I'm trying to figure out how to handle stemming for a lot of different languages.
I can only use a single index because of a parent-child join at query time.
I would like to stick to a single language per document as recommended on https://www.elastic.co/guide/en/elasticsearch/guide/current/one-lang-docs.html. I don't want to add all language analyzers to a multi field because that would imply too much analyzing overhead.
Isn't there some kind of solution to base the analyzer on the document language?
e.g. set a default analyzer on the document and reference that analyzer from the field.
What do you guys think?