Cover the generated directory being created, by making the directory injectable

[thedavidmeister]

Closes #64

This branch had been reduced to nothing, and is restored here

Read this first if you looked at this PR before today. Its previous head,
30478d4, was a merge of main that resolved by taking main wholesale and
discarded both of the branch's own commits. git diff b422d97 30478d4 was
empty: the head tree was byte-identical to main at b422d97, which is
pre-#138. Against today's main that head was not a no-op but a revert — it
would have put script/Build.sol, src/generated/CodeGennable.sol and
.github/workflows/build-pointers.yaml back.

The work has been reconstructed on top of current main and the branch now
carries a real diff again.

The gap is real, measured on main today

src/lib/LibFs.sol creates the generated directory for itself:

vm.createDir(GENERATED_DIR, true);

Comment that line out on main and the suite reports 145 tests passed, 0 failed. Re-run it after rm -rf src/generated, to rule out the directory
merely being inherited from an earlier run, and it is 145 passed again. The
mutant survives, so the bootstrap behaviour — and the NatSpec claim that "the
first generation in a repo does not need it committed already" — is unprobed.

The masker is the suite's own scaffolding. Both
test/lib/LibFs.buildFileForContract.t.sol:40 and
test/src/lib/LibFs.isPresent.t.sol:26 call vm.createDir(GENERATED_DIR, true)
in setUp, so the directory is always already there by the time the library
would have created it.

Those setUp calls are not the bug and are not removed here. They exist because
every generating test writes under GENERATED_DIR and forge runs test
contracts in parallel, so a test that deletes the directory to make it missing
races all of its siblings. That is the hazard issue #64 measured directly: 1
failure in 6 full-suite runs, 0 in 8 runs under -j 1.

What ships

Issue #64 offered two ways to land the test — serialize the suite, or make the
directory injectable. This takes the second, because the first taxes every run
forever to cover one line.

src/lib/LibFs.sol:

• pathForContractIn(string dir, string contractName), private. The name
  check is unchanged, so the path is still a direct child of dir for every
  name accepted at all.
• buildFileForContract(Vm, address, string dir, string contractName, string body), internal overload. dir is what gets created when missing and what
  the file is written a direct child of.

The existing pathForContract and four-argument buildFileForContract are now
each the new function applied to GENERATED_DIR and are otherwise untouched, so
no consumer sees a behaviour change. dir is interpolated verbatim and is not
checked — that is why pathForContractIn is private rather than internal, and
it is stated in the NatSpec of both new functions.

One NatSpec line is corrected while it is being touched: the directory is created
"along with any missing parent of it", which vm.createDir(dir, true) has always
done and the docstring did not say.

test/lib/LibFs.buildFileForContract.t.sol:

• testBuildFileForContractCreatesTheDirectory owns
  src/generated/LibFsBuildCreatesDir/nested and never touches GENERATED_DIR
  itself, so it does not race a single sibling. Two levels are missing, so a
  create that made the leaf but not the missing parent fails here too.

Mutant proven dead

With vm.createDir(dir, true) removed from the new overload, on this branch:

[FAIL: vm.writeFile: failed to open file ".../src/generated/LibFsBuildCreatesDir/nested/LibFsBuildCreatesDir.sol": No such file or directory (os error 2)] testBuildFileForContractCreatesTheDirectory()
Ran 19 test suites: 145 tests passed, 1 failed, 0 skipped (146 total tests)

Exactly one test fails, and it is the new one. That is the same mutant that
survives 145/145 on main.

QA

origin/main = 7394a42, merged in (merge, not rebase).

• nix develop -c forge test — Ran 19 test suites: 146 tests passed, 0 failed, 0 skipped (146 total tests). main is 145; the one added test is the whole
  difference.
• nix develop -c forge fmt --check — clean, no diff.
• git status --porcelain — empty after a full run.

[coderabbitai]

Warning

Review limit reached

@thedavidmeister, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 8 minutes

Limit details: You’ve used all 1 included review currently available under your plan.

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: d984ffb9-453d-4e86-af4b-c66246359d69

📥 Commits

Reviewing files that changed from the base of the PR and between 7394a42 and acf1bf4.

📒 Files selected for processing (2)

• src/lib/LibFs.sol
• test/lib/LibFs.buildFileForContract.t.sol

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}