I converted a csv containing customer information to a sqlite table, and it needed breaking down to several tables. Specifically, some columns where really about the employer, so it was better to normalize the employers in the table out and bring these columns along.

To normalize the Employer column I used sqlite-utils, a cool library that makes working with sqlite nicer. Extracting can be done with

sqlite-utils extract my.db customers Employer

which create a new table with unique values etc.

How can we bring Employer_State along? I used a relatively new (2020) sqlite feature called UPDATE FROM which allows you to update table values based on a different table.

UPDATE employers
SET STATE = customers.EMPLOYER_STATE
FROM customers
WHERE employers.id = customers.employers_id
  AND conditions...

Fairly straightforward. We update the employers table's column STATE with the column from customers. The WHERE clause is responsible for the join. You can also add other conditions to prevent updating some rows, or handle multiple values etc.