spark-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From marmb...@apache.org
Subject git commit: [SQL] Code Cleanup: Left Semi Hash Join
Date Wed, 11 Jun 2014 19:10:07 GMT
Repository: spark
Updated Branches:
  refs/heads/branch-1.0 cc004488d -> 597c7efdc


[SQL] Code Cleanup: Left Semi Hash Join

Some improvement for PR #837, add another case to white list and use `filter` to build result
iterator.

Author: Daoyuan <daoyuan.wang@intel.com>

Closes #1049 from adrian-wang/clean-LeftSemiJoinHash and squashes the following commits:

b314d5a [Daoyuan] change hashSet name
27579a9 [Daoyuan] add semijoin to white list and use filter to create new iterator in LeftSemiJoinBNL

Signed-off-by: Michael Armbrust <michael@databricks.com>
(cherry picked from commit ce6deb1e5b4cd40c97730fcf5dc89cb2f624bce2)
Signed-off-by: Michael Armbrust <michael@databricks.com>


Project: http://git-wip-us.apache.org/repos/asf/spark/repo
Commit: http://git-wip-us.apache.org/repos/asf/spark/commit/597c7efd
Tree: http://git-wip-us.apache.org/repos/asf/spark/tree/597c7efd
Diff: http://git-wip-us.apache.org/repos/asf/spark/diff/597c7efd

Branch: refs/heads/branch-1.0
Commit: 597c7efdcfc6879478b606d27fdb1a8d42372e1a
Parents: cc00448
Author: Daoyuan <daoyuan.wang@intel.com>
Authored: Wed Jun 11 12:08:28 2014 -0700
Committer: Michael Armbrust <michael@databricks.com>
Committed: Wed Jun 11 12:10:00 2014 -0700

