iiif_print 1.0.0 → 1.1.0

data/lib/iiif_print/configuration.rb

@@ -1,4 +1,5 @@
 module IiifPrint
+  # rubocop:disable Metrics/ClassLength
   class Configuration
     attr_writer :after_create_fileset_handler
 
@@ -12,6 +13,14 @@ module IiifPrint
       end
     end
 
+    attr_writer :ancestory_identifier_function
+    # The function, with arity 1, that receives a work and returns it's identifier for the purposes
+    # of object ancestry.
+    # @return [Proc]
+    def ancestory_identifier_function
+      @ancestory_identifier_function ||= ->(work) { work.id }
+    end
+
     attr_writer :excluded_model_name_solr_field_values
     # By default, this uses an array of human readable types
     #   ex: ['Generic Work', 'Image']
@@ -21,6 +30,45 @@ module IiifPrint
       @excluded_model_name_solr_field_values = []
     end
 
+    def skip_splitting_pdf_files_that_end_with_these_texts=(values)
+      @skip_splitting_pdf_files_that_end_with_these_texts = Array.wrap(values).map(&:downcase)
+    end
+
+    ##
+    #   @return [Array<String>] the file suffixes (e.g. [".reader.pdf"]) that we will skip.  Per
+    #           the implementation of {.split_for_path_suffix?}, these values are cast to
+    #           downcase.
+    def skip_splitting_pdf_files_that_end_with_these_texts
+      @skip_splitting_pdf_files_that_end_with_these_texts || []
+    end
+
+    attr_writer :unique_child_title_generator_function
+
+    # The function, with keywords (though maybe you'll want to splat ignore a few), is responsible
+    # for generating the child work file title.  of object ancestry.
+    #
+    # The keyword parameters that will be passed to this function are:
+    #
+    # :original_pdf_path - The fully qualified pathname to the original PDF from which the images
+    #                      were split.
+    # :image_path - The fully qualified pathname for an image of the single page from the PDF.
+    # :parent_work - The object in which we're "attaching" the image.
+    # :page_number - The image is of the N-th page_number of the original PDF
+    # :page_padding - A helper number that indicates the number of significant digits of pages
+    #                 (e.g. 150 pages would have a padding of 3).
+    #
+    # @return [Proc]
+    # rubocop:disable Lint/UnusedBlockArgument
+    def unique_child_title_generator_function
+      @unique_child_title_generator_function ||= lambda { |original_pdf_path:, image_path:, parent_work:, page_number:, page_padding:|
+        identifier = parent_work.id
+        filename = File.basename(original_pdf_path)
+        page_suffix = "Page #{(page_number.to_i + 1).to_s.rjust(page_padding.to_i, '0')}"
+        "#{identifier} - #{filename} #{page_suffix}"
+      }
+    end
+    # rubocop:enable Lint/UnusedBlockArgument
+
     # This method wraps Hyrax's configuration so we can sniff out the correct method to use.  The
     # {Hyrax::Configuration#whitelisted_ingest_dirs} is deprecated in favor of
     # {Hyrax::Configuration#registered_ingest_dirs}.
@@ -44,7 +92,7 @@ module IiifPrint
 
     attr_writer :default_iiif_manifest_version
     def default_iiif_manifest_version
-      @default_iiif_manifest_version || 2
+      @default_iiif_manifest_version.presence || 2
     end
 
     attr_writer :metadata_fields
@@ -81,19 +129,108 @@ module IiifPrint
     end
     # rubocop:enable Metrics/MethodLength
 
