Skip to content

Commit

Permalink
#1511 - Do not initialize POS mapping loader if mapping is disabled
Browse files Browse the repository at this point in the history 
- Do not initialize/configure mapper if the mapped feature is disabled
  • Loading branch information
@reckart
reckart committed Dec 31, 2022
1 parent f3d62c1 commit c217616
Show file tree
Hide file tree
Showing 14 changed files with 538 additions and 624 deletions.
83 changes: 39 additions & 44 deletions ...o-core-berkeleyparser-gpl/src/main/java/org/dkpro/core/berkeleyparser/BerkeleyParser.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -80,11 +80,8 @@
@ResourceMetaData(name = "Berkeley Parser")
@DocumentationResource("${docbase}/component-reference.html#engine-${shortClassName}")
@OperationalProperties(multipleDeploymentAllowed = false)
@TypeCapability(
inputs = {
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Token",
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence" },
outputs = {
@TypeCapability(inputs = { "de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Token",
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence" }, outputs = {
"de.tudarmstadt.ukp.dkpro.core.api.syntax.type.constituent.Constituent",
"de.tudarmstadt.ukp.dkpro.core.api.syntax.type.PennTree" })
public class BerkeleyParser
Expand All @@ -105,19 +102,20 @@ public class BerkeleyParser
protected String variant;

/**
* URI of the model artifact. This can be used to override the default model resolving
* mechanism and directly address a particular model.
* URI of the model artifact. This can be used to override the default model resolving mechanism
* and directly address a particular model.
*
* <p>The URI format is {@code mvn:${groupId}:${artifactId}:${version}}. Remember to set
* the variant parameter to match the artifact. If the artifact contains the model in
* a non-default location, you also have to specify the model location parameter, e.g.
* {@code classpath:/model/path/in/artifact/model.bin}.</p>
* <p>
* The URI format is {@code mvn:${groupId}:${artifactId}:${version}}. Remember to set the
* variant parameter to match the artifact. If the artifact contains the model in a non-default
* location, you also have to specify the model location parameter, e.g.
* {@code classpath:/model/path/in/artifact/model.bin}.
* </p>
*/
public static final String PARAM_MODEL_ARTIFACT_URI =
ComponentParameters.PARAM_MODEL_ARTIFACT_URI;
public static final String PARAM_MODEL_ARTIFACT_URI = ComponentParameters.PARAM_MODEL_ARTIFACT_URI;
@ConfigurationParameter(name = PARAM_MODEL_ARTIFACT_URI, mandatory = false)
protected String modelArtifactUri;

/**
* Load the model from this location instead of locating the model automatically.
*/
Expand All @@ -129,23 +127,20 @@ public class BerkeleyParser
* Enable/disable type mapping.
*/
public static final String PARAM_MAPPING_ENABLED = ComponentParameters.PARAM_MAPPING_ENABLED;
@ConfigurationParameter(name = PARAM_MAPPING_ENABLED, mandatory = true, defaultValue =
ComponentParameters.DEFAULT_MAPPING_ENABLED)
@ConfigurationParameter(name = PARAM_MAPPING_ENABLED, mandatory = true, defaultValue = ComponentParameters.DEFAULT_MAPPING_ENABLED)
protected boolean mappingEnabled;

/**
* Location of the mapping file for part-of-speech tags to UIMA types.
*/
public static final String PARAM_POS_MAPPING_LOCATION =
ComponentParameters.PARAM_POS_MAPPING_LOCATION;
public static final String PARAM_POS_MAPPING_LOCATION = ComponentParameters.PARAM_POS_MAPPING_LOCATION;
@ConfigurationParameter(name = PARAM_POS_MAPPING_LOCATION, mandatory = false)
protected String posMappingLocation;

/**
* Location of the mapping file for constituent tags to UIMA types.
*/
public static final String PARAM_CONSTITUENT_MAPPING_LOCATION =
ComponentParameters.PARAM_CONSTITUENT_MAPPING_LOCATION;
public static final String PARAM_CONSTITUENT_MAPPING_LOCATION = ComponentParameters.PARAM_CONSTITUENT_MAPPING_LOCATION;
@ConfigurationParameter(name = PARAM_CONSTITUENT_MAPPING_LOCATION, mandatory = false)
protected String constituentMappingLocation;

