Allow for build command artifacts to populate by adjusting filter for data:args:which = 'run'

[mrevelo-clickup]

I recently implemented this package and it has been working great; however, we recently switched away from dbt run to dbt build. In doing so, I noticed that artifacts were no longer being parsed out into the fact and dimension models during a dbt_artifacts run despite being in the dbt_artifacts table.

Upon investigation, the model execution and run results models have a filter to limit the artifacts to those which are from run commands only, and not builds. I propose the filter be adjusted to data:args:which in ('run', 'build') for these models (or at least do the reverse and filter out non-run commands for future scalability):

• stg_dbt__model_executions.sql
• stg_dbt__run_results
• stg_dbt__run_results_env_keys

Unless there is a reason in particular while builds are not included - please let me know. Thanks!

[NiallRees]

Hi @mrevelo-clickup, thanks for this write-up and the solution proposed. It sounds great to me :)

[mrevelo-clickup]

Only other concern is that the models may need to be re-architected around the build command since it packages together tests, snapshots, etc. Unless they are run as their own commands, I don't think they will get parsed out into their relevant test and snapshot models. Will push through the easy change to include build for now to get over the hump.

[alanmcruickshank]

@NiallRees I think this can be closed given #85 should fix the same issue.