active_triples-solrizer 0.3.0

data/solr/solrconfig.xml

@@ -0,0 +1,179 @@
+<?xml version="1.0" encoding="UTF-8" ?>
+<config>
+  <!-- NOTE: various comments and unused configuration possibilities have been purged
+     from this file.  Please refer to http://wiki.apache.org/solr/SolrConfigXml,
+     as well as the default solrconfig file included with Solr -->
+  
+  <abortOnConfigurationError>${solr.abortOnConfigurationError:true}</abortOnConfigurationError>
+  
+  <luceneMatchVersion>4.9</luceneMatchVersion>
+
+  <directoryFactory name="DirectoryFactory" class="${solr.directoryFactory:solr.StandardDirectoryFactory}"/>
+
+  <!-- solr lib dirs -->
+  <lib dir="../lib/contrib/analysis-extras/lib" />
+  <lib dir="../lib/contrib/analysis-extras/lucene-libs" />
+  
+  <dataDir>${solr.data.dir:}</dataDir>
+
+  <updateHandler class="solr.DirectUpdateHandler2">
+
+    <updateLog>
+      <str name="dir">${solr.ulog.dir:}</str>
+    </updateLog>
+
+    <autoCommit> 
+      <maxTime>${solr.autoCommit.maxTime:15000}</maxTime> 
+      <openSearcher>false</openSearcher> 
+    </autoCommit>
+
+    <autoSoftCommit> 
+      <maxTime>${solr.autoSoftCommit.maxTime:-1}</maxTime> 
+    </autoSoftCommit>
+
+  </updateHandler>
+
+  <requestHandler name="search" class="solr.SearchHandler" default="true">
+    <!-- default values for query parameters can be specified, these
+         will be overridden by parameters in the request
+      -->
+     <lst name="defaults">
+       <str name="defType">edismax</str>
+       <str name="echoParams">explicit</str>
+       <str name="q.alt">*:*</str>
+       <str name="mm">2&lt;-1 5&lt;-2 6&lt;90%</str>
+       <int name="qs">1</int>
+       <int name="ps">2</int>
+       <float name="tie">0.01</float>
+       <!-- this qf and pf are used by default, if not otherwise specified by
+            client. The default blacklight_config will use these for the
+            "keywords" search. See the author_qf/author_pf, title_qf, etc 
+            below, which the default blacklight_config will specify for
+            those searches. You may also be interested in:
+            http://wiki.apache.org/solr/LocalParams
+       -->
+        <str name="qf">
+          id
+          all_text_timv
+          active_fedora_model_ssi
+          object_type_si
+        </str>
+        <str name="pf">
+          all_text_timv^10
+        </str>
+
+       <str name="author_qf">
+       </str>
+       <str name="author_pf">
+       </str>
+       <str name="title_qf">
+       </str>
+       <str name="title_pf">
+       </str>
+       <str name="subject_qf">
+       </str>
+       <str name="subject_pf">
+       </str>
+       
+       <str name="fl">
+         *, 
+         score
+       </str>
+
+       <str name="facet">true</str>
+       <str name="facet.mincount">1</str>
+       <str name="facet.limit">10</str>
+       <str name="facet.field">active_fedora_model_ssi</str>
+       <str name="facet.field">object_type_si</str>
+       
+       <str name="spellcheck">true</str>
+       <str name="spellcheck.dictionary">default</str>
+       <str name="spellcheck.onlyMorePopular">true</str>
+       <str name="spellcheck.extendedResults">true</str>
+       <str name="spellcheck.collate">false</str>
+       <str name="spellcheck.count">5</str>
+
+     </lst>
+    <arr name="last-components">
+      <str>spellcheck</str>
+    </arr>
+  </requestHandler>
+
+  <requestHandler name="permissions" class="solr.SearchHandler" >
+    <lst name="defaults">
+      <str name="facet">off</str>
+      <str name="echoParams">all</str>
+      <str name="rows">1</str>
+      <str name="q">{!raw f=id v=$id}</str> <!-- use id=666 instead of q=id:666 -->
+      <str name="fl">
+        id,
+        access_ssim,
+        discover_access_group_ssim,discover_access_person_ssim,
+        read_access_group_ssim,read_access_person_ssim,
+        edit_access_group_ssim,edit_access_person_ssim,
+        depositor_ti,
+        embargo_release_date_dtsi
+        inheritable_access_ssim,
+        inheritable_discover_access_group_ssim,inheritable_discover_access_person_ssim,
+        inheritable_read_access_group_ssim,inheritable_read_access_person_ssim,
+        inheritable_edit_access_group_ssim,inheritable_edit_access_person_ssim,
+        inheritable_embargo_release_date_dtsi
+      </str>
+    </lst>
+  </requestHandler>
+  
+  <requestHandler name="standard" class="solr.SearchHandler">
+     <lst name="defaults">
+       <str name="echoParams">explicit</str>
+       <str name="defType">lucene</str>
+     </lst>
+  </requestHandler>
+
+  <!-- for requests to get a single document; use id=666 instead of q=id:666 -->
+  <requestHandler name="document" class="solr.SearchHandler" >
+    <lst name="defaults">
+      <str name="echoParams">all</str>
+      <str name="fl">*</str>
+      <str name="rows">1</str>
+      <str name="q">{!raw f=id v=$id}</str> <!-- use id=666 instead of q=id:666 -->
+    </lst>
+  </requestHandler>
+
+  <searchComponent name="spellcheck" class="solr.SpellCheckComponent">
+    <str name="queryAnalyzerFieldType">textSpell</str>
+    <!-- Multiple "Spell Checkers" can be declared and used by this component
+      (e.g. for title_spell field)
+      -->
+    <lst name="spellchecker">
+      <str name="name">default</str>
+      <str name="field">spell</str>
+      <str name="spellcheckIndexDir">./spell</str>
+      <str name="buildOnOptimize">true</str>
+    </lst>
+  </searchComponent>
+  
+  <requestHandler name="/replication" class="solr.ReplicationHandler" startup="lazy" /> 
+  
+  <requestDispatcher handleSelect="true" >
+    <requestParsers enableRemoteStreaming="true" multipartUploadLimitInKB="2048" />
+  </requestDispatcher>
+  
+  <requestHandler name="/analysis/field" startup="lazy" class="solr.FieldAnalysisRequestHandler" />
+  <requestHandler name="/update" class="solr.UpdateRequestHandler"  />
+  <requestHandler name="/admin/" class="org.apache.solr.handler.admin.AdminHandlers" />
+  
+  <requestHandler name="/admin/ping" class="solr.PingRequestHandler">
+    <lst name="invariants">
+      <str name="q">solrpingquery</str>
+    </lst>
+    <lst name="defaults">
+      <str name="echoParams">all</str>
+    </lst>
+  </requestHandler>
+  
+  <!-- config for the admin interface --> 
+  <admin>
+    <defaultQuery>search</defaultQuery>
+  </admin>
+  
+</config>

