No ticket for the SELECT NEXT VALUE FOR my_seq. Would you mind filing one?
On Tue, Apr 8, 2014 at 10:50 PM, Dan Di Spaltro <[email protected]>wrote: > Thanks good idea with using the system table. Regarding your first > suggestion James, It's pretty hard to use that when it's basically > impossible to get the autogenerated value back, that was alluded to in a > previous thread. Is there a ticket open for this already, I couldn't find > anything. > > Thanks, > > -Dan > > > On Tue, Apr 8, 2014 at 10:27 PM, James Taylor <[email protected]>wrote: > >> We should support statements without any FROM clause for cases like this. >> That way you could do something like this: >> {code} >> SELECT NEXT VALUE FOR my_seq; >> {code} >> In the meantime, there are a couple of workarounds: >> - do the NEXT VALUE FOR in the UPSERT statement, as this is the most >> common use case >> - use a table where you know there will be rows, like the SYSTEM.SEQUENCE >> table or the SYSTEM.CATALOG table. Best to put a LIMIT clause on your >> SELECT so you don't select the entire table each time. >> >> Thanks, >> James >> >> >> >> On Tue, Apr 8, 2014 at 6:26 PM, Dan Di Spaltro >> <[email protected]>wrote: >> >>> So I created a new table and sequence. My typical pattern for that is >>> selecting from the sequence using a table, but I noticed if the table is >>> empty it doesn't select anything. While in practice this won't really be a >>> problem but it's a pain when testing, is this intended behavior? >>> >>> Thanks! >>> >>> -Dan >>> >>> -- >>> Dan Di Spaltro >>> >> >> > > > -- > Dan Di Spaltro >
