4.1: an iterative canonical serializer

corpus/round-trip/combinations/carry-text-attributes.json

@@ -50,25 +50,6 @@
         }
       ],
       "type": "paragraph"
-    },
-    {
-      "content": [
-        {
-          "attrs": {
-            "localId": "01a0a067-68cf-78af-abd6-c660ec0d189b"
-          },
-          "text": "Two ",
-          "type": "text"
-        },
-        {
-          "attrs": {
-            "localId": "01a0a067-68d9-73fa-9b9a-caa42061993c"
-          },
-          "text": "anchors",
-          "type": "text"
-        }
-      ],
-      "type": "paragraph"
     }
   ],
   "type": "doc",

corpus/round-trip/combinations/carry-text-attributes.md

@@ -3,5 +3,3 @@
 Signed by :adf{json="{\"attrs\":{\"localId\":\"01a0a067-68d2-787e-afc5-2459776aa029\"},\"text\":\"the owner\",\"type\":\"text\"}"}
 
 :adf{json="{\"attrs\":{\"localId\":\"01a0a067-68d5-7372-9962-29a039654056\"},\"text\":\"One \",\"type\":\"text\"}"}:adf{json="{\"attrs\":{\"localId\":\"01a0a067-68d5-7372-9962-29a039654056\"},\"text\":\"anchor\",\"type\":\"text\"}"}
-
-:adf{json="{\"attrs\":{\"localId\":\"01a0a067-68cf-78af-abd6-c660ec0d189b\"},\"text\":\"Two \",\"type\":\"text\"}"}:adf{json="{\"attrs\":{\"localId\":\"01a0a067-68d9-73fa-9b9a-caa42061993c\"},\"text\":\"anchors\",\"type\":\"text\"}"}

src/adf/document.ts

@@ -72,7 +72,7 @@ export function isAdfMark(value: unknown): value is AdfMark {
   return !('attrs' in value) || isAttributes(value['attrs'])
 }
 
-export function nodeAttrs(node: { attrs?: AdfAttributes }): AdfAttributes {
+export function nodeAttrs(node: { attrs?: AdfAttributes }): Readonly<AdfAttributes> {
   return node.attrs ?? {}
 }
 

src/adf/editor-normal.test.ts

@@ -2,6 +2,7 @@ import assert from 'node:assert/strict'
 import test from 'node:test'
 
 import type { AdfNode } from './document.ts'
+import type { JsonValue } from '../json-value.ts'
 import { toEditorNormal } from './editor-normal.ts'
 
 test('merges adjacent text nodes carrying identical marks, at every level', () => {
@@ -14,6 +15,7 @@ test('merges adjacent text nodes carrying identical marks, at every level', () =
     { text: 'e', type: 'text' },
     { attrs: { localId: '01a0a06b-5281-7f27-9022-8d3a74b0ab0d' }, text: 'f', type: 'text' },
     { text: 'g', type: 'text' },
+    { attrs: {}, text: 'h', type: 'text' },
   ]
   assert.deepEqual(toEditorNormal({ content: [{ attrs: { panelType: 'info' }, content: [{ content, type: 'paragraph' }], type: 'panel' }], type: 'doc', version: 1 }), {
     content: [
@@ -27,7 +29,7 @@ test('merges adjacent text nodes carrying identical marks, at every level', () =
               { type: 'hardBreak' },
               { text: 'e', type: 'text' },
               { attrs: { localId: '01a0a06b-5281-7f27-9022-8d3a74b0ab0d' }, text: 'f', type: 'text' },
-              { text: 'g', type: 'text' },
+              { text: 'gh', type: 'text' },
             ],
             type: 'paragraph',
           },
@@ -58,7 +60,7 @@ test('reads an empty attrs object, marks array or content array as the absent ke
   assert.deepEqual(toEditorNormal({ content: [], type: 'doc', version: 1 }), { type: 'doc', version: 1 })
 })
 
