KStemmer

Namespace

Rowles.LeanCorpus.Analysis.Stemmers

Assembly

Rowles.LeanCorpus.dll

Lexicon-validated English stemmer inspired by Krovetz stemming.

public sealed class KStemmer : ISpanStemmer

KStemmer

Implements

ISpanStemmer

Constructors

KStemmer(IKStemLexicon)

Initialises a new KStemmer with the supplied base-form lexicon.

Fields

DerivationalRules

Exceptions

InflectionalRules

ProtectedWords

_derivationalBySuffix

_inflectionalBySuffix

_lexicon

Methods

ApplyRulesCore(Span<char>, int)

Applies inflectional and derivational rules without re-checking exceptions or the lexicon (caller has already done so).

BuildRuleLookup(MorphRule[])

Builds a lookup dictionary keyed by the last 2 characters of each rule's suffix.

ContainsVowel(string)

ContainsVowelSpan(ReadOnlySpan<char>)

IsVowel(char)

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 a stemmed string.

StemCore(Span<char>, int)

Core stemming logic operating on the already-lowered word in output. Called from Stem(ReadOnlySpan{char}, Span{char}) which hasn't pre-checked exceptions or the lexicon.

StemPreLowered(ReadOnlySpan<char>, Span<char>)

Same as Stem(ReadOnlySpan{char}, Span{char}) but assumes the input is already lowercased. Callers who have already lowercased via LowercaseFilter can use this to avoid a redundant per-character ToLowerInvariant pass.

TryRulesSpan(Span<char>, int, FrozenDictionary<string, MorphRule[]>, out int)

UndoubleTrailingConsonantSpan(Span<char>, int)