Add importer model, validations to paper model

This commit is contained in:
Andreas Haller 2015-05-17 17:49:09 +02:00
parent 568abb631f
commit 46de026234
8 changed files with 67 additions and 13 deletions

3
app/models/importer.rb Normal file
View file

@ -0,0 +1,3 @@
class Importer < ActiveRecord::Base
validates :url, presence: true, uniqueness: true
end

View file

@ -5,20 +5,28 @@ class Paper < ActiveRecord::Base
include Elasticsearch::Model include Elasticsearch::Model
include Elasticsearch::Model::Callbacks include Elasticsearch::Model::Callbacks
validates_presence_of :body, :content, :name, :originator, :paper_type, :published_at, :reference, :url
validates :url, uniqueness: true
class << self class << self
def import_from_json(json_string) def import_from_json(json_string)
old_count = count
JSON.parse(json_string).each do |record| JSON.parse(json_string).each do |record|
attributes = { attributes = {
body: record['body'],
content: record['content'],
name: record['name'], name: record['name'],
url: record['url'], resolution: record['resolution'],
reference: record['reference'],
paper_type: record['paper_type'],
originator: record['originator'], originator: record['originator'],
paper_type: record['paper_type'],
published_at: record['published_at'], published_at: record['published_at'],
reference: record['reference'],
url: record['url'],
} }
record = find_or_initialize_by(url: attributes[:url]) record = find_or_initialize_by(url: attributes[:url])
record.update_attributes!(attributes) record.update_attributes(attributes)
end end
puts "Imported #{count - old_count} Papers!"
end end
end end
end end

View file

@ -5,6 +5,7 @@ class CreatePapers < ActiveRecord::Migration
t.string :url t.string :url
t.string :reference t.string :reference
t.string :name t.string :name
t.string :body
t.datetime :published_at t.datetime :published_at
t.datetime :scraped_at t.datetime :scraped_at
t.string :paper_type t.string :paper_type
@ -14,5 +15,10 @@ class CreatePapers < ActiveRecord::Migration
t.timestamps t.timestamps
end end
add_index(:papers, :reference)
add_index(:papers, :originator)
add_index(:papers, :body)
add_index(:papers, [:reference, :body], unique: true)
end end
end end

View file

@ -0,0 +1,8 @@
class CreateImporters < ActiveRecord::Migration
def change
create_table :importers do |t|
t.string :url
t.timestamps null: false
end
end
end

View file

@ -11,12 +11,19 @@
# #
# It's strongly recommended that you check this file into your version control system. # It's strongly recommended that you check this file into your version control system.
ActiveRecord::Schema.define(version: 20150413193656) do ActiveRecord::Schema.define(version: 20150517152218) do
create_table "papers", force: true do |t| create_table "importers", force: :cascade do |t|
t.string "url"
t.datetime "created_at", null: false
t.datetime "updated_at", null: false
end
create_table "papers", force: :cascade do |t|
t.string "name" t.string "name"
t.string "url" t.string "url"
t.string "reference" t.string "reference"
t.string "body"
t.datetime "published_at" t.datetime "published_at"
t.datetime "scraped_at" t.datetime "scraped_at"
t.string "paper_type" t.string "paper_type"
@ -27,7 +34,12 @@ ActiveRecord::Schema.define(version: 20150413193656) do
t.datetime "updated_at" t.datetime "updated_at"
end end
create_table "users", force: true do |t| add_index "papers", ["body"], name: "index_papers_on_body"
add_index "papers", ["originator"], name: "index_papers_on_originator"
add_index "papers", ["reference", "body"], name: "index_papers_on_reference_and_body", unique: true
add_index "papers", ["reference"], name: "index_papers_on_reference"
create_table "users", force: :cascade do |t|
t.string "email" t.string "email"
t.datetime "created_at" t.datetime "created_at"
t.datetime "updated_at" t.datetime "updated_at"

View file

@ -1,11 +1,10 @@
namespace :import_papers do namespace :import_papers do
desc 'Import Paper records from CSV' desc 'Import Paper records from CSV'
task :from_morph => :environment do |t, args| task :from_morph => :environment do
require 'open-uri' require 'open-uri'
api_key = ENV['MORPH_API_KEY'] Importer.all.each do |importer|
uri = URI.parse "https://api.morph.io/ahx/city_council_leipzig_recent_papers/data.json?key=#{api_key}&query=select%20*%20from%20%27data%27" uri = URI.parse(importer.url)
puts "Download files from #{uri}"
Paper.import_from_json(uri.read) Paper.import_from_json(uri.read)
end end
end end
end

11
test/fixtures/importers.yml vendored Normal file
View file

@ -0,0 +1,11 @@
# Read about fixtures at http://api.rubyonrails.org/classes/ActiveRecord/FixtureSet.html
# This model initially had no columns defined. If you add columns to the
# model remove the '{}' from the fixture names and add the columns immediately
# below each fixture, per the syntax in the comments below
#
one: {}
# column: value
#
two: {}
# column: value

View file

@ -0,0 +1,7 @@
require 'test_helper'
class ImporterTest < ActiveSupport::TestCase
# test "the truth" do
# assert true
# end
end