Class JapaneseTokenizer
- All Implemented Interfaces:
Closeable
,AutoCloseable
This tokenizer sets a number of additional attributes:
BaseFormAttribute
containing base form for inflected adjectives and verbs.PartOfSpeechAttribute
containing part-of-speech.ReadingAttribute
containing reading and pronunciation.InflectionAttribute
containing additional part-of-speech information for inflected forms.
This tokenizer uses a rolling Viterbi search to find the least cost segmentation (path) of the
incoming characters. For tokens that appear to be compound (> length 2 for all Kanji, or >
length 7 for non-Kanji), we see if there is a 2nd best segmentation of that token after applying
penalties to the long tokens. If so, and the Mode is JapaneseTokenizer.Mode.SEARCH
, we output the alternate
segmentation as well.
-
Nested Class Summary
Nested ClassesModifier and TypeClassDescriptionprivate static final class
static enum
Tokenization mode: this determines how the tokenizer handles compound and unknown words.(package private) static final class
static enum
Token type reflecting the original source of this token(package private) static final class
Nested classes/interfaces inherited from class org.apache.lucene.util.AttributeSource
AttributeSource.State
-
Field Summary
FieldsModifier and TypeFieldDescriptionprivate final BaseFormAttribute
private final RollingCharBuffer
private final CharacterDefinition
private final ConnectionCosts
static final JapaneseTokenizer.Mode
Default tokenization mode.private final TokenInfoDictionary
private final EnumMap<JapaneseTokenizer.Type,
Dictionary> private final boolean
private GraphvizFormatter
private boolean
private final boolean
private final TokenInfoFST
private final FST.BytesReader
private final InflectionAttribute
private int
private int
private JapaneseTokenizer.Lattice
private static final int
private static final int
private int
private final OffsetAttribute
private final boolean
private boolean
private int
private final PartOfSpeechAttribute
private final PositionIncrementAttribute
private final JapaneseTokenizer.WrappedPositionArray
private final PositionLengthAttribute
private final ReadingAttribute
private static final int
private static final int
private static final int
private static final int
private final boolean
private final CharTermAttribute
private final UnknownDictionary
private final UserDictionary
private final TokenInfoFST
private final FST.BytesReader
private static final boolean
private final IntsRef
Fields inherited from class org.apache.lucene.analysis.TokenStream
DEFAULT_TOKEN_ATTRIBUTE_FACTORY
-
Constructor Summary
ConstructorsConstructorDescriptionJapaneseTokenizer
(UserDictionary userDictionary, boolean discardPunctuation, boolean discardCompoundToken, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer.JapaneseTokenizer
(UserDictionary userDictionary, boolean discardPunctuation, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer.JapaneseTokenizer
(AttributeFactory factory, TokenInfoDictionary systemDictionary, UnknownDictionary unkDictionary, ConnectionCosts connectionCosts, UserDictionary userDictionary, boolean discardPunctuation, boolean discardCompoundToken, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer, supplying a custom system dictionary and unknown dictionary.JapaneseTokenizer
(AttributeFactory factory, UserDictionary userDictionary, boolean discardPunctuation, boolean discardCompoundToken, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer using the system and unknown dictionaries shipped with Lucene.JapaneseTokenizer
(AttributeFactory factory, UserDictionary userDictionary, boolean discardPunctuation, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer using the system and unknown dictionaries shipped with Lucene. -
Method Summary
Modifier and TypeMethodDescriptionprivate void
add
(Dictionary dict, JapaneseTokenizer.Position fromPosData, int endPos, int wordID, JapaneseTokenizer.Type type, boolean addPenalty) private void
backtrace
(JapaneseTokenizer.Position endPosData, int fromIDX) private void
backtraceNBest
(JapaneseTokenizer.Position endPosData, boolean useEOS) int
calcNBestCost
(String examples) void
close()
Releases resources associated with this stream.private int
computePenalty
(int pos, int length) private int
computeSecondBestThreshold
(int pos, int length) void
end()
This method is called by the consumer after the last token has been consumed, afterTokenStream.incrementToken()
returnedfalse
(using the newTokenStream
API).private void