----------------------------------------------------------------------
 .../org/apache/spark/sql/execution/joins.scala  | 40 ++++---------------
 .../semijoin-0-1631b71327abf75b96116036b977b26c |  0
 .../semijoin-1-e7c99e1f46d502edbb0925d75aab5f0c | 11 +++++
 ...semijoin-10-ffd4fb3a903a6725ccb97d5451a3fec6 |  0
 .../semijoin-11-246a40dcafe077f02397e30d227c330 |  8 ++++
 ...semijoin-12-6d93a9d332ba490835b17f261a5467df |  0
 ...semijoin-13-18282d38b6efc0017089ab89b661764f |  0
 ...semijoin-14-19cfcefb10e1972bec0ffd421cd79de7 |  0
 .../semijoin-15-1de6eb3f357bd1c4d02ab4d19d43589 | 11 +++++
 .../semijoin-16-d3a72a90515ac4a8d8e9ac923bcda3d |  0
 ...semijoin-17-f0f8720cfd11fd71af17b310dc2d1019 |  3 ++
 ...semijoin-18-f7b2ce472443982e32d954cbb5c96765 |  0
 ...semijoin-19-b1f1c7f701abe81c01e72fb98f0bd13f |  3 ++
 .../semijoin-2-deb9c3286ae8e851b1fdb270085b16bc |  0
 ...semijoin-20-b7a8ebaeb42b2eaba7d97cadc3fd96c1 |  0
 ...semijoin-21-480418a0646cf7260b494b9eb4821bb6 |  0
 ...semijoin-22-b6aebd98f7636cda7b24e0bf84d7ba41 |  0
 ...semijoin-23-ed730ccdf552c07e7a82ba6b7fd3fbda |  5 +++
 ...semijoin-24-d16b37134de78980b2bf96029e8265c3 |  0
 ...semijoin-25-be2bd011cc80b480b271a08dbf381e9b | 19 +++++++++
 ...semijoin-26-f1d3bab29f1ebafa148dbe3816e1da25 |  0
 ...semijoin-27-391c256254d171973f02e7f33672ce1d |  4 ++
 ...semijoin-28-b56400f6d9372f353cf7292a2182e963 |  0
 ...semijoin-29-9efeef3d3c38e22a74d379978178c4f5 | 14 +++++++
 .../semijoin-3-b4d4317dd3a10e18502f20f5c5250389 | 11 +++++
 ...semijoin-30-dd901d00fce5898b03a57cbc3028a70a |  0
 .../semijoin-31-e5dc4d8185e63e984aa4e3a2e08bd67 | 14 +++++++
 ...semijoin-32-23017c7663f2710265a7e2a4a1606d39 |  0
 ...semijoin-33-ed499f94c6e6ac847ef5187b3b43bbc5 | 14 +++++++
 ...semijoin-34-5e1b832090ab73c141c1167d5b25a490 |  0
 ...semijoin-35-8d5731f26232f6e26dd8012461b08d99 | 26 ++++++++++++
 ...semijoin-36-b1159823dca8025926407f8aa921238d |  0
 ...semijoin-37-a15b9074f999ce836be5329591b968d0 | 29 ++++++++++++++
 .../semijoin-38-f37547c73a48ce3ba089531b176e6ba |  0
 ...semijoin-39-c22a6f11368affcb80a9c80e653a47a8 | 29 ++++++++++++++
 .../semijoin-4-dfdad5a2742f93e8ea888191460809c0 |  0
 ...semijoin-40-32071a51e2ba6e86b1c5e40de55aae63 |  0
 ...semijoin-41-cf74f73a33b1af8902b7970a9350b092 | 29 ++++++++++++++
 ...semijoin-42-6b4257a74fca627785c967c99547f4c0 |  0
 ...semijoin-43-e8a166ac2e94bf8d1da0fe91b0db2c81 | 31 +++++++++++++++
 ...semijoin-44-945aaa3a24359ef73acab1e99500d5ea |  0
 ...semijoin-45-3fd94ffd4f1eb6cf83dcc064599bf12b | 42 ++++++++++++++++++++
 ...semijoin-46-620e01f81f6e5254b4bbe8fab4043ec0 |  0
 ...semijoin-47-f0140e4ee92508ba241f91c157b7af9c | 35 ++++++++++++++++
 ...semijoin-48-8a04442e84f99a584c2882d0af8c25d8 |  0
 ...semijoin-49-df1d6705d3624be72036318a6b42f04c |  0
 .../semijoin-5-d3c2f84a12374b307c58a69aba4ec70d | 22 ++++++++++
 .../semijoin-6-90bb51b1330230d10a14fb7517457aa0 |  0
 .../semijoin-7-333d72e8bce6d11a35fc7a30418f225b |  0
 .../semijoin-8-d46607be851a6f4e27e98cbbefdee994 |  0
 .../semijoin-9-f7adaf0f77ce6ff8c3a4807f428d8de2 |  6 +++
 .../hive/execution/HiveCompatibilitySuite.scala |  1 +
 52 files changed, 374 insertions(+), 33 deletions(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/core/src/main/scala/org/apache/spark/sql/execution/joins.scala
----------------------------------------------------------------------
diff --git a/sql/core/src/main/scala/org/apache/spark/sql/execution/joins.scala b/sql/core/src/main/scala/org/apache/spark/sql/execution/joins.scala
index 88ff3d4..8d7a5ba 100644
--- a/sql/core/src/main/scala/org/apache/spark/sql/execution/joins.scala
+++ b/sql/core/src/main/scala/org/apache/spark/sql/execution/joins.scala
@@ -169,7 +169,7 @@ case class LeftSemiJoinHash(
   def execute() = {
 
     buildPlan.execute().zipPartitions(streamedPlan.execute()) { (buildIter, streamIter) =>
-      val hashTable = new java.util.HashSet[Row]()
+      val hashSet = new java.util.HashSet[Row]()
       var currentRow: Row = null
 
       // Create a Hash set of buildKeys
@@ -177,43 +177,17 @@ case class LeftSemiJoinHash(
         currentRow = buildIter.next()
         val rowKey = buildSideKeyGenerator(currentRow)
         if(!rowKey.anyNull) {
-          val keyExists = hashTable.contains(rowKey)
+          val keyExists = hashSet.contains(rowKey)
           if (!keyExists) {
-            hashTable.add(rowKey)
+            hashSet.add(rowKey)
           }
         }
       }
 
