How to pass around collation information
in [PgSql]
|
From: Peter Eisentraut on 3 Jun 2010 03:43 On ons, 2010-06-02 at 16:56 -0400, Robert Haas wrote: > But in the end the only purpose of setting it on a column is to set > which one will be used for operations on that column. And the user > might still override it for a particular query. Of course. I'm just saying that it *can* be useful to attach a collation to a column definition, rather than only allowing it to be specified along with the sort operation. -- Sent via pgsql-hackers mailing list (pgsql-hackers(a)postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
From: David Christensen on 3 Jun 2010 08:30 On Jun 3, 2010, at 2:43 AM, Peter Eisentraut wrote: > On ons, 2010-06-02 at 16:56 -0400, Robert Haas wrote: >> But in the end the only purpose of setting it on a column is to set >> which one will be used for operations on that column. And the user >> might still override it for a particular query. > > Of course. I'm just saying that it *can* be useful to attach a > collation to a column definition, rather than only allowing it to be > specified along with the sort operation. How does collation relate to per-table/column encodings? For that matter, are per-table/column encodings spec, and/or something that we're looking to implement down the line? Regards, David -- David Christensen End Point Corporation david(a)endpoint.com -- Sent via pgsql-hackers mailing list (pgsql-hackers(a)postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
From: Peter Eisentraut on 3 Jun 2010 10:41 On tor, 2010-06-03 at 07:30 -0500, David Christensen wrote: > How does collation relate to per-table/column encodings? There is some connection between collations and character sets or encodings, because a collation is tied to one of those, just as a necessity of implementation (depending on implementation details). You could have per-column (or per-some-other-small-unit) collation with a global encoding, but a per-column encoding with a global collation wouldn't work. > For that matter, are per-table/column encodings spec, Yes. > and/or something that we're looking to implement down the line? I don't think anyone is seriously planning that. But per-column collations would have to come first, anyway. Of course there is always the possibility that someone comes up with an alternative plan that invalidates the above, but the above represents the facts from the SQL standard and other implementations. -- Sent via pgsql-hackers mailing list (pgsql-hackers(a)postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
From: Robert Haas on 3 Jun 2010 11:46 On Thu, Jun 3, 2010 at 3:43 AM, Peter Eisentraut <peter_e(a)gmx.net> wrote: > On ons, 2010-06-02 at 16:56 -0400, Robert Haas wrote: >> But in the end the only purpose of setting it on a column is to set >> which one will be used for operations on that column. And the user >> might still override it for a particular query. > > Of course. I'm just saying that it *can* be useful to attach a > collation to a column definition, rather than only allowing it to be > specified along with the sort operation. Oh, I agree with that, for sure. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise Postgres Company -- Sent via pgsql-hackers mailing list (pgsql-hackers(a)postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
First
|
Prev
|
Pages: 1 2 3 4 Prev: [HACKERS] How to pass around collation information Next: [BUGS] dividing money by money |