making trimquery output immediately reexecutable + chainable

This isn't an issue as much as a feature enhancement.

BTW, Thanks for this project, very powerful stuff.

From the API docs, I'm not sure how to make this entire thing a "chainable"
operation, as in (working off of the 10min intro):

1) define a dataset and metadata objects
2) execute a statement
3) produces another dataset (we'll call it OUTPUT1)

If I wanted to take OUTPUT1, how would I compile a new schema to execute
statements against it? I could manually produce one beforehand since I know
the select statement I executed, but it would be nice if I could get the
SelectStatement object to produce a schema object for it's resultset, since
that encapsulates all the info on the query anyway, so I could immediately
compile a new schema and execute another select against the output.

I can probably work around it in the near term.

Original issue reported on code.google.com by carlemueller on 26 Jun 2007 at 9:39

x550j / trimpath

making trimquery output immediately reexecutable + chainable #1