You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@hive.apache.org by ha...@apache.org on 2014/09/05 17:59:50 UTC
svn commit: r1622729 [5/8] - in /hive/trunk:
ant/src/org/apache/hadoop/hive/ant/
ql/src/gen/vectorization/ExpressionTemplates/
ql/src/java/org/apache/hadoop/hive/ql/exec/vector/
ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/ ql/src/java...
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/PosModDoubleToDouble.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/PosModDoubleToDouble.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/PosModDoubleToDouble.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/PosModDoubleToDouble.java Fri Sep 5 15:59:47 2014
@@ -61,8 +61,8 @@ public class PosModDoubleToDouble extend
VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.LONG,
- VectorExpressionDescriptor.ArgumentType.DOUBLE)
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.FLOAT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.SCALAR).build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/PosModLongToLong.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/PosModLongToLong.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/PosModLongToLong.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/PosModLongToLong.java Fri Sep 5 15:59:47 2014
@@ -61,8 +61,8 @@ public class PosModLongToLong extends Ma
VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.LONG,
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.SCALAR).build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/RoundWithNumDigitsDoubleToDouble.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/RoundWithNumDigitsDoubleToDouble.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/RoundWithNumDigitsDoubleToDouble.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/RoundWithNumDigitsDoubleToDouble.java Fri Sep 5 15:59:47 2014
@@ -64,8 +64,8 @@ public class RoundWithNumDigitsDoubleToD
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.DOUBLE,
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.FLOAT_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.SCALAR);
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsFalse.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsFalse.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsFalse.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsFalse.java Fri Sep 5 15:59:47 2014
@@ -144,7 +144,7 @@ public class SelectColumnIsFalse extends
b.setMode(VectorExpressionDescriptor.Mode.FILTER)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsNotNull.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsNotNull.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsNotNull.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsNotNull.java Fri Sep 5 15:59:47 2014
@@ -113,7 +113,7 @@ public class SelectColumnIsNotNull exten
b.setMode(VectorExpressionDescriptor.Mode.FILTER)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.ANY)
+ VectorExpressionDescriptor.ArgumentType.ALL_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsNull.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsNull.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsNull.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsNull.java Fri Sep 5 15:59:47 2014
@@ -111,7 +111,7 @@ public class SelectColumnIsNull extends
b.setMode(VectorExpressionDescriptor.Mode.FILTER)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.ANY)
+ VectorExpressionDescriptor.ArgumentType.ALL_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsTrue.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsTrue.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsTrue.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/SelectColumnIsTrue.java Fri Sep 5 15:59:47 2014
@@ -143,7 +143,7 @@ public class SelectColumnIsTrue extends
b.setMode(VectorExpressionDescriptor.Mode.FILTER)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringExpr.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringExpr.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringExpr.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringExpr.java Fri Sep 5 15:59:47 2014
@@ -18,6 +18,10 @@
package org.apache.hadoop.hive.ql.exec.vector.expressions;
+import java.util.Arrays;
+
+import org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector;
+
/**
* String expression evaluation helper functions.
*/
@@ -32,6 +36,7 @@ public class StringExpr {
*/
public static int compare(byte[] arg1, int start1, int len1, byte[] arg2, int start2, int len2) {
for (int i = 0; i < len1 && i < len2; i++) {
+ // Note the "& 0xff" is just a way to convert unsigned bytes to signed integer.
int b1 = arg1[i + start1] & 0xff;
int b2 = arg2[i + start2] & 0xff;
if (b1 != b2) {
@@ -40,4 +45,259 @@ public class StringExpr {
}
return len1 - len2;
}
+
+ public static int characterCount(byte[] bytes) {
+ int end = bytes.length;
+
+ // count characters
+ int j = 0;
+ int charCount = 0;
+ while(j < end) {
+ // UTF-8 continuation bytes have 2 high bits equal to 0x80.
+ if ((bytes[j] & 0xc0) != 0x80) {
+ ++charCount;
+ }
+ j++;
+ }
+ return charCount;
+ }
+
+ public static int characterCount(byte[] bytes, int start, int length) {
+ int end = start + length;
+
+ // count characters
+ int j = start;
+ int charCount = 0;
+ while(j < end) {
+ // UTF-8 continuation bytes have 2 high bits equal to 0x80.
+ if ((bytes[j] & 0xc0) != 0x80) {
+ ++charCount;
+ }
+ j++;
+ }
+ return charCount;
+ }
+
+ // A setVal with the same function signature as rightTrim, leftTrim, truncate, etc, below.
+ // Useful for class generation via templates.
+ public static void assign(BytesColumnVector outV, int i, byte[] bytes, int start, int length) {
+ // set output vector
+ outV.setVal(i, bytes, start, length);
+ }
+
+ /*
+ * Right trim a slice of a byte array and return the new byte length.
+ */
+ public static int rightTrim(byte[] bytes, int start, int length) {
+ // skip trailing blank characters
+ int j = start + length - 1;
+ while(j >= start && bytes[j] == 0x20) {
+ j--;
+ }
+
+ return (j - start) + 1;
+ }
+
+ /*
+ * Right trim a slice of a byte array and place the result into element i of a vector.
+ */
+ public static void rightTrim(BytesColumnVector outV, int i, byte[] bytes, int start, int length) {
+ // skip trailing blank characters
+ int j = start + length - 1;
+ while(j >= start && bytes[j] == 0x20) {
+ j--;
+ }
+
+ // set output vector
+ outV.setVal(i, bytes, start, (j - start) + 1);
+ }
+
+ /*
+ * Truncate a slice of a byte array to a maximum number of characters and
+ * return the new byte length.
+ */
+ public static int truncate(byte[] bytes, int start, int length, int maxLength) {
+ int end = start + length;
+
+ // count characters forward
+ int j = start;
+ int charCount = 0;
+ while(j < end) {
+ // UTF-8 continuation bytes have 2 high bits equal to 0x80.
+ if ((bytes[j] & 0xc0) != 0x80) {
+ if (charCount == maxLength) {
+ break;
+ }
+ ++charCount;
+ }
+ j++;
+ }
+ return (j - start);
+ }
+
+ /*
+ * Truncate a slice of a byte array to a maximum number of characters and
+ * place the result into element i of a vector.
+ */
+ public static void truncate(BytesColumnVector outV, int i, byte[] bytes, int start, int length, int maxLength) {
+ int end = start + length;
+
+ // count characters forward
+ int j = start;
+ int charCount = 0;
+ while(j < end) {
+ // UTF-8 continuation bytes have 2 high bits equal to 0x80.
+ if ((bytes[j] & 0xc0) != 0x80) {
+ if (charCount == maxLength) {
+ break;
+ }
+ ++charCount;
+ }
+ j++;
+ }
+
+ // set output vector
+ outV.setVal(i, bytes, start, (j - start));
+ }
+
+ /*
+ * Truncate a byte array to a maximum number of characters and
+ * return a byte array with only truncated bytes.
