The official resource for SQLFluff related GitHub Actions and Workflows.
Workflows are listed below by contributing team with a brief description. To learn more and how to implement each, click the links below.
/models
conda
to setup a virtual environment and manage python
, dbt
, and sqlfluff
dependencies.templater = dbt
- this requires a dummy profiles.yml
and the ability to connect to your data warehouse from inside the workflow.
dbt/models
directory)/lint
(can be adjusted) in itrequirements.txt
based dependency managment rather than conda
.--write-output
option.GitHub Actions is a feature within GitHub. It allows you to...
Automate, customize, and execute your software development workflows right in your repository with GitHub Actions. You can discover, create, and share actions to perform any job you'd like, including CI/CD, and combine actions in a completely customized workflow.
These workflows, defined by YAML files placed in the .github/workflows/
directory of your repo, are event-driven, meaning that you can run a workflow after a specified event has occurred (e.g., opening a Pull request to the main
branch).
To add a layer of confusion, the GitHub Marketplace has individual "Actions" (e.g., the checkout action) that you can use as steps inside your workflow(s). Using Actions can reduce the code you need to write to define your worflow and accomplish your goal.
Conceptually, you can connect these poorly named pieces like so:
Often you will hear us (and others, including GitHub) use the term "GitHub Action" and "Workflow" interchangeably. If you simply remember that the YAML files are defining a list of steps to be executed when a certain event happens in your repo, you are well on your way to understanding and using the power of workflows in GitHub Actions.
See the Github docs to learn more about GitHub Actions and developing your own worflows.