Update Bento to use TIMDEX V2 fields

Why these changes are being introduced:

Bento currently uses deprecated fields in its TIMDEX searches:

* `identifier`
* `publicationDate`
* `source`

Relevant ticket(s):

* https://mitlibraries.atlassian.net/browse/TIMX-174

How this addresses that need:

This makes the following V1 to V2 replacements:

* `identifier` with `identifiers` (`kind` == 'Collection Identifer')
* `publicationDate` with `dates` (there is some nuance to this;
see the changes to the NormalizeTimdex model for details)
* `source` with `sourceFacet`

Side effects of this change:

* Some cassettes have been regenerated to use the new search params.
* Tests that anticipate results from non-ASpace sources have been
removed. These tests would fail if we were to regenerate cassettes
in the current codebase, since the ASpace source is hardcoded in the
search string.
* Some test assertions have been modified to match the latest
TIMDEX data load.

Author: jazairi

app/models/normalize_timdex.rb

@@ -24,7 +24,7 @@ def extract_results(results, norm)
       result_title = extract_title(record)
       result = Result.new(result_title, record['sourceLink'])
       result.blurb = extract_first_from_list(record['summary'])
-      result.year = extract_field(record['publicationDate'])
+      result.year = extract_dates(record['dates'])
       result.authors = extract_field(extract_list(record['contributors']))
       result.physical_description = extract_field(record['physicalDescription'])
       norm['results'] << result
@@ -33,10 +33,36 @@ def extract_results(results, norm)
 
   def extract_title(record)
     title = record['title']
-    title << (" (#{record['identifier'].to_s.gsub('MIT:archivespace:', '')})") if record['identifier'].present?
+    title << extract_collection_id(record['identifiers']) if record['identifiers'].present?
     title
   end
 
+  def extract_dates(dates)
+    # It is unlikely for a record to have more than one creation date, but just in case...
+    relevant_dates = dates.select { |date| date['kind'] == 'creation' }
+
+    # If the record *does* have more than one creation date, it's probably not worth determining which to display.
+    return if relevant_dates.count > 1
+
+    relevant_date = relevant_dates.first
+
+    # We are only concerned with creation dates that are ranges, since we harvest ASpace metadata at the collection
+    # level.
+    return unless relevant_date['kind'] == 'creation' && relevant_date['range'].present?
+
+    "#{relevant_date['range']['gte']}-#{relevant_date['range']['lte']}"
+  end
+
+  def extract_collection_id(identifiers)
+    relevant_ids = identifiers.map { |id| id['value'] if id['kind'] == 'Collection Identifier' }.compact
+
+    # In the highly unlikely event that there is more than one collection identifier, there's something weird going
+    # on with the record and we should skip it.
+    return if relevant_ids.count > 1
+
+    " (#{relevant_ids.first})"
+  end
+
   def extract_list(contributors)
     contributors&.map do |creator|
       [creator['value'], '&field[]=creators_text&q[]=' << URI.encode_www_form_component(creator['value'])]

app/models/search_timdex.rb

@@ -23,7 +23,7 @@ def initialize
   # @param term [string] The string we are searching for
   # @return [Hash] A Hash with search metadata and an Array of {Result}s
   def search(term)
-    @query = '{"query":"{search(searchterm: \"' + clean_term(term) + '\", source: \"MIT ArchivesSpace\") {hits records {sourceLink title identifier publicationDate physicalDescription summary contributors { value } } } }"}'
+    @query = '{ "query": "{ search(searchterm: \"' + clean_term(term) + '\", sourceFilter: \"MIT ArchivesSpace\") { hits records { sourceLink title identifiers { kind value } dates { kind value range { gte lte } } physicalDescription summary contributors { value } } } }"}'
     results = @timdex_http.timeout(http_timeout)
                           .post(TIMDEX_URL, body: @query)
     json_result = JSON.parse(results.to_s)

config/environments/test.rb

@@ -1,4 +1,4 @@
-require "active_support/core_ext/integer/time"
+require 'active_support/core_ext/integer/time'
 
