Rust: Implement type inference for ref expression as type equality

[paldepind]

I was debugging an issue where types where not inferred correctly, and realized that for a call like f(&a) the declared type of fs first parameters would not flow to a through the borrow.

This PR fixes that and also removes some logic for ref expressions that doesn't match Rust's type system. Due to implicit dereferencing &&&&a and &a can seem interchangeable, but the types are distinct. Removing this collapsing doesn't seem to affect the tests.

[Copilot]

Pull Request Overview

This PR enhances Rust expression type inference by treating &x as a true type equality and removes legacy reference-handling logic. Key changes include:

• Added new test cases in main.rs to verify boolean and custom-flag type inference through multiple borrows.
• Extended the typeEquality predicate to include RefExpr and simplified inferRefExprType in QL.
• Removed outdated logic for reference-expression inference in TypeInference.qll.

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File	Description
rust/ql/test/library-tests/type-inference/main.rs	Added MyFlag struct tests and **&&true inference
rust/ql/lib/codeql/rust/internal/TypeInference.qll	Updated typeEquality, simplified inferRefExprType and inference pipeline

Comments suppressed due to low confidence (2)

rust/ql/test/library-tests/type-inference/main.rs:1158

• [nitpick] The field name bool shadows the built-in type name and may be confusing; consider renaming it to something more descriptive like flag or state.

    struct MyFlag {

rust/ql/lib/codeql/rust/internal/TypeInference.qll:980

• [nitpick] Update this doc comment to mention that inferRefExprType no longer takes a path parameter and now only returns TRefType for the given RefExpr.

/** Gets the root type of the reference expression `re`. */

[hvitved]

LGTM. Let's see what DCA says before merging.

[paldepind]

DCA seems fine. Overall a reduction in missing call targets.