From: | Alexey Bashtanov <bashtanov(at)imap(dot)cc> |
---|---|
To: | Gajus Kuizinas <gajus(at)gajus(dot)com>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: A case for UPDATE DISTINCT attribute |
Date: | 2018-12-20 16:59:29 |
Message-ID: | 7afde60a-ec30-3c62-bbe2-6f73845a2cac@imap.cc |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Hello Gajus,
> I have observed that the following pattern is repeating in our data
> management programs:
>
> UPDATE
> event
> SET
> fuid = ${fuid},
> venue_id = ${venueId},
> url = ${url}
> WHERE
> id = ${id} AND
> fuid IS != ${fuid} AND
> venue_id IS != ${venueId} AND
> url IS DISTINCT FROM ${url};
>
...
> Meanwhile, a WHERE condition that excludes rows with matching values
> makes this into a noop in case of matching target column values.
For this to hold, you need your conditions in WHERE to be ORed, not ANDed.
>
> UPDATE DISTINCT
> event
> SET
> fuid = ${fuid},
> venue_id = ${venueId},
> url = ${url}
> WHERE
> id = ${id}
>
> would encourage greater adoption of such pattern.
>
> Is there a technical reason this does not existing already?
>
> ᐧ
At least a bunch of questions and concerns arise. Like these:
1) We cannot treat it as a syntactic sugar only and just expand it on
parsing stage,
as the expression to generate the value assigned may be volatile, like
UPDATE ... SET ... = random();
2) How should this interact with triggers? E.g. when NEW and OLD were
the same
before BEFORE UPDATE trigger execution, but would be different after. Or
visa versa.
Should they be included into transition tables?
3) Should RETURNING clause return the non-updated rows?
4) It must be not easy to guarantee anything if there is a FROM clause,
a target row is present in the join more than once.
5) We need to fail correctly if one of the column data types doesn't
have an equality operator.
Best regards,
Alexey
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2018-12-20 17:26:51 | Re: Switching to 64-bit Bitmapsets |
Previous Message | Bossart, Nathan | 2018-12-20 16:48:11 | Re: A few new options for vacuumdb |