+    attr_writer :additional_tesseract_options
+    ##
+    # The additional options to pass to the Tesseract configuration
+    #
+    # @see https://tesseract-ocr.github.io/tessdoc/Command-Line-Usage.html
+    # @return [String]
+    def additional_tesseract_options
+      @additional_tesseract_options || ""
+    end
+
+    attr_writer :uv_config_path
+    ##
+    # According to https://github.com/samvera/hyrax/wiki/Hyrax-Management-Guide#universal-viewer-config
+    # the name of the UV config file should be /uv/uv_config.json (with an _)
+    # However, in most applications, it is /uv/uv-config.json (with a -)
+    def uv_config_path
+      @uv_config_path || "/uv/uv-config.json"
+    end
+
+    attr_writer :uv_base_path
+    ##
+    # While we're at it, we're going to go ahead and make the base path configurable as well
+    def uv_base_path
+      @uv_base_path || "/uv/uv.html"
+    end
+
+    attr_writer :child_work_attributes_function
+    ##
+    # Here we allow for customization of the child work attributes
+    def child_work_attributes_function
+      @child_work_attributes_function ||= lambda do |parent_work:, admin_set_id:|
+        {
+          admin_set_id: admin_set_id.to_s,
+          creator: parent_work.creator.to_a,
+          rights_statement: parent_work.rights_statement.to_a,
+          visibility: parent_work.visibility.to_s,
+          is_child: true
+        }
+      end
+    end
+
     attr_writer :sort_iiif_manifest_canvases_by
+    ##
+    # Normally, the canvases are sorted by the `ordered_members` association.
+    # However, if you want it to be sorted by another property, you can set this
+    # configuration.  Change `nil` to something like `:title` or `:identifier`.
+    #
+    # Should you want to sort by the filename of the image, you
+    # set `nil` to `:label`.  This looks at the canvas label, which is typically set
+    # to the filename of the image.
     def sort_iiif_manifest_canvases_by
-      @sort_iiif_manifest_canvases_by || :title
+      @sort_iiif_manifest_canvases_by || nil
     end
 
-    attr_writer :additional_tessearct_options
+    attr_writer :ocr_coords_from_json_function
     ##
-    # The additional options to pass to the Tesseract configuration
+    # This is used to determine where to pull the OCR coordinates from.  By default, it will
+    # pull from the JSON file that is generated by the OCR engine.  However, if you have a
+    # different source, you can set this configuration.  Current implementation has access to
+    # the `file_set_id`` and the `document` [SolrDocument].
     #
-    # @see https://tesseract-ocr.github.io/tessdoc/Command-Line-Usage.html
-    # @return [String]
-    def additional_tessearct_options
-      @additional_tessearct_options || ""
+    # @see IiifPrint::BlacklightIiifSearch::AnnotationDecorator#fetch_and_parse_coords
+    def ocr_coords_from_json_function
+      @ocr_coords_from_json_function ||= lambda do |file_set_id:, **|
+        IiifPrint::Data::WorkDerivatives.data(from: file_set_id, of_type: 'json')
+      end
+    end
+
+    attr_writer :all_text_generator_function
+    ##
+    # This configuration determines where to pull the full text from.  By default, it will
+    # pull from the TXT file that is generated by the OCR engine.  However, if your
+    # application has its own implementation of generating the full text, then you can
+    # set your own configuration here.
+    def all_text_generator_function
+      @all_text_generator_function ||= lambda do |object:|
+        IiifPrint::Data::WorkDerivatives.data(from: object, of_type: 'txt')
+      end
+    end
+
+    attr_writer :iiif_metadata_field_presentation_order
+    ##
+    # This is the default sorter for the metadata.  It will sort by the order of the keys specificied.
+    # By default, this is turned off as it returns nil. If you want to turn it on, you can set this
+    # this to an array of symbols the properties on the work.
+    #
+    # @example [:title, :description, :date_created]
+    # @return [Array<Symbol>]
+    def iiif_metadata_field_presentation_order
+      @iiif_metadata_field_presentation_order || nil
+    end
+
+    def questioning_authority_fields=(fields)
+      @questioning_authority_fields = Array.wrap(fields).map(&:to_s)
+    end
+
+    ##
+    # This is used to explicitly set which fields should be rendered as a Questioning Authority in the UV.
+    # By default, we render `rights_statement` and `license` as QA fields.
+    def questioning_authority_fields
+      @questioning_authority_fields ||= ['rights_statement', 'license']
     end
   end
+  # rubocop:enable Metrics/ModuleLength
 end

