Amazon web services – How to preserve the key letter case in AWS Glue Transform node?

I am trying to add a new column called timestamp in AWS Glue.

My upstream data keys have capital letters.

However, after adding the column timestamp, the keys of the remaining columns got lowercased.

Experiment 1: Transform – SQL Query

select 
    cast((Absolute_Time - 2082844800) * 1000 as bigint) as timestamp,
    *
from
    engine_demo;

However, it lowercases all my keys.

Experiment 2: Transform – Custom code

def MyTransform (glueContext, dfc) -> DynamicFrameCollection:
    df = dfc.select(list(dfc.keys())[0]).toDF()
    
    from pyspark.sql import functions as F
    df = df.withColumn(
      "timestamp",
      ((F.col("Absolute_Time") - 2082844800) * 1000).cast("long")
    )
    dynamic_frame = DynamicFrame.fromDF(df, glueContext, "changed")
    return (DynamicFrameCollection({"result_dynf": dynamic_frame}, glueContext))

It also lowercases all my keys

Experiment 3

I found Enable case sensitivity for spark.sql globally

However, there is no place to set in "Transform – SQL Query" node, so I try to set the config in "Transform – Custom" code:

def MyTransform (glueContext, dfc) -> DynamicFrameCollection:
    glueContext.spark_session.conf.set("spark.sql.caseSensitive", "true")
    
    df = dfc.select(list(dfc.keys())[0]).toDF()
    from pyspark.sql import functions as F
    df = df.withColumn(
      "timestamp",
      ((F.col("Absolute_Time") - 2082844800) * 1000).cast("long")
    )
    dynamic_frame = DynamicFrame.fromDF(df, glueContext, "changed")
    return (DynamicFrameCollection({"dynamic_frame": dynamic_frame}, glueContext))

However, this above code gives same result.

Is there a way to preserve the key letter case in the AWS Glue Transform node? Thanks!