[SPARK-6379][SQL] Support a functon to call user-defined functions registered in SQLContext

[maropu]

This is useful for using pre-defined UDFs in SQLContext;

val df = Seq(("id1", 1), ("id2", 4), ("id3", 5)).toDF("id", "value")
val sqlctx = df.sqlContext
sqlctx.udf.register("simpleUdf", (v: Int) => v * v)
df.select($"id", sqlctx.callUdf("simpleUdf", $"value"))

[rxin]

Can't we make this explicit by using "callUDF"?

[maropu]

You mean df.select($"id", callUDF($"value"))?

[rxin]

I meant something like

df.select($"id", callUDF("simpleUDF", $"value"))

[maropu]

Ok, I'll try to refine my codes.

[maropu]

@rxin fixed, and plz check it.

[marmbrus]

This probably belongs in functions.scala with the other callUDF methods. Also, please explicitly specify return types for public methods.

[marmbrus]

Once you have updated, can you also update the PR description, as this becomes the commit message. Thanks!

[maropu]

The description updated and the patch fixed.

[marmbrus]

This probably belongs in functions.scala with the other callUDF methods.

This is still defined in the wrong file.

[maropu]

Oh, misunderstood.
I'll move the function into functions.scala.

[maropu]

Fixed, and plz re-check it.

[marmbrus]

ok to test

[SparkQA]

Test build #30095 has finished for PR 5061 at commit f858aff.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.
• This patch does not change any dependencies.

[marmbrus]

Thanks! Merged to master.