toward supporting [elton ls] in --prov-mode

globalbioticinteractions · Dec 24, 2024 · 372ab0a · 372ab0a
1 parent 9c5316f
commit 372ab0a
Show file tree

Hide file tree

Showing 3 changed files with 55 additions and 11 deletions.
diff --git a/src/main/java/org/globalbioticinteractions/elton/cmd/CmdList.java b/src/main/java/org/globalbioticinteractions/elton/cmd/CmdList.java
@@ -1,12 +1,15 @@
 package org.globalbioticinteractions.elton.cmd;
 
 import org.eol.globi.util.ResourceServiceRemote;
-
 import org.globalbioticinteractions.dataset.DatasetRegistry;
 import org.globalbioticinteractions.dataset.DatasetRegistryException;
 import org.globalbioticinteractions.dataset.DatasetRegistryGitHubArchive;
 import org.globalbioticinteractions.dataset.DatasetRegistryProxy;
 import org.globalbioticinteractions.dataset.DatasetRegistryZenodo;
+import org.globalbioticinteractions.elton.store.AccessLogger;
+import org.globalbioticinteractions.elton.store.ActivityListener;
+import org.globalbioticinteractions.elton.store.LocalPathToHashIRI;
+import org.globalbioticinteractions.elton.util.DatasetRegistryUtil;
 import picocli.CommandLine;
 
 import java.io.File;
@@ -37,21 +40,39 @@ public String getDescription() {
     public void run(PrintStream out) {
         InputStreamFactoryLogging inputStreamFactory = createInputStreamFactory();
 
-        DatasetRegistry registryLocal = getDatasetRegistry(inputStreamFactory, getActivityListener());
+        DatasetRegistry registryLocal = getDatasetRegistryWithProv();
+
+        ActivityListener activityListener =
+                getEnableProvMode()
+                        ? getActivityListenerWithProv()
+                        : new AccessLogger(DatasetRegistryUtil.NAMESPACE_LOCAL, getProvDir());
+
 
         File cacheDir = new File(getDataDir());
 
+        DatasetRegistryUtil.ResourceServiceListening resourceServiceRemote
+                = new DatasetRegistryUtil.ResourceServiceListening(
+                getActivityIdFactory(),
+                activityListener,
+                getActivityContext(),
+                new ResourceServiceRemote(inputStreamFactory, cacheDir),
+                new LocalPathToHashIRI(new File(getDataDir()))
+        );
+
+
         List<DatasetRegistry> onlineAndOffline = Arrays.asList(
-                new DatasetRegistryZenodo(new ResourceServiceRemote(inputStreamFactory, cacheDir)),
-                new DatasetRegistryGitHubArchive(new ResourceServiceRemote(inputStreamFactory, cacheDir)),
+                new DatasetRegistryZenodo(resourceServiceRemote),
+                new DatasetRegistryGitHubArchive(resourceServiceRemote),
                 registryLocal
         );
 
         List<DatasetRegistry> registries = isOnline() ? onlineAndOffline : Collections.singletonList(registryLocal);
 
+        PrintStream dataSink = getDataSink(out);
+
         DatasetRegistry registry = new DatasetRegistryProxy(registries);
         try {
-            registry.findNamespaces(out::println);
+            registry.findNamespaces(dataSink::println);
         } catch (
                 DatasetRegistryException e) {
             throw new RuntimeException(e);

diff --git a/src/main/java/org/globalbioticinteractions/elton/util/DatasetRegistryUtil.java b/src/main/java/org/globalbioticinteractions/elton/util/DatasetRegistryUtil.java
@@ -130,7 +130,7 @@ private static DatasetRegistry forCacheOrLocalDir(String dataDir,
                                                       ResourceService resourceServiceRemote,
                                                       ContentPathFactory contentPathFactory,
                                                       ProvenancePathFactory provenancePathFactory,
-                                                      ActivityListener dereferenceListener,
+                                                      ActivityListener activityListener,
                                                       ActivityContext ctx,
                                                       Supplier<IRI> activityIdFactory) {
         DatasetRegistry registry = forCache(
@@ -147,7 +147,7 @@ private static DatasetRegistry forCacheOrLocalDir(String dataDir,
                     contentPathFactory,
                     dataDir,
                     provDir,
-                    dereferenceListener,
+                    activityListener,
                     ctx,
                     activityIdFactory
             );

