dexin-content 0.1.2 → 0.2.1

package/package.json

@@ -1,69 +1,82 @@
-{
-  "name": "dexin-content",
-  "version": "0.1.2",
-  "description": "Structured content compilation & runtime layer — neutral Markdown → Content AST → PositiveArtifact + store/query/collection pipeline.",
-  "keywords": [
-    "markdown",
-    "content",
-    "compiler",
-    "ast",
-    "cms",
-    "remark",
-    "dexinlabs"
-  ],
-  "type": "module",
-  "license": "MIT",
-  "author": "得心实验室 <https://gitee.com/cuizhn>",
-  "repository": {
-    "type": "git",
-    "url": "git+https://gitee.com/cuizhn/dexin-content.git"
-  },
-  "homepage": "https://gitee.com/cuizhn/dexin-content#readme",
-  "bugs": {
-    "url": "https://gitee.com/cuizhn/dexin-content/issues"
-  },
-  "engines": {
-    "node": ">=20.0.0"
-  },
-  "files": [
-    "index.ts",
-    "collection.ts",
-    "store.ts",
-    "query.ts",
-    "diff.ts",
-    "core/**/*.ts",
-    "README.md",
-    "LICENSE"
-  ],
-  "exports": {
-    ".": "./index.ts",
-    "./core/types": "./core/types.ts",
-    "./core/compiler": "./core/compiler.ts",
-    "./core/frontmatter": "./core/frontmatter.ts",
-    "./core/discovery": "./core/discovery.ts",
-    "./core/markdown": "./core/markdown.ts",
-    "./store": "./store.ts",
-    "./collection": "./collection.ts",
-    "./query": "./query.ts",
-    "./diff": "./diff.ts"
-  },
-  "scripts": {
-    "typecheck": "tsc --noEmit",
-    "runtime:check": "tsx scripts/runtime-check.ts",
-    "pack": "npm pack --dry-run"
-  },
-  "dependencies": {
-    "yaml": "^2.5.1",
-    "remark-parse": "^11.0.0",
-    "remark-gfm": "^4.0.0",
-    "remark-math": "^6.0.0",
-    "remark-directive": "^3.0.0",
-    "unified": "^11.0.0"
-  },
-  "devDependencies": {
-    "@types/node": "^22.0.0",
-    "@types/mdast": "^4.0.4",
-    "typescript": "^5.5.0",
-    "tsx": "^4.7.0"
-  }
-}
+{
+  "name": "dexin-content",
+  "version": "0.2.1",
+  "description": "Open content toolchain — Markdown → LessonAST/ContentPackage compiler, CLI and validator (dexinlabs ecosystem).",
+  "keywords": [
+    "markdown",
+    "content",
+    "compiler",
+    "ast",
+    "remark",
+    "dexinlabs"
+  ],
+  "type": "module",
+  "license": "Apache-2.0",
+  "author": "得心实验室 <https://gitee.com/cuizhn>",
+  "repository": {
+    "type": "git",
+    "url": "git+https://gitee.com/cuizhn/dexin-content.git"
+  },
+  "homepage": "https://gitee.com/cuizhn/dexin-content#readme",
+  "bugs": {
+    "url": "https://gitee.com/cuizhn/dexin-content/issues"
+  },
+  "engines": {
+    "node": ">=20.0.0"
+  },
+  "bin": {
+    "dexin-content": "./bin/dexin-content.mjs"
+  },
+  "files": [
+    "index.ts",
+    "core/**/*.ts",
+    "cli/**/*.ts",
+    "domains/**/*.ts",
+    "bin/**",
+    "README.md",
+    "LICENSE"
+  ],
+  "exports": {
+    "./core/markdown": "./core/parser/markdown.ts",
+    ".": "./index.ts",
+    "./cli/validate": "./cli/validate.ts",
+    "./core/frontmatter": "./core/parser/frontmatter.ts",
+    "./diff": "./core/artifact/diff.ts",
+    "./cli/package": "./cli/package.ts",
+    "./collection": "./core/artifact/collection.ts",
+    "./cli/build": "./cli/build.ts",
+    "./query": "./core/artifact/query.ts",
+    "./core/compiler": "./core/compiler/compiler.ts",
+    "./core/types": "./core/types/index.ts",
