Skip to content

Latest commit

 

History

History
98 lines (78 loc) · 2.73 KB

README.md

File metadata and controls

98 lines (78 loc) · 2.73 KB
Raw
  • Under development *

GENGO'S HUMAN TRANSLATION API

Gengo makes it easy to plug human powered translation in to your service or platform.

Install

$npm install gengo-node

require

Gengo = require Gengo
gengoClient = new Gengo {public: YOUR_PUBLIC_KEY, private: your_private_key}

By default the client sends all requests to the Gengo Sandbox environment.

Send a couple of jobs for translation

First we create a couple of jobs that represent a blog post. I've only set a few of the options here, but check out the job payloads for a full list.

blog_post =
  title:
    lc_src: 'en'
    lc_tgt: 'ja'
    tier: 'standard'
    body_src: "This is the title of my blog post"
    custom_data: {blog_post_id: 2322, part: 'title'}
    callback_url: "http://mysite.com/gengo_callback/"
  body:
    lc_src: 'en'
    lc_tgt: 'ja'
    tier: 'standard'
    body_src: "This is the body content of my blog post"
    custom_data: {blog_post_id: 2322, part: 'body'}
    callback_url: "http://mysite.com/gengo_callback/"

There are two important concepts in this payload.

  1. custom_data: Here we've added 2 bits of information that will help us map the translated content back to our own system. In this example a blog post ID and the part of the post.

  2. callback_url: Since there will be real human translators working on the content it may take a bit of time. Once the translation is ready we'll post the translation to the URL provided along with the custom_data and order details.

Since the Gengo API is designed to support thousnads of jobs, there is a queueing mechanism placed in front of the API. This means that when a jobs are sent, we reply with an order_id and put the jobs in a queue.

blog_post_order_id = null
Gengo.postJobs blog_post, (res) ->
  blog_post_order_id = res.order_id
  console.log res

  ###
  {
    "order_id": "139370",
    "group_id": 23015,
    "job_count": "2",
    "credits_used": "3.50",
    "currency": "USD"
  }
  ###

Now that we have the order ID we can check on the status of the order.

blog_post_job_ids = null
Gengo.getOrder blog_post_order_id, (res) ->
  blog_post_job_ids = res.order.jobs_available
  console.log res

  ###
  {
    "order": {
      "order_id": "139370",
      "total_credits": "3.50",
      "currency": "USD",
      "total_units": 17,
      "as_group": 1,
      "jobs_available": [
        "243646",
        "243647",
      ],
      "jobs_pending": [],
      "jobs_reviewable": [],
      "jobs_approved": [],
      "jobs_queued": 0,
      "total_jobs": "2"
    }
  }
  ###