common/org.gnit.lucenekmp.analysis.morph

Package-level declarations

Types

abstract class BinaryDictionary<T : MorphData>(targetMapResource: IOSupplier<InputStream>, dictResource: IOSupplier<InputStream>, targetMapCodecHeader: String, dictCodecHeader: String, dictCodecVersion: Int) : Dictionary<T>

Abstract dictionary base class.

Abstract base dictionary writer class.

abstract class CharacterDefinition(charDefResource: IOSupplier<InputStream>, charDefCodecHeader: String, charDefCodecVersion: Int, classCount: Int)

Character category data.

Writes character definition file

abstract class ConnectionCosts(connectionCostResource: IOSupplier<InputStream>, connectionCostsCodecHeader: String, dictCodecVersion: Int)

n-gram connection cost data

class ConnectionCostsWriter<T : ConnectionCosts>(implClazz: KClass<T>, forwardSize: Int, backwardSize: Int)

Writes connection costs

High-level dictionary interface for morphological analyzers.

Abstract writer class to write dictionary entries.

Outputs the dot (graphviz) string for the viterbi lattice.

interface MorphData

High-level interface that represents morphological information in a dictionary

abstract class Token(val surfaceForm: CharArray, val offset: Int, val length: Int, val startOffset: Int, val endOffset: Int, val type: TokenType)

Analyzed token with morphological data.

abstract class TokenInfoFST(fst: FST<Long>, cacheCeiling: Int, cacheFloor: Int)

Thin wrapper around an FST with root-arc caching.

Token type reflecting the original source of this token

abstract class Viterbi<T : Token, U : Viterbi.Position>(fst: TokenInfoFST, fstReader: FST.BytesReader, dictionary: BinaryDictionary<out MorphData>, userFST: TokenInfoFST?, userFSTReader: FST.BytesReader?, userDictionary: Dictionary<out MorphData>?, costs: ConnectionCosts, positionFactory: () -> U)

Performs Viterbi algorithm for morphological Tokenizers.

Viterbi subclass for n-best path calculation.