arclight 0.1.0 → 1.6.2

data/solr/conf/solrconfig.xml

@@ -28,8 +28,9 @@
        that you fully re-index after changing this setting as it can
        affect both how text is indexed and queried.
   -->
-  <luceneMatchVersion>5.0.0</luceneMatchVersion>
+  <luceneMatchVersion>8.2.0</luceneMatchVersion>
 
+  <lib dir="${solr.install.dir:../../../..}/modules/analysis-extras/lib" />
   <lib dir="${solr.install.dir:../../../..}/contrib/analysis-extras/lib" />
   <lib dir="${solr.install.dir:../../../..}/contrib/analysis-extras/lucene-libs" />
 
@@ -74,7 +75,15 @@
        <int name="rows">10</int>
 
        <str name="q.alt">*:*</str>
-       <str name="mm">2&lt;-1 5&lt;-2 6&lt;90%</str>
+
+       <!-- Minimum Should Match parameter (mm) -->
+       <str name="mm">4&lt;90%</str>
+       <!-- 1-4 term query, all must match
+            5+ term query, 90% must match; rounded down, so e.g.:
+            * 5-10 term query, all but one must match
+            * 11-20 term query, all but two must match
+            * 21-29 term query, all but three must match, etc.
+       -->
 
        <!-- this qf and pf are used by default, if not otherwise specified by
             client. The default blacklight_config will use these for the
@@ -84,59 +93,106 @@
             http://wiki.apache.org/solr/LocalParams
        -->
 
+       <!-- In general, matches using the pf parameter should be boosted higher
+            than qf, so the proximity of multiple search terms in the matching
+            documents is significant in relevance scoring. See:
+            https://solr.apache.org/guide/solr/latest/query-guide/dismax-query-parser.html#pf-phrase-fields-parameter
+       -->
        <str name="qf">
+         collection_title_tesim^150
          title_tesim^100
-         text^10
+         normalized_title_teim^100
+         unitid_identifier_match^40
+         name_teim^10
+         place_teim^10
+         subject_teim^2
+         id
+         ead_ssi
+         ref_ssm
+         unitid_ssm
+         container_teim
+         parent_unittitles_tesim
+         text
        </str>
        <str name="pf">
-         title_tesim^100
-         text^10
+         collection_title_tesim^300
+         title_tesim^200
+         normalized_title_teim^200
+         unitid_identifier_match^80
+         name_teim^20
+         place_teim^20
+         subject_teim^5
+         id^2
+         ead_ssi^2
+         ref_ssm^2
+         unitid_ssm^2
+         container_teim^2
+         parent_unittitles_tesim^2
+         text^2
+       </str>
+
+       <str name="qf_container">
+         container_teim
+       </str>
+       <str name="pf_container">
+         container_teim^2
+       </str>
+       <str name="qf_identifier">
+         id
+         ead_ssi
+         ref_ssm
+         unitid_ssm
+         unitid_identifier_match
+       </str>
+       <str name="pf_identifier">
+         id^2
+         ead_ssi^2
+         ref_ssm^2
+         unitid_ssm^2
+         unitid_identifier_match^2
+       </str>
+       <str name="qf_name">
+         name_teim
+       </str>
+       <str name="pf_name">
+         name_teim^2
+       </str>
+       <str name="qf_place">
+         place_teim
+       </str>
+       <str name="pf_place">
+         place_teim^2
+       </str>
+       <str name="qf_subject">
+         subject_teim
+       </str>
+       <str name="pf_subject">
+         subject_teim^2
+       </str>
+       <str name="qf_title">
+         title_tesim
+         normalized_title_teim
+       </str>
+       <str name="pf_title">
+         title_tesim^2
+         normalized_title_teim^2
        </str>
 
        <int name="ps">3</int>
        <float name="tie">0.01</float>
 
-       <str name="fl">
-         id,
-         score,
-         abstract_ssm,
-         accessrestrict_ssm,
-         child_component_count_isim,
-         collection_ssm,
-         containers_ssim,
-         creator_ssm,
-         extent_ssm,
-         geogname_ssm,
-         has_online_content_ssim,
-         language_ssm,
-         level_ssm,
-         repository_ssm,
-         scopecontent_ssm,
-         title_ssm,
-         normalized_title_ssm,
-         normalized_date_ssm,
-         unitid_ssm,
-         parent_ssm,
-         parent_unittitles_ssm,
-         ead_ssi,
-         ref_ssm,
-         component_level_isim,
-         parent_access_restrict_ssm,
-         parent_access_terms_ssm,
-         names_coll_ssim
-         places_ssim
-       </str>
+       <str name="fl">*</str>
 
