@leonsilicon/moedict-data

教育部重編國語辭典修訂本 的資料檔，以 JSON 形式打包並附帶 TypeScript 型別。

Parsed JSON of the Ministry of Education Revised Mandarin Chinese Dictionary (Republic of China), packaged for Node.js with accurate TypeScript types. Source data is from g0v/moedict-data.

The dataset contains 161,197 entries including characters, words, idioms, with bopomofo, pinyin, definitions, examples, quotes, and synonyms/antonyms.

Install

npm install @leonsilicon/moedict-data

Usage

import dict from "@leonsilicon/moedict-data";

console.log(dict.length); // 161197

const entry = dict.find((e) => e.title === "中文");
console.log(entry?.heteronyms[0]?.pinyin); // "zhōng wén"

The JSON file is also exposed directly:

import dict from "@leonsilicon/moedict-data/dict-revised.json" with { type: "json" };

Schema

interface Entry {
  title: string;
  heteronyms: Heteronym[];
  radical?: string;
  stroke_count?: number;
  non_radical_stroke_count?: number;
}

interface Heteronym {
  definitions: Definition[];
  bopomofo?: string;
  pinyin?: string;
}

interface Definition {
  def: string;
  type?: string;       // 名 / 動 / 形 / ...
  quote?: string[];    // 古典引文
  example?: string[];  // 例句
  link?: string[];     // 相關連結
  synonyms?: string;   // 近義詞 (comma-separated)
  antonyms?: string;   // 反義詞 (comma-separated)
}

Building

The shipped dict-revised.json is generated from data/dict-revised.json.xz:

bun run parse

This requires the xz binary on $PATH.

License

MIT. Dictionary data is © 中華民國教育部 and distributed under the terms used by g0v/moedict-data.