use of org.apache.hadoop.hive.ql.exec.vector.ColumnVector in project hive by apache.
the class VectorUDFDateAddColCol method evaluate.
@Override
public void evaluate(VectorizedRowBatch batch) {
if (childExpressions != null) {
super.evaluateChildren(batch);
}
ColumnVector inputColVector1 = batch.cols[colNum1];
LongColumnVector inputColVector2 = (LongColumnVector) batch.cols[colNum2];
int[] sel = batch.selected;
int n = batch.size;
long[] vector2 = inputColVector2.vector;
LongColumnVector outV = (LongColumnVector) batch.cols[outputColumnNum];
long[] outputVector = outV.vector;
if (n <= 0) {
// Nothing to do
return;
}
/*
* Propagate null values for a two-input operator and set isRepeating and noNulls appropriately.
*/
NullUtil.propagateNullsColCol(inputColVector1, inputColVector2, outV, batch.selected, batch.size, batch.selectedInUse);
switch(primitiveCategory) {
case DATE:
// Now disregard null in second pass.
if ((inputColVector1.isRepeating) && (inputColVector2.isRepeating)) {
// All must be selected otherwise size would be zero
// Repeating property will not change.
outV.isRepeating = true;
outputVector[0] = evaluateDate(inputColVector1, 0, vector2[0]);
} else if (batch.selectedInUse) {
for (int j = 0; j != n; j++) {
int i = sel[j];
outputVector[i] = evaluateDate(inputColVector1, i, vector2[i]);
}
} else {
for (int i = 0; i != n; i++) {
outputVector[i] = evaluateDate(inputColVector1, i, vector2[i]);
}
}
break;
case TIMESTAMP:
// Now disregard null in second pass.
if ((inputColVector1.isRepeating) && (inputColVector2.isRepeating)) {
// All must be selected otherwise size would be zero
// Repeating property will not change.
outV.isRepeating = true;
outputVector[0] = evaluateTimestamp(inputColVector1, 0, vector2[0]);
} else if (batch.selectedInUse) {
for (int j = 0; j != n; j++) {
int i = sel[j];
outputVector[i] = evaluateTimestamp(inputColVector1, i, vector2[i]);
}
} else {
for (int i = 0; i != n; i++) {
outputVector[i] = evaluateTimestamp(inputColVector1, i, vector2[i]);
}
}
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
// Repeating property will not change.
outV.isRepeating = true;
evaluateString((BytesColumnVector) inputColVector1, outV, 0, vector2[0]);
} else if (batch.selectedInUse) {
for (int j = 0; j != n; j++) {
int i = sel[j];
evaluateString((BytesColumnVector) inputColVector1, outV, i, vector2[i]);
}
} else {
for (int i = 0; i != n; i++) {
evaluateString((BytesColumnVector) inputColVector1, outV, i, vector2[i]);
}
}
break;
default:
throw new Error("Unsupported input type " + primitiveCategory.name());
}
}
use of org.apache.hadoop.hive.ql.exec.vector.ColumnVector in project hive by apache.
the class VectorUDFDateAddColScalar method evaluate.
@Override
public void evaluate(VectorizedRowBatch batch) {
if (childExpressions != null) {
super.evaluateChildren(batch);
}
LongColumnVector outputColVector = (LongColumnVector) batch.cols[outputColumnNum];
ColumnVector inputCol = batch.cols[this.colNum];
/* every line below this is identical for evaluateLong & evaluateString */
final int n = inputCol.isRepeating ? 1 : batch.size;
int[] sel = batch.selected;
final boolean selectedInUse = (inputCol.isRepeating == false) && batch.selectedInUse;
boolean[] outputIsNull = outputColVector.isNull;
if (batch.size == 0) {
/* n != batch.size when isRepeating */
return;
}
// We do not need to do a column reset since we are carefully changing the output.
