chore: vendor bats test helpers instead of git submodules by jdx · Pull Request #859 · jdx/hk

jdx · 2026-04-23T18:34:05Z

Summary

Inlines bats-core, bats-assert, bats-file, and bats-support into the repo so cloning no longer needs --recurse-submodules or mise run init.
Drops submodules: recursive from every CI workflow (ci, docs, release, release-plz, autofix) and removes self.submodules = true from flake.nix.
Deletes .gitmodules and mise-tasks/init; updates CONTRIBUTING.md clone instructions.

Test plan

CI workflows succeed without submodule checkout
mise run test:bats runs end-to-end locally

🤖 Generated with Claude Code

Note

Medium Risk
Moderate risk because it replaces submodule-based test dependencies with vendored copies and updates CI/automation checkout behavior, which can break test execution or licensing/packaging expectations, but it does not change production runtime code.

Overview
Removes git submodule usage for Bats-based tests and vendors the Bats tooling and helper libraries directly into test/, eliminating the need for --recurse-submodules during cloning.

Updates automation to match this new layout: all GitHub Actions workflows drop submodules: recursive, flake.nix no longer enables submodules, mise build no longer depends on an init task, and contributor docs are updated; .gitmodules and mise-tasks/init are deleted accordingly.

^{Reviewed by Cursor Bugbot for commit 124429e. Bugbot is set up for automated code reviews on this repo. Configure here.}

greptile-apps · 2026-04-23T18:35:22Z

Greptile Summary

This PR vendors the bats-core, bats-assert, bats-file, and bats-support test helpers directly into the repository, eliminating all git submodule usage. All CI workflows, the Nix flake, the mise-tasks/build dependency list, and contributor documentation are updated accordingly. The change is clean and mechanical with no logic modifications.

Confidence Score: 5/5

Safe to merge — purely mechanical removal of git submodules in favour of vendored copies with no logic changes.

All findings are informational. Every submodule reference has been consistently removed across CI workflows, flake.nix, mise-tasks, and contributor docs. The vendored files replace the submodules at the same paths, so test execution paths are unchanged.

No files require special attention.

Important Files Changed

Filename	Overview
.github/workflows/ci.yml	Removes `submodules: recursive` from all six checkout steps; no other changes
.gitmodules	Deleted; all four submodule declarations removed as content is now vendored
flake.nix	Removes `self.submodules = true` from inputs; correct since test helpers are now vendored and no longer need submodule checkout
mise-tasks/build	Drops the `init` task dependency; correct now that submodule initialisation is no longer needed
mise-tasks/init	Deleted; the `git submodule update --init --recursive` script is no longer needed
CONTRIBUTING.md	Clone instruction updated to remove `--recurse-submodules` flag
test/bats/bin/bats	Vendored bats-core entrypoint script, replacing former submodule
test/test_helper/bats-assert/load.bash	Vendored bats-assert helper, replacing former submodule

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[Developer clones repo] --> B["git clone https://github.com/jdx/hk.git"]
    B --> C[test/bats/ already present\nbats-core vendored]
    B --> D[test/test_helper/ already present\nbats-assert, bats-file, bats-support vendored]
    C --> E[mise run test:bats]
    D --> E

    CI[CI Workflow] --> F["actions/checkout\nno submodules flag needed"]
    F --> E

    subgraph Before
        G["git clone --recurse-submodules ..."] --> H["mise run init\ngit submodule update --init --recursive"]
        H --> I[test helpers available]
    end

    subgraph After
        B
        C
        D
    end

_{Reviews (3): Last reviewed commit: "chore: vendor bats test helpers instead ..." | Re-trigger Greptile}

gemini-code-assist

Code Review

This pul l request vendors the bats-core testing framework and its helper libraries (bats-assert, bats-file, and bats-support) directly into the repository, removing the previous reliance on Git submodules. The review identified several technical issues within the vendored scripts: a bug in bats-exec-suite that incorrectly bundles com m and flags when running in parallel, a failure in the process-tracking logic of bats-exec-test that breaks timeouts, a regression in log filename form atting that introduces spaces, and a syntax error in the regex assertion logic. These issues should be addressed to ensure the testing infrastructure rem ains reliable and portable.

gemini-code-assist · 2026-04-23T18:41:34Z

+  # shellcheck disable=SC2086,SC2068
+  # we need to handle the quoting of ${flags[@]} ourselves,
+  # because parallel can only quote it as one
+  "${parallel_binary_name}" --keep-order --jobs "$num_jobs" -- "bats-exec-file" "$(printf "%q " "${flags[@]}")" "{#}" "{}" "$TESTS_LIST_FILE" <<< "$(printf "%s\n" "${BATS_UNIQUE_TEST_FILENAMES[@]}")" 2>&1 || bats_exec_suite_status=1


Bundling the flags array into a single quoted string using printf "%q " before passing it to parallel will cause all flags to be treated as a single argument by the worker processes. This will break option parsing in bats-exec-file. The flags should be passed as separate arguments to allow parallel to handle them correctly.

"${parallel_binary_name}" --keep-order --jobs "$num_jobs" -- "bats-exec-file" "${flags[@]}" "{#}" "{}" "$TESTS_LIST_FILE" <<< "$(printf "%s\n" "${BATS_UNIQUE_TEST_FILENAMES[@]}")" 2>&1 || bats_exec_suite_status=1

gemini-code-assist · 2026-04-23T18:41:34Z

+        printf "%d\n" "${row[$pid_col]}"
+      fi
+    done
+  } < <(ps -ef "$parent_pid")


The command ps -ef "$parent_pid" will only return information about the parent process itself on most systems, not its children. Since the subsequent loop checks for processes whose PPID matches $parent_pid, it will fail to find any child processes. This breaks the timeout mechanism which relies on identifying and killing child processes. It should be ps -ef to scan the full process list.