data/lib/iiif_print/data/fileset_helper.rb

@@ -7,7 +7,7 @@ module IiifPrint
         # if context is itself a string, presume it is a file set id
         return @work if @work.is_a? String
         # if context is not a String, presume a work or fileset context:
-        fileset.nil? ? nil : fileset.id
+        fileset&.id
       end
 
       def first_fileset

data/lib/iiif_print/data/work_derivatives.rb

@@ -42,16 +42,16 @@ module IiifPrint
       #
       # @return [String]
       def self.data(from:, of_type:)
-        new(from).data(of_type)
+        new(work: from).data(of_type)
       end
 
       # alternate constructor spelling:
       def self.of(work, fileset = nil, parent = nil)
-        new(work, fileset, parent)
+        new(work: work, fileset: fileset, parent: parent)
       end
 
       # Adapt work and either specific or first fileset
-      def initialize(work, fileset = nil, parent = nil)
+      def initialize(work: nil, fileset: nil, parent: nil)
         # adapted context usually work, may be string id of FileSet
         @work = work
         @fileset = fileset.nil? ? first_fileset : fileset

data/lib/iiif_print/engine.rb

@@ -1,6 +1,7 @@
 require 'active_fedora'
 require 'hyrax'
 require 'blacklight_iiif_search'
+require 'derivative_rodeo'
 
 module IiifPrint
   # module constants:
@@ -12,6 +13,7 @@ module IiifPrint
 
     # rubocop:disable Metrics/BlockLength
     config.to_prepare do
+      require "iiif_print/jobs/create_relationships_job"
       # We don't have a hard requirement of Bullkrax but in our experience, lingering on earlier
       # versions can introduce bugs of both Bulkrax and some of the assumptions that we've resolved.
       # Very early versions of Bulkrax do not have VERSION defined
@@ -41,26 +43,16 @@ module IiifPrint
       Hyrax::Renderers::FacetedAttributeRenderer.prepend(Hyrax::Renderers::FacetedAttributeRendererDecorator)
       Hyrax::WorksControllerBehavior.prepend(IiifPrint::WorksControllerBehaviorDecorator)
       Hyrax::WorkShowPresenter.prepend(IiifPrint::WorkShowPresenterDecorator)
+      Hyrax::IiifHelper.prepend(IiifPrint::IiifHelperDecorator)
 
       IiifPrint::ChildIndexer.decorate_work_types!
       IiifPrint::FileSetIndexer.decorate(Hyrax::FileSetIndexer)
 
       ::BlacklightIiifSearch::IiifSearchResponse.prepend(IiifPrint::IiifSearchResponseDecorator)
       ::BlacklightIiifSearch::IiifSearchAnnotation.prepend(IiifPrint::BlacklightIiifSearch::AnnotationDecorator)
+      ::BlacklightIiifSearch::IiifSearch.prepend(IiifPrint::IiifSearchDecorator)
       Hyrax::Actors::FileSetActor.prepend(IiifPrint::Actors::FileSetActorDecorator)
-
-      # Extending the presenter to the base url which includes the protocol.
-      # We need the base url to render the facet links and normalize the interface.
-      Hyrax::IiifManifestPresenter.send(:attr_accessor, :base_url)
-      Hyrax::IiifManifestPresenter::DisplayImagePresenter.send(:attr_accessor, :base_url)
-      # Extending this class because there is an #ability= but not #ability and this definition
-      # mirrors the Hyrax::IiifManifestPresenter#ability.
-      module Hyrax::IiifManifestPresenter::DisplayImagePresenterDecorator
-        def ability
-          @ability ||= NullAbility.new
-        end
-      end
-      Hyrax::IiifManifestPresenter::DisplayImagePresenter.prepend(Hyrax::IiifManifestPresenter::DisplayImagePresenterDecorator)
+      Hyrax::Actors::CleanupFileSetsActor.prepend(IiifPrint::Actors::CleanupFileSetsActorDecorator)
 
