Unname every return under src/, and stop the generators emitting named ones

[thedavidmeister]

Closes#129

Premise checked first

Every named return the issue lists at 369cfe9 was still there when this branched off ed91bfa, after #131 and #134. Nothing was already fixed. main then moved again — #136 landed the generated aggregate — and that merge is in here, so the sweep covers the code main has now rather than the code the issue was filed against.

What changed

Each function loses the name from its returns clause, gains an explicit return on every path, and keeps its @return tag with the name dropped from the tag text — the form zoltuAddress already used.

file	function	name dropped
src/abstract/RainDeploySuitesBase.sol	checkedCandidateSuites	candidates
src/abstract/RainDeploySuitesBase.sol	allSuites	suites
src/abstract/RainDeploySuitesBase.sol	suiteNames	names
src/abstract/RegistryDeploySuites.sol	candidateSuites	candidates
src/abstract/RainDeployVerifyBase.sol	deriveDeployment	derived
src/abstract/RainDeployVerifyBase.sol	deriveDeployments	derived
src/lib/LibRainDeploy.sol	isStartBlock	isStart
src/lib/LibRainDeploy.sol	findDeployBlock	deployBlock
src/lib/LibRainDeploy.sol	deployZoltu	deployedAddress
src/lib/LibRainDeploy.sol	deployToNetworks	deployedAddress
src/lib/LibRainDeploy.sol	deployAndBroadcast	deployedAddress
src/lib/LibRainDeploySnapshot.sol	frozenSnapshotPaths	paths
src/lib/LibRainDeploySnapshot.sol	sortedRecordPaths	sorted
src/lib/LibRainDeploySnapshot.sol	releasedImportBlock	imports
src/lib/LibRainDeploySnapshot.sol	newestFrozenTag	newest
src/lib/LibRainDeploySnapshot.sol	aggregateImportBlock	imports
src/interface/IAddressRegistryV1.sol	get	account
src/concrete/AddressRegistry.sol	get	account
script/Build.sol	generatedContracts	contracts
script/Build.sol	generatedContractNames	names

RegistryDeploySuites.releasedSuites is on the issue's list but is not in this diff: #136 replaced its body with return LibReleasedSuites.releasedSuites();, which is already unnamed.

No named return is left under src/ or script/. test/ is out of scope per the issue.

The two the issue calls the hard case

deployToNetworks assigned deployedAddress only inside the if/else arms of the loop over networks, safe solely because NoNetworks guarantees an iteration. Enumerating the paths shows both arms end with deployedAddress == expectedAddress — the deploy arm reverts UnexpectedDeployedAddress unless they are equal, the skip arm assigns it outright — so the accumulator carried nothing the caller had not already handed in. The deploy arm keeps a local for its own comparison, and the function returns expectedAddress after the loop.

newestFrozenTag used bytes(newest).length == 0 as an is-this-the-first-iteration flag on the default-initialised named return. It now refuses an empty record up front, seeds from paths[0] and scans from 1.

Both are also what slither's uninitialized-local detector named once the return was unnamed. Seeding either with the value its declaration already defaults to would be a line no test can kill — what the issue closed#103 over — so they are restructured rather than seeded, and no suppression is added. M21 and M22 in the table below are there because the restructure moved a live refusal's operand: they delete each refusal in turn to show it still fails a test.

The generators

Two emitters wrote named returns as string literals — the per-contract released lib and #136's aggregate over them. Both now emit an unnamed clause, a local declaration and a trailing return suites;, and the aggregate's empty-repo branch emits return new DeploySuite[](0);. src/lib/LibAddressRegistryReleased.sol, src/lib/LibMigrationRegistryReleased.sol and src/lib/LibReleasedSuites.sol are the output of forge script script/Build.sol against them, so the next build reproduces them rather than reverting this diff.

The deploy address moves

AddressRegistry.get is one of the unnamed functions, so the contract's runtime code changes — one byte longer, two jump destinations shifted. The rolling candidate snapshot is regenerated with it:

• BYTECODE_HASH0xef8355…4ea73 → 0x6c37a6…10097
• DEPLOYED_ADDRESS0x25aC2b82915f191dbE64e65BAeDDD68b97b68fe1 → 0x8cACfbD5d78b6D87080cE0839708ac2dA5461F78

src/generated/ holds no frozen tag directory, so no released suite and nothing on chain is contradicted by the move. checkCandidatesAnchoredToSource is what makes the regeneration mandatory rather than optional.

Beyond the issue's table

script/Build.sol is neither src/ nor the test/ the issue excludes. Its two named returns are unnamed here rather than left as the only counterexamples to a ruling that says this repo does not name return values.

