feat(surfaces): emit toilville/* Gutenberg blocks from ForgeBlock (layer 2 to 3) #104

Merged

peterswimm merged 2 commits from feat/forgeblock-toilville-gutenberg-emitter into main 2026-09-09 21:02:58 +00:00

Conversation 0 Commits 2 Files changed 35 +1244 -408

peterswimm commented 2026-09-09 21:02:30 +00:00

Owner

Copy link

Closes the layer Peter flagged on #102. Targets main.

The pipeline stopped one step short. ForgeBlock was rendered straight to core
Gutenberg blocks, so the patterns were a flat content skeleton — while the blocks
that actually carry the Toilville design (toilville/* in
Toilville_Rituals/wordpress/rituals-wp/) sat unused, even though
toilville-fse/templates/front-page.html already composes them by hand.

The adapter layer

forgeblock_toilville.py — each adapter recognises a run of ForgeBlocks and
emits the matching toilville/* block with attributes filled from the design
copy. Unrecognised runs still fall through to core blocks, so output stays valid
while coverage grows. Adapters are surface-aware: a surface with no hero
block (rituals/news/contact) falls through rather than borrowing another page's
hero.

Surface	Emits
landing	toilville/hero-home, toilville/joinband-cta
spelwork	toilville/spelwork-hero, toilville/spelwork-cta
rituals / news / contact	core fallback — no matching block exists yet

The Home hero now round-trips the design exactly:

{"eyebrow":"Toilville LLC · software for governed work",
 "headline1":"Governance should be a","emphasis":"primitive",
 "headline2":", not a plugin.","body":"We build software for organizations…",
 "cta1Label":"See Rituals","cta1Url":"#rituals",
 "cta2Label":"Build on Spelwork","cta2Url":"#spelwork"}

Two converter fixes this depended on

1. _collect_text() lost document order. handle_data() concatenated every
text run of a node into .text regardless of position, so
<h1>a <span>b</span> c</h1> flattened to "a c b" — that's the mangled
"Governance should be a , not a plugin. primitive" in the #102 renders. Node
now keeps parts, an ordered interleave of text runs and children. Joining is
punctuation-aware, so <span>primitive</span>, not… no longer yields
"primitive , not…".

2. Headings record emphasis — the inline <em>/<strong>/recoloured
<span>. A flat string can't say which words carry the accent, which is
exactly what hero-home's headline1/emphasis/headline2 split needs.

And a fix to my own previous pass

Annotating CTAs in #101 promoted the nearest enclosing <div> to
data-block="actions", which swallowed sibling headings and body copy — the
Home closing CTA "Let's build it." vanished entirely. Buttons are now wrapped
in their own button-only container. Landing goes 47 → 49 blocks.

Supporting changes

• theme.json gains six palette slugs so layer 3 has presets for the tokens
  added in #100: accent-spelwork, accent-rituals,
  neutral-warm-{ink,body,muted}, brand-surface. Each verified against
  dist/toilville/tokens.dtcg.json.
• align_check.py gains a layer-3 audit — raw hex in the toilville/* blocks,
  with the SPELWork-skin values (#7394c8/#587c6d/#c66c5a) reported
  separately since those are correct-but-untokenised, not drift.

Correction to something I said on #102

I reported "9× stale #cc5c37, 0× canonical" in the blocks. That was a plain
grep that didn't distinguish fallbacks — they are all
var(--wp--preset--color--primary, #cc5c37), so the blocks do reference the
preset and already render the corrected #c05734 in production. Only the
fallback literal is stale. The genuine raw hex is 13 values (#fff, #E1B659
in JS/PHP, #1a1a1a, #5a5a5a, #f4f4f2), which the new audit lists.

Verification

• just check — fully green
• pytest generator/tests/ tools/wp_theme_ingestor/test_*.py — 267 passed
• Other brands: only benign change is spelwork gaining a correct
  "emphasis": "legible." on its H1

Still open

No block yet for the Rituals hero, the Wish/Steward/Ritual triad, the
ritual-in-progress demo, the CLI terminal, the intent-registry console, or "Who
we are". products-grid/product-card/wish-form adapters need column_set
and form detection first. Tracked on #102.

