Re: junit failure?

On Sun, 16 Nov 2003, Larry Rosenman wrote:

> > Tough to tell what the problem is exactly.  The JDBC tests don't have the
> > equivalent of the main regression test's diffs that you can send around.
> > This test could fail if you had a table in the database you are running
> > this against that had a name starting with "test".  Please try out the
> > attached patch which adds some debug info for this particular failure.
> >
> > Kris Jurka
>
>     [junit] ------------- Standard Output ---------------
>     [junit] TABLE_CAT, TABLE_SCHEM, TABLE_NAME, COLUMN_NAME, DATA_TYPE,
> TYPE_NAME, COLUMN_SIZE, BUFFER_LENGTH, DECIMAL_DIGITS, NUM_PREC_RADIX,
> NULLABLE, REMARKS, COLUMN_DEF, SQL_DATA_TYPE, SQL_DATETIME_SUB,
> CHAR_OCTET_LENGTH, ORDINAL_POSITION, IS_NULLABLE
>     [junit] null, public, test_b, descr, 12, text, -1, null, 0, 10, 1,
> null, null, null, null, -1, 1, YES
>     [junit] null, public, test_b, imageid, 4, int4, 4, null, 0, 10, 1,
> null, null, null, null, 4, 2, YES
>     [junit] null, public, test_b, id, 4, int4, 4, null, 0, 10, 1, null,
> null, null, null, 4, 3, YES
>     [junit] null, public, testmetadata, id, 4, int4, 4, null, 0, 10, 1,
> this is a column comment, null, null, null, 4, 1, YES
>     [junit] null, public, testmetadata, name, 12, text, -1, null, 0, 10, 1,
> null, null, null, null, -1, 2, YES
>     [junit] null, public, testmetadata, updated, 93, timestamp, 8, null, 0,
> 10, 1, null, null, null, null, 8, 3, YES
>     [junit] ------------- ---------------- ---------------

This shows that a table named test_b is in the database.  This nearly
certainly came from you previously running example/metadata which creates
a table of that design.  I can't understand why you aren't seeing tables
test_a or test_c here though.  In any case it is not a jdbc problem.  If
you log into the test database and delete any objects you find there the
unit tests should run through with no errors.

Kris Jurka