feat(bestax-migrate): convert navbar dividers with classes, and navbar text colors

[allxsmith]

Part of #743, following two library fixes that just merged. Navbar.Divider now merges a className with its own class (#791), and Navbar.Menu, Navbar.Start and Navbar.End now render textColor (#792). The codemod refused or kept both as classes because of the old behavior, so each converts now:

<div className="navbar-end has-text-white">
  <hr className="navbar-divider my-0 site-divider" />
</div>
// becomes
<Navbar.End textColor="white">
  <Navbar.Divider className="my-0 site-divider" />
</Navbar.End>

What changed

• bulma-classes (feat(bestax-migrate)): the divider's entry loses ownClassOnly, and since nothing else used that field, the field and its attr:className refusal go too. attr:className still covers the wrappers and absorbed children it was already refused on. The navbar areas' textColor: null goes, so has-text-* becomes textColor there like everywhere else.
• bloomer (fix(bestax-migrate)): the same stale assumption lived in bloomer's source, which listed Navbar.Divider among the parts whose className replaces their own class. It's a helperless part now, so a helper on a NavbarDivider becomes its Bulma class in className instead of a TODO, and a spread converts instead of keeping bloomer's element.
• MCP (fix(bestax-mcp)): lookup_bulma_classes stopped saying a divider with another class stays markup, and the index drops the ownClassOnly field.
• Skill: the component map, prop map and unmappables lose the divider refusal and the three navbar areas from the has-text-* exceptions. bloomer's prop map loses the divider from its class-hostile list.

Checked

• The render truth test and the props e2e sweep the navbar entries like any other, so has-text-* on each area is rendered and typechecked through textColor.
• A DarkNavbar case in the navbar fixture pair, and a divider with a class and a text color on .navbar-end in the kitchen sink, render the same before and after. The divider left leftovers.tsx, since it converts now.
• A planner test for each change, a bloomer test for a divider with a helper and one with a spread, and the MCP lookup test now expects the divider's extra class to stay in className.
• pnpm all passes locally.

Merge note: squashing this keeps only the title's scope, so bestax-mcp's lookup fix waits for its next release unless the merge keeps the commits.

[coderabbitai]

Warning

Review limit reached

Next included review available in 1 minute.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

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

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: Repository: allxsmith/bestax/.coderabbit.yaml

Review profile: CHILL

Plan: Advanced

Run ID: a14ae326-2406-4388-858a-47de66c21b65

📥 Commits

Reviewing files that changed from the base of the PR and between 47ecdb6 and 03c938e.

📒 Files selected for processing (22)

• bestax-mcp/data/bulma-classes.json
• bestax-mcp/data/skills.json
• bestax-mcp/src/__tests__/server.test.ts
• bestax-mcp/src/bulma-classes.ts
• bestax-migrate/e2e/bulma-classes-kitchen-sink.test.ts
• bestax-migrate/fixtures/bulma-classes-kitchen-sink/src/layout.tsx
• bestax-migrate/fixtures/bulma-classes-kitchen-sink/src/leftovers.tsx
• bestax-migrate/src/sources/bloomer/__tests__/specials.test.ts
• bestax-migrate/src/sources/bloomer/mapping.ts
• bestax-migrate/src/sources/bloomer/transform.ts
• bestax-migrate/src/sources/bulma-classes/__testfixtures__/navbar.input.tsx
• bestax-migrate/src/sources/bulma-classes/__testfixtures__/navbar.output.tsx
• bestax-migrate/src/sources/bulma-classes/__tests__/plan.test.ts
• bestax-migrate/src/sources/bulma-classes/class-map.ts
• bestax-migrate/src/sources/bulma-classes/plan.ts
• bestax-migrate/src/sources/bulma-classes/rules.ts
• scripts/gen-mcp-index.mjs
• scripts/gen-mcp-index.test.mjs
• skills/bestax-migrate/references/bloomer/prop-map.md
• skills/bestax-migrate/references/bulma-classes/component-map.md
• skills/bestax-migrate/references/bulma-classes/prop-map.md
• skills/bestax-migrate/references/bulma-classes/unmappables.md

---

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.}

[allxsmith]

deep-review: fresh

A small PR, so keep the review small: read the three commits one at a time with git show, not the whole PR diff at once.

Priority: is each conversion now really identical markup? Navbar.Divider merging a className (bulma-ui's Navbar.tsx, since #791) and Navbar.Menu/Start/End rendering textColor (since #792). Then check that removing ownClassOnly left no refusal, message or doc claiming the old behavior, in the bulma-classes source, bloomer's source, the MCP lookup and the skill references.

