问题
I have a following table structure.
USERS
PROPERTY_VALUE
PROPERTY_NAME
USER_PROPERTY_MAP
I am trying to retrieve user/s from the users table who have matching properties in property_value table.
A single user can have multiple properties. The example data here has 2 properties for user '1', but there can be more than 2. I want to use all those user properties in the WHERE clause.
This query works if user has a single property but it fails for more than 1 properties:
SELECT * FROM users u
INNER JOIN user_property_map upm ON u.id = upm.user_id
INNER JOIN property_value pv ON upm.property_value_id = pv.id
INNER JOIN property_name pn ON pv.property_name_id = pn.id
WHERE (pn.id = 1 AND pv.id IN (SELECT id FROM property_value WHERE value like '101')
AND pn.id = 2 AND pv.id IN (SELECT id FROM property_value WHERE value like '102')) and u.user_name = 'user1' and u.city = 'city1'
I understand since the query has pn.id = 1 AND pn.id = 2 it will always fail because pn.id can be either 1 or 2 but not both at the same time. So how can I re-write it to make it work for n number of properties?
In above example data there is only one user with id = 1 that has both matching properties used in the WHERE clause. The query should return a single record with all columns of the USERS table.
To clarify my requirements
I am working on an application that has a users list page on the UI listing all users in the system. This list has information like user id, user name, city etc. - all columns of the in USERS table. Users can have properties as detailed in the database model above.
The users list page also provides functionality to search users based on these properties. When searching for users with 2 properties, 'property1' and 'property2', the page should fetch and display only matching rows. Based on the test data above, only user '1' fits the bill.
A user with 4 properties including 'property1' and 'property2' qualifies. But a user with only one property 'property1' would be excluded due to the missing 'property2'.
回答1:
This is a case of relational-division. I added the tag.
Indexes
Assuming a PK or UNIQUE constraint on USER_PROPERTY_MAP(property_value_id, user_id) - columns in this order to make my queries fast. Related:
- Is a composite index also good for queries on the first field?
You should also have an index on PROPERTY_VALUE(value, property_name_id, id). Again, columns in this order. Add the the last column id only if you get index-only scans out of it.
For a given number of properties
There are many ways to solve it. This should be one of the simplest and fastest for exactly two properties:
SELECT u.*
FROM users u
JOIN user_property_map up1 ON up1.user_id = u.id
JOIN user_property_map up2 USING (user_id)
WHERE up1.property_value_id =
(SELECT id FROM property_value WHERE property_name_id = 1 AND value = '101')
AND up2.property_value_id =
(SELECT id FROM property_value WHERE property_name_id = 2 AND value = '102')
-- AND u.user_name = 'user1' -- more filters?
-- AND u.city = 'city1'
Not visiting table PROPERTY_NAME, since you seem to have resolved property names to IDs already, according to your example query. Else you could add a join to PROPERTY_NAME in each subquery.
We have assembled an arsenal of techniques under this related question:
- How to filter SQL results in a has-many-through relation
For an unknown number of properties
@Mike and @Valera have very useful queries in their respective answers. To make this even more dynamic:
WITH input(property_name_id, value) AS (
VALUES -- provide n rows with input parameters here
(1, '101')
, (2, '102')
-- more?
)
SELECT *
FROM users u
JOIN (
SELECT up.user_id AS id
FROM input
JOIN property_value pv USING (property_name_id, value)
JOIN user_property_map up ON up.property_value_id = pv.id
GROUP BY 1
HAVING count(*) = (SELECT count(*) FROM input)
) sub USING (id);
Only add / remove rows from the VALUES expression. Or remove the WITH clause and the JOIN for no property filters at all.
The problem with this class of queries (counting all partial matches) is performance. My first query is less dynamic, but typically considerably faster. (Just test with EXPLAIN ANALYZE.) Especially for bigger tables and a growing number of properties.
Best of both worlds?
