Tom> No, it shouldn't; read the SQL spec. AFAICS the syntax
Tom> select interval '1' month is perfectly spec-compliant. The
Tom> variant select interval '1 month' is *not* per-spec, it is a
Tom> Postgres-ism.
That is my understanding, though I don't have a copy of the spec (my
r
> "Bruce" == Bruce Momjian writes:
Bruce> First, the fundamental issue with intervals is that they
Bruce> are not tied to a particular date, meaning there is no way
Bruce> to map a specific number of days to a number of months.
Bruce> (Some days are also 23 or 25 hours but tha
Roy> I don't have a copy of the spec, but according to "A guide to
Roy> the SQL standard" conversions like this that would discard
Roy> data are supposed to raise an exception.
Just to clarify, my understanding is that in ANSI SQL it is valid to
convert from the data type INTERVAL YEAR
Tom> AFAICS the syntax
Tom> select interval '1' month
Tom> is perfectly spec-compliant.
Well, it's not _perfectly_ spec compliant, because AIUI SELECTs
without FROM clauses are a postgres-ism, too. But I'm just
nitpicking...
-roy
---(end of br
> Given the error message, this seems to be the whole plpgsql caches
> query plans but we don't invalidate those plans when there are schema
> changes. In all currently released versions you pretty much need to
> use EXECUTE on any queries where the table may go away, for example,
> any use of te
> "Roy" == Roy Badami <[EMAIL PROTECTED]> writes:
Roy> FWIW, this is documented behaviour (section 35.1 of the
Roy> manual).
That's 35.1 in the 8.0 doc, BTW. It's 37.1 in the 7.4 doc.
Roy> -roy
---(end of broadcast)---
TIP
Roy Badami wrote:
>
> Tom> No, it shouldn't; read the SQL spec. AFAICS the syntax
> Tom> select interval '1' month is perfectly spec-compliant. The
> Tom> variant select interval '1 month' is *not* per-spec, it is a
> Tom> Postgres-ism.
>
> That is my understanding, though I don
Bruce> So, we have a few major problems with intervals. Let me
Bruce> think a little and I will summarize.
FWIW, AFAICT the problems I reported in bug 1517 and 1518 all relate
to undocumented features of PostgreSQL.
All the documented interval functionality works fine. The
undocumented
Roy Badami <[EMAIL PROTECTED]> writes:
> All the documented interval functionality works fine. The
> undocumented support for ANSI SQL interval data types and litereals
> doesn't :-/
I think the reason it's not documented is precisely that Tom never
finished it. It may not be very far away thoug
On Fri, 18 Mar 2005, Tom Lane wrote:
> Andrew - Supernews <[EMAIL PROTECTED]> writes:
> > It turns out that the scenario above is trivial to hit in 8.0 using
> > referential constraints; RI triggers cache their plans, and on 8.0 the RI
> > query is planned as a seqscan if the tables are freshly cr
On Sat, 19 Mar 2005, Roy Badami wrote:
> Roy> I don't have a copy of the spec, but according to "A guide to
> Roy> the SQL standard" conversions like this that would discard
> Roy> data are supposed to raise an exception.
>
> Just to clarify, my understanding is that in ANSI SQL it is
Tom> Feel like hacking the code?
Hmm, in principle I might take a look some time; in reality it's
unlikely I'll have time any time soon...
There are some design issues involved, though. If you have the type
modifier, do you isnist on SQL syntax in the string?
ie do you accept
interval '
Roy Badami <[EMAIL PROTECTED]> writes:
> ie do you accept
>interval '1 day 1 hour' day to second
I think we have to, and the reason is that this isn't different under
the hood from reading the external value '1 day 1 hour' and storing
it into a column that has the DAY TO SECOND typmod. If we
On 2005-03-19, Stephan Szabo <[EMAIL PROTECTED]> wrote:
> On Fri, 18 Mar 2005, Tom Lane wrote:
>> Hm. One thing we could do is to throw in some default values when we
>> see the table has exactly zero pages --- perhaps ye olde traditional
>> 1000/10, or possibly something else, but anyway not exac
> "Tom" == Tom Lane <[EMAIL PROTECTED]> writes:
>> ie do you accept interval '1 day 1 hour' day to second
Tom> I think we have to, and the reason is that this isn't
Tom> different under the hood from reading the external value '1
Tom> day 1 hour' and storing it into a column t
Similary the undocumented postgresism of interpreting
INTERVAL '1:02'
as 1 hour 2 minutes is consistent with the ANSI
INTERVAL '1:02' HOUR TO MINUTE
but not with the ANSI
INTERVAL '1:02' MINUTE TO SECOND
which of course means 1 minute 2 seconds.
The fact is that AN
Roy Badami <[EMAIL PROTECTED]> writes:
> Similary the undocumented postgresism of interpreting
>INTERVAL '1:02'
> as 1 hour 2 minutes is consistent with the ANSI
> INTERVAL '1:02' HOUR TO MINUTE
> but not with the ANSI
>INTERVAL '1:02' MINUTE TO SECOND
> which of course mea
17 matches
Mail list logo
