Add validation of metadata on record save

Gemfile

@@ -58,6 +58,9 @@ gem "faraday"
 gem "oj" # JSON parser
 gem "rabl"
 
+# JSON Schema Validation
+gem "json-schema"
+
 group :development, :test do
   # See https://guides.rubyonrails.org/debugging_rails_applications.html#debugging-with-the-debug-gem
   gem "debug", platforms: %i[mri windows], require: "debug/prelude"

Gemfile.lock

@@ -164,6 +164,8 @@ GEM
     jsbundling-rails (1.3.1)
       railties (>= 6.0.0)
     json (2.8.2)
+    json-schema (5.1.0)
+      addressable (~> 2.8)
     kamal (2.3.0)
       activesupport (>= 7.0)
       base64 (~> 0.2)
@@ -435,6 +437,7 @@ DEPENDENCIES
   govuk-components
   govuk_design_system_formbuilder
   jsbundling-rails
+  json-schema
   kamal
   oj
   pg (~> 1.1)

app/helpers/application_helper.rb

@@ -1,2 +1,10 @@
 module ApplicationHelper
+  def valid_record_tag(record)
+    text, colour = if record.json_valid?
+                     %w[Valid green]
+                   else
+                     %w[Invalid red]
+                   end
+    content_tag :strong, text, class: "govuk-tag govuk-tag--#{colour}"
+  end
 end

app/models/record.rb

@@ -1,5 +1,5 @@
 class Record < ApplicationRecord
-  before_validation :json_parse_metadata
+  before_validation :json_parse_metadata, :store_metadata_validation
 
   validates :name, :metadata, presence: true
 
@@ -11,4 +11,8 @@ def json_parse_metadata
   rescue JSON::ParserError => e
     errors.add :metadata, e.message
   end
+
+  def store_metadata_validation
+    self.json_valid = JsonValidator.valid?(metadata)
+  end
 end

app/services/json_validator.rb

@@ -0,0 +1,17 @@
+class JsonValidator
+  JSON_SCHEMA_PATH = "https://co-cddo.github.io/data-catalogue-metadata/schema/dataset_schema.json".freeze
+
+  def self.valid?(json)
+    new(json).valid?
+  end
+
+  attr_reader :json
+
+  def initialize(json)
+    @json = json
+  end
+
+  def valid?
+    JSON::Validator.validate JSON_SCHEMA_PATH, json
+  end
+end

app/views/records/index.html.erb

@@ -20,7 +20,7 @@
             <%= record.name %>
           </td>
           <td class="govuk-table__cell">
-            <%#= valid_record_tag(record) %>
+            <%= valid_record_tag(record) %>
           </td>
           <td class="govuk-table__cell">
             <%= link_to "Show this record", record %>

db/migrate/20241121094103_add_json_valid_to_records.rb

@@ -0,0 +1,5 @@
+class AddJsonValidToRecords < ActiveRecord::Migration[8.0]
+  def change
+    add_column :records, :json_valid, :boolean
+  end
+end

spec/fixtures/files/dataset.json

@@ -0,0 +1,36 @@
+{
+  "identifier": "8d085327-21b6-4d8b-9705-88faad231d22",
+  "supplierIdentifier": "acme-123",
+  "modified": "2023-12-09T16:09:53+00:00",
+  "status": "Published",
+  "title": "Advance Passenger Information",
+  "description": "Travel data and personal data given to airlines by passenger. API covers  both inbound and outbound air passengers. API includes the passenger’s full name, nationality, date of birth, gender and travel document number, type and country of issue.The data does not include those arriving by sea or rail routes, by private aircraft or via the Common Travel Area (CTA).",
+  "type": "Data Set",
+  "theme": [
+    "Transport and infrastructure",
+    "Population and society"
+  ],
+  "keyword": [
+    "Air travel",
+    "Passport",
+    "Airports",
+    "leaving UK",
+    "entering UK"
+  ],
+"contactPoint": [
+    {
+      "name": "Rob Nichols",
+      "email": "[email protected]"
+    }
+  ],
+  "publisher": "academy-for-social-justice",
+  "securityClassification": "OFFICIAL",
+  "accessRights": "INTERNAL",
+  "distribution": [
+    {
+      "accessService": ["8d085327-21b6-4d8b-9705-88faad231d23"],
+      "downloadURL": "http://example.com/path/to/file.csv",
+      "mediaType":["text/csv"]
+    }
+  ]
+}

spec/models/record_spec.rb

@@ -1,5 +1,20 @@
 require "rails_helper"
 
 RSpec.describe Record, type: :model do
-  pending "add some examples to (or delete) #{__FILE__}"
+  describe "metadata validated on save" do
+    let(:record) { create :record, metadata: { invalid: :content } }
+
+    it "sets json_valid as false if metadata doesn't match schema" do
+      expect(record.json_valid).to be_falsey
+    end
+
+    context "with valid data" do
+      let(:metadata) { json_from_fixture("dataset.json") }
+      let(:record) { create :record, metadata: }
+
+      it "sets json_valid as true" do
+        expect(record.json_valid).to be_truthy
+      end
+    end
+  end
 end

spec/services/json_validator_spec.rb

@@ -0,0 +1,19 @@
+require "rails_helper"
+
+RSpec.describe JsonValidator, type: :service do
+  describe ".valid?" do
+    let(:json) { json_from_fixture("dataset.json") }
+
+    it "is true for valid json" do
+      expect(described_class.valid?(json)).to be_truthy
+    end
+
+    context "with invalid json" do
+      let(:json) { { invalid: :content }.to_json }
+
+      it "is false" do
+        expect(described_class.valid?(json)).not_to be_truthy
+      end
+    end
+  end
+end

spec/support/fixture_helper.rb

@@ -0,0 +1,3 @@
+def json_from_fixture(path)
+  JSON.parse(File.read(fixture_file_upload(path)))
+end

spec/support/webmock.rb

@@ -1 +1,4 @@
 require "webmock/rspec"
+
+# Allow calls to JSON schema
+WebMock.disable_net_connect!(allow: "https://co-cddo.github.io")