gdor-indexer 0.1.0 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 5e1bc32ca2df58b2018dbc9f27aa2ce897cde0e4
-  data.tar.gz: c9a3378b959896f6007a7bdbb187069102a7897c
+  metadata.gz: 64f3819be05778daa18edefbef0db3c415d8a7af
+  data.tar.gz: 379426de04d4160b4f2334a54d3308fc8d798edd
 SHA512:
-  metadata.gz: c401ef958ff8bd02ce867cc066cbf9f40d6127bbc4d4916a79e08d158d694ea2ad4c27e633e2a0f27b8dccf5c623259b4f23df779692d0551f58979040ff0cf6
-  data.tar.gz: f0a797d7d3df7bfbdd4daa75d19a215682eb08ca885c437341a7c02b02e6bfca0b35fdc165d9116cbfb0f9b6eec1cd89de84501a724d1c5b53ca0d4969bc98c7
+  metadata.gz: d42cbf9c5d8ee8b9a2038b2ea77d0b51dc89f0fe7633ab2a6b5b17039fd25544c8e6397ae8c6e29b648cc5ae5ba3580bbe2f17ef9f708dc1134dc3414e13f8f6
+  data.tar.gz: 4c44ff23c66af8336d00ef80b747dc28a833a9bc1bbf564c0fd048c0cd4b33f022c7ef4338015f5e1a9181f107ae490eb821afd125f6c6eaa6551d6f8630880e

data/.rubocop.yml

@@ -1,3 +1,8 @@
 require: rubocop-rspec
 
 inherit_from: .rubocop_todo.yml
+
+Style/StringLiterals:
+  Enabled: true
+  EnforcedStyle: single_quotes
+  

data/.rubocop_todo.yml

@@ -1,6 +1,6 @@
 # This configuration was generated by
 # `rubocop --auto-gen-config`
-# on 2015-10-26 16:04:23 -0700 using RuboCop version 0.34.2.
+# on 2015-10-28 13:42:33 -0700 using RuboCop version 0.34.2.
 # The point is for the user to remove these configuration records
 # one by one as the offenses are removed from the code base.
 # Note that changes in the inspected code, or installation of new
@@ -37,9 +37,9 @@ Metrics/AbcSize:
 # Offense count: 1
 # Configuration parameters: CountComments.
 Metrics/ClassLength:
-  Max: 233
+  Max: 242
 
-# Offense count: 4
+# Offense count: 5
 Metrics/CyclomaticComplexity:
   Max: 9
 
@@ -67,7 +67,7 @@ RSpec/FilePath:
     - 'spec/unit/solr_doc_builder_spec.rb'
     - 'spec/unit/solr_doc_hash_spec.rb'
 
-# Offense count: 356
+# Offense count: 359
 RSpec/InstanceVariable:
   Exclude:
     - 'spec/unit/gdor_mods_fields_spec.rb'

data/Gemfile

@@ -2,11 +2,3 @@ source 'https://rubygems.org'
 
 # Specify your gem's dependencies in spotlight-dor-resources.gemspec
 gemspec
-
-group :deployment do
-  gem 'capistrano', '~> 3.2'
-  gem 'capistrano-bundler'
-  gem 'capistrano-rvm' # gdor-indexer used to need jruby for merged records
-  gem 'lyberteam-capistrano-devel'
-  gem 'rainbow' # for color output
-end

data/gdor-indexer.gemspec

@@ -40,4 +40,5 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency 'vcr'
   spec.add_development_dependency 'jettywrapper'
   spec.add_development_dependency 'webmock'
+  spec.add_development_dependency 'pry-byebug'
 end

data/lib/gdor/indexer.rb

@@ -17,7 +17,7 @@ module GDor
   class Indexer
     include Hooks
 
-    define_hooks :before_index, :before_merge
+    define_hooks :before_index
 
     # local files
     require 'gdor/indexer/version'
@@ -53,6 +53,15 @@ module GDor
     end
 
     def logger
+      config_level =
+        case config.log_level
+        when 'debug' then Logger::DEBUG
+        when 'info' then Logger::INFO
+        when 'warn' then Logger::WARN
+        when 'error' then Logger::ERROR
+        when 'fatal' then Logger::FATAL
+        end
+      harvestdor.logger.level = config_level ? config_level : Logger::INFO
       harvestdor.logger
     end
 
@@ -160,7 +169,7 @@ module GDor
         building_facet: 'Stanford Digital Repository' # INDEX-53 add building_facet = Stanford Digital Repository here for collection
       )
 
-      logger.info "Indexing collection object #{resource.druid} (unmerged)"
+      logger.info "Indexing collection object #{resource.druid}"
       doc_hash = coll_sdb.doc_hash
       doc_hash.combine fields_to_add
       validation_messages = doc_hash.validate_collection(config)
@@ -267,7 +276,7 @@ module GDor
 
       if @druids_failed_to_ix.size > 0
         body += "\n"
-        body += "records that may have failed to index (merged recs as druids, not ckeys): \n"
+        body += "records that may have failed to index: \n"
         body += @druids_failed_to_ix.join("\n") + "\n"
       end
 

data/lib/gdor/indexer/version.rb

@@ -1,5 +1,5 @@
 module GDor
   class Indexer
-    VERSION = '0.1.0'
+    VERSION = '0.2.0'
   end
 end

data/spec/config/walters_integration_spec.yml

@@ -42,3 +42,7 @@ solr:
   read_timeout: 60
   open_timeout: 60
   max_retries: 10
+
+# the severity level of messages to be logged.  Valid values are debug, info, warn, error, fatal
+# default: debug
+log_level: info

data/spec/unit/gdor_mods_fields_spec.rb

@@ -7,15 +7,16 @@ describe GDor::Indexer::ModsFields do
     @mods_xml = "<mods #{@ns_decl}><note>gdor_mods_fields testing</note></mods>"
   end
 
-  let :logger do
-    Logger.new StringIO.new
-  end
-
   def sdb_for_mods(m)
     resource = Harvestdor::Indexer::Resource.new(double, @fake_druid)
     allow(resource).to receive(:public_xml).and_return(nil)
     allow(resource).to receive(:mods).and_return(Nokogiri::XML(m))
-    GDor::Indexer::SolrDocBuilder.new(resource, logger)
+    i = Harvestdor::Indexer.new
+    i.logger.level = Logger::WARN
+    allow(resource).to receive(:indexer).and_return(i)
+    lgr = Logger.new(StringIO.new)
+    lgr.level = Logger::WARN
+    GDor::Indexer::SolrDocBuilder.new(resource, lgr)
   end
 
   context 'doc_hash_from_mods' do

data/spec/unit/indexer_spec.rb

@@ -28,6 +28,9 @@ describe GDor::Indexer do
     allow(r).to receive(:public_xml?).and_return true
     allow(r).to receive(:content_metadata).and_return nil
     allow(r).to receive(:collection?).and_return false
+    i = Harvestdor::Indexer.new
+    i.logger.level = Logger::WARN
+    allow(r).to receive(:indexer).and_return i
     r
   end
 
@@ -40,6 +43,9 @@ describe GDor::Indexer do
     allow(r).to receive(:content_metadata).and_return nil
     allow(r).to receive(:identity_md_obj_label).and_return ''
     allow(r).to receive(:collection?).and_return true
+    i = Harvestdor::Indexer.new
+    i.logger.level = Logger::WARN
+    allow(r).to receive(:indexer).and_return i
     r
   end
 
@@ -48,6 +54,19 @@ describe GDor::Indexer do
       @indexer.logger.info('walters_integration_spec logging test message')
       expect(File).to exist(File.join(@yaml['harvestdor']['log_dir'], @yaml['harvestdor']['log_name']))
     end
+    it 'logger level defaults to INFO' do
+      expect(@indexer.logger.level).to eq Logger::INFO
+    end
+    it 'logger level can be specified in config field' do
+      indexer = described_class.new(@config_yml_path) do |config|
+        config.log_level = 'debug'
+      end
+      expect(indexer.logger.level).to eq Logger::DEBUG
+      indexer = described_class.new(@config_yml_path) do |config|
+        config.log_level = 'warn'
+      end
+      expect(indexer.logger.level).to eq Logger::WARN
+    end
   end
 
   describe '#harvest_and_index' do
@@ -74,7 +93,9 @@ describe GDor::Indexer do
         end
 
         def logger
-          Logger.new(STDERR)
+          lgr = Logger.new(StringIO.new)
+          lgr.level = Logger::WARN
+          lgr
         end
       end.new(collection, resource))
 