data/spec/active_triples/solrizer/indexing_service_spec.rb

@@ -0,0 +1,79 @@
+require 'spec_helper'
+require 'support/dummy_resource.rb'
+
+describe ActiveTriples::Solrizer::SolrService do
+
+  describe "#generate_solr_document" do
+    include_context "shared dummy resource class"
+
+    context "when small number of properties" do
+      it "should produce solr doc with all indexed properties" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_short_all_values).generate_solr_document ).to eq expected_solr_doc_short_all_values
+      end
+
+      it "should produce solr doc with only properties with values indexed" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_short_partial_values).generate_solr_document ).to eq expected_solr_doc_short_partial_values
+      end
+    end
+
+    context "when all properties indexed" do
+      it "should produce solr doc with all indexed properties" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_indexed).generate_solr_document ).to eq expected_solr_doc_indexed
+      end
+    end
+
+    context "when all properties stored" do
+      it "should produce solr doc with all stored properties" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_stored).generate_solr_document ).to eq expected_solr_doc_stored
+      end
+    end
+
+    context "when all properties stored and indexed" do
+      it "should produce solr doc with all stored indexed properties" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_stored_indexed).generate_solr_document ).to eq expected_solr_doc_stored_indexed
+      end
+    end
+
+    context "when all properties indexed and multi-valued" do
+      it "should produce solr doc with all indexed properties" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_indexed_multi).generate_solr_document ).to eq expected_solr_doc_indexed_multi
+      end
+    end
+
+    context "when all properties stored and multi-valued" do
+      it "should produce solr doc with all stored properties" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_stored_multi).generate_solr_document ).to eq expected_solr_doc_stored_multi
+      end
+    end
+
+    context "when all properties stored and indexed and multi-valued" do
+      it "should produce solr doc with all stored indexed properties" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_stored_indexed_multi).generate_solr_document ).to eq expected_solr_doc_stored_indexed_multi
+      end
+    end
+
+    context "when all properties indexed and range" do
+      it "should produce solr doc with all indexed properties optimized for range" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_indexed_range).generate_solr_document ).to eq expected_solr_doc_indexed_range
+      end
+    end
+
+    context "when all properties indexed and sortable" do
+      it "should produce solr doc with all indexed properties set up for sorting" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_indexed_sort).generate_solr_document ).to eq expected_solr_doc_indexed_sort
+      end
+    end
+
+    context "when all properties indexed and vectored" do
+      it "should produce solr doc with all indexed properties and text fields vectored" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_indexed_vector).generate_solr_document ).to eq expected_solr_doc_indexed_vector
+      end
+    end
+
+    context "when all properties indexed and guessing the type" do
+      it "should produce solr doc with all indexed properties with appropriate types" do
+        expect( ActiveTriples::Solrizer::IndexingService.new(dr_indexed_guess).generate_solr_document ).to eq expected_solr_doc_indexed_guess
+      end
+    end
+  end
+end

