This email is a heads-up.
A couple months ago I announced that I was going to experiment with
cross-cpan refactoring where I would update a variety of DBI wrappers
to use my SQL::Routine module internally. Since then I have changed
my mind and decided not to do this, largely due to the complexity of
the task, including that many of the modules are regularly being
updated.
Instead, I will be going ahead with a plan I had much earlier on
which is to make a new Pure Perl module that is a wrapper for Rosetta
that emulates DBI, probably to be named Rosetta::Emulator::DBI. Then
any other programs or modules that use DBI (including all the
aforementioned wrapper modules) should be able to use this as a
drop-in replacement.
What will happen behind the scenes with your database instructions
that didn't happen before is that their SQL strings will be
completely parsed into a SQL::Routine model and then re-assembled
into new SQL strings which are actually executed, assuming your
actual database engine is unchanged.
The advantage to you of having interjected Rosetta into your
communications with the database is that you should be able to
substitute any other database engine for your original one and your
application will continue to work without any changes on its part.
Similarly, your application will then be able to work unchanged with
storage engines that it would have never worked with before except
with large changes; this includes databases that don't use SQL as
their native instruction set, or for which a DBI driver doesn't exist.
I expect to have this functional with a few database products by the
end of June.
As a side effect, since I am maintaining native Perl 6 versions of my
modules in parallel with the Perl 5 ones, porting your applications
to Perl 6 and/or Parrot later can also be assisted by using them.
Any feedback or questions or offers of assistence are welcome.
Have a good day. -- Darren Duncan
