Overview › Swift / Obj-C Interop & Modernization

Swift / Obj-C Interop & Modernization

Swift-first new code; ~108 .m + 109 .h legacy files in the main app.

0 findings in this category · iOS

SCORE

Summary

The project is in mid-migration: 3,435 Swift files (most under Costco-Digital/ SPM packages) vs 108 Obj-C .m + 109 .h files (concentrated in legacy Costco/Costco/ flows). The new modules are Swift-first; the older areas remain Obj-C heavy.

Findings

HIGH

108 Obj-C source files

Each Obj-C file imposes a maintenance tax: NS-prefixed types, manual nullability annotations, separate header files, slower compile times in mixed-language modules.

Recommendation: Identify Obj-C hot files (most-changed in last 90 days) and convert quarterly. Track LoC reduction.

MEDIUM

Bridging headers + module maps

Confirm bridging headers correctly annotate Obj-C with NS_ASSUME_NONNULL_BEGIN/END blocks; otherwise Swift sees imported APIs as platform types (SomeType!) which silently bypass null safety.

Recommendation: Audit every public Obj-C header for nullable/nonnull annotations.

MEDIUM

Custom RZ-prefixed Pods (Raizlabs legacy)

RZVinyl (networking + caching) and RZTransitions are Raizlabs-era forks vendored via custom git URLs. They still ship Obj-C; replacement reduces both the Obj-C surface and the unmaintained-fork risk.

Recommendation: Replace RZVinyl with URLSession + Codable patterns already used in CoreImplementation; replace RZTransitions with native UIKit / SwiftUI transitions.

LOW

Swift 5 — adopt Swift 6 strict concurrency when ready

Swift 6 (Xcode 16+) ships strict concurrency checking which surfaces data-race bugs at compile time.

Recommendation: Pilot SWIFT_STRICT_CONCURRENCY=complete on one SPM package; track warning count over time.

No findings logged in this category yet for iOS.

Costco iOS · Code Review Report · Generated 2026-05-07 · 88 machine-curated findings