https://claude.ai/code/session_01Kbzfscp4w5enUrZxMnDt78

Closes the layer Peter flagged on #102. Targets `main`. The pipeline stopped one step short. ForgeBlock was rendered straight to **core** Gutenberg blocks, so the patterns were a flat content skeleton — while the blocks that actually carry the Toilville design (`toilville/*` in `Toilville_Rituals/wordpress/rituals-wp/`) sat unused, even though `toilville-fse/templates/front-page.html` already composes them by hand. ## The adapter layer `forgeblock_toilville.py` — each adapter recognises a run of ForgeBlocks and emits the matching `toilville/*` block with attributes filled from the design copy. Unrecognised runs still fall through to core blocks, so output stays valid while coverage grows. Adapters are **surface-aware**: a surface with no hero block (rituals/news/contact) falls through rather than borrowing another page's hero. | Surface | Emits | |---|---| | landing | `toilville/hero-home`, `toilville/joinband-cta` | | spelwork | `toilville/spelwork-hero`, `toilville/spelwork-cta` | | rituals / news / contact | core fallback — no matching block exists yet | The Home hero now round-trips the design exactly: ```json {"eyebrow":"Toilville LLC · software for governed work", "headline1":"Governance should be a","emphasis":"primitive", "headline2":", not a plugin.","body":"We build software for organizations…", "cta1Label":"See Rituals","cta1Url":"#rituals", "cta2Label":"Build on Spelwork","cta2Url":"#spelwork"} ``` ## Two converter fixes this depended on **1. `_collect_text()` lost document order.** `handle_data()` concatenated every text run of a node into `.text` regardless of position, so `<h1>a <span>b</span> c</h1>` flattened to `"a c b"` — that's the mangled *"Governance should be a , not a plugin. primitive"* in the #102 renders. `Node` now keeps `parts`, an ordered interleave of text runs and children. Joining is punctuation-aware, so `<span>primitive</span>, not…` no longer yields `"primitive , not…"`. **2. Headings record `emphasis`** — the inline `<em>`/`<strong>`/recoloured `<span>`. A flat string can't say *which* words carry the accent, which is exactly what `hero-home`'s `headline1`/`emphasis`/`headline2` split needs. ## And a fix to my own previous pass Annotating CTAs in #101 promoted the nearest enclosing `<div>` to `data-block="actions"`, which **swallowed sibling headings and body copy** — the Home closing CTA *"Let's build it."* vanished entirely. Buttons are now wrapped in their own button-only container. Landing goes 47 → 49 blocks. ## Supporting changes - `theme.json` gains six palette slugs so layer 3 has presets for the tokens added in #100: `accent-spelwork`, `accent-rituals`, `neutral-warm-{ink,body,muted}`, `brand-surface`. Each verified against `dist/toilville/tokens.dtcg.json`. - `align_check.py` gains a layer-3 audit — raw hex in the `toilville/*` blocks, with the SPELWork-skin values (`#7394c8`/`#587c6d`/`#c66c5a`) reported separately since those are correct-but-untokenised, not drift. ## Correction to something I said on #102 I reported "9× stale `#cc5c37`, 0× canonical" in the blocks. That was a plain grep that didn't distinguish fallbacks — they are all `var(--wp--preset--color--primary, #cc5c37)`, so the blocks **do** reference the preset and already render the corrected `#c05734` in production. Only the fallback literal is stale. The genuine raw hex is 13 values (`#fff`, `#E1B659` in JS/PHP, `#1a1a1a`, `#5a5a5a`, `#f4f4f2`), which the new audit lists. ## Verification - `just check` — fully green - `pytest generator/tests/ tools/wp_theme_ingestor/test_*.py` — 267 passed - Other brands: only benign change is `spelwork` gaining a correct `"emphasis": "legible."` on its H1 ## Still open No block yet for the Rituals hero, the Wish/Steward/Ritual triad, the ritual-in-progress demo, the CLI terminal, the intent-registry console, or "Who we are". `products-grid`/`product-card`/`wish-form` adapters need `column_set` and form detection first. Tracked on #102. https://claude.ai/code/session_01Kbzfscp4w5enUrZxMnDt78

peterswimm added 2 commits 2026-09-09 21:02:35 +00:00

docs(review): side-by-side fidelity evidence for the toilville deck pipeline ... 15f5902d96

Renders the five toilville.com pages design-vs-generated so the gaps flagged in
PR #101 are visible rather than described: displaced inline <em> in headings,
feature bullets reduced to en-dashes, two- and three-column card grids
flattened to a single linear column, and the News/Contact structures lost.

Includes the render harness so the comparison is reproducible.

