Re: final patch - plpgsql: for-in-array
От | Pavel Stehule |
---|---|
Тема | Re: final patch - plpgsql: for-in-array |
Дата | |
Msg-id | AANLkTi=cwb-z_YbP61a-cAeGzY3aqqkKbbgPK4XLsiba@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: final patch - plpgsql: for-in-array (Tom Lane <tgl@sss.pgh.pa.us>) |
Список | pgsql-hackers |
2010/11/18 Tom Lane <tgl@sss.pgh.pa.us>: > Pavel Stehule <pavel.stehule@gmail.com> writes: >> this note was a different -only a few people use FOR IN SELECT UNNEST >> for iteration over array. So from Robert's question (what is important >> for current code?) perspective the more significant is access to >> individual fields via subscripts. For example: > >> for i in 1..10000 loop >> s := s + A[i]; >> end loop > >> is slow, when high limit of array is some bigger number > 1000. > > True, but inventing new FOR syntax isn't going to help people who are > used to doing that. sure - I don't try it. Any change of this mean significant plpgsql's refactoring and significant increasing the size and complexity of code. More there can be still some overhead, because subscript can be expression. And in almost all cases people dislike to write subscripts. > > regards, tom lane >
В списке pgsql-hackers по дате отправления: