Add checksumtype to the golangci-lint file. #111
+14
−50
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
gregory-nisbet
force-pushed
the
2025-10-05-golangci-lint-additions
branch
from
bcc14bb to
e0d4781
Compare
philhassey
requested changes
Oct 20, 2025
Collaborator
philhassey
left a comment
philhassey
left a comment
There was a problem hiding this comment.
I think there's a typo to fix.
Sorry for the delay on reviewing this, there was some trouble with notifications.
gregory-nisbet
force-pushed
the
2025-10-05-golangci-lint-additions
branch
from
4de67ff to
e8315d1
Compare
philhassey
reviewed
Nov 3, 2025
gregory-nisbet
force-pushed
the
2025-10-05-golangci-lint-additions
branch
from
e8315d1 to
11d1e87
Compare
Some of the types within cedar-go match the sum type idiom (an interface with an unexported method) described in the linter below. Let's add an explicit check for exhaustiveness for this specific linter. https://github.com/alecthomas/go-check-sumtype *Issue #, if available:* IDK what issue tracker you use. *Description of changes:* Add a sum type annotation, add two branches to convertNamespace, both of which are just comments. Let me know if the behavior in those branches should be a little smarter. Signed-off-by: Greg NISBET <[email protected]>
Some nonce types have an embedded Node member and are in the same package, so they appear to the sum type checker to be real variants based on the heuristics that that package uses. There are a few ways to handle this: 1) make the unknownNode and unknownType cases real variants 2) remove the tests that use unknownNode and unknownType 3) rewrite the tests that use unknownNode and unknownType to use a different strategy to detect the same defects. Signed-off-by: Greg NISBET <[email protected]>
Remove unknownType and unknownNode and the two tests that depend on them. Signed-off-by: Greg NISBET <[email protected]>
Signed-off-by: Greg NISBET <[email protected]>
gregory-nisbet
force-pushed
the
2025-10-05-golangci-lint-additions
branch
from
ffdcd12 to
685b2da
Compare
Contributor
Author
|
Rev'd PR. I think this will be the most controversial change (in |
philhassey
requested changes
Nov 17, 2025
| at.ContextRecord = t | ||
| case *Path: | ||
| at.ContextPath = t | ||
| case *SetType: |
Collaborator
There was a problem hiding this comment.
Can you explain why this needed to be added, etc? I see there is another case of panic in this file, which I'm not really excited about either ... Is there a way we can rephrase both to avoid those panics? I'd rather we use error handling in bad cases.
Either way, the PR can't be merged without 100% coverage, so this does need to be covered with a test to show when it happens.
gregory-nisbet
added a commit
to gregory-nisbet/cedar-go
that referenced
this pull request
Nov 22, 2025
gochecksumtype is a linter in golangci-lint that checks for switch statement exhaustiveness for types that look like sum types in other languages. More precisely, interface types with a //sumtype:decl annotation and an unexported nullary marker method are statically guaranteed to be closed (since new types that implement the interface can't be created). This means that we can check for exhaustiveness on type switches for values implementing the sumdecl-compatible interface, which is what the linter does. The initial attempt to add this to every type with a marker interface cedar-policy#111 has grown in complexity since a few of the tests use unknownNode and unknownType types for checking the error paths. gochecksumtype has no way of indicating test-only variants of a sum type. So, to keep the diff size small and the code coverage at 100% (except for the generated parser) at each step, I'm going to break PR cedar-policy#111 into pieces and land them separately. Signed-off-by: Greg NISBET <[email protected]>
Contributor
Author
|
I want to split this PR. Here's the first piece: #121 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Some of the types within cedar-go match the sum type idiom (an interface with an unexported method) described in the linter below.
Let's add an explicit check for exhaustiveness for this specific linter.
https://github.com/alecthomas/go-check-sumtype
Issue #, if available:
n/a
Description of changes:
Add a sum type annotation, add two branches to convertNamespace, both of which are just comments. Let me know if the behavior in those branches should be a little smarter.