Hi @Vernalis,
sometimes the column type gets lost, especially when using table reader / writer nodes during loop execution to store intermittent results. Though, chances are a collection is passed as a string in a CSV or any other file format.
Converting back to a collection type requires a few steps:
I’d like to suggest to clone the Collection to String node into a String to Collection node to allow, by configuring these options, to convert back into a collection with ease.
PS: Just for reference if anyone finds that topic. Here is a function for the column expression node that does all of it in one swipe:
split(replace(replaceChars(column(variable("currentColumnName")), "[]", ""), ", ", ","), ",")
Paired with a Column List Loop it is an unintuitive but usable approach:
Best
Mike