I have an Azure Storage Data Table that I would like to update based on records that were just streamed into a Delta Live Table. Below is example code:

@Dlt.create_table(
  comment="comment",
  table_properties={
    "pipelines.autoOptimize.managed": "true"
  },
  partition_cols = ["EventId"]
)
def live_table_test():
  return (
    spark.readStream.format("cloudFiles") 
        .options(**cloudfile) 
        .load(upload_path)
        .selectExpr("*", "_metadata.file_path as file_path")
  )

I want to update an Azure Storage Data Table based on records that were just streamed into that table. Would something like this work?

@Dlt.create_table(
  comment="comment",
  table_properties={
    "pipelines.autoOptimize.managed": "true"
  },
  partition_cols = ["EventId"]
)
def live_table_test():
  df = (
    spark.readStream.format("cloudFiles") 
        .options(**cloudfile) 
        .load(upload_path)
        .selectExpr("*", "_metadata.file_path as file_path")
  )
  update_azure_table(df)
  return df

I need to be able to get the list of distinct EventIds from the df and pass them into that function. How can I do this? And should I be using the "dlt.create_table" decorator? Or should I use the "dlt.table" decorator? I'm finding almost no documentation about this online.