JapaneseStemmer
- Namespace
- Rowles.LeanCorpus.Analysis.Stemmers
- Assembly
- Rowles.LeanCorpus.dll
Japanese stemmer — identity implementation.
public sealed class JapaneseStemmer : ISpanStemmer- Inheritance
-
JapaneseStemmer
- Implements
Remarks
Japanese morphology is handled by conjugation paradigms (活用), not detachable suffixes of the kind that a Snowball-style stemmer can reliably strip. Inflected verb and adjective forms are interleaved with auxiliary morphemes (e.g. 食べられる, 走っている) that require a full morphological analyser to decompose correctly.
Recommended pre-processing for Japanese search:
- Morphological analysis with MeCab, Kuromoji, or SudachiPy
- Lemmatisation using the analyser's dictionary base-form output
- Kana normalisation (hiragana ↔ katakana, full-width → half-width)
This class is provided so the ISpanStemmer pipeline compiles uniformly
across all supported languages.
Stem(ReadOnlySpan<char>, Span<char>)
Stems word into output.
Returns the length of the stemmed result, or -1 if output is too small.
The caller must ensure output.Length >= word.Length for guaranteed success.
Stem(string)
Convenience overload returning the stemmed string.