AWS Glue - Redshift 中具有 Json 结构的字段

Posted 2023-03-31

【中文标题】AWS Glue - Redshift 中具有 Json 结构的字段

【英文标题】：AWS Glue - Field with Json structure in Redshift

【发布时间】：2021-07-08 14:58:11

【问题描述】：

您好，我正在使用 AWS Glue 尝试将数据从 S3 中的 Json 文件加载到 Redshift。我正在使用路径为 $[*] 的 Json 爬虫，由于某种原因，其中一个字段（等级）以 Json 结构进入表：

关于如何使“等级”仅显示等级本身的值的任何想法？我是否需要针对这项工作调整 PySpark 脚本？

这是目前为止的脚本：

import sys
from awsglue.transforms import *
from awsglue.utils import getResolvedOptions
from pyspark.context import SparkContext
from awsglue.context import GlueContext
from awsglue.job import Job
from pyspark.sql.functions import current_date 
from awsglue.dynamicframe import DynamicFrame

## @params: [TempDir, JOB_NAME]
args = getResolvedOptions(sys.argv, ['TempDir','JOB_NAME'])

sc = SparkContext()
glueContext = GlueContext(sc)
spark = glueContext.spark_session
job = Job(glueContext)
job.init(args['JOB_NAME'], args)
## @type: DataSource
## @args: [database = "linkredshift", table_name = "uni3", transformation_ctx = "datasource0"]
## @return: datasource0
## @inputs: []
datasource0 = glueContext.create_dynamic_frame.from_catalog(database = "linkredshift", table_name = "uni3", transformation_ctx = "datasource0")
df=datasource0.toDF().withColumn('data_date',current_date())  
datasource0 = DynamicFrame.fromDF(df, glueContext, "datasource0")

## @type: ApplyMapping
## @args: [mapping = [("first_name", "string", "first_name", "string"), ("last_name", "string", "last_name", "string"), ("subject", "string", "subject", "string"), ("grade", "string", "grade", "string")], transformation_ctx = "applymapping1"]
## @return: applymapping1
## @inputs: [frame = datasource0]
applymapping1 = ApplyMapping.apply(frame = datasource0, mappings = [("first_name", "string", "first_name", "string"), ("last_name", "string", "last_name", "string"), ("subject", "string", "subject", "string"), ("grade", "string", "grade", "string"), ("data_date", "date", "data_date", "date")], transformation_ctx = "applymapping1")
## @type: SelectFields
## @args: [paths = ["subject", "grade", "last_name", "first_name"], transformation_ctx = "selectfields2"]
## @return: selectfields2
## @inputs: [frame = applymapping1]
selectfields2 = SelectFields.apply(frame = applymapping1, paths = ["subject", "grade", "last_name", "first_name", "data_date"], transformation_ctx = "selectfields2")
## @type: ResolveChoice
## @args: [choice = "MATCH_CATALOG", database = "linkredshift", table_name = "dev_public_students", transformation_ctx = "resolvechoice3"]
## @return: resolvechoice3
## @inputs: [frame = selectfields2]
resolvechoice3 = ResolveChoice.apply(frame = selectfields2, choice = "MATCH_CATALOG", database = "linkredshift", table_name = "dev_public_students", transformation_ctx = "resolvechoice3")
## @type: ResolveChoice
## @args: [choice = "make_cols", transformation_ctx = "resolvechoice4"]
## @return: resolvechoice4
## @inputs: [frame = resolvechoice3]
resolvechoice4 = ResolveChoice.apply(frame = resolvechoice3, choice = "make_cols", transformation_ctx = "resolvechoice4")
## @type: DataSink
## @args: [database = "linkredshift", table_name = "dev_public_students", redshift_tmp_dir = TempDir, transformation_ctx = "datasink5"]
## @return: datasink5
## @inputs: [frame = resolvechoice4]
datasink5 = glueContext.write_dynamic_frame.from_catalog(frame = resolvechoice4, database = "linkredshift", table_name = "dev_public_students", redshift_tmp_dir = args["TempDir"], transformation_ctx = "datasink5")
job.commit()

【参考方案1】：

由于等级字段可以是字符串或整数，因此需要进行带有强制转换的解析选择，使其成为其中之一。尝试将您的 applymapping1 行更改为这两行：

datasource0= datasource0.resolveChoice(specs=[("grade",'cast:int')])
applymapping1 = ApplyMapping.apply(frame = datasource0, mappings = [("first_name", "string", "first_name", "string"), ("last_name", "string", "last_name", "string"), ("subject", "string", "subject", "string"), ("grade", "int", "grade", "int"), ("data_date", "date", "data_date", "date")], transformation_ctx = "applymapping1")

请注意，这会将您的成绩字段更改为 int 类型（因为这是我在 resolveChoice 中强制转换的内容，以便使用 int 中的值）。以后可以随意将成绩字段转换为字符串，但 int 似乎是更好的选择。

参考这里：https://docs.aws.amazon.com/glue/latest/dg/aws-glue-programming-python-samples-medicaid.html

【讨论】：

由于某种原因，当我用上面的两行替换 applymapping1 时，数据没有被加载，我只剩下一个空表和 2 个等级列：“grade”和“grade_int”。你能确认一下吗？

糟糕，这应该是一个项目，而不是演员表。我更新了代码。

现在我改用“项目”运行，所有其他列都已加载但分级。此外，我仍然得到 2 列：“grade”和“grade_int”。你能确认一下吗？谢谢

嗯，你能在 resolvechoice 行之后添加一个datasource0.printSchema() 并告诉我架构日志中的内容吗？注意我恢复到解决方案中的强制转换，因为这应该可以工作，并且我添加了一个 AWS 链接作为一些背景。

当然可以，现在运行！