data/spec/active_triples/solrizer/profile_indexing_service_spec.rb

@@ -0,0 +1,149 @@
+require 'spec_helper'
+require 'support/dummy_resource.rb'
+
+describe ActiveTriples::Solrizer::ProfileIndexingService do
+  include_context "shared dummy resource class"
+
+  describe "#export" do
+
+    context "when small number of properties" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_short_all_values).export ).to eq expected_object_profile_short_all_values
+      end
+
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_short_partial_values).export ).to eq expected_object_profile_short_partial_values
+      end
+    end
+
+    context "when all properties indexed" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_indexed).export ).to eq expected_object_profile_indexed
+      end
+    end
+
+    context "when all properties stored" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_stored).export ).to eq expected_object_profile_stored
+      end
+    end
+
+    context "when all properties stored and indexed" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_stored_indexed).export ).to eq expected_object_profile_stored_indexed
+      end
+    end
+
+    context "when all properties indexed and multi-valued" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_indexed_multi).export ).to eq expected_object_profile_indexed_multi
+      end
+    end
+
+    context "when all properties stored and multi-valued" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_stored_multi).export ).to eq expected_object_profile_stored_multi
+      end
+    end
+
+    context "when all properties stored and indexed and multi-valued" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_stored_indexed_multi).export ).to eq expected_object_profile_stored_indexed_multi
+      end
+    end
+
+    context "when all properties indexed and range" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_indexed_range).export ).to eq expected_object_profile_indexed_range
+      end
+    end
+
+    context "when all properties indexed and sortable" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_indexed_sort).export ).to eq expected_object_profile_indexed_sort
+      end
+    end
+
+    context "when all properties indexed and vectored" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_indexed_vector).export ).to eq expected_object_profile_indexed_vector
+      end
+    end
+
+    context "when all properties indexed and guessing the type" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_indexed_guess).export ).to eq expected_object_profile_indexed_guess
+      end
+    end
+
+    context "when a property is a resource" do
+      before do
+        class DummyResourceB < ActiveTriples::Resource
+          configure :type => RDF::URI('http://example.org/SomeClass')
+        end
+        class DummyResourceA < ActiveTriples::Resource
+          configure :type => RDF::URI('http://example.org/SomeClass')
+          property :resource_b, :predicate => RDF::URI("http://example.org/ontology/related_object")
+        end
+      end
+
+      it "should put the object id in place of the resource" do
+        dr_b = DummyResourceB.new('http://www.example.org/dr_b')
+        dr_a = DummyResourceA.new('http://www.example.org/dr_a')
+        dr_a.resource_b = dr_b
+
+        expected_object_profile =
+          '{"id":"http://www.example.org/dr_a",'\
+           '"resource_b":"http://www.example.org/dr_b"}'
+
+        expect( ActiveTriples::Solrizer::ProfileIndexingService.new(dr_a).export ).to eq expected_object_profile
+      end
+    end
+  end
+
+  describe "#import" do
+    context "when all properties have values" do
+      it "should fill a resource with values by deserializing the object profile stored with the solr doc" do
+        filled_resource = ActiveTriples::Solrizer::ProfileIndexingService.new(dr_short_all_values).import( expected_object_profile_short_all_values )
+        expect( filled_resource.id ).to               eq 'http://www.example.org/dr1'
+        expect( filled_resource.title ).to            eq ['Test Title']
+        expect( filled_resource.description_si ).to   eq ['Test text description stored and indexed.']
+        expect( filled_resource.borrower_uri_i ).to   eq ['http://example.org/i/b2']
+        expect( filled_resource.clip_number_simr ).to eq [7,8,9,10]
+        expect( filled_resource.price_s ).to          eq [789.01]
+        expect( filled_resource.bookEdition ).to      eq ['Ed. 2']
+      end
+    end
+
+    context "when some properties do not have values" do
+      it "should fill a resource with values by deserializing the object profile stored with the solr doc" do
+        filled_resource = ActiveTriples::Solrizer::ProfileIndexingService.new(dr_short_partial_values).import( expected_object_profile_short_partial_values )
+        expect( filled_resource.id ).to               eq 'http://www.example.org/dr1'
+        expect( filled_resource.title ).to            eq ['Test Title']
+        expect( filled_resource.description_si ).to   eq ['Test text description stored and indexed.']
+        expect( filled_resource.borrower_uri_i ).to   eq []
+        expect( filled_resource.clip_number_simr ).to eq []
+        expect( filled_resource.price_s ).to          eq [789.01]
+        expect( filled_resource.bookEdition ).to      eq ['Ed. 2']
+      end
+    end
+
+    context "when properties have multiple values and all types represented" do
+      it "should fill a resource with values by deserializing the object profile stored with the solr doc" do
+        filled_resource = ActiveTriples::Solrizer::ProfileIndexingService.new(dr_stored_indexed_multi).import( expected_object_profile_stored_indexed_multi )
+        expect( filled_resource.id ).to              eq 'http://www.example.org/dr_sim'
+        expect( filled_resource.title ).to           eq ['Test Title','Title 2']
+        expect( filled_resource.description ).to     eq ['Test text description stored and indexed and multi values.','Desc 2','Desc 3']
+        expect( filled_resource.borrower_uri ).to    eq ['http://example.org/b_sim','http://example.org/b_sim3']
+        expect( filled_resource.clip_number ).to     eq [3,4,5,6]
+        expect( filled_resource.answer_count ).to    eq [12345678901234567893,32345678901234567893]
+        expect( filled_resource.price ).to           eq [123.43,323.43]
+        expect( filled_resource.max_price ).to       eq [12345678901234567.83,32345678901234567.83]
+        expect( filled_resource.modified_time ).to   eq ['1995-12-31T23:59:53Z','2015-12-31T23:59:53Z']
+        expect( filled_resource.is_proprietary ).to  eq [false,true]
+        # expect( filled_resource.latitude       =
+        expect( filled_resource.bookEdition ).to     eq ['Ed. 3','Ed. 3a']
+      end
+    end
+  end
+end