+ */
+ public static byte[] truncateScalar(byte[] bytes, int maxLength) {
+ int end = bytes.length;
+
+ // count characters forward
+ int j = 0;
+ int charCount = 0;
+ while(j < end) {
+ // UTF-8 continuation bytes have 2 high bits equal to 0x80.
+ if ((bytes[j] & 0xc0) != 0x80) {
+ if (charCount == maxLength) {
+ break;
+ }
+ ++charCount;
+ }
+ j++;
+ }
+ if (j == end) {
+ return bytes;
+ } else {
+ return Arrays.copyOf(bytes, j);
+ }
+ }
+
+ /*
+ * Right trim and truncate a slice of a byte array to a maximum number of characters and
+ * return the new byte length.
+ */
+ public static int rightTrimAndTruncate(byte[] bytes, int start, int length, int maxLength) {
+ int end = start + length;
+
+ // count characters forward and watch for final run of pads
+ int j = start;
+ int charCount = 0;
+ int padRunStart = -1;
+ while(j < end) {
+ // UTF-8 continuation bytes have 2 high bits equal to 0x80.
+ if ((bytes[j] & 0xc0) != 0x80) {
+ if (charCount == maxLength) {
+ break;
+ }
+ if (bytes[j] == 0x20) {
+ if (padRunStart == -1) {
+ padRunStart = j;
+ }
+ } else {
+ padRunStart = -1;
+ }
+ ++charCount;
+ } else {
+ padRunStart = -1;
+ }
+ j++;
+ }
+ if (padRunStart != -1) {
+ return (padRunStart - start);
+ } else {
+ return (j - start);
+ }
+ }
+
+ /*
+ * Right trim and truncate a slice of a byte array to a maximum number of characters and
+ * place the result into element i of a vector.
+ */
+ public static void rightTrimAndTruncate(BytesColumnVector outV, int i, byte[] bytes, int start, int length, int maxLength) {
+ int end = start + length;
+
+ // count characters forward and watch for final run of pads
+ int j = start;
+ int charCount = 0;
+ int padRunStart = -1;
+ while(j < end) {
+ // UTF-8 continuation bytes have 2 high bits equal to 0x80.
+ if ((bytes[j] & 0xc0) != 0x80) {
+ if (charCount == maxLength) {
+ break;
+ }
+ if (bytes[j] == 0x20) {
+ if (padRunStart == -1) {
+ padRunStart = j;
+ }
+ } else {
+ padRunStart = -1;
+ }
+ ++charCount;
+ } else {
+ padRunStart = -1;
+ }
+ j++;
+ }
+ // set output vector
+ if (padRunStart != -1) {
+ outV.setVal(i, bytes, start, (padRunStart - start));
+ } else {
+ outV.setVal(i, bytes, start, (j - start) );
+ }
+ }
+
+ /*
+ * Right trim and truncate a byte array to a maximum number of characters and
+ * return a byte array with only the trimmed and truncated bytes.
+ */
+ public static byte[] rightTrimAndTruncateScalar(byte[] bytes, int maxLength) {
+ int end = bytes.length;
+
+ // count characters forward and watch for final run of pads
+ int j = 0;
+ int charCount = 0;
+ int padRunStart = -1;
+ while(j < end) {
+ // UTF-8 continuation bytes have 2 high bits equal to 0x80.
+ if ((bytes[j] & 0xc0) != 0x80) {
+ if (charCount == maxLength) {
+ break;
+ }
+ if (bytes[j] == 0x20) {
+ if (padRunStart == -1) {
+ padRunStart = j;
+ }
+ } else {
+ padRunStart = -1;
+ }
+ ++charCount;
+ } else {
+ padRunStart = -1;
+ }
+ j++;
+ }
+ if (padRunStart != -1) {
+ return Arrays.copyOf(bytes, padRunStart);
+ } else if (j == end) {
+ return bytes;
+ } else {
+ return Arrays.copyOf(bytes, j);
+ }
+ }
}
Added: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatCharScalar.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatCharScalar.java?rev=1622729&view=auto
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatCharScalar.java (added)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatCharScalar.java Fri Sep 5 15:59:47 2014
@@ -0,0 +1,52 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.hive.ql.exec.vector.expressions;
+
+import org.apache.hadoop.hive.common.type.HiveChar;
+import org.apache.hadoop.hive.ql.exec.vector.VectorExpressionDescriptor;
+
+/**
+ * Vectorized instruction to concatenate a string column to a scalar and put
+ * the result in an output column.
+ */
+public class StringGroupColConcatCharScalar extends StringGroupColConcatStringScalar {
+ private static final long serialVersionUID = 1L;
+
+ public StringGroupColConcatCharScalar(int colNum, HiveChar value, int outputColumn) {
+ super(colNum, value.getStrippedValue().getBytes(), outputColumn);
+ }
+
+ public StringGroupColConcatCharScalar() {
+ super();
+ }
+
+ @Override
+ public VectorExpressionDescriptor.Descriptor getDescriptor() {
+ return (new VectorExpressionDescriptor.Builder())
+ .setMode(
+ VectorExpressionDescriptor.Mode.PROJECTION)
+ .setNumArguments(2)
+ .setArgumentTypes(
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.CHAR)
+ .setInputExpressionTypes(
+ VectorExpressionDescriptor.InputExpressionType.COLUMN,
+ VectorExpressionDescriptor.InputExpressionType.SCALAR).build();
+ }
+}
Added: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatStringScalar.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatStringScalar.java?rev=1622729&view=auto
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatStringScalar.java (added)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatStringScalar.java Fri Sep 5 15:59:47 2014
@@ -0,0 +1,164 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.hive.ql.exec.vector.expressions;
+
+import org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector;
+import org.apache.hadoop.hive.ql.exec.vector.VectorExpressionDescriptor;
+import org.apache.hadoop.hive.ql.exec.vector.VectorizedRowBatch;
+
+/**
+ * Vectorized instruction to concatenate a string column to a scalar and put
+ * the result in an output column.
+ */
+public class StringGroupColConcatStringScalar extends VectorExpression {
+ private static final long serialVersionUID = 1L;
+ private int colNum;
+ private int outputColumn;
+ private byte[] value;
+
+ public StringGroupColConcatStringScalar(int colNum, byte[] value, int outputColumn) {
+ this();
+ this.colNum = colNum;
+ this.outputColumn = outputColumn;
+ this.value = value;
+ }
+
+ public StringGroupColConcatStringScalar() {
+ super();
+ }
+
+ @Override
+ public void evaluate(VectorizedRowBatch batch) {
+
+ if (childExpressions != null) {
+ super.evaluateChildren(batch);
+ }
+
+ BytesColumnVector inputColVector = (BytesColumnVector) batch.cols[colNum];
+ BytesColumnVector outV = (BytesColumnVector) batch.cols[outputColumn];
+ int[] sel = batch.selected;
+ int n = batch.size;
+ byte[][] vector = inputColVector.vector;
+ int[] start = inputColVector.start;
+ int[] length = inputColVector.length;
+
+ if (n == 0) {
+
+ // Nothing to do
+ return;
+ }
+
+ // initialize output vector buffer to receive data
+ outV.initBuffer();
+
+ if (inputColVector.noNulls) {
+ outV.noNulls = true;
+ if (inputColVector.isRepeating) {
+ outV.isRepeating = true;
+ outV.setConcat(0, vector[0], start[0], length[0], value, 0, value.length);
+ } else if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ outV.setConcat(i, vector[i], start[i], length[i], value, 0, value.length);
+ }
+ outV.isRepeating = false;
+ } else {
+ for(int i = 0; i != n; i++) {
+ outV.setConcat(i, vector[i], start[i], length[i], value, 0, value.length);
+ }
+ outV.isRepeating = false;
+ }
+ } else {
+
+ /*
+ * Handle case with nulls. Don't do function if the value is null, to save time,
+ * because calling the function can be expensive.
