Controlling unique values, not a key

There is a table with the key “uniqueid (” UUID “)”.
This table has two text columns (not a key), the contents of which must be unique within this table.
What is the best way to do the uniqueness check at save time?

Try:

ISBLANK(
  FILTER(
    "this-table",
    AND(
      ([this-column] = [_THISROW].[this-column]),
      NOT([_ROWNUMBER] = [_THISROW].[_ROWNUMBER])
    )
  )
)

with appropriate replacements for this-table and this-column.

1 Like

Steve, thanks!
I’ve tried various options with select () + [uuid] <> [_ THISROW]. [Uuid], but I didn’t get it.
What are the fundamental differences between select () and filter ()? By returning unique values? Those. when I use select () do not duplicate values return?

FAQ: FILTER(), LOOKUP(), MAXROW(), MINROW(), REF_ROWS(), and SELECT() Tips & Tricks ?

How are FILTER(), LOOKUP(), MAXROW(), MINROW(), REF_ROWS(), and SELECT() related? FILTER(), LOOKUP(), MAXROW(), and MINROW() are all built upon SELECT(), so they all share similar capabilities and limitations. FILTER(“table”, select-expression) is equivalent to: SELECT( table[key-column], (select-expression) ) LOOKUP(match-expression, “table”, “match-column”, “result-column”) is equivalent to: ANY( SELECT( table[result-column], ([match-column] = (match-expression)) ) ) N…

2 Likes