-      new Iterator[Row] {
-        private[this] var currentStreamedRow: Row = _
-        private[this] var currentHashMatched: Boolean = false
-
-        private[this] val joinKeys = streamSideKeyGenerator()
-
-        override final def hasNext: Boolean =
-          streamIter.hasNext && fetchNext()
-
-        override final def next() = {
-          currentStreamedRow
-        }
-
-        /**
-         * Searches the streamed iterator for the next row that has at least one match in
hashtable.
-         *
-         * @return true if the search is successful, and false the streamed iterator runs
out of
-         *         tuples.
-         */
-        private final def fetchNext(): Boolean = {
-          currentHashMatched = false
-          while (!currentHashMatched && streamIter.hasNext) {
-            currentStreamedRow = streamIter.next()
-            if (!joinKeys(currentStreamedRow).anyNull) {
-              currentHashMatched = hashTable.contains(joinKeys.currentValue)
-            }
-          }
-          currentHashMatched
-        }
-      }
+      val joinKeys = streamSideKeyGenerator()
+      streamIter.filter(current => {
+        !joinKeys(current).anyNull && hashSet.contains(joinKeys.currentValue)
+      })
     }
   }
 }

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-0-1631b71327abf75b96116036b977b26c
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-0-1631b71327abf75b96116036b977b26c
b/sql/hive/src/test/resources/golden/semijoin-0-1631b71327abf75b96116036b977b26c
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-1-e7c99e1f46d502edbb0925d75aab5f0c
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-1-e7c99e1f46d502edbb0925d75aab5f0c
b/sql/hive/src/test/resources/golden/semijoin-1-e7c99e1f46d502edbb0925d75aab5f0c
new file mode 100644
index 0000000..2ed47ab
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-1-e7c99e1f46d502edbb0925d75aab5f0c
@@ -0,0 +1,11 @@
+0	val_0
+0	val_0
+0	val_0
+2	val_2
+4	val_4
+5	val_5
+5	val_5
+5	val_5
+8	val_8
+9	val_9
+10	val_10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-10-ffd4fb3a903a6725ccb97d5451a3fec6
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-10-ffd4fb3a903a6725ccb97d5451a3fec6
b/sql/hive/src/test/resources/golden/semijoin-10-ffd4fb3a903a6725ccb97d5451a3fec6
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-11-246a40dcafe077f02397e30d227c330
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-11-246a40dcafe077f02397e30d227c330
b/sql/hive/src/test/resources/golden/semijoin-11-246a40dcafe077f02397e30d227c330
new file mode 100644
index 0000000..a24bd8c
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-11-246a40dcafe077f02397e30d227c330
@@ -0,0 +1,8 @@
+0	val_0
+0	val_0
+0	val_0
+4	val_2
+8	val_4
+10	val_5
+10	val_5
+10	val_5

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-12-6d93a9d332ba490835b17f261a5467df
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-12-6d93a9d332ba490835b17f261a5467df
b/sql/hive/src/test/resources/golden/semijoin-12-6d93a9d332ba490835b17f261a5467df
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-13-18282d38b6efc0017089ab89b661764f
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-13-18282d38b6efc0017089ab89b661764f
b/sql/hive/src/test/resources/golden/semijoin-13-18282d38b6efc0017089ab89b661764f
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-14-19cfcefb10e1972bec0ffd421cd79de7
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-14-19cfcefb10e1972bec0ffd421cd79de7
b/sql/hive/src/test/resources/golden/semijoin-14-19cfcefb10e1972bec0ffd421cd79de7
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-15-1de6eb3f357bd1c4d02ab4d19d43589
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-15-1de6eb3f357bd1c4d02ab4d19d43589
b/sql/hive/src/test/resources/golden/semijoin-15-1de6eb3f357bd1c4d02ab4d19d43589
new file mode 100644
index 0000000..03c61a9
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-15-1de6eb3f357bd1c4d02ab4d19d43589
@@ -0,0 +1,11 @@
+val_0
+val_0
+val_0
+val_10
+val_2
+val_4
+val_5
+val_5
+val_5
+val_8
+val_9

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-16-d3a72a90515ac4a8d8e9ac923bcda3d
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-16-d3a72a90515ac4a8d8e9ac923bcda3d
b/sql/hive/src/test/resources/golden/semijoin-16-d3a72a90515ac4a8d8e9ac923bcda3d
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-17-f0f8720cfd11fd71af17b310dc2d1019
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-17-f0f8720cfd11fd71af17b310dc2d1019
b/sql/hive/src/test/resources/golden/semijoin-17-f0f8720cfd11fd71af17b310dc2d1019
new file mode 100644
index 0000000..2dcdfd1
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-17-f0f8720cfd11fd71af17b310dc2d1019
@@ -0,0 +1,3 @@
+0	val_0
+0	val_0
+0	val_0

