Almost, but not quite, entirely unlike tea.

Matches patterns against a dictionary of words and phrases mined from Russian Wikipedia. Text is normalized to lowercase letters (Latin and Cyrillic), numbers and spaces. More common results are returned first. Input is UTF-8.

Syntax

• a-z, а-я, 0-9, space - literal match (Latin, Cyrillic, digits, space)
• [], (), {}, |, ., ?, *, + - same as regexp
• "expr" - forbid word breaks without a space or hyphen
• expr&expr - both expressions must match (intersection)
• <aaagmnr>, <(gram)(ana)> - anagram of contents (note warnings)
• _ (underscore) - one letter/digit: [a-zа-я0-9] (UTF-8)
• # (number sign) - digit: [0-9]
• A - Latin letter: [a-z]
• C - Latin consonant (including y)
• V - Latin vowel ([aeiou], not y)
• R - Cyrillic letter: [а-я]
• S - Cyrillic consonant
• G - Cyrillic vowel (а, е, и, о, у, ы, э, ю, я)

Examples

• "SGSGSGS" - seven letters, alternating Cyrillic consonant / vowel (quoted = one word)
• <ватерполистка> - anagram
• 13п_*у - numbers and letters
• "_ ____ю __ ______" - partially filled mask
• "CC RRR" - Latin + Cyrillic

Original versions

• nutrimatic.org/2016 - “Classic” public Nutrimatic; dictionary from the English Wikipedia
• nutrimatic.org/2024 - Current public Nutrimatic; refreshed English‑Wikipedia index

More

• Usage guide (original English‑Wikipedia Nutrimatic on nutrimatic.org): usage tips, worked examples, why it's slow.
• Source code for this Russian Wikipedia fork on GitHub: not completely documented, but it's there!