@@ -115,120 +136,116 @@ describe GDor::Indexer do
   end
 
   context '#item_solr_document' do
-    context 'unmerged' do
-      it 'calls Harvestdor::Indexer.solr_add' do
-        doc_hash = @indexer.item_solr_document(resource)
-        expect(doc_hash).to include id: @fake_druid
-      end
-      it 'calls validate_item' do
-        expect_any_instance_of(GDor::Indexer::SolrDocHash).to receive(:validate_item).and_return([])
-        @indexer.item_solr_document resource
-      end
-      it 'calls GDor::Indexer::SolrDocBuilder.validate_mods' do
-        allow_any_instance_of(GDor::Indexer::SolrDocHash).to receive(:validate_item).and_return([])
-        expect_any_instance_of(GDor::Indexer::SolrDocHash).to receive(:validate_mods).and_return([])
-        @indexer.item_solr_document resource
-      end
-      it 'calls add_coll_info' do
-        expect(@indexer).to receive(:add_coll_info)
-        @indexer.item_solr_document resource
-      end
-      it 'has fields populated from the collection record' do
-        sdb = double
-        allow(sdb).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new)
-        allow(sdb).to receive(:display_type)
-        allow(sdb).to receive(:file_ids)
-        allow(sdb.doc_hash).to receive(:validate_mods).and_return([])
-        allow(GDor::Indexer::SolrDocBuilder).to receive(:new).and_return(sdb)
-        allow(resource).to receive(:collections).and_return([double(druid: 'foo', bare_druid: 'foo', identity_md_obj_label: 'bar')])
-        doc_hash = @indexer.item_solr_document resource
-        expect(doc_hash).to include druid: @fake_druid, collection: ['foo'], collection_with_title: ['foo-|-bar']
-      end
-      it 'has fields populated from the MODS' do
-        title = 'fake title in mods'
-        ng_mods = Nokogiri::XML("<mods #{@ns_decl}><titleInfo><title>#{title}</title></titleInfo></mods>")
-        allow(resource).to receive(:mods).and_return(ng_mods)
-        doc_hash = @indexer.item_solr_document resource
-        expect(doc_hash).to include id: @fake_druid, title_display: title
-      end
-      it 'populates url_fulltext field with purl page url' do
-        doc_hash = @indexer.item_solr_document resource
-        expect(doc_hash).to include id: @fake_druid, url_fulltext: "#{@yaml['harvestdor']['purl']}/#{@fake_druid}"
-      end
-      it 'populates druid and access_facet fields' do
-        doc_hash = @indexer.item_solr_document resource
-        expect(doc_hash).to include id: @fake_druid, druid: @fake_druid, access_facet: 'Online'
-      end
-      it 'populates display_type field by calling display_type method' do
-        expect_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:display_type).and_return('foo')
-        doc_hash = @indexer.item_solr_document resource
-        expect(doc_hash).to include id: @fake_druid, display_type: 'foo'
-      end
-      it 'populates file_id field by calling file_ids method' do
-        expect_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:file_ids).at_least(1).times.and_return(['foo'])
-        doc_hash = @indexer.item_solr_document resource
-        expect(doc_hash).to include id: @fake_druid, file_id: ['foo']
-      end
-      it 'populates building_facet field with Stanford Digital Repository' do
-        doc_hash = @indexer.item_solr_document resource
-        expect(doc_hash).to include id: @fake_druid, building_facet: 'Stanford Digital Repository'
-      end
-    end # unmerged item
+    it 'calls Harvestdor::Indexer.solr_add' do
+      doc_hash = @indexer.item_solr_document(resource)
+      expect(doc_hash).to include id: @fake_druid
+    end
+    it 'calls validate_item' do
+      expect_any_instance_of(GDor::Indexer::SolrDocHash).to receive(:validate_item).and_return([])
+      @indexer.item_solr_document resource
+    end
+    it 'calls GDor::Indexer::SolrDocBuilder.validate_mods' do
+      allow_any_instance_of(GDor::Indexer::SolrDocHash).to receive(:validate_item).and_return([])
+      expect_any_instance_of(GDor::Indexer::SolrDocHash).to receive(:validate_mods).and_return([])
+      @indexer.item_solr_document resource
+    end
+    it 'calls add_coll_info' do
+      expect(@indexer).to receive(:add_coll_info)
+      @indexer.item_solr_document resource
+    end
+    it 'has fields populated from the collection record' do
+      sdb = double
+      allow(sdb).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new)
+      allow(sdb).to receive(:display_type)
+      allow(sdb).to receive(:file_ids)
+      allow(sdb.doc_hash).to receive(:validate_mods).and_return([])
+      allow(GDor::Indexer::SolrDocBuilder).to receive(:new).and_return(sdb)
+      allow(resource).to receive(:collections).and_return([double(druid: 'foo', bare_druid: 'foo', identity_md_obj_label: 'bar')])
+      doc_hash = @indexer.item_solr_document resource
+      expect(doc_hash).to include druid: @fake_druid, collection: ['foo'], collection_with_title: ['foo-|-bar']
+    end
+    it 'has fields populated from the MODS' do
+      title = 'fake title in mods'
+      ng_mods = Nokogiri::XML("<mods #{@ns_decl}><titleInfo><title>#{title}</title></titleInfo></mods>")
+      allow(resource).to receive(:mods).and_return(ng_mods)
+      doc_hash = @indexer.item_solr_document resource
+      expect(doc_hash).to include id: @fake_druid, title_display: title
+    end
+    it 'populates url_fulltext field with purl page url' do
+      doc_hash = @indexer.item_solr_document resource
+      expect(doc_hash).to include id: @fake_druid, url_fulltext: "#{@yaml['harvestdor']['purl']}/#{@fake_druid}"
+    end
+    it 'populates druid and access_facet fields' do
+      doc_hash = @indexer.item_solr_document resource
+      expect(doc_hash).to include id: @fake_druid, druid: @fake_druid, access_facet: 'Online'
+    end
+    it 'populates display_type field by calling display_type method' do
+      expect_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:display_type).and_return('foo')
+      doc_hash = @indexer.item_solr_document resource
+      expect(doc_hash).to include id: @fake_druid, display_type: 'foo'
+    end
+    it 'populates file_id field by calling file_ids method' do
+      expect_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:file_ids).at_least(1).times.and_return(['foo'])
+      doc_hash = @indexer.item_solr_document resource
+      expect(doc_hash).to include id: @fake_druid, file_id: ['foo']
+    end
+    it 'populates building_facet field with Stanford Digital Repository' do
+      doc_hash = @indexer.item_solr_document resource
+      expect(doc_hash).to include id: @fake_druid, building_facet: 'Stanford Digital Repository'
+    end
   end # item_solr_document
 
