Skip to content

Anagram

Pick out the candidates that are anagrams of a word, ignoring case, where a word is not an anagram of itself.

Adapted from Exercism’s problem specification (MIT). Listed under Strings & Text.

// FILE: anagram.nomi
//# Picks the anagrams of a word out of a list of candidates.

/// The candidates that are anagrams of `subject`, in their original order.
/// Letter case does not matter, and a word is not an anagram of itself.
pub fn find(subject: String, candidates: List<String>) -> List<String>
    subject = String.fold_case(subject)
    letters = Iter.sort(subject)

    candidates
    |> Iter.filter(fn candidate ->
        candidate = String.fold_case(candidate)
        candidate != subject and Iter.sort(candidate) == letters
    end)
    |> Iter.to_list()
end

// FILE: anagram_test.nomi
import anagram

test "finds anagrams"
    [
        ("no matches", "diaper", ["hello", "world", "zombies", "pants"], []),
        ("detects two anagrams", "solemn", ["lemons", "cherry", "melons"], ["lemons", "melons"]),
        ("does not detect anagram subsets", "good", ["dog", "goody"], []),
        ("detects anagram", "listen", ["enlists", "google", "inlets", "banana"], ["inlets"]),
        (
            "detects three anagrams",
            "allergy",
            ["gallery", "ballerina", "regally", "clergy", "largely", "leading"],
            ["gallery", "regally", "largely"],
        ),
        (
            "detects multiple anagrams with different case",
            "nose",
            ["Eons", "ONES"],
            ["Eons", "ONES"],
        ),
        ("does not detect non-anagrams with identical checksum", "mass", ["last"], []),
        (
            "detects anagrams case-insensitively",
            "Orchestra",
            ["cashregister", "Carthorse", "radishes"],
            ["Carthorse"],
        ),
        (
            "detects anagrams using case-insensitive subject",
            "Orchestra",
            ["cashregister", "carthorse", "radishes"],
            ["carthorse"],
        ),
        (
            "detects anagrams using case-insensitive possible matches",
            "orchestra",
            ["cashregister", "Carthorse", "radishes"],
            ["Carthorse"],
        ),
        ("does not detect an anagram if the original word is repeated", "go",     ["goGoGO"], []),
        ("anagrams must use all letters exactly once",                  "tapper", ["patter"], []),
        ("words are not anagrams of themselves",                        "BANANA", ["BANANA"], []),
        (
            "words are not anagrams of themselves even if letter case is partially different",
            "BANANA",
            ["Banana"],
            [],
        ),
        (
            "words are not anagrams of themselves even if letter case is completely different",
            "BANANA",
            ["banana"],
            [],
        ),
        (
            "words other than themselves can be anagrams",
            "LISTEN",
            ["LISTEN", "Silent"],
            ["Silent"],
        ),
        ("handles case of greek letters", "ΑΒΓ", ["ΒΓΑ", "ΒΓΔ", "γβα", "αβγ"], ["ΒΓΑ", "γβα"]),
        ("different characters may have the same bytes", "a⬂", ["€a"], []),
    ]
    |> subtests (description, subject, candidates, expected), description
        assert anagram.find(subject, candidates) == expected
    end
end
  • The anagram key is Iter.sort(word): a String iterates by grapheme, so sorting it gives a List<String> to compare, with no codepoint conversion.
  • Case is ignored with String.fold_case on both sides, once for the subject and once per candidate. Folding is Unicode’s caseless matching, which String.lowercase only approximates: the two Greek sigmas ς and σ fold to one letter and lower-case to two.
  • All 18 current cases are one subtests table, since every row is the same call and comparison. The four cases that newer ones reimplement are left out.
  • Each row carries its canonical description as its first column and is named by it (subtests cases, (description, subject, candidates, expected), description), so a failing row reports under its description.