akkio-97 opened a new pull request #4073: URL: https://github.com/apache/carbondata/pull/4073 ### Why is this PR needed? ### What changes were proposed in this PR? ### Does this PR introduce any user interface change? - No - Yes. (please explain the change and update document) ### Is any new testcase added? - No - Yes ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
CarbonDataQA2 commented on pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#issuecomment-756396608 Build Success with Spark 2.3.4, Please check CI http://121.244.95.60:12444/job/ApacheCarbonPRBuilder2.3/5290/ ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
CarbonDataQA2 commented on pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#issuecomment-756399694 Build Success with Spark 2.4.5, Please check CI http://121.244.95.60:12444/job/ApacheCarbon_PR_Builder_2.4.5/3530/ ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
VenuReddy2103 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r556246616 ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -328,9 +328,29 @@ public BigDecimal getDecimal(Object valueToBeConverted) { public void fillVector(Object valuesToBeConverted, int size, ColumnVectorInfo vectorInfo, BitSet nullBitSet, DataType pageType) { CarbonColumnVector vector = getCarbonColumnVector(vectorInfo, nullBitSet); - //TODO handle complex child - int precision = vectorInfo.measure.getMeasure().getPrecision(); - int newMeasureScale = vectorInfo.measure.getMeasure().getScale(); + int precision; Review comment: These changes seem to be same as in`DecimalIntConverter.fillVector()`. Shall we reuse it instead of duplicating same logic again here ? Probably `DecimalUnscaledConverter` can extend `DecimalIntConverter` and overide fillVector like below ? > public void fillVector(Object valuesToBeConverted, int size, ColumnVectorInfo vectorInfo, BitSet nullBitSet, DataType pageType) { if (valuesToBeConverted instanceof byte[][]) { ... } else { super.fillVector(valuesToBeConverted, size, vectorInfo, nullBitSet, pageType); } } ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
VenuReddy2103 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r556246909 ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -348,6 +368,32 @@ public void fillVector(Object valuesToBeConverted, int size, vector.putDecimal(i, value, precision); } } + } else if (valuesToBeConverted instanceof byte[]) { Review comment: Shall we avoid this code and reuse ? Same as above. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
VenuReddy2103 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r556248112 ########## File path: integration/spark/src/test/scala/org/apache/carbondata/integration/spark/testsuite/dataload/SparkStoreCreatorForPresto.scala ########## @@ -365,6 +365,15 @@ class SparkStoreCreatorForPresto extends QueryTest with BeforeAndAfterAll{ sql(s"""LOAD DATA LOCAL INPATH '$resourcesPath/IUD/dest.csv' INTO TABLE streaming_table""") } + test("Test decimal unscaled converter") { + sql("drop table if exists array_decimal") + sql( + "CREATE TABLE IF NOT EXISTS array_decimal (salary array<decimal(20,3)>) STORED AS " + + "carbondata" + ) + sql("insert into array_decimal select array(922.580, 3.435) ") Review comment: Didn't drop this table in `afterAll()` like other testcases do ? ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
VenuReddy2103 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r556262687 ########## File path: integration/presto/src/test/scala/org/apache/carbondata/presto/integrationtest/PrestoTestUsingSparkStore.scala ########## @@ -325,4 +326,22 @@ class PrestoTestUsingSparkStore } + test("Test decimal unscaled converter") { + prestoServer + .execute( + "create table presto_spark_db.array_decimal(salary array(decimal(20,3)) ) with" + + "(format='CARBON') ") + copyStoreContents("array_decimal") + val result: List[Map[String, Any]] = prestoServer + .executeQuery("SELECT * FROM presto_spark_db.array_decimal") + assert(result.size == 1) + for (i <- 0 to 0) { Review comment: result size is only 1. Suggest to remove this redundant loop if not required. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
ajantha-bhat commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r559992716 ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -328,9 +328,29 @@ public BigDecimal getDecimal(Object valueToBeConverted) { public void fillVector(Object valuesToBeConverted, int size, ColumnVectorInfo vectorInfo, BitSet nullBitSet, DataType pageType) { CarbonColumnVector vector = getCarbonColumnVector(vectorInfo, nullBitSet); - //TODO handle complex child - int precision = vectorInfo.measure.getMeasure().getPrecision(); - int newMeasureScale = vectorInfo.measure.getMeasure().getScale(); + int precision; + int newMeasureScale; + if (vectorInfo.measure == null) { + // complex primitive decimal flow comes as dimension + precision = ((DecimalType) vector.getType()).getPrecision(); + newMeasureScale = ((DecimalType) vector.getType()).getScale(); + size = ColumnVectorInfo.getUpdatedPageSizeForChildVector(vectorInfo, size); + } else { + precision = vectorInfo.measure.getMeasure().getPrecision(); + newMeasureScale = vectorInfo.measure.getMeasure().getScale(); + } Review comment: can this be extracted into a common private method and reuse the same in all the places ? ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
