@v1hz/md2docx 1.6.0 → 2.1.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@v1hz/md2docx",
-  "version": "1.6.0",
+  "version": "2.1.0",
   "description": "基于 pandoc 的 Markdown 转 Word 工具，自动格式化 Markdown，支持自定义样式，AI 友好，用户友好。",
   "keywords": [
     "converter",
@@ -21,43 +21,48 @@
     "url": "git+https://github.com/WXY-V1hZ/md2docx.git"
   },
   "bin": {
-    "md2docx": "src/index.ts"
+    "md2docx": "dist/index.js"
   },
   "files": [
-    "src",
-    "config/config.json",
-    "config/config.schema.json",
-    "config/style.json",
+    "dist",
+    "config",
     "README.md",
     "LICENSE"
   ],
   "type": "module",
-  "module": "src/index.ts",
+  "module": "dist/index.js",
   "scripts": {
     "cli": "bun run src/index.ts",
+    "clean:dist": "bun -e \"import { rmSync } from 'node:fs'; rmSync('dist', { recursive: true, force: true })\"",
+    "build": "bun run clean:dist && bun build src/index.ts --target=node --outdir dist --entry-naming index.js --asset-naming [name].[ext]",
+    "build:exe": "bun run clean:dist && bun build src/index.ts --compile --outfile dist/md2docx.exe",
+    "prepack": "bun run build",
     "test": "bun test",
     "check": "bun tsc --noEmit && bun run oxlint && bun run oxfmt"
   },
   "dependencies": {
+    "@resvg/resvg-wasm": "2.6.2",
     "@xmldom/xmldom": "^0.9.10",
     "beautiful-mermaid": "^1.1.3",
+    "commander": "^15.0.0",
     "docx": "^9.7.1",
     "pizzip": "^3.2.0",
     "remark-frontmatter": "^5.0.0",
     "remark-gfm": "^4.0.1",
     "remark-parse": "^11.0.0",
     "remark-stringify": "^11.0.0",
-    "sharp": "^0.35.3",
     "unified": "^11.0.5",
     "xpath": "^0.0.34"
   },
   "devDependencies": {
     "@types/bun": "^1.3.14",
+    "@types/node": "^24.0.0",
     "oxfmt": "^0.58.0",
     "oxlint": "^1.73.0",
+    "sharp": "^0.35.3",
     "typescript": "^7.0.2"
   },
   "engines": {
-    "bun": ">=1.3"
+    "node": ">=22.12.0"
   }
 }

