Hi all.
I've a table with large objects, here is the definition:
PAFlow-emmebi=# \d documenti
Tabella "public.documenti"
Colonna | Tipo | Modificatori
---------+-----------------------+--------------
id_doc | character varying(50) | not null
file | oid |
kind | integer | not null
size | bigint |
Indici:
"documenti_pkey" chiave primaria, btree (id_doc, kind)
(sorry for the mixed language output)
I need to populate the size attribute with the size of the large object in file.
My first attempt was:
SELECT lo_lseek(lo_open(file, 262144), 0, 2) AS size FROM documenti ;
This opens the largeobject, and passes the result to the lo_lseek,
which goes up to the end of the largeobject, and thus computes the
size.
Now I could prepare an update which does the job. My doubt is that I
could use all the resources in the update (documenti table is quite
large).
I thought I could use something like this:
SELECT lo_lseek(lo_open(file, 262144), 0, 2) AS size, lo_close(0)
FROM documenti ;
which is quite nasty, and relies on side effects happening in the
proper order, but uses just one file descriptor for all the query.
Does anyone has any other suggestion?
Regards
Marco
--
Marco Bizzarri
http://iliveinpisa.blogspot.com/
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