   context '#collection_solr_document' do
-    context 'unmerged' do
-      it 'calls validate_collection' do
-        doc_hash = GDor::Indexer::SolrDocHash.new
-        allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(doc_hash) # speed up the test
-        expect(doc_hash).to receive(:validate_collection).and_return([])
-        doc_hash = @indexer.collection_solr_document collection
-      end
-      it 'calls GDor::Indexer::SolrDocBuilder.validate_mods' do
-        doc_hash = GDor::Indexer::SolrDocHash.new
-        allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(doc_hash) # speed up the test
-        expect(doc_hash).to receive(:validate_mods).and_return([])
-        doc_hash = @indexer.collection_solr_document collection
-      end
-      it 'populates druid and access_facet fields' do
-        doc_hash = @indexer.collection_solr_document collection
-        expect(doc_hash).to include druid: @coll_druid_from_test_config, access_facet: 'Online'
-      end
-      it 'populates url_fulltext field with purl page url' do
+    it 'calls validate_collection' do
+      doc_hash = GDor::Indexer::SolrDocHash.new
+      allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(doc_hash) # speed up the test
+      expect(doc_hash).to receive(:validate_collection).and_return([])
+      doc_hash = @indexer.collection_solr_document collection
+    end
+    it 'calls GDor::Indexer::SolrDocBuilder.validate_mods' do
+      doc_hash = GDor::Indexer::SolrDocHash.new
+      allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(doc_hash) # speed up the test
+      expect(doc_hash).to receive(:validate_mods).and_return([])
+      doc_hash = @indexer.collection_solr_document collection
+    end
+    it 'populates druid and access_facet fields' do
+      doc_hash = @indexer.collection_solr_document collection
+      expect(doc_hash).to include druid: @coll_druid_from_test_config, access_facet: 'Online'
+    end
+    it 'populates url_fulltext field with purl page url' do
+      doc_hash = @indexer.collection_solr_document collection
+      expect(doc_hash).to include url_fulltext: "#{@yaml['harvestdor']['purl']}/#{@coll_druid_from_test_config}"
+    end
+    it "collection_type should be 'Digital Collection'" do
+      allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new) # speed up the test
+
+      doc_hash = @indexer.collection_solr_document collection
+      expect(doc_hash).to include collection_type: 'Digital Collection'
+    end
+    context 'add format_main_ssim Archive/Manuscript' do
+      it 'no other values' do
+        allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new)
+
         doc_hash = @indexer.collection_solr_document collection
-        expect(doc_hash).to include url_fulltext: "#{@yaml['harvestdor']['purl']}/#{@coll_druid_from_test_config}"
+        expect(doc_hash).to include format_main_ssim: 'Archive/Manuscript'
       end
-      it "collection_type should be 'Digital Collection'" do
-        allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new) # speed up the test
+      it 'other values present' do
+        allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new({ format_main_ssim: %w(Image Video) }))
 
