- The docstring of the ObjStorage implementations have been updated to reference the one present in the base class.
- Make the base API ObjStorage an Abstract Base Class for python. This add more security when a new ObjStorage implementation is created.
- Also, the filtering object storage now use this feature to make its extension easier.
- Remove duplicate test for the PathSlicingObjectStorage.
- Factorize hash computation into objstorage
- Rename an argument that didn't follow the API
Details
Details
- Reviewers
olasd zack - Group Reviewers
Reviewers - Commits
- rDOBJSe3aa5c0a182b: Rename an argument that didn't follow the API
rDOBJS742b86e0b40c: Remove docstring in subclasses
rDOBJS5f2bf871ebe5: Factorize hash computation into objstorage
rDOBJS8de255fb5f54: Use the abc metaclass for some abstract classes
rDOBJSc27165fb372c: Remove duplicate test for the PathSlicingObjectStorage.
rDOBJSd53c6b90f34c: Reference base class in implementations docstrings
Diff Detail
Diff Detail
- Repository
- rDOBJS Object storage
- Branch
- D95-2
- Lint
No Linters Available - Unit
No Unit Test Coverage - Build Status
Buildable 363 Build 536: arc lint + arc unit
Event Timeline
Comment Actions
- Factorize hash computation into objstorage
- Rename an argument that didn't follow the API
swh/objstorage/objstorage.py | ||
---|---|---|
17–21 | If we really don't have a helper like this one into hashutil, we probably should add one there, and use it here. |
swh/objstorage/objstorage.py | ||
---|---|---|
17–21 | That's swh.core.hashutil.hashdata. |
swh/objstorage/objstorage.py | ||
---|---|---|
17–21 | swh.core.hashutil.hashdata computes the hashcode for each algorithm (sha1, sha256, sha1_git). |
swh/objstorage/objstorage.py | ||
---|---|---|
17–21 | You can control which algos hashdata uses with an argument, that should be enough? |
swh/objstorage/objstorage.py | ||
---|---|---|
17–21 | Didn't see the default arg; that's exactly what's needed! |