[github-actions]

Preview Deployment

Preview URL: https://c473b0f3.bestax.pages.dev

[github-actions]

Preview Deployment

Preview URL: https://b1ca5c68.bestax.pages.dev

[allxsmith]

deep-review: fresh

The runs before this one were stopped by the runner shutting down partway through ("The runner has received a shutdown signal"), which looks like the VM running out of resources. So keep the machine light: don't run pnpm all, pnpm build, a Storybook build or a whole package's test suite. CI already runs those on this head. If you want evidence from a test, run one file, for example pnpm -C bestax-migrate test -- src/sources/bulma-classes/__tests__/plan.test.ts.

Read the commits one at a time with git show, not the whole PR diff at once. The priority is the same as the earlier steer on this PR: does every conversion the change allows render exactly the markup it replaces, and is everything else refused?

[allxsmith]

deep-review: fresh

Runs on this PR keep getting stopped by the runner shutting down at about the 20-minute mark, whatever they're doing, so this review has to finish well inside that. Keep it to the core and post what you have by about 15 minutes:

• The navbar entries in bestax-migrate/src/sources/bulma-classes/class-map.ts (the divider without ownClassOnly, has-text-* on the menu, start and end), against bulma-ui's Navbar.tsx.
• NO_CLASSNAME_TARGETS and HELPERLESS_TARGETS in bestax-migrate/src/sources/bloomer/mapping.ts, and the transform branch that reads them.

Skip the MCP, docs and skill changes, and don't run builds or test suites: CI runs them on this head. Read with git show <sha> -- <path> for just those files.

[claude]

Deep review — 1 blocking · 2 advisory

No.	Severity	Area	Finding	Location
1	🟡 Minor	Correctness	The block comment introducing the four navbar-part entries still says Menu, Start and End "spread textColor onto the element rather than reading it, so their text color stays a class" — false as of #792, and it contradicts the textColor: null lines this PR deletes. Already carried by the open thread at class-map.ts:1228, so not re-posted inline.	bestax-migrate/src/sources/bulma-classes/class-map.ts:1214-1216
2	🔵 Advisory	Robustness	Navbar.Divider moving into HELPERLESS_TARGETS gives up the spread guard NO_CLASSNAME_TARGETS applied. A NavbarDivider carrying a spread now converts, and a spread holding a bloomer-only helper (isPulled, hasTextColor, …) lands on the <hr> as a DOM attribute rather than becoming a class — TypeScript does not flag excess props arriving through a JSX spread. Intended and correct given the className merge, and identical to every other member of that set, so on the record rather than a defect.	bestax-migrate/src/sources/bloomer/mapping.ts:166
3	🔵 Advisory	Coverage	Dropdown.Divider sits in both NO_CLASSNAME_TARGETS and HELPERLESS_TARGETS, and the if / else-if means its HELPERLESS_TARGETS row can never be reached. Pre-existing, and now the only such row — worth a look while the file is open.	bestax-migrate/src/sources/bloomer/mapping.ts:167

Overall: The change is sound, and each removal is grounded in library code I read rather than in the PR body. NavbarDivider renders classNames(usePrefixedClassNames("navbar-divider"), className) (bulma-ui/src/components/Navbar.tsx:833-840), so dropping ownClassOnly and the attr:className refusal it drove is right, and the field has no remaining reference anywhere in the repo (a repo-wide grep for ownClassOnly is empty). NavbarMenu, NavbarStart and NavbarEnd each feed color: textColor ?? color into useBulmaClasses (:387-397, :445-455, :477-487), and each types textColor to validColors — token for token the colour list the table converts — so removing textColor: null cannot emit a value the prop rejects. The bgColor: null rows correctly survive: NavbarMenuProps and NavbarStartEndProps omit backgroundColor from BulmaClassesProps, and there is no bgColor key to slip through. On the bloomer side both remaining NO_CLASSNAME_TARGETS claims still hold — PaginationEllipsis spreads its props after its own className (Pagination.tsx:339-347) and DropdownDivider is a bare React.FC declaring no props (Dropdown.tsx:544-546) — and Navbar.Divider genuinely belongs in HELPERLESS_TARGETS, since it extends only React.HTMLAttributes<HTMLHRElement>. The riskiest part is the widest-reaching one, has-text-* on three new targets, and that is exactly what e2e/bulma-classes-truth.test.ts:243-269 gates, by rendering raw markup against the converted component for every HELPER_TOKENS entry on every allowed tag. Start with the one open thread; the code itself needs nothing.

