commit af7702128621b57e91bffc96bca339960ca2193a
author Michael Blow <mblow@apache.org> Fri Dec 16 12:17:09 2016 -0800
committer Till Westmann <tillw@apache.org> Fri Dec 16 14:59:27 2016 -0800
tree 7bd49fde984c69c2b389f7a85b38436b2f8d8ac8
parent 2f9d4c3ab4d55598fe9a14fbf28faef12bed208b

Fix ASTERIXDB-1255

This changes enable dynamic hashtable size in SimilarityJaccardEvaluator

Change-Id: I08c550d35d864df08792369d38ff81012e3976af
Reviewed-on: https://asterix-gerrit.ics.uci.edu/586
Sonar-Qube: Jenkins <jenkins@fulliautomatix.ics.uci.edu>
Tested-by: Jenkins <jenkins@fulliautomatix.ics.uci.edu>
Integration-Tests: Jenkins <jenkins@fulliautomatix.ics.uci.edu>
Reviewed-by: Till Westmann <tillw@apache.org>

asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/common/SimilarityJaccardEvaluator.java

diff --git a/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/common/SimilarityJaccardEvaluator.java b/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/common/SimilarityJaccardEvaluator.java
index e9928a1..ab817bd 100644
--- a/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/common/SimilarityJaccardEvaluator.java
+++ b/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/common/SimilarityJaccardEvaluator.java
@@ -50,11 +50,11 @@
 public class SimilarityJaccardEvaluator implements IScalarEvaluator {
 
     // Parameters for hash table.
-    protected final int TABLE_SIZE = 100;
-    protected final int TABLE_FRAME_SIZE = 32768;
+    protected static final int MIN_TABLE_SIZE = 100;
+    protected static final int TABLE_FRAME_SIZE = 32768;
 
     // Assuming type indicator in serde format.
-    protected final int TYPE_INDICATOR_SIZE = 1;
+    protected static final int TYPE_INDICATOR_SIZE = 1;
 
     protected final ArrayBackedValueStorage resultStorage = new ArrayBackedValueStorage();
     protected final DataOutput out = resultStorage.getDataOutput();
@@ -89,6 +89,8 @@
     // Ignore case for strings. Defaults to true.
     protected final boolean ignoreCase = true;
 
+    protected int hashTableSize = MIN_TABLE_SIZE;
+
     public SimilarityJaccardEvaluator(IScalarEvaluatorFactory[] args, IHyracksTaskContext context)
             throws HyracksDataException {
         firstOrdListEval = args[0].createScalarEvaluator(context);
@@ -140,6 +142,10 @@
         if (firstListIter.size() == 0 || secondListIter.size() == 0) {
             return false;
         }
+
+        // Set the size of the table dynamically
+        hashTableSize = Math.max(Math.max(firstListIter.size(), secondListIter.size()), MIN_TABLE_SIZE);
+
         // TODO: Check item types are compatible.
         return true;
     }
@@ -229,7 +235,7 @@
                 .createBinaryHashFunction(probeItemTypeTag, ignoreCase);
         IBinaryComparator cmp = ListItemBinaryComparatorFactory.INSTANCE.createBinaryComparator(buildItemTypeTag,
                 probeItemTypeTag, ignoreCase);
-        hashMap = new BinaryHashMap(TABLE_SIZE, TABLE_FRAME_SIZE, putHashFunc, getHashFunc, cmp);
+        hashMap = new BinaryHashMap(hashTableSize, TABLE_FRAME_SIZE, putHashFunc, getHashFunc, cmp);
     }
 
     protected boolean checkArgTypes(ATypeTag typeTag1, ATypeTag typeTag2) throws HyracksDataException {