yrby 0.5.0-x86_64-linux → 0.6.0-x86_64-linux

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA256:
3
- metadata.gz: 231867b33dd1e931ddd90b1b8c6dc2ba1b57d2959038bf4bc5e3cf0c18e1974f
4
- data.tar.gz: 28a227ff609a340c2ec9e52eac3ff5fa0a8853969fbf2aa0b1d3a93d9274f4ef
3
+ metadata.gz: 79698fd2fb1bd27630f886cb29bb5f0bf6a5bacdcfaf0104cfa93e2b82ac0cce
4
+ data.tar.gz: 9b527fc536ca22c48b1800529d98d5887d4414d4d45c58fb82c6c4c8a1c5d3bc
5
5
  SHA512:
6
- metadata.gz: f1410e692049f9c97a34cda63c174bbd9cf3a72061c1a43db121e1e15b72b9120663129e1bea81c188d65d9255df161b7b19c4c9b3f055960d93faf0baf68ee5
7
- data.tar.gz: b1057a860265ce47641c11560a5c2eee9ce10acb676c161847484fb8247621288be44899cc95f86c2ccd49bbd8d5699858935e19a62791feff83c8e21eed91f6
6
+ metadata.gz: 580f47b1edb62176cec4d081a063a642fc6ebde8dee413c22dde243cdb430c430a118dc3d72123aba015661c58f3924c9cd56514fb6e8f57643ef38dc28295da
7
+ data.tar.gz: 3c64aeee5940a2183198362c04a2cdb508263fc24143bae33310f84a91a0a822c9fced780047da543288f4a3e25c8524edcc2d73828068136eebf4a02d12faff
data/CHANGELOG.md CHANGED
@@ -4,7 +4,68 @@ All notable changes to this project are documented here. The format is based on
4
4
  [Keep a Changelog](https://keepachangelog.com/en/1.1.0/), and this project aims
5
5
  to follow [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
6
6
 
7
- ## [Unreleased]
7
+ ## [0.6.0] - 2026-07-11
8
+
9
+ ### Added
10
+
11
+ - **Custom render rules for `Y::Lexical` and `Y::ProseMirror`.** Both
12
+ renderers now take a block registering rules per node type —
13
+ `rules.node "callout", tag: "aside"` for markup-as-data, a Ruby block for
14
+ logic — plus `nodes:`/`marks:` keywords as the equivalent data form, to
15
+ render node types the pinned schemas don't know or to override how a
16
+ built-in renders.
17
+ Declarative rules (`tag`/`attrs`/`text`/`contains`, with templates mixing
18
+ literals and attribute references) render natively at full speed. Callback
19
+ rules run a Ruby block per node, receiving its type, attributes,
20
+ already-rendered children, and `child_types` (its element/block children by
21
+ type — the structural facts behind gallery counts and nested-list classes).
22
+ The block runs after the document read has finished, never while the doc is
23
+ locked, so it can safely read or write the same doc. Blocks are proven
24
+ sufficient for whole schemas: the gem's own editor schemas (`Y::Lexxy`,
25
+ `Y::Tiptap`) ship through this API, and the fixture tests hold their
26
+ output byte-identical to a live editor's. With no callback rules the
27
+ render path is unchanged, byte for byte. See "Custom nodes and marks" in
28
+ the README.
29
+ - **`Y::Lexical#node_types` / `Y::ProseMirror#node_types` — schema
30
+ discovery.** Ask a real document which node types it holds and what they
31
+ look like: counts, attribute names as stored, child types, whether text
32
+ runs appear, and whether a builtin or one of your rules already handles
33
+ each ("handled" nil marks what still needs a rule). Editors store names
34
+ you'd never guess; this is how you find them.
35
+ - `Y::RenderRules.escape_text` / `escape_attr` — the exact escaping the
36
+ native renderers use, for blocks that build markup from stored values
37
+ (ERB's `html_escape` also rewrites apostrophes, which breaks byte parity
38
+ with editor output).
39
+
40
+ ### Changed
41
+
42
+ - **Lexical rendering is now two classes: `Y::Lexical` (core Lexical) and
43
+ `Y::Lexxy` (core plus the Lexxy schema as render rules).** Stock Lexical
44
+ has no canonical serializer, so the editor-specific class carries the
45
+ editor's name — `Y::Lexxy.new(doc).to_html` is the byte-parity call for
46
+ Lexxy/Rails apps, and `Y::Lexical` is the base any other Lexical editor
47
+ extends with its own rules. The
48
+ native side renders core structure — paragraphs, headings, quotes, code,
49
+ lists, tables, links, the full text-format model. Lexxy's own node types
50
+ (attachments, galleries, `early_escape_code`, `horizontal_divider`) and its
51
+ decorations of core nodes (the table figure wrapper, header-cell styling,
52
+ the nested-list-item class) are rules applied beneath the app's, on the
53
+ same extension API — the gem's Lexxy support is the API's first consumer.
54
+ Output is unchanged: the fixture tests still hold `to_html` byte-identical
55
+ to a live editor's serialized value, now through the extension path. An
56
+ unknown Lexical container also degrades better: its block children render
57
+ without an invented wrapper instead of being dropped.
58
+ - **ProseMirror rendering gets the same split: `Y::ProseMirror` (core
59
+ ProseMirror) and `Y::Tiptap` (core plus Tiptap's extension nodes as render
60
+ rules).** `Y::Tiptap.new(doc).to_html` is the byte-parity call for Tiptap
61
+ apps. The native side renders prosemirror-schema-basic plus the
62
+ prosemirror-tables family; Tiptap's extension nodes — task lists, mentions,
63
+ the details family — are `Y::Tiptap::NODES` rules. Marks stay native in the
64
+ base class: mark rendering (nesting order, `textStyle` CSS, `code`
65
+ exclusivity) runs through text-run machinery node rules don't reach, so
66
+ `Y::ProseMirror` still renders Tiptap's full mark set and `rules.mark`
67
+ overrides individual marks. Output through `Y::Tiptap` is unchanged, held
68
+ byte-identical to a live editor's `getHTML()` by the fixture tests.
8
69
 