Residual risk:

• Another source carrying the same stale divider assumption. Refuted: rbx/mapping.ts:699 and react-bulma-components/mapping.ts:665 map Navbar.Divider with no class-hostile list of their own, and a grep for the "replaces the class that makes it work" and "drops its own class" phrasings across ts, tsx, md, mjs and json now matches nothing outside the two comments this PR edited.
• A generated artefact still encoding ownClassOnly or the old textColor: null. Refuted for the ESLint plugin: scripts/gen-eslint-meta.mjs:232-255 reads only status, target, part and PRECEDENCE, so metadata.ts:242 (navbar-divider to Navbar.Divider, converts: true) is unaffected and cannot go stale. The MCP index is regenerated here; per the steer I did not review it.
• The conditionalStays half of the refusal deleted alongside ownClassOnly. Refuted: a conditional class on a divider now falls through to the leftover-joiner path, and classNames drops a falsy or empty result, so a joiner call that adds one class conditionally renders identically before and after.
• A dead className entry in the VOCABULARY of rules.ts once that refusal went. Refuted: the two refusals its rewritten comment now describes are live at plan.ts:641 (the child a target renders itself) and plan.ts:797 (the wrapper a target renders), so the token is still reachable and the comment matches them.
• Open: the HELPERLESS_TARGETS spread path in row 2. It is the accepted shape for that whole set rather than something this PR introduced, but a bloomer helper riding in through a spread on a NavbarDivider now reaches the DOM where the element used to be retained instead.
• Per the steer I ran no builds or suites and did not review the MCP, docs or skill changes; CI covers the first, and the skill and MCP tables were checked only where they bear on the two source files above.

🏄 Two library fixes landed upstream and this one just paddles out and takes the wave they made — dividers with classes and navbar text colors both convert clean now, and it is all backed by a render-truth sweep instead of vibes. Only flotsam left is a stale comment still arguing the old line; scrape that off and it is good to go.

[allxsmith]

deep-review: verify

Check the fix for the open thread (the stale comment above the navbar parts) in 03c938e. The same commit also drops the unreachable Dropdown.Divider row from bloomer's HELPERLESS_TARGETS.

Keep it short: runs here get stopped by the runner at about 20 minutes, so read with git show 03c938e2 and don't run builds or test suites (CI runs them on this head).

[github-actions]

Preview Deployment

Preview URL: https://f254b756.bestax.pages.dev

[claude]

Deep review (verify) — 1 resolved · 0 open

#	Severity	Area	Finding	Location
1	🟡 Minor	Correctness	Stale comment claiming Menu/Start/End keep their text color as a class — verified fixed, resolved	bestax-migrate/src/sources/bulma-classes/class-map.ts:1214

Overall: This pass settled the one open thread from the prior deep review and reviewed no commits. The comment is now reduced to // No background prop on these parts., which is the deletion I asked for and is true of all four entries (bgColor: null on each, textColor: null on none); grep -rn "rather than reading it" matches nothing repo-wide. I also confirmed the same commit's Dropdown.Divider removal from HELPERLESS_TARGETS is a dead-row deletion: NO_CLASSNAME_TARGETS (bloomer/mapping.ts:156-159) still holds it and transform.ts:350 tests that set in the leading if, with HELPERLESS_TARGETS only reachable through the else if at :405 — no test imports the set and skills/…/bloomer/prop-map.md:149-154 already lists the divider as the no-className exception, so no behaviour or doc moved. Nothing left for the human here beyond the normal squash-merge.

Residual risk: none identified for the settled thread — the claim it reported is gone from the source rather than reworded, and the only other change in the commit is provably unreachable code. Per the steer I did not run builds or test suites; CI on this head covers bestax-migrate's suite, and the two edits are a comment and a dead set member, neither of which has a runtime path a test could newly exercise.

🏄 Clean little clean-up set, dude — one lying comment trimmed to the one clause that's actually true, and a dead row swept out of a list where the first if always won anyway. Nothing gnarly left in the water; ship it.

[github-actions]

📸 Story screenshots at handoff — 03c938e

No Storybook stories map to this PR's changed files — nothing to screenshot. (workflow run)

[bestax-release-bot]

🎉 This PR is included in version 2.13.0 🎉

The release is available on:

• npm package (@latest dist-tag)
• GitHub release

Your semantic-release bot 📦🚀

[bestax-release-bot]

🎉 This PR is included in version 1.7.0 🎉

The release is available on:

• npm package (@latest dist-tag)
• GitHub release

Your semantic-release bot 📦🚀

[bestax-release-bot]

🎉 This PR is included in version 5.18.0 🎉

The release is available on:

• npm package (@latest dist-tag)
• GitHub release

Your semantic-release bot 📦🚀

[bestax-release-bot]

🎉 This PR is included in version 4.2.13 🎉

The release is available on:

• npm package (@latest dist-tag)
• GitHub release

Your semantic-release bot 📦🚀