JMdict Parser

It is a CLI application that parses a JMdict XML and spits out a JSON file.

Installation

npm i -g jmdict-parser

About

JMdict

JMdict is a Japanese to X (X being English, and a couple of other languages) dictionary whose files are in XML format.

How it works

It simply reads the JMdict file into a Javascript object and writes that object into a file.

Characteristics of the output

The JSON file can be read easily for future use, and it is even ~25% smaller than the original XML file!

Structure of JSON output

The JSON file is essentially an array of entry objects. Here's what one random entry object looks like:

{
   "ent_seq":[
      "1002340"
   ],
   "k_ele":[
      {
         "keb":[
            "お早うございます"
         ],
         "ke_pri":[
            "spec1"
         ]
      }
   ],
   "r_ele":[
      {
         "reb":[
            "おはようございます"
         ],
         "re_pri":[
            "spec1"
         ]
      }
   ],
   "sense":[
      {
         "pos":[
            "∫"
         ],
         "xref":[
            "お早おおう"
         ],
         "misc":[
            "&uk;",
            "&pol;"
         ],
         "s_inf":[
            "may be used more generally at any time of day"
         ],
         "gloss":[
            "good morning"
         ]
      }
   ]
}

For more information about the key names read the comments at the start of the original JMdict files.

Usage

• Head over to the JMdict website and download the version of the dictionary file suitable for you. Also, you would need to extract it as it would be archived.

• Install the CLI. See here.

• You may run the it as follows:

jmdict-parser <your_edict_file>

• This would generate a .json file in the same location as the original file.

Download a sample

I have already created one for the English only JMdict file. You may download it from here.