outputColVector.isRepeating = false;
switch(primitiveCategory) {
case DATE:
if (inputCol.isRepeating) {
if (inputCol.noNulls || !inputCol.isNull[0]) {
outputColVector.isNull[0] = false;
outputColVector.vector[0] = evaluateDate(inputCol, 0);
} else {
outputColVector.isNull[0] = true;
outputColVector.noNulls = false;
}
outputColVector.isRepeating = true;
} else if (inputCol.noNulls) {
if (batch.selectedInUse) {
if (!outputColVector.noNulls) {
for (int j = 0; j != n; j++) {
final int i = sel[j];
// Set isNull before call in case it changes it mind.
outputIsNull[i] = false;
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
} else {
for (int j = 0; j != n; j++) {
final int i = sel[j];
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
}
} else {
if (!outputColVector.noNulls) {
// Assume it is almost always a performance win to fill all of isNull so we can
// safely reset noNulls.
Arrays.fill(outputIsNull, false);
outputColVector.noNulls = true;
}
for (int i = 0; i != n; i++) {
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
}
} else /* there are nulls in the inputColVector */
{
// Carefully handle NULLs..
// Handle case with nulls. Don't do function if the value is null, to save time,
// because calling the function can be expensive.
outputColVector.noNulls = false;
if (selectedInUse) {
for (int j = 0; j < n; j++) {
int i = sel[j];
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
}
} else {
for (int i = 0; i < n; i++) {
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
}
}
}
break;
case TIMESTAMP:
if (inputCol.isRepeating) {
if (inputCol.noNulls || !inputCol.isNull[0]) {
outputColVector.isNull[0] = false;
outputColVector.vector[0] = evaluateTimestamp(inputCol, 0);
} else {
outputColVector.isNull[0] = true;
outputColVector.noNulls = false;
}
outputColVector.isRepeating = true;
} else if (inputCol.noNulls) {
if (batch.selectedInUse) {
if (!outputColVector.noNulls) {
for (int j = 0; j != n; j++) {
final int i = sel[j];
// Set isNull before call in case it changes it mind.
outputIsNull[i] = false;
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
} else {
for (int j = 0; j != n; j++) {
final int i = sel[j];
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
}
} else {
if (!outputColVector.noNulls) {
// Assume it is almost always a performance win to fill all of isNull so we can
// safely reset noNulls.
Arrays.fill(outputIsNull, false);
outputColVector.noNulls = true;
}
for (int i = 0; i != n; i++) {
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
}
} else /* there are nulls in the inputColVector */
{
// Carefully handle NULLs..
// Handle case with nulls. Don't do function if the value is null, to save time,
// because calling the function can be expensive.
outputColVector.noNulls = false;
if (batch.selectedInUse) {
for (int j = 0; j < n; j++) {
int i = sel[j];
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
}
} else {
for (int i = 0; i < n; i++) {
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
}
}
}
break;
case STRING:
case CHAR:
case VARCHAR:
if (inputCol.isRepeating) {
if (inputCol.noNulls || !inputCol.isNull[0]) {
outputColVector.isNull[0] = false;
evaluateString(inputCol, outputColVector, 0);
} else {
outputColVector.isNull[0] = true;
outputColVector.noNulls = false;
}
outputColVector.isRepeating = true;
} else if (inputCol.noNulls) {
if (batch.selectedInUse) {
if (!outputColVector.noNulls) {
for (int j = 0; j != n; j++) {
final int i = sel[j];
// Set isNull before call in case it changes it mind.
outputIsNull[i] = false;
evaluateString(inputCol, outputColVector, i);
}
} else {
for (int j = 0; j != n; j++) {
final int i = sel[j];
evaluateString(inputCol, outputColVector, i);
}
}
} else {
if (!outputColVector.noNulls) {
// Assume it is almost always a performance win to fill all of isNull so we can
// safely reset noNulls.
Arrays.fill(outputIsNull, false);
outputColVector.noNulls = true;
}
for (int i = 0; i != n; i++) {
evaluateString(inputCol, outputColVector, i);
}
}
} else /* there are nulls in the inputColVector */
{
// Carefully handle NULLs..