This solution with a recursive CTE should be a good compromise: fast and dynamic:
WITH RECURSIVE input AS (
SELECT count(*) OVER () AS ct
, row_number() OVER () AS rn
, *
FROM (
VALUES -- provide n rows with input parameters here
(1, '101')
, (2, '102')
-- more?
) i (property_name_id, value)
)
, rcte AS (
SELECT i.ct, i.rn, up.user_id AS id
FROM input i
JOIN property_value pv USING (property_name_id, value)
JOIN user_property_map up ON up.property_value_id = pv.id
WHERE i.rn = 1
UNION ALL
SELECT i.ct, i.rn, up.user_id
FROM rcte r
JOIN input i ON i.rn = r.rn + 1
JOIN property_value pv USING (property_name_id, value)
JOIN user_property_map up ON up.property_value_id = pv.id
AND up.user_id = r.id
)
SELECT u.*
FROM rcte r
JOIN users u USING (id)
WHERE r.ct = r.rn; -- has all matches
dbfiddle here
The manual about recursive CTEs.
The added complexity does not pay for small tables where the additional overhead outweighs any benefit or the difference is negligible to begin with. But it scales much better and is increasingly superior to "counting" techniques with growing tables and a growing number of property filters.
Counting techniques have to visit all rows in user_property_map for all given property filters, while this query (as well as the 1st query) can eliminate irrelevant users early.
Optimizing performance
With current table statistics (reasonable settings, autovacuum running), Postgres has knowledge about "most common values" in each column and will reorder joins in the 1st query to evaluate the most selective property filters first (or at least not the least selective ones). Up to a certain limit: join_collapse_limit. Related:
- Postgresql join_collapse_limit and time for query planning
- Why does a slight change in the search term slow down the query so much?
This "deus-ex-machina" intervention is not possible with the 3rd query (recursive CTE). To help performance (possibly a lot) you have to place more selective filters first yourself. But even with the worst-case ordering it will still outperform counting queries.
Related:
- Check statistics targets in PostgreSQL
Much more gory details:
- PostgreSQL partial index unused when created on a table with existing data
More explanation in the manual:
- Statistics Used by the Planner
回答2:
SELECT *
FROM users u
WHERE u.id IN(
select m.user_id
from property_value v
join USER_PROPERTY_MAP m
on v.id=m.property_value_id
where (v.property_name_id, v.value) in( (1, '101'), (2, '102') )
group by m.user_id
having count(*)=2
)
OR
SELECT u.id
FROM users u
INNER JOIN user_property_map upm ON u.id = upm.user_id
INNER JOIN property_value pv ON upm.property_value_id = pv.id
WHERE (pv.property_name_id=1 and pv.value='101')
OR (pv.property_name_id=2 and pv.value='102')
GROUP BY u.id
HAVING count(*)=2
No property_name table needed in query if propery_name_id are kown.
回答3:
If you want just to filter:
SELECT users.*
FROM users
where (
select count(*)
from user_property_map
left join property_value on user_property_map.property_value_id = property_value.id
left join property_name on property_value.property_name_id = property_name.id
where user_property_map.user_id = users.id -- join with users table
and (property_name.name, property_value.value) in (
values ('property1', '101'), ('property2', '102') -- filter properties by name and value
)
) = 2 -- number of properties you filter by
Or, if you need users ordered descending by number of matches, you could do:
select * from (
SELECT users.*, (
select count(*) as property_matches
from user_property_map
left join property_value on user_property_map.property_value_id = property_value.id
left join property_name on property_value.property_name_id = property_name.id
where user_property_map.user_id = users.id -- join with users table
and (property_name.name, property_value.value) in (
values ('property1', '101'), ('property2', '102') -- filter properties by name and value
)
)
FROM users
) t
order by property_matches desc
回答4:
SELECT * FROM users u
INNER JOIN user_property_map upm ON u.id = upm.user_id
INNER JOIN property_value pv ON upm.property_value_id = pv.id
INNER JOIN property_name pn ON pv.property_name_id = pn.id
WHERE (pn.id = 1 AND pv.id IN (SELECT id FROM property_value WHERE value
like '101') )
OR ( pn.id = 2 AND pv.id IN (SELECT id FROM property_value WHERE value like
'102'))
OR (...)
