Lodestar.Text 0.5.0. 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.

GermanSnowballStemmer.Stem

The Snowball stem of one German word.

public static string Stem(string word)

Parameters — word is a single German word. It is lowercased and NFC-normalised before the rules run, which matters more here than elsewhere: German nouns are capitalised in ordinary text, so most input arrives needing it.

Returns — string, the stem, always lowercase; the final step folds umlauts to their base vowel.

Exceptions — ArgumentNullException when word is null. An empty string, or a word of one character, is returned lowercased and otherwise untouched.

Example — the two spellings of ß meeting, and an umlaut removed on the way out.

using Lodestar.Text.Stemming;

string sharp = GermanSnowballStemmer.Stem("straße");  // => strass
string spelled = GermanSnowballStemmer.Stem("strasse");  // => strass
string umlaut = GermanSnowballStemmer.Stem("größe");  // => gross

Remarks — größe → gross is the result most likely to look like a bug. It is two substitutions, both deliberate: ß → ss before the rules, ö → o after them. The stem is a key, and it is the same key grosse produces, which is the whole intent.

Because umlauts are stripped last rather than first, a word whose suffix rule depends on ä or ü still sees it while the decision is being made. Folding them up front would change which suffixes match, and give different stems.

Applies to — net10.0, netstandard2.0.

See also — GermanSnowballStemmer, the stemming index.