commit 72abd30a585da23e007b4d22daa712eb751b426e
author Hussain Towaileb <Hussain.Towaileb@Couchbase.com> Thu Jul 16 12:02:17 2020 +0300
committer Hussain Towaileb <hussainht@gmail.com> Mon Jul 20 17:37:08 2020 +0000
tree 61e927e6c019e89fa75e98ec6afb92a3aada9d70
parent d4ca9254a95dd18c34cad491186d9961a372f0af

[ASTERIXDB-2685][FUN] regex-split avoid creating unnecessary objects if not needed

- user model changes: no
- storage format changes: no
- interface changes: no

Details:
- Switched to using the RegExpMatcher class to avoid creating
  unnecessary objects (compiling patterns) if the next tuple
  pattern is similar to the previous one.

Change-Id: Ie6920fb049f7b333e3d41de154839d83b5280926
Reviewed-on: https://asterix-gerrit.ics.uci.edu/c/asterixdb/+/7223
Integration-Tests: Jenkins <jenkins@fulliautomatix.ics.uci.edu>
Tested-by: Jenkins <jenkins@fulliautomatix.ics.uci.edu>
Reviewed-by: Hussain Towaileb <hussainht@gmail.com>
Reviewed-by: Dmitry Lychagin <dmitry.lychagin@couchbase.com>

asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/StringRegExpSplitDescriptor.java

diff --git a/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/StringRegExpSplitDescriptor.java b/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/StringRegExpSplitDescriptor.java
index da6a206..8de8c8a 100644
--- a/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/StringRegExpSplitDescriptor.java
+++ b/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/StringRegExpSplitDescriptor.java
@@ -30,6 +30,7 @@
 import org.apache.asterix.om.types.AbstractCollectionType;
 import org.apache.asterix.om.types.BuiltinType;
 import org.apache.asterix.runtime.evaluators.base.AbstractScalarFunctionDynamicDescriptor;
+import org.apache.asterix.runtime.evaluators.functions.utils.RegExpMatcher;
 import org.apache.hyracks.algebricks.core.algebra.functions.FunctionIdentifier;
 import org.apache.hyracks.algebricks.runtime.base.IEvaluatorContext;
 import org.apache.hyracks.algebricks.runtime.base.IScalarEvaluator;
@@ -65,10 +66,13 @@
                     private final AbstractCollectionType collectionType =
                             new AOrderedListType(BuiltinType.ASTRING, BuiltinType.ASTRING.getTypeName());
 
+                    private final RegExpMatcher matcher = new RegExpMatcher();
+
                     @Override
                     protected void process(UTF8StringPointable srcPtr, UTF8StringPointable patternPtr,
                             IPointable result) throws HyracksDataException {
-                        String[] splits = srcPtr.toString().split(patternPtr.toString());
+                        matcher.build(srcPtr, patternPtr);
+                        String[] splits = matcher.split();
 
                         // Result is a list of type strings
                         listBuilder.reset(collectionType);

asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/utils/RegExpMatcher.java

diff --git a/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/utils/RegExpMatcher.java b/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/utils/RegExpMatcher.java
index 778df5b..1a190cc 100644
--- a/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/utils/RegExpMatcher.java
+++ b/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/evaluators/functions/utils/RegExpMatcher.java
@@ -209,4 +209,13 @@
         matcher.appendTail(resultBuf);
         return resultBuf.toString();
     }
+
+    /**
+     * Splits the provided source string using the provided regular expression.
+     *
+     * @return the string split tokens
+     */
+    public String[] split() {
+        return pattern.split(charSeq);
+    }
 }