         doc_hash = @indexer.collection_solr_document collection
-        expect(doc_hash).to include collection_type: 'Digital Collection'
+        expect(doc_hash).to include format_main_ssim: ['Image', 'Video', 'Archive/Manuscript']
       end
-      context 'add format_main_ssim Archive/Manuscript' do
-        it 'no other values' do
-          allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new)
+      it 'already has values Archive/Manuscript' do
+        allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new({ format_main_ssim: 'Archive/Manuscript' }))
 
-          doc_hash = @indexer.collection_solr_document collection
-          expect(doc_hash).to include format_main_ssim: 'Archive/Manuscript'
-        end
-        it 'other values present' do
-          allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new({ format_main_ssim: %w(Image Video) }))
-
-          doc_hash = @indexer.collection_solr_document collection
-          expect(doc_hash).to include format_main_ssim: ['Image', 'Video', 'Archive/Manuscript']
-        end
-        it 'already has values Archive/Manuscript' do
-          allow_any_instance_of(GDor::Indexer::SolrDocBuilder).to receive(:doc_hash).and_return(GDor::Indexer::SolrDocHash.new({ format_main_ssim: 'Archive/Manuscript' }))
-
-          doc_hash = @indexer.collection_solr_document collection
-          expect(doc_hash).to include format_main_ssim: ['Archive/Manuscript']
-        end
-      end
-      it 'populates building_facet field with Stanford Digital Repository' do
         doc_hash = @indexer.collection_solr_document collection
-        expect(doc_hash).to include building_facet: 'Stanford Digital Repository'
+        expect(doc_hash).to include format_main_ssim: ['Archive/Manuscript']
       end
-    end # unmerged collection
+    end
+    it 'populates building_facet field with Stanford Digital Repository' do
+      doc_hash = @indexer.collection_solr_document collection
+      expect(doc_hash).to include building_facet: 'Stanford Digital Repository'
+    end
   end #  index_coll_obj_per_config
 
   context '#add_coll_info and supporting methods' do
@@ -273,7 +290,6 @@ describe GDor::Indexer do
         @indexer.add_coll_info(doc_hash, [double(druid: coll_druid1, bare_druid: coll_druid1, public_xml: @ng_pub_xml, identity_md_obj_label: 'foo'), double(druid: coll_druid2, bare_druid: coll_druid2, public_xml: @ng_pub_xml, identity_md_obj_label: 'bar')])
         expect(doc_hash[:collection_with_title]).to match_array ["#{coll_druid1}-|-foo", "#{coll_druid2}-|-bar"]
       end
-      # other tests show it uses druid when coll rec isn't merged
     end
 