9
70
  ## [0.5.0] - 2026-07-08
10
71
 
data/README.md CHANGED
@@ -129,6 +129,27 @@ The rest of the dev setup, plus the demo, is in [CONTRIBUTING.md](CONTRIBUTING.m
129
129
  and the test/load suites.
130
130
  - [CHANGELOG.md](CHANGELOG.md) and [CONTRIBUTING.md](CONTRIBUTING.md).
131
131
 
132
+ ## Editors
133
+
134
+ yrby syncs opaque Yjs updates, so it works with any editor that has a Yjs
135
+ binding. The demo app runs four, and CI drives each one in real Chrome:
136
+ concurrent typing with every keystroke accounted for, remote cursors,
137
+ local-only undo, and byte parity between the server-side renderers and the
138
+ editor's own serializer. Each page is a working integration to copy from:
139
+
140
+ | Editor | Yjs binding | Demo code |
141
+ |---|---|---|
142
+ | [Tiptap](https://tiptap.dev) (v2) | `@tiptap/extension-collaboration` | [`app.js`](examples/actioncable-demo/frontend/src/app.js) |
143
+ | [Lexxy](https://github.com/basecamp/lexxy) (Lexical) | [`lexxy-realtime`](https://www.npmjs.com/package/lexxy-realtime) | [`lexxy.js`](examples/actioncable-demo/frontend/src/lexxy.js) |
144
+ | [Rhino Editor](https://github.com/KonnorRogers/rhino-editor) (Tiptap 3) | `@tiptap/extension-collaboration` + `-caret` | [`rhino.js`](examples/actioncable-demo/frontend/src/rhino.js) |
145
+ | [CodeMirror 6](https://codemirror.net) | `y-codemirror.next` | [`codemirror.js`](examples/actioncable-demo/frontend/src/codemirror.js) |
146
+
147
+ The demo also syncs plain Yjs shapes with no editor at all — a whiteboard
148
+ on a `Y.Map`, a kanban board on a `Y.Array`, a co-filled form — over the
149
+ same channel. The demo README's "Using this in your own app" section has
150
+ the integration recipe, and its `NoteMaterializer` shows how to render a
151
+ document to ActionText server-side with `Y::Tiptap` or `Y::Lexxy`.
152
+
132
153
  ## Usage
133
154
 
134
155
  ### Doc (Low-Level Document Sync)
@@ -188,18 +209,20 @@ guarantees keep serving safe:
188
209
 
189
210
  ### Rendering to HTML
190
211
 
191
- Two schema-pinned renderers turn a collaborative document into HTML on the
192
- server, with no Node process or headless editor: `Y::ProseMirror` for
193
- ProseMirror/Tiptap documents and `Y::Lexical` for
194
- [Lexxy](https://github.com/basecamp/lexxy) (Lexical) documents. Each returns
195
- `nil` for a root that belongs to the other schema.
212
+ Schema-pinned renderers turn a collaborative document into HTML on the
213
+ server, with no Node process or headless editor. Each is an editor-specific
214
+ class (byte-for-byte with that editor's own serializer) built on a core base
215
+ any other editor extends with rules: `Y::Tiptap` on `Y::ProseMirror` for
216
+ ProseMirror documents, and `Y::Lexxy` (the
217
+ [Lexxy](https://github.com/basecamp/lexxy) editor) on `Y::Lexical`. Each
218
+ returns `nil` for a root that belongs to the other schema.
196
219
 
197
- #### `Y::ProseMirror`
220
+ #### `Y::Tiptap` (and `Y::ProseMirror`, its base)
198
221
 
199
222
  ```ruby
200
- prosemirror = Y::ProseMirror.new(doc)
201
- prosemirror.to_html # the "default" fragment (Tiptap's default root)
202
- prosemirror.to_html("content") # or another XML root
223
+ tiptap = Y::Tiptap.new(doc)
224
+ tiptap.to_html # the "default" fragment (Tiptap's default root)
225
+ tiptap.to_html("content") # or another XML root
203
226
  ```
204
227
 
205
228
  The output matches Tiptap's own `getHTML()`, checked byte-for-byte in the tests
@@ -214,17 +237,30 @@ tables, text styles (color, font family), and every text mark. A table renders
214
237
  as semantic `<table><tbody>`, without the column-width styling Tiptap's editor
215
238
  view adds.
216
239
 
217
- #### `Y::Lexical`
240
+ The support is layered like the Lexical side: `Y::ProseMirror` covers core
241
+ ProseMirror natively — prosemirror-schema-basic plus the prosemirror-tables
242
+ family — and Tiptap's extension nodes (task lists, mentions, the details
243
+ family) are `Y::Tiptap`'s rule set (`Y::Tiptap::NODES`), built on the
244
+ extension API below. Marks stay in the base: mark rendering (nesting order,
245
+ `textStyle` CSS, `code` exclusivity) runs through native text-run machinery
246
+ that node rules don't reach, so `Y::ProseMirror` renders Tiptap's mark set
247
+ as-is and `rules.mark` overrides individual marks.
248
+
249
+ #### `Y::Lexxy` (and `Y::Lexical`, its base)
218
250
 
219
251
  ```ruby
220
- lexical = Y::Lexical.new(doc)
221
- lexical.to_html # the "root" fragment (Lexical's default root name)
222
- lexical.to_html("notepad") # or another XML root
252
+ lexxy = Y::Lexxy.new(doc)
253
+ lexxy.to_html # the "root" fragment (Lexical's default root name)
254
+ lexxy.to_html("notepad") # or another XML root
223
255
  ```
224
256
 
225
257
  The HTML is identical to what a `lexxy-editor` submits to Rails (its `value`).
226
258
  The tests check this byte-for-byte against a document captured from a real
227
- editor.
259
+ editor. Stock Lexical has no canonical serializer — every editor configures
260
+ its own — so the editor-specific class carries the editor's name, and
261
+ `Y::Lexical` is the core-Lexical base: paragraphs, headings, quotes, code,
262
+ lists, tables, links, and the full text-format model, for any other Lexical
263
+ editor to extend with rules.
228
264
 
229
265
  It handles the whole Lexxy 0.9.x node set: paragraphs, headings, every text
230
266
  format and their combinations, links, the four list types and nesting,
@@ -233,8 +269,178 @@ header cells, image galleries, and ActionText attachments (uploads and
233
269
  mentions both emit `<action-text-attachment>` elements that ActionText can
234
270
  re-render).
235
271
 
236
- In both renderers an unknown node keeps its content — text falls back to a
237
- plain paragraph rather than disappearing.
272
+ Internally that support is layered: `Y::Lexical` covers core Lexical
273
+ structure natively, and everything Lexxy adds — its node types (attachments,
274
+ galleries) and its decorations of core nodes (the table wrapper, header-cell
275
+ styling, nested-list classes) — is `Y::Lexxy`'s rule set
276
+ (`Y::Lexxy::NODES`), built on the extension API below. The gem's own Lexxy
277
+ support is the API's first consumer: an app rule for one of those types
278
+ simply replaces it.
279
+
280
+ In both renderers an unknown node keeps its content — text and nested blocks
281
+ fall back to readable markup rather than disappearing.
282
+
283
+ #### Custom nodes and marks
284
+
285
+ The built-in schemas are pinned to what Tiptap and Lexxy ship, but apps add
286
+ their own node types. Both renderers take rules for them. A rule is checked
287
+ before the built-in schema, so it can add a node type or replace how a
288
+ built-in renders.
289
+
290
+ Rules register in a block — one `rules.node` call per type. A declarative
291
+ rule is markup as data, rendered natively:
292
+
293
+ ```ruby
294
+ tiptap = Y::Tiptap.new(doc) do |rules|
295
+ rules.node "callout", tag: "aside",
296
+ attrs: { "class" => ["callout callout--", :kind] },
297
+ contains: :blocks
298
+ end
299
+ ```
300
+
301
+ `tag` names the element. `attrs` values are templates: a string is a literal,
302
+ a symbol reads that attribute off the node, an array concatenates both kinds;
303
+ an attribute that resolves empty is left out. `text` (same template form)
304
+ emits literal text content. `contains` declares what lives inside the node — `:inline` (formatted text,
305
+ the default), `:blocks` (child block nodes — a container), or `:none` (a
306
+ leaf). `void: true` skips the closing tag.
307
+
308
+ You don't have to guess any of those names or shapes. Editors store types
309
+ and attributes under names you'd never predict (Rhino's strike mark is
310
+ `rhino-strike`; Lexical prefixes its own props `__`), so ask a real
311
+ document instead — make one in your editor using your custom node, then:
312
+
313
+ ```ruby
314
+ Y::Tiptap.new(doc).node_types
315
+ # => { "callout" => { "count" => 2, "attrs" => ["kind"],
316
+ # "children" => ["paragraph"], "text" => false,
317
+ # "handled" => nil },
318
+ # "paragraph" => { ..., "handled" => "builtin" } }
319
+ ```
320
+
321
+ `handled` nil marks the types that still need a rule; `attrs` are the stored
322
+ names your templates and blocks will read; `children` plus `text` is how you
323
+ pick `contains:` (child block types → `:blocks`; text → `:inline`).
324
+
325
+ When markup-as-data isn't enough, give the node a block:
326
+
327
+ ```ruby
328
+ lexical = Y::Lexical.new(doc) do |rules|
329
+ rules.node "video_embed" do |node|
330
+ src = ERB::Util.html_escape(node.attrs["__src"])
331
+ %(<video controls src="#{src}"></video>)
332
+ end
333
+ end
334
+ ```
335
+
336
+ The block gets the node's type, its stored attributes, `node.content` — the
337
+ children, already rendered to HTML — and `node.child_types`, the node's
338
+ element/block children by type, in document order. `child_types` answers the
339
+ structural questions attributes can't: how many images a gallery holds, or
340
+ whether a list item carries a nested list. Whatever the block returns is
341
+ spliced into the output as-is: it's trusted HTML, so escape any values you
342
+ interpolate. To set the content mode for a callback, give the node both —
343
+ `rules.node "embed", contains: :blocks do |node| ... end`.
344
+
345
+ Callbacks never run while the document is locked. The render finishes first
346
+ (inside one read transaction, GVL released), then the blocks run and their
347
+ output is spliced in — so a callback can safely read or even write the same
348
+ doc. With no callback rules, `to_html` skips the splicing entirely.
349
+
350
+ Blocks are the escape hatch for everything the declarative form can't say,
351
+ and they're proven sufficient: `Y::Lexxy` and `Y::Tiptap` are themselves
352
+ built on this API (`lib/y/lexxy.rb`, `lib/y/tiptap.rb`) — simple nodes as
353
+ declarative hashes, everything with logic as plain methods mapped by node
354
+ type (a `Method` responds to `call` like any lambda) — and the fixture tests
355
+ hold their output byte-identical to a live editor's.
356
+
357
+ The ProseMirror side also takes custom marks:
358
+
359
+ ```ruby
360
+ tiptap = Y::Tiptap.new(doc) do |rules|
361
+ rules.mark "comment", tag: "span", attrs: { "data-comment-id" => :id }
362
+ end
363
+ ```
364
+
365
+ Symbol refs resolve against the mark's own attributes. A custom mark wraps
366
+ outside every built-in mark; several on one run nest alphabetically. A rule
367
+ for a built-in mark name (`"bold"`) replaces its built-in tag.
368
+
369
+ ##### Worked examples
370
+
371
+ A video-embed node from an app's Tiptap extension — a type the pinned schema
372
+ has never heard of:
373
+
374
+ ```ruby
375
+ tiptap = Y::Tiptap.new(doc) do |rules|
376
+ rules.node "videoEmbed" do |node|
377
+ src = ERB::Util.html_escape(node.attrs["src"])
378
+ title = ERB::Util.html_escape(node.attrs["title"] || "Video")
379
+ %(<figure class="video"><iframe src="#{src}" title="#{title}" allowfullscreen></iframe></figure>)
380
+ end
381
+ end
382
+ ```
383
+
384
+ Resolving mentions against the database. Blocks run after the document read
385
+ has finished, so hitting ActiveRecord (or the doc itself) inside one is safe:
386
+
387
+ ```ruby
388
+ tiptap = Y::Tiptap.new(doc) do |rules|
389
+ rules.node "mention" do |node|
390
+ user = User.find_by(id: node.attrs["id"])
391
+ next "<span>@unknown</span>" unless user
392
+
393
+ %(<a class="mention" href="/users/#{user.id}">@#{ERB::Util.html_escape(user.handle)}</a>)
394
+ end
395
+ end
396
+ ```
397
+
398
+ Overriding a shipped rule — rendering Lexxy uploads as real image markup
399
+ instead of the `<action-text-attachment>` elements ActionText re-renders:
400
+
401
+ ```ruby
402
+ lexxy = Y::Lexxy.new(doc) do |rules|
403
+ rules.node "action_text_attachment" do |node|
404
+ src = ERB::Util.html_escape(node.attrs["src"])
405
+ alt = ERB::Util.html_escape(node.attrs["altText"].to_s)
406
+ caption = node.attrs["caption"].to_s
407
+ html = %(<img src="#{src}" alt="#{alt}" loading="lazy">)
408
+ html += "<figcaption>#{ERB::Util.html_escape(caption)}</figcaption>" unless caption.empty?
409
+ "<figure>#{html}</figure>"
410
+ end
411
+ end
412
+ ```
413
+
414
+ Markup that depends on structure — `node.child_types` lists the node's
415
+ element/block children in document order, so a layout container can size
416
+ itself by its column count while the columns themselves stay declarative:
417
+
418
+ ```ruby
419
+ tiptap = Y::Tiptap.new(doc) do |rules|
420
+ rules.node "columns", contains: :blocks do |node|
421
+ %(<div class="columns columns--#{node.child_types.length}">#{node.content}</div>)
422
+ end
423
+ rules.node "column", tag: "div", attrs: { "class" => "column" }, contains: :blocks
424
+ end
425
+ ```
426
+
427
+ Content-aware overrides — dropping the empty paragraphs an editor keeps
428
+ around the cursor, since `node.content` arrives already rendered:
429
+
430
+ ```ruby
431
+ lexical = Y::Lexical.new(doc) do |rules|
432
+ rules.node "paragraph" do |node|
433
+ node.content.empty? ? "" : "<p>#{node.content}</p>"
434
+ end
435
+ end
436
+ ```
437
+
438
+ For a larger reference, the gem's own editor schemas ship this way — see
439
+ `Y::Lexxy::NODES` in `lib/y/lexxy.rb` (declarative hashes for the simple
440
+ nodes, a plain method per node that needs logic — galleries, list items,
441
+ header cells, both attachment types — mapped with `method(:name)`) and
442
+ `Y::Tiptap::NODES` in `lib/y/tiptap.rb` (task lists, mentions, the details
443
+ family).
238
444
 
