Text 0.6.0 doublemetaphone encode

Lodestar.Text 0.6.0. This page is frozen at that release. Read the current documentation for what main says now. A link to a decision or a migration page follows main, and leaves the archive.

DoubleMetaphone.Encode

The two Double Metaphone codes of one word.

public static DoubleMetaphoneCode Encode(ReadOnlySpan<char> value)
public static DoubleMetaphoneCode Encode(string value)

Parameters — value is a single word. Only its ASCII letters are read; case does not matter, and everything else — accents, non-Latin scripts, digits, punctuation, spacing — is dropped rather than rejected. The string overload forwards to the span one.

Returns — DoubleMetaphoneCode, a pair of variable-length codes drawn from A F H J K L M N P R S T X 0. Primary is the empty string when value holds no ASCII letter; Secondary is the empty string when the word has no alternate pronunciation.

Exceptions — ArgumentNullException when value is null (the string overload only; a ReadOnlySpan<char> cannot be null). An empty string is accepted and encodes to two empty codes.

Example — one reading, then two.

using Lodestar.Text.Phonetics;

DoubleMetaphoneCode thomas = DoubleMetaphone.Encode("Thomas");
string only = thomas.Primary;  // => TMS
string none = thomas.Secondary;  // =>

DoubleMetaphoneCode knuth = DoubleMetaphone.Encode("Knuth");
string th = knuth.Primary;  // => N0
string t = knuth.Secondary;  // => NT

Remarks — match on either code. Two words are phonetic candidates when any of their codes agree, so the test is four comparisons and not one; taking Primary alone throws away the reason to use this encoder over Metaphone.

Knuth above is the ordinary shape of an alternate: TH is 0 to an English ear and a plain T to a German one, and the encoder declines to choose. Wright encodes to RT with no alternate — the silent W is not a disagreement, just English spelling, which this encoder models the same way Metaphone.Encode does.

Accents are dropped, not folded. élan encodes as lan does, to LN, because the reference keeps ASCII letters only. That is a real limitation on non-English names rather than a rounding of one, and it is pinned by the corpus rather than asserted here, so what the reference does stays the reference's to say.

Applies to — net10.0, netstandard2.0.

See also — DoubleMetaphone, DoubleMetaphoneCode, Metaphone.Encode, the phonetics index.