improve build process

[damonmcc]

motivations

• failure to import dof_dtm started a deep dive into how it's imported
• although multiple bash and SQL errors occurred (example 1, example 2), builds would continue and show green checkmarks. this is a confusing and not a desired behavior

changes

• add set -e to all bash scripts to exit if any command has a non-zero exit status
• add -v ON_ERROR_STOP=1 to SQL queries in bash files to ensure they stop on errors
• add and use a bash function run_sql_file to run all sql files and stop on errors
• fix preprocessing bug around preprocessing
  • was creating and deleting dcp_zoning_taxlot without using it and then renaming dcp_zoningtaxlots to dcp_zoning_taxlot (but dcp_zoningtaxlots table never exists)
  • now just creating dcp_zoning_taxlot
• add ST_MakeValid to preprocessing of dof_dtm.geom column to handle latest data (error in a build)
• autoformat config.sh
• add scripts used to re-project and convert dof_dtm source data due to data library failing to do so

notes

• theses changes were used in latest successful build and that data has been QAed and pushed to Bytes
• bash scripts related to dof_dtm have been added for potential future use. but ideally they're here for reference in a more long-term fix to the issue in data library (https://github.com/NYCPlanning/db-data-library/issues/381)

[fvankrieken]

A couple very minor questions, but looks good and I love it - I've been using set -e locally to make getting my environment set up much more sane, and personally would love to make that our standard practice to avoid these silent failures