126 lines
3.6 KiB
Java
126 lines
3.6 KiB
Java
package eu.dnetlib.data.collector.plugins.datacite;
|
|
|
|
|
|
import java.io.ByteArrayOutputStream;
|
|
import java.io.IOException;
|
|
import java.net.URL;
|
|
import java.util.ArrayDeque;
|
|
import java.util.Iterator;
|
|
import java.util.Objects;
|
|
import java.util.Queue;
|
|
import java.util.zip.DataFormatException;
|
|
import java.util.zip.Inflater;
|
|
|
|
import com.google.gson.Gson;
|
|
import com.google.gson.GsonBuilder;
|
|
import eu.dnetlib.data.collector.plugins.datacite.schema.DataciteSchema;
|
|
import eu.dnetlib.data.collector.plugins.datacite.schema.Result;
|
|
import org.apache.commons.codec.binary.Base64;
|
|
import org.apache.commons.io.IOUtils;
|
|
import org.apache.commons.lang3.StringUtils;
|
|
|
|
public class DataciteESIterator implements Iterator<String> {
|
|
|
|
|
|
private final long timestamp;
|
|
|
|
private String scrollId;
|
|
|
|
private Queue<String> currentPage;
|
|
|
|
private final Gson g = new GsonBuilder().create();
|
|
|
|
private String baseURL = "http://ip-90-147-167-25.ct1.garrservices.it:5000";
|
|
|
|
private static final String START_PATH = "new_scan";
|
|
private static final String NEXT_PATH = "scan/%s";
|
|
|
|
|
|
public DataciteESIterator(long timestamp, String baseUrl) throws Exception {
|
|
this.timestamp = timestamp;
|
|
this.baseURL = baseUrl;
|
|
currentPage = new ArrayDeque<>();
|
|
startRequest();
|
|
}
|
|
|
|
private static String decompression(final Result r) {
|
|
try {
|
|
byte[] byteArray = Base64.decodeBase64(r.getBody().getBytes());
|
|
Inflater decompresser = new Inflater();
|
|
decompresser.setInput(byteArray);
|
|
ByteArrayOutputStream bos = new ByteArrayOutputStream(byteArray.length);
|
|
byte[] buffer = new byte[8192];
|
|
while (!decompresser.finished()) {
|
|
int size = decompresser.inflate(buffer);
|
|
bos.write(buffer, 0, size);
|
|
}
|
|
byte[] unzippeddata = bos.toByteArray();
|
|
decompresser.end();
|
|
|
|
return new String(unzippeddata);
|
|
} catch (DataFormatException e) {
|
|
return null;
|
|
}
|
|
|
|
}
|
|
|
|
private void fillQueue(final String hits) {
|
|
if (StringUtils.isBlank(hits) || "[]".equalsIgnoreCase(hits.trim()))
|
|
return;
|
|
try {
|
|
DataciteSchema datacitepage = g.fromJson(hits, DataciteSchema.class);
|
|
this.scrollId = datacitepage.getScrollId();
|
|
datacitepage.getResult().stream().map(DataciteESIterator::decompression).filter(Objects::nonNull).forEach(this.currentPage::add);
|
|
} catch (Throwable e) {
|
|
System.out.println(hits);
|
|
e.printStackTrace();
|
|
}
|
|
}
|
|
|
|
private void startRequest() throws Exception {
|
|
String url = baseURL+"/"+START_PATH;
|
|
final URL startUrl = new URL(timestamp >0 ? url + "?timestamp="+timestamp : url);
|
|
fillQueue(IOUtils.toString(startUrl.openStream()));
|
|
}
|
|
|
|
private void getNextPage() throws IOException {
|
|
String url = baseURL+"/"+NEXT_PATH;
|
|
final URL startUrl = new URL(String.format(url,scrollId));
|
|
fillQueue(IOUtils.toString(startUrl.openStream()));
|
|
}
|
|
|
|
|
|
@Override
|
|
public boolean hasNext() {
|
|
return currentPage.size() >0;
|
|
}
|
|
|
|
@Override
|
|
public String next() {
|
|
|
|
if (currentPage.size() == 0) {
|
|
|
|
return null;
|
|
}
|
|
|
|
String nextItem = currentPage.remove();
|
|
if (currentPage.size() == 0) {
|
|
try {
|
|
getNextPage();
|
|
} catch (Throwable e) {
|
|
throw new RuntimeException(e);
|
|
}
|
|
}
|
|
|
|
return nextItem;
|
|
}
|
|
|
|
public String getBaseURL() {
|
|
return baseURL;
|
|
}
|
|
|
|
public void setBaseURL(final String baseURL) {
|
|
this.baseURL = baseURL;
|
|
}
|
|
}
|