I was just made aware of this topic, so sorry for the long delay.
The node description lists all available operators, regardless of input type. But the operators that can actually be used depend on the input column type. I think the node description should make this point clearer and explain, which operators only apply to which data types.
Based on your filter value, I believe SummaryStatistic is an XML column. In order to filter it with a pattern in 5.3.0-5.3.1, you must first convert it to a String column. We will add support to filter XML (and other “string-compatible” column types) with patterns in 5.4 and also in 5.3.2, due to forum feedback such as yours (internal ticket AP-22915).
Inclusion/exclusion by Row Number and Row ID is supported as special columns (the first two in the dropdown).
Range checking must be done by using two conditions, joined by “AND”, but we will add a “BETWEEN” operator in the future (internal ticket UIEXT-2006).
Could you clarify a bit what you mean in the third screenshot? That the new Row Filter does not offer to filter based on collection elements?
Yeap, the new row filter does not offer this option as the older one, where I used as a list/lib with keywords to match by values.
All information that you passed I could see as you said, selecting a column and then, options allowed for it. It’s not clear as soon now from the information tab, I know that was my mistake, but could show aout column type too to be more clear how to use it latter.
Ah, yes… I think we need to make the first two choices (RowID and Row Number) more visually distinct. I believe @DanielBog is aware of this already, since we got the feedback in another thread.
yes we are aware of this and want to improve that in the future. Furthermore, we plan to re-introduce the type indication to make it more clear what column has which type.
Hi @denisfi , I’m just adding reference here to the following similar thread which may be of interest, where features of this node and ticket were also discussed.