Suite

run	result
main at ed91bfa, no RPC env	259 passed / 51 failed, every failure *_RPC_URL not found; grep '[FAIL' | grep -vc '_RPC_URL. not found' = 0
this branch, no RPC env	CI's test job is green
this branch, public RPC endpoints bound to all five aliases	320 passed / 0 failed / 320 total
forge fmt --check	exit 0
CI rainix	test, legal and static (slither + fmt + single-contract) all green

Every mutation below runs against that configuration, so the fork-dependent functions in LibRainDeploy are measured rather than skipped. Two earlier runs against rate-limited public endpoints reported failures that were HTTP 429s rather than the mutation; those are re-run and the table counts only runs whose failures are assertions.

Mutation pass

One mutation at a time, each deleting a return this diff ADDS or emits, against the full suite with live RPC. tests run is the suite's own N total tests line, so a run that compiled nothing or filtered to nothing is visible rather than counted as a kill. infra failures counts RPC transport errors in the run and is 0 everywhere, so no verdict rests on a fork that failed to open.

id	function	mutation	tests run	assertion failures	infra failures	verdict	killing tests
M01	RainDeploySuitesBase.checkedCandidateSuites	delete return candidates;	319	16	0	KILLED	testAllSuitesIsReleasedThenCandidates,testCandidatesPresentAnswers
M02	RainDeploySuitesBase.allSuites	delete return suites;	319	11	0	KILLED	testAllSuitesIsReleasedThenCandidates,testDerivationLeavesNoCodeBehind
M03	RainDeploySuitesBase.suiteNames	delete return names;	319	5	0	KILLED	testEmptySuiteIsUnknown,testRunRefusesToBroadcastACandidateThatIsNotItsSource
M05	RegistryDeploySuites.candidateSuites	delete return candidates;	319	5	0	KILLED	testEveryCandidateArtifactPathResolves,testEveryCandidateHasASnapshot
M06	RainDeployVerifyBase.deriveDeployment	delete return derived;	319	13	0	KILLED	testChainCodeHashMismatchReverts,testChainIgnoresAnUndeployedCandidate
M07	RainDeployVerifyBase.deriveDeployments	delete return derived;	319	5	0	KILLED	testChainCodeHashMismatchReverts,testChainMatrixReachesTheLastSupportedNetwork
M08	LibRainDeploy.isStartBlock	delete return isStart;	319	3	0	KILLED	testFindDeployBlockZoltuFactory,testIsStartBlockAtDeployBlock
M09	LibRainDeploy.findDeployBlock	delete return low;	319	3	0	KILLED	testFindDeployBlockExactZoltuBaseDeployBlock,testFindDeployBlockZoltuFactory
M10	LibRainDeploy.deployZoltu	delete return deployedAddress;	319	43	0	KILLED	testChainCodeHashMismatchReverts,testChainIgnoresAnUndeployedCandidate
M11	LibRainDeploy.deployToNetworks	delete return expectedAddress;	319	6	0	KILLED	testDeployAndBroadcastHappyPath,testDeployToNetworksEveryDependencyPresentDeploys
M12	LibRainDeploy.deployAndBroadcast	drop return from the tail call	319	1	0	KILLED	testDeployAndBroadcastHappyPath
M13	LibRainDeploySnapshot.frozenSnapshotPaths	delete return paths;	319	11	0	KILLED	testCheckReleaseFollowsRecordAcceptsExactlyTheStrictlyGreater,testCheckReleaseFollowsRecordRefusesATagBelowTheNewestRelease
M14	LibRainDeploySnapshot.sortedRecordPaths	delete return sorted;	319	3	0	KILLED	testRecordPathsForContractSelectsOneContractInTagOrder,testSortedRecordPathsOrdersTagsAsVersions
M15	LibRainDeploySnapshot.releasedImportBlock	delete return imports;	319	1	0	KILLED	testReleasedImportBlockAliasesEveryRecord
M16	LibRainDeploySnapshot.newestFrozenTag	delete return newest;	319	5	0	KILLED	testCheckReleaseFollowsRecordAcceptsExactlyTheStrictlyGreater,testCheckReleaseFollowsRecordRefusesATagBelowTheNewestRelease
M17	LibRainDeploySnapshot.releasedLibraryBlock	stop emitting return suites;	319	1	0	KILLED	testReleasedLibraryBlockDeclaresEveryRecord
M18	AddressRegistry.get	delete return account;	319	24	0	KILLED	testCandidatesPresentAnswers,testCheckResolvedAddressesChecksEveryRead
M19	Build.generatedContracts	delete return contracts;	319	2	0	KILLED	testGeneratedContractsAreExactlyTheDeclaredCandidates,testTheCommittedAggregateIsInDeclarationOrder
M20	LibRainDeploySnapshot.newestFrozenTag	delete the empty-record return ""; guard	319	5	0	KILLED	testFreezeCopiesTheRegeneratedRollingSnapshot,testFreezeCutsEveryNamedContract
M21	LibRainDeploy.deployToNetworks	delete the post-broadcast UnexpectedDeployedAddress refusal	319	1	0	KILLED	testDeployToNetworksFactoryReportsOtherAddressReverts
M22	LibRainDeploy.deployToNetworks	delete the UnexpectedDeployedCodeHash refusal	319	1	0	KILLED	testUnexpectedDeployedCodeHashReverts
M23	Build.generatedContractNames	delete return names;	320	1	0	KILLED	testGeneratedContractNamesAreTheDeclarationInOrder
M24	LibRainDeploySnapshot.aggregateImportBlock	delete return imports;	319	2	0	KILLED	testAggregateImportBlockImportsEveryReleasedLib,testTheCommittedAggregateIsWhatTheGeneratorEmits
M25	LibRainDeploySnapshot.aggregateLibraryBlock	stop emitting return suites;	319	2	0	KILLED	testAggregateLibraryBlockConcatenatesEveryReleasedLib,testTheCommittedAggregateIsWhatTheGeneratorEmits
M26	LibRainDeploySnapshot.aggregateLibraryBlock	emit new DeploySuite[](1) for an empty repo	319	1	0	KILLED	testAggregateLibraryBlockDeclaresNothingForNoContracts