239
445
  ### Protocol codec (module functions)
240
446
 
data/lib/y/3.4/yrby.so CHANGED
Binary file
data/lib/y/4.0/yrby.so CHANGED
Binary file
data/lib/y/lexxy.rb ADDED
@@ -0,0 +1,100 @@
1
+ # frozen_string_literal: true
2
+
3
+ module Y
4
+ # The Lexxy renderer: Y::Lexical (core Lexical) plus the Lexxy-specific
5
+ # schema, applied beneath the app's rules — an app rule for one of these
6
+ # types simply replaces it. This is the byte-parity class: the fixture
7
+ # tests hold `Y::Lexxy.new(doc).to_html` identical to a live editor's own
8
+ # serialized value.
9
+ #
10
+ # The schema doubles as the reference for augmenting a renderer: simple
11
+ # nodes are declarative hashes, nodes with logic are plain methods mapped
12
+ # in NODES.
13
+ class Lexxy < Lexical
14
+ # A cursor-placement placeholder; empty ones export to nothing.
15
+ def self.provisional_paragraph(node)
16
+ node.content.empty? ? "" : "<p>#{node.content}</p>"
17
+ end
18
+
19
+ # Adjacent previewable images; the class carries the image count
20
+ # (ActionText's convention).
21
+ def self.gallery(node)
22
+ %(<div class="attachment-gallery attachment-gallery--#{node.child_types.length}">#{node.content}</div>)
23
+ end
24
+
25
+ # Lexxy wraps tables in a styled figure.
26
+ def self.table(node)
27
+ %(<figure class="lexxy-content__table-wrapper"><table><tbody>#{node.content}</tbody></table></figure>)
28
+ end
29
+
30
+ # Class + background match Lexxy's own header-cell export.
31
+ def self.table_cell(node)
32
+ header = node.attrs["__headerState"].is_a?(Numeric) && node.attrs["__headerState"].positive?
33
+ return "<td>#{node.content}</td>" unless header
34
+
35
+ style = %(style="background-color: rgb(242, 243, 245);")
36
+ %(<th class="lexxy-content__table-cell--header" #{style}>#{node.content}</th>)
37
+ end
38
+
39
+ # Attribute order follows Lexxy's export — checked items put aria-checked
40
+ # before value; items holding a nested list append the
41
+ # lexxy-nested-listitem class after value.
42
+ def self.list_item(node)
43
+ out = +"<li"
44
+ checked = node.attrs["__checked"]
45
+ out << %( aria-checked="#{checked}") unless checked.nil?
46
+ out << %( value="#{node.attrs["__value"] || 1}")
47
+ out << %( class="lexxy-nested-listitem") if node.child_types.include?("list")
48
+ "#{out}>#{node.content}</li>"
49
+ end
50
+
51
+ # An upload, in the exact shape ActionText round-trips: attribute order
52
+ # and presence mirror Lexxy's exportDOM (nulls omitted, `previewable`
53
+ # only when true, `presentation="gallery"` always).
54
+ def self.upload(node)
55
+ out = +"<action-text-attachment"
56
+ out << attachment_attr(node, "sgid", "sgid")
57
+ out << %( previewable="true") if node.attrs["previewable"] == true
58
+ [%w[url src], %w[alt altText], %w[caption caption],
59
+ %w[content-type contentType], %w[filename fileName],
60
+ %w[filesize fileSize], %w[width width], %w[height height]]
61
+ .each { |html_name, stored| out << attachment_attr(node, html_name, stored) }
62
+ %(#{out} presentation="gallery"></action-text-attachment>)
63
+ end
64
+
65
+ # A content attachment (mention, embed): `content` carries the escaped
66
+ # inner HTML; `plainText` is not exported.
67
+ def self.mention(node)
68
+ out = +"<action-text-attachment"
69
+ out << attachment_attr(node, "sgid", "sgid")
70
+ out << attachment_attr(node, "content", "innerHtml")
71
+ out << attachment_attr(node, "content-type", "contentType")
72
+ "#{out}></action-text-attachment>"
73
+ end
74
+
75
+ # A stored nil (unset) is skipped; a stored empty string still emits.
76
+ def self.attachment_attr(node, html_name, stored)
77
+ return "" if node.attrs[stored].nil?
78
+
79
+ %( #{html_name}="#{RenderRules.escape_attr(node.attrs[stored])}")
80
+ end
81
+
82
+ NODES = {
83
+ # Lexxy's replacement for Lexical's CodeNode.
84
+ "early_escape_code" => { tag: "pre", attrs: { "data-language" => :language } },
85
+ "horizontal_divider" => { tag: "hr", void: true },
86
+ "provisonal_paragraph" => method(:provisional_paragraph), # (sic: Lexxy's spelling)
87
+ "image_gallery" => method(:gallery),
88
+ "table" => { contains: :blocks, render: method(:table) },
89
+ "wrapped_table_node" => { contains: :blocks, render: method(:table) },
90
+ "tablecell" => { contains: :blocks, render: method(:table_cell) },
91
+ "listitem" => { contains: :blocks, render: method(:list_item) },
92
+ "action_text_attachment" => method(:upload),
93
+ "custom_action_text_attachment" => method(:mention)
94
+ }.freeze
95
+
96
+ def initialize(doc, nodes: {}, &)
97
+ super(doc, nodes: NODES.merge(nodes.transform_keys(&:to_s)), &)
98
+ end
99
+ end
100
+ end
@@ -0,0 +1,282 @@
1
+ # frozen_string_literal: true
2
+
3
+ require "json"
4
+
5
+ module Y
6
+ # Custom render rules for Y::Lexical and Y::ProseMirror.
7
+ #
8
