Add sub query explanations in DisjunctionMaxQuery#explain on no-match (…

…#13362)
apache · May 14, 2024 · a6d76fc · a6d76fc
1 parent b1470b9
commit a6d76fc
Show file tree

Hide file tree

Showing 4 changed files with 61 additions and 5 deletions.
diff --git a/lucene/CHANGES.txt b/lucene/CHANGES.txt
@@ -67,6 +67,8 @@ Improvements
 * GITHUB#12966: Move most of the responsibility from TaxonomyFacets implementations to TaxonomyFacets itself.
   This reduces code duplication and enables future development. (Stefan Vodita)
 
+* GITHUB#13362: Add sub query explanations to DisjunctionMaxQuery, if the overall query didn't match. (Tim Grein)
+
 Optimizations
 ---------------------
 

diff --git a/lucene/core/src/java/org/apache/lucene/search/DisjunctionMaxQuery.java b/lucene/core/src/java/org/apache/lucene/search/DisjunctionMaxQuery.java
@@ -213,19 +213,22 @@ public Explanation explain(LeafReaderContext context, int doc) throws IOExceptio
       boolean match = false;
       double max = 0;
       double otherSum = 0;
-      List<Explanation> subs = new ArrayList<>();
+      List<Explanation> subsOnMatch = new ArrayList<>();
+      List<Explanation> subsOnNoMatch = new ArrayList<>();
       for (Weight wt : weights) {
         Explanation e = wt.explain(context, doc);
         if (e.isMatch()) {
           match = true;
-          subs.add(e);
+          subsOnMatch.add(e);
           double score = e.getValue().doubleValue();
           if (score >= max) {
             otherSum += max;
             max = score;
           } else {
             otherSum += score;
           }
+        } else if (match == false) {
+          subsOnNoMatch.add(e);
         }
       }
       if (match) {
@@ -234,9 +237,9 @@ public Explanation explain(LeafReaderContext context, int doc) throws IOExceptio
             tieBreakerMultiplier == 0.0f
                 ? "max of:"
                 : "max plus " + tieBreakerMultiplier + " times others of:";
-        return Explanation.match(score, desc, subs);
+        return Explanation.match(score, desc, subsOnMatch);
       } else {
-        return Explanation.noMatch("No matching clause");
+        return Explanation.noMatch("No matching clause", subsOnNoMatch);
       }
     }
   } // end of DisjunctionMaxWeight inner class

diff --git a/lucene/core/src/test/org/apache/lucene/search/TestDisjunctionMaxQuery.java b/lucene/core/src/test/org/apache/lucene/search/TestDisjunctionMaxQuery.java
@@ -497,6 +497,57 @@ public void testRandomTopDocs() throws Exception {
     doTestRandomTopDocs(4, 1.0f, 0.5f, 0.05f, 0f);
   }
 
+  public void testExplainMatch() throws IOException {
+    // Both match
+    Query sub1 = tq("hed", "elephant");
+    Query sub2 = tq("dek", "elephant");
+
+    final DisjunctionMaxQuery dq = new DisjunctionMaxQuery(Arrays.asList(sub1, sub2), 0.0f);
+
+    final Weight dw = s.createWeight(s.rewrite(dq), ScoreMode.COMPLETE, 1);
+    LeafReaderContext context = (LeafReaderContext) s.getTopReaderContext();
+    Explanation explanation = dw.explain(context, 1);
+
+    assertEquals("max of:", explanation.getDescription());
+    // Two matching sub queries should be included in the explanation details
+    assertEquals(2, explanation.getDetails().length);
+  }
+
+  public void testExplainNoMatch() throws IOException {
+    // No match
+    Query sub1 = tq("abc", "elephant");
+    Query sub2 = tq("def", "elephant");
+
+    final DisjunctionMaxQuery dq = new DisjunctionMaxQuery(Arrays.asList(sub1, sub2), 0.0f);
+
+    final Weight dw = s.createWeight(s.rewrite(dq), ScoreMode.COMPLETE, 1);
+    LeafReaderContext context = (LeafReaderContext) s.getTopReaderContext();
+    Explanation explanation = dw.explain(context, 1);
+
+    assertEquals("No matching clause", explanation.getDescription());
+    // Two non-matching sub queries should be included in the explanation details
+    assertEquals(2, explanation.getDetails().length);
+  }
+
+  public void testExplainMatch_OneNonMatchingSubQuery_NotIncludedInExplanation()
+      throws IOException {
+    // Matches
+    Query sub1 = tq("hed", "elephant");
+
+    // Doesn't match
+    Query sub2 = tq("def", "elephant");
+
+    final DisjunctionMaxQuery dq = new DisjunctionMaxQuery(Arrays.asList(sub1, sub2), 0.0f);
+
+    final Weight dw = s.createWeight(s.rewrite(dq), ScoreMode.COMPLETE, 1);
+    LeafReaderContext context = (LeafReaderContext) s.getTopReaderContext();
+    Explanation explanation = dw.explain(context, 1);
+
+    assertEquals("max of:", explanation.getDescription());
+    // Only the matching sub query (sub1) should be included in the explanation details
+    assertEquals(1, explanation.getDetails().length);
+  }
+
   private void doTestRandomTopDocs(int numFields, double... freqs) throws IOException {
     assert numFields == freqs.length;
     Directory dir = newDirectory();

diff --git a/lucene/test-framework/src/java/org/apache/lucene/tests/search/CheckHits.java b/lucene/test-framework/src/java/org/apache/lucene/tests/search/CheckHits.java
@@ -417,7 +417,7 @@ public static void verifyExplanation(
     if (descr.startsWith("score based on ") && descr.contains("child docs in range")) {
       assertTrue("Child doc explanations are missing", detail.length > 0);
     }
-    if (detail.length > 0) {
+    if (detail.length > 0 && expl.isMatch()) {
       if (detail.length == 1 && COMPUTED_FROM_PATTERN.matcher(descr).matches() == false) {
         // simple containment, unless it's a freq of: (which lets a query explain how the freq is
         // calculated),