Re: Partial aggregates pushdown
От | Alexander Pyhalov |
---|---|
Тема | Re: Partial aggregates pushdown |
Дата | |
Msg-id | d6da71cbe09676cda43ad39e5cb791a2@postgrespro.ru обсуждение исходный текст |
Ответ на | RE: Partial aggregates pushdown ("Fujii.Yuki@df.MitsubishiElectric.co.jp" <Fujii.Yuki@df.MitsubishiElectric.co.jp>) |
Ответы |
RE: Partial aggregates pushdown
|
Список | pgsql-hackers |
Fujii.Yuki@df.MitsubishiElectric.co.jp писал 2023-09-28 07:40: >> I'm not sure that I like this mechanics of adding sort group clauses - >> it seems we do in core additional work, which is of use only for >> one extension, but at least it seems to be working. > We cannot deparse the original sort group clauses and pathtarget > when performing partial aggreggate pushdown by any FDWs. > So I think the additional sort group clauses and pathtarget are > needed by any FDWs, not only postgres_fdw. > Hi. It seems to me that *fdw postfixes don't clarify things, but just make naming more ugly. + * Adding these Vars and PlaceHolderVars to PathTarget, + * FDW cannot deparse this by the original List of SortGroupClauses. + * So, before this adding process, + * setGroupClausePartial generates another Pathtarget and another + * List of SortGroupClauses for FDW. It seems that something like: /* * Modified PathTarget cannot be used by FDW as-is to deparse this statement. * So, before modifying PathTarget, setGroupClausePartial generates * another Pathtarget and another list List of SortGroupClauses * to make deparsing possible. */ sounds better. -- Best regards, Alexander Pyhalov, Postgres Professional
В списке pgsql-hackers по дате отправления: