-
Notifications
You must be signed in to change notification settings - Fork 685
Description
I was thinking about #9199 and led myself down a rabbit hole.
In that issue, we discussed making sure that all backends create views when doing a read_* so they don't clobber existing data.
Then if the user wants to re-use an existing name, they can first drop the existing table object using drop_table or drop_view.
One hiccup there, how do they know whether it's a table or a view? Chances are, they don't, especially if we created it for them.
Given that, should we rethink the output of list_tables (and the TablesAccessor) to include information about what kind of abstract table a given object is?
This dovetails with the work on the DDL accessor thing, but we also can't expect users to know to look in con.ddl.list_views() to check if the existing name is there or not.
What if we fancy up the output of con.tables to also indicate whether a given object is a table or a view (or other) ?
Metadata
Metadata
Assignees
Labels
Type
Projects
Status