Sphinx does not normalize words. Why?

Question

Sphinx

index featurevalue {
  source = featurevaluesource
  path = /var/lib/sphinxsearch/data/featurevalue
  charset_type = utf-8
  rt_field = value
  expand_keywords = 1
  #enable_star    = 1
        min_prefix_len = 1
        min_word_len = 2
#	morphology = stem_enru
  morphology = lemmatize_ru_all
}

Here is the code for my sphinx config. Whether I check words are normalized and I am convinced that is not present.

mysql> call keywords('девушка девушек девушки', 'featurevalue' );
+----------------+----------------+
| tokenized      | normalized     |
+----------------+----------------+
| девушка        | девушка        |
| девушек        | девушек        |
| девушки        | девушки        |
+----------------+----------------+
3 rows in set (0.00 sec)

Same result with morphology = stem_enru enabled. What should be done to normalize? In the tuning examples, it turns out that it normalizes from the very beginning.
PS I don't know if this is important or not, but my index does not contain the words girl girl and so on. Checking the normalization from the tuning examples.

Answer 1

what are the examples?