data/spec/active_triples/solrizer/properties_indexing_service_spec.rb

@@ -0,0 +1,78 @@
+require 'spec_helper'
+require 'support/dummy_resource.rb'
+
+describe ActiveTriples::Solrizer::PropertiesIndexingService do
+  include_context "shared dummy resource class"
+
+  describe "#export" do
+    context "when small number of properties" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_short_all_values).export ).to eq expected_solr_properties_short_all_values
+      end
+
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_short_partial_values).export ).to eq expected_solr_properties_short_partial_values
+      end
+    end
+
+    context "when all properties indexed" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_indexed).export ).to eq expected_solr_properties_indexed
+      end
+    end
+
+    context "when all properties stored" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_stored).export ).to eq expected_solr_properties_stored
+      end
+    end
+
+    context "when all properties stored and indexed" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_stored_indexed).export ).to eq expected_solr_properties_stored_indexed
+      end
+    end
+
+    context "when all properties indexed and multi-valued" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_indexed_multi).export ).to eq expected_solr_properties_indexed_multi
+      end
+    end
+
+    context "when all properties stored and multi-valued" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_stored_multi).export ).to eq expected_solr_properties_stored_multi
+      end
+    end
+
+    context "when all properties stored and indexed and multi-valued" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_stored_indexed_multi).export ).to eq expected_solr_properties_stored_indexed_multi
+      end
+    end
+
+    context "when all properties indexed and range" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_indexed_range).export ).to eq expected_solr_properties_indexed_range
+      end
+    end
+
+    context "when all properties indexed and sortable" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_indexed_sort).export ).to eq expected_solr_properties_indexed_sort
+      end
+    end
+
+    context "when all properties indexed and vectored" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_indexed_vector).export ).to eq expected_solr_properties_indexed_vector
+      end
+    end
+
+    context "when all properties indexed and guessing the type" do
+      it "should produce object profile holding serialization of the resource" do
+        expect( ActiveTriples::Solrizer::PropertiesIndexingService.new(dr_indexed_guess).export ).to eq expected_solr_properties_indexed_guess
+      end
+    end
+  end
+end