Is it possible to inverse the select statement in SQL? - mysql

Question

When I want to select all columns expect foo and bar, what I normally do is just explicitly list all the other columns in select statement.
select a, b, c, d, ... from ...
But if table has dozen columns, this is tedious process for simple means. What I would like to do instead, is something like the following pseudo statement:
select * except(foo, bar) from ...
I would also like to know, if there is a function to filter out rows from the result consisting of multiple columns, if multiple rows has same content in all corresponding columns. In other words duplicate rows would be filtered out.
------------------------
A | B | C
------------------------ ====> ------------------------
A | B | C A | B | C
------------------------ ------------------------

Answer 1

You can query INFORMATION_SCHEMA db and get the list of columns (except two) for that table, e.g.:
SELECT REPLACE(GROUP_CONCAT(COLUMN_NAME), '<foo,bar>,', '')
FROM INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_NAME = '<your_table>' AND TABLE_SCHEMA = '<database>';
Once you get the list of columns, you can use that in your select query.

Answer 2

You can create view based on this table with all columns except these two columns and then use this view everytime with
select * from view
simple group by on all column will remove such duplicates. there are other options as well - distinct and row_number.

Answer 3

select * except(foo, bar) from
This is a frequently requested feature on SO. However, it has not made it to the SQL Standard and I don't know of any SQL products that support it. I guess when the product managers ask their developers, MVPs, usergroups, etc to measure enthusiasm for this prospective feature, they mostly hear, "SELECT * FROM is considered dangerous, we need to protect new users who don't know what they are doing, etc."
You may find it useful to use NATURAL JOIN rather than INNER JOIN etc which removes what would be duplicated columns from the resulting table expression e.g.
SELECT *
FROM Table1 t1
INNER JOIN Table2 t2
ON t1.foo = t2.foo
AND t1.bar = t2.bar;
will result in two columns named foo and two named bar (and possibly other repeated names), probably de-duplicated in some way e.g. by suffixing the range variable names t1 and t2 that INNER JOIN forced you into using.
Whereas:
SELECT *
FROM Table1 NATURAL JOIN Table2;
doesn't require the use of range variables (a good thing) because there will only be one column named foo and one named bar in the result.
And to remove duplicated rows as well as columns changed the implied SELECT ALL * into the explicit SELECT DISTINCT * e.g.
SELECT DISTINCT *
FROM Table1 NATURAL JOIN Table2;
Doing this may reduce your need for the SELECT ALL BUT { these columns } feature you desire.
Of course, if you do that you will be told, "NATURAL JOIN is considered dangerous, we need to protect you from yourself in case you don't know what you are doing, etc." :)