       Hyrax.config do |config|
         config.callback.set(:after_create_fileset) do |file_set, user|
@@ -71,6 +63,8 @@ module IiifPrint
 
     config.after_initialize do
       IiifPrint::Solr::Document.decorate(SolrDocument)
+      Hyrax::IiifManifestPresenter::DisplayImagePresenter
+        .prepend(IiifPrint::IiifManifestPresenterBehavior::DisplayImagePresenterBehavior)
     end
     # rubocop:enable Metrics/BlockLength
   end

data/lib/iiif_print/errors.rb

@@ -6,4 +6,22 @@ module IiifPrint
   # Data transformation or read-error:
   class DataError < IiifPrintError
   end
+
+  class MissingFileError < IiifPrintError
+  end
+
+  class WorkNotConfiguredToSplitFileSetError < IiifPrintError
+    def initialize(file_set:, work:)
+      message = "Expected that we would be splitting #{file_set.class} ID=#{file_set&.id} #to_param=#{file_set&.to_param} " \
+                "for work #{work.class} ID=#{work&.id} #to_param=#{work&.to_param}.  " \
+                "However it was not configured for PDF splitting."
+      super(message)
+    end
+  end
+
+  class UnexpectedMimeTypeError < IiifPrintError
+    def initialize(file_set:, mime_type:)
+      super "Unexpected mime_type #{mime_type} for #{file_set.class} ID=#{file_set.id.inspect}"
+    end
+  end
 end

data/lib/iiif_print/homepage_search_builder.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+# Overrides Hyrax to add show_parents_only to processor chain
+module IiifPrint
+  class HomepageSearchBuilder < Hyrax::HomepageSearchBuilder
+    self.default_processor_chain += [:show_parents_only]
+
+    def show_parents_only(solr_parameters)
+      query = if blacklight_params["include_child_works"] == 'true'
+                ActiveFedora::SolrQueryBuilder.construct_query(is_child_bsi: 'true')
+              else
+                ActiveFedora::SolrQueryBuilder.construct_query(is_child_bsi: nil)
+              end
+      solr_parameters[:fq] += [query]
+    end
+  end
+end

data/lib/iiif_print/image_tool.rb

@@ -3,11 +3,10 @@ require 'tmpdir'
 
 module IiifPrint
   class ImageTool
-    attr_accessor :path, :ftype
+    attr_accessor :path
 
     def initialize(path)
       @path = path
-      @ftype = magic
       @metadata = nil
     end
 
@@ -60,7 +59,7 @@ module IiifPrint
     end
 
     def im_line_select(lines, key)
-      line = lines.find { |l| l.scrub.downcase.strip.start_with?(key) }
+      line = lines.find { |l| l.scrub.downcase.strip.start_with?(key.downcase) }
       # Given "key: value" line, return the value as String stripped of
       #   leading and trailing whitespace
       return line if line.nil?
@@ -75,20 +74,25 @@ module IiifPrint
 
     # @return [Array<String>] lines of output from imagemagick `identify`
     def im_identify
-      cmd = "identify -verbose #{path}"
+      cmd = "identify -format 'Geometry: %G\nDepth: %[bit-depth]\nColorspace: %[colorspace]\nAlpha: %A\nMIME type: %m\n' #{path}"
       `#{cmd}`.lines
     end
 
     def im_mime(lines)
       return 'application/pdf' if pdf? # workaround older imagemagick bug
-      im_line_select(lines, 'mime type')
+
+      format = im_line_select(lines, 'mime type')
+      return if format.blank?
+
+      # `identify -format` with the `%m` switch only gives the format, we are coercing it into an image mime type
+      Mime::Type.lookup_by_extension(format.downcase).to_s
     end
 