+ */
+ outV.noNulls = false;
+ if (inputColVector.isRepeating) {
+ outV.isRepeating = true;
+ outV.isNull[0] = inputColVector.isNull[0];
+ if (!inputColVector.isNull[0]) {
+ outV.setConcat(0, vector[0], start[0], length[0], value, 0, value.length);
+ }
+ } else if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ if (!inputColVector.isNull[i]) {
+ outV.setConcat(i, vector[i], start[i], length[i], value, 0, value.length);
+ }
+ outV.isNull[i] = inputColVector.isNull[i];
+ }
+ outV.isRepeating = false;
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inputColVector.isNull[i]) {
+ outV.setConcat(i, vector[i], start[i], length[i], value, 0, value.length);
+ }
+ outV.isNull[i] = inputColVector.isNull[i];
+ }
+ outV.isRepeating = false;
+ }
+ }
+ }
+
+ @Override
+ public int getOutputColumn() {
+ return outputColumn;
+ }
+
+ @Override
+ public String getOutputType() {
+ return "StringGroup";
+ }
+
+ public int getColNum() {
+ return colNum;
+ }
+
+ public void setColNum(int colNum) {
+ this.colNum = colNum;
+ }
+
+ public byte[] getValue() {
+ return value;
+ }
+
+ public void setValue(byte[] value) {
+ this.value = value;
+ }
+
+ public void setOutputColumn(int outputColumn) {
+ this.outputColumn = outputColumn;
+ }
+
+ @Override
+ public VectorExpressionDescriptor.Descriptor getDescriptor() {
+ return (new VectorExpressionDescriptor.Builder())
+ .setMode(
+ VectorExpressionDescriptor.Mode.PROJECTION)
+ .setNumArguments(2)
+ .setArgumentTypes(
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.STRING)
+ .setInputExpressionTypes(
+ VectorExpressionDescriptor.InputExpressionType.COLUMN,
+ VectorExpressionDescriptor.InputExpressionType.SCALAR).build();
+ }
+}
Added: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatVarCharScalar.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatVarCharScalar.java?rev=1622729&view=auto
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatVarCharScalar.java (added)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupColConcatVarCharScalar.java Fri Sep 5 15:59:47 2014
@@ -0,0 +1,52 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.hive.ql.exec.vector.expressions;
+
+import org.apache.hadoop.hive.common.type.HiveVarchar;
+import org.apache.hadoop.hive.ql.exec.vector.VectorExpressionDescriptor;
+
+/**
+ * Vectorized instruction to concatenate a string column to a scalar and put
+ * the result in an output column.
+ */
+public class StringGroupColConcatVarCharScalar extends StringGroupColConcatStringScalar {
+ private static final long serialVersionUID = 1L;
+
+ public StringGroupColConcatVarCharScalar(int colNum, HiveVarchar value, int outputColumn) {
+ super(colNum, value.getValue().getBytes(), outputColumn);
+ }
+
+ public StringGroupColConcatVarCharScalar() {
+ super();
+ }
+
+ @Override
+ public VectorExpressionDescriptor.Descriptor getDescriptor() {
+ return (new VectorExpressionDescriptor.Builder())
+ .setMode(
+ VectorExpressionDescriptor.Mode.PROJECTION)
+ .setNumArguments(2)
+ .setArgumentTypes(
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.VARCHAR)
+ .setInputExpressionTypes(
+ VectorExpressionDescriptor.InputExpressionType.COLUMN,
+ VectorExpressionDescriptor.InputExpressionType.SCALAR).build();
+ }
+}
Added: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupConcatColCol.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupConcatColCol.java?rev=1622729&view=auto
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupConcatColCol.java (added)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringGroupConcatColCol.java Fri Sep 5 15:59:47 2014
@@ -0,0 +1,455 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.hive.ql.exec.vector.expressions;
+
+import org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector;
+import org.apache.hadoop.hive.ql.exec.vector.ColumnVector;
+import org.apache.hadoop.hive.ql.exec.vector.VectorExpressionDescriptor;
+import org.apache.hadoop.hive.ql.exec.vector.VectorizedRowBatch;
+
+/**
+ * Vectorized instruction to concatenate two string columns and put
+ * the output in a third column.
+ */
+public class StringGroupConcatColCol extends VectorExpression {
+ private static final long serialVersionUID = 1L;
+ private int colNum1;
+ private int colNum2;
+ private int outputColumn;
+
+ public StringGroupConcatColCol(int colNum1, int colNum2, int outputColumn) {
+ this();
+ this.colNum1 = colNum1;
+ this.colNum2 = colNum2;
+ this.outputColumn = outputColumn;
+ }
+
+ public StringGroupConcatColCol() {
+ super();
+ }
+
+ @Override
+ public void evaluate(VectorizedRowBatch batch) {
+
+ if (childExpressions != null) {
+ super.evaluateChildren(batch);
+ }
+
+ BytesColumnVector inV1 = (BytesColumnVector) batch.cols[colNum1];
+ BytesColumnVector inV2 = (BytesColumnVector) batch.cols[colNum2];
+ BytesColumnVector outV = (BytesColumnVector) batch.cols[outputColumn];
+ int[] sel = batch.selected;
+ int n = batch.size;
+ byte[][] vector1 = inV1.vector;
+ byte[][] vector2 = inV2.vector;
+ int[] len1 = inV1.length;
+ int[] len2 = inV2.length;
+ int[] start1 = inV1.start;
+ int[] start2 = inV2.start;
+
+ // return immediately if batch is empty
+ if (n == 0) {
+ return;
+ }
+
+ // prepare output buffer to accept results
+ outV.initBuffer();
+
+ /* Handle default case for isRepeating setting for output. This will be set to true
+ * later in the special cases where that is necessary.
+ */
+ outV.isRepeating = false;
+
+ if (inV1.noNulls && !inV2.noNulls) {
+
+ // propagate nulls
+
+ /* We'll assume that there *may* be nulls in the input if !noNulls is true
+ * for an input vector. This is to be more forgiving of errors in loading
+ * the vectors. A properly-written vectorized iterator will make sure that
+ * isNull[0] is set if !noNulls and isRepeating are true for the vector.
