# content-on-sx: Documents, blocks & collaborative editing on Smalltalk

> **DRAFT outline.** The CMS vertical — blog, WYSIWYG editor, Ghost sync. Depends
> on `persist-on-sx` (document history as an event log). Ghost/CMS sync stays a thin
> external adapter (Python/FFI) until a native replacement exists.

rose-ash's `blog` domain is content management: a block-based WYSIWYG editor,
navigation, Ghost CMS sync. A document is a tree of live blocks; editing is a
stream of operations; collaboration needs conflict-free merge. That is an object
model — blocks are objects, edits are messages, and a document is the object graph
responding to them. Smalltalk's "everything is an object responding to messages"
maps directly to a block/WYSIWYG model, and a semilattice (CRDT) merge keeps
concurrent edits conflict-free.

End-state: a Smalltalk-on-SX document model (typed blocks, structural ops),
operation log + CRDT merge for collaborative editing, versioning/history via the
event store, and a render boundary to HTML/SX. External CMS (Ghost) sync is an
injected adapter, not core.

## Status (rolling)

`bash lib/content/conformance.sh` → **657/657** (Phases 1–4 COMPLETE + ~29 extensions: HTML/SX escaping, Markdown render + import/export incl. tables & frontmatter (full round-trip), CRDT replication, tree-aware validation, snapshot cache, doc metadata, plain-text render, nested block trees + deep editing + flatten, doc stats, table + callout blocks, HTML page wrapper + SEO page, doc composition + id-remap, portable data + wire serialization, block query + transforms + find/replace, TOC + anchored headings + outline, normalization)

## Ground rules

- **Scope:** only `lib/content/**` and `plans/content-on-sx.md`. May **import**
  from `lib/smalltalk/`, and (once it exists) `lib/persist/`. Do not edit substrates.
- **Architecture:** a document is an ordered tree of blocks (objects); an edit is a
  message (`insert`/`update`/`move`/`delete`); concurrent edits merge via a
  commutative (CRDT/semilattice) operation so order doesn't matter. History is the
  `persist` event stream; any version is a replay.
- **Determinism:** merge must be commutative + idempotent (test: apply ops in any
  order / twice → same document).
- **Commits:** one feature per commit. Progress log + tick boxes.

## Architecture sketch

```
Edit op                                 Rendered document
  (insert block after id) ...             HTML / SX tree
        │                                       ▲
        ▼                                       │
lib/content/block.sx                    lib/content/render.sx
  — typed blocks as objects               — block tree → HTML/SX
  — heading/text/image/embed              — (reuses SX render boundary)
        │                                       ▲
        ▼                                       │
lib/content/doc.sx                      lib/content/merge.sx
  — ordered block tree                    — CRDT/semilattice op merge
  — apply op, structural moves            — concurrent-edit reconciliation
        │                                       ▲
        ▼                                       │
lib/content/api.sx ── (content/edit) (content/render) (content/history) ──┐
        │                                                                  │
        ├── op log + versions → persist                                      │
        └── Ghost/CMS sync  → injected external adapter (thin, non-core) ──┘
```

## Phase 1 — Block document model
- [x] `block.sx` — typed block objects
- [x] `doc.sx` — ordered tree, apply edit op, structural moves
- [x] `render.sx` — block tree → HTML/SX
- [x] `api.sx` + tests + scoreboard + conformance.sh

## Phase 2 — Op log + versioning
- [x] edit ops as `persist` events; replay to any version
- [x] `(content/history doc)`, diff between versions

## Phase 3 — Collaborative merge (CRDT)
- [x] commutative/idempotent op merge
- [x] concurrent-edit tests (any order, double-apply → identical)

## Phase 4 — External sync + federation
- [x] Ghost/CMS sync via injected adapter (import/export)
- [x] federated documents (peer-authored blocks) — trust-gated stub
- [x] tests: round-trip import/export, conflict on concurrent external edit

