D-Net

package eu.dnetlib.msro.eagle.workflows.nodes.transform.tmid;

import javax.xml.ws.wsaddressing.W3CEndpointReference;

import com.googlecode.sarasvati.Arc;

import com.googlecode.sarasvati.NodeToken;

import eu.dnetlib.enabling.resultset.MappedResultSetFactory;

import eu.dnetlib.enabling.resultset.client.utils.EPRUtils;

import eu.dnetlib.msro.rmi.MSROException;

import eu.dnetlib.msro.workflows.nodes.SimpleJobNode;

public class TrismegistosInjectionJobNode extends SimpleJobNode {

	private String inputEprParam;

	private String outputEprParam;

	private String tmIdListPath;

	private MappedResultSetFactory mappedResultSetFactory;

	@Override

	protected String execute(NodeToken token) throws Exception {

		final String inputEpr = token.getEnv().getAttribute(inputEprParam);

		if ((inputEpr == null) || inputEpr.isEmpty()) throw new MSROException("InputEprParam (" + inputEprParam + ") not found in ENV");

		final W3CEndpointReference epr = getMappedResultSetFactory().createMappedResultSet(new EPRUtils().getEpr(inputEpr), 

				new TrismegistosInjectionUnaryFunction(tmIdListPath));

		token.getEnv().setAttribute(outputEprParam, epr.toString());

		return Arc.DEFAULT_ARC;

	}

	public String getInputEprParam() {

		return inputEprParam;

	}

	public void setInputEprParam(String inputEprParam) {

		this.inputEprParam = inputEprParam;

	}

	public String getOutputEprParam() {

		return outputEprParam;

	}

	public void setOutputEprParam(String outputEprParam) {

		this.outputEprParam = outputEprParam;

	}

	public String getTmIdListPath() {

		return tmIdListPath;

	}

	public void setTmIdListPath(String tmIdListPath) {

		this.tmIdListPath = tmIdListPath;

	}

	public MappedResultSetFactory getMappedResultSetFactory() {

		return mappedResultSetFactory;

	}

	public void setMappedResultSetFactory(MappedResultSetFactory mappedResultSetFactory) {

		this.mappedResultSetFactory = mappedResultSetFactory;

	}

}

package eu.dnetlib.msro.eagle.workflows.nodes.transform.tmid;

import java.io.IOException;

import java.io.StringReader;

import java.util.ArrayList;

import java.util.List;

import java.util.Map;

import javax.xml.stream.XMLInputFactory;

import javax.xml.stream.XMLStreamConstants;

import javax.xml.stream.XMLStreamException;

import javax.xml.stream.XMLStreamReader;

import org.dom4j.Document;

import org.dom4j.DocumentException;

import org.dom4j.DocumentHelper;

import org.dom4j.Element;

import org.dom4j.Node;

import org.dom4j.io.SAXReader;

import org.springframework.core.io.FileSystemResource;

import eu.dnetlib.miscutils.functional.UnaryFunction;

public class TrismegistosInjectionUnaryFunction implements UnaryFunction<String, String> {

	private SAXReader reader;

	private static Map<String, List<String>> tmIdToLocalIdsMap;

	private static Map<String, String> localIdToTmIdMap;

	public TrismegistosInjectionUnaryFunction(String tmIdListPath) {

		this.reader = new SAXReader();

		FileSystemResource tmIds = new FileSystemResource(tmIdListPath);

		XMLInputFactory factory = XMLInputFactory.newInstance();

		try {

			XMLStreamReader reader = factory.createXMLStreamReader(tmIds.getInputStream());

			boolean isSafeToGetNextXmlElement = true;

			while (isSafeToGetNextXmlElement) {

				if (reader.getEventType() == XMLStreamConstants.START_ELEMENT) {

					if("tmid".equals(reader.getLocalName())){

				        parseTmEvent(reader);

				    }

				}

				if (reader.hasNext()) {

	                reader.next();

	            } else {

	                isSafeToGetNextXmlElement = false;

	                break;

