Skip to content

Commit

Permalink
MET-5806 sonarcloud improvement recommendations part II
Browse files Browse the repository at this point in the history
  • Loading branch information
@jeortizquan
jeortizquan committed Aug 7, 2024
1 parent 2a8d786 commit 5631f87
Show file tree
Hide file tree
Showing 6 changed files with 196 additions and 148 deletions.
305 changes: 176 additions & 129 deletions ...s-media-service/src/main/java/eu/europeana/metis/mediaprocessing/RdfDeserializerImpl.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -35,131 +35,89 @@
import org.xml.sax.SAXException;

/**
* This implements RDF deserialization functionality. The code that obtains the individual resources
* does not assume that we can convert the record to an EDM internal format. Link checking must also
* run on EDM external. We therefore use XPath expressions to obtain the required data.
*
* TODO use {@link eu.europeana.metis.schema.convert.RdfConversionUtils} - no org.jibx.runtime.*
* import should remain.
* This implements RDF deserialization functionality. The code that obtains the individual resources does not assume that we can
* convert the record to an EDM internal format. Link checking must also run on EDM external. We therefore use XPath expressions
* to obtain the required data.
* <p>
* TODO use {@link eu.europeana.metis.schema.convert.RdfConversionUtils} - no org.jibx.runtime.* import should remain.
*/
class RdfDeserializerImpl implements RdfDeserializer {

private final UnmarshallingContextWrapper unmarshallingContext = new UnmarshallingContextWrapper();

private static final String OEMBED_XPATH_CONDITION_IS_SHOWN_BY = "/rdf:RDF/ore:Aggregation/edm:isShownBy/@rdf:resource[/rdf:RDF/ore:Aggregation/edm:isShownBy/@rdf:resource =/rdf:RDF/edm:WebResource[svcs:has_service/@rdf:resource = /rdf:RDF/svcs:Service/@rdf:about and /rdf:RDF/svcs:Service/dcterms:conformsTo/@rdf:resource = \"https://oembed.com/\"]/@rdf:about]";
private static final String OEMBED_XPATH_CONDITION_HAS_VIEW = "/rdf:RDF/ore:Aggregation/edm:hasView/@rdf:resource[/rdf:RDF/ore:Aggregation/edm:hasView/@rdf:resource=/rdf:RDF/edm:WebResource[svcs:has_service/@rdf:resource = /rdf:RDF/svcs:Service/@rdf:about and /rdf:RDF/svcs:Service/dcterms:conformsTo/@rdf:resource = \"https://oembed.com/\"]/@rdf:about]";
private static final String XPATH_OBJECT = "/rdf:RDF/ore:Aggregation/edm:object/@rdf:resource";
private static final String XPATH_IS_SHOWN_BY = "/rdf:RDF/ore:Aggregation/edm:isShownBy/@rdf:resource";
private static final String XPATH_HAS_VIEW = "/rdf:RDF/ore:Aggregation/edm:hasView/@rdf:resource";
private static final String XPATH_IS_SHOWN_AT = "/rdf:RDF/ore:Aggregation/edm:isShownAt/@rdf:resource";
private static final String OEMBED_NAMESPACE = "https://oembed.com/";
private static final String XPATH_HAS_SERVICE =
"svcs:has_service/@rdf:resource = /rdf:RDF/svcs:Service/@rdf:about" +
" and /rdf:RDF/svcs:Service/dcterms:conformsTo/@rdf:resource";
private static final String XPATH_WEB_RESOURCE =
"/rdf:RDF/edm:WebResource[" + XPATH_HAS_SERVICE + " = \"" + OEMBED_NAMESPACE + "\"";
private static final String OEMBED_XPATH_CONDITION_IS_SHOWN_BY =
XPATH_IS_SHOWN_BY + "[" + XPATH_IS_SHOWN_BY + " = " + XPATH_WEB_RESOURCE + "]/@rdf:about]";
private static final String OEMBED_XPATH_CONDITION_HAS_VIEW = XPATH_HAS_VIEW
+ "[" + XPATH_HAS_VIEW + "=" + XPATH_WEB_RESOURCE + "]/@rdf:about]";