## Extensions (post-roadmap)
- [x] HTML escaping at the render boundary (`String>>htmlEscaped`: & < > ")
- [x] asSx wire string-escaping (`String>>sxEscaped`: \ and " in SX literals)
- [x] Markdown render mode (`asMarkdown:` / `content/render doc "md"`)
- [x] durable CRDT replication (`crdt-store.sx`: ops on persist, replay + converge)
- [x] document validation (`validate.sx`: ids, per-type fields, duplicate ids; tree-aware — descends into sections, tree-wide dup ids, section field check)
- [x] Markdown import adapter (`md-import.sx`: text → blocks, round-trips export; incl. pipe tables + frontmatter → metadata)
- [x] Markdown doc export (`md-doc.sx`: content/markdown-doc, frontmatter from metadata, full round-trip)
- [x] snapshot cache over replay (`snapshot.sx`: cache-not-primary, transparent)
- [x] document metadata (`meta.sx`: title/slug/tags + Ghost title plumbing)
- [x] plain-text render + excerpt (`text.sx`: asText, content/excerpt)
- [x] nested block trees (`section.sx`: CtSection container, recursive render, deep-find)
- [x] document statistics (`stats.sx`: word/char/block counts, reading time)
- [x] table block (`table.sx`: CtTable, renders html/sx/text/md, validated)
- [x] callout block (`callout.sx`: CtCallout note/warning/tip, renders html/sx/text/md, validated)
- [x] HTML page wrapper (`page.sx`: content/page, escaped title from metadata)
- [x] SEO page (`page-full.sx`: content/page-full, lang + meta description from excerpt)
- [x] document composition (`compose.sx`: concat/prepend/concat-all/wrap-section)
- [x] deep tree editing (`tree-edit.sx`: doc-deep-update/replace/delete/insert-into)
- [x] id remapping / clone (`clone.sx`: content/remap-ids + prefix-ids, collision-free compose)
- [x] block query + TOC (`query.sx`: content/select/select-type/count-type/headings)
- [x] block transforms (`transform.sx`: content/map-blocks/map-type/set-field-on)
- [x] TOC rendering (`toc.sx`: content/toc-markdown + toc-html from headings)
- [x] anchored-heading render (`anchor.sx`: content/html-anchored, functional TOC links)
- [x] document outline (`outline.sx`: content/outline, nested heading tree)
- [x] document flatten (`flatten.sx`: content/flatten, un-nest sections; inverse of wrap-section)
- [x] document normalization (`normalize.sx`: content/normalize, drop empty blocks/sections)
- [x] global find/replace (`find-replace.sx`: content/find-replace across text-bearing blocks)
- [x] portable data serialization (`data.sx`: content/to-data + from-data, round-trips tree)
- [x] wire serialization (`wire.sx`: content/to-wire + from-wire, SX-text on the wire)

## Progress log

- 2026-06-07 — Extension: callout/admonition block (`callout.sx`). `CtCallout`
  holds kind (note/warning/tip) + text; answers asHTML (`<aside class="callout
  callout-KIND">`, escaped), asSx, asText, asMarkdown: (`> **kind:** text`).
  Self-contained; `mk-callout`/`callout?`/`callout-kind`. validate.sx gained a
  `callout` field case. 12 tests; suite 657/657.
- 2026-06-07 — Extension: document flatten (`flatten.sx`). `content/flatten`
  un-nests a sectioned doc into a flat block sequence (each section replaced
  inline by its recursively-flattened children, wrapper dropped) — the inverse of
  content/wrap-section, for flat export targets. Inline tree handling; immutable.
  10 tests; suite 645/645.
- 2026-06-07 — Extension: nested document outline (`outline.sx`).
  `content/outline` builds a hierarchical heading tree from content/headings —
  each node `{:id :text :level :children}`, headings nesting under the nearest
  lower-level heading (recursive forest build). The structured companion to the
  flat TOC for nested nav. 14 tests; suite 635/635.
- 2026-06-07 — Extension: anchored-heading render (`anchor.sx`).
  `content/html-anchored` renders like asHTML but headings carry `id="<block-id>"`
  (tree-wide, sections recurse, text escaped), so the TOC's `#id` links resolve —
  completing the TOC feature end-to-end. A separate render; plain asHTML
  unchanged. 6 tests; suite 621/621.
- 2026-06-07 — Extension: global find/replace (`find-replace.sx`).
  `content/find-replace` replaces every occurrence of a substring in the text
  field of text/heading/code/quote blocks tree-wide (via the transform layer) —
  rename a term throughout a doc. Leaves non-text fields (image alt/src) alone;
  immutable, case-sensitive. 10 tests; suite 615/615.
- 2026-06-07 — Extension: document normalization (`normalize.sx`).
  `content/normalize` drops empty text blocks and empty sections tree-wide;
  sections are normalised first so one emptied by the pass is itself removed.
  For tidying imported/edited docs; non-text empties (dividers, blank-alt images)
  preserved. Inline tree handling; immutable. 11 tests; suite 605/605.
- 2026-06-07 — Extension: table-of-contents rendering (`toc.sx`).
  `content/toc-markdown` produces a Markdown bullet list indented by heading
  level with `[text](#id)` links; `content/toc-html` produces a `<ul>` of escaped
  anchor links (`#id`). Built from `content/headings`; the blog page's TOC
  artifact. 8 tests; suite 594/594.
- 2026-06-07 — Extension: tree-wide block transforms (`transform.sx`). The write
  counterpart to query: `content/map-blocks` (predicate) / `content/map-type` /
  `content/set-field-on` apply a function to every matching block across the tree
  (sections rebuilt), for bulk edits (cdn src rewrites, heading-level bumps, text
  sanitisation). Inline tree rebuild (no section.sx dep); immutable. 12 tests;
  suite 586/586.
- 2026-06-07 — Extension: block query + TOC (`query.sx`). `content/select`
  (predicate) / `content/select-type` / `content/count-type` / `content/select-ids`
  collect blocks across the whole tree (sections recurse); `content/headings`
  derives a table of contents (`{:id :level :text}` per heading, document order).
  Inline tree detection (no section.sx dep). 13 tests; suite 574/574.
- 2026-06-07 — Extension: id remapping / clone (`clone.sx`).
  `content/remap-ids` deep-rewrites every block id across the tree (sections
  recurse) via a function; `content/prefix-ids` prefixes them. Enables
  collision-free composition (prefix each doc before concat → validates clean,
  where the unprefixed concat has duplicate ids). Content unchanged, only ids;
  immutable. 10 tests; suite 561/561.
- 2026-06-07 — Extension: deep tree editing (`tree-edit.sx`). `doc-deep-update`
  / `doc-deep-replace` / `doc-deep-delete` / `doc-deep-insert-into` mutate blocks
  anywhere in the nested tree (descending into CtSection children), completing
  tree mutation to match the deep-find read path; all immutable. 17 tests; suite
  551/551.
- 2026-06-07 — Extension: on-the-wire serialization (`wire.sx`).
  `content/to-wire` serialises a document to a transmittable SX-text string (data
  form + SX serializer); `content/from-wire` parses it back into a live document.
  The whole-document format for persistence / HTTP / federation (distinct from
  the per-op persist log); round-trips nested trees + tables; reads externally