package/src/cli.ts

@@ -1,210 +0,0 @@
-import { type AppConfig } from "./config";
-
-type SchemaValueType = "boolean" | "integer" | "number" | "string";
-
-interface SchemaNode {
-  type?: SchemaValueType | "object";
-  description?: string;
-  default?: unknown;
-  enum?: unknown[];
-  minimum?: number;
-  properties?: Record<string, SchemaNode>;
-  required?: string[];
-}
-
-export interface ConfigOption {
-  path: string;
-  type: SchemaValueType;
-  description: string;
-  default?: unknown;
-  enum?: unknown[];
-  minimum?: number;
-  required: boolean;
-}
-
-export interface CliOptions {
-  help: boolean;
-  version: boolean;
-  web: boolean;
-  mdPath?: string;
-  outputPath?: string;
-  configPath?: string;
-  overrides: Map<string, unknown>;
-}
-
-export function getConfigOptions(schema: SchemaNode): ConfigOption[] {
-  const options: ConfigOption[] = [];
-
-  function walk(node: SchemaNode, prefix: string): void {
-    if (node.type !== "object" || !node.properties) return;
-    for (const [name, child] of Object.entries(node.properties)) {
-      const path = prefix ? `${prefix}.${name}` : name;
-      if (child.type === "object") {
-        walk(child, path);
-      } else if (isValueType(child.type)) {
-        options.push({
-          path,
-          type: child.type,
-          description: child.description ?? "",
-          default: child.default,
-          enum: child.enum,
-          minimum: child.minimum,
-          required: node.required?.includes(name) ?? false,
-        });
-      }
-    }
-  }
-
-  walk(schema, "");
-  return options;
-}
-
-export function parseCliArgs(args: string[], configOptions: ConfigOption[]): CliOptions {
-  const optionsByName = new Map(configOptions.map((option) => [option.path, option]));
-  const result: CliOptions = { help: false, version: false, web: false, overrides: new Map() };
-  const positional: string[] = [];
-
-  for (let i = 0; i < args.length; i++) {
-    const arg = args[i]!;
-    if (arg === "-h" || arg === "--help") {
-      result.help = true;
-      continue;
-    }
-    if (arg === "-v" || arg === "--version") {
-      result.version = true;
-      continue;
-    }
-    if (arg === "--web") {
-      result.web = true;
-      continue;
-    }
-    if (arg === "-o" || arg === "--config") {
-      const value = args[++i];
-      if (value === undefined || value.startsWith("--")) {
-        throw new Error(`${arg} 缺少路径`);
-      }
-      if (arg === "-o") result.outputPath = value;
-      else result.configPath = value;
-      continue;
-    }
-    if (arg.startsWith("--")) {
-      const name = arg.slice(2);
-      const option = optionsByName.get(name);
-      if (!option) throw new Error(`未知参数：${arg}`);
-      const value = args[++i];
-      if (value === undefined) throw new Error(`${arg} 缺少值`);
-      result.overrides.set(name, parseConfigValue(value, option));
-      continue;
-    }
-    if (arg.startsWith("-") && !isNumericArgument(arg)) {
-      throw new Error(`未知参数：${arg}`);
-    }
-    positional.push(arg);
-  }
-
-  if (!result.help && !result.version && !result.web) {
-    if (positional.length === 0) throw new Error("缺少 Markdown 文件路径");
-    if (positional.length > 1) throw new Error("只能指定一个 Markdown 文件路径");
-    result.mdPath = positional[0];
-  }
-  return result;
-}
-
-export function applyConfigOverrides(
-  config: AppConfig,
-  overrides: Map<string, unknown>,
-): AppConfig {
-  const merged = structuredClone(config) as unknown as Record<string, unknown>;
-  for (const [path, value] of overrides) {
-    const segments = path.split(".");
-    let target = merged;
-    for (const segment of segments.slice(0, -1)) {
-      const next = target[segment];
-      if (!isRecord(next)) throw new Error(`配置路径不存在：${path}`);
-      target = next;
-    }
-    target[segments.at(-1)!] = value;
-  }
-  return merged as unknown as AppConfig;
-}
-
-export function formatHelp(configOptions: ConfigOption[]): string {
-  const lines = [
-    "用法:",
-    "  md2docx <md-path> [选项]",
-    "  md2docx clean",
-    "",
-    "选项:",
-    "  -o <path>                         输出 docx 路径",
-    "  --config <path>                   配置文件路径，默认 config/config.json",
-    "  --web                             打开默认配置的网页编辑器",
-    "  -v, --version                     显示版本号",
-    "  -h, --help                        显示帮助",
-    "",
-    "子命令:",
-    "  clean                             清除 tmp/ 缓存",
-    "",
-    "配置覆盖:",
-  ];
-
-  for (const option of configOptions) {
-    const constraints = [
-      option.default === undefined ? undefined : `默认: ${String(option.default)}`,
-      option.enum ? `可选: ${option.enum.join(" | ")}` : undefined,
-      option.minimum === undefined ? undefined : `最小: ${option.minimum}`,
-    ].filter(Boolean);
-    lines.push(`  --${option.path} <${option.type}>`);
-    lines.push(
-      `      ${option.description}${constraints.length ? `（${constraints.join("；")}）` : ""}`,
-    );
-  }
-
-  return lines.join("\n");
-}
-
-function parseConfigValue(value: string, option: ConfigOption): unknown {
-  let parsed: unknown;
-  switch (option.type) {
-    case "boolean":
-      if (value !== "true" && value !== "false") {
-        throw new Error(`--${option.path} 需要 boolean（true 或 false），收到：${value}`);
-      }
-      parsed = value === "true";
-      break;
-    case "integer":
-      if (!/^-?\d+$/.test(value)) {
-        throw new Error(`--${option.path} 需要整数，收到：${value}`);
-      }
-      parsed = Number(value);
-      break;
-    case "number":
-      parsed = Number(value);
-      if (!Number.isFinite(parsed)) {
-        throw new Error(`--${option.path} 需要数字，收到：${value}`);
-      }
-      break;
-    case "string":
-      parsed = value;
-      break;
-  }
-
-  if (option.enum && !option.enum.includes(parsed)) {
-    throw new Error(`--${option.path} 必须是以下值之一：${option.enum.join(", ")}`);
-  }
-  if (typeof parsed === "number" && option.minimum !== undefined && parsed < option.minimum) {
-    throw new Error(`--${option.path} 不能小于 ${option.minimum}`);
-  }
-  return parsed;
-}
-
-function isValueType(type: SchemaNode["type"]): type is SchemaValueType {
-  return type === "boolean" || type === "integer" || type === "number" || type === "string";
-}
-
-function isRecord(value: unknown): value is Record<string, unknown> {
-  return typeof value === "object" && value !== null && !Array.isArray(value);
-}
-
-function isNumericArgument(value: string): boolean {
-  return /^-\d/.test(value);
-}

