False-positive "enum variant added on exhaustive enum"

[adamchalmers]

Which lint or lints are the issue

enum_marked_non_exhaustive

Known issues that might be causing this

• Is the flagged item defined in another crate, or a re-export of such an item? (#638)

Steps to reproduce the bug with the above code

git clone and checkout out this PR: KittyCAD/modeling-api#234
Run cargo semver-checks check-release -p kittycad-modeling-cmds

Actual Behaviour

I get two semver check failures:
--- failure enum_marked_non_exhaustive: enum marked #[non_exhaustive] ---
--- failure enum_variant_added: enum variant added on exhaustive enum ---

Expected Behaviour

I should not get any semver failures, as:

1. The enum was marked #[non_exhaustive] many commits ago (back in the 0.1.x of the crate)
2. The enum is only conditionally marked #[non_exhaustive], if the unstable_exhaustive crate feature is not used. This feature should be used -- it has the prefix "unstable_" and it's also not a default feature. So, by default the enum should be marked as non_exhaustive.
3. Because the enum is marked N.E. there should be no warning about enum variant added on exhaustive enum.

Verbose Lint Output

 Parsing kittycad-modeling-cmds v0.2.0 (current)
  Parsed [   2.130s] (current)
 Parsing kittycad-modeling-cmds v0.2.0 (baseline, cached)
  Parsed [   0.089s] (baseline)
Checking kittycad-modeling-cmds v0.2.0 -> v0.2.0 (no change)
Starting 67 checks, 0 unnecessary on 10 threads
    PASS [   0.052s]       major        auto_trait_impl_removed
    PASS [   0.017s]       major        constructible_struct_adds_field
    PASS [   0.003s]       major        constructible_struct_adds_private_field
    PASS [   0.002s]       major        constructible_struct_changed_type
    PASS [   0.059s]       major        derive_trait_impl_removed
    FAIL [   0.002s]       major        enum_marked_non_exhaustive
    PASS [   0.002s]       major        enum_missing
    PASS [   0.002s]       minor        enum_must_use_added
    PASS [   0.003s]       major        enum_now_doc_hidden
    PASS [   0.003s]       major        enum_repr_int_changed
    PASS [   0.005s]       major        enum_repr_int_removed
    PASS [   0.002s]       major        enum_repr_transparent_removed
    PASS [   0.004s]       major        enum_struct_variant_field_added
    PASS [   0.004s]       major        enum_struct_variant_field_missing
    PASS [   0.016s]       major        enum_struct_variant_field_now_doc_hidden
    PASS [   0.013s]       major        enum_tuple_variant_field_added
    PASS [   0.012s]       major        enum_tuple_variant_field_missing
    PASS [   0.010s]       major        enum_tuple_variant_field_now_doc_hidden
    FAIL [   0.007s]       major        enum_variant_added
    PASS [   0.014s]       major        enum_variant_missing
    PASS [   0.073s]       major        function_abi_no_longer_unwind
    PASS [   0.009s]       major        function_changed_abi
    PASS [   0.002s]       major        function_changed_export_name
    PASS [   0.002s]       major        function_const_removed
    PASS [   0.003s]       major        function_missing
    PASS [   0.004s]       minor        function_must_use_added
    PASS [   0.002s]       major        function_now_doc_hidden
    PASS [   0.002s]       major        function_parameter_count_changed
    PASS [   0.003s]       major        function_unsafe_added
    PASS [   0.003s]       major        inherent_method_const_removed
    PASS [   0.003s]       major        inherent_method_missing
    PASS [   0.004s]       minor        inherent_method_must_use_added
    PASS [   0.007s]       major        inherent_method_unsafe_added
    PASS [   0.007s]       major        method_parameter_count_changed
    PASS [   0.002s]       major        module_missing
    PASS [   0.003s]       major        pub_module_level_const_missing
    PASS [   0.001s]       major        pub_module_level_const_now_doc_hidden
    PASS [   0.001s]       major        pub_static_missing
    PASS [   0.002s]       major        repr_c_removed
    PASS [   0.001s]       major        repr_packed_added
    PASS [   0.002s]       major        repr_packed_removed
    PASS [   0.015s]       major        sized_impl_removed
    PASS [   0.002s]       major        struct_marked_non_exhaustive
    PASS [   0.007s]       major        struct_missing
    PASS [   0.003s]       minor        struct_must_use_added
    PASS [   0.013s]       major        struct_now_doc_hidden
    PASS [   0.018s]       major        struct_pub_field_missing
    PASS [   0.013s]       major        struct_pub_field_now_doc_hidden
    PASS [   0.002s]       major        struct_repr_transparent_removed
    PASS [   0.007s]       major        struct_with_pub_fields_changed_type
    PASS [   0.002s]       major        trait_method_missing
    PASS [   0.002s]       major        trait_method_now_doc_hidden
    PASS [   0.001s]       major        trait_method_unsafe_added
    PASS [   0.002s]       major        trait_method_unsafe_removed
    PASS [   0.001s]       major        trait_missing
    PASS [   0.002s]       minor        trait_must_use_added
    PASS [   0.001s]       major        trait_now_doc_hidden
    PASS [   0.001s]       major        trait_removed_associated_constant
    PASS [   0.001s]       major        trait_removed_associated_type
    PASS [   0.001s]       major        trait_removed_supertrait
    PASS [   0.001s]       major        trait_unsafe_added
    PASS [   0.001s]       major        trait_unsafe_removed
    PASS [   0.001s]       major        tuple_struct_to_plain_struct
    PASS [   0.003s]       minor        type_marked_deprecated
    PASS [   0.002s]       major        union_now_doc_hidden
    PASS [   0.001s]       major        unit_struct_changed_kind
    PASS [   0.020s]       major        variant_marked_non_exhaustive
 Checked [   0.084s] 67 checks; 65 passed, 2 failed, 0 unnecessary

