[Grok-dev] megrok.rdb, Oracle, and views.

[Martijn Faassen]

Hi there,

A bit I'm confused about is that you say SA can't reflect views but 
you're still using rdb.reflected().

If we had a spelling for rdb.dontcreate() (rdb.evolve is way too cute in 
my mind - and confusing, as people will think about schema evolution), 
would that be sufficient for you to stop using rdb.reflected() or would 
you still need it?

Concerning spelling, I see you use 'useexisting', which is a SQLAlchemy 
feature, correct? Perhaps the spelling should be 'rdb.useexisting'. In 
this case it'll always try to use an existing table and never try to 
create it.

I'm hoping your use case can be served by having a 'rdb.useexisting()', 
and a patch that listens to this in setup.py so it skips creating the table.

Regards,

Martijn

P.S. I see you use __table_args__. Any reason you can't use rdb.tableargs?