sabeel
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

implemented methods to extract fulltext link from an API call

This commit is contained in:
sab 2024-09-11 14:57:38 +02:00
parent df82f8beb9
commit ef6c90cc64
2 changed files with 36 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/transformation/xslt/DataFetcher.java
View File

@ -6,6 +6,7 @@ import net.sf.saxon.s9api.*;
import org.apache.commons.io.IOUtils;
import org.json.JSONObject;
import java.io.IOException;
import java.net.MalformedURLException;
import java.net.URL;
import java.util.ArrayList;
import java.util.List;
@ -18,16 +19,6 @@ import java.util.List;
public class DataFetcher implements ExtensionFunction, Serializable {
/**
* this method fetches JSON from a provided URL and returns it as Dublin Core
*/
public static List<String> fetchAndTransform(URL url) throws IOException {
JSONObject jsonObject = getJson(url);
List<String> authors = getAuthorsFromJson(jsonObject);
return transformListToDublinCore(authors);
}
/**
* This method fetches JSON object from a given URL
* @param url a url in the metadata for fetching authors in JSON format
@ -88,14 +79,39 @@ public class DataFetcher implements ExtensionFunction, Serializable {
return dublinCoreAuthors;
}
/**
* todo: add documentation
/**
* This is a public method which fetches authors and transform them into Dublin Core
*/
public static String getAndTransformAuthors(URL url) throws IOException{
return String.join(", ", transformListToDublinCore(getAuthorsFromJson(getJson(url))));
}
/**
* This method extracts link to fulltext from a given JSON
*
* @return
*/
static private String getLinkToFulltextFromJson(JSONObject jsonObject) throws MalformedURLException {
// note: Link to JSON containing fulltextlink is in "primary_file" attribute.
// And in the resultant JSON, “links->download” contains the URL to fulltext
return jsonObject
.getJSONObject("data")
.getJSONObject("links")
.getString("download");
}
/**
* This is a public method which fetches link to full text and returns it as a suitable format
*/
public static String getFullTextLinkAndTransform (URL url )throws IOException{
return getLinkToFulltextFromJson(getJson(url));
}
@Override
public QName getName() {
return null;

8
dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/transformation/xslt/DataFetcherTest.java
View File

@ -6,6 +6,7 @@ import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.Test;
import java.io.IOException;
import java.net.MalformedURLException;
import java.net.URI;
import java.net.URISyntaxException;
import java.net.URL;
@ -57,4 +58,11 @@ class DataFetcherTest {
}
@Test
void getLinkToFulltextFromJson() throws URISyntaxException, IOException {
URL linkToFullTextDocument = new URI("https://api.osf.io/v2/files/5de7c96f84c479000c7928af/?format=json").toURL();
System.out.println(DataFetcher.getFullTextLinkAndTransform(linkToFullTextDocument));
}
}