Has anyone developed a Latin stemmer-analyser (e.g. for Elasticsearch)? Learner of Latin and Ancient Greek here.