D-Net

package eu.dnetlib.data.mdstore.modular;

import java.io.ByteArrayInputStream;

import java.util.HashMap;

import java.util.Map;

import java.util.Stack;

import javax.xml.stream.XMLInputFactory;

import javax.xml.stream.XMLStreamConstants;

import javax.xml.stream.XMLStreamException;

import javax.xml.stream.XMLStreamReader;

import org.apache.commons.logging.Log;

import org.apache.commons.logging.LogFactory;

import static eu.dnetlib.data.mdstore.modular.MDStoreConstants.*;

/**

 * This method outperforms SimpleRecordParser by a vast amount, especially since we are just getting stuff in the

 * header.

 * 

 * @author marko

 * 

 */

public class StreamingRecordParser implements RecordParser {

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

	private long ts;

	@Override

	public Map<String, String> parseRecord(String record) {

		try {

			XMLInputFactory factory = XMLInputFactory.newInstance();

			XMLStreamReader parser = factory.createXMLStreamReader(new ByteArrayInputStream(record.getBytes()));

			HashMap<String, String> res = new HashMap<String, String>();

			res.put(TIMESTAMP, String.valueOf(getTimestamp()));

			Stack<String> elementStack = new Stack<String>();

			elementStack.push("/");

			while (parser.hasNext()) {

				int event = parser.next();

				if (event == XMLStreamConstants.END_ELEMENT) {

					elementStack.pop();

				} else if (event == XMLStreamConstants.START_ELEMENT) {

					final String localName = parser.getLocalName();

					elementStack.push(localName);

					if (OBJIDENTIFIER.equals(localName)) {

						parser.next();

						res.put(ID, parser.getText().trim());

					} else if ("identifier".equals(localName) && "efgEntity".equals(grandParent(elementStack))) {

						if (!res.containsKey("originalId")) {

							parser.next();

//							log.info("ZZZZZZ OK: found identifier at right depth " + elementStack);

							res.put("originalId", parser.getText().trim());

						}

					}

					else if ("identifier".equals(localName)) {

//						log.info("ZZZZZZ: found identifier not at right depth " + elementStack + " grand parent " + grandParent(elementStack));

					}

					if (res.containsKey(ID) && res.containsKey("originalId"))

						return res;

				}

			}

			return res;

		} catch (XMLStreamException e) {

			throw new IllegalStateException(e);

		}

	}

	private String grandParent(Stack<String> elementStack) {

		if (elementStack.size() <= 3)

			return "";

		return elementStack.get(elementStack.size() - 3);

	}

	@Override

	public void setTimestamp(final long ts) {

		this.ts = ts;

		log.debug("RecordParser date set to "+ts);

	}

	@Override

	public long getTimestamp() {

		return ts;

	}

}