[GitHub] [lucene] nik9000 commented on a change in pull request #415: LUCENE-10206 Implement O(1) count on query cache

[GitBox <gi...@apache.org>]

nik9000 commented on a change in pull request #415:
URL: https://github.com/apache/lucene/pull/415#discussion_r736693202



##########
File path: lucene/core/src/test/org/apache/lucene/search/TestLRUQueryCache.java
##########
@@ -1976,4 +1989,60 @@ public void testSkipCachingForTermQuery() throws IOException {
     reader.close();
     dir.close();
   }
+
+  public void testCacheHasFastCount() throws IOException {
+    Query query = new PhraseQuery("words", new BytesRef("alice"), new BytesRef("ran"));
+
+    Directory dir = newDirectory();
+    RandomIndexWriter w =
+        new RandomIndexWriter(
+            random(), dir, newIndexWriterConfig().setMergePolicy(NoMergePolicy.INSTANCE));
+    Document doc1 = new Document();
+    doc1.add(new TextField("words", "tom ran", Store.NO));
+    Document doc2 = new Document();
+    doc2.add(new TextField("words", "alice ran", Store.NO));
+    doc2.add(new StringField("f", "a", Store.NO));
+    Document doc3 = new Document();
+    doc3.add(new TextField("words", "alice ran", Store.NO));
+    doc3.add(new StringField("f", "b", Store.NO));
+    w.addDocuments(Arrays.asList(doc1, doc2, doc3));
+
+    try (IndexReader reader = w.getReader()) {
+      IndexSearcher searcher = newSearcher(reader);
+      searcher.setQueryCachingPolicy(ALWAYS_CACHE);
+      LRUQueryCache allCache =
+          new LRUQueryCache(1000000, 10000000, context -> true, Float.POSITIVE_INFINITY);
+      searcher.setQueryCache(allCache);
+      Weight weight = searcher.createWeight(query, ScoreMode.COMPLETE_NO_SCORES, 1);
+      LeafReaderContext context = getOnlyLeafReader(reader).getContext();
+      // We don't have a fast count before the cache is filled
+      assertEquals(weight.count(context), -1);
+      // Fetch the scorer to populate the cache
+      weight.scorer(context);
+      assertEquals(List.of(query), allCache.cachedQueries());
+      // Now we *do* have a fast count
+      assertEquals(weight.count(context), 2);
+    }
+
+    w.deleteDocuments(new TermQuery(new Term("f", new BytesRef("b"))));
+    try (IndexReader reader = w.getReader()) {
+      IndexSearcher searcher = newSearcher(reader);
+      searcher.setQueryCachingPolicy(ALWAYS_CACHE);
+      LRUQueryCache allCache =
+          new LRUQueryCache(1000000, 10000000, context -> true, Float.POSITIVE_INFINITY);
+      searcher.setQueryCache(allCache);
+      Weight weight = searcher.createWeight(query, ScoreMode.COMPLETE_NO_SCORES, 1);
+      LeafReaderContext context = getOnlyLeafReader(reader).getContext();
+      // We don't have a fast count before the cache is filled
+      assertEquals(weight.count(context), -1);
+      // Fetch the scorer to populate the cache
+      weight.scorer(context);
+      assertEquals(List.of(query), allCache.cachedQueries());
+      // We still don't have a fast count because we have deleted documents
+      assertEquals(weight.count(context), -1);

Review comment:
       Bah! I leave this same comment on other people's code. And yet I make the same mistake. Will fix.




-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org