+    "./store": "./core/artifact/store.ts",
+    "./cli/serve": "./cli/serve.ts",
+    "./cli/check": "./cli/check.ts",
+    "./core/discovery": "./core/compiler/discovery.ts",
+    "./core/types/lessonAST": "./core/types/lessonAST.ts",
+    "./domains/lesson": "./domains/lesson/index.ts",
+    "./domains/document": "./domains/document/index.ts"
+  },
+  "scripts": {
+    "typecheck": "tsc --noEmit",
+    "runtime:check": "tsx scripts/runtime-check.ts",
+    "pack": "npm pack --dry-run",
+    "build": "tsx cli/build.ts",
+    "package": "tsx cli/package.ts"
+  },
+  "dependencies": {
+    "tsx": "^4.23.0",
+    "yaml": "^2.5.1",
+    "remark-parse": "^11.0.0",
+    "remark-gfm": "^4.0.0",
+    "remark-math": "^6.0.0",
+    "remark-directive": "^3.0.0",
+    "unified": "^11.0.0",
+    "remark": "^15.0.0",
+    "micromark-extension-directive": "^3.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^22.0.0",
+    "@types/mdast": "^4.0.4",
+    "typescript": "^5.5.0"
+  }
+}

package/core/discovery.ts

@@ -1,110 +0,0 @@
-// ─────────────────────────────────────────────────────────────
-// dexin-content/core/discovery.ts
-// Source adapter + collection routing + identity derivation helpers.
-//
-// Document path convention → DocumentIdentity:
-//   id   = relPath minus .md, minus trailing `/index`
-//   path = '/' + id
-//   file = relPath WITH .md extension
-//   + auxiliary fields: collection, (optionally) index_file:true
-//
-// Hosts needing domain-specific identity shapes define their own types
-// and construct CompileInput identities themselves; core only knows the
-// document convention.
-// ─────────────────────────────────────────────────────────────
-
-import path from 'node:path'
-import { readFileSync, existsSync, statSync } from 'node:fs'
-import type { DocumentIdentity, ParseError } from './types'
-
-// ── Shared BOM / CRLF sanitisation primitives ─────────────
-//    Used by both sync (readSourceFile) and async (createLocalSource.read)
-//    IO entry points to guarantee byte-identical failure behaviour.
-
-/**
- * Build a LINE_ENDING_CONTAMINATION ParseError for a given relative path.
- * Caller supplies a short remediation hint matching its context.
- */
-export function mkLineEndingError (
-  relPath: string,
-  hint: string = 'Normalize to LF before processing.'
-): ParseError & Error {
-  const err = new Error(
-    `[LINE_ENDING_CONTAMINATION] Source file '${relPath}' contains CRLF line endings. ` +
-    hint
-  ) as ParseError & Error
-  err.code = 'LINE_ENDING_CONTAMINATION'
-  err.file = relPath
-  return err
-}
-
-/**
- * Strip a leading UTF-8 BOM (U+FEFF) if present and guard against CRLF
- * contamination. Returns the cleaned source string; throws with
- * LINE_ENDING_CONTAMINATION code on CR presence.
- *
- * This is intentionally a pure string-in / string-out (or throw) helper so
- * both sync and async IO paths share one canonical sanitisation.
- */
-export function sanitizeSourceText (raw: string, relPath: string): string {
-  const source = raw.charCodeAt(0) === 0xfeff ? raw.slice(1) : raw
-  if (source.includes('\r')) throw mkLineEndingError(relPath)
-  return source
-}
-
-/**
- * Normalise a relative path to forward slashes and strip any leading `./`.