Expand All @@ -163,7 +158,7 @@ public class BerkeleyParser
public static final String PARAM_READ_POS = ComponentParameters.PARAM_READ_POS;
@ConfigurationParameter(name = PARAM_READ_POS, mandatory = true, defaultValue = "true")
private boolean readPos;

/**
* Sets whether to create or not to create POS tags. The creation of constituent tags must be
* turned on for this to work.
Expand Down Expand Up @@ -234,35 +229,37 @@ public class BerkeleyParser
private MappingProvider constituentMappingProvider;

@Override
public void initialize(UimaContext aContext)
throws ResourceInitializationException
public void initialize(UimaContext aContext) throws ResourceInitializationException
{
super.initialize(aContext);

modelProvider = new BerkeleyParserModelProvider();

posMappingProvider = createPosMappingProvider(this, posMappingLocation, language,
modelProvider);
if (writePos) {
posMappingProvider = createPosMappingProvider(this, posMappingLocation, language,
modelProvider);
}

constituentMappingProvider = createConstituentMappingProvider(this,
constituentMappingLocation, language, modelProvider);
}

@Override
public void process(JCas aJCas)
throws AnalysisEngineProcessException
public void process(JCas aJCas) throws AnalysisEngineProcessException
{
CAS cas = aJCas.getCas();

modelProvider.configure(cas);
posMappingProvider.configure(cas);
if (writePos) {
posMappingProvider.configure(cas);
}
constituentMappingProvider.configure(cas);

for (Sentence sentence : select(aJCas, Sentence.class)) {
List<Token> tokens = selectCovered(aJCas, Token.class, sentence);
List<String> tokenText = tokens.stream().map(t ->
t.getText()).collect(Collectors.toList());
List<String> tokenText = tokens.stream().map(t -> t.getText())
.collect(Collectors.toList());

List<String> posTags = null;
if (readPos) {
posTags = new ArrayList<String>(tokens.size());
Expand All @@ -271,15 +268,15 @@ public void process(JCas aJCas)
}
}

Tree<String> parseOutput = modelProvider.getResource().getBestConstrainedParse(
tokenText, posTags, false);
Tree<String> parseOutput = modelProvider.getResource()
.getBestConstrainedParse(tokenText, posTags, false);

// Check if the sentence could be parsed or not
if (parseOutput.getChildren().isEmpty()) {
getLogger().warn("Unable to parse sentence: [" + sentence.getCoveredText() + "]");
continue;
}

if (!binarize) {
parseOutput = TreeAnnotations.unAnnotateTree(parseOutput, keepFunctionLabels);
}
Expand Down Expand Up @@ -361,8 +358,7 @@ private Annotation createConstituentAnnotationFromTree(JCas aJCas, Tree<String>

// Now that we know how many children we have, link annotation of
// current node with its children
FSArray childArray = FSCollectionFactory.createFSArray(aJCas,
childAnnotations);
FSArray childArray = FSCollectionFactory.createFSArray(aJCas, childAnnotations);
constAnno.setChildren(childArray);

// write annotation for current node to index
Expand Down Expand Up @@ -391,8 +387,7 @@ private class BerkeleyParserModelProvider
}

@Override
protected CoarseToFineMaxRuleParser produceResource(URL aUrl)
throws IOException
protected CoarseToFineMaxRuleParser produceResource(URL aUrl) throws IOException
{
try (ObjectInputStream is = new ObjectInputStream(
new GZIPInputStream(aUrl.openStream()))) {
Expand All @@ -405,10 +400,10 @@ protected CoarseToFineMaxRuleParser produceResource(URL aUrl)
double threshold = 1.0;

Properties metadata = getResourceMetaData();
SingletonTagset posTags = new SingletonTagset(
POS.class, metadata.getProperty("pos.tagset"));
SingletonTagset constTags = new SingletonTagset(
Constituent.class, metadata.getProperty("constituent.tagset"));
SingletonTagset posTags = new SingletonTagset(POS.class,
metadata.getProperty("pos.tagset"));
SingletonTagset constTags = new SingletonTagset(Constituent.class,
metadata.getProperty("constituent.tagset"));