     def populate_im_color!(lines, result)
       bpc = im_line_select(lines, 'depth').split('-')[0].to_i # '1-bit' -> 1
       colorspace = im_line_select(lines, 'colorspace')
       color = colorspace == 'Gray' ? 'gray' : 'color'
-      has_alpha = !im_line_select(lines, 'Alpha').nil?
+      has_alpha = !im_line_select(lines, 'alpha') == 'Undefined'
       result[:num_components] = (color == 'gray' ? 1 : 3) + (has_alpha ? 1 : 0)
       result[:color] = bpc == 1 ? 'monochrome' : color
       result[:bits_per_component] = bpc
@@ -105,11 +109,11 @@ module IiifPrint
     end
 
     def magic
-      File.read(@path, 23, 0)
+      @magic ||= File.read(@path, 23, 0)
     end
 
     def jp2?
-      @ftype.end_with?('ftypjp2')
+      magic.end_with?('ftypjp2')
     end
 
     def pdf?

data/lib/iiif_print/jobs/child_works_from_pdf_job.rb

@@ -1,20 +1,38 @@
 module IiifPrint
   module Jobs
+    # @deprecated
     class ChildWorksFromPdfJob < IiifPrint::Jobs::ApplicationJob
+      ##
       # Break a pdf into individual pages
-      # @param parent_work
+      #
+      # @param candidate_for_parency [FileSet, Hydra::PCDM::Work]
       # @param pdf_paths: [<Array => String>] paths to pdfs
       # @param user: [User]
       # @param admin_set_id: [<String>]
-      # @param prior_pdfs: [<Integer>] count of pdfs already on parent work
-      def perform(parent_work, pdf_paths, user, admin_set_id, prior_pdfs)
-        @parent_work = parent_work
+      # rubocop:disable Metrics/MethodLength
+      def perform(candidate_for_parency, pdf_paths, user, admin_set_id, *)
+        ##
+        # We know that we have cases where parent_work is nil, this will definitely raise an
+        # exception; which is fine because we were going to do it later anyway.
+        @parent_work = if candidate_for_parency.work?
+                         pdf_file_set = nil
+                         candidate_for_parency
+                       else
+                         # We likely have a file set
+                         pdf_file_set = candidate_for_parency
+                         IiifPrint.parent_for(candidate_for_parency)
+                       end
         @child_admin_set_id = admin_set_id
         child_model = @parent_work.iiif_print_config.pdf_split_child_model
 
-        # handle each input pdf
-        pdf_paths.each_with_index do |path, pdf_idx|
-          split_pdf(path, pdf_idx, user, prior_pdfs, child_model)
+        # When working with remote files, we have put the PDF file into the correct path before submitting this job.
+        # However, there seem to be cases where we still don't have the file when we get here, so to be sure, we
+        # re-do the same command that was previously used to prepare the file path. If the file is already here, it
+        # simply returns the path, but if not it will copy the file there, giving us one more chance to have what we need.
+        pdf_paths = [Hyrax::WorkingDirectory.find_or_retrieve(pdf_file_set.files.first.id, pdf_file_set.id, pdf_paths.first)] if pdf_file_set
+        # handle each input pdf (when input is a file set, we will only have one).
+        pdf_paths.each do |original_pdf_path|
+          split_pdf(original_pdf_path, user, child_model, pdf_file_set)
         end
 
         # Link newly created child works to the parent
@@ -31,15 +49,25 @@ module IiifPrint
 
         # TODO: clean up image_files and pdf_paths
       end
+      # rubocop:enable Metrics/MethodLength
 
       private
 
-      def split_pdf(path, pdf_idx, user, prior_pdfs_count, child_model)
-        image_files = @parent_work.iiif_print_config.pdf_splitter_service.new(path).to_a
-        return if image_files.blank?
+      # rubocop:disable Metrics/ParameterLists
+      # rubocop:disable Metrics/MethodLength
+      def split_pdf(original_pdf_path, user, child_model, pdf_file_set)
+        image_files = @parent_work.iiif_print_config.pdf_splitter_service.call(original_pdf_path, file_set: pdf_file_set)
 