Claude-Session: https://claude.ai/code/session_01Kbzfscp4w5enUrZxMnDt78

feat(surfaces): emit toilville/* Gutenberg blocks from ForgeBlock (layer 2 -> 3) ... a6dd67df72

The pipeline stopped one layer short. ForgeBlock was being rendered straight to
CORE Gutenberg blocks (paragraph/heading/group/buttons), so the generated
patterns were a flat content skeleton — while the blocks that actually carry the
Toilville design (toilville/* in Toilville_Rituals/wordpress/rituals-wp/) sat
unused, even though toilville-fse/templates/front-page.html already composes
them by hand.

forgeblock_toilville.py adds an adapter layer: each adapter recognises a run of
ForgeBlocks and emits the matching toilville/* block with attributes filled from
the design copy. Unrecognised runs still fall through to core blocks, so output
stays valid while coverage grows. Adapters are surface-aware — a surface with no
hero block (rituals/news/contact) falls through rather than borrowing another
page's hero.

Covered so far:
  landing   -> toilville/hero-home + toilville/joinband-cta
  spelwork  -> toilville/spelwork-hero + toilville/spelwork-cta

Two converter fixes this depended on:

1. _collect_text() lost document order. handle_data() concatenated every text
   run of a node into .text regardless of position, so
   `<h1>a <span>b</span> c</h1>` flattened to "a c b" — the mangled
   "Governance should be a , not a plugin. primitive" in the review renders.
   Node now keeps `parts`, an ordered interleave of text runs and children, and
   _collect_text walks that. Joining is punctuation-aware so
   `<span>primitive</span>, not…` no longer yields "primitive , not…".

2. Headings now record `emphasis` — the inline <em>/<strong>/recoloured <span>.
   A flat string cannot say WHICH words carry the accent, which is exactly what
   hero-home's headline1/emphasis/headline2 split needs.

Also fixes an error in the previous pass: annotating CTAs promoted the nearest
enclosing <div> to data-block="actions", which swallowed sibling headings and
body copy (the Home closing CTA "Let's build it." vanished entirely). Buttons
are now wrapped in their own button-only container. Landing goes 47 -> 49 blocks.

theme.json gains six palette slugs so layer 3 has presets for the tokens added
in #100 — accent-spelwork, accent-rituals, neutral-warm-{ink,body,muted},
brand-surface. Each verified against dist/toilville/tokens.dtcg.json.

align_check.py gains a layer-3 audit: raw hex in the toilville/* blocks, with
the SPELWork-skin values (#7394c8/#587c6d/#c66c5a) reported separately since
those are correct-but-untokenised, not drift.

Claude-Session: https://claude.ai/code/session_01Kbzfscp4w5enUrZxMnDt78

peterswimm merged commit 35416cdb60 into main 2026-09-09 21:02:58 +00:00

peterswimm deleted branch feat/forgeblock-toilville-gutenberg-emitter 2026-09-09 21:02:58 +00:00

peterswimm referenced this pull request from a commit 2026-09-09 21:02:59 +00:00

Merge pull request 'feat(surfaces): emit toilville/* Gutenberg blocks from ForgeBlock (layer 2 to 3)' (#104) from feat/forgeblock-toilville-gutenberg-emitter into main

forge-ci referenced this pull request from a commit 2026-09-09 21:13:36 +00:00

docs(review): after-shots rendered through the real toilville/* blocks

peterswimm referenced this pull request 2026-09-09 21:14:10 +00:00

toilville.com: generated patterns are a content skeleton — fidelity gaps vs the design #102

forge-ci referenced this pull request from a commit 2026-09-09 21:20:34 +00:00

feat(theme): expose primary/secondary container slugs to WP presets

peterswimm referenced this pull request 2026-09-09 21:27:43 +00:00

feat(theme): expose primary/secondary container slugs to WP presets #105

forge-ci referenced this pull request from a commit 2026-09-09 21:29:20 +00:00

feat(theme): expose primary/secondary container slugs to WP presets

peterswimm referenced this pull request 2026-09-09 21:29:28 +00:00

feat(theme): expose primary and secondary container slugs to WP presets #106

Sign in to join this conversation.

Reviewers

No reviewers

Labels

Clear labels   

priority

P1

Urgent — no working icon yet (custom art pending) or a shipping asset

  

priority

P2

High — has an accessible stand-in; bespoke art wanted

  

priority

P3

Medium — stage visual headers

  

priority

P4

Low — behavioral dimension family

  

bug

Something is not working

  

duplicate

This issue or pull request already exists

  

enhancement

New feature

  

help wanted

Need some help

  

invalid

Something is wrong

  

question

More information is needed

  

wontfix

This won't be fixed

No labels

priority

P1

priority

P2

priority

P3

priority

P4

bug

duplicate

enhancement

help wanted

invalid

question

wontfix

Milestone

Clear milestone

No items

No milestone

Projects

Clear projects

No items

No project

Assignees

Clear assignees

No assignees

1 participant

Notifications

Subscribe

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference

Toilville/toilville-design!104

No description provided.