Improve InferFiltersFromConstraints rule performance when parsing spark sql

test("Expression explosion when analyze test") {
    RuleExecutor.resetMetrics()
    Seq((1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14))
      .toDF("a", "b", "c", "d", "e", "f", "g", "h", "i", "j",
        "k", "l", "m", "n")
      .write.saveAsTable("test")
    val df = spark.table("test")
    val df2 = df.filter("a+b+c+d+e+f+g+h+i+j+k+l+m+n > 100")
    val df3 = df2.select('a as 'a1, 'b as 'b1,
      'c as 'c1, 'd as 'd1, 'e as 'e1, 'f as 'f1,
      'g as 'g1, 'h as 'h1, 'i as 'i1, 'j as 'j1,
      'k as 'k1, 'l as 'l1, 'm as 'm1, 'n as 'n1)
    val df4 = df3.join(df2, df3("a1") === df2("a"))
    df4.explain(true)
    logWarning(RuleExecutor.dumpTimeSpent())
  }