private final UnmarshallingContextWrapper unmarshallingContext = new UnmarshallingContextWrapper();
private final XPathExpressionWrapper getObjectExpression = new XPathExpressionWrapper(
xPath -> xPath.compile("/rdf:RDF/ore:Aggregation/edm:object/@rdf:resource"));
xPath -> xPath.compile(XPATH_OBJECT));
private final XPathExpressionWrapper getHasViewExpression = new XPathExpressionWrapper(
xPath -> xPath.compile("/rdf:RDF/ore:Aggregation/edm:hasView/@rdf:resource | " + OEMBED_XPATH_CONDITION_HAS_VIEW));
xPath -> xPath.compile(XPATH_HAS_VIEW + " | " + OEMBED_XPATH_CONDITION_HAS_VIEW));
private final XPathExpressionWrapper getIsShownAtExpression = new XPathExpressionWrapper(
xPath -> xPath.compile("/rdf:RDF/ore:Aggregation/edm:isShownAt/@rdf:resource"));
xPath -> xPath.compile(XPATH_IS_SHOWN_AT));
private final XPathExpressionWrapper getIsShownByExpression = new XPathExpressionWrapper(
xPath -> xPath.compile("/rdf:RDF/ore:Aggregation/edm:isShownBy/@rdf:resource | " + OEMBED_XPATH_CONDITION_IS_SHOWN_BY));
xPath -> xPath.compile(XPATH_IS_SHOWN_BY + " | " + OEMBED_XPATH_CONDITION_IS_SHOWN_BY));

private static class XPathExpressionWrapper extends
AbstractThreadSafeWrapper<XPathExpression, RdfDeserializationException> {

XPathExpressionWrapper(
ThrowingFunction<XPath, XPathExpression, XPathExpressionException> expressionCreator) {
super(() -> {
final XPathFactory factory;
synchronized (XPathFactory.class) {
factory = XPathFactory.newInstance();
}
final XPath xPath = factory.newXPath();
xPath.setNamespaceContext(new RdfNamespaceContext());
try {
return expressionCreator.apply(xPath);
} catch (XPathExpressionException e) {
throw new RdfDeserializationException("Could not initialize xpath expression.", e);
}
});
}

NodeList evaluate(Document document) throws RdfDeserializationException {
return process(compiledExpression -> {
try {
return (NodeList) compiledExpression.evaluate(document, XPathConstants.NODESET);
} catch (XPathExpressionException e) {
throw new RdfDeserializationException("Problem with deserializing RDF.", e);
}
});
}
private static List<RdfResourceEntry> convertToResourceEntries(
Map<String, Set<UrlType>> urlWithTypes) {
return urlWithTypes.entrySet().stream().map(RdfDeserializerImpl::convertToResourceEntry)
.toList();
}

private static class UnmarshallingContextWrapper extends
AbstractThreadSafeWrapper<IUnmarshallingContext, RdfDeserializationException> {

public UnmarshallingContextWrapper() {
super(() -> {
try {
return RdfBindingFactoryProvider.getBindingFactory().createUnmarshallingContext();
} catch (JiBXException e) {
throw new RdfDeserializationException("Problem creating deserializer.", e);
}
});
}
private static RdfResourceEntry convertToResourceEntry(Map.Entry<String, Set<UrlType>> entry) {
return new RdfResourceEntry(entry.getKey(), entry.getValue());
}

public RDF deserializeToRdf(InputStream inputStream) throws RdfDeserializationException {
return process(context -> {
try {
return (RDF) context.unmarshalDocument(inputStream, "UTF-8");
} catch (JiBXException e) {
throw new RdfDeserializationException("Problem with deserializing record to RDF.", e);
}
});
private static <R> R performDeserialization(byte[] input, DeserializationOperation<R> operation)
throws RdfDeserializationException {
try (InputStream inputStream = new ByteArrayInputStream(input)) {
return operation.performDeserialization(inputStream);
} catch (IOException e) {
throw new RdfDeserializationException("Problem with reading byte array - Shouldn't happen.", e);
}
}

@Override
public RdfResourceEntry getMainThumbnailResourceForMediaExtraction(byte[] input)
throws RdfDeserializationException {
throws RdfDeserializationException {
return performDeserialization(input, this::getMainThumbnailResourceForMediaExtraction);
}

@Override
public RdfResourceEntry getMainThumbnailResourceForMediaExtraction(InputStream inputStream)
throws RdfDeserializationException {
throws RdfDeserializationException {
return getMainThumbnailResourceForMediaExtraction(deserializeToDocument(inputStream))
.orElse(null);
}

private Optional<RdfResourceEntry> getMainThumbnailResourceForMediaExtraction(Document record)
throws RdfDeserializationException {

// Get the entries of the required types.
final Map<String, Set<UrlType>> resourceEntries = getResourceEntries(record,
Collections.singleton(UrlType.URL_TYPE_FOR_MAIN_THUMBNAIL_RESOURCE));

// If there is not exactly one, we return an empty optional.
if (resourceEntries.size() != 1) {
return Optional.empty();
}

// So there is exactly one. Convert and return.
return Optional.of(convertToResourceEntries(resourceEntries).get(0));
.orElse(null);
}

@Override
public List<RdfResourceEntry> getRemainingResourcesForMediaExtraction(byte[] input)
throws RdfDeserializationException {
throws RdfDeserializationException {
return performDeserialization(input, this::getRemainingResourcesForMediaExtraction);
}

@Override
public List<RdfResourceEntry> getRemainingResourcesForMediaExtraction(InputStream inputStream)
throws RdfDeserializationException {
throws RdfDeserializationException {

// Get all the resource entries.
final Document record = deserializeToDocument(inputStream);
final Map<String, Set<UrlType>> allResources = getResourceEntries(record,
UrlType.URL_TYPES_FOR_MEDIA_EXTRACTION);
final Document deserializedDocument = deserializeToDocument(inputStream);
final Map<String, Set<UrlType>> allResources = getResourceEntries(deserializedDocument,
UrlType.URL_TYPES_FOR_MEDIA_EXTRACTION);

// Find the main thumbnail resource if it exists and remove it from the result.
getMainThumbnailResourceForMediaExtraction(record).map(RdfResourceEntry::getResourceUrl)
.ifPresent(allResources::remove);
getMainThumbnailResourceForMediaExtraction(deserializedDocument).map(RdfResourceEntry::getResourceUrl)
.ifPresent(allResources::remove);

// Done.
return convertToResourceEntries(allResources);
Expand All @@ -175,46 +133,51 @@ public List<RdfResourceEntry> getResourceEntriesForLinkChecking(byte[] input)
public List<RdfResourceEntry> getResourceEntriesForLinkChecking(InputStream inputStream)
throws RdfDeserializationException {
return convertToResourceEntries(getResourceEntries(deserializeToDocument(inputStream),
UrlType.URL_TYPES_FOR_LINK_CHECKING));
UrlType.URL_TYPES_FOR_LINK_CHECKING));
}

private static List<RdfResourceEntry> convertToResourceEntries(
Map<String, Set<UrlType>> urlWithTypes) {
return urlWithTypes.entrySet().stream().map(RdfDeserializerImpl::convertToResourceEntry)
.toList();
@Override
public EnrichedRdf getRdfForResourceEnriching(byte[] input) throws RdfDeserializationException {
return performDeserialization(input, this::getRdfForResourceEnriching);
}

private static RdfResourceEntry convertToResourceEntry(Map.Entry<String, Set<UrlType>> entry) {
return new RdfResourceEntry(entry.getKey(), entry.getValue());
@Override
public EnrichedRdf getRdfForResourceEnriching(InputStream inputStream)
throws RdfDeserializationException {
return new EnrichedRdfImpl(unmarshallingContext.deserializeToRdf(inputStream));
}

Map<String, Set<UrlType>> getResourceEntries(Document document,
Set<UrlType> allowedUrlTypes) throws RdfDeserializationException {
final Map<String, Set<UrlType>> urls = new HashMap<>();
for (UrlType type : allowedUrlTypes) {
final Set<String> urlsForType = getUrls(document, type);
for (String url : urlsForType) {
urls.computeIfAbsent(url, k -> new HashSet<>()).add(type);
}
private Optional<RdfResourceEntry> getMainThumbnailResourceForMediaExtraction(Document document)
throws RdfDeserializationException {

// Get the entries of the required types.
final Map<String, Set<UrlType>> resourceEntries = getResourceEntries(document,
Collections.singleton(UrlType.URL_TYPE_FOR_MAIN_THUMBNAIL_RESOURCE));

// If there is not exactly one, we return an empty optional.
if (resourceEntries.size() != 1) {
return Optional.empty();
}
return urls;

// So there is exactly one. Convert and return.
return Optional.of(convertToResourceEntries(resourceEntries).get(0));
}

private Set<String> getUrls(Document document, UrlType type) throws RdfDeserializationException {

// Determine the right expression to apply.
final XPathExpressionWrapper expression =
switch (type) {
case OBJECT -> getObjectExpression;
case HAS_VIEW -> getHasViewExpression;
case IS_SHOWN_AT -> getIsShownAtExpression;
case IS_SHOWN_BY -> getIsShownByExpression;
};
switch (type) {
case OBJECT -> getObjectExpression;
case HAS_VIEW -> getHasViewExpression;
case IS_SHOWN_AT -> getIsShownAtExpression;
case IS_SHOWN_BY -> getIsShownByExpression;
};

// Evaluate the expression and convert the node list to a set of attribute values.
final NodeList nodes = expression.evaluate(document);
return IntStream.range(0, nodes.getLength()).mapToObj(nodes::item).map(Node::getNodeValue)
.collect(Collectors.toSet());
.collect(Collectors.toSet());
}

private Document deserializeToDocument(InputStream inputStream) throws RdfDeserializationException {
Expand All @@ -233,29 +196,113 @@ private Document deserializeToDocument(InputStream inputStream) throws RdfDeseri
}
}

@Override
public EnrichedRdf getRdfForResourceEnriching(byte[] input) throws RdfDeserializationException {
return performDeserialization(input, this::getRdfForResourceEnriching);
}
@FunctionalInterface
private interface DeserializationOperation<R> {

@Override
public EnrichedRdf getRdfForResourceEnriching(InputStream inputStream)
throws RdfDeserializationException {
return new EnrichedRdfImpl(unmarshallingContext.deserializeToRdf(inputStream));
/**
* Perform deserialization r.
*
* @param inputStream the input stream
* @return the r
* @throws RdfDeserializationException the rdf deserialization exception
*/
R performDeserialization(InputStream inputStream) throws RdfDeserializationException;
}

private static <R> R performDeserialization(byte[] input, DeserializationOperation<R> operation)
throws RdfDeserializationException {
try (InputStream inputStream = new ByteArrayInputStream(input)) {
return operation.performDeserialization(inputStream);
} catch (IOException e) {
throw new RdfDeserializationException("Problem with reading byte array - Shouldn't happen.", e);
private static class XPathExpressionWrapper extends
AbstractThreadSafeWrapper<XPathExpression, RdfDeserializationException> {

/**
* Instantiates a new X path expression wrapper.
*
* @param expressionCreator the expression creator
*/
XPathExpressionWrapper(
ThrowingFunction<XPath, XPathExpression, XPathExpressionException> expressionCreator) {
super(() -> {
final XPathFactory factory;
synchronized (XPathFactory.class) {
factory = XPathFactory.newInstance();
}
final XPath xPath = factory.newXPath();
xPath.setNamespaceContext(new RdfNamespaceContext());
try {
return expressionCreator.apply(xPath);
} catch (XPathExpressionException e) {
throw new RdfDeserializationException("Could not initialize xpath expression.", e);
}
});
}

/**
* Evaluate node list.
*
* @param document the document
* @return the node list
* @throws RdfDeserializationException the rdf deserialization exception
*/
NodeList evaluate(Document document) throws RdfDeserializationException {
return process(compiledExpression -> {
try {
return (NodeList) compiledExpression.evaluate(document, XPathConstants.NODESET);
} catch (XPathExpressionException e) {
throw new RdfDeserializationException("Problem with deserializing RDF.", e);
}
});
}
}

@FunctionalInterface
private interface DeserializationOperation<R> {
private static class UnmarshallingContextWrapper extends
AbstractThreadSafeWrapper<IUnmarshallingContext, RdfDeserializationException> {

R performDeserialization(InputStream inputStream) throws RdfDeserializationException;
/**
* Instantiates a new Unmarshalling context wrapper.
*/
public UnmarshallingContextWrapper() {
super(() -> {
try {
return RdfBindingFactoryProvider.getBindingFactory().createUnmarshallingContext();
} catch (JiBXException e) {
throw new RdfDeserializationException("Problem creating deserializer.", e);
}
});
}

/**
* Deserialize to rdf rdf.
*
* @param inputStream the input stream
* @return the rdf
* @throws RdfDeserializationException the rdf deserialization exception
*/
public RDF deserializeToRdf(InputStream inputStream) throws RdfDeserializationException {
return process(context -> {
try {
return (RDF) context.unmarshalDocument(inputStream, "UTF-8");
} catch (JiBXException e) {
throw new RdfDeserializationException("Problem with deserializing record to RDF.", e);
}
});
}
}

/**
* Gets resource entries.
*
* @param document the document
* @param allowedUrlTypes the allowed url types
* @return the resource entries
* @throws RdfDeserializationException the rdf deserialization exception
*/
Map<String, Set<UrlType>> getResourceEntries(Document document,
Set<UrlType> allowedUrlTypes) throws RdfDeserializationException {
final Map<String, Set<UrlType>> urls = new HashMap<>();
for (UrlType type : allowedUrlTypes) {
final Set<String> urlsForType = getUrls(document, type);
for (String url : urlsForType) {
urls.computeIfAbsent(url, k -> new HashSet<>()).add(type);
}
}
return urls;
}
}
Loading

0 comments on commit 5631f87

Please sign in to comment.