From: | Andrew Dunstan <andrew(at)dunslane(dot)net> |
---|---|
To: | Ivan Voras <ivoras(at)freebsd(dot)org> |
Cc: | "tomas(at)tuxteam(dot)de" <tomas(at)tuxteam(dot)de>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Simple row serialization? |
Date: | 2008-01-27 17:38:52 |
Message-ID: | 479CC1AC.4090101@dunslane.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Ivan Voras wrote:
> On 27/01/2008, tomas(at)tuxteam(dot)de <tomas(at)tuxteam(dot)de> wrote:
>
>
>> On Sat, Jan 26, 2008 at 07:32:35PM +0100, Ivan Voras wrote:
>>
>>> Andrew Dunstan wrote:
>>>
>> [...]
>>
>>>> Or you could use an hstore (see contrib).
>>>>
>>> Doesn't seem applicable.
>>>
>> Have a closer look: it might :-)
>>
>
> I found these documents on hstore:
>
> http://developer.postgresql.org/pgdocs/postgres/hstore.html
> http://www.sai.msu.su/~megera/wiki/Hstore
>
> >From them, it seems "hstore" is another datatype (as in:
> 'a=>b'::hstore), which sounds good (though if it was me I'd have
> picked a different name for it, like "dict" or "hashmap" :) ) for
> storing both field names and their values, but I don't see a way to
> convert a row/record passed to a trigger to a hstore.
>
>
>
It's trivial to do in a plperl trigger, since it gets the new and old
records as hashes with the field names as keys.
cheers
andrew
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2008-01-27 17:45:16 | Re: CLUSTER and synchronized scans and pg_dump et al |
Previous Message | Gregory Stark | 2008-01-27 15:02:17 | CLUSTER and synchronized scans and pg_dump et al |