Merge pull request #228 from MITLibraries/use-66-tacos

matt-bernhardt · web-flow · commit 5ec7532ffb36 · 2025-10-07T13:14:38.000-04:00
Enable TACOS integration
diff --git a/.env.test b/.env.test
@@ -7,6 +7,9 @@ PRIMO_SCOPE=cdi
 PRIMO_TAB=all
 PRIMO_VID=01MIT_INST:MIT
 SYNDETICS_PRIMO_URL=https://syndetics.com/index.php?client=primo
+TACOS_HOST=FAKE_TACOS_HOST
+TACOS_URL=http://FAKE_TACOS_HOST/graphql
+TACOS_SOURCE=FAKE_TACOS_SOURCE
 TIMDEX_GRAPHQL=https://FAKE_TIMDEX_HOST/graphql
 TIMDEX_HOST=FAKE_TIMDEX_HOST
 TIMDEX_INDEX=FAKE_TIMDEX_INDEX
diff --git a/README.md b/README.md
@@ -128,6 +128,7 @@ mode (e.g., `GDT=false` will still enable GDT features). Note that this is curre
 may have unexpected consequences if applied to other TIMDEX UI apps.
 - `GLOBAL_ALERT`: The main functionality for this comes from our theme gem, but when set the value will be rendered as
   safe html above the main header of the site.
+- `ORIGINS`: sets origins for CORS (currently used only for TACOS API calls).
 - `PLATFORM_NAME`: The value set is added to the header after the MIT Libraries logo. The logic and CSS for this comes from our theme gem.
 - `PRIMO_TIMEOUT`: The number of seconds before a Primo request times out (default 6).
 - `REQUESTS_PER_PERIOD` - number of requests that can be made for general throttles per `REQUEST_PERIOD`
@@ -136,13 +137,21 @@ may have unexpected consequences if applied to other TIMDEX UI apps.
 - `REDIRECT_REQUEST_PERIOD`- time in minutes used along with `REDIRECT_REQUEST_PERIOD`
 - `SENTRY_DSN`: Client key for Sentry exception logging.
 - `SENTRY_ENV`: Sentry environment for the application. Defaults to 'unknown' if unset.
+- `TACOS_SOURCE`: If set, this value is sent to TACOS (as the `sourceSystem` value) to distinguish which application
+instance is sending what search traffic. Defaults to "unset" if not defined.
+- `TACOS_URL`: The GraphQL endpoint for the [TACOS API](https://github.com/mitlibraries/tacos/). When set, the
+  application will log search terms to TACOS (and eventually return suggested resources that TACOS detects).
 - `TIMDEX_INDEX`: Name of the index, or alias, to provide to the GraphQL endpoint. Defaults to `nil` which will let TIMDEX determine the best index to use. Wildcard values can be set, for example `rdi*` would search any indexes that begin with `rdi` in the underlying OpenSearch instance behind TIMDEX.
 - `TIMDEX_SOURCES`: Comma-separated list of sources to display in the advanced-search source selection element. This
   overrides the default which is set in ApplicationHelper.
 
 #### Test Environment-only Variables
 
 - `SPEC_REPORTER`: Optional variable. If set, enables spec reporter style output from tests rather than minimal output.
+- `TACOS_HOST`: Test Env only. Used to ensure the VCR cassettes can properly scrub specific host data to make sure we
+get the same cassettes regardless of which host was used to generate the cassettes. This should be set to the hostname
+that matches `TACOS_URL`. Ex: If `TACOS_URL` is `http://localhost:3001/graphql` then `TACOS_HOST` should be
+`localhost:3001`.
 - `TIMDEX_HOST`: Test Env only. Used to ensure the VCR cassettes can properly scrub specific host data to make sure we get the same cassettes regardless of which host was used to generate the cassettes. This should be set to the host name that matches `TIMDEX_GRAPHQL`. Ex: If `TIMDEX_GRAPHQL` is `https://www.example.com/graphql` then `TIMDEX_HOST` should be `www.example.com`.
 
