I see that typesense has locale-specific customizations for tokenization of field values. As per docs at https://typesense.org/docs/26.0/api/collections.html#schema-parameters , this is applicable only for ja,zh,ko,th,el,ru,sr,uk, and be. Is there a separate tokenizer in the works for ar(Arabic) ?