[FEA] Support array_repeat

[viadea]

I wish we can support array_repeat.

eg:

from pyspark.sql.functions import *
df = spark.createDataFrame([(["a", "b", "a"], ["b", "c"]), (["a","a"], ["b", "c"]), (["aa"], ["b", "c"])    ], ['x', 'y'])
df.write.format("parquet").mode("overwrite").save("/tmp/testparquet")
df = spark.read.parquet("/tmp/testparquet")
df.select(array_repeat(df.x, 3).alias("repeat")).collect()

    ! <ArrayRepeat> array_repeat(x#72, 3) cannot run on GPU because GPU does not currently support the operator class org.apache.spark.sql.catalyst.expressions.ArrayRepeat

[revans2]

This is something that we probably can do without any help from cudf. The simplest way I can think of is to create the data column using cudf::repeat and the offsets buffer using a scan SUM. Ww will need to do some bounds checking to make sure that we don't overflow (Spark has similar limits, but ours are going to be a lot smaller). We might also need to fix up the nulls as a null input count results in a null array.