introduce default namespace for streaming globi.json; refactor namesp…

…ace constants
globalbioticinteractions · Jul 29, 2024 · 516a43e · 516a43e
1 parent 07f5bcc
commit 516a43e
Show file tree

Hide file tree

Showing 6 changed files with 32 additions and 28 deletions.
diff --git a/src/main/java/org/globalbioticinteractions/elton/cmd/CmdReview.java b/src/main/java/org/globalbioticinteractions/elton/cmd/CmdReview.java
@@ -114,7 +114,7 @@ public void run() {
                         new ResourceServiceLocalAndRemote(factory)
                 );
 
-                review("local", registryLocal, factory);
+                review(DatasetRegistryUtil.NAMESPACE_LOCAL, registryLocal, factory);
             }
 
             reviewCachedOrRemote(remoteNamespaces, factory);

diff --git a/src/main/java/org/globalbioticinteractions/elton/cmd/CmdStream.java b/src/main/java/org/globalbioticinteractions/elton/cmd/CmdStream.java
@@ -6,6 +6,7 @@
 import org.apache.commons.io.FileUtils;
 import org.apache.commons.io.IOUtils;
 import org.apache.commons.lang.StringUtils;
+import org.globalbioticinteractions.elton.util.DatasetRegistryUtil;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 import picocli.CommandLine;
@@ -39,25 +40,23 @@ public void run() {
             while ((line = reader.readLine()) != null) {
                 try {
                     JsonNode jsonNode = new ObjectMapper().readTree(line);
-                    if (jsonNode.has("namespace")) {
-                        String namespace = jsonNode.get("namespace").asText();
-                        if (StringUtils.isNotBlank(namespace)) {
-                            try {
-                                StreamingNamespaceConfigHandler namespaceHandler = new StreamingNamespaceConfigHandler(
-                                        jsonNode,
-                                        this.createInputStreamFactory(),
-                                        this.getCacheDir(),
-                                        this.getStderr(),
-                                        this.getStdout()
-                                );
-                                namespaceHandler.setShouldWriteHeader(isFirst.get());
-                                namespaceHandler.onNamespace(namespace);
-                                isFirst.set(false);
-                            } catch (Exception e) {
-                                LOG.error("failed to add dataset associated with namespace [" + namespace + "]", e);
-                            } finally {
-                                FileUtils.forceDelete(new File(this.getCacheDir()));
-                            }
+                    String namespace = jsonNode.at("/namespace").asText(DatasetRegistryUtil.NAMESPACE_LOCAL);
+                    if (StringUtils.isNotBlank(namespace)) {
+                        try {
+                            StreamingNamespaceConfigHandler namespaceHandler = new StreamingNamespaceConfigHandler(
+                                    jsonNode,
+                                    this.createInputStreamFactory(),
+                                    this.getCacheDir(),
+                                    this.getStderr(),
+                                    this.getStdout()
+                            );
+                            namespaceHandler.setShouldWriteHeader(isFirst.get());
+                            namespaceHandler.onNamespace(namespace);
+                            isFirst.set(false);
+                        } catch (Exception e) {
+                            LOG.error("failed to add dataset associated with namespace [" + namespace + "]", e);
+                        } finally {
+                            FileUtils.forceDelete(new File(this.getCacheDir()));
                         }
                     }
                 } catch (JsonProcessingException e) {

diff --git a/src/main/java/org/globalbioticinteractions/elton/cmd/DatasetRegistryFactoryImpl.java b/src/main/java/org/globalbioticinteractions/elton/cmd/DatasetRegistryFactoryImpl.java
@@ -10,6 +10,7 @@
 import org.globalbioticinteractions.dataset.DatasetRegistryGitHubArchive;
 import org.globalbioticinteractions.dataset.DatasetRegistryZenodo;
 import org.globalbioticinteractions.elton.util.DatasetRegistrySingleDir;
+import org.globalbioticinteractions.elton.util.DatasetRegistryUtil;
 
 import javax.annotation.Resource;
 import java.lang.reflect.Constructor;
@@ -22,9 +23,9 @@
 
