- Query optimization:
- Level 0: default level, nothing is optimized
- Level 1: facets can be optimized
- Level 2: facets AND fulltext can be optimized
- AFS now supports Chinese language:
- Use of ICU tokenizer to split Chinese characters
- Allows pinyin queries without using accentuation
- Translitteration between simplified and traditional characters
- Numbers written in Chinese characters return digits.
- Performance of exact queries containing empty words has been improved on large corpus.
- The compound splitter has been adapted to split Japanese language (katakana).