 ### Generating VCR Cassettes
diff --git a/app/controllers/tacos_controller.rb b/app/controllers/tacos_controller.rb
@@ -0,0 +1,9 @@
+class TacosController < ApplicationController
+  layout false
+
+  def analyze
+    return unless ApplicationHelper.tacos_enabled?
+
+    Tacos.analyze(params[:q])
+  end
+end
diff --git a/app/helpers/application_helper.rb b/app/helpers/application_helper.rb
@@ -1,4 +1,9 @@
 module ApplicationHelper
+  def tacos_enabled?
+    ENV.fetch('TACOS_URL', '').present?
+  end
+  module_function :tacos_enabled?
+
   def timdex_sources
     ENV.fetch('TIMDEX_SOURCES', timdex_source_defaults).split(',')
   end
diff --git a/app/models/tacos.rb b/app/models/tacos.rb
@@ -0,0 +1,49 @@
+class Tacos
+  # The tacos_client argument here is unused in production - it is provided for
+  # our test suite so that we can mock various error conditions to ensure that
+  # error handling happens as we intend.
+  def self.analyze(term, tacos_client = nil)
+    tacos_http = setup(tacos_client)
+    query = '{ "query": "{ logSearchEvent(searchTerm: \"' + clean_term(term) + '\", sourceSystem: \"' + tacos_source + '\" ) { phrase source detectors { suggestedResources { title url } } } }" }'
+    begin
+      raw_response = tacos_http.timeout(http_timeout).post(tacos_url, body: query)
+      JSON.parse(raw_response.to_s)
+    rescue HTTP::Error
+      {"error" => "A connection error has occurred"}
+    rescue JSON::ParserError
+      {"error" => "A parsing error has occurred"}
+    end
+  end
+
+  private
+
+  def self.clean_term(term)
+    term.gsub('"', '\'')
+  end
+
+  def self.http_timeout
+    ENV.fetch('TIMDEX_TIMEOUT', 6).to_f
+  end
+
+  def self.origins
+    ENV.fetch('ORIGINS', nil)
+  end
+
+  # We define the HTTP connection this way so that it can be overridden during
+  # testing, to make sure that the .analyze method can handle specific error
+  # conditions.
+  def self.setup(tacos_client)
+    tacos_client || HTTP.persistent(tacos_url)
+                        .headers(accept: 'application/json',
+                                 'Content-Type': 'application/json',
+                                 origin: origins)
+  end
+
+  def self.tacos_source
+    ENV.fetch('TACOS_SOURCE', 'timdexui_unset')
+  end
+
+  def self.tacos_url
+    ENV.fetch('TACOS_URL', nil)
+  end
+end
diff --git a/app/views/search/_trigger_tacos.html.erb b/app/views/search/_trigger_tacos.html.erb
@@ -0,0 +1,8 @@
+<% return unless (tacos_enabled? and @enhanced_query[:q].present?) %>
+
+<% data_url = "/analyze?q=#{URI.encode_www_form_component(@enhanced_query[:q])}" %>
+
+<div class="tacos-container"
+     data-controller="content-loader"
+     data-content-loader-url-value=<%= data_url %>>
+</div>
diff --git a/app/views/search/results.html.erb b/app/views/search/results.html.erb
@@ -68,4 +68,6 @@
   <% end %>
 </div>
 
+<%= render(partial: 'trigger_tacos') if tacos_enabled? %>
+
 <%= javascript_include_tag "filters" %>
diff --git a/app/views/tacos/analyze.html.erb b/app/views/tacos/analyze.html.erb
@@ -0,0 +1 @@
+<!-- Result of TACOS analysis would go here -->
diff --git a/config/routes.rb b/config/routes.rb
@@ -8,6 +8,8 @@
   get 'issn', to: 'fact#issn'
   get 'pmid', to: 'fact#pmid'
 
+  get 'analyze', to: 'tacos#analyze'
+
   get 'record/(:id)',
       to: 'record#view',
       as: 'record',
diff --git a/test/controllers/search_controller_test.rb b/test/controllers/search_controller_test.rb
@@ -275,7 +275,7 @@ def setup
       end
       assert_response :success
 
-      assert_select('div[data-content-loader-url-value]', 0)
+      assert_select('div[data-content-loader-url-value].fact-container', 0)
     end
   end
 
@@ -305,7 +305,7 @@ def setup
 
       assert_response :success
 
-      assert_select('div[data-content-loader-url-value]', 0)
+      assert_select('div[data-content-loader-url-value].fact-container', 0)
     end
   end
 
@@ -333,7 +333,7 @@ def setup
       end
       assert_response :success
 
-      assert_select('div[data-content-loader-url-value]', 0)
+      assert_select('div[data-content-loader-url-value].fact-container', 0)
     end
   end
 
@@ -361,7 +361,31 @@ def setup
       end
       assert_response :success
 
-      assert_select('div[data-content-loader-url-value]', 0)
+      assert_select('div[data-content-loader-url-value].fact-container', 0)
+    end
+  end
+
+  test 'TACOS intervention is inserted when TACOS enabled' do
+    VCR.use_cassette('tacos',
+                     allow_playback_repeats: true) do
+      get '/results?q=tacos'
+
+      assert_response :success
+
+      tacos_div = assert_select('div[data-content-loader-url-value].tacos-container')
+      assert_equal '/analyze?q=tacos', tacos_div.attribute('data-content-loader-url-value').value
+    end
+  end
+
+  test 'TACOS intervention not inserted when TACOS not enabled' do
+    VCR.use_cassette('tacos',
+                     allow_playback_repeats: true) do
+      ClimateControl.modify(TACOS_URL: '') do
+        get '/results?q=tacos'
+      end
+      assert_response :success
+
+      assert_select('div[data-content-loader-url-value].tacos-container', 0)
     end
   end
 
