在 MySQL 查询中插入来自多行的连接值作为新行

Question

I currently have a table with columns for "field_type" (username, city, state), "user_id", and "value". The user_id column obviously has lots of repeats. I'd like to merge the city and state data into a single "location" field_type value. I need something that will:

for every integer in the user_id column: -check if there exist corresponding (not null) table rows for field_type "city" and "state" -if yes, insert a new row into the table with field_type "location" which concatenates the corresponding city and state values for that user_id

I haven't worked with MySQL much so I don't really know where to start. I've tried to simplify the problem a bit - it's actually a somewhat more complicated wordpress table and I'm trying to reformat the data to be compatible with a new plugin, but this covers the basics of what has to happen so I should hopefully be able to extrapolate an actual solution from the answers. Thanks for any pointers!

Edit: Current structure looks like this: |-id (key)-|- field_type -|- user_id -|- value -| | 1 | username | 1 | Joe | | 2 | city | 1 | Albany | | 3 | state | 1 | NY | | 4 | username | 2 | Bob | | 5 | city | 2 | Toledo | | 6 | state | 2 | OH | |-id (key)-|- field_type -|- user_id -|- value -| | 1 | username | 1 | Joe | | 2 | city | 1 | Albany | | 3 | state | 1 | NY | | 4 | username | 2 | Bob | | 5 | city | 2 | Toledo | | 6 | state | 2 | OH |

And I would like to get something like this: |-id (key)-|- field_type -|- user_id -|- value ---------| | 1 | username | 1 | Joe | | 2 | city | 1 | Albany | | 3 | state | 1 | NY | | 4 | username | 2 | Bob | | 5 | city | 2 | Toledo | | 6 | state | 2 | OH | | 7 | location | 1 | Albany, NY | | 8 | location | 2 | Toledo, OH | |-id (key)-|- field_type -|- user_id -|- value ---------| | 1 | username | 1 | Joe | | 2 | city | 1 | Albany | | 3 | state | 1 | NY | | 4 | username | 2 | Bob | | 5 | city | 2 | Toledo | | 6 | state | 2 | OH | | 7 | location | 1 | Albany, NY | | 8 | location | 2 | Toledo, OH |

Duplicate user_id values are how it's supposed to work, so they don't need to be removed.

Answer 1

You could use an INSERT... SELECT query to do this:

INSERT INTO yourtable
SELECT 'location' AS field_type, t1.user_id, CONCAT(t1.value, ' ', t2.value) AS value
FROM yourtable t1
JOIN yourtable t2 ON t2.user_id = t1.user_id AND t1.field_type = 'city' AND t2.field_type = 'state'
WHERE t1.value IS NOT NULL AND t2.value IS NOT NULL

Demo on dbfiddle

Answer 2

Since you are mentioned that there are duplicates with the same user_id. I don't think inserting a new row will be a good idea. So have written an update query to update the existing data. You can later cleanup the duplciates.

UPDATE your_table SET locations = CONCAT_WS('-', state, city) where city is not null or state is not null