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-18-f7b2ce472443982e32d954cbb5c96765
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-18-f7b2ce472443982e32d954cbb5c96765
b/sql/hive/src/test/resources/golden/semijoin-18-f7b2ce472443982e32d954cbb5c96765
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-19-b1f1c7f701abe81c01e72fb98f0bd13f
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-19-b1f1c7f701abe81c01e72fb98f0bd13f
b/sql/hive/src/test/resources/golden/semijoin-19-b1f1c7f701abe81c01e72fb98f0bd13f
new file mode 100644
index 0000000..a367051
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-19-b1f1c7f701abe81c01e72fb98f0bd13f
@@ -0,0 +1,3 @@
+val_10
+val_8
+val_9

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-2-deb9c3286ae8e851b1fdb270085b16bc
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-2-deb9c3286ae8e851b1fdb270085b16bc
b/sql/hive/src/test/resources/golden/semijoin-2-deb9c3286ae8e851b1fdb270085b16bc
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-20-b7a8ebaeb42b2eaba7d97cadc3fd96c1
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-20-b7a8ebaeb42b2eaba7d97cadc3fd96c1
b/sql/hive/src/test/resources/golden/semijoin-20-b7a8ebaeb42b2eaba7d97cadc3fd96c1
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-21-480418a0646cf7260b494b9eb4821bb6
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-21-480418a0646cf7260b494b9eb4821bb6
b/sql/hive/src/test/resources/golden/semijoin-21-480418a0646cf7260b494b9eb4821bb6
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-22-b6aebd98f7636cda7b24e0bf84d7ba41
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-22-b6aebd98f7636cda7b24e0bf84d7ba41
b/sql/hive/src/test/resources/golden/semijoin-22-b6aebd98f7636cda7b24e0bf84d7ba41
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-23-ed730ccdf552c07e7a82ba6b7fd3fbda
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-23-ed730ccdf552c07e7a82ba6b7fd3fbda
b/sql/hive/src/test/resources/golden/semijoin-23-ed730ccdf552c07e7a82ba6b7fd3fbda
new file mode 100644
index 0000000..72bc6a6
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-23-ed730ccdf552c07e7a82ba6b7fd3fbda
@@ -0,0 +1,5 @@
+4	val_2
+8	val_4
+10	val_5
+10	val_5
+10	val_5

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-24-d16b37134de78980b2bf96029e8265c3
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-24-d16b37134de78980b2bf96029e8265c3
b/sql/hive/src/test/resources/golden/semijoin-24-d16b37134de78980b2bf96029e8265c3
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-25-be2bd011cc80b480b271a08dbf381e9b
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-25-be2bd011cc80b480b271a08dbf381e9b
b/sql/hive/src/test/resources/golden/semijoin-25-be2bd011cc80b480b271a08dbf381e9b
new file mode 100644
index 0000000..d89ea17
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-25-be2bd011cc80b480b271a08dbf381e9b
@@ -0,0 +1,19 @@
+0
+0
+0
+0
+0
+0
+2
+4
+4
+5
+5
+5
+8
+8
+9
+10
+10
+10
+10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-26-f1d3bab29f1ebafa148dbe3816e1da25
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-26-f1d3bab29f1ebafa148dbe3816e1da25
b/sql/hive/src/test/resources/golden/semijoin-26-f1d3bab29f1ebafa148dbe3816e1da25
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-27-391c256254d171973f02e7f33672ce1d
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-27-391c256254d171973f02e7f33672ce1d
b/sql/hive/src/test/resources/golden/semijoin-27-391c256254d171973f02e7f33672ce1d
new file mode 100644
index 0000000..dbbdae7
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-27-391c256254d171973f02e7f33672ce1d
@@ -0,0 +1,4 @@
+0	val_0
+0	val_0
+0	val_0
+8	val_8

