postgresql - spark 将字符串转换为 TimestampType

Question

我有一个数据框，我想将其插入到 Spark 中的 Postgresql 中。在 spark 中，DateTimestamp 列为字符串格式。在 postgreSQL 中，它是没有时区的 TimeStamp。

在日期时间列上插入数据库时​​引发错误。我确实尝试更改数据类型，但插入仍然出错。我无法弄清楚为什么转换不起作用。如果我将相同的插入字符串粘贴到 PgAdmin 中并运行，插入语句运行正常。

import java.text.SimpleDateFormat;
import java.util.Calendar
object EtlHelper {
 // Return the current time stamp

  def getCurrentTime() : String = {    
    val now = Calendar.getInstance().getTime()   
    val hourFormat = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss")   
    return hourFormat.format(now)   
  }
 }  

在另一个文件中

object CreateDimensions {

def createDimCompany(spark:SparkSession, location:String, propsLocation :String):Unit = {      
import spark.implicits._    

val dimCompanyStartTime = EtlHelper.getCurrentTime()
val dimcompanyEndTime = EtlHelper.getCurrentTime()
val prevDimCompanyId = 2
val numRdd = 27
val AuditDF = spark.createDataset(Array(("dim_company", prevDimCompanyId,numRdd,dimCompanyStartTime,dimcompanyEndTime))).toDF("audit_tbl_name","audit_tbl_id","audit_no_rows","audit_tbl_start_date","audit_tbl_end_date")//.show()

AuditDF.withColumn("audit_tbl_start_date",AuditDF.col("audit_tbl_start_date").cast(DataTypes.TimestampType))
AuditDF.withColumn("audit_tbl_end_date",AuditDF.col("audit_tbl_end_date").cast(DataTypes.TimestampType))

AuditDF.printSchema()
}  
}

root
 |-- audit_tbl_name: string (nullable = true)
 |-- audit_tbl_id: long (nullable = false)
 |-- audit_no_rows: long (nullable = false)
 |-- audit_tbl_start_date: string (nullable = true)
 |-- audit_tbl_end_date: string (nullable = true)

这是我得到的错误

INSERT INTO etl.audit_master ("audit_tbl_name","audit_tbl_id","audit_no_rows","audit_tbl_start_date","audit_tbl_end_date") VALUES ('dim_company',27,2,'2018-05-02 12:15:54','2018-05-02 12:15:59') was aborted: ERROR: column "audit_tbl_start_date" is of type timestamp without time zone but expression is of type character varying
  Hint: You will need to rewrite or cast the expression.

感谢任何帮助。

谢谢

Answer 1

AuditDF.printSchema() 采用原始 AuditDF 数据帧，因为您没有通过分配保存 .withColumn 的转换。 数据帧是不可变的对象，可以转换为另一个数据帧但不能改变自身。因此，您始终需要一个分配来保存您应用的转换。

所以正确的方法是分配以保存更改

val transformedDF = AuditDF.withColumn("audit_tbl_start_date",AuditDF.col("audit_tbl_start_date").cast(DataTypes.TimestampType))
                          .withColumn("audit_tbl_end_date",AuditDF.col("audit_tbl_end_date").cast("timestamp"))

transformedDF.printSchema()

你会看到变化

root
 |-- audit_tbl_name: string (nullable = true)
 |-- audit_tbl_id: integer (nullable = false)
 |-- audit_no_rows: integer (nullable = false)
 |-- audit_tbl_start_date: timestamp (nullable = true)
 |-- audit_tbl_end_date: timestamp (nullable = true)

.cast(DataTypes.TimestampType) 和 .cast("timestamp") 都是一样的