diff --git a/src/test/java/org/globalbioticinteractions/elton/cmd/CmdListTest.java b/src/test/java/org/globalbioticinteractions/elton/cmd/CmdListTest.java
@@ -1,5 +1,6 @@
 package org.globalbioticinteractions.elton.cmd;
 
+import org.apache.commons.lang3.StringUtils;
 import org.junit.Rule;
 import org.junit.Test;
 import org.junit.rules.TemporaryFolder;
@@ -9,10 +10,13 @@
 import java.io.IOException;
 import java.io.PrintStream;
 import java.net.URISyntaxException;
+import java.util.Arrays;
 
+import static org.hamcrest.CoreMatchers.containsString;
 import static org.hamcrest.CoreMatchers.startsWith;
 import static org.hamcrest.MatcherAssert.assertThat;
 import static org.hamcrest.core.Is.is;
+import static org.hamcrest.core.IsNot.not;
 
 public class CmdListTest {
 
@@ -22,27 +26,46 @@ public class CmdListTest {
     @Test
     public void listOffline() throws URISyntaxException, IOException {
         String cacheDir = CmdTestUtil.cacheDirTest(tmpFolder);
-        ByteArrayOutputStream out = runCmd(cacheDir);
+        ByteArrayOutputStream out = runCmd(cacheDir, false);
         assertThat(out.toString(), startsWith("globalbioticinteractions/template-dataset"));
     }
 
+    @Test
+    public void listOfflineWithProv() throws URISyntaxException, IOException {
+        String cacheDir = CmdTestUtil.cacheDirTest(tmpFolder);
+        ByteArrayOutputStream out = runCmd(cacheDir, true);
+        String actual = out.toString();
+        assertThat(actual, not(startsWith("globalbioticinteractions/template-dataset")));
+
+        String[] lines = StringUtils.split(actual, '\n');
+        assertThat(lines.length, is(19));
+        assertThat(lines[0], startsWith("<https://globalbioticinteractions.org/elton> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.w3.org/ns/prov#SoftwareAgent>"));
+        assertThat(lines[lines.length - 1], containsString("<http://www.w3.org/ns/prov#endedAtTime>"));
+
+        long numberOfWasDerivedFromCount = Arrays.stream(lines).filter(line -> StringUtils.contains(line, "wasDerivedFrom")).count();
+
+        assertThat(numberOfWasDerivedFromCount, not(is(0L)));
+    }
+
     @Test
     public void listOfflineNonExistingCacheDir() throws URISyntaxException {
         String cacheDirNonExisting = "this/does/not/exist";
         assertThat(new File(cacheDirNonExisting).exists(), is(false));
-        ByteArrayOutputStream out = runCmd(cacheDirNonExisting);
+        ByteArrayOutputStream out = runCmd(cacheDirNonExisting, false);
         assertThat(out.toString(), is("local\n"));
         assertThat(new File(cacheDirNonExisting).exists(), is(false));
     }
 
-    private ByteArrayOutputStream runCmd(String cacheDir) {
+    private ByteArrayOutputStream runCmd(String cacheDir, boolean enableProvMode) {
         CmdList cmd = new CmdList();
         cmd.setDataDir(cacheDir);
         cmd.setProvDir(cacheDir);
+        cmd.setEnableProvMode(enableProvMode);
 
         ByteArrayOutputStream out1 = new ByteArrayOutputStream();
         PrintStream out = new PrintStream(out1);
-        cmd.run(out);
+        cmd.setStdout(out);
+        cmd.run();
         return out1;
     }