-test('normalizes a document nesting far past the levels a recursive walk survives', () => {
+test('normalizes blocks and mark attributes nesting far past the levels a recursive walk survives', () => {
   const levels = 100000
   let node: AdfNode = { content: [], type: 'paragraph' }
   for (let level = 0; level < levels; level += 1) node = { content: [node], type: 'blockquote' }
@@ -67,4 +69,9 @@ test('normalizes a document nesting far past the levels a recursive walk survive
   for (; normal?.content !== undefined; depth += 1) normal = normal.content[0]
   assert.equal(depth, levels)
   assert.deepEqual(normal, { type: 'paragraph' })
+  let deep: JsonValue = 1
+  for (let level = 0; level < 2 * levels; level += 1) deep = [deep]
+  const marks = [{ attrs: { deep }, type: 'textColor' }]
+  const merged = toEditorNormal({ content: [{ content: [{ marks, text: 'a', type: 'text' }, { marks, text: 'b', type: 'text' }], type: 'paragraph' }], type: 'doc', version: 1 })
+  assert.deepEqual(merged.content?.[0]?.content?.map((text) => text.text), ['ab'])
 })

src/adf/editor-normal.ts

@@ -11,7 +11,6 @@ export function sameMark(candidate: AdfMark, mark: AdfMark): boolean {
   return markKey(candidate) === markKey(mark)
 }
 
-// AGENTS.md §2: adjacent text nodes carrying identical marks and no attributes are one node.
 export function mergeAdjacentText(nodes: readonly AdfNode[]): AdfNode[] {
   const merged: AdfNode[] = []
   for (const node of nodes) {

src/canonical-json.test.ts

@@ -1,6 +1,7 @@
 import assert from 'node:assert/strict'
 import test from 'node:test'
 
+import type { JsonValue } from './json-value.ts'
 import { serializeCanonicalJson } from './canonical-json.ts'
 
 test('sorts object keys recursively', () => {
@@ -39,3 +40,15 @@ test('leaves non-ASCII raw', () => {
 test('spells scalars in canonical JSON', () => {
   assert.equal(serializeCanonicalJson([null, true, false, 0, -1.5, 'a"b'], 'compact'), '[null,true,false,0,-1.5,"a\\"b"]')
 })
+
+test('spells a value nesting far past the levels a recursive walk survives', () => {
+  const levels = 200000
+  let array: JsonValue = 1
+  let object: JsonValue = 1
+  for (let level = 0; level < levels; level += 1) {
+    array = [array]
+    object = { a: object }
+  }
+  assert.equal(serializeCanonicalJson(array, 'compact'), `${'['.repeat(levels)}1${']'.repeat(levels)}`)
+  assert.equal(serializeCanonicalJson(object, 'compact'), `${'{"a":'.repeat(levels)}1${'}'.repeat(levels)}`)
+})

src/canonical-json.ts

@@ -2,28 +2,42 @@ import type { JsonValue } from './json-value.ts'
 
 export type JsonSpelling = 'compact' | 'two-space'
 
+type Member = { label: string; value: JsonValue }
+
+type Pending = string | { depth: number; value: JsonValue }
+
 export function serializeCanonicalJson(value: JsonValue, spelling: JsonSpelling): string {
-  return serialize(value, spelling === 'compact' ? '' : '  ', 0)
+  const indent = spelling === 'compact' ? '' : '  '
+  const text: string[] = []
+  const pending: Pending[] = [{ depth: 0, value }]
+  for (let next = pending.pop(); next !== undefined; next = pending.pop()) {
+    if (typeof next === 'string') {
+      text.push(next)
+      continue
+    }
+    const { depth, value: held } = next
+    if (Array.isArray(held)) schedule(pending, '[', held.map((item) => ({ label: '', value: item })), ']', indent, depth)
+    else if (held !== null && typeof held === 'object') schedule(pending, '{', objectMembers(held, indent), '}', indent, depth)
+    else text.push(JSON.stringify(held))
+  }
+  return text.join('')
 }
 
