spark dataframe 新增列

import org.apache.spark.sql.functions._

//新增一列以oldColName * 2 为新列，$"oldColName" 相当于 col("oldColName")
df.withColumn("newColName", $"oldColName"*2)

val cmsUrlPatten = """(^http[s]?://.+_)[0-9]+\.(shtml|html){1}.*""".r

def url2cms(url: String): String = {
val tagUrl = url.split("\\?")(0)
if (cmsUrlPatten.findAllIn(tagUrl).isEmpty) url
else {
val cmsUrlPatten(urlPre, urlEnd) = tagUrl
s"${urlPre}1.$urlEnd"
}
}

val sqlfunc = udf((arg: String) => url2cms(arg))

trafficDF.withColumn("cmsUrl", sqlfunc($"url"))