use of org.wikidata.wdtk.datamodel.interfaces.EntityIdValue in project OpenRefine by OpenRefine.
the class QualifierCompatibilityScrutinizer method scrutinize.
@Override
public void scrutinize(Statement statement, EntityIdValue entityId, boolean added) {
if (!added) {
// not scrutinizing deleted statements
return;
}
PropertyIdValue statementProperty = statement.getClaim().getMainSnak().getPropertyId();
Set<PropertyIdValue> qualifiers = statement.getClaim().getQualifiers().stream().map(e -> e.getProperty()).collect(Collectors.toSet());
Set<PropertyIdValue> missingQualifiers = mandatoryQualifiers(statementProperty).stream().filter(p -> !qualifiers.contains(p)).collect(Collectors.toSet());
Set<PropertyIdValue> disallowedQualifiers = qualifiers.stream().filter(p -> !qualifierIsAllowed(statementProperty, p)).collect(Collectors.toSet());
for (PropertyIdValue missing : missingQualifiers) {
QAWarning issue = new QAWarning(missingMandatoryQualifiersType, statementProperty.getId() + "-" + missing.getId(), QAWarning.Severity.WARNING, 1);
issue.setProperty("statement_property_entity", statementProperty);
issue.setProperty("missing_property_entity", missing);
issue.setProperty("example_item_entity", entityId);
addIssue(issue);
}
for (PropertyIdValue disallowed : disallowedQualifiers) {
QAWarning issue = new QAWarning(disallowedQualifiersType, statementProperty.getId() + "-" + disallowed.getId(), QAWarning.Severity.WARNING, 1);
issue.setProperty("statement_property_entity", statementProperty);
issue.setProperty("disallowed_property_entity", disallowed);
issue.setProperty("example_item_entity", entityId);
addIssue(issue);
}
}
use of org.wikidata.wdtk.datamodel.interfaces.EntityIdValue in project OpenRefine by OpenRefine.
the class WbStatementExpr method evaluate.
public StatementEdit evaluate(ExpressionContext ctxt, EntityIdValue subject, PropertyIdValue propertyId) throws SkipSchemaExpressionException {
Snak mainSnak = null;
if (mainSnakValueExpr != null) {
Value mainSnakValue = mainSnakValueExpr.evaluate(ctxt);
mainSnak = Datamodel.makeValueSnak(propertyId, mainSnakValue);
} else {
// hack to make sure we have a non-null snak
mainSnak = Datamodel.makeNoValueSnak(propertyId);
}
// evaluate qualifiers
List<Snak> qualifiers = new ArrayList<Snak>(getQualifiers().size());
for (WbSnakExpr qExpr : getQualifiers()) {
try {
qualifiers.add(qExpr.evaluate(ctxt));
} catch (SkipSchemaExpressionException e) {
QAWarning warning = new QAWarning("ignored-qualifiers", null, QAWarning.Severity.INFO, 1);
warning.setProperty("example_entity", subject);
warning.setProperty("example_property_entity", mainSnak.getPropertyId());
ctxt.addWarning(warning);
}
}
List<SnakGroup> groupedQualifiers = groupSnaks(qualifiers);
Claim claim = Datamodel.makeClaim(subject, mainSnak, groupedQualifiers);
// evaluate references
List<Reference> references = new ArrayList<Reference>();
for (WbReferenceExpr rExpr : getReferences()) {
try {
references.add(rExpr.evaluate(ctxt));
} catch (SkipSchemaExpressionException e) {
QAWarning warning = new QAWarning("ignored-references", null, QAWarning.Severity.INFO, 1);
warning.setProperty("example_entity", subject);
warning.setProperty("example_property_entity", mainSnak.getPropertyId());
ctxt.addWarning(warning);
}
}
StatementRank rank = StatementRank.NORMAL;
return new StatementEdit(Datamodel.makeStatement(claim, references, rank, ""), merger, mode);
}
use of org.wikidata.wdtk.datamodel.interfaces.EntityIdValue in project OpenRefine by OpenRefine.
the class QuickStatementsUpdateScheduler method splitUpdate.
/**
* Separates out the statements which refer to new entities from the rest of the
* update. The resulting updates are stored in {@link referencingUpdates} and
* {@link updatesWithoutReferences}.
