I tested with Jaybird 1.5 and I could reproduce this bug. To be
sure this bug had nothing to do with my front-end tool, I used
another sql front-end (hsqldb manager), and I still could reproduce
this error.

To reproduce this error consistently, my previous test case has
to be revised as:

drop table test;
create table test (news varchar(255));

Insert about 200 rows for test. Some rows have columns longer than
80 characters.

Then, do
select substr(news,1,2) from test;

This first select will issue an exception as expected.

Then issue the same sql again,
select substr(news,1,2) from test;

This time, Jaybird will complain with the following:
java.lang.NullPointerException
at org.firebirdsql.jdbc.field.FBField.isNull(FBField.java:177)
at org.firebirdsql.jdbc.FBResultSet.getField
(FBResultSet.java:336)
at org.firebirdsql.jdbc.FBResultSet.getObject
(FBResultSet.java:295)
at Venus.doShowResultset(Venus.java:4570)
at Venus$11.run(Venus.java:4341)
...


Trying to narrow down the problem, I observer that the first
select statement caused a SQLException, but the second select
statement does NOT cause any exception. As a result, since the
resultset is not null, the programs tries to display it. ==>
Error.

BEn






--- In Firebird-Java@yahoogroups.com, "Roman Rokytskyy"
<rrokytskyy@a...> wrote:

> Hi,
>
> > > I have reported the bug in the substr and other string functions
> > > in the Firebird general group. This occurs when the string

argument

> > > is longer than 80 characters. This bug helps expose another

bug

> > > (???) in the JDBC driver.
> >
> > Thanks a lot for the report! I will check this bug.
>
> I tried to reproduce bug with JayBird 1.5 and could not. Here's the
> report I get from driver:
>
> org.firebirdsql.jdbc.FBSQLException: GDS Exception. 335544321.
> arithmetic exception, numeric overflow, or string truncation
>
> Can you create a reproducable test case for JayBird 1.5?
>
> Thanks!
> Roman