Is there any significant performance penalty for using
case_insensitive: true in term queries?
If I use keyword mapping with lowercase normalizer, highlighter would emit lowercased results which I would like to avoid. I can index it as
keyword tokenizer and
lowercase filter but the nature of the field is keyword and I would like to keep it as such while being able to match case insensitive preserve case in highlighter.
case_insensitive: true would take care of all this but I could not find any info on how it is implemented and if it has any considerable performance penalty