*
* @param update
* @throws ImpossibleSchedulingException
* if two new entity ids are referred to in the same statement
*/
protected void splitUpdate(TermedStatementEntityEdit update) throws ImpossibleSchedulingException {
TermedStatementEntityEditBuilder remainingUpdateBuilder = new TermedStatementEntityEditBuilder(update.getEntityId()).addLabels(update.getLabels(), true).addLabels(update.getLabelsIfNew(), false).addDescriptions(update.getDescriptions(), true).addDescriptions(update.getDescriptionsIfNew(), false).addAliases(update.getAliases());
Map<EntityIdValue, TermedStatementEntityEditBuilder> referencingUpdates = new HashMap<>();
for (StatementEdit statement : update.getStatementEdits()) {
Set<ReconEntityIdValue> pointers = extractor.extractPointers(statement.getStatement());
if (pointers.isEmpty()) {
remainingUpdateBuilder.addStatement(statement);
} else if (pointers.size() == 1 && !update.isNew()) {
EntityIdValue pointer = pointers.stream().findFirst().get();
TermedStatementEntityEditBuilder referencingBuilder = referencingUpdates.get(pointer);
if (referencingBuilder == null) {
referencingBuilder = new TermedStatementEntityEditBuilder(update.getEntityId());
}
referencingBuilder.addStatement(statement);
referencingUpdates.put(pointer, referencingBuilder);
} else if (pointers.size() == 1 && pointers.stream().findFirst().get().equals(update.getEntityId())) {
remainingUpdateBuilder.addStatement(statement);
} else {
throw new ImpossibleSchedulingException();
}
}
// Add the update that is not referring to anything to the schedule
TermedStatementEntityEdit pointerFree = remainingUpdateBuilder.build();
if (!pointerFree.isNull()) {
pointerFreeUpdates.add(pointerFree);
}
// Add the other updates to the map
for (Entry<EntityIdValue, TermedStatementEntityEditBuilder> entry : referencingUpdates.entrySet()) {
TermedStatementEntityEdit pointerUpdate = entry.getValue().build();
UpdateSequence pointerUpdatesForKey = pointerUpdates.get(entry.getKey());
if (pointerUpdatesForKey == null) {
pointerUpdatesForKey = new UpdateSequence();
}
pointerUpdatesForKey.add(pointerUpdate);
pointerUpdates.put(entry.getKey(), pointerUpdatesForKey);
}
}
use of org.wikidata.wdtk.datamodel.interfaces.EntityIdValue in project OpenRefine by OpenRefine.
the class QuickStatementsUpdateScheduler method schedule.
@Override
public List<TermedStatementEntityEdit> schedule(List<TermedStatementEntityEdit> updates) throws ImpossibleSchedulingException {
pointerUpdates = new HashMap<>();
pointerFreeUpdates = new UpdateSequence();
for (TermedStatementEntityEdit update : updates) {
splitUpdate(update);
}
// Reconstruct
List<TermedStatementEntityEdit> fullSchedule = new ArrayList<>();
Set<EntityIdValue> mentionedNewEntities = new HashSet<>(pointerUpdates.keySet());
for (TermedStatementEntityEdit update : pointerFreeUpdates.getUpdates()) {
fullSchedule.add(update);
UpdateSequence backPointers = pointerUpdates.get(update.getEntityId());
if (backPointers != null) {
fullSchedule.addAll(backPointers.getUpdates());
}
mentionedNewEntities.remove(update.getEntityId());
}
// as the entities would remain blank in this batch).
for (EntityIdValue missingId : mentionedNewEntities) {
fullSchedule.add(new TermedStatementEntityEditBuilder(missingId).build());
fullSchedule.addAll(pointerUpdates.get(missingId).getUpdates());
}
return fullSchedule;
}
use of org.wikidata.wdtk.datamodel.interfaces.EntityIdValue in project OpenRefine by OpenRefine.
the class UpdateSequence method add.
/**
* Adds a new update to the list, merging it with any existing one with the same
* subject.
*
* @param update
*/
public void add(TermedStatementEntityEdit update) {
EntityIdValue subject = update.getEntityId();
if (index.containsKey(subject)) {
int i = index.get(subject);
TermedStatementEntityEdit oldUpdate = updates.get(i);
updates.set(i, oldUpdate.merge(update));
} else {
index.put(subject, updates.size());
updates.add(update);
}
}
Aggregations