(nutch) branch master updated: [NUTCH-3160] Remove System.exit(..) from reusable code (#903)

[lewismc]

This is an automated email from the ASF dual-hosted git repository.

lewismc pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/nutch.git


The following commit(s) were added to refs/heads/master by this push:
     new da5570826 [NUTCH-3160] Remove System.exit(..) from reusable code (#903)
da5570826 is described below

commit da557082662227cde52959652f406446496044da
Author: Luca <15426+lfoppi...@users.noreply.github.com>
AuthorDate: Fri Feb 27 00:25:16 2026 +0000

    [NUTCH-3160] Remove System.exit(..) from reusable code (#903)
---
 src/java/org/apache/nutch/metrics/NutchMetrics.java |  3 +++
 .../apache/nutch/tools/CommonCrawlDataDumper.java   |  8 +++++++-
 .../nutch/tools/TestCommonCrawlDataDumper.java      | 21 +++++++++++++++++++++
 3 files changed, 31 insertions(+), 1 deletion(-)

diff --git a/src/java/org/apache/nutch/metrics/NutchMetrics.java 
b/src/java/org/apache/nutch/metrics/NutchMetrics.java
index c65a4f0ce..ef4fe79e5 100644
--- a/src/java/org/apache/nutch/metrics/NutchMetrics.java
+++ b/src/java/org/apache/nutch/metrics/NutchMetrics.java
@@ -82,6 +82,9 @@ public final class NutchMetrics {
   /** Counter group for WARC export operations. */
   public static final String GROUP_WARC_EXPORTER = "nutch_warc_exporter";
 
+  /** Counter group for Common Crawl data dumper tool. */
+  public static final String GROUP_COMMONCRAWL_DUMPER = 
"nutch_commoncrawl_dumper";
+
   /** Counter group for domain statistics operations. */
   public static final String GROUP_DOMAIN_STATS = "nutch_domain_stats";
 
diff --git a/src/java/org/apache/nutch/tools/CommonCrawlDataDumper.java 
b/src/java/org/apache/nutch/tools/CommonCrawlDataDumper.java
index d5d5035e8..8e37c21fc 100644
--- a/src/java/org/apache/nutch/tools/CommonCrawlDataDumper.java
+++ b/src/java/org/apache/nutch/tools/CommonCrawlDataDumper.java
@@ -67,6 +67,8 @@ import org.apache.nutch.crawl.Inlinks;
 import org.apache.nutch.crawl.LinkDbReader;
 import org.apache.nutch.metadata.Metadata;
 import org.apache.nutch.metadata.Nutch;
+import org.apache.nutch.metrics.ErrorTracker;
+import org.apache.nutch.metrics.NutchMetrics;
 import org.apache.nutch.protocol.Content;
 import org.apache.nutch.util.DumpFileUtil;
 import org.apache.nutch.util.NutchConfiguration;
@@ -188,6 +190,7 @@ public class CommonCrawlDataDumper extends NutchTool 
implements Tool {
   private GzipCompressorOutputStream gzipOutput = null;
   private TarArchiveOutputStream tarOutput = null;
   private ArrayList<String> fileList = null;
+  private ErrorTracker errorTracker;
 
   /**
    * Main method for invoking this tool
@@ -210,6 +213,7 @@ public class CommonCrawlDataDumper extends NutchTool 
implements Tool {
    * @param config A populated {@link CommonCrawlConfig}
    */
   public CommonCrawlDataDumper(CommonCrawlConfig config) {
+    this();
     this.config = config;
   }
 
@@ -217,6 +221,7 @@ public class CommonCrawlDataDumper extends NutchTool 
implements Tool {
    * Constructor
    */
   public CommonCrawlDataDumper() {
+    this.errorTracker = new 
ErrorTracker(NutchMetrics.GROUP_COMMONCRAWL_DUMPER);
   }
 
   /**
@@ -274,7 +279,8 @@ public class CommonCrawlDataDumper extends NutchTool 
implements Tool {
     if (parts == null || parts.size() == 0) {
       LOG.error( "No segment directories found in {} ",
           segmentRootDir.getAbsolutePath());
-      System.exit(1);
+      this.errorTracker.recordError(ErrorTracker.ErrorType.OTHER);
+      return;
     }
     LOG.info("Found {} segment parts", parts.size());
     if (gzip && !warc) {
diff --git a/src/test/org/apache/nutch/tools/TestCommonCrawlDataDumper.java 
b/src/test/org/apache/nutch/tools/TestCommonCrawlDataDumper.java
index fee72b65a..8124fe20b 100644
--- a/src/test/org/apache/nutch/tools/TestCommonCrawlDataDumper.java
+++ b/src/test/org/apache/nutch/tools/TestCommonCrawlDataDumper.java
@@ -17,14 +17,19 @@
 package org.apache.nutch.tools;
 
 import java.io.File;
+import java.lang.reflect.Field;
 import java.nio.file.Files;
 import java.util.Collection;
 
 import org.apache.commons.io.FileUtils;
 import org.apache.commons.io.filefilter.FileFilterUtils;
+import org.apache.nutch.metrics.ErrorTracker;
 import org.junit.jupiter.api.Test;
 
 import static org.junit.jupiter.api.Assertions.assertTrue;
+import static org.mockito.Mockito.mock;
+import static org.mockito.Mockito.times;
+import static org.mockito.Mockito.verify;
 
 /**
  * 
@@ -106,6 +111,22 @@ public class TestCommonCrawlDataDumper {
 
   }
 
+  @Test
+  public void testDumpWithNoSegmentDirectoriesRecordsOtherError() throws 
Exception {
+    File emptySegmentDir = 
Files.createTempDirectory("empty-segments").toFile();
+    File outputDir = Files.createTempDirectory("dump-output").toFile();
+
+    ErrorTracker mockErrorTracker = mock(ErrorTracker.class);
+    CommonCrawlDataDumper dumper = new CommonCrawlDataDumper();
+    Field errorTrackerField = 
CommonCrawlDataDumper.class.getDeclaredField("errorTracker");
+    errorTrackerField.setAccessible(true);
+    errorTrackerField.set(dumper, mockErrorTracker);
+
+    dumper.dump(outputDir, emptySegmentDir, null, false, null, false, "", 
false);
+
+    verify(mockErrorTracker, 
times(1)).recordError(ErrorTracker.ErrorType.OTHER);
+  }
+
   private boolean hasFile(String fileName, Collection<File> files) {
     for (File f : files) {
       if (f.getName().equals(fileName)) {