Numberer tagNumberer = (Numberer) pData.getNumbs().get("tags");
for (int i = 0; i < tagNumberer.size(); i++) {
Expand Down
33 changes: 13 additions & 20 deletions dkpro-core-cogroo-asl/src/main/java/org/dkpro/core/cogroo/CogrooPosTagger.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@
import static java.util.Arrays.asList;
import static org.apache.uima.fit.util.JCasUtil.select;
import static org.apache.uima.fit.util.JCasUtil.selectCovered;
import static org.dkpro.core.api.resources.MappingProviderFactory.createPosMappingProvider;

import java.io.IOException;
import java.net.URL;
Expand Down Expand Up @@ -50,7 +51,6 @@
import org.dkpro.core.api.parameter.ComponentParameters;
import org.dkpro.core.api.resources.CasConfigurableProviderBase;
import org.dkpro.core.api.resources.MappingProvider;
import org.dkpro.core.api.resources.MappingProviderFactory;
import org.dkpro.core.api.resources.ModelProviderBase;

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS;
Expand All @@ -67,10 +67,8 @@
@ResourceMetaData(name = "CoGrOO POS-Tagger")
@DocumentationResource("${docbase}/component-reference.html#engine-${shortClassName}")
@LanguageCapability("pt")
@TypeCapability(
inputs = {
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Token",
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence" })
@TypeCapability(inputs = { "de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Token",
"de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence" })
public class CogrooPosTagger
extends JCasAnnotator_ImplBase
{
Expand All @@ -85,29 +83,27 @@ public class CogrooPosTagger
* Enable/disable type mapping.
*/
public static final String PARAM_MAPPING_ENABLED = ComponentParameters.PARAM_MAPPING_ENABLED;
@ConfigurationParameter(name = PARAM_MAPPING_ENABLED, mandatory = true, defaultValue =
ComponentParameters.DEFAULT_MAPPING_ENABLED)
@ConfigurationParameter(name = PARAM_MAPPING_ENABLED, mandatory = true, defaultValue = ComponentParameters.DEFAULT_MAPPING_ENABLED)
protected boolean mappingEnabled;

/**
* Load the part-of-speech tag to UIMA type mapping from this location instead of locating
* the mapping automatically.
* Load the part-of-speech tag to UIMA type mapping from this location instead of locating the
* mapping automatically.
*/
public static final String PARAM_POS_MAPPING_LOCATION =
ComponentParameters.PARAM_POS_MAPPING_LOCATION;
public static final String PARAM_POS_MAPPING_LOCATION = ComponentParameters.PARAM_POS_MAPPING_LOCATION;
@ConfigurationParameter(name = PARAM_POS_MAPPING_LOCATION, mandatory = false)
protected String posMappingLocation;

private CasConfigurableProviderBase<Analyzer> modelProvider;
private MappingProvider mappingProvider;

@Override
public void initialize(UimaContext aContext)
throws ResourceInitializationException
public void initialize(UimaContext aContext) throws ResourceInitializationException
{
super.initialize(aContext);

modelProvider = new ModelProviderBase<Analyzer>() {
modelProvider = new ModelProviderBase<Analyzer>()
{
{
setContextObject(CogrooPosTagger.this);

Expand All @@ -116,8 +112,7 @@ public void initialize(UimaContext aContext)
}

@Override
protected Analyzer produceResource(URL aUrl)
throws IOException
protected Analyzer produceResource(URL aUrl) throws IOException
{
Properties props = getAggregatedProperties();

Expand All @@ -133,13 +128,11 @@ protected Analyzer produceResource(URL aUrl)
}
};

mappingProvider = MappingProviderFactory.createPosMappingProvider(this, posMappingLocation,
"bosque", language);
mappingProvider = createPosMappingProvider(this, posMappingLocation, "bosque", language);
}

@Override
public void process(JCas aJCas)
throws AnalysisEngineProcessException
public void process(JCas aJCas) throws AnalysisEngineProcessException
{
CAS cas = aJCas.getCas();
modelProvider.configure(cas);
Expand Down
Loading

0 comments on commit c217616

Please sign in to comment.