SQL: Complicated Select Statement and returned radius - mysql

Question

This query is driving me insane and I honestly just don't know how to accomplish it.
I need to grab all USERS and SHOPS longitude and latitude that are within a certain radius, and return back some other information attached to those resulting IDs... the radius part, I have figured out:
( 3959 * acos( cos( radians('48.453541') ) * cos( radians( lat ) ) * cos( radians( lng ) - radians('-123.491765') ) + sin( radians('48.453541') ) * sin( radians( lat ) ) ) ) AS distance
^This radians in the statement above are just a random address that I want to query against.
But grabbing results that essentially are two different rows AS ONE result confuses me right out. Both tables are constructed in this manner:
id
meta_key
meta_value
So I need to grab the "meta_value" of two keys for both a USER and a SHOP ... the "meta_key" for the USER is "bid_user_lat" and "bid_user_lng" ... and the "meta_key" for the SHOP are "bid_resource_lat" and "bid_resource_lng"
Ideal results (just showing one of each, but should be multiple, depending on radius):
id - 223
first_name - "Mark"
last_name - "Johnson"
bid_user_lat - 45.0000
bid_user_lng - -150.0000
id - 688
company_name - "Joe's Shop"
bid_resource_lat - 45.0000
bid_resource_lng - -150.0000
I really hope I'm explaining this correctly, because this is just driving me nuts!
My updated query is as follow, yet it's still not producing any results and it should:
SET #lat = '48.453541';
SET #lng = '-123.491765';
SET #radius = '10000';
SELECT
wp_usermeta.user_id,
bid_user_lat.meta_value,
bid_user_lng.meta_value,
( 3959 * acos( cos( radians( #lat ) ) * cos( radians( bid_user_lat.meta_value ) ) * cos( radians( bid_user_lng.meta_value ) - radians( #lng ) ) + sin( radians( #lat ) ) * sin( radians( bid_user_lat.meta_value ) ) ) ) AS distance
FROM
wp_usermeta
JOIN
wp_usermeta bid_user_lat
ON wp_usermeta.user_id = bid_user_lat.user_id
AND wp_usermeta.meta_key = "bid_user_lat"
JOIN
wp_usermeta bid_user_lng
ON wp_usermeta.user_id = bid_user_lng.user_id
AND wp_usermeta.meta_key = "bid_user_lng"
HAVING distance < #radius ORDER BY distance LIMIT 0 , 20;

Answer 1

You can get this information by using self-joins:
SELECT
table.id,
bid_user_lat.meta_value,
bid_user_lng.meta_value
FROM
table
JOIN
table bid_user_lat
ON table.id = bid_user_lat.id
AND table.meta_key = "bid_user_lat"
JOIN
table bid_user_lng
ON table.id = bid_user_lng.id
AND table.meta_key = "bid_user_lng"
WHERE
{ distance clause }
...the distance clause is where you ensure that it's within the proper distance. You can use bid_user_lat.meta_value as the bid_user_lat value, and bid_user_lng.meta_value as the bid_user_lng value in your calculation.
If you want to return the User results in the same query as the Shop results, then you can do the same query as the above one (using Shops instead of Users, of course), and connect them with a UNION or UNION ALL.
Side note:
If you have any control over the database, you may want to re-visit the design. The solution above is not ideal by any means... If a User always has a bid_user_lat and a bid_user_lng field, or even if it's just a common attribute of User, then it really should have its own column in the User table.
Correction:
I was JOINing incorrectly. The ON clause should use bid_user_lng.meta_key = "bid_user_lat" instead of wp_usermeta.meta_key = "bid_user_lat". You also need to add a GROUP BY clause so that you don't have duplicate records. Applying these to your query above, you get:
SET #lat = '48.453541';
SET #lng = '-123.491765';
SET #radius = '10000';
SELECT
wp_usermeta.user_id,
bid_user_lat.meta_value,
bid_user_lng.meta_value,
( 3959 * acos( cos( radians( #lat ) ) * cos( radians( bid_user_lat.meta_value ) ) * cos( radians( bid_user_lng.meta_value ) - radians( #lng ) ) + sin( radians( #lat ) ) * sin( radians( bid_user_lat.meta_value ) ) ) ) AS distance
FROM
wp_usermeta
JOIN
wp_usermeta bid_user_lat
ON wp_usermeta.user_id = bid_user_lat.user_id
AND bid_user_lat.meta_key = "bid_user_lat"
JOIN
wp_usermeta bid_user_lng
ON wp_usermeta.user_id = bid_user_lng.user_id
AND bid_user_lng.meta_key = "bid_user_lng"
GROUP BY wp_usermeta.user_id
HAVING distance < #radius ORDER BY distance LIMIT 0 , 20;
Here's confirmation that it works:
mysql> CREATE TABLE wp_usermeta (user_id INTEGER UNSIGNED, meta_key VARCHAR(255), meta_value VARCHAR(255));
Query OK, 0 rows affected (0.25 sec)
mysql> INSERT INTO wp_usermeta (user_id, meta_key, meta_value)
-> VALUES
-> (1, "bid_user_lat", "45.000"), (1, "bid_user_lng", "-150.000"),
-> (2, "bid_user_lat", "20.000"), (2, "bid_user_lng", "20.000"),
-> (3, "bid_user_lat", "-300.000"), (3, "bid_user_lng", "70.000");
Query OK, 6 rows affected (0.16 sec)
Records: 6 Duplicates: 0 Warnings: 0
mysql> SELECT
-> wp_usermeta.user_id,
-> bid_user_lat.meta_value,
-> bid_user_lng.meta_value,
->
-> ( 3959 * acos( cos( radians( #lat ) ) * cos( radians( bid_user_lat.meta_value ) ) * cos( radians( bid_user_lng.meta_value ) - radians( #lng ) ) + sin( radians( #lat ) ) * sin( radians( bid_user_lat.meta_value ) ) ) ) AS distance
->
-> FROM
-> wp_usermeta
-> JOIN
-> wp_usermeta bid_user_lat
-> ON wp_usermeta.user_id = bid_user_lat.user_id
-> AND bid_user_lat.meta_key = "bid_user_lat"
-> JOIN
-> wp_usermeta bid_user_lng
-> ON wp_usermeta.user_id = bid_user_lng.user_id
-> AND bid_user_lng.meta_key = "bid_user_lng"
->
-> GROUP BY wp_usermeta.user_id
->
-> HAVING distance < #radius ORDER BY distance LIMIT 0 , 20;
+---------+------------+------------+--------------------+
| user_id | meta_value | meta_value | distance |
+---------+------------+------------+--------------------+
| 1 | 45.000 | -150.000 | 1271.3329043047352 |
| 3 | -300.000 | 70.000 | 4905.4310014631055 |
| 2 | 20.000 | 20.000 | 7198.549954690863 |
+---------+------------+------------+--------------------+
3 rows in set (0.05 sec)