remove size from default analysisTypes list for segmentMetadata query (#3773)

docs/content/querying/segmentmetadataquery.md

@@ -32,7 +32,7 @@ There are several main parts to a segment metadata query:
 |toInclude|A JSON Object representing what columns should be included in the result. Defaults to "all".|no|
 |merge|Merge all individual segment metadata results into a single result|no|
 |context|See [Context](../querying/query-context.html)|no|
-|analysisTypes|A list of Strings specifying what column properties (e.g. cardinality, size) should be calculated and returned in the result. Defaults to ["cardinality", "size", "interval", "minmax"]. See section [analysisTypes](#analysistypes) for more details.|no|
+|analysisTypes|A list of Strings specifying what column properties (e.g. cardinality, size) should be calculated and returned in the result. Defaults to ["cardinality", "interval", "minmax"]. See section [analysisTypes](#analysistypes) for more details.|no|
 |lenientAggregatorMerge|If true, and if the "aggregators" analysisType is enabled, aggregators will be merged leniently. See below for details.|no|
 
 The format of the result is:

integration-tests/src/test/resources/hadoop/batch_hadoop_queries.json

@@ -16,7 +16,7 @@
         "columns": {
           "location": {
             "type": "STRING",
-            "size": 10140,
+            "size": 0,
             "hasMultipleValues": false,
             "minValue": "location_1",
             "maxValue": "location_5",
@@ -43,7 +43,7 @@
           },
           "__time": {
             "type": "LONG",
-            "size": 10140,
+            "size": 0,
             "hasMultipleValues": false,
             "minValue": null,
             "maxValue": null,
@@ -52,7 +52,7 @@
           },
           "product": {
             "type": "STRING",
-            "size": 9531,
+            "size": 0,
             "hasMultipleValues": false,
             "minValue": "product_1",
             "maxValue": "product_9",
@@ -60,7 +60,7 @@
             "errorMessage": null
           }
         },
-        "size": 34881,
+        "size": 0,
         "numRows": 1014,
         "aggregators": null,
         "timestampSpec": null,

integration-tests/src/test/resources/queries/twitterstream_queries.json

@@ -598,14 +598,14 @@
                     "has_links": {
                         "type": "STRING",
                         "hasMultipleValues": false,
-                        "size": 7773438,
+                        "size": 0,
                         "cardinality": 2,
                         "minValue":"No",
                         "maxValue":"Yes",
                         "errorMessage": null
                     }
                 },
-                "size": 747056474,
+                "size": 0,
                 "numRows": 3702583,
                 "aggregators": null,
                 "timestampSpec": null,
@@ -619,14 +619,14 @@
                     "has_links": {
                         "type": "STRING",
                         "hasMultipleValues": false,
-                        "size": 7901000,
+                        "size": 0,
                         "cardinality": 2,
                         "minValue":"No",
                         "maxValue":"Yes",
                         "errorMessage": null
                     }
                 },
-                "size": 755796690,
+                "size": 0,
                 "numRows": 3743002,
                 "aggregators": null,
                 "timestampSpec": null,
@@ -640,14 +640,14 @@
                     "has_links": {
                         "type": "STRING",
                         "hasMultipleValues": false,
-                        "size": 7405654,
+                        "size": 0,
                         "cardinality": 2,
                         "minValue":"No",
                         "maxValue":"Yes",
                         "errorMessage": null
                     }
                 },
-                "size": 706893542,
+                "size": 0,
                 "numRows":3502959,
                 "aggregators": null,
                 "timestampSpec": null,

integration-tests/src/test/resources/queries/wikipedia_editstream_queries.json

@@ -1048,7 +1048,7 @@
                     "country_name": {
                         "type": "STRING",
                         "hasMultipleValues": false,
-                        "size": 41922148,
+                        "size": 0,
                         "cardinality": 208,
                         "minValue":"",
                         "maxValue":"mmx._unknown",
@@ -1057,14 +1057,14 @@
                     "language": {
                         "type": "STRING",
                         "hasMultipleValues": false,
-                        "size": 8924222,
+                        "size": 0,
                         "cardinality": 36,
                         "minValue":"ar",
                         "maxValue":"zh",
                         "errorMessage": null
                     }
                 },
-                "size": 902457341,
+                "size": 0,
                 "numRows": 4462111,
                 "aggregators": null,
                 "timestampSpec": null,

processing/src/main/java/io/druid/query/metadata/metadata/SegmentMetadataQuery.java