(package private) Dictionary
boolean
Consumers (i.e.,IndexWriter
) use this method to advance the stream to the next token.private static boolean
isPunctuation
(char ch) private void
parse()
private int
probeDelta
(String inText, String requiredToken) private void
pruneAndRescore
(int startPos, int endPos, int bestStartIDX) private void
registerNode
(int node, char[] fragment) void
reset()
This method is called by a consumer before it begins consumption usingTokenStream.incrementToken()
.private void
void
Expert: set this to produce graphviz (dot) output of the Viterbi latticevoid
setNBestCost
(int value) Methods inherited from class org.apache.lucene.analysis.Tokenizer
correctOffset, setReader, setReaderTestPoint
Methods inherited from class org.apache.lucene.util.AttributeSource
addAttribute, addAttributeImpl, captureState, clearAttributes, cloneAttributes, copyTo, endAttributes, equals, getAttribute, getAttributeClassesIterator, getAttributeFactory, getAttributeImplsIterator, hasAttribute, hasAttributes, hashCode, reflectAsString, reflectWith, removeAllAttributes, restoreState, toString
-
Field Details
-
DEFAULT_MODE
Default tokenization mode. Currently this isJapaneseTokenizer.Mode.SEARCH
. -
VERBOSE
private static final boolean VERBOSE- See Also:
-
SEARCH_MODE_KANJI_LENGTH
private static final int SEARCH_MODE_KANJI_LENGTH- See Also:
-
SEARCH_MODE_OTHER_LENGTH
private static final int SEARCH_MODE_OTHER_LENGTH- See Also:
-
SEARCH_MODE_KANJI_PENALTY
private static final int SEARCH_MODE_KANJI_PENALTY- See Also:
-
SEARCH_MODE_OTHER_PENALTY
private static final int SEARCH_MODE_OTHER_PENALTY- See Also:
-
MAX_UNKNOWN_WORD_LENGTH
private static final int MAX_UNKNOWN_WORD_LENGTH- See Also:
-
MAX_BACKTRACE_GAP
private static final int MAX_BACKTRACE_GAP- See Also:
-
dictionaryMap
-
fst
-
dictionary
-
unkDictionary
-
costs
-
userDictionary
-
characterDefinition
-
arc
-
fstReader
-
wordIdRef
-
userFSTReader
-
userFST
-
buffer
-
positions
-
discardPunctuation
private final boolean discardPunctuation -
searchMode
private final boolean searchMode -
extendedMode
private final boolean extendedMode -
outputCompounds
private final boolean outputCompounds -
outputNBest
private boolean outputNBest -
nBestCost
private int nBestCost -
end
private boolean end -
lastBackTracePos
private int lastBackTracePos -
lastTokenPos
private int lastTokenPos -
pos
private int pos -
pending
-
termAtt
-
offsetAtt
-
posIncAtt
-
posLengthAtt
-
basicFormAtt
-
posAtt
-
readingAtt
-
inflectionAtt
-
dotOut
-
lattice
-
-
Constructor Details
-
JapaneseTokenizer
public JapaneseTokenizer(UserDictionary userDictionary, boolean discardPunctuation, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer.Uses the default AttributeFactory.
- Parameters:
userDictionary
- Optional: if non-null, user dictionary.discardPunctuation
- true if punctuation tokens should be dropped from the output.mode
- tokenization mode.
-
JapaneseTokenizer
public JapaneseTokenizer(UserDictionary userDictionary, boolean discardPunctuation, boolean discardCompoundToken, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer.Uses the default AttributeFactory.
- Parameters:
userDictionary
- Optional: if non-null, user dictionary.discardPunctuation
- true if punctuation tokens should be dropped from the output.discardCompoundToken
- true if compound tokens should be dropped from the output when tokenization mode is not NORMAL.mode
- tokenization mode.
-
JapaneseTokenizer
public JapaneseTokenizer(AttributeFactory factory, UserDictionary userDictionary, boolean discardPunctuation, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer using the system and unknown dictionaries shipped with Lucene.- Parameters:
factory
- the AttributeFactory to useuserDictionary
- Optional: if non-null, user dictionary.discardPunctuation
- true if punctuation tokens should be dropped from the output.mode
- tokenization mode.
-
JapaneseTokenizer
public JapaneseTokenizer(AttributeFactory factory, UserDictionary userDictionary, boolean discardPunctuation, boolean discardCompoundToken, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer using the system and unknown dictionaries shipped with Lucene.- Parameters:
factory
- the AttributeFactory to useuserDictionary
- Optional: if non-null, user dictionary.discardPunctuation
- true if punctuation tokens should be dropped from the output.discardCompoundToken
- true if compound tokens should be dropped from the output when tokenization mode is not NORMAL.mode
- tokenization mode.
