Super-simple Elm app to read JSON from a REST API, convert it to a nested Elm data structure, and print the result in the browser.

example.json

[
  {
    "japanese": "ボブは魚が好きだ。",
    "english": "Bob likes fish.",
    "tags": [],
    "ve": [
      {
        "word": "ボブ",
        "lemma": "ボブ",
        "part_of_speech": "proper noun",
        "reading": "ボブ",
        "transcription": "ボブ",
        "tokens": [
          {
            "literal": "ボブ",
            "lemma": "ボブ",
            "pos": "名詞"
          }
        ]
      },
      {
        "word": "は",
        "lemma": "は",
        "part_of_speech": "postposition",
        "reading": "ハ",
        "transcription": "ワ",
        "tokens": [
          {
            "literal": "は",
            "lemma": "は",
            "pos": "助詞"
          }
        ]
      },
      {
        "word": "魚",
        "lemma": "魚",
        "part_of_speech": "noun",
        "reading": "サカナ",
        "transcription": "サカナ",
        "tokens": [
          {
            "literal": "魚",
            "lemma": "魚",
            "pos": "名詞"
          }
        ]
      },
      {
        "word": "が",
        "lemma": "が",
        "part_of_speech": "postposition",
        "reading": "ガ",
        "transcription": "ガ",
        "tokens": [
          {
            "literal": "が",
            "lemma": "が",
            "pos": "助詞"
          }
        ]
      },
      {
        "word": "好き",
        "lemma": "好き",
        "part_of_speech": "adjective",
        "reading": "スキ",
        "transcription": "スキ",
        "tokens": [
          {
            "literal": "好き",
            "lemma": "好き",
            "pos": "名詞"
          }
        ]
      },
      {
        "word": "だ",
        "lemma": "だ",
        "part_of_speech": "verb",
        "reading": "ダ",
        "transcription": "ダ",
        "tokens": [
          {
            "literal": "だ",
            "lemma": "だ",
            "pos": "助動詞"
          }
        ]
      },
      {
        "word": "。",
        "lemma": "。",
        "part_of_speech": "symbol",
        "reading": "。",
        "transcription": "。",
        "tokens": [
          {
            "literal": "。",
            "lemma": "。",
            "pos": "記号"
          }
        ]
      }
    ]
  }
]

getsentences.elm

import Json.Decode as Json exposing ((:=))

import Http
import Html exposing (Html)
import Task exposing (Task, andThen)

import Graphics.Element as Element

main =
  Signal.map Element.show sentenceMailbox.signal

-- This is just a plain variable
-- sentenceMailbox : Signal.Mailbox (List String)
sentenceMailbox =
  Signal.mailbox []

-- This is a function
-- report : (List String) -> Task x ()
report input =
  Signal.send sentenceMailbox.address input

-- This is a port, which NEEDS its type signature—can't be inferred
port fetch : Task Http.Error ()
port fetch =
  getSentence 23 `andThen` report

-- This is a function too
--getSentence : Number -> Task Http.Error (List Sentence)
getSentence number =
  Http.get (Json.list sentenceDecoder) ("http://localhost:3003/" ++ toString number)

-- These are my awesome record types ya!
type alias MeCabMorpheme = { literal : String
                           , lemma : String
                           , pos : String}
type alias VeLemma = { word : String
                     , lemma : String
                     , pos : String
                     , reading : String
                     , transcription : String
                     , morphemes : (List MeCabMorpheme)}
type alias Sentence = { japanese : String 
                      , english : String 
                      , lemmas : (List VeLemma)}

-- This is just a function, see http://www.troikatech.com/blog/2015/08/17/decoding-larger-json-objects-in-elm
apply func value =
    Json.object2 (<|) func value

-- These here are variables, building up a nested set of JSON decoders
meCabDecoder =
  Json.map MeCabMorpheme ("literal" := Json.string)
    `apply` ("lemma" := Json.string)
    `apply` ("pos" := Json.string)

veDecoder =
  Json.map VeLemma ("word" := Json.string)
  `apply` ("lemma" := Json.string)
  `apply` ("part_of_speech" := Json.string)
  `apply` ("reading" := Json.string)
  `apply` ("transcription" := Json.string)
  `apply` ("tokens" := (Json.list meCabDecoder))

sentenceDecoder = 
  Json.map Sentence ("japanese" := Json.string) 
  `apply` ("english" := Json.string) 
  `apply` ("ve" := (Json.list veDecoder))

output.txt

[{ english = "Bob likes fish.", japanese = "ボブは魚が好きだ。", lemmas = [{ lemma = "ボブ", morphemes = [{ lemma = "ボブ", literal = "ボブ", pos = "名詞" }], pos = "proper noun", reading = "ボブ", transcription = "ボブ", word = "ボブ" },{ lemma = "は", morphemes = [{ lemma = "は", literal = "は", pos = "助詞" }], pos = "postposition", reading = "ハ", transcription = "ワ", word = "は" },{ lemma = "魚", morphemes = [{ lemma = "魚", literal = "魚", pos = "名詞" }], pos = "noun", reading = "サカナ", transcription = "サカナ", word = "魚" },{ lemma = "が", morphemes = [{ lemma = "が", literal = "が", pos = "助詞" }], pos = "postposition", reading = "ガ", transcription = "ガ", word = "が" },{ lemma = "好き", morphemes = [{ lemma = "好き", literal = "好き", pos = "名詞" }], pos = "adjective", reading = "スキ", transcription = "スキ", word = "好き" },{ lemma = "だ", morphemes = [{ lemma = "だ", literal = "だ", pos = "助動詞" }], pos = "verb", reading = "ダ", transcription = "ダ", word = "だ" },{ lemma = "。", morphemes = [{ lemma = "。", literal = "。", pos = "記号" }], pos = "symbol", reading = "。", transcription = "。", word = "。" }] }]