-function serialize(value: JsonValue, indent: string, depth: number): string {
-  if (Array.isArray(value)) {
-    if (value.length === 0) return '[]'
-    const items = value.map((item) => serialize(item, indent, depth + 1))
-    return `[${join(items, indent, depth)}]`
-  }
-  if (value !== null && typeof value === 'object') {
-    const keys = Object.keys(value).sort()
-    if (keys.length === 0) return '{}'
+function objectMembers(value: { [key: string]: JsonValue }, indent: string): Member[] {
   const separator = indent === '' ? ':' : ': '
-    const entries = keys.map((key) => `${JSON.stringify(key)}${separator}${serialize(value[key] ?? null, indent, depth + 1)}`)
-    return `{${join(entries, indent, depth)}}`
-  }
-  return JSON.stringify(value)
+  return Object.keys(value)
+    .sort()
+    .map((key) => ({ label: `${JSON.stringify(key)}${separator}`, value: value[key] ?? null }))
 }
 
-function join(parts: readonly string[], indent: string, depth: number): string {
-  if (indent === '') return parts.join(',')
-  const inner = `\n${indent.repeat(depth + 1)}`
-  return `${inner}${parts.join(`,${inner}`)}\n${indent.repeat(depth)}`
+function schedule(pending: Pending[], open: string, members: readonly Member[], close: string, indent: string, depth: number): void {
+  if (members.length === 0) {
+    pending.push(`${open}${close}`)
+    return
+  }
+  const inner = indent === '' ? '' : `\n${indent.repeat(depth + 1)}`
+  const scheduled: Pending[] = []
+  for (const [index, member] of members.entries()) scheduled.push(`${index === 0 ? open : ','}${inner}${member.label}`, { depth: depth + 1, value: member.value })
+  scheduled.push(indent === '' ? close : `\n${indent.repeat(depth)}${close}`)
+  for (const item of scheduled.reverse()) pending.push(item)
 }

src/corpus.test.ts

@@ -96,7 +96,7 @@ for (const directory of roundTripDirectories) {
       assert.ok(isAdfDocument(expected), `${name}.json is not an ADF document`)
       const result = markdownToAdf(readFileSync(join(roundTripRoot, directory, `${name}.md`), 'utf8'))
       assert.ok(result.ok, result.ok ? '' : `${result.error.code}: ${result.error.message}`)
-      assert.deepEqual(toEditorNormal(result.value), toEditorNormal(expected))
+      assert.deepEqual(toEditorNormal(result.value), expected)
     })
   }
 }
@@ -185,12 +185,12 @@ for (const name of pairedNames(normalizationRoot, '.md', '.json')) {
     assert.ok(isAdfDocument(expected), `${name}.json is not an ADF document`)
     const result = markdownToAdf(readFileSync(join(normalizationRoot, `${name}.md`), 'utf8'))
     assert.ok(result.ok, result.ok ? '' : `${result.error.code}: ${result.error.message}`)
-    assert.deepEqual(toEditorNormal(result.value), toEditorNormal(expected))
+    assert.deepEqual(toEditorNormal(result.value), expected)
     const emitted = adfToMarkdown(result.value)
     assert.ok(emitted.ok, emitted.ok ? '' : `${emitted.error.code}: ${emitted.error.message}`)
     const again = markdownToAdf(emitted.value)
     assert.ok(again.ok, again.ok ? '' : `${again.error.code}: ${again.error.message}`)
-    assert.deepEqual(toEditorNormal(again.value), toEditorNormal(expected))
+    assert.deepEqual(toEditorNormal(again.value), expected)
   })
 }
 

src/markdown/emit/adf-to-markdown.test.ts

@@ -356,7 +356,7 @@ test('refuses marks and attributes nested deeper than the emitter carries', () =
     assert.ok(spelled.ok, spelled.ok ? '' : spelled.error.message)
     const read = markdownToAdf(spelled.value)
     assert.ok(read.ok, read.ok ? '' : read.error.message)
-    assert.deepEqual(toEditorNormal(read.value), toEditorNormal(document(node)))
+    assert.deepEqual(toEditorNormal(read.value), document(node))
   }
 
   assert.equal(markdown(adfToMarkdown(document(paragraph({ marks: [{ attrs, type: 'em' }], text: 'x', type: 'text' })))), deeper('depth', 'em', largestNesting - 3))