SQL task: how to retrieve users with names in Cyrillic and Latin, sorted so that Cyrillic comes first, and within groups lexicographically?

To extract users with names in Cyrillic and Latin scripts, sorted so that Cyrillic comes first, and within groups sorted lexicographically, you can use SQL with a Unicode range check.

Example for PostgreSQL:

SELECT username
FROM users
ORDER BY
  CASE
    WHEN username ~ '^[\u0400-\u04FF]' THEN 0  -- Cyrillic
    ELSE 1  -- Latin and others
  END,
  username;

Explanation:

• The regular expression ^[\u0400-\u04FF] checks if the name starts with a character in the Cyrillic range.
• In ORDER BY, first sort by this condition (0 — Cyrillic, 1 — others), then lexicographically by the name.

If the database does not support Unicode in regex, you can use a function to check the first character by its code, for example:

ORDER BY
  CASE
    WHEN ascii(substring(username from 1 for 1)) BETWEEN 192 AND 255 THEN 0
    ELSE 1
  END,
  username

But this is less universal and depends on the encoding.