Question

I’m working on a webapp with a set of data that users can browse and edit.

Users want to customize one of the fields that appear for each item, say “product_name”.

Each product has a set of default values in a table, which is the same for all users:

The product table would be like (skipping syntax cruft, you get the idea):

CREATE TABLE products (
  pid     int,            # product id
  pname   varchar         # product name (users want to customize this)
);

I need to display the user’s customized product name in all tables/lists on my site whenever there is a customized value for that product, but otherwise display the default value.

For space efficiency I thought of using an extra table that sort of “overrides” the default values, for any given user, for any given product. The rows are added in this table as needed. A user that did not customize any product name would not have any rows in there:

CREATE TABLE custom (
  pid     int,
  userid  int,
  pname   varchar
);

I can easily join these two tables on product id, and could pick the default (products.pname) , or user-edited value (custom.pname) for each row in php, but I’m hoping there is a more efficient way to do this in MySQL.

Is there an expression that lets me pick one column if not null otherwise pick another (accross two tables) ?

Would you suggest a different way to handle this ?

Thank you.

PS: maybe a type of JOIN that would allow a column of table 2 with the same name to override a column in table 1 for those rows where it exists in table 2 ?

Answer 1

You are looking for the COALESCE function.

SELECT COALESCE(the_field, "This literal") FROM YOUR_TABLE

results in the_field if the_field is not null, otherwise it is “This literal”. You can mix fields and literals as much as you want .. so SELECT COALESCE(this_field, that_field) works too. You can have more than 2 parameters … e.g. SELECT COALESCE(this_field,that_field, theother_field)