dic

Documentation

Overview

Package dic implements the dictionary of the morph analyzer.

Index

• Constants
• func NewContents(b []byte) [][]string
• type ConnectionTable
  • func LoadConnectionTable(r io.Reader) (t ConnectionTable, err error)
  • func (t *ConnectionTable) At(row, col int) int16
  • func (t ConnectionTable) WriteTo(w io.Writer) (n int64, err error)
• type Contents
  • func (c Contents) WriteTo(w io.Writer) (n int64, err error)
• type Dic
  • func Load(path string) (d *Dic, err error)
  • func SysDic() *Dic
  • func SysDicIPA() *Dic
  • func SysDicUni() *Dic
  • func (d Dic) CharacterCategory(r rune) byte
• type IndexTable
  • func BuildIndexTable(sortedKeywords []string) (IndexTable, error)
  • func ReadIndexTable(r io.Reader) (IndexTable, error)
  • func (idx IndexTable) CommonPrefixSearch(input string) (lens []int, ids [][]int)
  • func (idx IndexTable) CommonPrefixSearchCallback(input string, callback func(id, l int))
  • func (idx IndexTable) Search(input string) []int
  • func (idx IndexTable) WriteTo(w io.Writer) (n int64, err error)
• type Morph
  • func LoadMorphSlice(r io.Reader) ([]Morph, error)
• type MorphSlice
  • func (m MorphSlice) WriteTo(w io.Writer) (n int64, err error)
• type Trie
• type UserDic
  • func NewUserDic(path string) (udic *UserDic, err error)
• type UserDicContent

Constants

const (
	// IPADicPath represents the internal IPA dictionary path.
	IPADicPath = "dic/ipa"
	// UniDicPath represents the internal UniDic dictionary path.
	UniDicPath = "dic/uni"
)

const UserDicColumnSize = 4

UserDicColumnSize is the column size of the user dictionary.

Functions

func NewContents

func NewContents(b []byte) [][]string

NewContents creates dictionary contents from byte slice

Types

type ConnectionTable

type ConnectionTable struct {
	Row, Col int64
	Vec      []int16
}

ConnectionTable represents a connection matrix of morphs.

func LoadConnectionTable

func LoadConnectionTable(r io.Reader) (t ConnectionTable, err error)

LoadConnectionTable loads ConnectionTable from io.Reader.

func (*ConnectionTable) At

func (t *ConnectionTable) At(row, col int) int16

At returns the connection cost of matrix[row, col].

func (ConnectionTable) WriteTo

func (t ConnectionTable) WriteTo(w io.Writer) (n int64, err error)

WriteTo implements the io.WriterTo interface

type Contents

type Contents [][]string

Contents represents dictionary contents.

func (Contents) WriteTo

func (c Contents) WriteTo(w io.Writer) (n int64, err error)

WriteTo implements the io.WriterTo interface

type Dic

type Dic struct {
	Morphs       []Morph
	Contents     [][]string
	Connection   ConnectionTable
	Index        IndexTable
	CharClass    []string
	CharCategory []byte
	InvokeList   []bool
	GroupList    []bool
	UnkMorphs    []Morph
	UnkIndex     map[int32]int32
	UnkIndexDup  map[int32]int32
	UnkContents  [][]string
}

Dic represents a dictionary of a tokenizer.

func Load

func Load(path string) (d *Dic, err error)

Load loads a dictionary from a file.

func SysDic

func SysDic() *Dic

SysDic returns the kagome system dictionary.

func SysDicIPA

func SysDicIPA() *Dic

SysDicIPA returns the IPA system dictionary.

func SysDicUni

func SysDicUni() *Dic

SysDicUni returns the UniDic system dictionary.

func (Dic) CharacterCategory

func (d Dic) CharacterCategory(r rune) byte

CharacterCategory returns the category of a rune.

type IndexTable

type IndexTable struct {
	Da  da.DoubleArray
	Dup map[int32]int32
}

IndexTable represents a dictionary index.

func BuildIndexTable

func BuildIndexTable(sortedKeywords []string) (IndexTable, error)

BuildIndexTable constructs a index table from keywords.

func ReadIndexTable

func ReadIndexTable(r io.Reader) (IndexTable, error)

ReadIndexTable loads a index table.

func (IndexTable) CommonPrefixSearch

func (idx IndexTable) CommonPrefixSearch(input string) (lens []int, ids [][]int)

CommonPrefixSearch finds keywords sharing common prefix in an input and returns the ids and it's lengths if found.

func (IndexTable) CommonPrefixSearchCallback

func (idx IndexTable) CommonPrefixSearchCallback(input string, callback func(id, l int))

CommonPrefixSearchCallback finds keywords sharing common prefix in an input and callback with id and length.

func (IndexTable) Search

func (idx IndexTable) Search(input string) []int

Search finds the given keyword and returns the id if found.

func (IndexTable) WriteTo

func (idx IndexTable) WriteTo(w io.Writer) (n int64, err error)

WriteTo saves a index table.

type Morph

type Morph struct {
	LeftID, RightID, Weight int16
}

Morph represents part of speeches and an occurrence cost.

func LoadMorphSlice

func LoadMorphSlice(r io.Reader) ([]Morph, error)

LoadMorphSlice loads morph data from io.Reader

type MorphSlice

type MorphSlice []Morph

MorphSlice represents a slice of morphs.

func (MorphSlice) WriteTo

func (m MorphSlice) WriteTo(w io.Writer) (n int64, err error)

WriteTo implements the io.WriterTo interface

type Trie

type Trie interface {
	Search(input string) []int32
	PrefixSearch(input string) (length int, output []int32)
	CommonPrefixSearch(input string) (lens []int, outputs [][]int32)
	CommonPrefixSearchCallback(input string, callback func(id, l int))
}

Trie is an interface representing retrieval ability.

type UserDic

type UserDic struct {
	Index    IndexTable
	Contents []UserDicContent
}

UserDic represents a user dictionary.

func NewUserDic

func NewUserDic(path string) (udic *UserDic, err error)

NewUserDic build a user dictionary from a file.

type UserDicContent

type UserDicContent struct {
	Tokens []string
	Yomi   []string
	Pos    string
}

UserDicContent represents contents of a word in a user dictionary.