Re: Count(*) Question
От | Joel Burton |
---|---|
Тема | Re: Count(*) Question |
Дата | |
Msg-id | JGEPJNMCKODMDHGOBKDNGEFBCMAA.joel@joelburton.com обсуждение исходный текст |
Ответ на | Re: Count(*) Question (Martijn van Oosterhout <kleptog@svana.org>) |
Ответы |
Re: Count(*) Question
|
Список | pgsql-general |
> > I was reading through the Rules section of the online docs, > and noticed the > > following note: (* is just an abbreviation for all the > attribute names of a > > relation. It is expanded by the parser into the individual > attributes, so > > the rule system never sees it.) > > Does this mean that count(*) may return less than the total > number of > > records if all the fields in a record are NULL? > > Yes, I beleive so. > > > If this is true, is there a better way to get a count of records? > > I think count(1) is the common suggestion. Interesting. In 7.3devel, it does not fail to count the completely-null rows in count(*). Does it actually do this for any version?
В списке pgsql-general по дате отправления: