public abstract class TokenSearchManager extends java.lang.Object implements SearchManager
Modifier and Type | Field and Description |
---|---|
protected boolean |
buildWordLattice |
protected boolean |
keepAllTokens |
static java.lang.String |
PROP_BUILD_WORD_LATTICE
The property that specifies whether to build a word lattice.
|
static java.lang.String |
PROP_KEEP_ALL_TOKENS
The property that controls whether or not we keep all tokens.
|
Constructor and Description |
---|
TokenSearchManager() |
Modifier and Type | Method and Description |
---|---|
protected Token |
getResultListPredecessor(Token token)
Find the token to use as a predecessor in resultList given a candidate
predecessor.
|
void |
newProperties(PropertySheet ps)
This method is called when this configurable component needs to be reconfigured.
|
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
allocate, deallocate, recognize, startRecognition, stopRecognition
@S4Boolean(defaultValue=true) public static final java.lang.String PROP_BUILD_WORD_LATTICE
@S4Boolean(defaultValue=false) public static final java.lang.String PROP_KEEP_ALL_TOKENS
protected boolean buildWordLattice
protected boolean keepAllTokens
public void newProperties(PropertySheet ps) throws PropertyException
Configurable
newProperties
in interface Configurable
ps
- a property sheet holding the new dataPropertyException
- if there is a problem with the properties.protected Token getResultListPredecessor(Token token)
keepAllTokens
is set to true and we just store everything that
was built before.
buildWordLattice
is set to true.
In this case we want to insert intermediate token to store the score and
this token will be used during lattice path collapse to get score on
edge. See Lattice
for details of resultList
compression.
token
- the token of interest