 # The test environment is used exclusively to run your application's
 # test suite. You never need to work with it otherwise. Remember that
@@ -14,6 +14,7 @@
   ENV['FLIPFLOP_KEY'] = 'yoyo'
   ENV['ALMA_OPENURL'] = 'https://na06.alma.exlibrisgroup.com/view/uresolver/01MIT_INST/openurl?'
   ENV['ALMA_SRU'] = 'https://mit.alma.exlibrisgroup.com/view/sru/01MIT_INST?version=1.2&operation=searchRetrieve&recordSchema=marcxml&query=alma.all_for_ui='
+  ENV['ASPACE_SEARCH_URI'] = 'https://archivesspace.mit.edu/search?op[]=&field[]=creators_text&q[]='
   ENV['EXL_INST_ID'] = '01MIT_INST'
   ENV['PRIMO_SEARCH'] = 'true'
   ENV['PRIMO_API_URL'] = 'https://another_fake_server.example.com/v1'
@@ -26,7 +27,7 @@
   ENV['PRIMO_SPLASH_PAGE'] = 'https://libraries.mit.edu/news/library-platform-before/32066/'
   ENV['MIT_PRIMO_URL'] = 'https://mit.primo.exlibrisgroup.com'
   ENV['SYNDETICS_PRIMO_URL'] = 'https://syndetics.com/index.php?client=primo'
-  ENV['TIMDEX_URL']='https://timdex.mit.edu/graphql'
+  ENV['TIMDEX_URL'] = 'https://timdex.mit.edu/graphql'
 
   # Turn false under Spring and add config.action_view.cache_template_loading = true.
   config.cache_classes = true

test/integration/search_test.rb

@@ -70,10 +70,13 @@ def teardown
   end
 
   test 'timdex results are populated' do
-    VCR.use_cassette('popcorn timdex',
+    VCR.use_cassette('aspace timdex',
                      allow_playback_repeats: true) do
-      get '/search/search_boxed?q=popcorn&target=timdex'
+      get '/search/search_boxed?q=archives&target=timdex'
       assert_response :success
+      assert_select 'a.bento-link' do |value|
+        assert value.text.include? 'Paul Earls'
+      end
     end
   end
 

test/models/normalize_timdex_test.rb

@@ -1,14 +1,6 @@
 require 'test_helper'
 
 class NormalizeTimdexTest < ActiveSupport::TestCase
-  def timdex_records
-    VCR.use_cassette('popcorn timdex',
-                     allow_playback_repeats: true) do
-      raw_query = SearchTimdex.new.search('popcorn')
-      NormalizeTimdex.new.to_result(raw_query, 'popcorn')
-    end
-  end
-
   def aspace_records
     VCR.use_cassette('aspace timdex',
                      allow_playback_repeats: true) do
@@ -17,30 +9,9 @@ def aspace_records
     end
   end
 
-  test 'normalized timdex records have expected title' do
-    assert_equal(
-      'Popcorn Venus /',
-      timdex_records['results'][0].title
-    )
-  end
-
-  test 'normalized timdex records have expected url' do
-    assert_equal(
-      'https://library.mit.edu/item/000346597',
-      timdex_records['results'][0].url
-    )
-  end
-
-  test 'normalized timdex records have expected record count' do
-    assert_equal(
-      115,
-      timdex_records['total']
-    )
-  end
-
   test 'normalized aspace records have an identifier in the title' do
     assert_match(
-      'AC.0129',
+      'MC.0552',
       aspace_records['results'][0].title
     )
   end
@@ -51,4 +22,18 @@ def aspace_records
       aspace_records['results'][0].physical_description
     )
   end
+
+  test 'normalized aspace records have a creation date range' do
+    assert_equal '1914-1998', aspace_records['results'][0].year
+  end
+
+  test 'normalized aspace records have an array of contributors' do
+    assert aspace_records['results'][0].authors.is_a? Array
+    assert_equal 'Earls, Paul, 1934-1998', aspace_records['results'][0].authors[0][0]
+  end
+
+  test 'normalized aspace records have a summary' do
+    assert_match 'The Paul Earls archives contains a large number of music manuscript',
+                 aspace_records['results'][0].blurb
+  end
 end

test/models/search_timdex_test.rb

@@ -10,10 +10,11 @@ def after
   end
 
   test 'can search timdex' do
-    VCR.use_cassette('popcorn timdex',
+    VCR.use_cassette('aspace timdex',
                      allow_playback_repeats: true) do
-      query = SearchTimdex.new.search('popcorn')
-      assert_equal(Hash, query.class)
+      response = SearchTimdex.new.search('archives')
+      assert_equal Hash, response.class
+      assert response['data']['search']['hits'] > 0
     end
   end