--- failure enum_marked_non_exhaustive: enum marked #[non_exhaustive] ---

Description:
A public enum has been marked #[non_exhaustive]. Pattern-matching on it outside of its crate must now include a wildcard pattern like _, or it will fail to compile.
ref: https://doc.rust-lang.org/cargo/reference/semver.html#attr-adding-non-exhaustive
impl: https://github.com/obi1kenobi/cargo-semver-checks/tree/v0.30.0/src/lints/enum_marked_non_exhaustive.ron

Failed in:
enum ModelingCmd in /Users/adamchalmers/kc-repos/modeling-api/modeling-cmds/src/def_enum.rs:9

--- failure enum_variant_added: enum variant added on exhaustive enum ---

Description:
A publicly-visible enum without #[non_exhaustive] has a new variant.
ref: https://doc.rust-lang.org/cargo/reference/semver.html#enum-variant-new
impl: https://github.com/obi1kenobi/cargo-semver-checks/tree/v0.30.0/src/lints/enum_variant_added.ron

Failed in:
variant ModelingCmdEndpoint:TestDoNotMerge in /Users/adamchalmers/kc-repos/modeling-api/modeling-cmds/src/def_enum.rs:810
Summary semver requires new major version: 2 major and 0 minor checks failed
Finished [ 2.387s] kittycad-modeling-cmds

Generated System Information

Software version

cargo-semver-checks 0.30.0

Operating system

macOS 13.4.1 (Darwin 22.5.0)

Command-line

/Users/adamchalmers/.cargo/bin/cargo-semver-checks semver-checks --bugreport 

cargo version

> cargo -V 
cargo 1.76.0 (c84b36747 2024-01-18)

Compile time information

• Profile: release
• Target triple: aarch64-apple-darwin
• Family: unix
• OS: macos
• Architecture: aarch64
• Pointer width: 64
• Endian: little
• CPU features: aes,crc,dit,dotprod,dpb,dpb2,fcma,fhm,flagm,fp16,frintts,jsconv,lor,lse,neon,paca,pacg,pan,pmuv3,ras,rcpc,rcpc2,rdm,sb,sha2,sha3,ssbs,vh
• Host: aarch64-apple-darwin

Build Configuration

None

Additional Context

I want to mark the enum as #[non_exhaustive] for most consumers, but for my own internal projects, I am OK opting into breaking my code every time I add a new enum variant.

[adamchalmers]

OK I figured out one of the lints is my fault, still think the other is legit but will open a new issue

[obi1kenobi]

I'm having trouble reproducing this:

$ git checkout achalmers/test-semver
branch 'achalmers/test-semver' set up to track 'origin/achalmers/test-semver'.
Switched to a new branch 'achalmers/test-semver'

$ cargo semver-checks -p kittycad-modeling-cmds
     Parsing kittycad-modeling-cmds v0.2.2 (current)
      Parsed [   4.063s] (current)
     Parsing kittycad-modeling-cmds v0.2.2 (baseline, cached)
      Parsed [   0.145s] (baseline)
    Checking kittycad-modeling-cmds v0.2.2 -> v0.2.2 (no change)
     Checked [   0.197s] 67 checks; 67 passed, 0 unnecessary
    Finished [   4.642s] kittycad-modeling-cmds

Though I do notice that in my checkout, the library seems to be at v0.2.2 and in the issue's repro that was at v0.2.0. I re-ran the branch's code against v0.2.0 as well (cargo semver-checks -p kittycad-modeling-cmds --baseline-version 0.2.0), and that didn't flag any issues either.

The repro branch was force-pushed a few times, so maybe something about it changed that broke the repro? Or maybe there's some non-standard configuration somewhere in your local checkout, or in your CI? If the latter, I'd love to get to the bottom of that so I can make cargo-semver-checks robust against that.

Regarding the unstable feature, it won't be enabled by default during semver checking. By default we use a heuristic that enables more than just the default features, but de-selects features that are unstable or nightly, and the unstable_exhaustive feature name would be excluded by default. More info on the heuristic here: https://github.com/obi1kenobi/cargo-semver-checks#what-features-does-cargo-semver-checks-enable-in-the-tested-crates

So I agree that your expectation of what the tool should do here seems accurate based on the info you gave here. As far as I can tell at the moment, that seems to be happening. If you still feel there's an issue, I'd love to work it out with you.