	            }

			}

		} catch (IOException e) {

			// TODO Auto-generated catch block

			e.printStackTrace();

		} catch (XMLStreamException e) {

			// TODO Auto-generated catch block

			e.printStackTrace();

		}

	}

	private void parseTmEvent(XMLStreamReader reader) {

		String tmId = reader.getAttributeValue(null, "id");

		List<String> localIds = new ArrayList<String>();

		boolean isSafeToGetNextXmlElement = true;

		while(isSafeToGetNextXmlElement) {

			if (reader.getEventType() == XMLStreamConstants.END_ELEMENT) {

                break;

            } else {

            	if (reader.getEventType() == XMLStreamConstants.START_ELEMENT) {

            		if("link".equals(reader.getLocalName())){

				        String localId = parseLinkEvent(reader);

				        localIdToTmIdMap.put(localId, tmId);

				        localIds.add(localId);

				    }

            	}

            }

			try{

				if (reader.hasNext()) {

	                reader.next();

	            } else {

	                isSafeToGetNextXmlElement = false;

	                break;

	            }

			} catch (XMLStreamException e) {

				e.printStackTrace();

			}

		}

		tmIdToLocalIdsMap.put(tmId, localIds);

	}

	private String parseLinkEvent(XMLStreamReader reader) {

		String cp = reader.getAttributeValue(null, "cp");

		String val = "";

		boolean isSafeToGetNextXmlElement = true;

        while(isSafeToGetNextXmlElement) {

            if (reader.getEventType() == XMLStreamConstants.END_ELEMENT) {

                break;

            } else if (reader.getEventType() == XMLStreamConstants.CHARACTERS) {

                val = reader.getText();

            }

            try{

				if (reader.hasNext()) {

	                reader.next();

	            } else {

	                isSafeToGetNextXmlElement = false;

	                break;

	            }

			} catch (XMLStreamException e) {

				e.printStackTrace();

			}

        }

        return (cp + "::" + val);

	}

	@Override

	public String evaluate(String input) {

		try {

			Document doc = reader.read(new StringReader(input));

			Node entityType = doc.selectSingleNode("//*[local-name()='entityType']");

			// TMid injection on Artifacts/Inscriptions

			if ("artifact".equals(entityType.getText())) {

				Node recordSourceInfo = doc.selectSingleNode("//*[local-name()='eagleObject']/*[local-name()='recordSourceInfo']");

				String localIdKey = (recordSourceInfo.selectSingleNode("@providerAcronym")).getText() + "::" + (recordSourceInfo.selectSingleNode("./text()")).getText();

				String tmId = localIdToTmIdMap.get(localIdKey);

				if (tmId != null) {

					// TMid found! Prepare for injection..

					List<String> alternateIds = tmIdToLocalIdsMap.get(tmId);

					Element inscription = (Element) doc.selectSingleNode("//*[local-name()='inscription']");

					Element injectedHasTmId = prepareInjectedElement(tmId, alternateIds);

					replaceElement(inscription, (Element) inscription.selectSingleNode("hasTmId"), injectedHasTmId);

					return doc.asXML();

				} else {

					// There is no TM ID for this localID. Just return the input string here..

					return input;

				}

			}

			// TMid injection on other items with rel to Artifacts

			if ("visual".equals(entityType.getText()) || "documental".equals(entityType.getText())) {

				Node recordSourceInfo = doc.selectSingleNode("//*[local-name()='hasArtifact']/*[local-name()='recordSourceInfo']");

				String localIdKey = (recordSourceInfo.selectSingleNode("@providerAcronym")).getText() + "::" + (recordSourceInfo.selectSingleNode("./text()")).getText();

				String tmId = localIdToTmIdMap.get(localIdKey);

				if (tmId != null) {

					// TMid found! Prepare for injection..

