Copy-paste detector of the source code

What it does

I think it is a good idea to traverse the AST parsed from the source code and try looking for duplicates with some degree of similarity (or rather a full copy) from another place. I have recently had a PR made into one of my projects which had precisely the same copied from a hundred lines of code above, even with the doc comments. If I hadn't been reviewing it well and thoroughly, I'd have missed it. I think it is a good idea to automate this and have such checks done in CI, and I think it is possible to do. Besides all, there are other projects already doing almost that, for example, jscpd or cpd.

Lint Name

copy-paste

Category

correctness, suspicious, style

Advantage

• Remove the duplicate code either in parts or full code branches (functions, methods, impls, comments, parts of a function body).
• Increase the code quality and readability.
• Possibly allow for early refactoring and/or optimisations made by the copy-paste elimination.

Drawbacks

• Might be comparatively difficult or performance/time-wise costly to find duplicates.
• Full duplicates are rare but would be helpful to reduce; however, partial code duplication might be linted false-positively. In this case, it would require a lint suppression (if it was enabled for the project).

Example

struct A;
#[cfg(target_os = "linux")]
impl A {
    fn do(&self) {
        println!("Hello world!");
    }

    fn linux_specific_function(&self) {
        println!("Hello from Linux!");
    }
}

#[cfg(target_os = "windows")]
impl A {
    fn do(&self) {
        println!("Hello world!");
    }

    fn windows_specific_function(&self) {
        println!("Hello from Windows!");
    }
}

Could be written as:

struct A;

impl A {
    fn do(&self) {
        println!("Hello world!");
    }
}

#[cfg(target_os = "linux")]
impl A {
    fn linux_specific_function(&self) {
        println!("Hello from Linux!");
    }
}

#[cfg(target_os = "windows")]
impl A {
    fn windows_specific_function(&self) {
        println!("Hello from Windows!");
    }
}

No source files, tests, or entry points are named. Start by reviewing the proposed AST traversal, similarity detection, CI integration, and false-positive concerns, then inspect the existing Clippy lint structure. Done would require a defined scope and an accepted implementation plan for the copy-paste lint.