Adds dataset.is_public column and populates it from metadata
Fixes an issue in update_es_index.py - if there is an error during the update, it will exit with both sm-yin and sm-yang still existing. This causes subsequent runs immediately fail because there is an assertion that only one index should exist at a time. The fix is to delete the broken index if there's an error
Caution:
This includes a database schema change that will have to be manually applied until we get a migrations system: ALTER TABLE dataset ADD COLUMN is_public boolean not null default(true);
Because is_public is non-nullable, this needs the feature/private-prototype branches in sm-graphql and sm-webapp to be merged at the same time.
Asana task: https://app.asana.com/0/597480600201879/662047100332398
dataset.is_publiccolumn and populates it from metadataupdate_es_index.py- if there is an error during the update, it will exit with bothsm-yinandsm-yangstill existing. This causes subsequent runs immediately fail because there is an assertion that only one index should exist at a time. The fix is to delete the broken index if there's an errorCaution:
ALTER TABLE dataset ADD COLUMN is_public boolean not null default(true);is_publicis non-nullable, this needs thefeature/private-prototypebranches in sm-graphql and sm-webapp to be merged at the same time.