// Handle case with nulls. Don't do function if the value is null, to save time,
// because calling the function can be expensive.
outputColVector.noNulls = false;
if (batch.selectedInUse) {
for (int j = 0; j < n; j++) {
int i = sel[j];
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
evaluateString(inputCol, outputColVector, i);
}
}
} else {
for (int i = 0; i < n; i++) {
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
evaluateString(inputCol, outputColVector, i);
}
}
}
}
break;
default:
throw new Error("Unsupported input type " + primitiveCategory.name());
}
}
use of org.apache.hadoop.hive.ql.exec.vector.ColumnVector in project hive by apache.
the class VectorUDFDateDiffColCol method evaluate.
@Override
public void evaluate(VectorizedRowBatch batch) {
if (childExpressions != null) {
super.evaluateChildren(batch);
}
ColumnVector inputColVector1 = batch.cols[colNum1];
ColumnVector inputColVector2 = batch.cols[colNum2];
int[] sel = batch.selected;
int n = batch.size;
LongColumnVector outV = (LongColumnVector) batch.cols[outputColumnNum];
long[] outputVector = outV.vector;
if (n <= 0) {
// Nothing to do
return;
}
/*
* Propagate null values for a two-input operator and set isRepeating and noNulls appropriately.
*/
NullUtil.propagateNullsColCol(inputColVector1, inputColVector2, outV, batch.selected, batch.size, batch.selectedInUse);
LongColumnVector convertedVector1 = toDateArray(batch, inputTypeInfos[0], inputColVector1, dateVector1);
LongColumnVector convertedVector2 = toDateArray(batch, inputTypeInfos[1], inputColVector2, dateVector2);
// Now disregard null in second pass.
if ((inputColVector1.isRepeating) && (inputColVector2.isRepeating)) {
// All must be selected otherwise size would be zero
// Repeating property will not change.
outV.isRepeating = true;
if (convertedVector1.isNull[0] || convertedVector2.isNull[0]) {
outV.isNull[0] = true;
} else {
outputVector[0] = convertedVector1.vector[0] - convertedVector2.vector[0];
}
} else if (inputColVector1.isRepeating) {
if (batch.selectedInUse) {
for (int j = 0; j != n; j++) {
int i = sel[j];
if (convertedVector1.isNull[0] || convertedVector2.isNull[i]) {
outV.isNull[i] = true;
} else {
outputVector[i] = convertedVector1.vector[0] - convertedVector2.vector[i];
}
}
} else {
for (int i = 0; i != n; i++) {
if (convertedVector1.isNull[0] || convertedVector2.isNull[i]) {
outV.isNull[i] = true;
} else {
outputVector[i] = convertedVector1.vector[0] - convertedVector2.vector[i];
}
}
}
} else if (inputColVector2.isRepeating) {
if (batch.selectedInUse) {
for (int j = 0; j != n; j++) {
int i = sel[j];
if (convertedVector1.isNull[i] || convertedVector2.isNull[0]) {
outV.isNull[i] = true;
} else {
outputVector[i] = convertedVector1.vector[i] - convertedVector2.vector[0];
}
}
} else {
for (int i = 0; i != n; i++) {
if (convertedVector1.isNull[i] || convertedVector2.isNull[0]) {
outV.isNull[i] = true;
} else {
outputVector[i] = convertedVector1.vector[i] - convertedVector2.vector[0];
}
}
}
} else {
if (batch.selectedInUse) {
for (int j = 0; j != n; j++) {
int i = sel[j];
if (convertedVector1.isNull[i] || convertedVector2.isNull[i]) {
outV.isNull[i] = true;
} else {
outputVector[i] = convertedVector1.vector[i] - convertedVector2.vector[i];
}
}
} else {
for (int i = 0; i != n; i++) {
if (convertedVector1.isNull[i] || convertedVector2.isNull[i]) {
outV.isNull[i] = true;
} else {
outputVector[i] = convertedVector1.vector[i] - convertedVector2.vector[i];
}
}
}
}
}
use of org.apache.hadoop.hive.ql.exec.vector.ColumnVector in project hive by apache.
the class VectorUDFDateDiffScalarCol method evaluate.
@Override
public void evaluate(VectorizedRowBatch batch) {
if (childExpressions != null) {
super.evaluateChildren(batch);
}
LongColumnVector outputColVector = (LongColumnVector) batch.cols[outputColumnNum];
ColumnVector inputCol = batch.cols[this.colNum];
/* every line below this is identical for evaluateLong & evaluateString */
final int n = inputCol.isRepeating ? 1 : batch.size;
int[] sel = batch.selected;
final boolean selectedInUse = (inputCol.isRepeating == false) && batch.selectedInUse;
boolean[] outputIsNull = outputColVector.isNull;
if (batch.size == 0) {
/* n != batch.size when isRepeating */
return;
}
// We do not need to do a column reset since we are carefully changing the output.
outputColVector.isRepeating = false;
PrimitiveCategory primitiveCategory0 = ((PrimitiveTypeInfo) inputTypeInfos[0]).getPrimitiveCategory();
switch(primitiveCategory0) {
case DATE:
baseDate = (int) longValue;
break;
case TIMESTAMP:
date.setTime(timestampValue.getTime());
baseDate = DateWritable.dateToDays(date);
break;
case STRING:
case CHAR:
case VARCHAR:
try {
date.setTime(formatter.parse(new String(stringValue, "UTF-8")).getTime());
baseDate = DateWritable.dateToDays(date);
break;
} catch (Exception e) {
outputColVector.noNulls = false;
if (selectedInUse) {
for (int j = 0; j < n; j++) {
int i = sel[j];
outputColVector.isNull[i] = true;
}
} else {
for (int i = 0; i < n; i++) {
outputColVector.isNull[i] = true;
}
}
return;
}
default:
throw new Error("Unsupported input type " + primitiveCategory0.name());
}
PrimitiveCategory primitiveCategory1 = ((PrimitiveTypeInfo) inputTypeInfos[1]).getPrimitiveCategory();
switch(primitiveCategory1) {
case DATE:
if (inputCol.isRepeating) {
if (inputCol.noNulls || !inputCol.isNull[0]) {
outputColVector.isNull[0] = false;
outputColVector.vector[0] = evaluateDate(inputCol, 0);
} else {
outputColVector.isNull[0] = true;
outputColVector.noNulls = false;
}
outputColVector.isRepeating = true;
} else if (inputCol.noNulls) {
if (batch.selectedInUse) {
if (!outputColVector.noNulls) {
for (int j = 0; j != n; j++) {
final int i = sel[j];
// Set isNull before call in case it changes it mind.
outputIsNull[i] = false;
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
} else {
for (int j = 0; j != n; j++) {
final int i = sel[j];
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
}
} else {
if (!outputColVector.noNulls) {
// Assume it is almost always a performance win to fill all of isNull so we can
// safely reset noNulls.
Arrays.fill(outputIsNull, false);
outputColVector.noNulls = true;
}
for (int i = 0; i != n; i++) {
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
}
} else /* there are NULLs in the inputColVector */
{
// Carefully handle NULLs..
// Handle case with nulls. Don't do function if the value is null, to save time,
// because calling the function can be expensive.
outputColVector.noNulls = false;
if (selectedInUse) {
for (int j = 0; j < n; j++) {
int i = sel[j];
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
}
} else {
for (int i = 0; i < n; i++) {
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
outputColVector.vector[i] = evaluateDate(inputCol, i);
}
}
}
}
break;
case TIMESTAMP:
if (inputCol.isRepeating) {
if (inputCol.noNulls || !inputCol.isNull[0]) {
outputColVector.isNull[0] = false;
outputColVector.vector[0] = evaluateTimestamp(inputCol, 0);
} else {
outputColVector.isNull[0] = true;
outputColVector.noNulls = false;
}
outputColVector.isRepeating = true;
} else if (inputCol.noNulls) {
if (batch.selectedInUse) {
if (!outputColVector.noNulls) {
for (int j = 0; j != n; j++) {
final int i = sel[j];
// Set isNull before call in case it changes it mind.
outputIsNull[i] = false;
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
} else {
for (int j = 0; j != n; j++) {
final int i = sel[j];
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
}
} else {
if (!outputColVector.noNulls) {
// Assume it is almost always a performance win to fill all of isNull so we can
// safely reset noNulls.
