discovery-indexer 1.0.1 → 2.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 8e437973bb7eef26cf7064d979158f73eaf966ff
-  data.tar.gz: e3ae3969e054bc29968c7b3eae6a6265408fdf07
+  metadata.gz: 6cda10dff64c547a94d5623b8d4eb99adc38fe55
+  data.tar.gz: 03f8ab54ee1cb990d54260f1d883b92869cc5d82
 SHA512:
-  metadata.gz: 67ff874812cd6d183f98de76fb2e39eff6089bf58f9fb8ba9a45c06274f97d994adb05002fc1638049a1fb5419a87ee5f2e1d5eaca60d87e75a142054c20d779
-  data.tar.gz: c2002f4a1d5609c76b3b1d2ff0b6082efe61e8a170d9658d8e715d210de6deeaa46f7755b62bb5ad07b6b4eb39549da132a48f0c3ff84ce2d110ade84db08db3
+  metadata.gz: d0ad8a756ae0904c5a9d61426a93895cbf1af6d46d4f61b36ff1ba94cf4beae0b671dcbe7e2c6ff419a811c3e71462f09db20e504e4949ee1da1075f8a16c1e4
+  data.tar.gz: b1c95061edb64ce7d585f583a3c932d0fc4fd220c591152decdb908ce8b1e40e9d0915cfb13ab0eaa9ab26061038b177f355fb37242700628d4e098b9ea6edf1

data/lib/discovery-indexer/reader/purlxml_model.rb

@@ -69,6 +69,14 @@ module DiscoveryIndexer
       #  ["pc0065_b08_f10_i031.jp2","pc0065_b08_f10_i032.jp2"]
       attr_accessor :image_ids
 
+      # @!attribute [rw] sw_image_ids
+      # @return [Array] a list of the image ids and corresponding druids with %2F
+      #  separator from the file and externalFile tags in the content_metadata
+      #  specifically for Searchworks
+      # @example
+      #  ["xx888yy9999%2Fpc0065_b08_f10_i031.jp2","aa111bb2222%2Fpc0065_b08_f10_i032.jp2"]
+      attr_accessor :sw_image_ids
+
       # @!attribute [rw] catkey
       # @return [String] the catkey attribute in identity_metadata
       attr_accessor :catkey

data/lib/discovery-indexer/reader/purlxml_parser_strict.rb

@@ -31,6 +31,7 @@ module DiscoveryIndexer
         purlxml_model.release_tags_hash = parse_release_tags_hash
         purlxml_model.file_ids          = parse_file_ids
         purlxml_model.image_ids         = parse_image_ids
+        purlxml_model.sw_image_ids      = parse_sw_image_ids
         purlxml_model.catkey            = parse_catkey
         purlxml_model.barcode           = parse_barcode
         purlxml_model.label             = parse_label
@@ -142,14 +143,28 @@ module DiscoveryIndexer
         dct
       end
 
-      # the @id attribute of resource/file elements that match the image type, including extension
+      # the @id attribute of resource/file elements that match the image, page, or thumb resource type, including extension
       # @return [Array<String>] filenames
       def parse_image_ids
         content_md = parse_content_metadata
-        return nil if content_md.nil?
+        return [] if content_md.nil?
         content_md.xpath('//resource[@type="page" or @type="image" or @type="thumb"]/file[@mimetype="image/jp2"]/@id').map(&:to_s)
       end
 
+      # the druid and id attribute of resource/file and objectId and fileId of the
+      # resource/externalFile elements that match the image, page, or thumb resource type, including extension
+      # Also, prepends the corresponding druid and %2F specifically for Searchworks use
+      # @return [Array<String>] filenames
+      def parse_sw_image_ids
+        content_md = parse_content_metadata
+        return [] if content_md.nil?
+        content_md.xpath('//resource[@type="page" or @type="image" or @type="thumb"]').map do |node|
+          node.xpath('./file[@mimetype="image/jp2"]/@id').map{ |x| "#{@druid}%2F" + x } << node.xpath('./externalFile[@mimetype="image/jp2"]').map do |y|
+            "#{y.attributes['objectId'].text.split(':').last}" + "%2F" + "#{y.attributes['fileId']}"
+          end
+        end.flatten
+      end
+
       def parse_sourceid
         get_value(@purlxml_ng_doc.css('//identityMetadata/sourceId'))
       end
@@ -165,14 +180,9 @@ module DiscoveryIndexer
       # the @id attribute of resource/file elements, including extension
       # @return [Array<String>] filenames
       def parse_file_ids
-        ids = []
         content_md = parse_content_metadata
-        return nil if content_md.nil?
-        content_md.xpath('//resource/file/@id').each do |node|
-          ids << node.text unless node.text.empty?
-        end
-        return nil if ids.empty?
-        ids
+        return [] if content_md.nil?
+        content_md.xpath('//resource/file/@id').map { |x| x.text }.compact
       end
 
       # @return catkey value from the DOR identity_metadata, or nil if there is no catkey

data/lib/discovery-indexer/version.rb

@@ -1,3 +1,3 @@
 module DiscoveryIndexer
-  VERSION = '1.0.1'
+  VERSION = '2.0.0'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: discovery-indexer
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 2.0.0
 platform: ruby
 authors:
 - Ahmed AlSum
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-02-19 00:00:00.000000000 Z
+date: 2016-02-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri