I meant to get back to this when I got back to my JOOQ code...sorry to let 
so much time pass. To answer your question, I generate my SQL by creating a 
Query object using a DSLContext created using a null connection and a 
dialect, and in this case the select fields were a Collection of 
Field<Object> (passed into select()) generated by a bit of code getting the 
list of desired attributes and aliases from another source and performing a 
DSL.fieldByName() to create them. Only after constructing the entire Query 
object am I calling getSQL() on it.

Would it really be better to render the various parts of my query to 
separate strings and patch them together after? That strikes me as taking a 
certain element of the knowledge of how to construct SQL  syntax and 
putting it into my code rather than having JOOQ do it for me. Though I 
perhaps since my code has to "know how" put together the JOOQ syntax calls, 
perhaps this is an illusary distinction. What is the reason you think I 
should prefer rendering unattached QueryParts?

I should note that this code is working fine so far, other than this one 
unimportant encounter with a very old PostgreSQL server...and I've since 
eliminated the "alias all columns" bit which was convenient for reasons too 
convoluted to explain but which became undesirable for other reasons. But 
more complexity will likely be going into my code and the SQL that it 
generates as I develop it, so if something about my current approach will 
make that prone to breaking I'd certainly like to head that off...

-- 
You received this message because you are subscribed to the Google Groups "jOOQ 
User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to