SQLcl

I have stream of alien csv files which I want to be staged as-is to database for further processing. Basically known things are:

1. filename
2. maximum amount of columns any file can have
3. the used separator

So how to have all load related metadata so that there is no need to check logs or investigate alien csv-files after they have been imported, but just sql the data?

Example of additional columns per import could be like:

1. sqlcl - load parameters used, e.g. : column_names off, delimiter ;, enclosure left “, enclosure right ”, encoding UTF8 .. etc. whatever was used to make the load to happen.
2. in the target table populate during import extra columns to help processing the data later on:
   1. filename
   2. csv_row_number
   3. imported_sysdate
   4. .. any other suggestions for clever fields helping later on ?

rgrds Paavo