Add multi-extractor combination tests with format validation

haniyakonain · haniyakonain · commit b6ce78ee94ad · 2025-10-30T02:13:49.000+05:30
- Added test-extraction-combinations.sh script to test different extractor combinations
- Tests mappings-only, all-enabled, and PageId+Label combinations
- Validates both N-Triples and Turtle output formats
- Integrated into GitHub Actions workflow
- TriX format warnings already implemented in Extraction.scala

How to run manually:
  cd scripts/src/main/bash
  chmod +x test-extraction-combinations.sh
  ./test-extraction-combinations.sh

Note: Server must be running on localhost:9999 before running tests
diff --git a/.github/workflows/server-web-api-test.yml b/.github/workflows/server-web-api-test.yml
@@ -51,6 +51,10 @@ jobs:
       run: |
         cd scripts/src/main/bash
         ./coords-integration-test.sh
+    - name: Run Multi-Extractor Combination Tests
+      run: |
+          cd scripts/src/main/bash
+          ./test-extraction-combinations.sh
     - name: Stop DBpedia Server
       run: |
         if [ -f server.pid ]; then
diff --git a/scripts/src/main/bash/test-extraction-combinations.sh b/scripts/src/main/bash/test-extraction-combinations.sh
@@ -0,0 +1,21 @@
+#!/bin/bash
+# Test Multi-Extractor API (PR #794)
+
+API="http://localhost:9999/extraction/en"
+TITLE="Albert_Einstein"
+
+echo "Testing Multi-Extractor API"
+
+# Test 1: Mappings-only + N-Triples
+curl -s "$API/extract?title=$TITLE&format=n-triples&extractors=mappings" | grep -q "^<" && echo "✓ Mappings N-Triples" || echo "✗ FAIL"
+
+# Test 2: All-enabled + Turtle
+curl -s "$API/extract?title=$TITLE&format=turtle-triples&extractors=custom" | grep -qE "(@prefix|^<)" && echo "✓ All-enabled Turtle" || echo "✗ FAIL"
+
+# Test 3: PageId + Label + N-Triples
+curl -s "$API/extract?title=$TITLE&format=n-triples&extractors=PageIdExtractor&extractors=LabelExtractor" | grep -q "^<" && echo "✓ PageId+Label N-Triples" || echo "✗ FAIL"
+
+# Test 4: PageId + Label + Turtle
+curl -s "$API/extract?title=$TITLE&format=turtle-triples&extractors=PageIdExtractor&extractors=LabelExtractor" | grep -qE "(@prefix|^<)" && echo "✓ PageId+Label Turtle" || echo "✗ FAIL"
+
+echo "Done"
diff --git a/server/src/main/scala/org/dbpedia/extraction/server/resources/Extraction.scala b/server/src/main/scala/org/dbpedia/extraction/server/resources/Extraction.scala
@@ -302,14 +302,22 @@ name match {
         val writer = new StringWriter
         val formatter = createFormatter(finalFormat, writer)
 
-        if (errors.nonEmpty && (finalFormat.contains("turtle") || finalFormat.contains("n-triples") || finalFormat.contains("n-quads"))) {
-          writer.write("# Warning: Partial results - some extractors failed:\n")
-          errors.foreach(err => writer.write(s"# - $err\n"))
-          writer.write("\n")
-        }
-
         val finalDestination = new DeduplicatingDestination(new WriterDestination(() => writer, formatter))
         finalDestination.open()
+        
+if (errors.nonEmpty) {
+  if (finalFormat.contains("turtle") || finalFormat.contains("n-triples") || finalFormat.contains("n-quads")) {          writer.write("# Warning: Partial results - some extractors failed:\n")
+          errors.foreach(err => writer.write(s"# - $err\n"))
+          writer.write("\n")  // Show extractor errors as comments (RDF-safe, won't break parsing)
+        } else if (finalFormat == "trix") {
+    // XML comments for TriX format
+    writer.write("<!-- WARNING: Partial extraction results\n")
+    writer.write("     Some extractors failed:\n")
+    errors.foreach(err => writer.write(s"     - $err\n"))
+    writer.write("-->\n")
+  }
+}
+
         finalDestination.write(collector.quads)
         finalDestination.close()
 
