EMR/Zeppelin set interpreter and run all notebooks

emr-zeppelin-run.rb

#!/usr/bin/env ruby

require 'json'
require 'net/http'
require 'open-uri'
require 'pp'
require 'uri'

ZEP_HOST = 'localhost'
ZEP_PORT = 8890
ZEP_ADDR = "http://#{ZEP_HOST}:#{ZEP_PORT}"

# list notebooks
notebooks = JSON.parse(open("#{ZEP_ADDR}/api/notebook").read)['body']

# get interpreter ids
interpreter_ids = JSON.parse(open("#{ZEP_ADDR}/api/interpreter/setting").read)['body'].map {|i| i['id']}

notebooks.each do |notebook|
  next if notebook['name'] =~ /^(test|sandbox|scratch|quick)/
  next if notebook['name'] !~ /prestaging\.(transform|compute)\./
  puts "--------------------------------------------------"
  pp notebook

  # set interpreter
  uri = URI("#{ZEP_ADDR}/api/notebook/interpreter/bind/#{notebook['id']}")
  request = Net::HTTP::Put.new(uri, 'Content-Type' => 'application/json')
  request.body = interpreter_ids.to_json
  response = Net::HTTP.new(uri.host, uri.port).start {|http| http.request(request) }

  # start notebook
  uri = URI("#{ZEP_ADDR}/api/notebook/job/#{notebook['id']}")
  req = Net::HTTP::Post.new(uri)
  res = Net::HTTP.start(uri.hostname, uri.port) do |http|
    http.request(req)
  end

  pp res
end