@@ -86,7 +86,6 @@ public class SegmentMetadataQuery extends BaseQuery<SegmentAnalysis>
 
   public static final EnumSet<AnalysisType> DEFAULT_ANALYSIS_TYPES = EnumSet.of(
       AnalysisType.CARDINALITY,
-      AnalysisType.SIZE,
       AnalysisType.INTERVAL,
       AnalysisType.MINMAX
   );

processing/src/test/java/io/druid/query/metadata/SegmentAnalyzerTest.java

@@ -91,11 +91,7 @@ public class SegmentAnalyzerTest
       final ColumnAnalysis columnAnalysis = columns.get(metric);
 
       Assert.assertEquals(metric, ValueType.FLOAT.name(), columnAnalysis.getType());
-      if (analyses == null) {
-        Assert.assertTrue(metric, columnAnalysis.getSize() > 0);
-      } else {
-        Assert.assertEquals(metric, 0, columnAnalysis.getSize());
-      }
+      Assert.assertEquals(metric, 0, columnAnalysis.getSize());
       Assert.assertNull(metric, columnAnalysis.getCardinality());
     }
   }
@@ -131,12 +127,11 @@ public class SegmentAnalyzerTest
         Assert.assertNull(columnAnalysis);
       } else {
         Assert.assertEquals(dimension, ValueType.STRING.name(), columnAnalysis.getType());
+        Assert.assertEquals(dimension, 0, columnAnalysis.getSize());
         if (analyses == null) {
-          Assert.assertTrue(dimension, columnAnalysis.getSize() > 0);
           Assert.assertTrue(dimension, columnAnalysis.getCardinality() > 0);
         } else {
           Assert.assertEquals(dimension, 0, columnAnalysis.getCardinality().longValue());
-          Assert.assertEquals(dimension, 0, columnAnalysis.getSize());
         }
       }
     }
@@ -145,11 +140,7 @@ public class SegmentAnalyzerTest
       final ColumnAnalysis columnAnalysis = columns.get(metric);
 
       Assert.assertEquals(metric, ValueType.FLOAT.name(), columnAnalysis.getType());
-      if (analyses == null) {
-        Assert.assertTrue(metric, columnAnalysis.getSize() > 0);
-      } else {
-        Assert.assertEquals(metric, 0, columnAnalysis.getSize());
-      }
+      Assert.assertEquals(metric, 0, columnAnalysis.getSize());
       Assert.assertNull(metric, columnAnalysis.getCardinality());
     }
   }

processing/src/test/java/io/druid/query/metadata/SegmentMetadataQueryQueryToolChestTest.java

@@ -63,7 +63,7 @@ public class SegmentMetadataQueryQueryToolChestTest
         new SegmentMetadataQueryQueryToolChest(null).getCacheStrategy(query);
 
     // Test cache key generation
-    byte[] expectedKey = {0x04, 0x01, (byte) 0xFF, 0x00, 0x01, 0x02, 0x04};
+    byte[] expectedKey = {0x04, 0x01, (byte) 0xFF, 0x00, 0x02, 0x04};
     byte[] actualKey = strategy.computeCacheKey(query);
     Assert.assertArrayEquals(expectedKey, actualKey);
 

processing/src/test/java/io/druid/query/metadata/SegmentMetadataQueryTest.java

@@ -154,7 +154,12 @@ public class SegmentMetadataQueryTest
                       .dataSource("testing")
                       .intervals("2013/2014")
                       .toInclude(new ListColumnIncluderator(Arrays.asList("__time", "index", "placement")))
-                      .analysisTypes(null)
+                      .analysisTypes(
+                          SegmentMetadataQuery.AnalysisType.CARDINALITY,
+                          SegmentMetadataQuery.AnalysisType.SIZE,
+                          SegmentMetadataQuery.AnalysisType.INTERVAL,
+                          SegmentMetadataQuery.AnalysisType.MINMAX
+                      )
                       .merge(true)
                       .build();
 

processing/src/test/java/io/druid/query/metadata/SegmentMetadataUnionQueryTest.java

@@ -121,6 +121,12 @@ public class SegmentMetadataUnionQueryTest
         .dataSource(QueryRunnerTestHelper.unionDataSource)
         .intervals(QueryRunnerTestHelper.fullOnInterval)
         .toInclude(new ListColumnIncluderator(Lists.newArrayList("placement")))
+        .analysisTypes(
+            SegmentMetadataQuery.AnalysisType.CARDINALITY,
+            SegmentMetadataQuery.AnalysisType.SIZE,
+            SegmentMetadataQuery.AnalysisType.INTERVAL,
+            SegmentMetadataQuery.AnalysisType.MINMAX
+        )
         .build();
     List result = Sequences.toList(runner.run(query, Maps.newHashMap()), Lists.<SegmentAnalysis>newArrayList());
     TestHelper.assertExpectedObjects(ImmutableList.of(expected), result, "failed SegmentMetadata union query");