Skip to content

rules.md

Latest commit

 

History

History
41 lines (30 loc) · 1.38 KB

rules.md

File metadata and controls

41 lines (30 loc) · 1.38 KB
Raw

Rules Reference

This page summarizes the rule families used by the analyzer. The exact rule list evolves over time, but the categories below are stable.

Rule families

  • Naming quality: terrible names, abbreviations, single-letter variables
  • Structure quality: deep nesting, long functions, god functions
  • Maintainability: duplication, commented-out code, stale TODO comments
  • Debug leftovers: println and similar print-based debugging
  • Language-specific hygiene: Go, Python, Rust, Java, C, C++, Ruby, Swift, and Zig rules

Commonly reported issues

  • magic-number
  • deep-nesting
  • long-function
  • god-function
  • println-debugging
  • commented-code
  • todo-comment
  • duplication

Language coverage

  • Rust, Go, Python, JavaScript, TypeScript, Java, C, C++, Ruby, Swift, and Zig are supported
  • Tree-sitter grammars are available for the full supported set used by source parsing
  • File discovery recognizes the corresponding extensions listed in src/language/mod.rs

Tuning tips

  • Use .garbage-code-hunter.toml to whitelist accepted names and numbers
  • Use exclude-patterns for generated code
  • Use directories to reduce sensitivity in noisy folders
  • Use CLI --exclude for one-off skips

Notes

  • Rule names and severities are designed to be humorous rather than authoritative
  • The tool does not claim to detect bugs, security issues, or performance regressions