+ */
+ outV.noNulls = false;
+ if (inV2.isRepeating) {
+ if (inV2.isNull[0]) {
+
+ // Output will also be repeating and null
+ outV.isNull[0] = true;
+ outV.isRepeating = true;
+
+ //return as no further processing is needed
+ return;
+ }
+ } else {
+ propagateNulls(batch.selectedInUse, n, sel, inV2, outV);
+ }
+
+ // perform data operation
+ if (inV1.isRepeating && inV2.isRepeating) {
+
+ /* All must be selected otherwise size would be zero.
+ * Repeating property will not change.
+ */
+ if (!inV2.isNull[0]) {
+ outV.setConcat(0, vector1[0], start1[0], len1[0], vector2[0], start2[0], len2[0]);
+ }
+ outV.isRepeating = true;
+ } else if (inV1.isRepeating) {
+ if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ if (!inV2.isNull[i]) {
+ outV.setConcat(i, vector1[0], start1[0], len1[0], vector2[i], start2[i], len2[i]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV2.isNull[0]) {
+ outV.setConcat(i, vector1[0], start1[0], len1[0], vector2[i], start2[i], len2[i]);
+ }
+ }
+ }
+ } else if (inV2.isRepeating) {
+ if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ if (!inV2.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[0], start2[0], len2[0]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV2.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[0], start2[0], len2[0]);
+ }
+ }
+ }
+ } else {
+ if (batch.selectedInUse) {
+ for(int j=0; j != n; j++) {
+ int i = sel[j];
+ if (!inV2.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[i], start2[i], len2[i]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV2.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[i], start2[i], len2[i]);
+ }
+ }
+ }
+ }
+ } else if (!inV1.noNulls && inV2.noNulls) {
+
+ // propagate nulls
+ outV.noNulls = false;
+ if (inV1.isRepeating) {
+
+ //Output will also be repeating and null
+ outV.isRepeating = true;
+ outV.isNull[0] = true;
+
+ //return as no further processing is needed
+ return;
+ } else {
+ propagateNulls(batch.selectedInUse, n, sel, inV1, outV);
+ }
+
+ // perform data operation
+ if (inV1.isRepeating && inV2.isRepeating) {
+ //All must be selected otherwise size would be zero
+ //Repeating property will not change.
+ if (!inV1.isNull[0]) {
+ outV.setConcat(0, vector1[0], start1[0], len1[0], vector2[0], start2[0], len2[0]);
+ }
+ outV.isRepeating = true;
+ } else if (inV1.isRepeating) {
+ if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ if (!inV1.isNull[0]) {
+ outV.setConcat(i, vector1[0], start1[0], len1[0], vector2[i], start2[i], len2[i]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV1.isNull[0]) {
+ outV.setConcat(i, vector1[0], start1[0], len1[0], vector2[i], start2[i], len2[i]);
+ }
+ }
+ }
+ } else if (inV2.isRepeating) {
+ if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ if (!inV1.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[0], start2[0], len2[0]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV1.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[0], start2[0], len2[0]);
+ }
+ }
+ }
+ } else {
+ if (batch.selectedInUse) {
+ for(int j=0; j != n; j++) {
+ int i = sel[j];
+ if (!inV1.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[i], start2[i], len2[i]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV1.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[i], start2[i], len2[i]);
+ }
+ }
+ }
+ }
+ } else if (!inV1.noNulls && !inV2.noNulls) {
+
+ // propagate nulls
+ outV.noNulls = false;
+ if (inV1.isRepeating && inV2.isRepeating) {
+ outV.isNull[0] = inV1.isNull[0] || inV2.isNull[0];
+
+ //Output will also be repeating
+ outV.isRepeating = true;
+
+ // return if output is null because no additional work is needed
+ if (outV.isNull[0]) {
+ return;
+ }
+ } else if (inV1.isRepeating) {
+ if (inV1.isNull[0]) { // then all output will be null
+ outV.isRepeating = true;
+ outV.isNull[0] = true;
+ return;
+ } else {
+ outV.isRepeating = false;
+ propagateNulls(batch.selectedInUse, n, sel, inV2, outV);
+ }
+ } else if (inV2.isRepeating) {
+ if (inV2.isNull[0]) {
+ outV.isRepeating = true;
+ outV.isNull[0] = true;
+ return;
+ } else {
+ outV.isRepeating = false;
+ propagateNulls(batch.selectedInUse, n, sel, inV1, outV);
+ }
+ } else {
+ propagateNullsCombine(batch.selectedInUse, n, sel, inV1, inV2, outV);
+ }
+
+ // perform data operation
+ if (inV1.isRepeating && inV2.isRepeating) {
+
+ // All must be selected otherwise size would be zero. Repeating property will not change.
+ if (!inV1.isNull[0] && !inV2.isNull[0]) {
+ outV.setConcat(0, vector1[0], start1[0], len1[0], vector2[0], start2[0], len2[0]);
+ }
+ outV.isRepeating = true;
+ } else if (inV1.isRepeating) {
+ if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ if (!inV1.isNull[0] && !inV2.isNull[i]) {
+ outV.setConcat(i, vector1[0], start1[0], len1[0], vector2[i], start2[i], len2[i]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV1.isNull[0] && !inV2.isNull[i]) {
+ outV.setConcat(i, vector1[0], start1[0], len1[0], vector2[i], start2[i], len2[i]);
+ }
+ }
+ }
+ } else if (inV2.isRepeating) {
+ if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ if (!inV1.isNull[i] && !inV2.isNull[0]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[0], start2[0], len2[0]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV1.isNull[i] && !inV2.isNull[0]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[0], start2[0], len2[0]);
+ }
+ }
+ }
+ } else {
+ if (batch.selectedInUse) {
+ for(int j=0; j != n; j++) {
+ int i = sel[j];
+ if (!inV1.isNull[i] && !inV2.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[i], start2[i], len2[i]);
+ }
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inV1.isNull[i] && !inV2.isNull[i]) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[i], start2[i], len2[i]);
+ }
+ }
+ }
+ }
+ } else { // there are no nulls in either input vector
+
+ // propagate null information
+ outV.noNulls = true;
+
+ // perform data operation
+ if (inV1.isRepeating && inV2.isRepeating) {
+
+ // All must be selected otherwise size would be zero. Repeating property will not change.
+ outV.setConcat(0, vector1[0], start1[0], len1[0], vector2[0], start2[0], len2[0]);
+ outV.isRepeating = true;
+ } else if (inV1.isRepeating) {
+ if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ outV.setConcat(i, vector1[0], start1[0], len1[0], vector2[i], start2[i], len2[i]);
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ outV.setConcat(i, vector1[0], start1[0], len1[0], vector2[i], start2[i], len2[i]);
+ }
+ }
+ } else if (inV2.isRepeating) {
+ if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[0], start2[0], len2[0]);
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[0], start2[0], len2[0]);
+ }
+ }
+ } else {
+ if (batch.selectedInUse) {
+ for(int j=0; j != n; j++) {
+ int i = sel[j];
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[i], start2[i], len2[i]);
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ outV.setConcat(i, vector1[i], start1[i], len1[i], vector2[i], start2[i], len2[i]);
+ }
+ }
+ }
+ }
+ }
+
+ /**
+ * Propagate the logic OR of null vectors from two inputs to output.