        <str name="facet">true</str>
        <str name="facet.mincount">1</str>
-       <str name="facet.field">level_sim</str>
-       <str name="facet.field">creator_sim</str>
-       <str name="facet.field">date_range_sim</str>
-       <str name="facet.field">names_sim</str>
-       <str name="facet.field">geogname_sim</str>
-       <str name="facet.field">access_subjects_sim</str>
-       <str name="facet.field">repository_sim</str>
-       <str name="facet.field">collection_sim</str>
+       <str name="facet.field">level_ssim</str>
+       <str name="facet.field">creator_ssim</str>
+       <str name="facet.field">date_range_isim</str>
+       <str name="facet.field">names_ssim</str>
+       <str name="facet.field">geogname_ssim</str>
+       <str name="facet.field">access_subjects_ssim</str>
+       <str name="facet.field">repository_ssim</str>
+       <str name="facet.field">collection_ssim</str>
 
        <str name="spellcheck">true</str>
        <str name="spellcheck.dictionary">default</str>
@@ -150,127 +206,6 @@
     </arr>
   </requestHandler>
 
-  <!-- Field-based searches -->
-  <requestHandler name="name_search" class="solr.SearchHandler">
-    <lst name="defaults">
-      <str name="defType">edismax</str>
-      <str name="echoParams">explicit</str>
-      <int name="rows">10</int>
-      <str name="q.alt">*:*</str>
-      <str name="mm">2&lt;-1 5&lt;-2 6&lt;90%</str>
-      <str name="df">name_tesim</str>
-      <str name="q.op">AND</str>
-
-      <str name="qf">
-        name_tesim
-      </str>
-      <str name="pf">
-        name_tesim
-      </str>
-
-      <str name="fl">*</str>
-      <str name="facet">true</str>
-      <str name="facet.mincount">1</str>
-      <str name="facet.field">level_sim</str>
-      <str name="facet.field">creator_sim</str>
-      <str name="facet.field">date_range_sim</str>
-      <str name="facet.field">names_sim</str>
-      <str name="facet.field">geogname_sim</str>
-      <str name="facet.field">access_subjects_sim</str>
-      <str name="facet.field">repository_sim</str>
-    </lst>
-  </requestHandler>
-
-  <requestHandler name="place_search" class="solr.SearchHandler">
-    <lst name="defaults">
-      <str name="defType">edismax</str>
-      <str name="echoParams">explicit</str>
-      <int name="rows">10</int>
-      <str name="q.alt">*:*</str>
-      <str name="mm">2&lt;-1 5&lt;-2 6&lt;90%</str>
-      <str name="df">place_tesim</str>
-      <str name="q.op">AND</str>
-
-      <str name="qf">
-        place_tesim
-      </str>
-      <str name="pf">
-        place_tesim
-      </str>
-
-      <str name="fl">*</str>
-      <str name="facet">true</str>
-      <str name="facet.mincount">1</str>
-      <str name="facet.field">level_sim</str>
-      <str name="facet.field">creator_sim</str>
-      <str name="facet.field">date_range_sim</str>
-      <str name="facet.field">names_sim</str>
-      <str name="facet.field">geogname_sim</str>
-      <str name="facet.field">access_subjects_sim</str>
-      <str name="facet.field">repository_sim</str>
-    </lst>
-  </requestHandler>
-
-  <requestHandler name="subject_search" class="solr.SearchHandler">
-    <lst name="defaults">
-      <str name="defType">edismax</str>
-      <str name="echoParams">explicit</str>
-      <int name="rows">10</int>
-      <str name="q.alt">*:*</str>
-      <str name="mm">2&lt;-1 5&lt;-2 6&lt;90%</str>
-      <str name="df">subject_tesim</str>
-      <str name="q.op">AND</str>
-
-      <str name="qf">
-        subject_tesim
-      </str>
-      <str name="pf">
-        subject_tesim
-      </str>
-
-      <str name="fl">*</str>
-      <str name="facet">true</str>
-      <str name="facet.mincount">1</str>
-      <str name="facet.field">level_sim</str>
-      <str name="facet.field">creator_sim</str>
-      <str name="facet.field">date_range_sim</str>
-      <str name="facet.field">names_sim</str>
-      <str name="facet.field">geogname_sim</str>
-      <str name="facet.field">access_subjects_sim</str>
-      <str name="facet.field">repository_sim</str>
-    </lst>
-  </requestHandler>
-
-  <requestHandler name="title_search" class="solr.SearchHandler">
-    <lst name="defaults">
-      <str name="defType">edismax</str>
-      <str name="echoParams">explicit</str>
-      <int name="rows">10</int>
-      <str name="q.alt">*:*</str>
-      <str name="mm">2&lt;-1 5&lt;-2 6&lt;90%</str>
-      <str name="df">title_tesim</str>
-      <str name="q.op">AND</str>
-
-      <str name="qf">
-        title_tesim
-      </str>
-      <str name="pf">
-        title_tesim
-      </str>
-
-      <str name="fl">*</str>
-      <str name="facet">true</str>
-      <str name="facet.mincount">1</str>
-      <str name="facet.field">level_sim</str>
-      <str name="facet.field">creator_sim</str>
-      <str name="facet.field">date_range_sim</str>
-      <str name="facet.field">names_sim</str>
-      <str name="facet.field">geogname_sim</str>
-      <str name="facet.field">access_subjects_sim</str>
-      <str name="facet.field">repository_sim</str>
-    </lst>
-  </requestHandler>
-
   <!-- for requests to get a single document; use id=666 instead of q=id:666 -->
   <requestHandler name="document" class="solr.SearchHandler" >
     <lst name="defaults">

