js-mdict

mdict (*.mdd *.mdx) file reader based on jeka-kiselyov/mdict .

Very thanks to fengdh and jeka-kiselyov.

Latest Version

v6.0.0-alpha-2

Usage

NOTE: Not Support Browser!

npm install [email protected]

ESM

file: demo.mjs

import { Mdict } from "js-mdict";

const mdict = new Mdict("resources/oald7.mdx");

const def = mdict.lookup("ask");
console.log(def.definition);


/*
<head><link rel="stylesheet" type="text/css" href="O7.css"/></head><body><span class="hw"> ask </span hw><span class="i_g"> <img src="key.gif"/>  /<a class="i_phon" href="sound://aask_ggv_r1_oa013910.spx">ɑ:sk</a i_phon><span class="z">; </span z><i>NAmE</i> <a class="y_phon" href="sound://aask_ggx_r1_wpu01057.spx">æsk</a y_phon>/ </span i_g><span class="cls"> verb</span cls><br><span class="sd">QUESTION<span class="chn"> 问题</span chn></span sd>
<div class="define"><span class="numb">1</span numb><span class="cf"> ~ <span class="bra">(</span bra>sb<span class="bra">)</span bra> <span class="bra">(</span bra>about sb/ sth<span class="bra">)</span bra> </span cf><span class="d">to say or write sth in the form of a question, in order to get information<span class="chn"> 问；询问</span chn></span d></div define>
<span class="phrase"><span class="pt">  [<span class="pt_inside">V <span class="pt_bold">speech</span></span><span>]</span> </span pt></span phrase>
<span class="sentence_eng">'Where are you going?' she asked. </span sentence_eng>
<span class="sentence_chi">"你去哪里？"她问道。</span sentence_chi>
<span class="phrase"><span class="pt"> [<span class="pt_inside">VN <span class="pt_bold">speech</span></span><span>]</span> </span pt></span phrase>
<span class="sentence_eng">'Are you sure?' he asked her. </span sentence_eng>
...
</body>
  */

CommonJS

file: demo.mjs

const { Mdict } = require('js-mdict');

const mdict = new Mdict('resources/oald7.mdx');

const def = mdict.lookup('ask');
console.log(def.definition);

/*
<head><link rel="stylesheet" type="text/css" href="O7.css"/></head><body><span class="hw"> ask </span hw><span class="i_g"> <img src="key.gif"/>  /<a class="i_phon" href="sound://aask_ggv_r1_oa013910.spx">ɑ:sk</a i_phon><span class="z">; </span z><i>NAmE</i> <a class="y_phon" href="sound://aask_ggx_r1_wpu01057.spx">æsk</a y_phon>/ </span i_g><span class="cls"> verb</span cls><br><span class="sd">QUESTION<span class="chn"> 问题</span chn></span sd>
<div class="define"><span class="numb">1</span numb><span class="cf"> ~ <span class="bra">(</span bra>sb<span class="bra">)</span bra> <span class="bra">(</span bra>about sb/ sth<span class="bra">)</span bra> </span cf><span class="d">to say or write sth in the form of a question, in order to get information<span class="chn"> 问；询问</span chn></span d></div define>
<span class="phrase"><span class="pt">  [<span class="pt_inside">V <span class="pt_bold">speech</span></span><span>]</span> </span pt></span phrase>
<span class="sentence_eng">'Where are you going?' she asked. </span sentence_eng>
<span class="sentence_chi">"你去哪里？"她问道。</span sentence_chi>
<span class="phrase"><span class="pt"> [<span class="pt_inside">VN <span class="pt_bold">speech</span></span><span>]</span> </span pt></span phrase>
<span class="sentence_eng">'Are you sure?' he asked her. </span sentence_eng>
...
</body>
  */

Benchmark

use oale8.mdx as an example:

Mdict#loading time: 0 sec
Mdict#lookup x 1,967 ops/sec ±0.95% (97 runs sampled)
Mdict#prefix x 1,800 ops/sec ±0.70% (96 runs sampled)
Mdict#fuzzy_search x 285 ops/sec ±14.03% (77 runs sampled)
Mdict#associate x 2,059 ops/sec ±4.53% (89 runs sampled)
Fastest is Mdict#associate

depreciate if you use js-mdict @2.0.3, you can use api shown below:

Note: 2.0.3 not supports mdd file, and record info encrypted file

import path from 'path';
import Mdict from 'js-mdict';

const dictPath = path.join(__dirname, '../resource/Collins.mdx');
const mdict = new Mdict(dictPath);
mdict
  .build()
  .then((_mdict) => {
    console.log('hello', _mdict.lookup('hello'));
    console.log('world', _mdict.lookup('world'));
    console.log(_mdict.attr());
  })
  .catch((err) => {
    console.error(err);
  });

Release

v6.0.0-alpha

implements with TypeScript

BREAKING CHANGES:

delete large files in history
delete suggest, associate etc. functions.

v5.0.0

NOTE: This version has a lot of changes, please read the example code example carefully!

注意：本版本改动了多个 API, 请仔细阅读上方的示例代码。

BREAKING CHANGES:

Refactor the key search algorithm: build the keyList first, and resort internally, then search the word
The suggest method has been deprecated.
The parse_definition method has been deprecated, use fetch_definition instead
Delete nspell and en-dictionary-us dependencies
Cleanup unstable compare method and key record compare, won't use keyInfoBlock.firstKey/lastKey to locate recordBlock
Use built-in TextDecoder and zlib, instead of third party TextDecoder and pako.
use pnpm instead of yarn to manage the dependencies.
Have fun!

v4.0.20

upgrade xmldom to v0.7.5

v4.0.19

support rangeKey API

v4.0.18

add rangeKeyBlock interface
enhance performance of readKeyBlock

v4.0.17

fix associate can't find special character bug

v4.0.16

fix accurate record start offset parse_definition bug

v4.0.15

fix findList return undefined will crash the associate and prefix method bug

v4.0.14

fix babel-runtime dependencies issue

v4.0.13

fix UpperCase key sensitive options logic, details #41
fix 1.2 mdx keyblock read bug
correct some Header properties (StripKey..)

very thanks to @songxiaocheng

v4.0.12

fix typings declaring and reformat codebase

v4.0.11

fix some .mdd file reading issues, and if you search mdd file, use lookup method, and it will return base64 data

v4.0.10

rewrite typings/mdict.d.ts

v4.0.9

rename typings/Mdict.d.ts to typings/mdict.d.ts

v4.0.8

fix uppercase words comparing missed bug
fix out of index error when cannot locate word offset
if cannot find the word key block, return undefined

v4.0.7

rename Mdict.js to mdict.js , rename MdictBase.js to mdict-base.js, fix import error on ubuntu.

v4.0.6

support search words by prefix associate (the phrase as the words' prefix, not the phrase's prefix as search token just like prefix function)
some security updates