D-Net

package eu.dnetlib.oai.mongo;

import java.io.ByteArrayInputStream;

import java.io.IOException;

import java.io.StringReader;

import java.util.Date;

import java.util.List;

import java.util.zip.ZipEntry;

import java.util.zip.ZipInputStream;

import javax.annotation.Resource;

import org.apache.commons.io.IOUtils;

import org.apache.commons.lang3.StringEscapeUtils;

import org.dom4j.Document;

import org.dom4j.DocumentException;

import org.dom4j.io.SAXReader;

import com.google.common.collect.Sets;

import com.mongodb.DBObject;

import eu.dnetlib.oai.conf.OAIConfigurationReader;

import eu.dnetlib.oai.info.RecordInfo;

import eu.dnetlib.rmi.provision.OaiPublisherRuntimeException;

/**

 * Helper class to generate a RecordInfo object from a Mongo DBObject.

 *

 * @author alessia

 */

public class RecordInfoGenerator {

	@Resource

	private MetadataExtractor metadataExtractor;

	@Resource

	private ProvenanceExtractor provenanceExtractor;

	@SuppressWarnings("unchecked")

	public RecordInfo transformDBObject(final DBObject object, final boolean includeBody) {

		if ((object == null) || object.keySet().isEmpty()) { return null; }

		String id = (String) object.get(OAIConfigurationReader.ID_FIELD);

		// need to escape the identifier, otherwise the XML breaks

		id = StringEscapeUtils.escapeXml11(id);

		final boolean deleted = (Boolean) object.get("deleted");

		final RecordInfo record = new RecordInfo();

		record.setIdentifier(id);

		record.setInternalId(object.get("_id").toString());

		record.setDatestamp((Date) object.get(OAIConfigurationReader.DATESTAMP_FIELD));

		record.setDeleted(deleted);

		final List<String> sets = (List<String>) object.get(OAIConfigurationReader.SET_FIELD);

		if (sets != null) {

			record.setSetspecs(Sets.newHashSet(sets));

		}

		if (includeBody && !deleted) {

			final String body = decompressRecord((byte[]) object.get(OAIConfigurationReader.BODY_FIELD));

			final SAXReader reader = new SAXReader();

			Document doc;

			try {

				doc = reader.read(new StringReader(body));

				record.setMetadata(this.metadataExtractor.apply(doc));

				record.setProvenance(this.provenanceExtractor.apply(doc));

			} catch (final DocumentException e) {

				throw new OaiPublisherRuntimeException(e);

			}

		}

		return record;

	}

	public String decompressRecord(final byte[] input) {

		try {

			final ByteArrayInputStream bis = new ByteArrayInputStream(input);

			final ZipInputStream zis = new ZipInputStream(bis);

			ZipEntry ze;

			ze = zis.getNextEntry();

			if (ze == null) { throw new OaiPublisherRuntimeException("cannot decompress null zip entry "); }

			if (!ze.getName()

					.equals(OAIConfigurationReader.BODY_FIELD)) { throw new OaiPublisherRuntimeException("cannot decompress zip entry name :" + ze.getName()); }

			return IOUtils.toString(zis);

		} catch (final IOException e) {

			throw new OaiPublisherRuntimeException(e);

		}

	}

	public MetadataExtractor getMetadataExtractor() {

		return this.metadataExtractor;

	}

	public void setMetadataExtractor(final MetadataExtractor metadataExtractor) {

		this.metadataExtractor = metadataExtractor;

	}

	public ProvenanceExtractor getProvenanceExtractor() {

		return this.provenanceExtractor;

	}

	public void setProvenanceExtractor(final ProvenanceExtractor provenanceExtractor) {

		this.provenanceExtractor = provenanceExtractor;

	}

}