Darren Duncan wrote:
For a loose DBI-alike example:
use FooDB;
my $dbms = FooDB->select(); # the DBMS is a singleton
my $sth = $dbms->prepare(<query code here>); # also "connect" with
query code
$sth->execute(<query params go here>);
...
From an initial response, I see that this code example may be confusing, and so
I'll provide one that's a bit more accurate, but still a bit pseudo-codey:
use FooDB;
my $machine = FooDB::select_machine(); # the VM is a singleton
my $process = $machine->new_process(); # a "connection" to the VM
$process->proc_invo( 'create_depot_mount', { # like DBI->connect()
name => 'db' , # namespace to mount this db schema under
we_may_update => 1 , # db access not read-only
details => [ ... ], # server/user/pass etc for this db schema
} );
$process->proc_invo( 'create_procedure', { # like DBI prepare() or DDL
depot => 'app' , # namespace to install query in
name => 'get_x', # name of the query
material => ... , # query code here, defined like an sproc
} );
my $gotten_x;
$process->proc_invo( 'fed.lib.app.get_x', { # like DBI execute()
gotten_x => \$gotten_x, # subject-to-update/INOUT param
name_of_x => 'joe' , # read-only/IN param
} );
my $x = $gotten_x->as_perl();
So hopefully it is more clear now how the module is supposed to work.
Note the uniformity of interface. Anything that you could conceivably have done
as a SQL statement in some DBMS you can do via proc_invo, which includes
"CONNECT" SQL.
The purpose of mount-points here, among other things, is to effectively support
cross-database queries. A direct analogy in SQLite is called ATTACH DATABASE.
About the only bit of this code that is somewhat implementation-specific is what
goes into "details", in the same way that connection strings in DBI vary per driver.
I could also define "routine handles" but haven't yet; if I did it could be like
this:
my $get_x = $process->proc_handle( 'fed.lib.app.get_x' );
my $gotten_x;
$get_x( { # like DBI execute()
gotten_x => \$gotten_x, # subject-to-update/INOUT param
name_of_x => 'joe' , # read-only/IN param
} );
Or I could add support for anonymous procedures to the Perl module, which are
even more like DBI's statement handles; they still have names, but the computer
generates the name, and practically speaking you'd only call it by the handle.
-- Darren Duncan
_______________________________________________
muldis-db-users mailing list
muldis-db-users@mm.darrenduncan.net
http://mm.darrenduncan.net/mailman/listinfo/muldis-db-users