+ # Both renderers accept a `nodes:` hash mapping a node type (Lexical's
9
+ # `__type`, ProseMirror's node name) to a rule; Y::ProseMirror also accepts
10
+ # `marks:`. A rule is consulted before the built-in schema, so it can add a
11
+ # custom node or override a built-in one.
12
+ #
13
+ # The usual way in is the block form — one `rules.node` call per type,
14
+ # keyword options for markup-as-data, a Ruby block for logic (see Builder
15
+ # below). `contains:` says what's inside the node: :inline (formatted
16
+ # text, the default), :blocks (child block nodes), or :none (a leaf). The
17
+ # nodes:/marks: keywords take the same rules as plain hashes, for shipping
18
+ # rule sets as data.
19
+ #
20
+ # Two kinds of rule:
21
+ #
22
+ # - Declarative (keyword options / a Hash): markup as data, rendered
23
+ # natively at full speed.
24
+ # Y::ProseMirror.new(doc) do |rules|
25
+ # rules.node "callout", tag: "aside",
26
+ # attrs: { "class" => ["callout callout--", :kind] },
27
+ # contains: :blocks
28
+ # end
29
+ # `tag` is the element; `attrs` values are templates — a String literal,
30
+ # a Symbol referencing one of the node's stored attributes, or an Array
31
+ # mixing both (an attribute that resolves empty is omitted); `text` is a
32
+ # template for literal text content; `void: true` emits no closing tag;
33
+ # `contains` declares what lives inside the node and renders there:
34
+ # :inline (default) for formatted text, :blocks for child block nodes
35
+ # (a container), :none for a leaf.
36
+ #
37
+ # - Callback (a block; in the hash form, a callable or `render:` plus
38
+ # `contains`):
39
+ # Y::Lexical.new(doc) do |rules|
40
+ # rules.node "video_embed" do |node|
41
+ # %(<video src="#{ERB::Util.html_escape(node.attrs["src"])}"></video>)
42
+ # end
43
+ # end
44
+ # The block runs after the document read has finished (never while the
45
+ # document is locked) and receives a RenderRules::Node with the node's
46
+ # type, stored attributes, children already rendered to HTML, and
47
+ # child_types (its element/block children by type). Its return value is
48
+ # spliced in verbatim — it is trusted HTML, so escape any attribute
49
+ # values you interpolate.
50
+ #
51
+ # Mark rules (ProseMirror only) are declarative: `tag` plus `attrs`
52
+ # templates whose Symbol refs resolve against the mark's own attributes. A
53
+ # custom mark wraps outside every built-in mark; several custom marks nest
54
+ # alphabetically. A rule for a built-in mark's stored name replaces its
55
+ # wrap (the markup changes, the semantics don't — an overridden code mark
56
+ # still excludes the other formatting).
57
+ module RenderRules
58
+ # What a callback receives. `attrs` keys are as stored (Lexical's own
59
+ # props keep their "__" prefix); `content` is the node's children,
60
+ # already rendered to an HTML string; `child_types` lists the node's
61
+ # element/block children by type, in document order — the structural
62
+ # facts attrs and content can't answer (a gallery's image count, whether
63
+ # a list item holds a nested list).
64
+ Node = Data.define(:type, :attrs, :content, :child_types)
65
+
66
+ module_function
67
+
68
+ # Compile the user-facing config into [rules_json, callbacks]. Structural
69
+ # validation happens in the native parser, which raises ArgumentError.
70
+ def compile(nodes, marks)
71
+ callbacks = {}
72
+ spec = {}
73
+ unless nodes.empty?
74
+ spec["nodes"] = nodes.to_h do |type, rule|
75
+ [type.to_s, compile_node(type.to_s, rule, callbacks)]
76
+ end
77
+ end
78
+ unless marks.empty?
79
+ spec["marks"] = marks.to_h do |name, rule|
80
+ [name.to_s, compile_mark(name.to_s, rule)]
81
+ end
82
+ end
83
+ [JSON.generate(spec), callbacks]
84
+ end
85
+
86
+ def compile_node(type, rule, callbacks)
87
+ if rule.respond_to?(:call)
88
+ callbacks[type] = rule
89
+ return { "callback" => true }
90
+ end
91
+ raise ArgumentError, "rule for #{type.inspect} must be a Hash or a callable" unless rule.is_a?(Hash)
92
+
93
+ if rule[:render]
94
+ callbacks[type] = rule[:render]
95
+ compiled = { "callback" => true }
96
+ compiled["content"] = rule[:contains].to_s if rule[:contains]
97
+ return compiled
98
+ end
99
+ compile_declarative_node(rule)
100
+ end
101
+
102
+ def compile_declarative_node(rule)
103
+ compiled = {}
104
+ compiled["tag"] = rule[:tag].to_s if rule[:tag]
105
+ compiled["void"] = true if rule[:void]
106
+ compiled["attrs"] = compile_attrs(rule[:attrs]) if rule[:attrs]
107
+ compiled["text"] = compile_parts(rule[:text]) if rule[:text]
108
+ compiled["content"] = rule[:contains].to_s if rule[:contains]
109
+ compiled
110
+ end
111
+
112
+ def compile_mark(name, rule)
113
+ raise ArgumentError, "mark rule for #{name.inspect} must be a Hash" unless rule.is_a?(Hash)
114
+
115
+ compiled = {}
116