					List<String> alternateIds = tmIdToLocalIdsMap.get(tmId);

					Element inscription = (Element) doc.selectSingleNode("//*[local-name()='hasArtifact']");

					Element injectedHasTmId = prepareInjectedElement(tmId, alternateIds);

					replaceElement(inscription, (Element) inscription.selectSingleNode("hasTmId"), injectedHasTmId);

					return doc.asXML();

				} else {

					// There is no TM ID for this localID. Just return the input string here..

					return input;

				}

			}

		} catch (DocumentException e) {

			// TODO Auto-generated catch block

			e.printStackTrace();

		}

		return input;

	}

	private Element prepareInjectedElement(String tmId, List<String> alternateIds) {

		Document document = DocumentHelper.createDocument();

		Element hasTmId = document.addElement( "hasTmId" );

		hasTmId.addElement("tmId").addText(tmId);

		for (String alternateId : alternateIds) {

			String[] tokens = alternateId.split("::");

			hasTmId.addElement("alternateId").addAttribute("providerAcronym", tokens[0]).addAttribute("localId", tokens[1]);

		}

		return (Element) document.selectSingleNode("hasTmId");

	}

	private void replaceElement(Element parent, Element oldElement, Element newElement) {

		List parentContent = parent.content();

		int index = parentContent.indexOf(oldElement);

		parentContent.set(index, newElement);

	}

}

package eu.dnetlib.msro.eagle.workflows.nodes.transform.editions;

import java.io.IOException;

import java.io.StringReader;

import java.util.Map;

import org.apache.commons.logging.Log;

import org.apache.commons.logging.LogFactory;

import org.dom4j.Document;

import org.dom4j.DocumentException;

import org.dom4j.Element;

import org.dom4j.Node;

import org.dom4j.io.SAXReader;

import org.springframework.core.io.ClassPathResource;

import eu.dnetlib.miscutils.functional.UnaryFunction;

import eu.dnetlib.miscutils.functional.xml.ApplyXslt;

public class EditionXsltUnaryFunction implements UnaryFunction<String, String> {

	private static final Log log = LogFactory.getLog(EditionXsltUnaryFunction.class);

	private ApplyXslt xslt;

	private SAXReader reader;

	public EditionXsltUnaryFunction(final String xsltClasspath, Map<String, String> xsltParams) throws IOException {

		this.xslt = new ApplyXslt(new ClassPathResource(xsltClasspath), xsltParams);

		this.reader = new SAXReader();

	}

	/**

	 * xPath the <metadata> node and applies xslt 2.0 transformation

	 */

	@Override

	public String evaluate(String input) {

		try {

			Document doc = reader.read(new StringReader(input));

			Node result = doc.selectSingleNode("//*[local-name()='metadata']/*");

			String edition = xslt.evaluate(result.asXML());

			result.detach();

			((Element) doc.selectSingleNode("//*[local-name()='metadata']")).add(reader.read(new StringReader(edition)).getRootElement());

			return doc.asXML();

		} catch (DocumentException e) {

			log.error("Problem creating dom4j document from \n" + input, e);

			throw new RuntimeException(e);

		}

	}

}

package eu.dnetlib.msro.eagle.workflows.nodes.transform.editions;

import java.util.HashMap;

import java.util.Map;

import javax.xml.ws.wsaddressing.W3CEndpointReference;

import com.googlecode.sarasvati.Arc;

import com.googlecode.sarasvati.NodeToken;

import eu.dnetlib.enabling.resultset.MappedResultSetFactory;

import eu.dnetlib.enabling.resultset.client.utils.EPRUtils;

import eu.dnetlib.msro.rmi.MSROException;

import eu.dnetlib.msro.workflows.nodes.SimpleJobNode;

/**

 * This class is intended to process records coming in MDStore compliant form (with header) 

 * and apply the xslt located at http://sourceforge.net/projects/epidoc/files/Example%20Stylesheets/

 * to the <metadata> blocks of each record.

