public class GazetteerListsCollector extends AbstractLanguageAnalyser
AbstractProcessingResource.InternalStatusListener, AbstractProcessingResource.IntervalProgressListener
corpus, document
interrupted
name
features
ANNOTATION_COREF_FEATURE_NAME, DATE_ANNOTATION_TYPE, DATE_POSTED_ANNOTATION_TYPE, DEFAULT_FILE, DOCUMENT_COREF_FEATURE_NAME, JOB_ID_ANNOTATION_TYPE, LOCATION_ANNOTATION_TYPE, LOOKUP_ANNOTATION_TYPE, LOOKUP_CLASS_FEATURE_NAME, LOOKUP_INSTANCE_FEATURE_NAME, LOOKUP_LANGUAGE_FEATURE_NAME, LOOKUP_MAJOR_TYPE_FEATURE_NAME, LOOKUP_MINOR_TYPE_FEATURE_NAME, LOOKUP_ONTOLOGY_FEATURE_NAME, MONEY_ANNOTATION_TYPE, ORGANIZATION_ANNOTATION_TYPE, PERSON_ANNOTATION_TYPE, PERSON_GENDER_FEATURE_NAME, PLUGIN_DIR, SENTENCE_ANNOTATION_TYPE, SPACE_TOKEN_ANNOTATION_TYPE, TOKEN_ANNOTATION_TYPE, TOKEN_CATEGORY_FEATURE_NAME, TOKEN_KIND_FEATURE_NAME, TOKEN_LENGTH_FEATURE_NAME, TOKEN_ORTH_FEATURE_NAME, TOKEN_STRING_FEATURE_NAME
Constructor and Description |
---|
GazetteerListsCollector() |
Modifier and Type | Method and Description |
---|---|
protected boolean |
alreadyPresentInGazetteer(String token)
The idea is to have this method check if an item is already present
in the gazetteer under this type, and if so, not to add it.
|
protected void |
collectLists(AnnotationSet annots,
String annotType) |
void |
execute()
Run the resource.
|
List<String> |
getAnnotationTypes()
get the types of the annotation
|
Gazetteer |
getGazetteer() |
String |
getMarkupASName() |
String |
getTheLanguage() |
protected void |
printStats() |
void |
setAnnotationTypes(List<String> newType)
set the types of the annotations
|
void |
setGazetteer(Gazetteer theGaz) |
void |
setMarkupASName(String newMarkupASName) |
void |
setTheLanguage(String language) |
getCorpus, getDocument, setCorpus, setDocument
addProgressListener, addStatusListener, cleanup, fireProcessFinished, fireProgressChanged, fireStatusChanged, getRuntimeParameterValues, getRuntimeParameterValues, init, interrupt, isInterrupted, reInit, removeProgressListener, removeStatusListener
checkParameterValues, flushBeanInfoCache, forgetBeanInfo, getBeanInfo, getInitParameterValues, getInitParameterValues, getName, getParameterValue, getParameterValue, getParameterValues, removeResourceListeners, setName, setParameterValue, setParameterValue, setParameterValues, setParameterValues, setResourceListeners, toString
getFeatures, setFeatures
clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait
reInit
cleanup, getParameterValue, init, setParameterValue, setParameterValues
getFeatures, setFeatures
getName, setName
interrupt, isInterrupted
public GazetteerListsCollector()
public void execute() throws ExecutionException
AbstractProcessingResource
execute
in interface Executable
execute
in class AbstractProcessingResource
ExecutionException
public void setMarkupASName(String newMarkupASName)
public String getMarkupASName()
public List<String> getAnnotationTypes()
public void setAnnotationTypes(List<String> newType)
newType
- public Gazetteer getGazetteer()
public void setGazetteer(Gazetteer theGaz)
public void setTheLanguage(String language)
public String getTheLanguage()
protected void collectLists(AnnotationSet annots, String annotType)
protected void printStats()
protected boolean alreadyPresentInGazetteer(String token)