On Tue, 5 May 2009 15:54:48 +0100, Tim Bunce <[email protected]>
wrote:

> On Tue, May 05, 2009 at 01:49:10PM +0200, H.Merijn Brand wrote:
> > On Tue, 5 May 2009 12:08:53 +0100, Tim Bunce <[email protected]>
> > wrote:
> > 
> > > On Mon, May 04, 2009 at 04:06:47PM +0200, H.Merijn Brand wrote:
> > > > I've hit a failure, which I am not about if it is my fault or that I
> > > > have to blame something else
> > > > 
> > > >   $dbh->do ("drop table $_") for $dbh->tables ();
> > > > 
> > > > Now that DBI returns the tables quoted, I expect that to work.
> > > > 
> > > > DBD::CSV uses the current user name as schema name, so the table list
> > > > looks like
> > > > 
> > > >   $VAR1 = [
> > > >     '"merijn".testaa.csv',
> > > >     '"merijn".testab',
> > > >     '"merijn".testac',
> > > >     '"merijn".testad.txt',
> > > >     '"merijn".testae.csv',
> > > >     ];
> > > 
> > > [Ignoring the issue you're refering to]
> > > 
> > > Why should DBD::CSV support schemas at all?
> > 
> > Because it was in there from the start. (In fact it was in DBD::File).
> > Not that I /like/ it, nor see the use of it, but the default
> > schema-name for DBD-File is the owner of the folder in which the
> > datafile resides
> 
> That seems wrong. Schemas primarily provide namespaces. The association
> with username is mostly an accident of history.

Which IMHO I cannot remove bluntly as a judgement error.

> > > What value does that give?
> > 
> > beats me, but I'm sure I break things if I simply remove it
> > 
> > > What use-cases does it help?
> > 
> > I have just implemented f_schema in DBD::File, so I can disable the
> > darn thing with { f_schema => undef }
> 
> Ah, ok. That's a good thing :)
> 
> > > What are the semantics of 'schemas' in DBD::CSV?
> > 
> > none that I am aware of
> 
> There must be some...
> 1. when reading metadata, the default schema-name is the owner of the
>     folder in which the datafile resides

With DBD::CSV, I bet all the files are owned by the same person (that
also owns the folder). If applications rely on folder ownership for
schema names, I have lost it there ... :)

> 2. when refering to a table name in an SQL statement
>     the schema portion is... ignored?

yes .... I think
/me quickly adds a todo test for DBD::CSV to check that

> p.s. You've missed DBI 1.608 but hopefully the gap before 1.609 won't be
> as large.

I'll sync and commit then
I have to wait for SQL::Statement anyway

Still writing tests and tests and tests (otherwise I would never have
found these stupid errors)

-- 
H.Merijn Brand  http://tux.nl      Perl Monger  http://amsterdam.pm.org/
using & porting perl 5.6.2, 5.8.x, 5.10.x, 5.11.x on HP-UX 10.20, 11.00,
11.11, 11.23, and 11.31, OpenSuSE 10.3, 11.0, and 11.1, AIX 5.2 and 5.3.
http://mirrors.develooper.com/hpux/           http://www.test-smoke.org/
http://qa.perl.org      http://www.goldmark.org/jeff/stupid-disclaimers/

Reply via email to