diff --git a/test/controllers/tacos_controller_test.rb b/test/controllers/tacos_controller_test.rb
@@ -0,0 +1,12 @@
+require 'test_helper'
+
+class TacosControllerTest < ActionDispatch::IntegrationTest
+  test 'analyze route exists but returns an HTML comment for now' do
+    VCR.use_cassette('tacos direct') do
+      get '/analyze?q=direct'
+
+      assert_response :success
+      assert_equal "<!-- Result of TACOS analysis would go here -->\n", response.body
+    end
+  end
+end
diff --git a/test/helpers/application_helper_test.rb b/test/helpers/application_helper_test.rb
@@ -3,6 +3,16 @@
 class ApplicationHelperTest < ActionView::TestCase
   include ApplicationHelper
 
+  test 'tacos_enabled? requires TACOS_URL env variable to have non-zero length' do
+    ClimateControl.modify TACOS_URL: 'foo' do
+      assert_equal true, tacos_enabled?
+    end
+
+    ClimateControl.modify TACOS_URL: '' do
+      assert_equal false, tacos_enabled?
+    end
+  end
+
   test 'index_page_title renders as expected' do
     assert_equal 'Search | MIT Libraries', index_page_title
   end
diff --git a/test/models/tacos_test.rb b/test/models/tacos_test.rb
@@ -0,0 +1,62 @@
+require 'test_helper'
+
+class TacosConnectionError
+  def timeout(_)
+    self
+  end
+
+  def post(_url, body:)
+    raise HTTP::ConnectionError, "forced connection failure"
+  end
+end
+
+class TacosParsingError
+  def timeout(_)
+    self
+  end
+
+  def post(_url, body:)
+    'This is not valid json'
+  end
+end
+
+class TacosTest < ActiveSupport::TestCase
+  test 'TACOS model has a call method that reflects a search term back' do
+    VCR.use_cassette('tacos popcorn') do
+      searchterm = 'popcorn'
+
+      result = Tacos.analyze(searchterm)
+
+      assert_instance_of Hash, result
+      assert_equal searchterm, result['data']['logSearchEvent']['phrase']
+    end
+  end
+
+  test 'TACOS model will use ENV to populate the sourceSystem value' do
+    VCR.use_cassette('tacos fake system') do
+      ClimateControl.modify(TACOS_SOURCE: 'faked') do
+        result = Tacos.analyze('popcorn')
+
+        assert_equal 'faked', result['data']['logSearchEvent']['source']
+      end
+    end
+  end
+
+  test 'TACOS model catches connection errors' do
+    tacos_client = TacosConnectionError.new
+
+    result = Tacos.analyze('popcorn', tacos_client)
+
+    assert_instance_of Hash, result
+    assert_equal 'A connection error has occurred', result['error']
+  end
+
+  test 'TACOS model catches parsing errors' do
+    tacos_client = TacosParsingError.new
+
+    result = Tacos.analyze('popcorn', tacos_client)
+
+    assert_instance_of Hash, result
+    assert_equal 'A parsing error has occurred', result['error']
+  end
+end
diff --git a/test/test_helper.rb b/test/test_helper.rb
@@ -18,14 +18,16 @@
 end
 
 VCR.configure do |config|
-  config.ignore_localhost = true
+  config.ignore_localhost = false
   config.cassette_library_dir = 'test/vcr_cassettes'
   config.hook_into :webmock
   config.allow_http_connections_when_no_cassette = false
   config.filter_sensitive_data('FAKE_TIMDEX_HOST') { ENV.fetch('TIMDEX_HOST').to_s }
   config.filter_sensitive_data('http://FAKE_TIMDEX_HOST/graphql/') { ENV.fetch('TIMDEX_GRAPHQL').to_s }
   config.filter_sensitive_data('FAKE_TIMDEX_INDEX') { ENV.fetch('TIMDEX_INDEX').to_s }
   config.filter_sensitive_data('FAKE_PRIMO_API_KEY') { ENV.fetch('PRIMO_API_KEY').to_s }
+  config.filter_sensitive_data('FAKE_TACOS_HOST') { ENV.fetch('TACOS_HOST').to_s }
+  config.filter_sensitive_data('http://FAKE_TACOS_HOST/graphql/') { ENV.fetch('TACOS_URL').to_s }
 end
 
 module ActiveSupport
diff --git a/test/vcr_cassettes/tacos.yml b/test/vcr_cassettes/tacos.yml
diff --git a/test/vcr_cassettes/tacos_direct.yml b/test/vcr_cassettes/tacos_direct.yml
diff --git a/test/vcr_cassettes/tacos_fake_system.yml b/test/vcr_cassettes/tacos_fake_system.yml
diff --git a/test/vcr_cassettes/tacos_popcorn.yml b/test/vcr_cassettes/tacos_popcorn.yml

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	`+<!-- Result of TACOS analysis would go here -->`