  authored wire strings. 11 tests; suite 534/534.
- 2026-06-07 — Extension: portable data serialization (`data.sx`).
  `content/to-data` converts a document to plain SX data
  (`{:id :title :slug :tags :blocks [{:id :type :fields}]}`, sections recursing);
  `content/from-data` reconstructs real block objects (section/table handled
  specially, others generically via mk-block). Round-trips the full tree +
  metadata (render-equal), decoupling storage/transport from the Smalltalk
  instance shape. 21 tests; suite 523/523.
- 2026-06-07 — Extension: document composition (`compose.sx`). `content/concat`
  / `content/prepend` / `content/concat-all` combine documents (keeping the
  first's id + metadata, concatenating blocks, immutable); `content/wrap-section`
  collapses a doc's blocks into a single nested section. For assembling pages
  from header/body/footer parts and templates. 17 tests; suite 502/502.
- 2026-06-07 — Extension: SEO-complete page (`page-full.sx`). `content/page-full`
  extends content/page with `<html lang="en">` and a `<meta name="description">`
  drawn from the document excerpt (plain text, escaped, 160 chars), composing the
  page/metadata/text layers into the SEO-ready artifact. 4 tests; suite 485/485.
- 2026-06-07 — Extension: Markdown document export (`md-doc.sx`).
  `content/markdown-doc` emits a `---` frontmatter block from metadata
  (title/slug/tags, only present fields) ahead of the Markdown body, or plain
  asMarkdown when there's no metadata. Completes the metadata round-trip:
  `md/import ∘ content/markdown-doc` preserves title/slug/tags + blocks. 12
  tests; suite 481/481.
- 2026-06-07 — Extension: Markdown frontmatter. `md/import` parses a leading
  `---` / `key: value` / `---` block into document metadata (title, slug,
  comma-separated tags via `doc-with-meta`) before parsing the body; a `---`
  elsewhere stays a divider. Ties the Markdown importer to the metadata layer the
  way real blog posts work. +9 tests; suite 469/469.
- 2026-06-07 — Extension: Markdown table import. `md-import.sx` now recognizes a
  `| … |` header row followed by a `| --- |` separator and parses a `CtTable`
  (cells trimmed, mixed with other blocks via blank-line separation), completing
  the Markdown table round-trip (import∘export == identity). +5 tests; suite
  460/460.
- 2026-06-07 — Extension: HTML page wrapper (`page.sx`). `content/page` composes
  metadata + render into a minimal valid HTML5 document — escaped `<title>` from
  doc metadata (falling back to id) and the rendered blocks as the body.
  `content/page-title`. The shippable artifact the blog serves. 7 tests; suite
  455/455.
- 2026-06-07 — Extension: table block (`table.sx`). `CtTable` holds headers +
  rows (string lists); answers asHTML (escaped `<table>`), asSx, asText, and
  asMarkdown: (pipe table with dashed separator row) by folding rows×cells via
  nested `inject:into:`. Self-contained (no edits to block.sx/render.sx);
  `mk-table`, `table?`, `table-headers/rows`. validate.sx gained a `table` field
  case (headers/rows must be lists). 15 tests; suite 448/448.
- 2026-06-07 — Extension: document statistics (`stats.sx`). `content/stats`
  returns `{:words :chars :blocks :reading-minutes}`; word/char counts derive
  from the tree-accurate `asText` projection, block count from an inline tree
  walk (no section.sx dep), reading time at 200 wpm rounded up. Counts descend
  into nested sections. 17 tests; suite 433/433.
- 2026-06-07 — Refinement: tree-aware validation. `validate.sx` now flattens the
  whole block tree (descending into `CtSection` children, guarding malformed
  non-list children) so field checks and duplicate-id detection cover nested
  blocks and span section boundaries; added a `section` field-type case. Inline
  tree detection (class + st-iv-get) keeps it free of a section.sx dependency.
  +6 tests; suite 416/416.
- 2026-06-07 — Extension: nested block trees (`section.sx`). `CtSection` is a
  block whose `children` ivar is a list of blocks (incl. nested sections →
  arbitrary depth), turning the flat document into the ordered TREE from the
  architecture sketch. Self-contained: it answers asHTML/asSx/asText/asMarkdown:
  by folding children's renderings (pure polymorphic recursion — no changes to
  block.sx/render.sx). `mk-section`, `section-children`, `section-append` (cow),
  and tree traversal `doc-deep-find` / `doc-tree-ids` / `doc-tree-count` that
  descend into sections. 25 tests; suite 410/410.