- */
-export function normaliseRel (p: string): string {
-  let out = p.split(path.sep).join('/')
-  while (out.startsWith('./')) out = out.slice(2)
-  return out
-}
-
-/**
- * Build DocumentIdentity from a `.md` file path relative
- * to the collection source root.
- */
-export function buildDocumentIdentity (
-  relPath: string,
-  collection: string
-): DocumentIdentity {
-  const norm = normaliseRel(relPath)
-  const file = norm
-
-  let id = file.endsWith('.md') ? file.slice(0, -'.md'.length) : file
-  if (id.endsWith('.markdown')) id = id.slice(0, -'.markdown'.length)
-
-  const isIndex = id.endsWith('/index') || id === 'index'
-  if (isIndex) {
-    if (id === 'index') id = ''
-    else id = id.slice(0, -'/index'.length)
-  }
-
-  const pathUrl = '/' + id
-  const identity: DocumentIdentity = {
-    id,
-    path: pathUrl,
-    file,
-    collection
-  }
-  if (isIndex) identity.index_file = true
-  return identity
-}
-
-/** Read a single UTF-8 file and pair it with its relative path. */
-export interface SourceFile {
-  absPath: string
-  relPath: string
-  source: string
-}
-
-export function readSourceFile (absPath: string, sourceRoot: string): SourceFile {
-  if (!existsSync(absPath) || !statSync(absPath).isFile()) {
-    throw new Error(`[discovery] Not a file or missing: ${absPath}`)
-  }
-  const raw = readFileSync(absPath, 'utf8')
-  const rel = normaliseRel(path.relative(sourceRoot, absPath))
-  const source = sanitizeSourceText(raw, rel)
-  return { absPath, relPath: rel, source }
-}

package/core/types.ts

@@ -1,162 +0,0 @@
-// ─────────────────────────────────────────────────────────────
-// dexin-content/core/types.ts — Generic shared types.
-// ─────────────────────────────────────────────────────────────
-// This file defines the shared type system used across domains.
-// It MUST remain business-agnostic. All names, literals, and
-// comment text are written in generic, domain-neutral vocabulary.
-
-// ── Inline content (shared across the neutral AST and all domains) ──
-
-export type Inline =
-  | TextInline
-  | BoldInline
-  | ItalicInline
-  | CodeInline
-  | LinkInline
-  | FormulaInline
-
-export interface TextInline { type: 'text'; value: string }
-export interface BoldInline { type: 'bold'; children: Inline[] }
-export interface ItalicInline { type: 'italic'; children: Inline[] }
-export interface CodeInline { type: 'code'; value: string }
-export interface LinkInline { type: 'link'; url: string; children: Inline[] }
-/** Inline LaTeX-format formula span (single-paragraph). */
-export interface FormulaInline { type: 'math'; latex: string }
-
-// ── Document-level blocks (neutral AST — no domain mapping) ──
-
-/**
- * HeadingBlock in the neutral AST: level is clamped to 1–4
- * (h1→1, h2→2, h3→3, h4→4). `mdDepth` is the ORIGINAL markdown heading
- * depth (1..6 inclusive) captured at parse time. It is OPTIONAL and exists
- * ONLY so active domain parsers can run their own fail-fast checks for
- * out-of-range depths. `mdDepth` MUST NOT leak into the serialised
- * canonical Artifact (the `content` output): domain parsers' clone/render
- * paths must drop it.
