spark scala extraneous input '>' ParseException in transform-CodePudding

I'm trying to execute the query bellow:

finvInventoryAllDf
      .groupBy("Site_siteId")
      .agg(
        collect_set(
          array(
            "InstalledOffer_applicationSource",
            "InstalledOffer_standardStatus", "InstalledOffer_installedOfferId"
          )
        ).as("array")
      )
      .withColumn("indicator", expr(transformExpr))

But I get an error in the expr(transformExpr), the value of transformExpr I'm trying to execute is :

val transformExpr = "transform(array, x -> array_contains(x, 'CIBASE') and array_contains(x, 'ACTIVE'))"

But I get the error saying that the input '>' is unexpected in the -> operator above.

bellow is a screen shot of the console log:

the version of scala I'm using is 2.11.8 and the spark version is 3.17.2

CodePudding user response：

You named one of your columns array, but array is also a built-in function in Spark SQL.

Just rename your column to something else and your code will work:

val transformExpr = "transform(ar, x -> array_contains(x, 'CIBASE') and array_contains(x, 'ACTIVE'))"

finvInventoryAllDf
      .groupBy("Site_siteId")
      .agg(
        collect_set(
          array(
            "InstalledOffer_applicationSource",
            "InstalledOffer_standardStatus", "InstalledOffer_installedOfferId"
          )
        ).as("ar")
      )
      .withColumn("indicator", expr(transformExpr))