我有spark嵌套框架df 1。我想过滤的基础上failed_rules列有单引号。我试过下面的代码，但得到错误。
import spark.implicits._

val columns = Seq("failed_rules","CMF_BANKRPT_IND","year")
val tupleSeq: Seq[(String, String, Int)] = Seq(("`name: isComplete,args:{column: cmf_lf_assess_num_hist,where_cond: CMF_BANKRPT_IND == 'B'}`", "B" , 1981),("`name: isComplete,Failure_Flag: Error,args:{column: cmf_lf_assess_num_hist,where_cond: CMF_BANKRPT_IND == 'B'}`", "A", 1989),("4", "John", 1991))

val df1=spark.createDataFrame(tupleSeq).toDF(columns:_*)
val extraCols = List("`name: isComplete,args:{column: cmf_lf_assess_num_hist,where_cond: CMF_BANKRPT_IND == 'B'}`")
val rules_wherecond = extraCols.filter( name => name.contains("where_cond"))

for (i <- 0 until rules_wherecond.length) {
  df1.filter(s"failed_rules == ${rules_wherecond(i)}").show(false)
}

错误消息：
Exception：无法解析'name: isComplete,args:{column: cmf_lf_assess_num_hist,where_cond: CMF_BANKRPT_IND == 'B'}'给定输入列：[failed_rules，CMF_BANKRPT_IND，year];第1行pos 16;