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-28-b56400f6d9372f353cf7292a2182e963
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-28-b56400f6d9372f353cf7292a2182e963
b/sql/hive/src/test/resources/golden/semijoin-28-b56400f6d9372f353cf7292a2182e963
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-29-9efeef3d3c38e22a74d379978178c4f5
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-29-9efeef3d3c38e22a74d379978178c4f5
b/sql/hive/src/test/resources/golden/semijoin-29-9efeef3d3c38e22a74d379978178c4f5
new file mode 100644
index 0000000..07c61af
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-29-9efeef3d3c38e22a74d379978178c4f5
@@ -0,0 +1,14 @@
+0	val_0	0	val_0
+0	val_0	0	val_0
+0	val_0	0	val_0
+0	val_0	0	val_0
+0	val_0	0	val_0
+0	val_0	0	val_0
+0	val_0	0	val_0
+0	val_0	0	val_0
+0	val_0	0	val_0
+4	val_4	4	val_2
+8	val_8	8	val_4
+10	val_10	10	val_5
+10	val_10	10	val_5
+10	val_10	10	val_5

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-3-b4d4317dd3a10e18502f20f5c5250389
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-3-b4d4317dd3a10e18502f20f5c5250389
b/sql/hive/src/test/resources/golden/semijoin-3-b4d4317dd3a10e18502f20f5c5250389
new file mode 100644
index 0000000..bf51e8f
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-3-b4d4317dd3a10e18502f20f5c5250389
@@ -0,0 +1,11 @@
+0	val_0
+0	val_0
+0	val_0
+4	val_2
+8	val_4
+10	val_5
+10	val_5
+10	val_5
+16	val_8
+18	val_9
+20	val_10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-30-dd901d00fce5898b03a57cbc3028a70a
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-30-dd901d00fce5898b03a57cbc3028a70a
b/sql/hive/src/test/resources/golden/semijoin-30-dd901d00fce5898b03a57cbc3028a70a
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-31-e5dc4d8185e63e984aa4e3a2e08bd67
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-31-e5dc4d8185e63e984aa4e3a2e08bd67
b/sql/hive/src/test/resources/golden/semijoin-31-e5dc4d8185e63e984aa4e3a2e08bd67
new file mode 100644
index 0000000..d6283e3
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-31-e5dc4d8185e63e984aa4e3a2e08bd67
@@ -0,0 +1,14 @@
+0	val_0
+0	val_0
+0	val_0
+0	val_0
+0	val_0
+0	val_0
+2	val_2
+4	val_4
+5	val_5
+5	val_5
+5	val_5
+8	val_8
+9	val_9
+10	val_10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-32-23017c7663f2710265a7e2a4a1606d39
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-32-23017c7663f2710265a7e2a4a1606d39
b/sql/hive/src/test/resources/golden/semijoin-32-23017c7663f2710265a7e2a4a1606d39
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-33-ed499f94c6e6ac847ef5187b3b43bbc5
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-33-ed499f94c6e6ac847ef5187b3b43bbc5
b/sql/hive/src/test/resources/golden/semijoin-33-ed499f94c6e6ac847ef5187b3b43bbc5
new file mode 100644
index 0000000..080180f
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-33-ed499f94c6e6ac847ef5187b3b43bbc5
@@ -0,0 +1,14 @@
+0
+0
+0
+0
+0
+0
+4
+4
+8
+8
+10
+10
+10
+10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-34-5e1b832090ab73c141c1167d5b25a490
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-34-5e1b832090ab73c141c1167d5b25a490
b/sql/hive/src/test/resources/golden/semijoin-34-5e1b832090ab73c141c1167d5b25a490
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-35-8d5731f26232f6e26dd8012461b08d99
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-35-8d5731f26232f6e26dd8012461b08d99
b/sql/hive/src/test/resources/golden/semijoin-35-8d5731f26232f6e26dd8012461b08d99
new file mode 100644
index 0000000..4a64d5c
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-35-8d5731f26232f6e26dd8012461b08d99
@@ -0,0 +1,26 @@
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+4
+4
+8
+8
+10
+10
+10
+10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-36-b1159823dca8025926407f8aa921238d
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-36-b1159823dca8025926407f8aa921238d
b/sql/hive/src/test/resources/golden/semijoin-36-b1159823dca8025926407f8aa921238d
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-37-a15b9074f999ce836be5329591b968d0
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-37-a15b9074f999ce836be5329591b968d0
b/sql/hive/src/test/resources/golden/semijoin-37-a15b9074f999ce836be5329591b968d0
new file mode 100644
index 0000000..1420c78
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-37-a15b9074f999ce836be5329591b968d0
@@ -0,0 +1,29 @@
+NULL
+NULL
+NULL
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+4
+4
+8
+8
+10
+10
+10
+10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-38-f37547c73a48ce3ba089531b176e6ba
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-38-f37547c73a48ce3ba089531b176e6ba
b/sql/hive/src/test/resources/golden/semijoin-38-f37547c73a48ce3ba089531b176e6ba
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-39-c22a6f11368affcb80a9c80e653a47a8
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-39-c22a6f11368affcb80a9c80e653a47a8
b/sql/hive/src/test/resources/golden/semijoin-39-c22a6f11368affcb80a9c80e653a47a8
new file mode 100644
index 0000000..1420c78
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-39-c22a6f11368affcb80a9c80e653a47a8
@@ -0,0 +1,29 @@
+NULL
+NULL
+NULL
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+4
+4
+8
+8
+10
+10
+10
+10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-4-dfdad5a2742f93e8ea888191460809c0
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-4-dfdad5a2742f93e8ea888191460809c0
b/sql/hive/src/test/resources/golden/semijoin-4-dfdad5a2742f93e8ea888191460809c0
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-40-32071a51e2ba6e86b1c5e40de55aae63
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-40-32071a51e2ba6e86b1c5e40de55aae63
b/sql/hive/src/test/resources/golden/semijoin-40-32071a51e2ba6e86b1c5e40de55aae63
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-41-cf74f73a33b1af8902b7970a9350b092
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-41-cf74f73a33b1af8902b7970a9350b092
b/sql/hive/src/test/resources/golden/semijoin-41-cf74f73a33b1af8902b7970a9350b092
new file mode 100644
index 0000000..aef9483
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-41-cf74f73a33b1af8902b7970a9350b092
@@ -0,0 +1,29 @@
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+4
+4
+8
+8
+10
+10
+10
+10
+16
+18
+20

