[Bug]: Does not work with Snowflake and quoting true

[ghost]

Overview

When used in a project with snowflake as target and quoting:identifier:true it fails to load the models. Setting quoting:identifier:false on the model dbt_artifacts in dbt_project.yml does not work.

How to reproduce

Create a project with target on Snowflake and set the quoting to true on identifiers. Add the dbt_artifacts package and run it.

Expected behaviour

That it would work with quoting:true but it seems that there is inconsistent usa of case in the package code.

Environment

Core:

• installed: 1.5.0
  Plugins:
• snowflake: 1.5.0

# Add the results here

Please paste the contents of your packages.yml file here:

packages:
  - git: "https://github.com/dbt-labs/dbt-utils.git" # git URL
    revision: 1.1.0 # tag or branch name
  - package: brooklyn-data/dbt_artifacts
    version: 2.4.2

[QuentinCoviaux]

Hey, we are also facing the same issue due to the quoting being enabled in our project.

We found a temporary workaround though, until we can have proper support for this.

1. In dbt_project.yml, temporarily change quoting.identifier to false (to make the deploy easier, we did this in a local dev environment).
2. Create the dbt_artifacts tables/views in Production with the following command: dbt run -m dbt_artifacts --target prod . This will create the models without casing enforced.
3. In Production, add the upload_results() to the on-run-end and make sure the quoting.identifier is set to true.
   3.1. To note that upload_sources throws an error if you have some freshness test filters. I believe this used to be a problem and was fixed in Source freshness filters with string quotations break upload_sources macro in Snowflake (v1.0.0 beta prelease) #141 for non-quoted projects.
   In such cases, the output from the macro will be of sort: '{"warn_after": {"count": null, "period": null}, "error_after": {"count": 2, "period": "hour"}, "filter": "datediff(\'day\', \"EventLoadedAt\", current_timestamp()) < 2"}', which is creating a parsing error: 100069 (22P02): Error parsing JSON: missing comma, pos 124.
   To bypass this error, create a custom upload_results macro in your project that will overwrite the package macro and remove the upload_sources call - maybe one could find a more elegant solution but we didn't really need the sources for now since we're mostly interested in execution times.

[glsdown]

Hi @nc-ckb Thanks for raising this and sorry it's taken a while.

As noted in the dbt docs, due to the way snowflake handles the capitalisation of tables without quoting it was causing errors with the DML we were using. However I have a PR open #381 that should resolve this issue.