Arrays.fill(outputIsNull, false);
outputColVector.noNulls = true;
}
for (int i = 0; i != n; i++) {
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
}
} else /* there are nulls in the inputColVector */
{
// Carefully handle NULLs..
// Handle case with nulls. Don't do function if the value is null, to save time,
// because calling the function can be expensive.
outputColVector.noNulls = false;
if (selectedInUse) {
for (int j = 0; j < n; j++) {
int i = sel[j];
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
}
} else {
for (int i = 0; i < n; i++) {
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
outputColVector.vector[i] = evaluateTimestamp(inputCol, i);
}
}
}
}
break;
case STRING:
case CHAR:
case VARCHAR:
if (inputCol.isRepeating) {
if (inputCol.noNulls || !inputCol.isNull[0]) {
outputColVector.isNull[0] = false;
evaluateString(inputCol, outputColVector, 0);
} else {
outputColVector.isNull[0] = true;
outputColVector.noNulls = false;
}
outputColVector.isRepeating = true;
} else if (inputCol.noNulls) {
if (batch.selectedInUse) {
if (!outputColVector.noNulls) {
for (int j = 0; j != n; j++) {
final int i = sel[j];
// Set isNull before call in case it changes it mind.
outputIsNull[i] = false;
evaluateString(inputCol, outputColVector, i);
}
} else {
for (int j = 0; j != n; j++) {
final int i = sel[j];
evaluateString(inputCol, outputColVector, i);
}
}
} else {
if (!outputColVector.noNulls) {
// Assume it is almost always a performance win to fill all of isNull so we can
// safely reset noNulls.
Arrays.fill(outputIsNull, false);
outputColVector.noNulls = true;
}
for (int i = 0; i != n; i++) {
evaluateString(inputCol, outputColVector, i);
}
}
} else /* there are nulls in the inputColVector */
{
// Carefully handle NULLs..
// Handle case with nulls. Don't do function if the value is null, to save time,
// because calling the function can be expensive.
outputColVector.noNulls = false;
if (selectedInUse) {
for (int j = 0; j < n; j++) {
int i = sel[j];
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
evaluateString(inputCol, outputColVector, i);
}
}
} else {
for (int i = 0; i < n; i++) {
outputColVector.isNull[i] = inputCol.isNull[i];
if (!inputCol.isNull[i]) {
evaluateString(inputCol, outputColVector, i);
}
}
}
}
break;
default:
throw new Error("Unsupported input type " + primitiveCategory1.name());
}
}
use of org.apache.hadoop.hive.ql.exec.vector.ColumnVector in project hive by apache.
the class VectorUDFMapIndexBaseScalar method evaluate.
@Override
public void evaluate(VectorizedRowBatch batch) {
if (childExpressions != null) {
super.evaluateChildren(batch);
}
ColumnVector outV = batch.cols[outputColumnNum];
MapColumnVector mapV = (MapColumnVector) batch.cols[mapColumnNum];
/*
* Do careful maintenance of the outputColVector.noNulls flag.
*/
int[] mapValueIndex;
if (mapV.isRepeating) {
if (mapV.isNull[0]) {
outV.isNull[0] = true;
outV.noNulls = false;
} else {
mapValueIndex = getMapValueIndex(mapV, batch);
if (mapValueIndex[0] == -1) {
// the key is not found in MapColumnVector, set the output as null ColumnVector
outV.isNull[0] = true;
outV.noNulls = false;
} else {
// the key is found in MapColumnVector, set the value
outV.setElement(0, (int) (mapV.offsets[0] + mapValueIndex[0]), mapV.values);
}
}
outV.isRepeating = true;
} else {
mapValueIndex = getMapValueIndex(mapV, batch);
for (int i = 0; i < batch.size; i++) {
int j = (batch.selectedInUse) ? batch.selected[i] : i;
if (mapV.isNull[j] || mapValueIndex[j] == -1) {
outV.isNull[j] = true;
outV.noNulls = false;
} else {
outV.isNull[j] = false;
outV.setElement(j, (int) (mapV.offsets[j] + mapValueIndex[j]), mapV.values);
}
}
outV.isRepeating = false;
}
}
Aggregations