sql-template-tag VS dbt-core

You can avoid this entirely with JavaScript's tagged template literals. Here is an example library: https://github.com/blakeembrey/sql-template-tag

If you prefer SQL and you use postgres you can check out slonik. If you want to keep using knex you can use https://github.com/blakeembrey/sql-template-tag. They will add placeholders for all the values and expose helper for query building via strings (e.g. a special join). At the same time they are typesafe.

it's fun for things you're ok with throwing away.

For example, I wanted a dbt[0] like tool, but written in rust, specifically focused on duckdb. Claude Code knocked it out[1] it without much guidance.

Also added support for all duckdb output options (e.g. write to a partitioned parquet instead of a table).

0 - SQL transformation tool (https://github.com/dbt-labs/dbt-core)

1 - https://github.com/definite-app/crabwalk

Just noticed 404s on dbt core - was this announced recently?

https://github.com/dbt-labs/dbt-core

dbt integration was one of our major goals early on but we found that the interaction wasn't as straightforward as had hoped.

There is an open PR in the dbt repo: https://github.com/dbt-labs/dbt-core/pull/5982#issuecomment-...

I have some ideas about future directions in this space where I believe PRQL could really shine. I will only be able to write those down in a couple of hours. I think this could be a really exciting direction for the project to grow into if anyone would like to collaborate and contribute!

I really dislike SQL, but recognize its importance for many organizations. I also understand that SQL is definitely testable, particularly if managed by environments such as DBT (https://github.com/dbt-labs/dbt-core). Those who arrived here with preference to python will note that dbt is largely implemented in python, adds Jinja macros and iterative forms to SQL, and adds code testing capabilities.

3). Data Build Tool Repository.

dbt Cloud rightfully gets a lot of credit for creating dbt Core and for being the first managed dbt Core platform, but there are several entrants in the market; from those who just run dbt jobs like Fivetran to platforms that offer more like EL + T like Mozart Data and Datacoves which also has hosted VS Code editor for dbt development and Airflow.

Ahh I misunderstood, yes column level lineage is useful. DBT prefers leveraging macros which sort of breaks this pattern. I think the DBT way would be to better separate fields into upstream models and use table tracking https://github.com/dbt-labs/dbt-core/discussions/4458