Dean Arnold wrote:
Jeff Zucker wrote:
I am preparing to add the ability to attach metadata to DBDs that
subclass DBD::File.
Will there be a std. i/f to get at that info from within
SQL::Statement (and its subclasses) ?
Yes definitely. though probably to start you'll just have access to the
SQL::Translator object, there won't be specific S:S methods to access it
for now.
(One of these days)
I hope to finish up an extension that wants to get at column type
info for better type conversion and esp. datetime support.
And presumably PK/FK info might be exploited for improved JOIN
processing.
Those things would be great to have.
Which leads to the question: will there be any hooks to enforce
referential integrity if PK/FK support is provided ?
That's certainly the long term plan. I'm not going to try to implement
that at the same time because um (insert some good reason other than
that I plain just don't have the time now). This will be some nice
juicy hacking for interested parties, hint, hint. :-)
Can I assume the YAML files are persistent ? If so, is there
any mechanism to assure they get deleted when a table is dropped ?
Sigh, yeah I was planning that.
--
Jeff