package/src/config.ts

@@ -1,52 +0,0 @@
-export interface CaptionStyle {
-  enabled: boolean;
-  format: string;
-  separator: string;
-}
-
-export interface MermaidConfig {
-  enabled: boolean;
-  theme: string;
-  density: number;
-}
-
-export interface BooleanConfig {
-  enabled: boolean;
-}
-
-export interface HeadingNumberingConfig {
-  enabled: boolean;
-  detectExisting: boolean;
-  existingPattern?: string;
-  useBuiltinRules?: boolean;
-}
-
-export type TitleExtractStrategy = "first-h1" | "single-h1" | "filename" | "none";
-
-export interface TitleConfig {
-  enabled: boolean;
-  strategy: TitleExtractStrategy;
-}
-
-export interface PandocConfig {
-  enabled: boolean;
-  outputName: string;
-}
-
-export interface AppConfig {
-  figureCaption: CaptionStyle;
-  tableCaption: CaptionStyle;
-  normalizeHeadings: BooleanConfig;
-  numberHeadings: HeadingNumberingConfig;
-  renderMermaid: MermaidConfig;
-  detectTitle: TitleConfig;
-  pandoc: PandocConfig;
-}
-
-export async function loadConfig(path: string): Promise<AppConfig> {
-  const exists = await Bun.file(path).exists();
-  if (!exists) throw new Error(`找不到配置文件：${path}`);
-  const raw = JSON.parse(await Bun.file(path).text()) as Record<string, unknown>;
-  delete raw.$schema;
-  return raw as unknown as AppConfig;
-}

package/src/index.ts

@@ -1,99 +0,0 @@
-#!/usr/bin/env bun
-
-import { existsSync, mkdirSync, rmSync, writeFileSync } from "fs";
-import { dirname, parse, resolve } from "path";
-import { $ } from "bun";
-
-import { applyConfigOverrides, formatHelp, getConfigOptions, parseCliArgs } from "./cli";
-import { loadConfig } from "./config";
-import {
-  CONFIG_PATH,
-  CONFIG_SCHEMA_PATH,
-  PKG_DIR,
-  STYLE_CONFIG,
-  TMP_DIR,
-  formattedMdPath,
-  preprocessDir,
-} from "./paths";
-import { preprocess } from "./preprocess/index";
-import { startWebEditor } from "./web";
-import { ensureTemplateDocx } from "./style/generate";
-
-async function cleanCache(): Promise<number> {
-  const resolved = resolve(TMP_DIR);
-  if (existsSync(resolved)) {
-    rmSync(resolved, { recursive: true, force: true });
-    console.log(`已清除缓存：${resolved}`);
-  } else {
-    console.log("缓存目录不存在，无需清理。");
-  }
-  return 0;
-}
-
-export async function run(args: string[]): Promise<number> {
-  try {
-    // 检查子命令
-    const subcommand = args[0];
-    if (subcommand === "clean") {
-      return await cleanCache();
-    }
-    const schema = JSON.parse(await Bun.file(CONFIG_SCHEMA_PATH).text());
-    const configOptions = getConfigOptions(schema);
-    const cli = parseCliArgs(args, configOptions);
-
-    if (cli.version) {
-      const { version } = JSON.parse(await Bun.file(resolve(PKG_DIR, "package.json")).text()) as {
-        version: string;
-      };
-      console.log(`@v1hz/md2docx v${version}`);
-      return 0;
-    }
-    if (cli.help) {
-      console.log(formatHelp(configOptions));
-      return 0;
-    }
-    if (cli.web) {
-      await startWebEditor();
-      return 0;
-    }
-
-    const mdPath = resolve(cli.mdPath!);
-    if (!existsSync(mdPath)) throw new Error(`找不到 Markdown 文件：${mdPath}`);
-
-    const config = await loadConfig(cli.configPath ?? CONFIG_PATH);
-    const cfg = applyConfigOverrides(config, cli.overrides);
-    const baseName = parse(mdPath).name;
-    const outDir = preprocessDir(baseName);
-    mkdirSync(outDir, { recursive: true });
-
-    const formattedMd = await preprocess(mdPath, cfg, outDir);
-    const mdOutput = formattedMdPath(baseName);
-    writeFileSync(mdOutput, formattedMd, "utf-8");
-
-    if (cfg.pandoc.enabled) {
-      const templatePath = await ensureTemplateDocx(STYLE_CONFIG);
-      const configuredName = cfg.pandoc.outputName.replaceAll("{file_name}", baseName);
-      const docxOutput = resolve(cli.outputPath ?? configuredName);
-      mkdirSync(dirname(docxOutput), { recursive: true });
-      const luaFilter = resolve(PKG_DIR, "config/lua/add-inline-code.lua");
-      const luaFlag = existsSync(luaFilter) ? `--lua-filter=${luaFilter}` : "";
-      const result =
-        await $`pandoc ${mdOutput} -o ${docxOutput} --reference-doc=${templatePath} ${luaFlag}`.nothrow();
-      if (result.exitCode !== 0) {
-        console.error(`pandoc 转换失败 (exit code ${result.exitCode}):`, result.stderr.toString());
-        return 1;
-      }
-      console.log(`已生成：${docxOutput}`);
-    }
-    return 0;
-  } catch (error) {
-    const message = error instanceof Error ? error.message : String(error);
-    console.error(`错误：${message}`);
-    console.error("使用 --help 查看帮助");
-    return 1;
-  }
-}
-
-if (import.meta.main) {
-  process.exitCode = await run(Bun.argv.slice(2));
-}

