LUCENE-9680 - Re-add IndexWriter::getFieldNames

Author: orenovadia

lucene/CHANGES.txt

@@ -7,6 +7,9 @@ http://s.apache.org/luceneversions
 
 New Features
 
+* LUCENE-9680: IndexWriter#getFieldNames() method added to get fields present in index.
+  This method was removed in LUCENE-8909. (Oren Ovadia)
+
 * LUCENE-9322: Vector-valued fields, Lucene90 Codec (Mike Sokolov, Julie Tibshirani, Tomoko Uchida)
 
 * LUCENE-9004: Approximate nearest vector search via NSW graphs

lucene/core/src/java/org/apache/lucene/index/FieldInfos.java

@@ -25,6 +25,7 @@
 import java.util.List;
 import java.util.Map;
 import java.util.Objects;
+import java.util.Set;
 import java.util.stream.Collectors;
 import java.util.stream.StreamSupport;
 import org.apache.lucene.util.ArrayUtil;
@@ -681,6 +682,10 @@ synchronized boolean contains(String fieldName, DocValuesType dvType) {
       }
     }
 
+    synchronized Set<String> getFieldNames() {
+      return Set.copyOf(nameToNumber.keySet());
+    }
+
     synchronized void clear() {
       numberToName.clear();
       nameToNumber.clear();

lucene/core/src/java/org/apache/lucene/index/IndexWriter.java

@@ -1987,6 +1987,17 @@ private DocValuesUpdate[] buildDocValuesUpdate(Term term, Field[] updates) {
     return dvUpdates;
   }
 
+  /**
+   * Return an unmodifiable set of all field names as visible from this IndexWriter, across all
+   * segments of the index.
+   *
+   * @lucene.experimental
+   */
+  public Set<String> getFieldNames() {
+    // FieldNumbers#getFieldNames() returns an unmodifiableSet
+    return globalFieldNumberMap.getFieldNames();
+  }
+
   // for test purpose
   final synchronized int getSegmentCount() {
     return segmentInfos.size();

lucene/core/src/test/org/apache/lucene/index/TestIndexWriter.java

@@ -4635,4 +4635,52 @@ public void testIndexWriterBlocksOnStall() throws IOException, InterruptedExcept
       }
     }
   }
+
+  public void testGetFieldNames() throws IOException {
+    Directory dir = newDirectory();
+
+    IndexWriter writer = new IndexWriter(dir, newIndexWriterConfig(new MockAnalyzer(random())));
+
+    assertEquals(Set.of(), writer.getFieldNames());
+
+    addDocWithField(writer, "f1");
+    assertEquals(Set.of("f1"), writer.getFieldNames());
+
+    // should be unmodifiable:
+    final Set<String> fieldSet = writer.getFieldNames();
+    assertThrows(UnsupportedOperationException.class, () -> fieldSet.add("cannot modify"));
+    assertThrows(UnsupportedOperationException.class, () -> fieldSet.remove("f1"));
+
+    addDocWithField(writer, "f2");
+    assertEquals(Set.of("f1", "f2"), writer.getFieldNames());
+
+    // set from a previous call is an independent immutable copy, cannot be modified.
+    assertEquals(Set.of("f1"), fieldSet);
+
+    // flush should not have an effect on field names
+    writer.flush();
+    assertEquals(Set.of("f1", "f2"), writer.getFieldNames());
+
+    // commit should not have an effect on field names
+    writer.commit();
+    assertEquals(Set.of("f1", "f2"), writer.getFieldNames());
+
+    writer.close();
+
+    // new writer should identify committed fields
+    writer = new IndexWriter(dir, newIndexWriterConfig(new MockAnalyzer(random())));
+    assertEquals(Set.of("f1", "f2"), writer.getFieldNames());
+
+    writer.deleteAll();
+    assertEquals(Set.of(), writer.getFieldNames());
+
+    writer.close();
+    dir.close();
+  }
+
+  private static void addDocWithField(IndexWriter writer, String field) throws IOException {
+    Document doc = new Document();
+    doc.add(newField(field, "value", storedTextType));
+    writer.addDocument(doc);
+  }
 }