solrizer 2.0.0.rc4 → 2.0.0.rc5

data/History.txt

@@ -1,12 +1,10 @@
 h2. 2.0.0
 HYDRA-827 DO NOT index terms by default
 HYDRA-863 Null pointer exception fixed.
-
-BUG: Replaced data_type method with type which was causing the :type option in OM terms not to work.  This makes certain tests
-     not pass under Ruby 1.8.7, specifically when using :type option with a proxied term in OM.
-
-Test fixes such as properly testing solr field suffixes when using the :type option in OM as well as using correct
-ISO.8601 dates in our fixtures
+BUG: Replaced data_type method with type which was causing the :type option in OM terms not to work.  This makes certain tests not pass under Ruby 1.8.7, specifically when using :type option with a proxied term in OM.
+BUG: Fixed TerminologyBasedSolrizer#solrize_node instance method to call class method correctly, and changed #solrize_term to call instance method #solrize_node so that it can be overridden by subclasses
+Test fixes such as properly testing solr field suffixes when using the :type option in OM as well as using correct ISO.8601 dates in our fixtures
+HYDRA-876 When solrizing OM documents, term_values is called instead of find_by_terms which ensures data types are deserialized correctly.  Dates are also converted to ISO 8601 format, both when intgrating with OM or when using solrized independently.
 
 h2. 1.2.2
 BUG: RSolr::Client doesn't have a delete method.  Changed to delete_by_id. (only affects solrizer shell script)  Thanks to mkorcy.

data/lib/solrizer/field_mapper.rb

@@ -330,7 +330,9 @@ module Solrizer
       id_field 'id'
       index_as :searchable, :default => true do |t|
         t.default :suffix => '_t'
-        t.date    :suffix => '_dt'
+        t.date :suffix => '_dt' do |value|
+          value.is_a?(Date) ? Time.parse(value.to_s).utc.iso8601 : Time.parse(value).utc.iso8601
+        end
         t.string  :suffix => '_t'
         t.text    :suffix => '_t'
         t.symbol  :suffix => '_s'

data/lib/solrizer/version.rb

@@ -1,3 +1,3 @@
 module Solrizer
-  VERSION = "2.0.0.rc4"
+  VERSION = "2.0.0.rc5"
 end

data/lib/solrizer/xml/terminology_based_solrizer.rb

@@ -14,32 +14,36 @@ module Solrizer::XML::TerminologyBasedSolrizer
     unless doc.class.terminology.nil?
       doc.class.terminology.terms.each_pair do |term_name,term|
         doc.solrize_term(term, solr_doc, field_mapper)
-        # self.solrize_by_term(accessor_name, accessor_info, :solr_doc=>solr_doc)
       end
     end
 
     return solr_doc
   end
   
-  # Populate a solr document with fields based on nodes in +xml+ corresponding to the 
-  # term identified by +term_pointer+ within +terminology+
+  # Populate a solr document with fields based on nodes in +xml+ 
+  # Values for a term are gathered by to +term_pointer+ using OM::XML::TermValueOperators.term_values 
+  # and are deserialized by OM according to :type, as determined in its terminology.
+  # The content of the actual field in solr is each +node+ of the +nodeset+ returned by OM,
+  # rendered to a string.
   # @param [OM::XML::Document] doc xml document to extract values from
   # @param [OM::XML::Term] term corresponding to desired xml values
   # @param [Hash] (optional) solr_doc (values hash) to populate
   def self.solrize_term(doc, term, solr_doc = Hash.new, field_mapper = nil, opts={})
-    terminology = doc.class.terminology
     parents = opts.fetch(:parents, [])
-
     term_pointer = parents+[term.name]
-    nodeset = doc.find_by_terms(*term_pointer)
+    nodeset = doc.term_values(*term_pointer)
     
-    nodeset.each do |node|
-      # create solr fields
+    nodeset.each do |n|
+      
+      # TODO: Solrizer::FieldMapper::Default is supposed to translate dates into full ISO 8601 formatted strings.
+      # However, there an integration issue with ActiveFedora using OM: it ignores the default field mapper given
+      # in this gem that does this. So, the following is a workaround until it is fixed.
+      node = n.is_a?(Date) ? Time.parse(n.to_s).utc.iso8601 : n.to_s      
       
