lways users
> who would try to
> specify both and than a behavior must exist for such a situation. If
> the 'f_dir_list'
> is preferred - such a behavior must be defined, too.
Is this problem DBD::File-specific?
What about SQLite3 databases?
--
Reinier Post
reatment
of character sets. These issues can be dealt with in a similar way.
The nice thing about such a matrix is that it will already be useful
on its own, even when it's unfinished.
--
Reinier Post
TU Eindhoven
Op di 13 december 2011 om 15:35:14 (+0100) schreef rehs...@googlemail.com (Jens
Rehsack):
> 2011/10/4 Reinier Post :
[...]
> > For instance, I have a simple script that needs to count the number of
> > distinct values in columns. This seems like a pretty elementary thing
> &g
Op do 05 april 2012 om 10:02:35 (+0200) schreef rehs...@googlemail.com (Jens
Rehsack):
> > I have included the devel list to see how others think
>
> So did I - good suggestion.
But I replied in RT - details there. Summary:
+ a 0-column CSV table with headers still has a(n empty) header line
e DBI backends that use it.
What I really want is something like this as a generic,
backend-independent interface to SQL, and that is not at all
what SQL::Statement, or DBI for that matter, tries to provide.
Some other DBI-based modules do take some steps in that direction.
--
Reinier Post
P.S.
> $stmt->from->join[1]->column = 'salary';
should be
$stmt->from->join[1]->table = 'salary';
Sorry about that.
The idea is to have something like what the CGI module provides for HTML.
--
Reinier
On Sat, Aug 11, 2012 at 10:44:14PM +0100, Tim Bunce wrote:
> The main function of the module, the 'clever' part, is the generation of
> test data that has referential integrity. Inserting the generated data
> into the table is relatively trivial.
>
> The name should focus on the clever bit more t
