public final class ShingleFilter extends TokenFilter
A ShingleFilter constructs shingles (token n-grams) from a token stream. In other words, it creates combinations of tokens as a single token.
For example, the sentence "please divide this sentence into shingles" might be tokenized into shingles "please divide", "divide this", "this sentence", "sentence into", and "into shingles".
This filter handles position increments > 1 by inserting filler tokens (tokens with termtext "_"). It does not handle a position increment of 0.
Modifier and Type | Class and Description |
---|---|
private class |
ShingleFilter.CircularSequence
An instance of this class is used to maintain the number of input
stream tokens that will be used to compose the next unigram or shingle:
gramSize . |
private static class |
ShingleFilter.InputWindowToken |
AttributeSource.State
Modifier and Type | Field and Description |
---|---|
static java.lang.String |
DEFAULT_FILLER_TOKEN
filler token for when positionIncrement is more than 1
|
static int |
DEFAULT_MAX_SHINGLE_SIZE
default maximum shingle size is 2.
|
static int |
DEFAULT_MIN_SHINGLE_SIZE
default minimum shingle size is 2.
|
static java.lang.String |
DEFAULT_TOKEN_SEPARATOR
The default string to use when joining adjacent tokens to form a shingle
|
static java.lang.String |
DEFAULT_TOKEN_TYPE
default token type attribute value is "shingle"
|
private AttributeSource.State |
endState
Holds the State after input.end() was called, so we can
restore it in our end() impl.
|
private boolean |
exhausted |
private char[] |
fillerToken
The string to insert for each position at which there is no token
(i.e., when position increment is greater than one).
|
private java.lang.StringBuilder |
gramBuilder
Shingle and unigram text is composed here.
|
private ShingleFilter.CircularSequence |
gramSize
The number of input tokens in the next output token.
|
private java.util.LinkedList<ShingleFilter.InputWindowToken> |
inputWindow
The sequence of input stream tokens (or filler tokens, if necessary)
that will be composed to form output shingles.
|
private boolean |
isNextInputStreamToken
Whether or not there is a next input stream token.
|
private boolean |
isOutputHere
Whether at least one unigram or shingle has been output at the current
position.
|
private int |
maxShingleSize
maximum shingle size (number of tokens)
|
private int |
minShingleSize
minimum shingle size (number of tokens)
|
private AttributeSource |
nextInputStreamToken
When the next input stream token has a position increment greater than
one, it is stored in this field until sufficient filler tokens have been
inserted to account for the position increment.
|
(package private) boolean |
noShingleOutput
true if no shingles have been output yet (for outputUnigramsIfNoShingles).
|
private int |
numFillerTokensToInsert
The remaining number of filler tokens to be inserted into the input stream
from which shingles are composed, to handle position increments greater
than one.
|
private OffsetAttribute |
offsetAtt |
private boolean |
outputUnigrams
By default, we output unigrams (individual tokens) as well as shingles
(token n-grams).
|
private boolean |
outputUnigramsIfNoShingles
By default, we don't override behavior of outputUnigrams.
|
private PositionIncrementAttribute |
posIncrAtt |
private PositionLengthAttribute |
posLenAtt |
private CharTermAttribute |
termAtt |
private java.lang.String |
tokenSeparator
The string to use when joining adjacent tokens to form a shingle
|
private java.lang.String |
tokenType
The token type attribute value to use - default is "shingle"
|
private TypeAttribute |
typeAtt |
input
DEFAULT_TOKEN_ATTRIBUTE_FACTORY
Constructor and Description |
---|
ShingleFilter(TokenStream input)
Construct a ShingleFilter with default shingle size: 2.
|
ShingleFilter(TokenStream input,
int maxShingleSize)
Constructs a ShingleFilter with the specified shingle size from the
TokenStream input |
ShingleFilter(TokenStream input,
int minShingleSize,
int maxShingleSize)
Constructs a ShingleFilter with the specified shingle size from the
TokenStream input |
ShingleFilter(TokenStream input,
java.lang.String tokenType)
Construct a ShingleFilter with the specified token type for shingle tokens
and the default shingle size: 2
|
Modifier and Type | Method and Description |
---|---|
void |
end()
This method is called by the consumer after the last token has been
consumed, after
TokenStream.incrementToken() returned false
(using the new TokenStream API). |
private ShingleFilter.InputWindowToken |
getNextToken(ShingleFilter.InputWindowToken target)
Get the next token from the input stream.
|
boolean |
incrementToken()
Consumers (i.e.,
IndexWriter ) use this method to advance the stream to
the next token. |
void |
reset()
This method is called by a consumer before it begins consumption using
TokenStream.incrementToken() . |
void |
setFillerToken(java.lang.String fillerToken)
Sets the string to insert for each position at which there is no token
(i.e., when position increment is greater than one).
|
void |
setMaxShingleSize(int maxShingleSize)
Set the max shingle size (default: 2)
|
void |
setMinShingleSize(int minShingleSize)
Set the min shingle size (default: 2).
|
void |
setOutputUnigrams(boolean outputUnigrams)
Shall the output stream contain the input tokens (unigrams) as well as
shingles? (default: true.)
|
void |
setOutputUnigramsIfNoShingles(boolean outputUnigramsIfNoShingles)
Shall we override the behavior of outputUnigrams==false for those
times when no shingles are available (because there are fewer than
minShingleSize tokens in the input stream)? (default: false.)