ajantha-bhat commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r559993362 ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -348,6 +368,32 @@ public void fillVector(Object valuesToBeConverted, int size, vector.putDecimal(i, value, precision); } } + } else if (valuesToBeConverted instanceof byte[]) { Review comment: yes, check and resue the code for this also ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
CarbonDataQA2 commented on pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#issuecomment-763603712 Build Failed with Spark 2.4.5, Please check CI http://121.244.95.60:12444/job/ApacheCarbon_PR_Builder_2.4.5/3560/ ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
CarbonDataQA2 commented on pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#issuecomment-763604307 Build Failed with Spark 2.3.4, Please check CI http://121.244.95.60:12444/job/ApacheCarbonPRBuilder2.3/5320/ ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
CarbonDataQA2 commented on pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#issuecomment-763710397 Build Success with Spark 2.3.4, Please check CI http://121.244.95.60:12444/job/ApacheCarbonPRBuilder2.3/5321/ ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
CarbonDataQA2 commented on pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#issuecomment-763714150 Build Success with Spark 2.4.5, Please check CI http://121.244.95.60:12444/job/ApacheCarbon_PR_Builder_2.4.5/3561/ ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
akkio-97 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r561605603 ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -328,9 +328,29 @@ public BigDecimal getDecimal(Object valueToBeConverted) { public void fillVector(Object valuesToBeConverted, int size, ColumnVectorInfo vectorInfo, BitSet nullBitSet, DataType pageType) { CarbonColumnVector vector = getCarbonColumnVector(vectorInfo, nullBitSet); - //TODO handle complex child - int precision = vectorInfo.measure.getMeasure().getPrecision(); - int newMeasureScale = vectorInfo.measure.getMeasure().getScale(); + int precision; Review comment: done ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -348,6 +368,32 @@ public void fillVector(Object valuesToBeConverted, int size, vector.putDecimal(i, value, precision); } } + } else if (valuesToBeConverted instanceof byte[]) { Review comment: done ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
akkio-97 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r561605668 ########## File path: integration/spark/src/test/scala/org/apache/carbondata/integration/spark/testsuite/dataload/SparkStoreCreatorForPresto.scala ########## @@ -365,6 +365,15 @@ class SparkStoreCreatorForPresto extends QueryTest with BeforeAndAfterAll{ sql(s"""LOAD DATA LOCAL INPATH '$resourcesPath/IUD/dest.csv' INTO TABLE streaming_table""") } + test("Test decimal unscaled converter") { + sql("drop table if exists array_decimal") + sql( + "CREATE TABLE IF NOT EXISTS array_decimal (salary array<decimal(20,3)>) STORED AS " + + "carbondata" + ) + sql("insert into array_decimal select array(922.580, 3.435) ") Review comment: done ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
akkio-97 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r561606064 ########## File path: integration/presto/src/test/scala/org/apache/carbondata/presto/integrationtest/PrestoTestUsingSparkStore.scala ########## @@ -325,4 +326,22 @@ class PrestoTestUsingSparkStore } + test("Test decimal unscaled converter") { + prestoServer + .execute( + "create table presto_spark_db.array_decimal(salary array(decimal(20,3)) ) with" + + "(format='CARBON') ") + copyStoreContents("array_decimal") + val result: List[Map[String, Any]] = prestoServer + .executeQuery("SELECT * FROM presto_spark_db.array_decimal") + assert(result.size == 1) + for (i <- 0 to 0) { Review comment: done ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
akkio-97 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r561605603 ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -328,9 +328,29 @@ public BigDecimal getDecimal(Object valueToBeConverted) { public void fillVector(Object valuesToBeConverted, int size, ColumnVectorInfo vectorInfo, BitSet nullBitSet, DataType pageType) { CarbonColumnVector vector = getCarbonColumnVector(vectorInfo, nullBitSet); - //TODO handle complex child - int precision = vectorInfo.measure.getMeasure().getPrecision(); - int newMeasureScale = vectorInfo.measure.getMeasure().getScale(); + int precision; Review comment: done ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -348,6 +368,32 @@ public void fillVector(Object valuesToBeConverted, int size, vector.putDecimal(i, value, precision); } } + } else if (valuesToBeConverted instanceof byte[]) { Review comment: done ########## File path: integration/spark/src/test/scala/org/apache/carbondata/integration/spark/testsuite/dataload/SparkStoreCreatorForPresto.scala ########## @@ -365,6 +365,15 @@ class SparkStoreCreatorForPresto extends QueryTest with BeforeAndAfterAll{ sql(s"""LOAD DATA LOCAL INPATH '$resourcesPath/IUD/dest.csv' INTO TABLE streaming_table""") } + test("Test decimal unscaled converter") { + sql("drop