+ *
+ * @param selectedInUse true/false flag to tell if sel[] is in use
+ * @param n number of qualifying rows
+ * @param sel selected value position array
+ * @param inV1 input vector 1
+ * @param inV2 input vector 2
+ * @param outV output vector
+ */
+ private static void propagateNullsCombine(boolean selectedInUse, int n, int[] sel,
+ ColumnVector inV1, ColumnVector inV2, BytesColumnVector outV) {
+ if (selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ outV.isNull[i] = inV1.isNull[i] || inV2.isNull[i];
+ }
+ } else {
+ for(int i = 0; i != n; i++) {
+ outV.isNull[i] = inV1.isNull[i] || inV2.isNull[i];
+ }
+ }
+ }
+
+ /**
+ * Propagate nulls from input vector inV to output vector outV.
+ *
+ * @param selectedInUse true/false flag to tell if sel[] is in use
+ * @param sel selected value position array
+ * @param n number of qualifying rows
+ * @param inV input vector
+ * @param outV ouput vector
+ */
+ private static void propagateNulls(boolean selectedInUse, int n, int[] sel, ColumnVector inV,
+ ColumnVector outV) {
+ if (selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ outV.isNull[i] = inV.isNull[i];
+ }
+ } else {
+ System.arraycopy(inV.isNull, 0, outV.isNull, 0, n);
+ }
+ }
+
+ @Override
+ public int getOutputColumn() {
+ return outputColumn;
+ }
+
+ @Override
+ public String getOutputType() {
+ return "StringGroup";
+ }
+
+ public int getColNum1() {
+ return colNum1;
+ }
+
+ public void setColNum1(int colNum1) {
+ this.colNum1 = colNum1;
+ }
+
+ public int getColNum2() {
+ return colNum2;
+ }
+
+ public void setColNum2(int colNum2) {
+ this.colNum2 = colNum2;
+ }
+
+ public void setOutputColumn(int outputColumn) {
+ this.outputColumn = outputColumn;
+ }
+
+ @Override
+ public VectorExpressionDescriptor.Descriptor getDescriptor() {
+ return (new VectorExpressionDescriptor.Builder())
+ .setMode(
+ VectorExpressionDescriptor.Mode.PROJECTION)
+ .setNumArguments(2)
+ .setArgumentTypes(
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY)
+ .setInputExpressionTypes(
+ VectorExpressionDescriptor.InputExpressionType.COLUMN,
+ VectorExpressionDescriptor.InputExpressionType.COLUMN).build();
+ }
+}
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringLength.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringLength.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringLength.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringLength.java Fri Sep 5 15:59:47 2014
@@ -162,7 +162,7 @@ public class StringLength extends Vector
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.STRING)
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Added: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringScalarConcatStringGroupCol.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringScalarConcatStringGroupCol.java?rev=1622729&view=auto
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringScalarConcatStringGroupCol.java (added)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringScalarConcatStringGroupCol.java Fri Sep 5 15:59:47 2014
@@ -0,0 +1,164 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.hive.ql.exec.vector.expressions;
+
+import org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector;
+import org.apache.hadoop.hive.ql.exec.vector.VectorExpressionDescriptor;
+import org.apache.hadoop.hive.ql.exec.vector.VectorizedRowBatch;
+
+/**
+ * Vectorized instruction to concatenate a scalar to a string column and put
+ * the result in an output column.
+ */
+public class StringScalarConcatStringGroupCol extends VectorExpression {
+ private static final long serialVersionUID = 1L;
+ private int colNum;
+ private int outputColumn;
+ private byte[] value;
+
+ public StringScalarConcatStringGroupCol(byte[] value, int colNum, int outputColumn) {
+ this();
+ this.colNum = colNum;
+ this.outputColumn = outputColumn;
+ this.value = value;
+ }
+
+ public StringScalarConcatStringGroupCol() {
+ super();
+ }
+
+ @Override
+ public void evaluate(VectorizedRowBatch batch) {
+
+ if (childExpressions != null) {
+ super.evaluateChildren(batch);
+ }
+
+ BytesColumnVector inputColVector = (BytesColumnVector) batch.cols[colNum];
+ BytesColumnVector outV = (BytesColumnVector) batch.cols[outputColumn];
+ int[] sel = batch.selected;
+ int n = batch.size;
+ byte[][] vector = inputColVector.vector;
+ int[] start = inputColVector.start;
+ int[] length = inputColVector.length;
+
+ if (n == 0) {
+
+ // Nothing to do
+ return;
+ }
+
+ // initialize output vector buffer to receive data
+ outV.initBuffer();
+
+ if (inputColVector.noNulls) {
+ outV.noNulls = true;
+ if (inputColVector.isRepeating) {
+ outV.isRepeating = true;
+ outV.setConcat(0, value, 0, value.length, vector[0], start[0], length[0]);
+ } else if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ outV.setConcat(i, value, 0, value.length, vector[i], start[i], length[i]);
+ }
+ outV.isRepeating = false;
+ } else {
+ for(int i = 0; i != n; i++) {
+ outV.setConcat(i, value, 0, value.length, vector[i], start[i], length[i]);
+ }
+ outV.isRepeating = false;
+ }
+ } else {
+
+ /*
+ * Handle case with nulls. Don't do function if the value is null, to save time,
+ * because calling the function can be expensive.
