Determine naming convention for issues/violations/errors/checks
fpgmaas opened this issue · 3 comments
Question
Right now, we use a combination of the words issue, violation, check, error code. It would be good to determine a standard naming convention. My proposal would be to use "rule" to refer to DEP001, DEP002, etc. and "violation"(as already implemented in the code by @mkniewallner) to refer to specific violations in the code/configuration files.
Agree, we should be more consistent in our naming. I must say that we didn't really discuss if "violation" is what we want to go with when starting using that over the codebase, but I think that it goes well with the usage of "rule" (at least compared to using "issue"), so using "rule" and "violation" is something I would agree with.
If we go for 'rule' and 'violation', we need to make some small changes. For example, a 'missing dependency' is an issue, whereas 'Project should not contain missing dependencies' is a rule that can be violated. So e.g. in the docs, we'd replace this:
- Missing dependencies (DEP001)
- Unused dependencies (DEP002)
- Transitive dependencies (DEP003)
- Misplaced development dependencies (DEP004)
with
Code | Description | More information |
---|---|---|
DEP001 | Project should not contain missing dependencies | link |
DEP002 | Project should not contain unused dependencies | link |
DEP003 | Project should not contain unnecessary transitive dependencies | link |
DEP004 | Project should not use development dependencies in non-development code | link |
see here.
Going for "rule" and "violation".