Re: DBM/DBD::File/SQL::Statement failures

On Wed, Feb 29, 2012 at 10:49:09AM +0000, Martin J. Evans wrote:
> On 26/02/12 12:27, Tim Bunce wrote:
> >On Sat, Feb 25, 2012 at 02:18:56PM -0800, Jonathan Leffler wrote:
> >
> >>    I'm slightly puzzled by a couple of the tests skipping because 'Not running with SQL::Statement', when
> >>    in fact I have SQL::Statement installed.
> >>
> >>    t/52dbm_complex.t ............... ok
> >>    t/zvg_52dbm_complex.t ........... ok
> >>    t/zvn_52dbm_complex.t ........... skipped: Not running with SQL::Statement
> >>    t/zvp_52dbm_complex.t ........... ok
> >>    t/zvxgn_52dbm_complex.t ......... skipped: Not running with SQL::Statement
> >>    t/zvxgnp_52dbm_complex.t ........ skipped: Not running with SQL::Statement
> >>    t/zvxgp_52dbm_complex.t ......... ok
> >>    t/zvxnp_52dbm_complex.t ......... skipped: Not running with SQL::Statement
> >>
> >>    Maybe it is the message that needs tweakiing?
> >
> >Yes, they should be changed to "requires SQL::Statement module".
> 
> I thought the "Not running with SQL::Statement" meant the test code
> uses SQL::Statement but the test code was not running with
> SQL::Statement e.g., t/zvn_52dbm_complex.t uses Nano and zvp* is Pure
> Perl so they don't use SQL::Statement.  For those tests it is
> irrelevant if you have SQL::Statement or not - no?

Ah, right. The "n" variants set the DBI_SQL_NANO env var in order to
test the bundled DBI::SQL::Nano SQL engine. So a better phrasing
would be something like:

    skipped: DBI::SQL::Nano is being used

I'll change it. Thanks Martin.

Tim.