Is there an easy way to compare two null columns?

Question

I need to compare a significant number of columns between records in two tables for a data conversion. I really need a comparison that will return true if: 1) both columns are null; or 2) both columns are not null and equal. I want it to return false if: 3) one column is null and the other is not; or 4) both columns are not null and are not equal. I am trying to find records which are not exact matches.
I found documentation at oracle-base.com about the SYS_OP_MAP_NONNULL function that would do what I want, but I don't want to use this since it's an undocumented feature and my code will be in production for a period of time.
I would rather not have to use a construct like this for each and every column I'm comparing:
          a.col is null
          and b.col is null
     or (
          a.col = b.col
)Also, I know about the NVL function, but I'm comparing columns which are entered by users, and I'm not comfortable substituting any values for null because those values might actually exist in the data.

Answer 1

Performance wasn't the issue but they will be about the same anyway. The issue was avoiding the messy syntax needed to do the job.
Which of these looks easier to you if you had to understand and maintain the code or add new columns
>
I would rather not have to use a construct like this for each and every column I'm comparing:
          a.col is null
          and b.col is null
     or (
          a.col = b.col
>
or
1 = DECODE (a.col, b.col, 1)So if, like the OP, you had to 'compare a significant number of columns' which syntax would you use?