Giovanni, guess this patch must help. could you plz verify? --Noble On Tue, Mar 31, 2009 at 1:36 AM, Shalin Shekhar Mangar <shalinman...@gmail.com> wrote: > On Mon, Mar 30, 2009 at 11:39 PM, Giovanni De Stefano < > giovanni.destef...@gmail.com> wrote: > >> I was using a nightly build DIH 1.4 from last week, then I upgraded to the >> latest nightly build and still the same problem. :-( >> >> > I remembered an old email related to Oracle timestamps: > > http://www.lucidimagination.com/search/document/4a2fe28f099c3f15/custom_transformer_to_handle_timestamp > > But now I'm confused. > > 1. The initial stack trace has a ClassCastException with java.sql.Date when > used with DateFormatTransformer. That is expected because Jdbc will give > that type if the database field is of date type. Such values can be indexed > without the need for any conversion. > 2. However, when you tried without DateFormatTransformer, the error seems to > suggest that the return type of the row is not java.sql.Date because it is > being converted into a string and then DateField complains that it is not a > valid format. > 3. The older email I mentioned above links to the javadoc page which > suggests that the oracle timestamp does not inherit from java.sql.Date > > Three conflicts! > > I don't have an oracle instance to experiment with. Can you please provide > stack trace on the last nightly build or trunk with DateFormatTransformer > and without? You can also try the workaround I had suggested in the email I > mentioned above to see if that solves the problem. > > -- > Regards, > Shalin Shekhar Mangar. >
-- --Noble Paul
