use SQL inside AWS Glue pySpark script

Question

I want to use AWS Glue to convert some csv data to orc.
The ETL job I created generated the following PySpark script:

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

args = getResolvedOptions(sys.argv, ['JOB_NAME'])

sc = SparkContext()
glueContext = GlueContext(sc)
spark = glueContext.spark_session
job = Job(glueContext)
job.init(args['JOB_NAME'], args)

datasource0 = glueContext.create_dynamic_frame.from_catalog(database = "tests", table_name = "test_glue_csv", transformation_ctx = "datasource0")

applymapping1 = ApplyMapping.apply(frame = datasource0, mappings = [("id", "int", "id", "int"), ("val", "string", "val", "string")], transformation_ctx = "applymapping1")

resolvechoice2 = ResolveChoice.apply(frame = applymapping1, choice = "make_struct", transformation_ctx = "resolvechoice2")

dropnullfields3 = DropNullFields.apply(frame = resolvechoice2, transformation_ctx = "dropnullfields3")

datasink4 = glueContext.write_dynamic_frame.from_options(frame = dropnullfields3, connection_type = "s3", connection_options = {"path": "s3://glue/output"}, format = "orc", transformation_ctx = "datasink4")
job.commit()

It takes the csv data (from the location of which the Athena table tests.test_glue_csv points to) and outputs to s3://glue/output/.

How can I insert in this script some SQL manipulations?

Thanks

Solution

You should first create a temp view/table from your dynamic frame

dyf.toDF().createOrReplaceTempView("view_dyf")

Here, dyf is your dynamic frame.

Then, use your spark object to apply sql queries on it

sqlDF = spark.sql("select * from view_dyf")
sqlDF.show()