Repository navigation
[duplicate-code] Duplicate Code: YAML block-walking state machine duplicated across codemod files #69116
Description
🔍 Duplicate Code Detected: Codemod YAML Block-Walking State Machine
Analysis of commit 5bf0b44 (github/gh-aw)
Assignee: @copilot
Summary
The pkg/cli/codemod_*.go family re-implements the same line-by-line YAML block-walking state machine. Each codemod iterates over lines, tracks inTools / nested-block booleans, detects block entry via strings.HasPrefix(trimmed, "...:"), captures indentation with the shared getIndentation, and resets state via the shared hasExitedBlock. The helpers are already factored into pkg/cli/yaml_frontmatter_utils.go, but the ~40-50 line control-flow scaffold around them is copy-pasted.
Duplication Details
Pattern: line-walk + block-state scaffold
- Severity: High
- Occurrences: 9+ files use the
inTools-style walk; 30 references tohasExitedBlockacross codemods - Representative Locations:
pkg/cli/codemod_github_repos.go(lines 62-110,renameGitHubReposToAllowedRepos)pkg/cli/codemod_toolset_singular.go(lines ~63-110)pkg/cli/codemod_allowed_repos_current.go(lines ~68-115)pkg/cli/codemod_github_app.go(lines ~99-139)pkg/cli/codemod_checkout_persist_credentials_false.go/codemod_steps_run_secrets_env.go(~25-line blocks repeated)pkg/cli/codemod_mcp_network.go/codemod_playwright_domains.go
- Code Sample:
for i, line := range lines { trimmed := strings.TrimSpace(line) if trimmed == "" { result = append(result, line); continue } if !strings.HasPrefix(trimmed, "#") { if inToolsGithub && hasExitedBlock(line, toolsGithubIndent) { inToolsGithub = false } if inTools && hasExitedBlock(line, toolsIndent) { inTools = false; inToolsGithub = false } } if strings.HasPrefix(trimmed, "tools:") { inTools = true; toolsIndent = getIndentation(line); ... } }
Impact Analysis
- Maintainability: A fix to block-exit/indentation edge cases must be applied in ~9 places.
- Bug Risk: High risk of inconsistent handling of comments/blank lines between codemods.
- Code Bloat: Hundreds of near-identical lines in
pkg/cli.
Refactoring Recommendations
- Extract a reusable block walker into
pkg/cli/yaml_frontmatter_utils.go, e.g.walkYAMLBlock(lines []string, path []string, onLine func(ctx blockCtx, line string) (string, bool)) ([]string, bool)that drives the enter/exit state machine and lets each codemod supply only the per-line transform.- Estimated effort: medium (4-8h)
- Benefits: single source of truth for indentation/exit semantics; each codemod shrinks to its unique rewrite logic.
- Add table-driven tests for the extracted walker covering comments, blank lines, and nested exits.
Implementation Checklist
- Review duplication findings
- Design shared walker signature
- Migrate codemods incrementally
- Update/centralize tests
- Verify
make recompile+ unit tests pass
Analysis Metadata
- Analyzed Files: pkg/cli/codemod_*.go (72 files)
- Detection Method: Serena activation + normalized sliding-window block hashing
- Commit: 5bf0b44
Generated by 🔍 Duplicate Code Detector · pi · opus48 · 139.4 AIC · ⌖ 42.8 AIC · ⊞ 1.6K · ◷
- expires on Oct 17, 2026, 1:56 PM UTC-08:00
Activity