+ */
+ outV.noNulls = false;
+ if (inputColVector.isRepeating) {
+ outV.isRepeating = true;
+ outV.isNull[0] = inputColVector.isNull[0];
+ if (!inputColVector.isNull[0]) {
+ outV.setConcat(0, value, 0, value.length, vector[0], start[0], length[0]);
+ }
+ } else if (batch.selectedInUse) {
+ for(int j = 0; j != n; j++) {
+ int i = sel[j];
+ if (!inputColVector.isNull[i]) {
+ outV.setConcat(i, value, 0, value.length, vector[i], start[i], length[i]);
+ }
+ outV.isNull[i] = inputColVector.isNull[i];
+ }
+ outV.isRepeating = false;
+ } else {
+ for(int i = 0; i != n; i++) {
+ if (!inputColVector.isNull[i]) {
+ outV.setConcat(i, value, 0, value.length, vector[i], start[i], length[i]);
+ }
+ outV.isNull[i] = inputColVector.isNull[i];
+ }
+ outV.isRepeating = false;
+ }
+ }
+ }
+
+ @Override
+ public int getOutputColumn() {
+ return outputColumn;
+ }
+
+ @Override
+ public String getOutputType() {
+ return "StringGroup";
+ }
+
+ public int getColNum() {
+ return colNum;
+ }
+
+ public void setColNum(int colNum) {
+ this.colNum = colNum;
+ }
+
+ public byte[] getValue() {
+ return value;
+ }
+
+ public void setValue(byte[] value) {
+ this.value = value;
+ }
+
+ public void setOutputColumn(int outputColumn) {
+ this.outputColumn = outputColumn;
+ }
+
+ @Override
+ public VectorExpressionDescriptor.Descriptor getDescriptor() {
+ return (new VectorExpressionDescriptor.Builder())
+ .setMode(
+ VectorExpressionDescriptor.Mode.PROJECTION)
+ .setNumArguments(2)
+ .setArgumentTypes(
+ VectorExpressionDescriptor.ArgumentType.STRING,
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY)
+ .setInputExpressionTypes(
+ VectorExpressionDescriptor.InputExpressionType.SCALAR,
+ VectorExpressionDescriptor.InputExpressionType.COLUMN).build();
+ }
+}
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringSubstrColStart.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringSubstrColStart.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringSubstrColStart.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringSubstrColStart.java Fri Sep 5 15:59:47 2014
@@ -250,8 +250,8 @@ public class StringSubstrColStart extend
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.STRING,
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.SCALAR);
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringSubstrColStartLen.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringSubstrColStartLen.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringSubstrColStartLen.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringSubstrColStartLen.java Fri Sep 5 15:59:47 2014
@@ -277,9 +277,9 @@ public class StringSubstrColStartLen ext
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(3)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.STRING,
- VectorExpressionDescriptor.ArgumentType.LONG,
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.SCALAR,
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringUnaryUDF.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringUnaryUDF.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringUnaryUDF.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringUnaryUDF.java Fri Sep 5 15:59:47 2014
@@ -200,7 +200,7 @@ public class StringUnaryUDF extends Vect
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.STRING)
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringUnaryUDFDirect.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringUnaryUDFDirect.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringUnaryUDFDirect.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/StringUnaryUDFDirect.java Fri Sep 5 15:59:47 2014
@@ -142,7 +142,7 @@ abstract public class StringUnaryUDFDire
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.STRING)
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Added: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/TruncStringOutput.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/TruncStringOutput.java?rev=1622729&view=auto
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/TruncStringOutput.java (added)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/TruncStringOutput.java Fri Sep 5 15:59:47 2014
@@ -0,0 +1,25 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.hive.ql.exec.vector.expressions;
+
+public interface TruncStringOutput {
+ abstract public int getMaxLength();
+
+ abstract public void setMaxLength(int maxLength);
+}
\ No newline at end of file
Added: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VarCharScalarConcatStringGroupCol.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VarCharScalarConcatStringGroupCol.java?rev=1622729&view=auto
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VarCharScalarConcatStringGroupCol.java (added)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VarCharScalarConcatStringGroupCol.java Fri Sep 5 15:59:47 2014
@@ -0,0 +1,52 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.hadoop.hive.ql.exec.vector.expressions;
+
+import org.apache.hadoop.hive.common.type.HiveVarchar;
+import org.apache.hadoop.hive.ql.exec.vector.VectorExpressionDescriptor;
+
+/**
+ * Vectorized instruction to concatenate a scalar to a string column and put
+ * the result in an output column.
+ */
+public class VarCharScalarConcatStringGroupCol extends StringScalarConcatStringGroupCol {
+ private static final long serialVersionUID = 1L;
+
+ public VarCharScalarConcatStringGroupCol(HiveVarchar value, int colNum, int outputColumn) {
+ super(value.getValue().getBytes(), colNum, outputColumn);
+ }
+
+ public VarCharScalarConcatStringGroupCol() {
+ super();
+ }
+
+ @Override
+ public VectorExpressionDescriptor.Descriptor getDescriptor() {
+ return (new VectorExpressionDescriptor.Builder())
+ .setMode(
+ VectorExpressionDescriptor.Mode.PROJECTION)
+ .setNumArguments(2)
+ .setArgumentTypes(
+ VectorExpressionDescriptor.ArgumentType.VARCHAR,
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY)
+ .setInputExpressionTypes(
+ VectorExpressionDescriptor.InputExpressionType.SCALAR,
+ VectorExpressionDescriptor.InputExpressionType.COLUMN).build();
+ }
+}
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorExpression.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorExpression.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorExpression.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorExpression.java Fri Sep 5 15:59:47 2014
@@ -30,16 +30,22 @@ import org.apache.hadoop.hive.ql.exec.ve
*/
public abstract class VectorExpression implements Serializable {
public enum Type {
- STRING, TIMESTAMP, DATE, OTHER;
+ STRING, CHAR, VARCHAR, TIMESTAMP, DATE, LONG, DOUBLE, DECIMAL, OTHER;
private static Map<String, Type> types = ImmutableMap.<String, Type>builder()
.put("string", STRING)
+ .put("char", CHAR)
+ .put("varchar", VARCHAR)
.put("timestamp", TIMESTAMP)
.put("date", DATE)
+ .put("long", LONG)
+ .put("double", DOUBLE)
+ .put("decimal", DECIMAL)
.build();
public static Type getValue(String name) {
- if (types.containsKey(name.toLowerCase())) {
- return types.get(name);
+ String nameLower = name.toLowerCase();
+ if (types.containsKey(nameLower)) {
+ return types.get(nameLower);
}
return OTHER;
}
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorExpressionWriterFactory.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorExpressionWriterFactory.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorExpressionWriterFactory.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorExpressionWriterFactory.java Fri Sep 5 15:59:47 2014
@@ -28,6 +28,7 @@ import java.util.List;
import org.apache.commons.lang.ArrayUtils;
import org.apache.commons.lang.StringUtils;
import org.apache.hadoop.hive.common.type.Decimal128;
+import org.apache.hadoop.hive.common.type.HiveChar;
import org.apache.hadoop.hive.common.type.HiveDecimal;
import org.apache.hadoop.hive.common.type.HiveVarchar;
import org.apache.hadoop.hive.ql.exec.vector.*;
@@ -46,6 +47,7 @@ import org.apache.hadoop.hive.serde2.obj
import org.apache.hadoop.hive.serde2.objectinspector.primitive.SettableDateObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.SettableDoubleObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.SettableFloatObjectInspector;
+import org.apache.hadoop.hive.serde2.objectinspector.primitive.SettableHiveCharObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.SettableHiveDecimalObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.SettableHiveVarcharObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.SettableIntObjectInspector;
@@ -403,6 +405,9 @@ public final class VectorExpressionWrite
case STRING:
return genVectorExpressionWritableString(
(SettableStringObjectInspector) fieldObjInspector);
+ case CHAR:
+ return genVectorExpressionWritableChar(
+ (SettableHiveCharObjectInspector) fieldObjInspector);
case VARCHAR:
return genVectorExpressionWritableVarchar(
(SettableHiveVarcharObjectInspector) fieldObjInspector);
@@ -558,6 +563,46 @@ public final class VectorExpressionWrite
}.init(fieldObjInspector);
}
+ private static VectorExpressionWriter genVectorExpressionWritableChar(
+ SettableHiveCharObjectInspector fieldObjInspector) throws HiveException {
+ return new VectorExpressionWriterBytes() {
+ private Object obj;
+ private Text text;
+
+ public VectorExpressionWriter init(SettableHiveCharObjectInspector objInspector)
+ throws HiveException {
+ super.init(objInspector);
+ this.text = new Text();
+ this.obj = initValue(null);
+ return this;
+ }
+
+ @Override
+ public Object writeValue(byte[] value, int start, int length) throws HiveException {
+ text.set(value, start, length);
+ ((SettableHiveCharObjectInspector) this.objectInspector).set(this.obj, text.toString());
+ return this.obj;
+ }
+
+ @Override
+ public Object setValue(Object field, byte[] value, int start, int length)
+ throws HiveException {
+ if (null == field) {
+ field = initValue(null);
+ }
+ text.set(value, start, length);
+ ((SettableHiveCharObjectInspector) this.objectInspector).set(field, text.toString());
+ return field;
+ }
+
+ @Override
+ public Object initValue(Object ignored) {
+ return ((SettableHiveCharObjectInspector) this.objectInspector)
+ .create(new HiveChar(StringUtils.EMPTY, -1));
+ }
+ }.init(fieldObjInspector);
+ }
+
private static VectorExpressionWriter genVectorExpressionWritableVarchar(
SettableHiveVarcharObjectInspector fieldObjInspector) throws HiveException {
return new VectorExpressionWriterBytes() {
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddColCol.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddColCol.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddColCol.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddColCol.java Fri Sep 5 15:59:47 2014
@@ -127,6 +127,8 @@ public class VectorUDFDateAddColCol exte
break;
case STRING:
+ case CHAR:
+ case VARCHAR:
// Now disregard null in second pass.