-
JapaneseTokenizer
public JapaneseTokenizer(AttributeFactory factory, TokenInfoDictionary systemDictionary, UnknownDictionary unkDictionary, ConnectionCosts connectionCosts, UserDictionary userDictionary, boolean discardPunctuation, boolean discardCompoundToken, JapaneseTokenizer.Mode mode) Create a new JapaneseTokenizer, supplying a custom system dictionary and unknown dictionary. This constructor provides an entry point for users that want to construct custom language models that can be used as input toDictionaryBuilder
.- Parameters:
factory
- the AttributeFactory to usesystemDictionary
- a custom known token dictionaryunkDictionary
- a custom unknown token dictionaryconnectionCosts
- custom token transition costsuserDictionary
- Optional: if non-null, user dictionary.discardPunctuation
- true if punctuation tokens should be dropped from the output.discardCompoundToken
- true if compound tokens should be dropped from the output when tokenization mode is not NORMAL.mode
- tokenization mode.
-
-
Method Details
-
setGraphvizFormatter
Expert: set this to produce graphviz (dot) output of the Viterbi lattice -
close
Description copied from class:Tokenizer
Releases resources associated with this stream.If you override this method, always call
super.close()
, otherwise some internal state will not be correctly reset (e.g.,Tokenizer
will throwIllegalStateException
on reuse).NOTE: The default implementation closes the input Reader, so be sure to call
super.close()
when overriding this method.- Specified by:
close
in interfaceAutoCloseable
- Specified by:
close
in interfaceCloseable
- Overrides:
close
in classTokenizer
- Throws:
IOException
-
reset
Description copied from class:TokenStream
This method is called by a consumer before it begins consumption usingTokenStream.incrementToken()
.Resets this stream to a clean state. Stateful implementations must implement this method so that they can be reused, just as if they had been created fresh.
If you override this method, always call
super.reset()
, otherwise some internal state will not be correctly reset (e.g.,Tokenizer
will throwIllegalStateException
on further usage).- Overrides:
reset
in classTokenizer
- Throws:
IOException
-
resetState
private void resetState() -
end
Description copied from class:TokenStream
This method is called by the consumer after the last token has been consumed, afterTokenStream.incrementToken()
returnedfalse
(using the newTokenStream
API). Streams implementing the old API should upgrade to use this feature.This method can be used to perform any end-of-stream operations, such as setting the final offset of a stream. The final offset of a stream might differ from the offset of the last token eg in case one or more whitespaces followed after the last token, but a WhitespaceTokenizer was used.
Additionally any skipped positions (such as those removed by a stopfilter) can be applied to the position increment, or any adjustment of other attributes where the end-of-stream value may be important.
If you override this method, always call
super.end()
.- Overrides:
end
in classTokenStream
- Throws:
IOException
- If an I/O error occurs
-
computeSecondBestThreshold
- Throws:
IOException
-
computePenalty
- Throws:
IOException
-
add
private void add(Dictionary dict, JapaneseTokenizer.Position fromPosData, int endPos, int wordID, JapaneseTokenizer.Type type, boolean addPenalty) throws IOException - Throws:
IOException
-
incrementToken
Description copied from class:TokenStream
Consumers (i.e.,IndexWriter
) use this method to advance the stream to the next token. Implementing classes must implement this method and update the appropriateAttributeImpl
s with the attributes of the next token.The producer must make no assumptions about the attributes after the method has been returned: the caller may arbitrarily change it. If the producer needs to preserve the state for subsequent calls, it can use
AttributeSource.captureState()
to create a copy of the current attribute state.This method is called for every token of a document, so an efficient implementation is crucial for good performance. To avoid calls to
AttributeSource.addAttribute(Class)
andAttributeSource.getAttribute(Class)
, references to allAttributeImpl
s that this stream uses should be retrieved during instantiation.To ensure that filters and consumers know which attributes are available, the attributes must be added during instantiation. Filters and consumers are not required to check for availability of attributes in
TokenStream.incrementToken()
.- Specified by:
incrementToken
in classTokenStream
- Returns:
- false for end of stream; true otherwise
- Throws:
IOException
-
parse
- Throws:
IOException
-
pruneAndRescore
- Throws:
IOException
-
registerNode
private void registerNode(int node, char[] fragment) -
fixupPendingList
private void fixupPendingList() -
probeDelta
- Throws:
IOException
-
calcNBestCost
-
setNBestCost
public void setNBestCost(int value) -
backtraceNBest
private void backtraceNBest(JapaneseTokenizer.Position endPosData, boolean useEOS) throws IOException - Throws:
IOException
-
backtrace
- Throws:
IOException
-
getDict
-
isPunctuation
private static boolean isPunctuation(char ch)
-