Question

I have following legacy database setup:

CREATE TABLE `categories` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `name` varchar(50) NOT NULL,
  PRIMARY KEY (`id`)
)

CREATE TABLE `items` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `name` varchar(50) NOT NULL,
  `category_ids` varchar(50) NOT NULL,
  PRIMARY KEY (`id`)
)

in which category_ids is a string of category id’s separated by comas: 1, 10, 15, 6. Is there a way to convert this database to more conventional one (using three tables, one for storing relationships), using only SQL and no other scripts?

Answer 1

MySQL doesn’t have CROSS APPLY or recursive CTEs, which would be the simplest routes.

But you’re only dong this once, so you only need a quick hack.

First, find out the maximum number of items in the category list…

SELECT
  MAX(LEN(category_ids) - LEN(REPLACE(category_ids, ',', '')) + 1) AS max_items
FROM
  items

Then you can do something like this…

SELECT
  items.id,
  SUBSTRING_INDEX(
    SUBSTRING_INDEX(
      items.category_ids,
      ',',
      map.id  -- Get the first 'n' items from the list
    ),
    ',',
    -1        -- Get the last item from (the first 'n' items from the list)
  ) AS category_id
FROM
  items
INNER JOIN
(
            SELECT 1 as id
  UNION ALL SELECT 2 as id
  UNION ALL SELECT 3 as id
  etc, etc, up to the max number of items found previously
)
  AS map
    ON LEN(items.category_ids) - LEN(REPLACE(items.category_ids, ',', '')) + 1 >= map.id

I haven’t tested it, but I’m assuming that SUBSTRING_INDEX('last', ',', -1) returns 'last'.

I’m no MySQL expert, so this may not be optimal, but as a one time quick win thistype of structure should work…