Hi,
I am uncertain if that was ever the case but I believe not since I catered for missing using toEmpty() before. If I recall correctly there i a major regression since the presence of a missing cell will always cause a missing result given back.
I also noticed that in the String Replace note but shove it off due to time constrains. If anyone agrees this being a bug / regression, I will upload the test workflow and extend it by the string replacer example I mentioned.
Best
Mike
Hi @mwiegand ,
I checked in 4.7.8. and 5.3.2 – the behavior is the same.
If there is a missing cell in the expression, the result is missing as well. You can let missing cells in the input be converted to null, which does not result in the “short-circuiting” behavior, by using the checkbox at “Insert Missing as Null”.
How nulls are handled depends on each function. For example, join already converts null to the empty string. toEmpty also talks about null and not missing. I would assume the expression is not even evaluated if an input cell is missing, unless the checkbox is checked.
Best,
Manuel
Thanks for checking @hotzm. But shouldn’t the result, regardless of null or missing, yield the same results … ideally with a string and in worst case, like division by zero, missing (as it is).
Though, I personally feel that getting a missing result should be considered a bug when converting missing to null results in one by comparison.
