feedback: Data storage | Tarantool

<…>s and use cases, use the “unicode” and “unicode_ci” collations.
|If you need Cyrillic letters ‘Е’ and ‘Ё’ to have the same level-1 weights,
try the Kyrgyz collation.|

https://www.tarantool.io/en/doc/latest/platform/ddl_dml/value_store/#collations

It seems you meant the opposite.

https://www.unicode.org/Public/17.0.0/uca/allkeys.txt

0435 ; [.282E.0020.0002] # CYRILLIC SMALL LETTER IE
0451 ; [.282E.0020.0002][.0000.002B.0002] # CYRILLIC SMALL LETTER IO

The primary weights are the same for these letters in the DUCET.

Kyrgyz collation, however, defines different primary weights for them.

https://github.com/unicode-org/cldr/blob/release-27/common/collation/ky.xml

&е<ё<<<Ё

(е<ё means exactly that the primary weight of е is less than the primary weight of ё, see the notation in the Table 14 in the Unicode Collation Algorithm.)

NB: There was an old issue regarding it, it may bring some context: https://github.com/tarantool/tarantool/issues/4007

Start at the Collations section of the Tarantool documentation page linked in the issue and compare its Kyrgyz-collation statement with Unicode UCA and CLDR references. Update the misleading wording so it accurately describes the primary weights of Cyrillic Е and Ё, then verify the rendered documentation page.