ref() when generating multiple tables in Dataform

[fuzzy-unicorn]

I want to set up a step in my pipeline, where I 

• export all tables in my "source" layer dataset to GCS (as a daily snapshot)
• load the most recent file in GCS for each table, to my "staging" layer dataset

I've written this operation to achieve this:

config { type: "operations" }

js {
 const now = new Date()
 now.setHours(now.getHours() + 9)
 const year = `${now.getFullYear()}`
 const month = `${String(now.getMonth() + 1).padStart(2, '0')}`
 const day = `${String(now.getDate()).padStart(2, '0')}`
 const hour = `${String(now.getHours()).padStart(2, '0')}`
 const minute = `${String(now.getMinutes()).padStart(2, '0')}`
 const second = `${String(now.getSeconds()).padStart(2, '0')}`
 const timestamp = `${year}${month}${day}_${hour}${minute}${second}`
}
for record in (
 select
   concat("source_dataset.", table_name) as source_table_id
   , concat("stg_dataset.", table_name) as stg_table_id
   , concat(
     'gs://${dataform.projectConfig.vars.snapshot_bucket_name}/'
     , table_name
     , '/${timestamp}/*.parquet'
) as gs_uri
 from `region-asia-northeast1`.INFORMATION_SCHEMA.TABLES
 where table_schema = "source_dataset"
)
do
  execute immediate ("""
    export data
    options (
    uri = '""" || record.gs_uri || """',
    format = 'parquet',
    overwrite = false)
    as (
      select *
      from """ || record.source_table_id || """
    );
  """);

 execute immediate ("""
   load data overwrite""" || record.stg_table_id || """
   from files (
   format = 'parquet',
   uris = ['""" || record.gs_uri || """']
   );
 """);
end for

However, I am wondering how I can reference each staging table in my downstream files. Since I do not generate each table with a SELECT statement, I cannot use `publish()` to define them, which means I cannot use `ref()` in my downstream SQLX files.

Currently I am using `dependencies` in the config block to manage the graph, but I am wondering if there is a better way.