- 2026-06-07 — Extension: plain-text render + excerpts (`text.sx`). Fourth
  boundary format via polymorphic `asText` (heading/text/code/quote→text,
  image→alt, embed/divider→"", list→", "-joined); the document joins non-empty
  child texts with a space. `content/render doc "text"`, `content/text`,
  `content/excerpt doc n` (first n chars + "…" if truncated). For previews,
  meta-descriptions, search indexing. 20 tests; suite 385/385.
- 2026-06-07 — Extension: document metadata (`meta.sx`). CtDoc gained optional
  title/slug/tags ivars (declared in doc.sx, default nil/empty, no effect on
  block ops). Reads via message dispatch; copy-on-write setters
  (`doc-with-title/slug/tags`, `doc-add-tag`, `doc-with-meta`, `doc-new-meta`)
  and `content/*` aliases; `doc-meta` returns the metadata dict. Ghost adapter
  now carries `:title` through import/export/round-trip. 27 tests; suite 365/365.
- 2026-06-07 — Extension: snapshot cache over op-log replay (`snapshot.sx`).
  Snapshots are a cache, never primary state — the log stays the source of truth.
  `content/snapshot!` stores a materialised head at a seq in the persist KV;
  `content/head-cached` / `content/at-cached` start from the nearest snapshot and
  replay only the tail, returning a document IDENTICAL to a full replay (tests
  assert transparency before/after snapshot, across versions, and after
  drop-snapshot fallback). `content/has-snapshot?` / `snapshot-seq` /
  `drop-snapshot!`. 20 tests; suite 338/338.
- 2026-06-07 — Extension: Markdown import adapter (`md-import.sx`), inverse of
  asMarkdown. Line-based parser: ATX headings, fenced code (```lang), blockquotes,
  unordered/ordered lists (grouping consecutive items), thematic breaks,
  paragraphs (consecutive plain lines joined with a space). Sequential ids
  b0,b1…. `md/import` / `content/from-markdown` / `markdown-adapter` (import +
  asMarkdown export). Round-trips canonical Markdown (import∘export == identity);
  imported docs pass validation. 24 tests; suite 318/318.
- 2026-06-07 — Extension: document validation (`validate.sx`). `content/validate`
  returns issue dicts `{:id :kind :detail}` (empty = valid); `content/valid?`
  and `content/issue-kinds` convenience. Checks block id (non-empty string),
  per-type required fields/types (heading level number, image src/alt strings,
  list ordered boolean + items list, etc.), unknown block types, and
  document-level duplicate ids. Guards imports/edits/federated input. 17 tests;
  suite 294/294.
- 2026-06-07 — Extension: durable CRDT replication (`crdt-store.sx`), uniting
  Phase 2 (persist) + Phase 3 (CvRDT). Each replica appends its CRDT ops to its
  own stream (`crdt:<doc>:<replica>`); `crdt/replay` folds one log into a state,
  `crdt/converge` merges every replica's replayed state, `crdt/document` /
  `crdt/order` materialise. Converged result is identical regardless of replica
  order or duplicate delivery (join + idempotent apply) → offline-capable,
  eventually-consistent editing. 14 tests; suite 277/277.
- 2026-06-07 — Extension: Markdown render mode (`markdown.sx`). Third boundary
  format alongside asHTML/asSx via the same polymorphic dispatch; blocks answer
  `asMarkdown: nl` (boundary supplies the newline — this Smalltalk dialect has
  no Character newline ctor). `content/render doc "md"`/`"markdown"`/`:md`,
  `content/markdown`, `asMarkdown`. headings (`#`×level), fenced code, `> ` quote,
  `![alt](src)`, `- `/`1. ` lists, `---`; doc joins blocks with a blank line. No
  MD escaping yet. 20 tests; suite 263/263.
- 2026-06-07 — Extension: asSx wire string-escaping. Added `String>>sxEscaped`
  (escapes `\`→`\\` then `"`→`\"`) and routed every `asSx` text/attr/list-item
  through it, so the SX wire format stays valid when content contains quotes or
  backslashes. +5 render tests (expected strings built from `q`/`bs` helpers to
  avoid escaping miscounts). Suite 243/243.
- 2026-06-07 — Extension: HTML escaping at the render boundary. Added
  `String>>htmlEscaped` (recursive char walk escaping & < > ", order-safe so &
  isn't double-escaped) and routed every `asHTML` text/attr through it — heading,
  text, code body + language, quote, image src/alt, embed url, list items.
  Render stays fully polymorphic in Smalltalk; escaping lives at the boundary.
  +8 render tests (incl. `<script>` payloads, attr breakout, ampersand-once).
  asSx wire-escaping deferred to next. Suite 238/238.