data/tasks/arclight.rake

@@ -5,7 +5,14 @@ require 'engine_cart/rake_task'
 require 'rspec/core/rake_task'
 require 'arclight'
 
-EngineCart.fingerprint_proc = EngineCart.rails_fingerprint_proc
+class DependencyNotInstalled < StandardError; end
+
+# Build with our opinionated defaults if none are provided.
+rails_options = ENV.fetch('ENGINE_CART_RAILS_OPTIONS', '')
+rails_options = "#{rails_options} -a propshaft" unless rails_options.match?(/-a\s|--asset-pipeline/)
+rails_options = "#{rails_options} -j importmap" unless rails_options.match?(/-j\s|--javascript/)
+rails_options = "#{rails_options} --css bootstrap" unless rails_options.match?(/--css/)
+ENV['ENGINE_CART_RAILS_OPTIONS'] = rails_options
 
 desc 'Run test suite'
 task ci: %w[arclight:generate] do
@@ -13,7 +20,7 @@ task ci: %w[arclight:generate] do
     solr.with_collection do
       Rake::Task['arclight:seed'].invoke
       within_test_app do
-        ## Do stuff inside arclight app here
+        system 'bin/rake spec:prepare'
       end
       Rake::Task['spec'].invoke
     end
@@ -22,7 +29,9 @@ end
 
 desc 'Run Eslint'
 task :eslint do
-  system './node_modules/.bin/eslint app/assets/javascripts'
+  raise DependencyNotInstalled, 'ESLint not found.  Please run yarn install.' unless File.exist?('./node_modules/.bin/eslint')
+
+  exit 1 unless system './node_modules/.bin/eslint app/assets/javascripts'
 end
 
 namespace :arclight do
@@ -45,7 +54,7 @@ namespace :arclight do
       solr.with_collection do
         Rake::Task['arclight:seed'].invoke
         within_test_app do
-          system "bundle exec rails s #{args[:rails_server_args]}"
+          system "bin/dev #{args[:rails_server_args]}"
         end
       end
     end
@@ -54,15 +63,17 @@ namespace :arclight do
   desc 'Seed fixture data to Solr'
   task :seed do
     puts 'Seeding index with data from spec/fixtures/ead...'
-    Dir.glob('spec/fixtures/ead/*.xml').each do |file|
-      system("FILE=#{file} rake arclight:index") # no REPOSITORY_ID
-    end
     Dir.glob('spec/fixtures/ead/*').each do |dir|
       next unless File.directory?(dir)
-      system("REPOSITORY_ID=#{File.basename(dir)} " \
-             'REPOSITORY_FILE=spec/fixtures/config/repositories.yml ' \
-             "DIR=#{dir} " \
-             'rake arclight:index_dir')
+
+      within_test_app do
+        # Sets the REPOSITORY_ID to the name of the file's containing directory
+        system("REPOSITORY_ID=#{File.basename(dir)} " \
+               "REPOSITORY_FILE=#{Arclight::Engine.root}/spec/fixtures/config/repositories.yml " \
+               "DIR=#{Arclight::Engine.root}/#{dir} " \
+               'SOLR_URL=http://127.0.0.1:8983/solr/blacklight-core ' \
+               'rake arclight:index_dir')
+      end
     end
   end
 end

data/template.rb

@@ -1,15 +1,10 @@
 # frozen_string_literal: true
 
-# We can remove `blacklight` and `blacklight_range_limit` once they are released
-# since they are already dependents of Arclight (but we need to pin to particular branches)
-gem 'blacklight', github: 'projectblacklight/blacklight'
-gem 'blacklight_range_limit', github: 'projectblacklight/blacklight_range_limit', branch: 'blacklight-7'
+gem 'arclight'
 
-gem 'arclight', github: 'sul-dlss/arclight'
+after_bundle do
+  generate 'blacklight:install', '--devise'
+  generate 'arclight:install', '-f'
 
-run 'bundle install'
-
-generate 'blacklight:install', '--devise'
-generate 'arclight:install', '-f'
-
-rake 'db:migrate'
+  rake 'db:migrate'
+end