package/src/paths.ts

@@ -1,34 +0,0 @@
-import { resolve } from "path";
-
-const PKG_DIR = resolve(import.meta.dir, "..");
-
-/** 包根目录 */
-export { PKG_DIR };
-
-/** 默认配置文件路径（相对于包安装目录） */
-export const CONFIG_PATH = resolve(PKG_DIR, "config/config.json");
-
-/** 默认配置 schema 路径（相对于包安装目录） */
-export const CONFIG_SCHEMA_PATH = resolve(PKG_DIR, "config/config.schema.json");
-
-/** 临时文件根目录（相对于当前工作目录） */
-export const TMP_DIR = "tmp";
-
-/** 预处理模块的输出目录（相对于当前工作目录） */
-export function preprocessDir(baseName: string): string {
-  return `${TMP_DIR}/preprocess/${baseName}`;
-}
-
-/** 预处理后的 Markdown 文件路径 */
-export function formattedMdPath(baseName: string): string {
-  return `${preprocessDir(baseName)}/${baseName}_formatted.md`;
-}
-
-/** 样式模块目录 */
-export const STYLE_DIR = `${TMP_DIR}/style`;
-
-/** 样式配置文件（相对于包安装目录），用户维护的样式定义 */
-export const STYLE_CONFIG = resolve(PKG_DIR, "config/style.json");
-
-/** 生成的模板 docx 文件路径 */
-export const STYLE_TEMPLATE_DOCX = `${STYLE_DIR}/style.docx`;

package/src/preprocess/caption.ts

@@ -1,107 +0,0 @@
-import { type Image, type Paragraph, type Root, type Text } from "mdast";
-import { type AppConfig } from "../config";
-
-export function numberTables(root: Root, config: AppConfig) {
-  const {
-    tableCaption: { format, separator },
-  } = config;
-  let counter = 0;
-  const inserts: { at: number; text: string }[] = [];
-  const used = new Set<number>();
-  for (let i = 0; i < root.children.length; i++) {
-    const child = root.children[i]!;
-    if (child?.type !== "table") continue;
-
-    counter++;
-    const label = format.replace("{n}", String(counter));
-
-    const prevIdx = i - 1;
-    const prev = prevIdx >= 0 ? root.children[prevIdx] : undefined;
-    if (prev && isCaption(prev) && !used.has(prevIdx)) {
-      used.add(prevIdx);
-      const existing = stripTableNum(captionText(prev), format);
-      const first = prev.children[0] as Text;
-      first.value = `Table: ${label}${existing ? `${separator}${existing}` : ""}`;
-      continue;
-    }
-
-    const nextIdx = i + 1;
-    const next = nextIdx < root.children.length ? root.children[nextIdx] : undefined;
-    if (next && isCaption(next) && !used.has(nextIdx)) {
-      used.add(nextIdx);
-      const existing = stripTableNum(captionText(next), format);
-      const first = next.children[0] as Text;
-      first.value = `Table: ${label}${existing ? `${separator}${existing}` : ""}`;
-      continue;
-    }
-
-    inserts.push({ at: i + 1, text: `Table: ${label}` });
-  }
-
-  for (let j = inserts.length - 1; j >= 0; j--) {
-    const { at, text } = inserts[j]!;
-    root.children.splice(at, 0, {
-      type: "paragraph",
-      children: [{ type: "text", value: text }],
-    } as Paragraph);
-  }
-}
-
-export function numberPictures(root: Root, config: AppConfig) {
-  const {
-    figureCaption: { format, separator },
-  } = config;
-  let counter = 0;
-  for (let i = 0; i < root.children.length; i++) {
-    const child = root.children[i]!;
-    if (
-      child?.type === "paragraph" &&
-      child.children.length === 1 &&
-      child.children[0]?.type === "image"
-    ) {
-      counter++;
-      const img = child.children[0] as Image;
-      const cleaned = stripPictureNum(img.title ?? img.alt ?? "", format);
-      const label = cleaned || fileNameFromUrl(img.url) || "";
-      const prefix = format.replace("{n}", String(counter));
-      img.alt = `${prefix}${label ? `${separator}${label}` : ""}`;
-    }
-  }
-}
-
-/** 从 URL/路径中提取不带扩展名的文件名 */
-function fileNameFromUrl(url: string): string | null {
-  const match = url.match(/\/([^/]+?)(?:\.[^/.]+)?$/);
-  return match ? match[1]! : null;
-}
-
-function isCaption(node: unknown): node is Paragraph {
-  return (
-    node != null &&
-    typeof node === "object" &&
-    (node as { type?: string }).type === "paragraph" &&
-    (node as { children?: { type?: string; value?: string }[] }).children?.length === 1 &&
-    (node as { children: { type?: string; value?: string }[] }).children[0]?.type === "text" &&
-    (
-      (node as { children: { type?: string; value?: string }[] }).children[0]?.value ?? ""
-    ).startsWith("Table: ")
-  );
-}
-
-function captionText(node: Paragraph): string {
-  const first = node.children[0];
-  if (first?.type === "text") {
-    return first.value.replace(/^Table: /, "");
-  }
-  return "";
-}
-
-function stripTableNum(text: string, format: string): string {
-  const escaped = format.replace(/[.*+?^${}()|[\\]\\]/g, "\\$&").replace("{n}", "\\d+");
-  return text.replace(new RegExp(`^${escaped}`), "");
-}
-
-function stripPictureNum(text: string, format: string): string {
-  const escaped = format.replace(/[.*+?^${}()|[\\]\\]/g, "\\$&").replace("{n}", "\\d+");
-  return text.replace(new RegExp(`^${escaped}`), "");
-}

package/src/preprocess/index.ts

@@ -1,49 +0,0 @@
-import { type Heading } from "mdast";
-import { unified } from "unified";
-import remarkParse from "remark-parse";
-import remarkStringify from "remark-stringify";
-import remarkGfm from "remark-gfm";
-import remarkFrontmatter from "remark-frontmatter";
-import { visit } from "unist-util-visit";
-
-import { type AppConfig } from "../config";
-import { addTitle, normalizeHeadings, numberHeadings } from "./title";
-import { numberTables, numberPictures } from "./caption";
-import { renderMermaid } from "./mermaid";
-
-export async function preprocess(mdPath: string, cfg: AppConfig, outDir: string): Promise<string> {
-  const md = await Bun.file(mdPath).text();
-
-  const processor = unified()
-    .use(remarkParse)
-    .use(remarkFrontmatter, ["yaml"])
-    .use(remarkGfm)
-    .use(remarkStringify, { resourceLink: true, bullet: "-" });
-  const ast = processor.parse(md);
-
-  const headings: Heading[] = [];
-  visit(ast, "heading", (heading: Heading) => {
-    headings.push(heading);
-  });
-
-  addTitle(mdPath, ast, headings, cfg.detectTitle);
-
-  if (cfg.normalizeHeadings.enabled || cfg.numberHeadings.enabled) {
-    normalizeHeadings(headings);
-  }
-  if (cfg.numberHeadings.enabled) {
-    numberHeadings(headings, cfg.numberHeadings);
-  }
-
-  if (cfg.tableCaption.enabled) {
-    numberTables(ast, cfg);
-  }
-  if (cfg.renderMermaid.enabled) {
-    await renderMermaid(ast, outDir, cfg.renderMermaid.theme, cfg.renderMermaid.density);
-  }
-  if (cfg.figureCaption.enabled) {
-    numberPictures(ast, cfg);
-  }
-
-  return processor.stringify(ast);
-}