- 2026-06-07 — Phase 4 `fed.sx` (**Phase 4 COMPLETE — roadmap done**):
  trust-gated federation. Peer ops carry provenance (`:author`, `:sig` stub);
  none are auto-accepted. The trust gate is a pluggable predicate (acl-on-sx
  hook) with a trusted-actor-list convenience stub. `content/merge-peer[-with]`
  applies only accepted ops through the CvRDT and quarantines the rest
  (`{:state :accepted :rejected}`). Concurrent local/external edits reconcile
  deterministically: same-field LWW by (ts,actor), commutative, idempotent;
  untrusted ops never touch state. 20 tests; suite 230/230.
- 2026-06-07 — Phase 4 `sync.sx` (cb1): external CMS sync via an injected
  adapter. Core defines the shape — `{:import :export}` — and delegates;
  `content/import` / `content/export` / `content/round-trip` know nothing about
  Ghost. A Ghost-flavoured adapter confines all format translation (post
  `:sections` ↔ content blocks, all 8 kinds). Swapping in a stub `raw-adapter`
  works identically. Round-trip (export∘import and import∘export) preserves ids,
  types, fields, order. 14 tests; suite 210/210. Next: trust-gated federation +
  concurrent-external-edit conflict (via CRDT).
- 2026-06-07 — Phase 3 `crdt.sx` (**Phase 3 complete**): collaborative merge as
  a state-based CvRDT. Merge is a join (lub) on a semilattice → commutative,
  associative, idempotent by construction. Ordering = unique dense Logoot
  position keys (cell = (digit actor), lexicographic); presence = OR-tombstones
  (remove-wins); each field = an LWW-Register keyed by logical (ts, actor). Every
  op contributes a PARTIAL element and per-id state is their join, so
  update-/delete-before-insert are not lost. `crdt-materialize` bridges back to a
  Phase-1 `CtDoc` (sort live elements by pos → blocks). Tests prove: ops in any
  order converge, double-apply is a no-op, merge commutes/associates/is
  idempotent, concurrent inserts order deterministically, same-field LWW by
  (ts,actor), disjoint fields both survive, two divergent replicas converge both
  ways. 34 tests; suite 196/196.
- 2026-06-07 — Phase 2 `store.sx` (**Phase 2 complete**): op log + versioning
  over the persist event stream. `content/commit!` appends an edit op as a
  persist event to the doc's stream (`content:<id>`); the log is the source of
  truth. `content/head` / `content/at b id seq` replay the op stream to the
  latest / any version (materialised doc is a cache, never primary state).
  `content/history` returns per-version metadata; `content/diff` /
  `content/diff-versions` report added/removed/changed block ids. Backend is
  injected via `(persist/open)` — content knows nothing about which backend.
  Minimal persist load (event/backend/log/kv/api). 29 tests; suite 162/162.
- 2026-06-07 — Phase 1 `api.sx` (**Phase 1 complete**): `content/*` facade over
  block + doc + render. `content/bootstrap!` registers the hierarchy;
  `content/edit` applies one op or an op stream; `content/render` picks the
  boundary format ("html"/"sx" or keyword). Re-exports `content/new`,
  `content/append`, `content/insert|update|move|delete`, `content/find`, etc.
  `content/op?` distinguishes a single op from a list/block. 26 tests; suite
  133/133. content/history deferred to Phase 2 (needs the persist op log).
- 2026-06-07 — Phase 1 `render.sx`: render boundary as polymorphic message
  dispatch. Every block and `CtDoc` answers `asHTML` / `asSx`; the document
  folds children via Smalltalk `inject:into:` (works on raw SX lists), so
  `(asHTML doc)` / `(asSx doc)` are pure sends with zero type-switching in SX.
  Lists/headings render in Smalltalk source. No HTML escaping yet (noted in
  render.sx — boundary concern before untrusted content). 29 tests; suite
  107/107.
- 2026-06-06 — Phase 1 `doc.sx`: ordered block document (`CtDoc`) as a
  Smalltalk object holding an ordered block sequence. Edit ops are data dicts
  (`insert`/`update`/`move`/`delete`) with `op-*` constructors; `doc-apply` /
  `doc-apply-all` interpret an op stream, each returning a NEW document (input
  never mutated → replay-safe). Structural moves, insert-after/at, find/index,
  immutability all tested. 40 tests; suite 78/78.
- 2026-06-06 — Phase 1 `block.sx`: typed block objects as Smalltalk instances
  (`CtBlock` hierarchy: text/heading/code/quote/image/embed/divider/list).
  Type tag + accessors are message sends (polymorphic dispatch); fields are
  immutable copy-on-write via functional `st-iv-set!` (history-safe). Added
  `mk-*` constructors, `block?` predicate, `lib/content/conformance.sh` +
  scoreboard. 38/38.

## Blockers

- Smalltalk-only load chain (tokenizer/parser/runtime/eval) does **not** load
  `lib/r7rs.sx`/`spec/stdlib.sx`, so r7rs aliases (`car`/`cdr`/`null?`) are
  absent. Use base SX primitives (`first`/`rest`/`(= (len x) 0)`) in
  `lib/content/**`. Not a substrate bug — just the load surface.