Add FileCollectorPlugin and respective test

dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/collection/CollectorWorker.java

@@ -7,6 +7,7 @@ import java.io.IOException;
 import java.util.Optional;
 import java.util.concurrent.atomic.AtomicInteger;
 
+import eu.dnetlib.dhp.collection.plugin.file.FileCollectorPlugin;
 import eu.dnetlib.dhp.collection.plugin.file.FileGZipCollectorPlugin;
 import org.apache.hadoop.fs.FileSystem;
 import org.apache.hadoop.fs.Path;
@@ -115,6 +116,8 @@ public class CollectorWorker extends ReportingJob {
 				return new OaiCollectorPlugin(clientParams);
 			case rest_json2xml:
 				return new RestCollectorPlugin(clientParams);
+			case file:
+				return new FileCollectorPlugin();
 			case fileGZip:
 				return new FileGZipCollectorPlugin();
 			case other:

dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/collection/plugin/CollectorPlugin.java

@@ -10,7 +10,7 @@ import eu.dnetlib.dhp.common.collection.CollectorException;
 public interface CollectorPlugin {
 
 	enum NAME {
-		oai, other, rest_json2xml, fileGZip;
+		oai, other, rest_json2xml, file, fileGZip;
 
 		public enum OTHER_NAME {
 			mdstore_mongodb_dump, mdstore_mongodb

dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/collection/plugin/file/FileCollectorPlugin.java

@@ -0,0 +1,25 @@
+package eu.dnetlib.dhp.collection.plugin.file;
+
+import eu.dnetlib.dhp.common.collection.CollectorException;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.BufferedInputStream;
+import java.io.FileInputStream;
+
+public class FileCollectorPlugin extends AbstractSplittedRecordPlugin {
+
+    private static final Logger log = LoggerFactory.getLogger(FileCollectorPlugin.class);
+
+    @Override
+    protected BufferedInputStream getBufferedInputStream(final String baseUrl) throws CollectorException {
+
+        log.info("baseUrl: {}", baseUrl);
+
+        try {
+            return new BufferedInputStream(new FileInputStream(baseUrl));
+        } catch (Exception e) {
+            throw new CollectorException("Error reading file " + baseUrl, e);
+        }
+    }
+}

dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/collection/plugin/file/FileGZipCollectorPlugin.java

@@ -22,8 +22,7 @@ public class FileGZipCollectorPlugin extends AbstractSplittedRecordPlugin {
             GZIPInputStream stream = new GZIPInputStream(new FileInputStream(baseUrl));
             return new BufferedInputStream(stream);
         } catch (Exception e) {
-            e.printStackTrace();
-            throw new CollectorException(e);
+            throw new CollectorException("Error reading file " + baseUrl, e);
         }
     }
 }

dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/collection/plugin/file/FileCollectorPluginTest.java

@@ -0,0 +1,53 @@
+package eu.dnetlib.dhp.collection.plugin.file;
+
+import eu.dnetlib.dhp.collection.ApiDescriptor;
+import eu.dnetlib.dhp.common.aggregation.AggregatorReport;
+import eu.dnetlib.dhp.common.collection.CollectorException;
+import org.junit.jupiter.api.Assertions;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.HashMap;
+import java.util.stream.Stream;
+
+public class FileCollectorPluginTest {
+
+    private static final Logger log = LoggerFactory.getLogger(FileGZipCollectorPluginTest.class);
+
+    private final ApiDescriptor api = new ApiDescriptor();
+    private FileCollectorPlugin plugin;
+
+    private static final String SPLIT_ON_ELEMENT = "repository";
+
+    @BeforeEach
+    public void setUp() {
+
+        final String gzipFile = this
+                .getClass()
+                .getResource("/eu/dnetlib/dhp/collection/plugin/file/opendoar.xml")
+                .getFile();
+
+        api.setBaseUrl(gzipFile);
+
+        HashMap<String, String> params = new HashMap<>();
+        params.put("splitOnElement", SPLIT_ON_ELEMENT);
+
+        api.setParams(params);
+
+        plugin = new FileCollectorPlugin();
+    }
+
+    @Test
+    void test() throws CollectorException {
+
+        final Stream<String> stream = plugin.collect(api, new AggregatorReport());
+
+        stream.limit(10).forEach(s -> {
+            Assertions.assertTrue(s.length() > 0);
+            log.info(s);
+        });
+    }
+}
+

dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/collection/plugin/file/FileGZipCollectorPluginTest.java

@@ -26,7 +26,7 @@ public class FileGZipCollectorPluginTest {
 
         final String gzipFile = this
                 .getClass()
-                .getResource("/eu/dnetlib/dhp/collection/plugin/file/gzip/opendoar.xml.gz")
+                .getResource("/eu/dnetlib/dhp/collection/plugin/file/opendoar.xml.gz")
                 .getFile();
 
         api.setBaseUrl(gzipFile);