[jira] [Updated] (CARBONDATA-2515) Filter OR Expression not working properly in Presto integration

[Akash R Nilugal (Jira)]

     [ https://issues.apache.org/jira/browse/CARBONDATA-2515?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sourabh Verma updated CARBONDATA-2515:
--------------------------------------
    Description:
Scenario - carbon-data Table 'load_table' with columns 'integer', 'datetime'

//table creation and load code (spark)
 val random = new Random()
 val df = spark.sparkContext.parallelize(1 to (365 * 24 * 360))
 .map(x => (random.nextInt(200), new Timestamp(currentMillis - (x * 1000l))))
 .toDF("integer", "datetime")

// Saves dataframe to carbondata file
 df.write.format("carbondata")
 .option("tableName", "load_table")
 .option("compress", "true")
 .option("tempCSV", "false")
 .mode(SaveMode.Overwrite)
 .save()

SQL (through Presto CLI) - select * from load_table where integer < 10 or integer > 50;

Actual result - 0 rows.
 Expected result - rows with integer value less than 10 and greater than 50.

cause - PrestoFilterUtil is creating AND Expressions.

  was:
Scenario - carbon-data Table 'load_table' with columns 'integer', 'datetime'

//table creation and load code (spark)
val random = new Random()
val df = spark.sparkContext.parallelize(1 to (365 * 24 * 360))
 .map(x => (random.nextInt(200), new Timestamp(currentMillis - ((x) * 1000l))))
 .toDF("integer", "datetime")

// Saves dataframe to carbondata file
df.write.format("carbondata")
 .option("tableName", "load_table")
 .option("compress", "true")
 .option("tempCSV", "false")
 .mode(SaveMode.Overwrite)
 .save()

SQL (through Presto CLI) - select * from load_table where integer < 10 or integer > 50;

Actual result - 0 rows.
Expected result - rows with integer value less than 10 and greater than 50.

cause - PrestoFilterUtil is creating AND Expressions.

> Filter OR Expression not working properly in Presto integration
> ---------------------------------------------------------------
>
>                 Key: CARBONDATA-2515
>                 URL: https://issues.apache.org/jira/browse/CARBONDATA-2515
>             Project: CarbonData
>          Issue Type: Bug
>          Components: presto-integration
>    Affects Versions: 1.4.0
>         Environment: Spark 2.1, Presto 0.187
>            Reporter: Sourabh Verma
>            Priority: Major
>             Fix For: 1.4.0
>
>
> Scenario - carbon-data Table 'load_table' with columns 'integer', 'datetime'
> //table creation and load code (spark)
>  val random = new Random()
>  val df = spark.sparkContext.parallelize(1 to (365 * 24 * 360))
>  .map(x => (random.nextInt(200), new Timestamp(currentMillis - (x * 1000l))))
>  .toDF("integer", "datetime")
> // Saves dataframe to carbondata file
>  df.write.format("carbondata")
>  .option("tableName", "load_table")
>  .option("compress", "true")
>  .option("tempCSV", "false")
>  .mode(SaveMode.Overwrite)
>  .save()
> SQL (through Presto CLI) - select * from load_table where integer < 10 or integer > 50;
> Actual result - 0 rows.
>  Expected result - rows with integer value less than 10 and greater than 50.
> cause - PrestoFilterUtil is creating AND Expressions.

--
This message was sent by Atlassian JIRA
(v7.6.3#76005)