-        pdf_sequence = pdf_idx + prior_pdfs_count
-        prepare_import_data(pdf_sequence, image_files, user)
+        # give as much info as possible if we don't have image files to work with.
+        if image_files.blank?
+          raise "#{@parent_work.class} (ID=#{@parent_work.id} " /
+                "to_param:#{@parent_work.to_param}) " /
+                "original_pdf_path #{original_pdf_path.inspect} " /
+                "pdf_file_set #{pdf_file_set.inspect}"
+        end
+
+        @split_from_pdf_id = pdf_file_set&.id
+        prepare_import_data(original_pdf_path, image_files, user)
 
         # submit the job to create all the child works for one PDF
         # @param [User] user
@@ -56,30 +84,54 @@ module IiifPrint
                                      @child_work_titles,
                                      {},
                                      @uploaded_files,
-                                     attributes.merge!(model: child_model.to_s).with_indifferent_access,
+                                     attributes.merge!(model: child_model.to_s, split_from_pdf_id: @split_from_pdf_id).with_indifferent_access,
                                      operation)
       end
+      # rubocop:enable Metrics/MethodLength
+      # rubocop:enable Metrics/ParameterLists
 
-      def prepare_import_data(pdf_sequence, image_files, user)
+      # rubocop:disable Metrics/MethodLength
+      def prepare_import_data(original_pdf_path, image_files, user)
         @uploaded_files = []
         @child_work_titles = {}
-        image_files.each_with_index do |image_path, idx|
+        number_of_pages_in_pdf = image_files.size
+        image_files.each_with_index do |image_path, page_number|
           file_id = create_uploaded_file(user, image_path).to_s
-          file_title = set_title(@parent_work.title.first, pdf_sequence, idx)
+
+          child_title = IiifPrint.config.unique_child_title_generator_function.call(
+            original_pdf_path: original_pdf_path,
+            image_path: image_path,
+            parent_work: @parent_work,
+            page_number: page_number,
+            page_padding: number_of_digits(nbr: number_of_pages_in_pdf)
+          )
+
           @uploaded_files << file_id
-          @child_work_titles[file_id] = file_title
+          @child_work_titles[file_id] = child_title
           # save child work info to create the member relationships
-          PendingRelationship.create!(child_title: file_title,
+          PendingRelationship.create!(child_title: child_title,
                                       parent_id: @parent_work.id,
-                                      child_order: sort_order(pdf_sequence, idx))
+                                      child_order: child_title,
+                                      parent_model: @parent_work.class,
+                                      child_model: @parent_work.iiif_print_config.pdf_split_child_model,
+                                      file_id: @split_from_pdf_id)
+
+          begin
+            # Clean up the temporary image path.
+            FileUtils.rm_f(image_path) if File.exist?(image_path)
+          rescue
+            # If we can't delete, let's move on.  Maybe it was already cleaned-up.
+          end
         end
       end
+      # rubocop:enable Metrics/MethodLength
 
-      def sort_order(pdf_sequence, idx)
-        "#{pdf_sequence} #{idx}"
+      def number_of_digits(nbr:)
+        nbr.to_s.size
       end
 
       def create_uploaded_file(user, path)
+        # TODO: Could we create a remote path?
         uf = Hyrax::UploadedFile.new
         uf.user_id = user.id
         uf.file = CarrierWave::SanitizedFile.new(path)
@@ -87,20 +139,9 @@ module IiifPrint
         uf.id
       end
 
-      def set_title(title, pdf_sequence, idx)
-        pdf_index = "Pdf Nbr #{pdf_sequence + 1}"
-        page_number = "Page #{idx + 1}"
-        "#{title}: #{pdf_index}, #{page_number}"
-      end
-
       # TODO: what attributes do we need to fill in from the parent work? What about AllinsonFlex?
       def attributes
-        {
-          admin_set_id: @child_admin_set_id.to_s,
-          creator: @parent_work.creator.to_a,
-          rights_statement: @parent_work.rights_statement.to_a,
-          visibility: @parent_work.visibility.to_s
-        }
+        IiifPrint.config.child_work_attributes_function.call(parent_work: @parent_work, admin_set_id: @child_admin_set_id)
       end
     end
   end