+ compiled["tag"] = rule[:tag].to_s if rule[:tag]
117
+ compiled["attrs"] = compile_attrs(rule[:attrs]) if rule[:attrs]
118
+ compiled
119
+ end
120
+
121
+ def compile_attrs(attrs)
122
+ attrs.map { |name, template| [name.to_s, compile_parts(template)] }
123
+ end
124
+
125
+ # A template: String literal, Symbol attribute reference, or an Array of
126
+ # both.
127
+ def compile_parts(template)
128
+ Array(template).map do |part|
129
+ case part
130
+ when Symbol then { "ref" => part.to_s }
131
+ else { "lit" => part.to_s }
132
+ end
133
+ end
134
+ end
135
+
136
+ # The escaping the native renderers use, for blocks that build markup
137
+ # from stored values. Text content escapes `&`, `<`, `>` (quotes stay
138
+ # literal, matching the browser serializer); attribute values also
139
+ # escape `"`. Prefer these over ERB::Util.html_escape when byte parity
140
+ # with editor output matters — html_escape also rewrites apostrophes.
141
+ def escape_text(value)
142
+ value.to_s.gsub("&", "&amp;").gsub("<", "&lt;").gsub(">", "&gt;")
143
+ end
144
+
145
+ def escape_attr(value)
146
+ escape_text(value).gsub('"', "&quot;")
147
+ end
148
+
149
+ # Yielded by Y::Lexical.new / Y::ProseMirror.new: register rules one
150
+ # call per node type. Keyword options are the markup-as-data; a block is
151
+ # the logic; both together say what a callback node contains and how to
152
+ # render it:
153
+ #
154
+ # Y::ProseMirror.new(doc) do |rules|
155
+ # rules.node "callout", tag: "aside", contains: :blocks
156
+ # rules.node "video" do |node|
157
+ # %(<video src="#{RenderRules.escape_attr(node.attrs["src"])}"></video>)
158
+ # end
159
+ # rules.node "columns", contains: :blocks do |node|
160
+ # %(<div class="cols--#{node.child_types.length}">#{node.content}</div>)
161
+ # end
162
+ # rules.mark "comment", tag: "span", attrs: { "data-comment-id" => :id }
163
+ # end
164
+ #
165
+ # It compiles to the same rule hashes the nodes:/marks: keywords take, so
166
+ # both forms mean the same thing; the keywords remain the data form for
167
+ # shipping rule sets (Y::Lexxy::NODES is one).
168
+ class Builder
169
+ attr_reader :nodes, :marks
170
+
171
+ def initialize(marks_allowed:)
172
+ @nodes = {}
173
+ @marks = {}
174
+ @marks_allowed = marks_allowed
175
+ end
176
+
177
+ def node(type, **options, &block)
178
+ @nodes[type.to_s] =
179
+ if block && options.empty?
180
+ block
181
+ elsif block
182
+ options.merge(render: block)
183
+ else
184
+ options
185
+ end
186
+ end
187
+
188
+ def mark(name, **options)
189
+ raise ArgumentError, "marks are ProseMirror-only" unless @marks_allowed
190
+
191
+ @marks[name.to_s] = options
192
+ end
193
+ end
194
+
195
+ # Resolve callback segments depth-first, so a callback's `node.content`
196
+ # is finished HTML even when callback nodes nest.
197
+ def splice(segments, callbacks)
198
+ segments.map do |segment|
199
+ next segment if segment.is_a?(String)
200
+
201
+ type, attrs_json, content, child_types = segment
202
+ node = Node.new(type: type, attrs: JSON.parse(attrs_json),
203
+ content: splice(content, callbacks),
204
+ child_types: child_types)
205
+ callbacks.fetch(type).call(node).to_s
206
+ end.join
207
+ end
208
+ end
209
+
210
+ # Y::Lexical and Y::ProseMirror are plain Ruby facades over the native
211
+ # renderers (Y::NativeLexical / Y::NativeProseMirror, private constants):
212
+ # they compile the rules config, hold the callbacks, and splice deferred
213
+ # segments after a render. The native handle does everything else.
214
+ class Lexical
215
+ # `Y::Lexical.new(doc, nodes: { "type" => rule })` — see Y::RenderRules
216
+ # for the rule forms. This is core Lexical only: paragraphs, headings,
217
+ # quotes, code, lists, tables, links, text formatting. Editor-specific
218
+ # nodes arrive as rules — Y::Lexxy subclasses this with the Lexxy schema;
219
+ # a different Lexical editor brings its own rule set the same way.
220
+ def initialize(doc, nodes: {})
221
+ builder = RenderRules::Builder.new(marks_allowed: false)
222
+ yield builder if block_given?
223
+ nodes = nodes.transform_keys(&:to_s).merge(builder.nodes)
224
+ rules_json, @render_callbacks = RenderRules.compile(nodes, {})
225
+ @native = NativeLexical.new(doc, rules_json)
226
+ end
227
+
228
+ def to_html(root = nil)
229
+ result = root.nil? ? @native.to_html : @native.to_html(root)
230
+ return result unless result.is_a?(Array)
231
+
232
+ RenderRules.splice(result, @render_callbacks)
233
+ end
234
+
235
+ # What node types this document actually contains — the discovery aid
236
+ # for writing rules. Facts per type: "count", "attrs" (names as stored),
237
+ # "children" (child node types), "text" (whether it holds text runs),
238
+ # and "handled" ("builtin", "rule", or nil — nil marks the types you
239
+ # still need a rule for). Children plus text is how you pick contains:.