} < <(ps -ef)

gemini-code-assist · 2026-04-23T18:41:34Z

+  export BATS_RUNLOG_FILE='/dev/null'
+else
+  # use UTC (-u) to avoid problems with TZ changes
+  BATS_RUNLOG_DATE=$(date -u '+%Y-%m-%d %H:%M:%S UTC')


The date format used for log filenames contains spaces and special characters. This can lead to issues with filename handling in scripts and is a regression from the upstream bats-core format (%Y-%m-%dT%H:%M:%S). It is safer to avoid spaces in filenames to ensure portability and prevent potential parsing errors in other shell utilities.

BATS_RUNLOG_DATE=$(date -u +%Y-%m-%dT%H:%M:%S)

gemini-code-assist · 2026-04-23T18:41:34Z

+	local -r value="${1}"
+	local -r pattern="${2}"
+
+	if [[ '' =~ ${pattern} ]]; (( ${?} == 2 )); then


There is a semicolon instead of a logical OR (||) between the regex test and the exit status check. This is inconsistent with the implementation in refute_regex.bash and appears to be a typo introduced during vendoring. Using || ensures that the error block is only entered if the regex evaluation fails with an exit status of 2 (invalid regex).

if [[ '' =~ ${pattern} ]] || (( ${?} == 2 )); then

Inline bats-core, bats-assert, bats-file, and bats-support into the repo instead of pulling them in as git submodules. This removes the need for `git clone --recurse-submodules` (or `mise-tasks/init`) and drops the `submodules: recursive` checkout step from every CI workflow. - Remove `.gitmodules` and delete the submodule gitlinks; commit the vendored source directly under `test/bats` and `test/test_helper/*`. - Remove `mise-tasks/init` (only ran `git submodule update`). - Strip `submodules: recursive` / `self.submodules = true` from workflows, `flake.nix`, and `CONTRIBUTING.md`. Co-Authored-By: Claude Opus 4.7 (1M context) <[email protected]>

@jdx

### 🐛 Bug Fixes - **(git)** skip untracked scan when HK_STASH_UNTRACKED=false by [@jdx](https://github.com/jdx) in [#861](#861) - **(run)** add post-commit and pre-rebase subcommands by [@jdx](https://github.com/jdx) in [#858](#858) ### 📚 Documentation - **(install)** recommend global hooks as primary setup path by [@jdx](https://github.com/jdx) in [#855](#855) - add cross-site announcement banner by [@jdx](https://github.com/jdx) in [#857](#857) - respect banner expires field by [@jdx](https://github.com/jdx) in [#862](#862) ### 🔍 Other Changes - vendor bats test helpers instead of git submodules by [@jdx](https://github.com/jdx) in [#859](#859) ### 📦️ Dependency Updates - bump communique to 1.0.3 by [@jdx](https://github.com/jdx) in [#863](#863) - update anthropics/claude-code-action digest to e58dfa5 by [@renovate[bot]](https://github.com/renovate[bot]) in [#864](#864)  --- > [!NOTE] > **Low Risk** > Low risk release bookkeeping only: version numbers, generated CLI/docs, and lockfile dependency bumps with no runtime logic changes. > > **Overview** > **Release prep for `v1.44.1`.** Updates `Cargo.toml`/`Cargo.lock` (including `libc`) and all version references in generated CLI docs (`docs/cli/*`, `hk.usage.kdl`). > > Refreshes documentation examples to reference the new `v1.44.1` Pkl package URLs and adds the `1.44.1` entry to `CHANGELOG.md`. > > <sup>Reviewed by [Cursor Bugbot](https://cursor.com/bugbot) for commit d7637d4. Bugbot is set up for automated code reviews on this repo. Configure [here](https://www.cursor.com/dashboard/bugbot).</sup>  Co-authored-by: mise-en-dev <[email protected]>

gemini-code-assist Bot reviewed Apr 23, 2026

View reviewed changes

jdx force-pushed the chore/remove-submodules branch from d81ae2d to 3db4dc0 Compare April 23, 2026 18:43

jdx force-pushed the chore/remove-submodules branch from 3db4dc0 to 124429e Compare April 23, 2026 18:46

jdx enabled auto-merge (squash) April 23, 2026 18:49

jdx merged commit 25d55a2 into main Apr 23, 2026
20 checks passed

jdx deleted the chore/remove-submodules branch April 23, 2026 19:10

jdx mentioned this pull request Apr 23, 2026

chore: release v1.44.1 #856

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

chore: vendor bats test helpers instead of git submodules#859

chore: vendor bats test helpers instead of git submodules#859
jdx merged 1 commit intomainfrom
chore/remove-submodules

jdx commented Apr 23, 2026 •

edited by cursor Bot

Loading

Uh oh!

greptile-apps Bot commented Apr 23, 2026 •

edited

Loading

Uh oh!

gemini-code-assist Bot left a comment

Uh oh!

gemini-code-assist Bot Apr 23, 2026

Uh oh!

gemini-code-assist Bot Apr 23, 2026

Uh oh!

gemini-code-assist Bot Apr 23, 2026

Uh oh!

gemini-code-assist Bot Apr 23, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Conversation

jdx commented Apr 23, 2026 • edited by cursor Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Uh oh!

greptile-apps Bot commented Apr 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Greptile Summary

Confidence Score: 5/5

Important Files Changed

Flowchart

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist Bot Apr 23, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Apr 23, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Apr 23, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Apr 23, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

jdx commented Apr 23, 2026 •

edited by cursor Bot

Loading

greptile-apps Bot commented Apr 23, 2026 •

edited

Loading