- */
-export interface HeadingBlock {
-  type: 'heading'
-  level: 1 | 2 | 3 | 4
-  children: Inline[]
-  /** Original markdown heading depth (1..6). Transient, non-canonical. */
-  mdDepth?: number
-}
-export interface ParagraphBlock { type: 'paragraph'; children: Inline[] }
-export interface QuoteBlock { type: 'quote'; children: DocumentBlock[] }
-export interface DividerBlock { type: 'divider' }
-export interface ListBlock { type: 'list'; ordered: boolean; items: Inline[][] }
-export type TableCell = Inline[]
-export interface TableBlock { type: 'table'; headers: TableCell[]; rows: TableCell[][] }
-export interface ImageBlock { type: 'image'; src: string; alt: string; caption?: string }
-export interface CodeBlock { type: 'code'; language: string; code: string }
-/** Block-level LaTeX-format formula node; display=true for centered fences. */
-export interface FormulaBlock { type: 'formula'; latex: string; display: boolean }
-
-/**
- * Neutral ContainerBlock:
- * Produced for every directive container encountered in the markdown body.
- * The active domain decides whether an unknown name is acceptable;
- * the generic layer always preserves the structure as-is.
- */
-export interface ContainerBlock {
-  type: 'container'
-  name: string
-  attrs: Record<string, string>
-  children: DocumentBlock[]
-}
-
-export type DocumentBlock =
-  | ParagraphBlock
-  | HeadingBlock
-  | QuoteBlock
-  | DividerBlock
-  | ListBlock
-  | TableBlock
-  | ImageBlock
-  | CodeBlock
-  | FormulaBlock
-  | ContainerBlock
-
-export interface DocumentContent {
-  version: 1
-  blocks: DocumentBlock[]
-}
-
-// ── Artifact shapes (shared five-field header) ──
-
-/**
- * Generic domain tag. Literal string values are chosen by the host runner
- * and flow through to the Artifact output untouched. Core itself does not
- * interpret them beyond registry lookup keys.
- */
-export type DomainName = string & { readonly __brand?: unique symbol }
-
-/**
- * Plain-document identity:
- *   id    — relative to content root; strip `.md` and trailing `/index`
- *   path  — URL semantics: `'/' + id`
- *   file  — physical path relative to content root, INCLUDING `.md` extension
- *
- * Domain-specific routing fields may be attached by hosts via the index
- * signature; core never interprets them.
- */
-export interface DocumentIdentity {
-  id: string
-  path: string        // URL semantic: '/' + id
-  file: string        // physical path relative to content root WITH ext
-  collection: string
-  [k: string]: unknown   // auxiliary fields (e.g. index_file:true)
-}
-
-/** Frontmatter scalar-only projection (SCHEMA-FREE RULE) */
-export type Meta = Record<string, string | number | boolean>
-
-export interface ArtifactHeader {
-  id: string               // Stable document id (unique within a store)
-  domain: DomainName | string
-  identity: DocumentIdentity
-  meta: Meta
-}
-
-export interface PositiveArtifact extends ArtifactHeader {
-  content: { version: 1; blocks: unknown[] }
-}
-
-/** Compile-failure output — captures first fail-fast error. */
-export interface ErrorArtifact extends ArtifactHeader {
-  kind: 'COMPILE_ERROR'
-  error: { code: string; message: string }
-}
-
-export type Artifact = PositiveArtifact | ErrorArtifact
-
-// ── Unified error object ──
-
-export interface ParseError extends Error {
-  code: string
-  message: string
-  file: string
-  loc?: { container?: string; headingDepth?: number; invalidEnumValue?: string; allowed?: string }
-}
-
-// ── Parser contract — domains implement; core dispatches ──
-
-export interface ParseContext {
-  /** Relative path to the source file, used in error messages. */
-  file: string
-  /** Collection-level schema (required fields + scalar type checks). */
-  schema?: Schema
-  /** Frontmatter already projected as scalar-only meta. */
-  meta: Meta
-}
-
-export interface Schema {
-  required?: string[]
-  types?: Record<string, 'string' | 'number' | 'boolean'>
-}
-
-export interface DomainParser {
-  domain: DomainName | string
-  /** Convert neutral AST → domain-specific Artifact content.blocks shape. May throw ParseError. */
-  parse(content: DocumentContent, ctx: ParseContext): { version: 1; blocks: unknown[] }
-}