On 2025-Jan-20, Sami Imseih wrote: > Patch looks good to me, but I am not sure about this part of the comment: > > "Only the topmost one is to be considered here; the child constraints > must be left alone," > > In this case, none of the pg_constraint entries are actually considered. > right?
Right. The parent one because it has conparentid = 0, the others because their parent constraint is in the list. -- Álvaro Herrera Breisgau, Deutschland — https://www.EnterpriseDB.com/