 * @author Andrea Mannocci

 *

 */

public class ApplyEditionXsltJobNode extends SimpleJobNode {

	private String inputEprParam;

	private String outputEprParam;

	private String xsltClasspath;

	private MappedResultSetFactory mappedResultSetFactory;

	private final Map<String, String> xsltParams = new HashMap<String, String>();

	@Override

	protected String execute(final NodeToken token) throws Exception {

		final String inputEpr = token.getEnv().getAttribute(inputEprParam);

		if ((inputEpr == null) || inputEpr.isEmpty()) throw new MSROException("InputEprParam (" + inputEprParam + ") not found in ENV");

		for (String name : token.getFullEnv().getAttributeNames()) {

			xsltParams.put(name, token.getFullEnv().getAttribute(name));

		}

		for (String name : token.getEnv().getAttributeNames()) {

			xsltParams.put(name, token.getEnv().getAttribute(name));

		}

		xsltParams.putAll(parseJsonParameters(token));

		final W3CEndpointReference epr = mappedResultSetFactory.createMappedResultSet(new EPRUtils().getEpr(inputEpr), 

				new EditionXsltUnaryFunction(xsltClasspath, xsltParams));

		token.getEnv().setAttribute(outputEprParam, epr.toString());

		return Arc.DEFAULT_ARC;

	}

	public String getInputEprParam() {

		return inputEprParam;

	}

	public void setInputEprParam(String inputEprParam) {

		this.inputEprParam = inputEprParam;

	}

	public String getOutputEprParam() {

		return outputEprParam;

	}

	public void setOutputEprParam(String outputEprParam) {

		this.outputEprParam = outputEprParam;

	}

	public String getXsltClasspath() {

		return xsltClasspath;

	}

	public void setXsltClasspath(String xsltClasspath) {

		this.xsltClasspath = xsltClasspath;

	}

	public MappedResultSetFactory getMappedResultSetFactory() {

		return mappedResultSetFactory;

	}

	public void setMappedResultSetFactory(

			MappedResultSetFactory mappedResultSetFactory) {

		this.mappedResultSetFactory = mappedResultSetFactory;

	}

}

		class="eu.dnetlib.msro.eagle.workflows.nodes.transform.editions.ApplyEditionXsltJobNode"

	<bean id="wfNodeTrismegistosInject"

		class="eu.dnetlib.msro.eagle.workflows.nodes.transform.tmid.TrismegistosInjectionJobNode"

		scope="prototype" p:mappedResultSetFactory-ref="mappedResultSetFactory" />

		<ARC to="trismegistosInjection"/>

	</ARCS>

</NODE>

<NODE name="trismegistosInjection" type="TrismegistosInject">

	<DESCRIPTION>Unpack objects</DESCRIPTION>

	<PARAMETERS>

		<PARAM required="true" type="string" name="inputEprParam" managedBy="system">eagle_epr</PARAM>

		<PARAM required="true" type="string" name="outputEprParam" managedBy="system">injected_epr</PARAM>

		<PARAM required="true" type="string" name="tmIdListPath" managedBy="user">/var/lib/eagle/content/Trismegistos/tmid.xml</PARAM>

	</PARAMETERS>

	<ARCS>

		<PARAM required="true" type="string" name="eprParam" managedBy="system">injected_epr</PARAM>

		<PARAM required="true" type="string" name="xsltClasspath" managedBy="system">/eu/dnetlib/msro/eagle/workflows/xslt/editions/start-edition.xsl</PARAM>

		<ARC to="trismegistosInjection"/>

	</ARCS>

</NODE>

<NODE name="trismegistosInjection" type="TrismegistosInject">

	<DESCRIPTION>Unpack objects</DESCRIPTION>

	<PARAMETERS>

		<PARAM required="true" type="string" name="inputEprParam" managedBy="system">eagle_epr</PARAM>

		<PARAM required="true" type="string" name="outputEprParam" managedBy="system">injected_epr</PARAM>

		<PARAM required="true" type="string" name="tmIdListPath" managedBy="user">/var/lib/eagle/content/Trismegistos/tmid.xml</PARAM>

	</PARAMETERS>

	<ARCS>

		<PARAM required="true" type="string" name="eprParam" managedBy="system">injected_epr</PARAM>