240
+ def node_types(root = nil)
241
+ json = root.nil? ? @native.node_types : @native.node_types(root)
242
+ json && JSON.parse(json)
243
+ end
244
+ end
245
+
246
+ class ProseMirror
247
+ # `Y::ProseMirror.new(doc, nodes: {...}, marks: {...})` — see
248
+ # Y::RenderRules for the rule forms. This is core ProseMirror only:
249
+ # prosemirror-schema-basic plus the prosemirror-tables family, and the
250
+ # full mark set (marks are native — see `rules.mark` for overrides).
251
+ # Editor-specific nodes arrive as rules — Y::Tiptap subclasses this with
252
+ # Tiptap's extension nodes; a different ProseMirror editor brings its
253
+ # own rule set the same way.
254
+ def initialize(doc, nodes: {}, marks: {})
255
+ builder = RenderRules::Builder.new(marks_allowed: true)
256
+ yield builder if block_given?
257
+ nodes = nodes.transform_keys(&:to_s).merge(builder.nodes)
258
+ marks = marks.transform_keys(&:to_s).merge(builder.marks)
259
+ rules_json, @render_callbacks = RenderRules.compile(nodes, marks)
260
+ @native = NativeProseMirror.new(doc, rules_json)
261
+ end
262
+
263
+ def to_html(root = nil)
264
+ result = root.nil? ? @native.to_html : @native.to_html(root)
265
+ return result unless result.is_a?(Array)
266
+
267
+ RenderRules.splice(result, @render_callbacks)
268
+ end
269
+
270
+ # What node types this document actually contains — the discovery aid
271
+ # for writing rules. Facts per type: "count", "attrs" (names as stored),
272
+ # "children" (child node types), "text" (whether it holds text runs),
273
+ # and "handled" ("builtin", "rule", or nil — nil marks the types you
274
+ # still need a rule for). Children plus text is how you pick contains:.
275
+ def node_types(root = nil)
276
+ json = root.nil? ? @native.node_types : @native.node_types(root)
277
+ json && JSON.parse(json)
278
+ end
279
+ end
280
+
281
+ private_constant :NativeLexical, :NativeProseMirror
282
+ end
data/lib/y/tiptap.rb ADDED
@@ -0,0 +1,63 @@
1
+ # frozen_string_literal: true
2
+
3
+ module Y
4
+ # The Tiptap renderer: Y::ProseMirror (core ProseMirror — schema-basic plus
5
+ # tables) plus Tiptap's extension nodes, applied beneath the app's rules —
6
+ # an app rule for one of these types simply replaces it. This is the
7
+ # byte-parity class: the fixture tests hold `Y::Tiptap.new(doc).to_html`
8
+ # identical to a live editor's own `getHTML()`.
9
+ #
10
+ # Tiptap's marks (underline, highlight, sub/superscript, textStyle) render
11
+ # natively in the base class — mark serialization is text-run machinery
12
+ # the rule system can't express.
13
+ class Tiptap < ProseMirror
14
+ # Tiptap's TaskItem markup: the data-checked flag (false when unset), a
15
+ # label wrapping the checkbox, and the item body in a div.
16
+ def self.task_item(node)
17
+ checked = node.attrs["checked"] == true
18
+ out = %(<li data-checked="#{checked}" data-type="taskItem">)
19
+ out << %(<label><input type="checkbox")
20
+ out << %( checked="checked") if checked
21
+ out << "><span></span></label><div>"
22
+ "#{out}#{node.content}</div></li>"
23
+ end
24
+
25
+ # Tiptap's Mention extension (no app-configured HTMLAttributes): data
26
+ # attributes when present, the suggestion char, and @label (falling back
27
+ # to @id) as the text.
28
+ def self.mention(node)
29
+ char = node.attrs["mentionSuggestionChar"] || "@"
30
+ out = +%(<span data-type="mention")
31
+ %w[id label].each do |key|
32
+ next if node.attrs[key].nil?
33
+
34
+ out << %( data-#{key}="#{RenderRules.escape_attr(node.attrs[key])}")
35
+ end
36
+ out << %( data-mention-suggestion-char="#{RenderRules.escape_attr(char)}">)
37
+ out << RenderRules.escape_text(char)
38
+ out << RenderRules.escape_text(node.attrs["label"] || node.attrs["id"] || "")
39
+ "#{out}</span>"
40
+ end
41
+
42
+ # The details family follows tiptap-php's renderHTML (the Tiptap
43
+ # extension is Pro-only, so there's no free getHTML() to capture
44
+ # against).
45
+ def self.details(node)
46
+ open = node.attrs["open"] == true ? %( open="open") : ""
47
+ "<details#{open}>#{node.content}</details>"
48
+ end
49
+
50
+ NODES = {
51
+ "taskList" => { tag: "ul", attrs: { "data-type" => "taskList" }, contains: :blocks },
52
+ "taskItem" => { contains: :blocks, render: method(:task_item) },
53
+ "mention" => method(:mention),
54
+ "details" => { contains: :blocks, render: method(:details) },
55
+ "detailsSummary" => { tag: "summary" },
56
+ "detailsContent" => { tag: "div", attrs: { "data-type" => "detailsContent" }, contains: :blocks }
57
+ }.freeze
58
+
59
+ def initialize(doc, nodes: {}, marks: {}, &)
60
+ super(doc, nodes: NODES.merge(nodes.transform_keys(&:to_s)), marks: marks, &)
61
+ end
62
+ end
63
+ end
data/lib/y/version.rb CHANGED
@@ -1,5 +1,5 @@
1
1
  # frozen_string_literal: true
2
2
 