if ((inputColVector1.isRepeating) && (inputColVector2.isRepeating)) {
// All must be selected otherwise size would be zero
@@ -144,6 +146,8 @@ public class VectorUDFDateAddColCol exte
}
}
break;
+ default:
+ throw new Error("Unsupported input type " + inputTypes[0].name());
}
}
@@ -235,8 +239,8 @@ public class VectorUDFDateAddColCol exte
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.ANY,
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.COLUMN);
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddColScalar.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddColScalar.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddColScalar.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddColScalar.java Fri Sep 5 15:59:47 2014
@@ -165,6 +165,8 @@ public class VectorUDFDateAddColScalar e
break;
case STRING:
+ case CHAR:
+ case VARCHAR:
if (inputCol.noNulls) {
outV.noNulls = true;
if (batch.selectedInUse) {
@@ -199,6 +201,8 @@ public class VectorUDFDateAddColScalar e
}
}
break;
+ default:
+ throw new Error("Unsupported input type " + inputTypes[0].name());
}
}
@@ -286,8 +290,8 @@ public class VectorUDFDateAddColScalar e
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.ANY,
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.SCALAR);
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddScalarCol.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddScalarCol.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddScalarCol.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateAddScalarCol.java Fri Sep 5 15:59:47 2014
@@ -84,6 +84,8 @@ public class VectorUDFDateAddScalarCol e
break;
case STRING:
+ case CHAR:
+ case VARCHAR:
try {
baseDate = formatter.parse(new String(stringValue, "UTF-8"));
break;
@@ -101,6 +103,8 @@ public class VectorUDFDateAddScalarCol e
}
return;
}
+ default:
+ throw new Error("Unsupported input type " + inputTypes[0].name());
}
if(batch.size == 0) {
@@ -213,8 +217,8 @@ public class VectorUDFDateAddScalarCol e
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.ANY,
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.INT_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.SCALAR,
VectorExpressionDescriptor.InputExpressionType.COLUMN);
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffColCol.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffColCol.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffColCol.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffColCol.java Fri Sep 5 15:59:47 2014
@@ -191,12 +191,14 @@ public class VectorUDFDateDiffColCol ext
return dateVector;
case STRING:
+ case CHAR:
+ case VARCHAR:
BytesColumnVector bcv = (BytesColumnVector) inputColVector;
copySelected(bcv, batch.selectedInUse, batch.selected, batch.size, dateVector);
return dateVector;
+ default:
+ throw new Error("Unsupported input type " + colType.name());
}
-
- return null;
}
// Copy the current object contents into the output. Only copy selected entries,
@@ -314,8 +316,8 @@ public class VectorUDFDateDiffColCol ext
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.ANY,
- VectorExpressionDescriptor.ArgumentType.ANY)
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.COLUMN);
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffColScalar.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffColScalar.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffColScalar.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffColScalar.java Fri Sep 5 15:59:47 2014
@@ -18,6 +18,7 @@
package org.apache.hadoop.hive.ql.exec.vector.expressions;
+import org.apache.hadoop.hive.metastore.parser.ExpressionTree.Operator;
import org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector;
import org.apache.hadoop.hive.ql.exec.vector.ColumnVector;
import org.apache.hadoop.hive.ql.exec.vector.LongColumnVector;
@@ -90,6 +91,8 @@ public class VectorUDFDateDiffColScalar
break;
case STRING:
+ case CHAR:
+ case VARCHAR:
try {
date.setTime(formatter.parse(new String(stringValue, "UTF-8")).getTime());
baseDate = DateWritable.dateToDays(date);
@@ -108,6 +111,8 @@ public class VectorUDFDateDiffColScalar
}
return;
}
+ default:
+ throw new Error("Invalid input type #1: " + inputTypes[1].name());
}
switch (inputTypes[0]) {
@@ -184,6 +189,8 @@ public class VectorUDFDateDiffColScalar
break;
case STRING:
+ case CHAR:
+ case VARCHAR:
if (inputCol.noNulls) {
outV.noNulls = true;
if (batch.selectedInUse) {
@@ -218,6 +225,8 @@ public class VectorUDFDateDiffColScalar
}
}
break;
+ default:
+ throw new Error("Invalid input type #0: " + inputTypes[0].name());
}
}
@@ -287,8 +296,8 @@ public class VectorUDFDateDiffColScalar
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.ANY,
- VectorExpressionDescriptor.ArgumentType.ANY)
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN,
VectorExpressionDescriptor.InputExpressionType.SCALAR);
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffScalarCol.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffScalarCol.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffScalarCol.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateDiffScalarCol.java Fri Sep 5 15:59:47 2014
@@ -90,6 +90,8 @@ public class VectorUDFDateDiffScalarCol
break;
case STRING:
+ case CHAR:
+ case VARCHAR:
try {
date.setTime(formatter.parse(new String(stringValue, "UTF-8")).getTime());
baseDate = DateWritable.dateToDays(date);
@@ -108,6 +110,8 @@ public class VectorUDFDateDiffScalarCol
}
return;
}
+ default:
+ throw new Error("Unsupported input type " + inputTypes[0].name());
}
switch (inputTypes[1]) {
@@ -184,6 +188,8 @@ public class VectorUDFDateDiffScalarCol
break;
case STRING:
+ case CHAR:
+ case VARCHAR:
if (inputCol.noNulls) {
outV.noNulls = true;
if (batch.selectedInUse) {
@@ -218,6 +224,8 @@ public class VectorUDFDateDiffScalarCol
}
}
break;
+ default:
+ throw new Error("Unsupported input type " + inputTypes[1].name());
}
}
@@ -287,8 +295,8 @@ public class VectorUDFDateDiffScalarCol
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(2)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.ANY,
- VectorExpressionDescriptor.ArgumentType.ANY)
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY,
+ VectorExpressionDescriptor.ArgumentType.STRING_DATETIME_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.SCALAR,
