Adds rattler-build recipe compatibility (and adds jolt-physics) #27243
Conversation
|
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( I do have some suggestions for making it better though... For recipes/example-new-recipe/recipe.yaml: This is a rattler-build recipe and not yet lintable. We are working on it! For recipes/jolt-physics/recipe.yaml: This is a rattler-build recipe and not yet lintable. We are working on it! |
|
Hi! Thanks for your contribution to conda-forge. If these changes are intentional (and you aren't submitting a recipe), please attach a |
|
Im unsure what is wrong with the WIndows build? There doesnt seem to be an actual error? |
There was a problem hiding this comment.
This PR does too many things IMO.
- format
build_all.py - add rattler-build support
- add a new rattler example recipe
- add a genuinely new recipe
Proving the functionality would already with actually building the example recipe (perhaps with some additional safeguard that we don't create a feedstock), and then a follow-up can add jolt.
W.r.t. to the error on windows, it seems to be failing while trying to call inspect_artifacts.
Inspecting artifacts
"inspect_artifacts needs conda-forge-ci-setup >=4.6.0"
##[error]Cmd.exe exited with code '1'.
|
Hi! Thanks for your contribution to conda-forge. If these changes are intentional (and you aren't submitting a recipe), please attach a |
|
Thanks for checking out this PR @h-vetinari Note that this was already reviewed in #27008 by both @jaimegp and @wolfv but happy to make more changes. The problem with the current setup is that there is no easy way to test this. The creation of the feedstock is performed by another repository so without merging this that code won't run and that code won't actually create a feedstock if there is no recipe, therefor we must add a recipe. The The example recipe is there as an example similar to the existing example recipe. I would be happy to add it in another PR but in the previous round of reviews, it was mentioned that this was missing from the PR. I undid the formatting. |
| rmtree(os.path.join(recipes_dir, example_recipe), ignore_errors=True) | ||
|
|
||
| # Determine the locations for the variant config files. | ||
| specs = OrderedDict() |
There was a problem hiding this comment.
This is less a question for this PR and more a topic overall across our infrastructure, but at least I thought I'd note it here: conda-forge/conda-smithy#2033
There was a problem hiding this comment.
| specs = OrderedDict() | |
| specs = {} |
Yeah, fair enough, I don't want to complicate this process for you any further. Jaime would actually be a great person to ask for figuring out what (if anything) is going wrong with the artefact inspection. ;-) |
Co-authored-by: h-vetinari <[email protected]>
|
Hi! Thanks for your contribution to conda-forge. If these changes are intentional (and you aren't submitting a recipe), please attach a |
|
Hi! Thanks for your contribution to conda-forge. If these changes are intentional (and you aren't submitting a recipe), please attach a |
|
The issue with |
|
YOLO 🚀 |
| - python | ||
|
|
||
| tests: | ||
| # Some packages might need a `test/commands` key to check CLI. |
There was a problem hiding this comment.
this would now be, presumably, tests/0/script/. tests/0/files/source was also surprising: the glob needs e.g. tests/**/*, not just tests now.
There was a problem hiding this comment.
True, that comment looks wrong. It is also script not command in the test section
Checklist
url) rather than a repo (e.g.git_url) is used in your recipe (see here for more details).This PR adds support for the new recipe format and adds the first such recipe for https://github.com/jrouwe/JoltPhysics
We already did a previous attempt with #27008 which we reverted because of some issues with conda-smithy. A new release is available which should fix this issue.