added support to parse labels in dockerfile #3987
Conversation
VarshaUN
force-pushed
the
support-OCI-labels
branch
from
edd1a8d to
c90069a
Compare
There was a problem hiding this comment.
@VarshaUN Thanks++
Looking good, need another round of changes and a lot more tests here for the added functionality, and this should be much better.
src/packagedcode/dockerfileLABEL.py
Outdated
|
|
||
| import io | ||
| from pathlib import Path | ||
| from dockerfile_parse import DockerfileParser |
There was a problem hiding this comment.
https://pypi.org/project/dockerfile-parse/ is nice to use, also note that we use this at container inspector https://github.com/aboutcode-org/container-inspector/blob/main/src/container_inspector/dockerfile.py#L14C8-L14C24 as we parse dockerfiles there too. container-inspector is also a scancode dependency, so we have to think about whether to move this code there. But overall I think this is fine here, as packagedata creation/datafile handlers should be in scancode.
I don't see this used anywhere else in scancode-toolkit, we have this in the requirements at https://github.com/aboutcode-org/scancode-toolkit/blob/develop/requirements.txt#L18 to pin the version when we import container-inspector. But we need to add dockerfile-parse here: https://github.com/aboutcode-org/scancode-toolkit/blob/develop/setup.cfg#L73 as this is not constrained at https://github.com/aboutcode-org/container-inspector/blob/main/setup.cfg#L62
src/packagedcode/dockerfileLABEL.py
Outdated
| class DockerfileHandler(models.DatafileHandler): | ||
| datasource_id = 'dockerfile' | ||
| default_package_type = 'docker-image' | ||
| path_patterns = ('Dockerfile', 'containerfile', '*.dockerfile') |
There was a problem hiding this comment.
We need to test the is_datafile() here with examples from here:
- https://github.com/search?q=org.opencontainers.image+path%3A**%2FDockerfile&type=code&ref=advsearch
- https://github.com/search?q=org.opencontainers.image+path%3A**%2F*containerfile*&type=code&ref=advsearch
- https://github.com/search?q=org.opencontainers.image+path%3A**%2F*.dockerfile&type=code&ref=advsearch
Also have these as tests for parsers, choosing from more popular repos/more diverse examples of usage.
See for example other tests like this: https://github.com/aboutcode-org/scancode-toolkit/blob/develop/tests/packagedcode/test_npm.py#L445
Signed-off-by: Varsha U N <[email protected]>
Co-authored-by: Ayan Sinha Mahapatra <[email protected]> Signed-off-by: Varsha U N <[email protected]>
Signed-off-by: Varsha U N <[email protected]>
Co-authored-by: Ayan Sinha Mahapatra <[email protected]> Signed-off-by: Varsha U N <[email protected]>
Signed-off-by: Varsha U N <[email protected]>
Signed-off-by: Varsha U N <[email protected]>
Signed-off-by: Varsha U N <[email protected]>
Signed-off-by: Varsha U N <[email protected]>
VarshaUN
force-pushed
the
support-OCI-labels
branch
from
d60d0c8 to
328639b
Compare
Fixes #3561
Tasks
Signed-off-by: Varsha U N [email protected]