.failNotEquals(Assert.java:282)
at junit.framework.Assert.assertEquals(Assert.java:64)
at junit.framework.Assert.assertEquals(Assert.java:201)
at org.apache.cayenne.merge.MergeCase.assertTokens(MergeCase.java:177)
at
org.apache.cayenne.merge.MergeCase.assertTokensAndExecute(MergeCase.java:188
)
at
org.apache.cayenne.merge.CreateTableToModelTest.testAddTable(CreateTableToMo
delTest.java:87)
>
>> Derby: FAIL (9 failed tests)
>
> Oups, mostly merging. I'll have to work on this..
Great. I'll hold up.
>
>> For example, I see a lot of messages about duplicate PK tables
>> coming from
>> org.apache.cayenne.merge.ExecutingMeregContext#executeSql():60
>
> I am using PkGenerator.createAutoPkStatements. This method returns
> "CREATE TABLE AUTO_PK_SUPPORT" (for some adapters) even though the
> table allready exist. It should be possible to create the table if it
> exist, otherwise just insert the missing rows. For adapters using
> sequences, this should not be the problem.
>
> The merging still works fine, so this is not a blocker for me. I will
> however look into it.
Yeah, as I said, it was more of an alarm than anything else. I get worried
when I see stack traces with passing tests. But, it's nothing that should
hold up the release.
--
Kevin
