Reputation: 2228
I have 7 tables db_category
, db_city
, db_locality
, db_shop
, db_shop_locality
, db_shop_recommended
, db_shop_views
.
db_category
category_id category_name publish
1 Apparel 1
db_city
city_id city_name publish
1 bangalore 1
db_locality
locality_id locality_name publish
1 kalyan nagar 1
2 Madiwala 1
db_shop
shop_id category_id city_id locality_id shop_name publish
1 1 1 1 name 1
2 1 1 2 name1 1
3 1 1 1 name2 1
4 1 1 1 name3 1
db_shop_recommended
recommended_id category_id shop_id priority publish
1 1 1 1 1
2 1 2 2 1
db_shop_views
views_id shop_id ip_addr publish
1 1 127.0.0.1 1
2 2 ::1 1
3 4 127.0.0.1 1
4 4 ::1 1
5 3 ::1 1
I want to join all the above tables. And the conditions for the join are,
display all the rows from db_shop
when publish=1
and join db_category
, db_city
and db_locality
with db_shop
in db_shop_recommended
who's priority=1
comes first and so on.
and shop_id
not in db_shop_recommended
, count shop_id
not in db_shop_recommended
from db_shop_views
who's shop_id
got more views.
My result should looks like below,
shop_id category_name city_name locality_name shop_name
1 Apparel bangalore kalyan nagar name (result based on `db_shop_recommended` who's priority is more)
2 Apparel bangalore Madiwala name1 (result based on `db_shop_recommended` who's priority is more)
4 Apparel bangalore kalyan nagar name4 (result based on `db_shop_views` who's view count is more)
3 Apparel bangalore kalyan nagar name3 (result based on `db_shop_views`)
I have no idea how to join and count the views. Is it possible to join in above method.
Upvotes: 1
Views: 42
Reputation: 2228
After lots of searching. Finally i got the answer.
SELECT DISTINCT s.shop_id, c.city_name, l.locality_name, ca.category_name, s.shop_name, s.shop_logo, s.cart_url, s.shop_about FROM db_shop s INNER JOIN db_city c ON c.city_id = s.city_id INNER JOIN db_category ca ON ca.category_id = s.category_id INNER JOIN db_locality l ON l.locality_id = s.locality_id LEFT JOIN (SELECT * FROM db_shop_recommended r ORDER BY r.priority ASC) r1 ON s.shop_id = r1.shop_id LEFT JOIN (select distinct g.shop_id, g.cnt from (select distinct shop_id, count(shop_id) cnt from db_shop_views group by shop_id) g inner join (select max(s.cnt) max_cnt from(select distinct shop_id, count(shop_id) cnt from db_shop_views group by shop_id) s) m on m.max_cnt = g.cnt) v ON s.shop_id = v.shop_id WHERE s.publish = 1
Upvotes: 1