 public class DatasetRegistryFactoryImpl implements DatasetRegistryFactory {
     private static final Map<String, Class<? extends DatasetRegistry>> REGISTRY_LOOKUP = MapUtils.unmodifiableMap(new TreeMap<String, Class<? extends DatasetRegistry>>() {{
-        put("local", DatasetRegistrySingleDir.class);
-        put("zenodo", DatasetRegistryZenodo.class);
-        put("github", DatasetRegistryGitHubArchive.class);
+        put(DatasetRegistryUtil.NAMESPACE_LOCAL, DatasetRegistrySingleDir.class);
+        put(DatasetRegistryUtil.NAMESPACE_ZENODO, DatasetRegistryZenodo.class);
+        put(DatasetRegistryUtil.NAMESPACE_GITHUB, DatasetRegistryGitHubArchive.class);
     }});
 
     private final InputStreamFactory inputStreamFactory;

diff --git a/src/main/java/org/globalbioticinteractions/elton/util/DatasetRegistryUtil.java b/src/main/java/org/globalbioticinteractions/elton/util/DatasetRegistryUtil.java
@@ -15,6 +15,8 @@
 public class DatasetRegistryUtil {
 
     public static final String NAMESPACE_LOCAL = "local";
+    public static final String NAMESPACE_ZENODO = "zenodo";
+    public static final String NAMESPACE_GITHUB = "github";
 
     public static DatasetRegistry forLocalDir(final URI localArchiveDir,
                                               final String cacheDir,

diff --git a/src/test/java/org/globalbioticinteractions/elton/cmd/CmdLogTest.java b/src/test/java/org/globalbioticinteractions/elton/cmd/CmdLogTest.java
@@ -3,6 +3,7 @@
 import bio.guoda.preston.HashType;
 import bio.guoda.preston.RefNodeConstants;
 import org.apache.commons.lang3.StringUtils;
+import org.globalbioticinteractions.elton.util.DatasetRegistryUtil;
 import org.junit.Rule;
 import org.junit.Test;
 import org.junit.rules.TemporaryFolder;
@@ -50,12 +51,12 @@ public void logRSSTemplate() throws URISyntaxException, IOException {
         CmdUpdate pull = new CmdUpdate();
         pull.setCacheDir(cacheDir);
         pull.setWorkDir(new File(getClass().getResource("/dataset-rss-cache/globi.json").toURI()).getParent());
-        pull.setRegistryNames(Collections.singletonList("local"));
+        pull.setRegistryNames(Collections.singletonList(DatasetRegistryUtil.NAMESPACE_LOCAL));
         pull.run();
 
         CmdLog cmd = new CmdLog();
         cmd.setCacheDir(cacheDir);
-        cmd.setNamespaces(Collections.singletonList("local"));
+        cmd.setNamespaces(Collections.singletonList(DatasetRegistryUtil.NAMESPACE_LOCAL));
         ByteArrayOutputStream out1 = new ByteArrayOutputStream();
         PrintStream out = new PrintStream(out1);
         cmd.setStdout(out);

diff --git a/src/test/java/org/globalbioticinteractions/elton/cmd/CmdUpdateTest.java b/src/test/java/org/globalbioticinteractions/elton/cmd/CmdUpdateTest.java
@@ -2,6 +2,7 @@
 
 import org.apache.commons.io.IOUtils;
 import org.apache.commons.lang3.StringUtils;
+import org.globalbioticinteractions.elton.util.DatasetRegistryUtil;
 import org.junit.Rule;
 import org.junit.Test;
 import org.junit.rules.TemporaryFolder;
@@ -36,12 +37,12 @@ public void updateLocalDataset() throws IOException, URISyntaxException {
         File file = tmpDir.newFolder();
         cmd.setCacheDir(file.getAbsolutePath());
         cmd.setWorkDir(localWorkDir.getAbsolutePath());
-        cmd.setRegistryNames(Arrays.asList("local"));
-        cmd.setNamespaces(Collections.singletonList("local"));
+        cmd.setRegistryNames(Arrays.asList(DatasetRegistryUtil.NAMESPACE_LOCAL));
+        cmd.setNamespaces(Collections.singletonList(DatasetRegistryUtil.NAMESPACE_LOCAL));
 
         cmd.run();
 
-        File local = new File(file, "local");
+        File local = new File(file, DatasetRegistryUtil.NAMESPACE_LOCAL);
         assertThat(local.exists(), is(true));
 
         File provenanceLog = new File(local, "access.tsv");