AddressRegistry.get's mutant (M18) is the one where the source-anchor tests fire for free: any edit to that contract's source breaks testSnapshotMatchesSource whatever the edit does. Its kill does not rest on them — testGetReturnsRegistered, testGetOpaqueNames, testResolveRegistered and testResolveUnregistered are among the 24, and each of those reads a binding back rather than a pin.

M23 survived the first run and is the reason testGeneratedContractNamesAreTheDeclarationInOrder is in this PR. generatedContractNames() is the list cutRelease freezes and the list the aggregate is emitted from, and both reach it only through forge script, so nothing in the suite could tell it from an empty array. The new test asserts it positionally against generatedContracts() and now kills the mutant.

QA

• Discriminating tests: testGeneratedContractNamesAreTheDeclarationInOrder — new, fails on the M23 mutant (320 tests ran, it is the only failure) and passes on this branch. testReleasedLibraryBlockDeclaresEveryRecord and testAggregateLibraryBlockConcatenatesEveryReleasedLib — existing, expectations updated to the new emitted text; each fails against the pre-change emitter, verified by M17 and M25 which revert exactly that literal.
• Mutations applied: 25, each a return this diff adds or emits, deleted one at a time against the full suite with live RPC — 25 killed, 0 survived, 0 runs with an RPC transport failure. Full table above with the killing test for each.
• Oracle: the emitted text is asserted character for character against the form the issue specifies, spelled literally in the test rather than computed the way the emitter computes it. Behaviour has no oracle beyond "unchanged": every function keeps its old semantics, which is what the 320/320 suite and the unchanged MigrationRegistry pins check.
• Category check: the issue asks for (a) every named return under src/ unnamed with an explicit return on every path, (b) @return kept with the name dropped, (c) the generator changed so it stops emitting them, (d) the generated files regenerated from it. All four are covered, including the aggregate generator Move the release version under foundry's [external] section #136 added after the issue was filed and the two named returns in script/Build.sol. test/ is the issue's own exclusion and is untouched apart from the expectations that pin generated text and the one new test.

[coderabbitai]

Warning

Review limit reached

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

Next review available in:40 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: 5c9575e8-678a-4e18-982a-0b7b689d5585

📥 Commits

Reviewing files that changed from the base of the PR and between 74260ae and a1bcccb.

⛔ Files ignored due to path filters (1)

• src/generated/candidate/AddressRegistry.sol is excluded by !**/generated/**

📒 Files selected for processing (14)

• script/Build.sol
• src/abstract/RainDeploySuitesBase.sol
• src/abstract/RainDeployVerifyBase.sol
• src/abstract/RegistryDeploySuites.sol
• src/concrete/AddressRegistry.sol
• src/interface/IAddressRegistryV1.sol
• src/lib/LibAddressRegistryReleased.sol
• src/lib/LibMigrationRegistryReleased.sol
• src/lib/LibRainDeploy.sol
• src/lib/LibRainDeploySnapshot.sol
• src/lib/LibReleasedSuites.sol
• test/concrete/BuildHarness.sol
• test/script/Build.t.sol
• test/src/lib/LibRainDeploySnapshot.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.}