OR (...)
You can't do AND because there is no such a case where id is 1 and 2 for the SAME ROW, you specify the where condition for each row!
If you run a simple test, like
SELECT * FROM users where id=1 and id=2
you will get 0 results. To achieve that use
id in (1,2)
or
id=1 or id=2
That query can be optimised more but this is a good start I hope.
回答5:
you are using AND operator between two pn.id=1 and pn.id=2. then how you getting the answer is between that:
(SELECT id FROM property_value WHERE value like '101') and
(SELECT id FROM property_value WHERE value like '102')
So like above comments , Use or operator.
Update 1:
SELECT * FROM users u
INNER JOIN user_property_map upm ON u.id = upm.user_id
INNER JOIN property_value pv ON upm.property_value_id = pv.id
INNER JOIN property_name pn ON pv.property_name_id = pn.id
WHERE pn.id in (1,2) AND pv.id IN (SELECT id FROM property_value WHERE value like '101' or value like '102');
回答6:
If you just want the distinct columns in U, it is:
SELECT DISTINCT u.*
FROM Users u INNER JOIN USER_PROPERTY_MAP upm ON u.id = upm.[user_id]
INNER JOIN PROPERTY_VALUE pv ON upm.property_value_id = pv.id
INNER JOIN PROPERTY_NAME pn ON pv.property_name_id = pn.id
WHERE (pn.id = 1 AND pv.[value] = '101')
OR (pn.id = 2 AND pv.[value] = '102')
Notice I used pv.[value] = instead of the subquery to reacquire id... this is simplification.
回答7:
If I understand your question correctly I would do it like this.
SELECT u.id, u.user_name, u.city FROM users u
WHERE (SELECT count(*) FROM property_value v, user_property_map m
WHERE m.user_id = u.id AND m.property_value_id = v.id AND v.value IN ('101', '102')) = 2
This should return a list of users that have all the properties listed in the IN clause. The 2 represents the number of properties searched for.
回答8:
Assuming you want to select all the fields in the USERS table
SELECT u.*
FROM USERS u
INNER JOIN
(
SELECT USERS.id as user_id, COUNT(*) as matching_property_count
FROM USERS
INNER JOIN (
SELECT m.user_id, n.name as property_name, v.value
FROM PROPERTY_NAME n
INNER JOIN PROPERTY_VALUE v ON n.id = v.property_name_id
INNER JOIN USER_PROPERTY_MAP m ON m.property_value_id = v.property_value_id
WHERE (n.id = @property_id_1 AND v.value = @property_value_1) -- Property Condition 1
OR (n.id = @property_id_2 AND v.value = @property_value_2) -- Property Condition 2
OR (n.id = @property_id_3 AND v.value = @property_value_3) -- Property Condition 3
OR (n.id = @property_id_N AND v.value = @property_value_N) -- Property Condition N
) USER_PROPERTIES ON USER_PROPERTIES.user_id = USERS.id
GROUP BY USERS.id
HAVING COUNT(*) = N --N = the number of Property Condition in the WHERE clause
-- Note :
-- Use HAVING COUNT(*) = N if property matches will be "MUST MATCH ALL"
-- Use HAVING COUNT(*) > 0 if property matches will be "MUST MATCH AT LEAST ONE"
) USER_MATCHING_PROPERTY_COUNT ON u.id = USER_MATCHING_PROPERTY_COUNT.user_id
来源：https://stackoverflow.com/questions/47351766/using-same-column-multiple-times-in-where-clause