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-42-6b4257a74fca627785c967c99547f4c0
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-42-6b4257a74fca627785c967c99547f4c0
b/sql/hive/src/test/resources/golden/semijoin-42-6b4257a74fca627785c967c99547f4c0
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-43-e8a166ac2e94bf8d1da0fe91b0db2c81
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-43-e8a166ac2e94bf8d1da0fe91b0db2c81
b/sql/hive/src/test/resources/golden/semijoin-43-e8a166ac2e94bf8d1da0fe91b0db2c81
new file mode 100644
index 0000000..0bc413e
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-43-e8a166ac2e94bf8d1da0fe91b0db2c81
@@ -0,0 +1,31 @@
+NULL
+NULL
+NULL
+NULL
+NULL
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+4
+4
+8
+8
+10
+10
+10
+10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-44-945aaa3a24359ef73acab1e99500d5ea
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-44-945aaa3a24359ef73acab1e99500d5ea
b/sql/hive/src/test/resources/golden/semijoin-44-945aaa3a24359ef73acab1e99500d5ea
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-45-3fd94ffd4f1eb6cf83dcc064599bf12b
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-45-3fd94ffd4f1eb6cf83dcc064599bf12b
b/sql/hive/src/test/resources/golden/semijoin-45-3fd94ffd4f1eb6cf83dcc064599bf12b
new file mode 100644
index 0000000..3131e64
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-45-3fd94ffd4f1eb6cf83dcc064599bf12b
@@ -0,0 +1,42 @@
+NULL
+NULL
+NULL
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+2
+4
+4
+5
+5
+5
+8
+8
+9
+10
+10
+10
+10
+10
+10
+10
+10
+10
+10
+10
+10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-46-620e01f81f6e5254b4bbe8fab4043ec0
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-46-620e01f81f6e5254b4bbe8fab4043ec0
b/sql/hive/src/test/resources/golden/semijoin-46-620e01f81f6e5254b4bbe8fab4043ec0
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-47-f0140e4ee92508ba241f91c157b7af9c
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-47-f0140e4ee92508ba241f91c157b7af9c
b/sql/hive/src/test/resources/golden/semijoin-47-f0140e4ee92508ba241f91c157b7af9c
new file mode 100644
index 0000000..ff30bed
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-47-f0140e4ee92508ba241f91c157b7af9c
@@ -0,0 +1,35 @@
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+4
+4
+8
+8
+10
+10
+10
+10
+10
+10
+10
+10
+10
+10
+16
+18
+20

