I know "normalizer doesn't like trim filter", but I've also found out I can't use length filter either.
Is there any way of knowing in advance which filters are accepted in a normalizer definition?
What I try to accomplish with the length filter is to avoid empty string values. I did that with the "length" filter with this:
only tokens of at least 1 character would go through.
I'm trying to accomplish the same using a normalizer, but it doesn't like "length" filter.