-      doc.solrize_node(node, term_pointer, term, solr_doc, field_mapper)
+      doc.solrize_node(node.to_s, term_pointer, term, solr_doc, field_mapper)
       unless term.kind_of? OM::XML::NamedTermProxy
         term.children.each_pair do |child_term_name, child_term|
-          doc.solrize_term(child_term, solr_doc, field_mapper, opts={:parents=>parents+[{term.name=>nodeset.index(node)}]})
+          doc.solrize_term(child_term, solr_doc, field_mapper, opts={:parents=>parents+[{term.name=>nodeset.index(node.to_s)}]})
         end
       end
     end
@@ -55,16 +59,9 @@ module Solrizer::XML::TerminologyBasedSolrizer
   # @param [Term] term the term to be solrized
   # @param [Hash] (optional) solr_doc (values hash) to populate
   # @return [Hash] the solr doc
-  def self.solrize_node(node, doc, term_pointer, term, solr_doc = Hash.new, field_mapper = nil, opts = {})
+  def self.solrize_node(node_value, doc, term_pointer, term, solr_doc = Hash.new, field_mapper = nil, opts = {})
     return solr_doc unless term.index_as && !term.index_as.empty?
     field_mapper ||= self.default_field_mapper
-    terminology = doc.class.terminology
-    
-    if term.path.kind_of?(Hash) && term.path.has_key?(:attribute)
-      node_value = node.value
-    else
-      node_value = node.text
-    end
     
     generic_field_name_base = OM::XML::Terminology.term_generic_name(*term_pointer)
     

data/spec/units/field_mapper_spec.rb

@@ -201,7 +201,13 @@ describe Solrizer::FieldMapper do
     it "should apply mappings for searchable by default" do
       # Just sanity check a couple; copy & pasting all data types is silly
       @mapper.solr_names_and_values('foo', 'bar', :string, []).should == { 'foo_t' => ['bar'] }
-      @mapper.solr_names_and_values('foo', 'bar', :date, []).should == { 'foo_dt' => ['bar'] }
+      @mapper.solr_names_and_values('foo', "1", :integer, []).should == { 'foo_i' =>["1"] }
+    end
+
+    it "should support full ISO 8601 dates" do
+      @mapper.solr_names_and_values('foo', "2012-11-06",              :date, []).should == { 'foo_dt' =>["2012-11-06T05:00:00Z"] }
+      @mapper.solr_names_and_values('foo', "November 6th, 2012",      :date, []).should == { 'foo_dt' =>["2012-11-06T05:00:00Z"] }
+      @mapper.solr_names_and_values('foo', Date.parse("6 Nov. 2012"), :date, []).should == { 'foo_dt' =>["2012-11-06T05:00:00Z"] }
     end
     
     it "should support displayable, facetable, sortable, unstemmed" do

data/spec/units/xml_terminology_based_solrizer_spec.rb

@@ -55,7 +55,7 @@ describe Solrizer::XML::TerminologyBasedSolrizer do
       solr_doc["topic_tag_t"].sort.should == ["CONTROLLED TERM", "TOPIC 1", "TOPIC 2"]
       
       # These are a holdover from an old verison of OM
-      solr_doc['journal_0_issue_0_publication_date_dt'].should == ["2007-02-01"]
+      solr_doc['journal_0_issue_0_publication_date_dt'].should == ["2007-02-01T05:00:00Z"]
 
       
     end
@@ -89,14 +89,14 @@ describe Solrizer::XML::TerminologyBasedSolrizer do
       unless RUBY_VERSION.match("1.8.7")
         solr_doc = Hash.new
         result = @mods_article.solrize_term(Samples::ModsArticle.terminology.retrieve_term(:pub_date), solr_doc)
-        solr_doc["pub_date_dt"].should == ["2007-02-01"]
+        solr_doc["pub_date_dt"].should == ["2007-02-01T05:00:00Z"]
       end
     end
 
     it "should add fields based on type using ref" do
       solr_doc = Hash.new
       result = @mods_article.solrize_term(Samples::ModsArticle.terminology.retrieve_term(:issue_date), solr_doc)
-      solr_doc["issue_date_dt"].should == ["2007-02-15"]
+      solr_doc["issue_date_dt"].should == ["2007-02-15T05:00:00Z"]
     end
 
     it "shouldn't index terms where index_as is an empty array" do

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: solrizer
 version: !ruby/object:Gem::Version
-  version: 2.0.0.rc4
+  version: 2.0.0.rc5
   prerelease: 6
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-11-02 00:00:00.000000000 Z
+date: 2012-11-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri
@@ -264,4 +264,3 @@ test_files:
 - spec/units/field_name_mapper_spec.rb
 - spec/units/xml_extractor_spec.rb
 - spec/units/xml_terminology_based_solrizer_spec.rb
-has_rdoc: