dnet-core/dnet-data-services/src/main/java/eu/dnetlib/data/mdstore/modular/BulkRecordMapperFactory.java

134 lines
4.4 KiB
Java

package eu.dnetlib.data.mdstore.modular;
import java.io.StringReader;
import java.io.StringWriter;
import java.util.List;
import javax.xml.namespace.QName;
import javax.xml.stream.XMLEventFactory;
import javax.xml.stream.XMLEventReader;
import javax.xml.stream.XMLEventWriter;
import javax.xml.stream.XMLInputFactory;
import javax.xml.stream.XMLOutputFactory;
import javax.xml.stream.XMLStreamException;
import javax.xml.stream.events.Namespace;
import javax.xml.stream.events.StartElement;
import javax.xml.stream.events.XMLEvent;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import com.google.common.base.Function;
import com.google.common.collect.Lists;
import eu.dnetlib.miscutils.collections.Pair;
import eu.dnetlib.miscutils.factory.Factory;
public class BulkRecordMapperFactory implements Factory<Function<String, Pair<String, String>>> {
private static final Log log = LogFactory.getLog(BulkRecordMapperFactory.class); // NOPMD by marko on 11/24/08 5:02 PM
protected static final String MD_RECORD = "mdRecord";
protected static final String MD_ID = "mdId";
protected static final String RECORD = "record";
protected ThreadLocal<XMLInputFactory> inputFactory = new ThreadLocal<XMLInputFactory>() {
@Override
protected XMLInputFactory initialValue() {
return XMLInputFactory.newInstance();
}
};
protected ThreadLocal<XMLOutputFactory> outputFactory = new ThreadLocal<XMLOutputFactory>() {
@Override
protected XMLOutputFactory initialValue() {
return XMLOutputFactory.newInstance();
}
};
protected ThreadLocal<XMLEventFactory> eventFactory = new ThreadLocal<XMLEventFactory>() {
@Override
protected XMLEventFactory initialValue() {
return XMLEventFactory.newInstance();
}
};
@Override
public Function<String, Pair<String, String>> newInstance() {
return new Function<String, Pair<String, String>>() {
private String mdId = null;
private String record = null;
@Override
public Pair<String, String> apply(String embeddedRecord) {
try {
final XMLEventReader parser = inputFactory.get().createXMLEventReader(new StringReader(embeddedRecord));
while (parser.hasNext()) {
final XMLEvent event = parser.nextEvent();
if (event != null && event.isStartElement()) {
final String localName = event.asStartElement().getName().getLocalPart();
if (MD_RECORD.equals(localName)) {
mdId = event.asStartElement().getAttributeByName(new QName(MD_ID)).getValue();
} else if (RECORD.equals(localName)) {
record = getRecord(embeddedRecord, parser);
}
}
}
} catch (final XMLStreamException e) {
log.error("error parsing record: " + embeddedRecord);
}
return new Pair<String, String>(mdId, record);
}
};
}
/**
* Copy the /indexRecord/result element and children, preserving namespace declarations etc.
*
* @param indexDocument
* @param results
* @param parser
* @throws XMLStreamException
*/
protected String getRecord(final String record, final XMLEventReader parser) throws XMLStreamException {
StringWriter results = new StringWriter();
final XMLEventWriter writer = outputFactory.get().createXMLEventWriter(results);
// TODO: newRecord should copy all the namespace prefixes setup in parents
// fortunately the only parent of the result element is the 'indexrecord', so it should be easy to get
// the namespaces declared on the root element (and fast)
final List<Namespace> namespaces = Lists.newArrayList(
eventFactory.get().createNamespace("dri", "http://www.driver-repository.eu/namespace/dri"),
eventFactory.get().createNamespace("dr", "http://www.driver-repository.eu/namespace/dr"),
eventFactory.get().createNamespace("xsi", "http://www.w3.org/2001/XMLSchema-instance"),
eventFactory.get().createNamespace("dc", "http://purl.org/dc/elements/1.1/"));
StartElement newRecord = eventFactory.get().createStartElement("", null, RECORD, null, namespaces.iterator());
// new root record
writer.add(newRecord);
// copy the rest as it is
while (parser.hasNext()) {
final XMLEvent resultEvent = parser.nextEvent();
// TODO: replace with depth tracking instead of close tag tracking.
if (resultEvent.isEndElement() && resultEvent.asEndElement().getName().getLocalPart().equals(RECORD)) {
writer.add(resultEvent);
break;
}
writer.add(resultEvent);
}
writer.close();
return results.toString();
}
}