table if exists array_decimal") + sql( + "CREATE TABLE IF NOT EXISTS array_decimal (salary array<decimal(20,3)>) STORED AS " + + "carbondata" + ) + sql("insert into array_decimal select array(922.580, 3.435) ") Review comment: done ########## File path: integration/presto/src/test/scala/org/apache/carbondata/presto/integrationtest/PrestoTestUsingSparkStore.scala ########## @@ -325,4 +326,22 @@ class PrestoTestUsingSparkStore } + test("Test decimal unscaled converter") { + prestoServer + .execute( + "create table presto_spark_db.array_decimal(salary array(decimal(20,3)) ) with" + + "(format='CARBON') ") + copyStoreContents("array_decimal") + val result: List[Map[String, Any]] = prestoServer + .executeQuery("SELECT * FROM presto_spark_db.array_decimal") + assert(result.size == 1) + for (i <- 0 to 0) { Review comment: done ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
Indhumathi27 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r562541775 ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -88,11 +88,34 @@ void fillVector(Object valuesToBeConverted, int size, ColumnVectorInfo info, Bit DecimalConverterType getDecimalConverterType(); + default int getPrecisionForDimension(CarbonColumnVector vector) { Review comment: I can see, these methods are called from a single place only. Then, why it is required to create these methods ?. Instead, modify caller method like below: """ public void initializeValues(int size, CarbonColumnVector vector, ColumnVectorInfo vectorInfo) { if (vectorInfo.measure == null) { DecimalType decimalType = (DecimalType) vector.getType(); // complex primitive decimal flow comes as dimension this.precision = decimalType.getPrecision(); this.newMeasureScale = decimalType.getScale(); this.updatedSize = ColumnVectorInfo.getUpdatedPageSizeForChildVector(vectorInfo, size); } else { CarbonMeasure carbonMeasure = vectorInfo.measure.getMeasure(); this.precision = carbonMeasure.getPrecision(); this.newMeasureScale = carbonMeasure.getScale(); } } """ ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
Indhumathi27 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r562543817 ########## File path: integration/spark/src/test/scala/org/apache/carbondata/integration/spark/testsuite/dataload/SparkStoreCreatorForPresto.scala ########## @@ -365,6 +366,15 @@ class SparkStoreCreatorForPresto extends QueryTest with BeforeAndAfterAll{ sql(s"""LOAD DATA LOCAL INPATH '$resourcesPath/IUD/dest.csv' INTO TABLE streaming_table""") } + test("Test decimal unscaled converter") { + sql("drop table if exists array_decimal") + sql( + "CREATE TABLE IF NOT EXISTS array_decimal (salary array<decimal(20,3)>) STORED AS " + Review comment: can you add a struct column with decimal type also ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
In reply to this post by GitBox
Indhumathi27 commented on a change in pull request #4073: URL: https://github.com/apache/carbondata/pull/4073#discussion_r562546331 ########## File path: core/src/main/java/org/apache/carbondata/core/metadata/datatype/DecimalConverterFactory.java ########## @@ -141,100 +165,116 @@ public void fillVector(Object valuesToBeConverted, int size, // for string, varchar, binary, date, decimal types lengthStoredInBytes = shortSizeInBytes; } - byte[] data = (byte[]) valuesToBeConverted; - if (pageType == DataTypes.BYTE) { - for (int i = 0; i < size; i++) { - if (nullBitSet.get(i)) { - vector.putNull(i); - } else { - BigDecimal value = BigDecimal.valueOf(data[i], scale); - if (value.scale() < newMeasureScale) { - value = value.setScale(newMeasureScale); - } - vector.putDecimal(i, value, precision); - } - } - } else if (pageType == DataTypes.SHORT) { + if (this instanceof DecimalUnscaledConverter && scale < newMeasureScale) { + scale = newMeasureScale; + } + + if (valuesToBeConverted instanceof byte[][]) { + byte[][] data = (byte[][]) valuesToBeConverted; for (int i = 0; i < size; i++) { if (nullBitSet.get(i)) { vector.putNull(i); } else { - BigDecimal value = BigDecimal - .valueOf(ByteUtil.toShortLittleEndian(data, i * shortSizeInBytes), - scale); + BigInteger bigInteger = new BigInteger(data[i]); + BigDecimal value = new BigDecimal(bigInteger, scale); if (value.scale() < newMeasureScale) { value = value.setScale(newMeasureScale); } vector.putDecimal(i, value, precision); } } - } else if (pageType == DataTypes.SHORT_INT) { - int shortIntSizeInBytes = DataTypes.SHORT_INT.getSizeInBytes(); - for (int i = 0; i < size; i++) { - if (nullBitSet.get(i)) { - vector.putNull(i); - } else { - BigDecimal value = BigDecimal - .valueOf(ByteUtil.valueOf3Bytes(data, i * shortIntSizeInBytes), - scale); - if (value.scale() < newMeasureScale) { - value = value.setScale(newMeasureScale); + } else if (valuesToBeConverted instanceof byte[]) { + byte[] data = (byte[]) valuesToBeConverted; + if (pageType == DataTypes.BYTE) { + for (int i = 0; i < size; i++) { + if (nullBitSet.get(i)) { + vector.putNull(i); + } else { + BigDecimal value = BigDecimal.valueOf(data[i], scale); Review comment: I think, we can still refactor the code for Cases: Byte, Short, Short_Int, Int and Long, As only difference is getting Bigdecimal value. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [hidden email] |
Free forum by Nabble | Edit this page |