Hi all,
I have the following tables (parent and two children)
CREATE SEQUENCE person_seq;
CREATE TABLE person(
_id integer DEFAULT nextval('person_seq') NOT NULL,
_timestamp TIMESTAMP NOT NULL,
_lastModified TIMESTAMP NOT NULL,
name VARCHAR(255) NOT NULL,
age INTEGER DEFAULT NULL,
PRIMARY KEY (_id)
);
CREATE TABLE person_nationality(
_id serial NOT NULL,
_parent INTEGER NOT NULL,
nationality VARCHAR(255) NOT NULL,
FOREIGN KEY (_parent) REFERENCES person (_id) ON DELETE CASCADE,
PRIMARY KEY (_id)
);
CREATE TABLE person_variables(
_id serial NOT NULL,
_parent INTEGER NOT NULL,
variable VARCHAR(255) DEFAULT NULL,
value VARCHAR(255) DEFAULT NULL,
FOREIGN KEY (_parent) REFERENCES person (_id) ON DELETE CASCADE,
PRIMARY KEY (_id)
);
I'm trying to generate XML element for each person which also contains
the person nationality and variables in one result set
SELECT
XMLROOT (
XMLELEMENT (
NAME information,
XMLATTRIBUTES (
person._id AS pid
),
XMLAGG(
XMLELEMENT(
name "nationality",
person_nationality.nationality
)
),
XMLAGG(
XMLELEMENT(
name "value",
person_variables.value
)
)
),
VERSION '1.0',
STANDALONE YES
)
FROM person
INNER JOIN person_nationality ON person_nationality._parent = person._id
INNER JOIN person_variables ON person_variables._parent = person._id
GROUP BY person._id
LIMIT 100 OFFSET 10000;
The above query return number of variables * nationality for each
person which is expected (for me) because of the join logic. Also i
can't use DISTINCT keyword in XMLAGG function.
Any idea how to do this? or is there is any aggregate function returns
array from row set, i didn't find any in the documenation.
Thanks in advance
Mina.
--
Sent via pgsql-sql mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-sql
