It would be good to have a dockerized container that allows executing starlark runtime locally.
This would be useful for local development as well as running other projects such as argo workflows.
Here's an example of how it could be used in argo
script:
image: starlarky:1.0
command:
- python
source: |
load("@stdlib//re2j", "re2j")
def process(input, ctx):
# do your code here
I guess the concept here is that we'd provide an entrypoint or cmd which takes a script as an input and then allows it to be invoked by passing data via stdin/stdout.
I'm not sure exactly how the api would work but it would be useful for executing starlarky anywhere.
It would be good to have a dockerized container that allows executing starlark runtime locally.
This would be useful for local development as well as running other projects such as argo workflows.
Here's an example of how it could be used in argo
I guess the concept here is that we'd provide an entrypoint or cmd which takes a script as an input and then allows it to be invoked by passing data via stdin/stdout.
I'm not sure exactly how the api would work but it would be useful for executing starlarky anywhere.