Architecture (Quick Overview)

中文

Layers

• crates/core: pure Rust core logic (scan files, normalize, compute fingerprints, produce results)
• crates/cli: Rust CLI binary (entrypoint, argument parsing, output formatting, exit codes)

Data flow (CLI → results)

1. dup-code-check parses CLI args → builds roots + options
2. Rust core:
   • collects file paths (respects .gitignore by default; skips common dependency/output dirs)
   • reads file contents; skips huge/binary files
   • runs detectors and returns results (optionally with ScanStats)
3. CLI formats output (text or JSON) and decides exit code based on --strict / stats

Key abstractions (conceptual)

• roots: multiple scan roots (cross-repo / cross-directory)
• ScanOptions: ignore rules, budgets, thresholds, output limits
• ScanStats: scan statistics and “scan completeness” signal for CI/strict mode

Related docs:

• Options: Scan Options
• Detectors: Detectors & Algorithms

Local build (via npm)

npm run build:

• cargo build --release -p dup-code-check
• copies the binary to bin/dup-code-check

Where to look in code

• CLI: crates/cli/src/main.rs
  • arg parsing: parse_args()
  • exit code policy: --strict + ScanStats
• Core: crates/core/src/lib.rs
  • file collection: collect_repo_files*
  • duplicate files: find_duplicate_files*
  • code spans: find_duplicate_code_spans*
  • report: generate_duplication_report*

Extensibility

Suggested path to add a new detector:

1. add core logic under crates/core (preferably with unit tests)
2. expose it via CLI (flags/output) under crates/cli when needed
3. update docs (docs/) and CHANGELOG.md