     context '#coll_display_types_from_items' do
@@ -301,14 +317,14 @@ describe GDor::Indexer do
 
   context '#num_found_in_solr' do
     before :each do
-      @unmerged_collection_response = { 'response' => { 'numFound' => '1', 'docs' => [{ 'id' => 'dm212rn7381', 'url_fulltext' => ['https://purl.stanford.edu/dm212rn7381'] }] } }
+      @collection_response = { 'response' => { 'numFound' => '1', 'docs' => [{ 'id' => 'dm212rn7381', 'url_fulltext' => ['https://purl.stanford.edu/dm212rn7381'] }] } }
       @item_response = { 'response' => { 'numFound' => '265', 'docs' => [{ 'id' => 'dm212rn7381' }] } }
     end
 
     it 'counts the items and the collection object in the solr index after indexing' do
       allow(@indexer.solr_client.client).to receive(:get) do |_wt, params|
         if params[:params][:fq].include?('id:"dm212rn7381"')
-          @unmerged_collection_response
+          @collection_response
         else
           @item_response
         end
@@ -340,7 +356,7 @@ describe GDor::Indexer do
 
     it 'email body includes failed to index druids' do
       @indexer.instance_variable_set(:@druids_failed_to_ix, %w(a b))
-      expect(subject).to match /records that may have failed to index \(merged recs as druids, not ckeys\): \na\nb\n\n/
+      expect(subject).to match /records that may have failed to index: \na\nb\n\n/
     end
 
     it 'email body include validation messages' do

data/spec/unit/public_xml_fields_spec.rb

@@ -8,9 +8,7 @@ describe GDor::Indexer::PublicXmlFields do
     @empty_pub_xml = "<publicObject id='druid:#{@fake_druid}'></publicObject>"
   end
 
-  let :logger do
-    Logger.new(StringIO.new)
-  end
+  let(:logger) { Logger.new(StringIO.new) }
 
   def sdb_for_pub_xml(m)
     resource = Harvestdor::Indexer::Resource.new(double, @fake_druid)

data/spec/unit/solr_doc_builder_spec.rb

@@ -9,13 +9,18 @@ describe GDor::Indexer::SolrDocBuilder do
   end
 
   let :logger do
-    Logger.new(StringIO.new)
+    lgr = Logger.new(StringIO.new)
+    lgr.level = Logger::WARN
+    lgr
   end
 
   def sdb_for_data(mods, pub_xml)
     resource = Harvestdor::Indexer::Resource.new(double, @fake_druid)
     allow(resource).to receive(:mods).and_return(Nokogiri::XML(mods))
     allow(resource).to receive(:public_xml).and_return(Nokogiri::XML(pub_xml))
+    i = Harvestdor::Indexer.new
+    i.logger.level = Logger::WARN
+    allow(resource).to receive(:indexer).and_return i
     GDor::Indexer::SolrDocBuilder.new(resource, logger)
   end
 
@@ -34,7 +39,7 @@ describe GDor::Indexer::SolrDocBuilder do
     before(:each) do
       @doc_hash = doc_hash
     end
-    it 'id field should be set to druid for non-merged record' do
+    it 'id field should be set to druid' do
       expect(@doc_hash[:id]).to eq(@fake_druid)
     end
     it 'does not have the gdor fields set in indexer.rb' do
@@ -44,7 +49,7 @@ describe GDor::Indexer::SolrDocBuilder do
       expect(@doc_hash).to_not have_key(:display_type)
       expect(@doc_hash).to_not have_key(:file_id)
     end
-    it 'has the full MODS in the modsxml field for non-merged record' do
+    it 'has the full MODS in the modsxml field' do
       expect(@doc_hash[:modsxml]).to be_equivalent_to @mods_xml
     end
   end # doc hash

