Question

This is the code outline I was thinking of using:

CREATE function get_n_chars ( dfrom varchar(9) ) returns varchar(9)
READS SQL DATA
begin
declare d varchar(9);
dfrom REGEXP '^.(...)';
d = $1;
limit 1;
return d;

I was not surprised to get an error, does anyone know the correct syntax, or of this is even possible?

Thanks

Answer 1

Per your comment above:

The basic problem is that a have a table that has company names that vary slightly but are the same company. Example: ‘Apple Inc.’, ‘Apple Computers’, and Apple Incorporate’. I am supposed to clean this table up. I need to find duplicates based on similar names. So all three ‘Apples’ are ‘duplicates’, after identifying I need to delete the two rows with most incomplete data. So, I need to display these three rows with their id numbers then delete two.

To retrieve all suspected duplicates, how about something like:

SELECT my_table.* FROM my_table JOIN (
  SELECT   LEFT(company_name, 4) AS abbr
  FROM     my_table
  GROUP BY abbr
  HAVING   COUNT(*) > 1
) t ON LEFT(my_table.company_name, 4) = t.abbr
ORDER BY my_table.company_name