3
3
  module Y
4
- VERSION = "0.5.0"
4
+ VERSION = "0.6.0"
5
5
  end
data/lib/y.rb CHANGED
@@ -12,6 +12,10 @@ rescue LoadError
12
12
  require_relative "y/yrby"
13
13
  end
14
14
 
15
+ require_relative "y/rendering"
16
+ require_relative "y/lexxy"
17
+ require_relative "y/tiptap"
18
+
15
19
  module Y
16
20
  # Doc, Error, and the protocol module functions are defined in the Rust
17
21
  # extension. The ActionCable integration (Y::ActionCable::Sync) lives in the
metadata CHANGED
@@ -1,14 +1,14 @@
1
1
  --- !ruby/object:Gem::Specification
2
2
  name: yrby
3
3
  version: !ruby/object:Gem::Version
4
- version: 0.5.0
4
+ version: 0.6.0
5
5
  platform: x86_64-linux
6
6
  authors:
7
7
  - JP Camara
8
8
  autorequire:
9
9
  bindir: bin
10
10
  cert_chain: []
11
- date: 2026-07-09 00:00:00.000000000 Z
11
+ date: 2026-07-12 00:00:00.000000000 Z
12
12
  dependencies:
13
13
  - !ruby/object:Gem::Dependency
14
14
  name: minitest
@@ -68,6 +68,9 @@ files:
68
68
  - lib/y.rb
69
69
  - lib/y/3.4/yrby.so
70
70
  - lib/y/4.0/yrby.so
71
+ - lib/y/lexxy.rb
72
+ - lib/y/rendering.rb
73
+ - lib/y/tiptap.rb
71
74
  - lib/y/version.rb
72
75
  - lib/yrby.rb
73
76
  homepage: https://github.com/jpcamara/yrby