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: gdor-indexer
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
 platform: ruby
 authors:
 - Naomi Dushay
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-10-26 00:00:00.000000000 Z
+date: 2015-10-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: harvestdor-indexer
@@ -334,6 +334,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: pry-byebug
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description: 
 email:
 - cabeer@stanford.edu
@@ -347,18 +361,12 @@ files:
 - ".rubocop.yml"
 - ".rubocop_todo.yml"
 - ".yardopts"
-- Capfile
 - Gemfile
 - LICENSE.txt
 - README.md
 - Rakefile
 - VERSION
 - bin/indexer
-- config/deploy.rb
-- config/deploy/dev.rb
-- config/deploy/fetcher.rb
-- config/deploy/prod.rb
-- config/deploy/stage.rb
 - gdor-indexer.gemspec
 - lib/gdor/indexer.rb
 - lib/gdor/indexer/mods_fields.rb

data/Capfile

@@ -1,26 +0,0 @@
-# Load DSL and Setup Up Stages
-require 'capistrano/setup'
-
-# Includes default deployment tasks
-require 'capistrano/deploy'
-
-# Includes tasks from other gems included in your Gemfile
-#
-# For documentation on these, see for example:
-#
-# https://github.com/capistrano/rvm
-# https://github.com/capistrano/rbenv
-# https://github.com/capistrano/chruby
-# https://github.com/capistrano/bundler
-# https://github.com/capistrano/rails
-#
-# require 'capistrano/rbenv'
-# require 'capistrano/chruby'
-require 'capistrano/bundler'
-# require 'capistrano/rails'
-require 'capistrano/rvm' # gdor-indexer needs jruby until merge-manager
-
-require 'dlss/capistrano'
-
-# Loads custom tasks from `lib/capistrano/tasks' if you have any defined.
-Dir.glob('lib/capistrano/tasks/*.cap').each { |r| import r }

data/config/deploy.rb

@@ -1,31 +0,0 @@
-set :application, 'gdor-indexer'
-set :repo_url, 'https://github.com/sul-dlss/gdor-indexer.git'
-
-# Default branch is :master
-# ask :branch, proc { `git rev-parse --abbrev-ref HEAD`.chomp }
-
-# gdor-indexer needs jruby until merge-manager
-set :rvm_ruby_version, 'jruby-1.7.10'
-
-set :user, 'lyberadmin'
-set :deploy_to, "/home/#{fetch(:user)}/#{fetch(:application)}"
-
-set :linked_dirs, %w(logs config/collections tmp solrmarc-sw)
-set :linked_files, %w(.ruby-version config/solr.yml bin/index-prod-image.sh bin/index-prod-hydrus.sh config/dor-fetcher-client.yml)
-
-set :stages, %w(dev stage prod fetcher)
-
-# Default value for :log_level is :debug
-set :log_level, :info
-
-# Default value for :format is :pretty
-# set :format, :pretty
-
-# Default value for :pty is false
-# set :pty, true
-
-# Default value for default_env is {}
-# set :default_env, { path: "/opt/ruby/bin:$PATH" }
-
-# Default value for keep_releases is 5
-set :keep_releases, 10

data/config/deploy/dev.rb

@@ -1,41 +0,0 @@
-# Simple Role Syntax
-# ==================
-# Supports bulk-adding hosts to roles, the primary
-# server in each group is considered to be the first
-# unless any hosts have the primary property set.
-# Don't declare `role :all`, it's a meta role
-# role :app, %w{deploy@example.com}
-# role :web, %w{deploy@example.com}
-# role :db, %w{deploy@example.com}
-
-# Extended Server Syntax
-# ======================
-# This can be used to drop a more detailed server
-# definition into the server list. The second argument
-# something that quacks like a hash can be used to set
-# extended properties on the server.
-server 'harvestdor-dev.stanford.edu', user: 'lyberadmin', roles: %w(web app db)
-
-Capistrano::OneTimeKey.generate_one_time_key!
-
-# you can set custom ssh options
-# it's possible to pass any option but you need to keep in mind that net/ssh understand limited list of options
-# you can see them in [net/ssh documentation](http://net-ssh.github.io/net-ssh/classes/Net/SSH.html#method-c-start)
-# set it globally
-# set :ssh_options, {
-# keys: %w(/home/rlisowski/.ssh/id_rsa),
-# forward_agent: false,
-# auth_methods: %w(password)
-# }
-# and/or per server
-# server 'example.com',
-# user: 'user_name',
-# roles: %w{web app},
-# ssh_options: {
-# user: 'user_name', # overrides user setting above
-# keys: %w(/home/user_name/.ssh/id_rsa),
-# forward_agent: false,
-# auth_methods: %w(publickey password)
-# password: 'please use keys'
-# }
-# setting per server overrides global ssh_options