VectorExpressionDescriptor.InputExpressionType.COLUMN);
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateLong.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateLong.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateLong.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFDateLong.java Fri Sep 5 15:59:47 2014
@@ -19,6 +19,7 @@
package org.apache.hadoop.hive.ql.exec.vector.expressions;
import org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector;
+import org.apache.hadoop.hive.ql.exec.vector.VectorExpressionDescriptor;
import org.apache.hadoop.hive.serde2.io.DateWritable;
import java.io.UnsupportedEncodingException;
@@ -49,6 +50,8 @@ public class VectorUDFDateLong extends L
case TIMESTAMP:
date.setTime(vector[i] / 1000000);
break;
+ default:
+ throw new Error("Unsupported input type " + inputTypes[0].name());
}
try {
byte[] bytes = formatter.format(date).getBytes("UTF-8");
@@ -58,4 +61,16 @@ public class VectorUDFDateLong extends L
outV.isNull[i] = true;
}
}
+
+ @Override
+ public VectorExpressionDescriptor.Descriptor getDescriptor() {
+ VectorExpressionDescriptor.Builder b = new VectorExpressionDescriptor.Builder();
+ b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
+ .setNumArguments(1)
+ .setArgumentTypes(
+ VectorExpressionDescriptor.ArgumentType.DATETIME_FAMILY)
+ .setInputExpressionTypes(
+ VectorExpressionDescriptor.InputExpressionType.COLUMN);
+ return b.build();
+ }
}
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFTimestampFieldLong.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFTimestampFieldLong.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFTimestampFieldLong.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFTimestampFieldLong.java Fri Sep 5 15:59:47 2014
@@ -179,6 +179,8 @@ public abstract class VectorUDFTimestamp
}
}
break;
+ default:
+ throw new Error("Unsupported input type " + inputTypes[0].name());
}
}
@@ -218,7 +220,7 @@ public abstract class VectorUDFTimestamp
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.LONG)
+ VectorExpressionDescriptor.ArgumentType.DATETIME_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFTimestampFieldString.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFTimestampFieldString.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFTimestampFieldString.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/expressions/VectorUDFTimestampFieldString.java Fri Sep 5 15:59:47 2014
@@ -181,7 +181,7 @@ public abstract class VectorUDFTimestamp
b.setMode(VectorExpressionDescriptor.Mode.PROJECTION)
.setNumArguments(1)
.setArgumentTypes(
- VectorExpressionDescriptor.ArgumentType.STRING)
+ VectorExpressionDescriptor.ArgumentType.STRING_FAMILY)
.setInputExpressionTypes(
VectorExpressionDescriptor.InputExpressionType.COLUMN);
return b.build();
Modified: hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/udf/VectorUDFAdaptor.java
URL: http://svn.apache.org/viewvc/hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/udf/VectorUDFAdaptor.java?rev=1622729&r1=1622728&r2=1622729&view=diff
==============================================================================
--- hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/udf/VectorUDFAdaptor.java (original)
+++ hive/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/vector/udf/VectorUDFAdaptor.java Fri Sep 5 15:59:47 2014
@@ -23,6 +23,7 @@ import java.sql.Timestamp;
import org.apache.hadoop.hive.common.type.HiveDecimal;
import org.apache.hadoop.hive.ql.exec.UDFArgumentException;
import org.apache.hadoop.hive.ql.exec.vector.*;
+import org.apache.hadoop.hive.ql.exec.vector.expressions.StringExpr;
import org.apache.hadoop.hive.ql.exec.vector.expressions.VectorExpression;
import org.apache.hadoop.hive.ql.exec.vector.expressions.VectorExpressionWriter;
import org.apache.hadoop.hive.ql.exec.vector.expressions.VectorExpressionWriterFactory;
@@ -30,8 +31,12 @@ import org.apache.hadoop.hive.ql.metadat
import org.apache.hadoop.hive.ql.plan.ExprNodeGenericFuncDesc;
import org.apache.hadoop.hive.ql.udf.generic.GenericUDF;
import org.apache.hadoop.hive.serde2.io.DateWritable;
+import org.apache.hadoop.hive.serde2.io.HiveCharWritable;
+import org.apache.hadoop.hive.serde2.io.HiveVarcharWritable;
import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.*;
+import org.apache.hadoop.hive.serde2.typeinfo.CharTypeInfo;
+import org.apache.hadoop.hive.serde2.typeinfo.VarcharTypeInfo;
import org.apache.hadoop.io.Text;
/**
@@ -215,6 +220,35 @@ public class VectorUDFAdaptor extends Ve
t = ((WritableStringObjectInspector) outputOI).getPrimitiveWritableObject(value);
}
bv.setVal(i, t.getBytes(), 0, t.getLength());
+ } else if (outputOI instanceof WritableHiveCharObjectInspector) {
+ WritableHiveCharObjectInspector writableHiveCharObjectOI = (WritableHiveCharObjectInspector) outputOI;
+ int maxLength = ((CharTypeInfo) writableHiveCharObjectOI.getTypeInfo()).getLength();
+ BytesColumnVector bv = (BytesColumnVector) colVec;
+
+ HiveCharWritable hiveCharWritable;
+ if (value instanceof HiveCharWritable) {
+ hiveCharWritable = ((HiveCharWritable) value);
+ } else {
+ hiveCharWritable = writableHiveCharObjectOI.getPrimitiveWritableObject(value);
+ }
+ Text t = hiveCharWritable.getTextValue();
+
+ // In vector mode, we stored CHAR as unpadded.
+ StringExpr.rightTrimAndTruncate(bv, i, t.getBytes(), 0, t.getLength(), maxLength);
+ } else if (outputOI instanceof WritableHiveVarcharObjectInspector) {
+ WritableHiveVarcharObjectInspector writableHiveVarcharObjectOI = (WritableHiveVarcharObjectInspector) outputOI;
+ int maxLength = ((VarcharTypeInfo) writableHiveVarcharObjectOI.getTypeInfo()).getLength();
+ BytesColumnVector bv = (BytesColumnVector) colVec;
+
+ HiveVarcharWritable hiveVarcharWritable;
+ if (value instanceof HiveVarcharWritable) {
+ hiveVarcharWritable = ((HiveVarcharWritable) value);
+ } else {
+ hiveVarcharWritable = writableHiveVarcharObjectOI.getPrimitiveWritableObject(value);
+ }
+ Text t = hiveVarcharWritable.getTextValue();
+
+ StringExpr.truncate(bv, i, t.getBytes(), 0, t.getLength(), maxLength);
} else if (outputOI instanceof WritableIntObjectInspector) {
LongColumnVector lv = (LongColumnVector) colVec;
if (value instanceof Integer) {