SOLR-6085: Suggester crashes when prefixToken is longer than surface form

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1638711 13f79535-47bb-0310-9956-ffa450edef68

lucene/suggest/src/java/org/apache/lucene/search/suggest/analyzing/AnalyzingInfixSuggester.java

@@ -695,13 +695,15 @@ public class AnalyzingInfixSuggester extends Lookup implements Closeable {
   protected void addPrefixMatch(StringBuilder sb, String surface, String analyzed, String prefixToken) {
     // TODO: apps can try to invert their analysis logic
     // here, e.g. downcase the two before checking prefix:
+    if (prefixToken.length() >= surface.length()) {
+      addWholeMatch(sb, surface, analyzed);
+      return;
+    }
     sb.append("<b>");
     sb.append(surface.substring(0, prefixToken.length()));
     sb.append("</b>");
-    if (prefixToken.length() < surface.length()) {
     sb.append(surface.substring(prefixToken.length()));
   }
-  }
 
   @Override
   public boolean store(DataOutput in) throws IOException {

lucene/suggest/src/test/org/apache/lucene/search/suggest/analyzing/AnalyzingInfixSuggesterTest.java

@@ -41,9 +41,11 @@ import org.apache.lucene.search.BooleanClause;
 import org.apache.lucene.search.suggest.Input;
 import org.apache.lucene.search.suggest.InputArrayIterator;
 import org.apache.lucene.search.suggest.Lookup.LookupResult;
+import org.apache.lucene.store.Directory;
 import org.apache.lucene.util.BytesRef;
 import org.apache.lucene.util.LuceneTestCase;
 import org.apache.lucene.util.TestUtil;
+import org.junit.Test;
 
 public class AnalyzingInfixSuggesterTest extends LuceneTestCase {
 
@@ -1076,4 +1078,28 @@ public class AnalyzingInfixSuggesterTest extends LuceneTestCase {
       suggester.close();
     }
   }
+
+  @Test
+  public void testAddPrefixMatch() throws IOException {
+    Analyzer a = new MockAnalyzer(random(), MockTokenizer.WHITESPACE, false);
+    Directory dir = newDirectory();
+    AnalyzingInfixSuggester suggester = new AnalyzingInfixSuggester(dir, a);
+
+    assertEquals("<b>Sol</b>r", pfmToString(suggester, "Solr", "Sol"));
+    assertEquals("<b>Solr</b>", pfmToString(suggester, "Solr", "Solr"));
+
+    // Test SOLR-6085 - the analyzed tokens match due to ss->ß normalization
+    assertEquals("<b>daß</b>", pfmToString(suggester, "daß", "dass"));
+
+    dir.close();
+    suggester.close();
+  }
+
+  private String pfmToString(AnalyzingInfixSuggester suggester, String surface, String prefix) throws IOException {
+    StringBuilder sb = new StringBuilder();
+    suggester.addPrefixMatch(sb, surface, "", prefix);
+    return sb.toString();
+  }
+
+
 }

solr/CHANGES.txt

@@ -400,6 +400,8 @@ Bug Fixes
 
 * SOLR-6704: TrieDateField type drops schema properties in branch 4.10 (Tomás Fernández Löbbe)
 
+* SOLR-6085: Suggester crashes when prefixToken is longer than surface form (janhoy)
+
 ==================  4.10.2 ==================
 
 Bug Fixes