LUCENE-3345: add another test

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1151346 13f79535-47bb-0310-9956-ffa450edef68

lucene/src/java/org/apache/lucene/index/codecs/pulsing/PulsingCodec.java

@@ -66,6 +66,11 @@ public class PulsingCodec extends Codec {
     this(1);
   }
   
+  /** @lucene.internal */
+  public int getFreqCutoff() {
+    return freqCutoff;
+  }
+
   /** Terms with freq <= freqCutoff are inlined into terms
    *  dict. */
   public PulsingCodec(int freqCutoff) {

lucene/src/test/org/apache/lucene/index/codecs/pulsing/Test10KPulsings.java

@@ -32,6 +32,7 @@ import org.apache.lucene.index.IndexReader;
 import org.apache.lucene.index.MultiFields;
 import org.apache.lucene.index.RandomIndexWriter;
 import org.apache.lucene.index.TermsEnum;
+import org.apache.lucene.index.codecs.Codec;
 import org.apache.lucene.index.codecs.CodecProvider;
 import org.apache.lucene.search.DocIdSetIterator;
 import org.apache.lucene.store.MockDirectoryWrapper;
@@ -92,4 +93,65 @@ public class Test10KPulsings extends LuceneTestCase {
     _TestUtil.checkIndex(dir);
     dir.close();
   }
+  
+  /** a variant, that uses pulsing, but uses a high TF to force pass thru to the underlying codec
+   * creates a broken index (triggers a different assert) than test10kPulsed, with this:
+   * ant test -Dtestcase=Test10KPulsings -Dtestmethod=test10kNotPulsed -Dtests.seed=7065174228571869719:2545882165086224608!!!!
+   */
+  public void test10kNotPulsed() throws Exception {
+    // we always run this test with pulsing codec.
+    CodecProvider cp = _TestUtil.alwaysCodec(new PulsingCodec(1));
+    
+    File f = _TestUtil.getTempDir("10kpulsings");
+    MockDirectoryWrapper dir = newFSDirectory(f);
+    dir.setCheckIndexOnClose(false); // we do this ourselves explicitly
+    RandomIndexWriter iw = new RandomIndexWriter(random, dir, 
+        newIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(random)).setCodecProvider(cp));
+    
+    Document document = new Document();
+    Field field = newField("field", "", Field.Store.YES, Field.Index.ANALYZED);
+    
+    switch(_TestUtil.nextInt(random, 0, 2)) {
+      case 0: field.setIndexOptions(IndexOptions.DOCS_ONLY); break;
+      case 1: field.setIndexOptions(IndexOptions.DOCS_AND_FREQS); break;
+      default: field.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS); break; 
+    }
+
+    document.add(field);
+    
+    NumberFormat df = new DecimalFormat("00000", new DecimalFormatSymbols(Locale.ENGLISH));
+
+    Codec codec = cp.lookup(cp.getFieldCodec("field"));
+    assertTrue(codec instanceof PulsingCodec);
+    PulsingCodec pulsing = (PulsingCodec) codec;
+    final int freq = pulsing.getFreqCutoff() + 1;
+    
+    for (int i = 0; i < 10050; i++) {
+      StringBuilder sb = new StringBuilder();
+      for (int j = 0; j < freq; j++) {
+        sb.append(df.format(i));
+        sb.append(' '); // whitespace
+      }
+      field.setValue(sb.toString());
+      iw.addDocument(document);
+    }
+    
+    IndexReader ir = iw.getReader();
+    iw.close();
+
+    TermsEnum te = MultiFields.getTerms(ir, "field").iterator();
+    DocsEnum de = null;
+    
+    for (int i = 0; i < 10050; i++) {
+      String expected = df.format(i);
+      assertEquals(expected, te.next().utf8ToString());
+      de = te.docs(null, de);
+      assertTrue(de.nextDoc() != DocIdSetIterator.NO_MORE_DOCS);
+      assertEquals(DocIdSetIterator.NO_MORE_DOCS, de.nextDoc());
+    }
+    ir.close();
+
+    _TestUtil.checkIndex(dir);
+    dir.close();
+  }
 }