Module org.apache.lucene.analysis.kuromoji

Package org.apache.lucene.analysis.ja.util

Class TokenInfoDictionaryBuilder

java.lang.Object

org.apache.lucene.analysis.ja.util.TokenInfoDictionaryBuilder

class TokenInfoDictionaryBuilder
extends Object

Field Summary

Fields

Modifier and Type	Field	Description
private final String	encoding
private final DictionaryBuilder.DictionaryFormat	format
private final Normalizer.Form	normalForm
private int	offset	Internal word id - incrementally assigned as entries are read and added.

Constructor Summary

Constructors

Constructor	Description
TokenInfoDictionaryBuilder(DictionaryBuilder.DictionaryFormat format, String encoding, boolean normalizeEntries)

Method Summary

Modifier and Type	Method	Description
TokenInfoDictionaryWriter	build(Path dir)
private TokenInfoDictionaryWriter	buildDictionary(List<Path> csvFiles)
private String[]	formatEntry(String[] features)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

encoding

private final String encoding

normalForm

private final Normalizer.Form normalForm

format

private final DictionaryBuilder.DictionaryFormat format

offset

private int offset

Internal word id - incrementally assigned as entries are read and added. This will be byte offset of dictionary file

Constructor Details

TokenInfoDictionaryBuilder

public TokenInfoDictionaryBuilder(DictionaryBuilder.DictionaryFormat format,
 String encoding,
 boolean normalizeEntries)

Method Details

build

public TokenInfoDictionaryWriter build(Path dir)
                                throws IOException

Throws:

IOException

buildDictionary

private TokenInfoDictionaryWriter buildDictionary(List<Path> csvFiles)
                                           throws IOException

Throws:

IOException

formatEntry

private String[] formatEntry(String[] features)