data/config/deploy/fetcher.rb

@@ -1,6 +0,0 @@
-# Temporary deployment target for DorFetcher work
-server 'harvestdor-dev.stanford.edu', user: 'lyberadmin', roles: %w(app)
-
-Capistrano::OneTimeKey.generate_one_time_key!
-
-set :deploy_to, "/home/#{fetch(:user)}/gdor-indexer-fetcher"

data/config/deploy/prod.rb

@@ -1,41 +0,0 @@
-# Simple Role Syntax
-# ==================
-# Supports bulk-adding hosts to roles, the primary
-# server in each group is considered to be the first
-# unless any hosts have the primary property set.
-# Don't declare `role :all`, it's a meta role
-# role :app, %w{deploy@example.com}
-# role :web, %w{deploy@example.com}
-# role :db, %w{deploy@example.com}
-
-# Extended Server Syntax
-# ======================
-# This can be used to drop a more detailed server
-# definition into the server list. The second argument
-# something that quacks like a hash can be used to set
-# extended properties on the server.
-server 'harvestdor-prod.stanford.edu', user: 'lyberadmin', roles: %w(web app db)
-
-Capistrano::OneTimeKey.generate_one_time_key!
-
-# you can set custom ssh options
-# it's possible to pass any option but you need to keep in mind that net/ssh understand limited list of options
-# you can see them in [net/ssh documentation](http://net-ssh.github.io/net-ssh/classes/Net/SSH.html#method-c-start)
-# set it globally
-# set :ssh_options, {
-# keys: %w(/home/rlisowski/.ssh/id_rsa),
-# forward_agent: false,
-# auth_methods: %w(password)
-# }
-# and/or per server
-# server 'example.com',
-# user: 'user_name',
-# roles: %w{web app},
-# ssh_options: {
-# user: 'user_name', # overrides user setting above
-# keys: %w(/home/user_name/.ssh/id_rsa),
-# forward_agent: false,
-# auth_methods: %w(publickey password)
-# password: 'please use keys'
-# }
-# setting per server overrides global ssh_options

data/config/deploy/stage.rb

@@ -1,41 +0,0 @@
-# Simple Role Syntax
-# ==================
-# Supports bulk-adding hosts to roles, the primary
-# server in each group is considered to be the first
-# unless any hosts have the primary property set.
-# Don't declare `role :all`, it's a meta role
-# role :app, %w{deploy@example.com}
-# role :web, %w{deploy@example.com}
-# role :db, %w{deploy@example.com}
-
-# Extended Server Syntax
-# ======================
-# This can be used to drop a more detailed server
-# definition into the server list. The second argument
-# something that quacks like a hash can be used to set
-# extended properties on the server.
-server 'harvestdor-stage.stanford.edu', user: 'lyberadmin', roles: %w(web app db)
-
-Capistrano::OneTimeKey.generate_one_time_key!
-
-# you can set custom ssh options
-# it's possible to pass any option but you need to keep in mind that net/ssh understand limited list of options
-# you can see them in [net/ssh documentation](http://net-ssh.github.io/net-ssh/classes/Net/SSH.html#method-c-start)
-# set it globally
-# set :ssh_options, {
-# keys: %w(/home/rlisowski/.ssh/id_rsa),
-# forward_agent: false,
-# auth_methods: %w(password)
-# }
-# and/or per server
-# server 'example.com',
-# user: 'user_name',
-# roles: %w{web app},
-# ssh_options: {
-# user: 'user_name', # overrides user setting above
-# keys: %w(/home/user_name/.ssh/id_rsa),
-# forward_agent: false,
-# auth_methods: %w(publickey password)
-# password: 'please use keys'
-# }
-# setting per server overrides global ssh_options