I've set up a taxonomy tree as nested set with left and right integer columns. If I've set it up correctly, the contents of either left or right should never be repeated in any other left or right column.

Is it possible to set up a single unique index for the contents of the two separate columns?

For example, given the valid (truncated and modified) table ...
Code:
> select id, lft, rgt from nodes limit 5; 
+----+------+------+ 
| id | lft  | rgt  | 
+----+------+------+ 
|  1 |    1 |  300 | 
|  2 |   11 |  200 | 
|  6 |   50 |   51 | 
|  7 |  137 |  138 | 
|  9 |  189 |  190 | 
+----+------+------+
The following command would error as 137 is the lft value of id=7.
Code:
 
UPDATE nodes SET rgt=137 WHERE id=6;
For additional clarification, I am NOT asking for an index of a combination of both columns like the following.
CREATE UNIQUE INDEX left_and_right ON nodes (left,right)