|
void |
setTokenSeparator(java.lang.String tokenSeparator)
Sets the string to use when joining adjacent tokens to form a shingle
|
void |
setTokenType(java.lang.String tokenType)
Set the type of the shingle tokens produced by this filter.
|
private void |
shiftInputWindow()
Fills
inputWindow with input stream tokens, if available,
shifting to the right if the window was previously full. |
close
addAttribute, addAttributeImpl, captureState, clearAttributes, cloneAttributes, copyTo, endAttributes, equals, getAttribute, getAttributeClassesIterator, getAttributeFactory, getAttributeImplsIterator, hasAttribute, hasAttributes, hashCode, reflectAsString, reflectWith, removeAllAttributes, restoreState, toString
public static final java.lang.String DEFAULT_FILLER_TOKEN
public static final int DEFAULT_MAX_SHINGLE_SIZE
public static final int DEFAULT_MIN_SHINGLE_SIZE
public static final java.lang.String DEFAULT_TOKEN_TYPE
public static final java.lang.String DEFAULT_TOKEN_SEPARATOR
private java.util.LinkedList<ShingleFilter.InputWindowToken> inputWindow
private ShingleFilter.CircularSequence gramSize
private java.lang.StringBuilder gramBuilder
private java.lang.String tokenType
private java.lang.String tokenSeparator
private char[] fillerToken
private boolean outputUnigrams
private boolean outputUnigramsIfNoShingles
private int maxShingleSize
private int minShingleSize
private int numFillerTokensToInsert
private AttributeSource nextInputStreamToken
private boolean isNextInputStreamToken
private boolean isOutputHere
boolean noShingleOutput
private AttributeSource.State endState
private final CharTermAttribute termAtt
private final OffsetAttribute offsetAtt
private final PositionIncrementAttribute posIncrAtt
private final PositionLengthAttribute posLenAtt
private final TypeAttribute typeAtt
private boolean exhausted
public ShingleFilter(TokenStream input, int minShingleSize, int maxShingleSize)
TokenStream
input
input
- input streamminShingleSize
- minimum shingle size produced by the filter.maxShingleSize
- maximum shingle size produced by the filter.public ShingleFilter(TokenStream input, int maxShingleSize)
TokenStream
input
input
- input streammaxShingleSize
- maximum shingle size produced by the filter.public ShingleFilter(TokenStream input)
input
- input streampublic ShingleFilter(TokenStream input, java.lang.String tokenType)
input
- input streamtokenType
- token type for shingle tokenspublic void setTokenType(java.lang.String tokenType)
tokenType
- token tokenTypepublic void setOutputUnigrams(boolean outputUnigrams)
outputUnigrams
- Whether or not the output stream shall contain
the input tokens (unigrams)public void setOutputUnigramsIfNoShingles(boolean outputUnigramsIfNoShingles)
Shall we override the behavior of outputUnigrams==false for those times when no shingles are available (because there are fewer than minShingleSize tokens in the input stream)? (default: false.)
Note that if outputUnigrams==true, then unigrams are always output, regardless of whether any shingles are available.
outputUnigramsIfNoShingles
- Whether or not to output a single
unigram when no shingles are available.public void setMaxShingleSize(int maxShingleSize)
maxShingleSize
- max size of output shinglespublic void setMinShingleSize(int minShingleSize)
Set the min shingle size (default: 2).
This method requires that the passed in minShingleSize is not greater than maxShingleSize, so make sure that maxShingleSize is set before calling this method.
The unigram output option is independent of the min shingle size.
minShingleSize
- min size of output shinglespublic void setTokenSeparator(java.lang.String tokenSeparator)
tokenSeparator
- used to separate input stream tokens in output shinglespublic void setFillerToken(java.lang.String fillerToken)
fillerToken
- string to insert at each position where there is no tokenpublic boolean incrementToken() throws java.io.IOException
TokenStream
IndexWriter
) use this method to advance the stream to
the next token. Implementing classes must implement this method and update
the appropriate AttributeImpl
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)
and AttributeSource.getAttribute(Class)
,
references to all AttributeImpl
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()
.
incrementToken
in class TokenStream
java.io.IOException
private ShingleFilter.InputWindowToken getNextToken(ShingleFilter.InputWindowToken target) throws java.io.IOException
Get the next token from the input stream.
If the next token has positionIncrement > 1
,
positionIncrement - 1
fillerToken
s are
inserted first.
target
- Where to put the new token; if null, a new instance is created.java.io.IOException
- if the input stream has a problempublic void end() throws java.io.IOException
TokenFilter
TokenStream.incrementToken()
returned false
(using the new TokenStream
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()
.
NOTE:
The default implementation chains the call to the input TokenStream, so
be sure to call super.end()
first when overriding this method.
end
in class TokenFilter
java.io.IOException
- If an I/O error occursprivate void shiftInputWindow() throws java.io.IOException
Fills inputWindow
with input stream tokens, if available,
shifting to the right if the window was previously full.
Resets gramSize
to its minimum value.
java.io.IOException
- if there's a problem getting the next tokenpublic void reset() throws java.io.IOException
TokenFilter
TokenStream.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
throw IllegalStateException
on further usage).
NOTE:
The default implementation chains the call to the input TokenStream, so
be sure to call super.reset()
when overriding this method.
reset
in class TokenFilter
java.io.IOException