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-48-8a04442e84f99a584c2882d0af8c25d8
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-48-8a04442e84f99a584c2882d0af8c25d8
b/sql/hive/src/test/resources/golden/semijoin-48-8a04442e84f99a584c2882d0af8c25d8
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-49-df1d6705d3624be72036318a6b42f04c
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-49-df1d6705d3624be72036318a6b42f04c
b/sql/hive/src/test/resources/golden/semijoin-49-df1d6705d3624be72036318a6b42f04c
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-5-d3c2f84a12374b307c58a69aba4ec70d
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-5-d3c2f84a12374b307c58a69aba4ec70d
b/sql/hive/src/test/resources/golden/semijoin-5-d3c2f84a12374b307c58a69aba4ec70d
new file mode 100644
index 0000000..60f6eac
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-5-d3c2f84a12374b307c58a69aba4ec70d
@@ -0,0 +1,22 @@
+0	val_0
+0	val_0
+0	val_0
+0	val_0
+0	val_0
+0	val_0
+2	val_2
+4	val_2
+4	val_4
+5	val_5
+5	val_5
+5	val_5
+8	val_4
+8	val_8
+9	val_9
+10	val_10
+10	val_5
+10	val_5
+10	val_5
+16	val_8
+18	val_9
+20	val_10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-6-90bb51b1330230d10a14fb7517457aa0
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-6-90bb51b1330230d10a14fb7517457aa0
b/sql/hive/src/test/resources/golden/semijoin-6-90bb51b1330230d10a14fb7517457aa0
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-7-333d72e8bce6d11a35fc7a30418f225b
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-7-333d72e8bce6d11a35fc7a30418f225b
b/sql/hive/src/test/resources/golden/semijoin-7-333d72e8bce6d11a35fc7a30418f225b
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-8-d46607be851a6f4e27e98cbbefdee994
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-8-d46607be851a6f4e27e98cbbefdee994
b/sql/hive/src/test/resources/golden/semijoin-8-d46607be851a6f4e27e98cbbefdee994
new file mode 100644
index 0000000..e69de29

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/resources/golden/semijoin-9-f7adaf0f77ce6ff8c3a4807f428d8de2
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/resources/golden/semijoin-9-f7adaf0f77ce6ff8c3a4807f428d8de2
b/sql/hive/src/test/resources/golden/semijoin-9-f7adaf0f77ce6ff8c3a4807f428d8de2
new file mode 100644
index 0000000..5baaac9
--- /dev/null
+++ b/sql/hive/src/test/resources/golden/semijoin-9-f7adaf0f77ce6ff8c3a4807f428d8de2
@@ -0,0 +1,6 @@
+0	val_0
+0	val_0
+0	val_0
+4	val_4
+8	val_8
+10	val_10

http://git-wip-us.apache.org/repos/asf/spark/blob/597c7efd/sql/hive/src/test/scala/org/apache/spark/sql/hive/execution/HiveCompatibilitySuite.scala
----------------------------------------------------------------------
diff --git a/sql/hive/src/test/scala/org/apache/spark/sql/hive/execution/HiveCompatibilitySuite.scala
b/sql/hive/src/test/scala/org/apache/spark/sql/hive/execution/HiveCompatibilitySuite.scala
index fb8f272..3581617 100644
--- a/sql/hive/src/test/scala/org/apache/spark/sql/hive/execution/HiveCompatibilitySuite.scala
+++ b/sql/hive/src/test/scala/org/apache/spark/sql/hive/execution/HiveCompatibilitySuite.scala
@@ -597,6 +597,7 @@ class HiveCompatibilitySuite extends HiveQueryFileTest with BeforeAndAfter
{
     "select_unquote_and",
     "select_unquote_not",
     "select_unquote_or",
+    "semijoin",
     "serde_regex",
     "serde_reported_schema",
     "set_variable_sub",


Mime
View raw message