fix(scripts): count subdirectory-only dirs as non-empty in PowerShell (parity with bash)

[jawwad-ali]

Description

Test-DirHasFiles in scripts/powershell/common.ps1 (the documented PowerShell twin of bash check_dir) tested directory non-emptiness with:

Get-ChildItem -Path $Path | Where-Object { -not $_.PSIsContainer } | Select-Object -First 1

The Where-Object { -not $_.PSIsContainer } filter counts only top-level files and ignores subdirectories. But the three sibling code paths all count any entry:

• bash check_dir (common.sh): [[ -d "$1" && -n $(ls -A "$1") ]]
• PowerShell JSON contracts checks (check-prerequisites.ps1, setup-tasks.ps1): Get-ChildItem ... | Select-Object -First 1 (no filter)

So a contracts/ directory whose only contents are subdirectories (e.g. a versioned/grouped layout contracts/v1/openapi.yaml) was reported present by bash, by bash JSON, and by PowerShell JSON — but [FAIL]/absent by PowerShell text mode. Test-DirHasFiles is the lone outlier, and it's internally inconsistent with the JSON path in its own files.

Fix

Drop the PSIsContainer filter so Test-DirHasFiles counts any entry, matching the other three paths (one-line change).

Testing

• Tested locally with uv run specify --help (exit 0)

• Ran existing tests with uv sync && uv run pytest

• uvx ruff check clean.

• uv run pytest tests/test_setup_tasks.py — 10 passed, 16 skipped (no regressions; bash + pwsh-only tests skip on my host).

• Verified directly with PowerShell: before, a subdir-only contracts/ → [FAIL]; after → [OK]; a genuinely empty dir and a missing dir still → [FAIL].

• New parity tests in tests/test_setup_tasks.py (bash check_dir + PowerShell Test-DirHasFiles) assert a subdir-only contracts/ is reported non-empty in both shells. The PowerShell test fails on main ([FAIL] contracts/) and passes with this fix.

AI Disclosure

• I did not use AI assistance for this contribution
• I did use AI assistance (describe below)

Found and fixed with Claude Code (Claude Opus 4.8) under my direction. AI located the divergence across the four code paths; I reproduced the [FAIL] vs present discrepancy under PowerShell, confirmed the one-line fix restores parity (and that empty/missing dirs still report [FAIL]), and reviewed the diff before submitting. I'll disclose if any review responses are AI-assisted as well.

[Copilot]

Pull request overview

This PR fixes a parity bug between the bash and PowerShell task scripts by making PowerShell’s Test-DirHasFiles treat a directory containing only subdirectories as “non-empty,” aligning behavior with the other prerequisite-check code paths.

Changes:

• Update Test-DirHasFiles in scripts/powershell/common.ps1 to treat any directory entry (including subdirectories) as non-empty.
• Add regression tests to assert bash check_dir and PowerShell Test-DirHasFiles both treat “subdir-only” directories as non-empty.

Show a summary per file

File	Description
scripts/powershell/common.ps1	Removes the subdirectory-excluding filter so non-emptiness counts any child entry.
tests/test_setup_tasks.py	Adds bash + PowerShell parity tests for subdirectory-only directory layouts.

Copilot's findings

Tip

Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

• Files reviewed: 2/2 changed files
• Comments generated: 2

[jawwad-ali]

Thanks for the review! Addressed in 6810368:

• Reworded the Test-DirHasFiles comment so it no longer claims exact ls -A parity (Get-ChildItem omits hidden entries without -Force). It now points at the in-repo PowerShell JSON contracts checks (check-prerequisites.ps1 / setup-tasks.ps1) as the matching reference, and keeps the subdir-only-counts-as-non-empty rationale.
• Renamed test_test_dir_has_files_ps_counts_subdir_only_contracts → test_dir_has_files_ps_counts_subdir_only_contracts to drop the doubled test_ prefix.

ruff clean; the PS parity test passes (bash leg skips on my host).

AI disclosure: prepared with Claude Code (Claude Opus 4.8) under my direction; I reviewed the diff before pushing.

[Copilot]

Copilot's findings

• Files reviewed: 2/2 changed files
• Comments generated: 2

[mnriem]

Please address Copilot feedback

[jawwad-ali]

@mnriem addressed Copilot's feedback in c1fb25c — the two tests were asserting result.returncode == 0, which proves nothing here (check_dir always exits 0; Test-DirHasFiles returns a boolean so pwsh exits 0 even on $false). Dropped that assertion and now rely on the ✓ / [OK] marker in stdout (the real behavioral signal), with an inline comment explaining why. Both review threads resolved. ruff clean; the PS parity test passes.

[Copilot]

Copilot's findings

• Files reviewed: 2/2 changed files
• Comments generated: 0 new

[mnriem]

Please address test & lint errors