URL | Source Branch | Target Branch | Workflow | Description |
---|---|---|---|---|
https://dataweek.de/ | main |
page-live |
build.yml |
Live |
https://staging.dataweek.de/ | develop |
page-staging |
preview-yml |
Staging/development preview for the Live page |
https://next.dataweek.de/ | next |
page-next |
New Layout development area | |
https://2022.dataweek.de/ | past/2022 |
page-2022 |
Archived page from 2022 | |
feature/newLayout |
Static branch of the new layout |
A source branch is built with an actions workflow and the result is pushed to the respective target branch.
The setup of the build workflow is as follows:
- Git
- Task
- Docker or Podman
- Optionally:
- Python (for
serve
resp.watch
andpictures:scale
resp.pictures:scale:preconditions
) - entr (for
watch
resp.build:watch
)
- Python (for
task build
task serve
Since jekyll only watches pages but not the layouts and data there is no automatic rebuild. You have to run task build serve
again, if you perform changes.
Additional requirement: entr
task watch
Serves the page and rebuilds it when changes happen in the directory.
To work with the data you need to clone https://github.com/AKSW/leipzig.dataweek.de-model/. This is done with
task sync-data
Use the restriction_file
feature (Jekyll RDF Documentation).
If you work on a single resource respectively the corresponding layout the build time might be to long for active development.
In this case the restriction:
in the _config.yml
file can be out-commented and replaced by a restriction_file: restriction_file.txt
setting. The restriction file would just contain the resource(s) you are working on, e.g.:
<https://2023.dataweek.de/2023-06-26/>