Refactor ETL

[alistairewj]

This is a big PR with lots of changes so apologies for that. I essentially re-ran the entire ETL locally and fixed many things along the way. I will highlight changes:

• My editor's auto-formatting tends to remove trailing whitespaces, etc, so you may see a lot of formatting changes
• Added a README in the build-omop/postgresql folder to explain how to prepare the OMOP schema and vocab
• Added a README-run-etl.md file which explains how translate MIMIC into OMOP
• Directly clones DDL from the OMOP repository and runs some scripts afterward to modify the schema - hopefully this is easier to maintain as OMOP changes their schema in the future
• Corrected some minor typos in the concept CSVs
• Reorganize some of the import/export files, and delete what I thought were superfluous files
• Explicitly writing out the insert columns so that we don't accidentally insert data into the wrong column
• Clean up the check ETL - it runs but it picks up errors which we will have to deal with. I tried to standardize it so that "have" was the OMOP schema and "want" was either a fixed number or from MIMIC.

There may be some regressions because for some reason my dev branch started early in the master commit history and I had to rebase. Hopefully none though, I tried to clean them all up.

[alistairewj]

Going to merge this and use dev to track updates in making ETL consistent with HIAE.