You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We once encountered a bug that went unnnoticed..where variable_id was missing in the catalog, so this is an error that could have only been captured after when open the catalog.
we need a simple regression test.
1- Have a "golden" standard catalog that comes out of the catalog builder right now. Name it appropriately to give it the version, tag or something to know which version of the code generated that for documentation purposes.
2- extend the create-gfdl-catalog CI workflow to include the following after a code change.
--- simply check the output json and csv with the golden standard in the github repo (ignoring the url to the csv).
this will quickly reveal if a column or value was left out , then gives a chance to assess if that was intended or not.
3- leverage pytest to do this and actions workflow to do this. Will need the addition of the golden standard catalogs, which can be accessed from github from a recent create-gfdl-catalog output (called manifests in github where i save the json and csv)
The text was updated successfully, but these errors were encountered:
We once encountered a bug that went unnnoticed..where variable_id was missing in the catalog, so this is an error that could have only been captured after when open the catalog.
we need a simple regression test.
1- Have a "golden" standard catalog that comes out of the catalog builder right now. Name it appropriately to give it the version, tag or something to know which version of the code generated that for documentation purposes.
2- extend the create-gfdl-catalog CI workflow to include the following after a code change.
--- simply check the output json and csv with the golden standard in the github repo (ignoring the url to the csv).
this will quickly reveal if a column or value was left out , then gives a chance to assess if that was intended or not.
3- leverage pytest to do this and actions workflow to do this. Will need the addition of the golden standard catalogs, which can be accessed from github from a recent create-gfdl-catalog output (called manifests in github where i save the json and csv)
The text was updated successfully, but these errors were encountered: