curation_concerns-models 0.1.0 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 51998b654c32654d086cbeb388436dff6db4c186
-  data.tar.gz: 59caa819f6745297082809edb2cf38b50ca43c4e
+  metadata.gz: 4b866be7465155c67d0157b20159c75f53cc6d2c
+  data.tar.gz: def664a51f3bff81f3ca2396787883ae6b3142cc
 SHA512:
-  metadata.gz: 3c0af2087c9ec611684b8f6c12bcc62386d7c44837338f08659ca0a2bd6fcbff6538f9e25abc42479f38147553087fd84381deb91aedfaa4637b4a79b6ee7fa0
-  data.tar.gz: 5079fdbb1d714c8a9b8ef18124da0777e407a0ce877ac748a5bf90fde98617ba88eb6f30480f88435f8b51d2820dc4baef2e8765a6fdc8f34900689ee26b53f7
+  metadata.gz: 399bdc3d1ea0bb4722605fa486662e1e33b4c01872f0fe3fe47d7edd36583c0a744a4bfa0198f39e288f4943c69ac00ec712e5ca985cfde379770c510c052516
+  data.tar.gz: 1bc3ca8a712bb6d141d4c8411c959db37c66931b86b0c1e3a7a969d42d41402ffb59397ab7d5da1e1c4eec84228a09c50fd9dce5932d37344368193f5a03ea86

data/README.md

@@ -4,6 +4,8 @@ Models extracted from worthwhile-models and sufia-models.
 
 ## Dependencies
 
+Checkout the readme for [hydra-derivatives](https://github.com/projecthydra/hydra-derivatives#dependencies) for additional dependencies for generating derivatives.
+
 ### FITS 0.6.2
 
 To install FITS:

data/Rakefile

@@ -1 +1 @@
-require "bundler/gem_tasks"
+require 'bundler/gem_tasks'

data/app/actors/concerns/curation_concerns/manages_embargoes_actor.rb

@@ -9,11 +9,11 @@ module CurationConcerns
   #     include Worthwile::ManagesEmbargoesActor
   #
   #     def create
-  #       interpret_visibility && super && copy_visibility
+  #       interpret_visibility && super
   #     end
   #
   #     def update
-  #       interpret_visibility && super && copy_visibility
+  #       interpret_visibility && super
   #     end
   #  end
   #
@@ -22,17 +22,17 @@ module CurationConcerns
 
     # Interprets embargo & lease visibility if necessary
     # returns false if there are any errors
-    def interpret_visibility(attributes=self.attributes)
+    def interpret_visibility(attributes = self.attributes)
       should_continue = interpret_embargo_visibility(attributes) && interpret_lease_visibility(attributes)
       if attributes[:visibility]
         curation_concern.visibility = attributes[:visibility]
       end
-      return should_continue
+      should_continue
     end
 
     # If user has set visibility to embargo, interprets the relevant information and applies it
     # Returns false if there are any errors and sets an error on the curation_concern
-    def interpret_embargo_visibility(attributes=self.attributes)
+    def interpret_embargo_visibility(attributes = self.attributes)
       if attributes[:visibility] == Hydra::AccessControls::AccessRight::VISIBILITY_TEXT_VALUE_EMBARGO
         if !attributes[:embargo_release_date]
           curation_concern.errors.add(:visibility, 'When setting visibility to "embargo" you must also specify embargo release date.')
@@ -42,9 +42,8 @@ module CurationConcerns
           curation_concern.apply_embargo(attributes[:embargo_release_date], attributes.delete(:visibility_during_embargo),
                                          attributes.delete(:visibility_after_embargo))
           if curation_concern.embargo
-            curation_concern.embargo.save  # See https://github.com/projecthydra/hydra-head/issues/226
+            curation_concern.embargo.save # See https://github.com/projecthydra/hydra-head/issues/226
           end
-          @needs_to_copy_visibility = true
           should_continue = true
         end
       else
@@ -57,23 +56,22 @@ module CurationConcerns
       attributes.delete(:visibility_during_embargo)
       attributes.delete(:visibility_after_embargo)
 
-      return should_continue
+      should_continue
     end
 
     # If user has set visibility to lease, interprets the relevant information and applies it
     # Returns false if there are any errors and sets an error on the curation_concern
-    def interpret_lease_visibility(attributes=self.attributes)
+    def interpret_lease_visibility(attributes = self.attributes)
       if attributes[:visibility] == Hydra::AccessControls::AccessRight::VISIBILITY_TEXT_VALUE_LEASE
         if !attributes[:lease_expiration_date]
           curation_concern.errors.add(:visibility, 'When setting visibility to "lease" you must also specify lease expiration date.')
           should_continue = false
         else
           curation_concern.apply_lease(attributes[:lease_expiration_date], attributes.delete(:visibility_during_lease),
-                                         attributes.delete(:visibility_after_lease))
+                                       attributes.delete(:visibility_after_lease))
           if curation_concern.lease
-            curation_concern.lease.save  # See https://github.com/projecthydra/hydra-head/issues/226
+            curation_concern.lease.save # See https://github.com/projecthydra/hydra-head/issues/226
           end
-          @needs_to_copy_visibility = true
           attributes.delete(:visibility)
           should_continue = true
         end
@@ -87,13 +85,7 @@ module CurationConcerns
       attributes.delete(:visibility_during_lease)
       attributes.delete(:visibility_after_lease)
 
-      return should_continue
-    end
-
-
-    def copy_visibility
-      CurationConcerns.queue.push(VisibilityCopyWorker.new(curation_concern.id)) if @needs_to_copy_visibility
-      true
+      should_continue
     end
   end
 end

data/app/actors/curation_concerns/base_actor.rb

@@ -11,7 +11,7 @@ module CurationConcerns
       @user = user
       @attributes = input_attributes.dup.with_indifferent_access
       @visibility = attributes[:visibility]
-      @cloud_resources= attributes.delete(:cloud_resources.to_s)
+      @cloud_resources = attributes.delete(:cloud_resources.to_s)
     end
 
     attr_reader :visibility
@@ -32,59 +32,55 @@ module CurationConcerns
     end
 
     protected
-    def apply_creation_data_to_curation_concern
-      apply_depositor_metadata
-      apply_deposit_date
-    end
 
-    def apply_update_data_to_curation_concern
-      true
-    end
+      def apply_creation_data_to_curation_concern
+        apply_depositor_metadata
+        apply_deposit_date
+      end
 
-    def apply_depositor_metadata
-      curation_concern.apply_depositor_metadata(user.user_key)
-      curation_concern.edit_users += [user.user_key]
-    end
+      def apply_update_data_to_curation_concern
+        true
+      end
 
-    def apply_deposit_date
-      curation_concern.date_uploaded = Date.today
-    end
+      def apply_depositor_metadata
+        curation_concern.apply_depositor_metadata(user.user_key)
+        curation_concern.edit_users += [user.user_key]
+      end
 
-    def save
-      curation_concern.save
-    end
+      def apply_deposit_date
+        curation_concern.date_uploaded = CurationConcerns::TimeService.time_in_utc
+      end
 
-    def apply_save_data_to_curation_concern
-      attributes[:rights] = Array(attributes[:rights]) if attributes.key? :rights
-      remove_blank_attributes!
-      curation_concern.attributes = attributes.symbolize_keys
-      curation_concern.date_modified = Date.today
-    end
+      def save
+        curation_concern.save
+      end
 
-    def attach_file(generic_file, file_to_attach)
-      ActiveSupport::Deprecation.warn("removing #{self.class}#attach_file, use CurationConcern.attach_file instead")
-      CurationConcern.attach_file(generic_file, user, file_to_attach)
-    end
+      def apply_save_data_to_curation_concern
+        attributes[:rights] = Array(attributes[:rights]) if attributes.key? :rights
+        remove_blank_attributes!
+        curation_concern.attributes = attributes.symbolize_keys
+        curation_concern.date_modified = CurationConcerns::TimeService.time_in_utc
+      end
 
-    # If any attributes are blank remove them
-    # e.g.:
-    #   self.attributes = { 'title' => ['first', 'second', ''] }
-    #   remove_blank_attributes!
-    #   self.attributes
-    # => { 'title' => ['first', 'second'] }
-    def remove_blank_attributes!
-      multivalued_form_attributes.each_with_object(attributes) do |(k, v), h|
-        if v.instance_of? Array
-          h[k] = v.select(&:present?)
-        else
-          h[k] = v
+      # If any attributes are blank remove them
+      # e.g.:
+      #   self.attributes = { 'title' => ['first', 'second', ''] }
+      #   remove_blank_attributes!
+      #   self.attributes
+      # => { 'title' => ['first', 'second'] }
+      def remove_blank_attributes!
+        multivalued_form_attributes.each_with_object(attributes) do |(k, v), h|
+          if v.instance_of? Array
+            h[k] = v.select(&:present?)
+          else
+            h[k] = v
+          end
         end
       end
-    end
 
-    # Return the hash of attributes that are multivalued and not uploaded files
-    def multivalued_form_attributes
-      attributes.select {|_, v| v.respond_to?(:select) && !v.respond_to?(:read) }
-    end
+      # Return the hash of attributes that are multivalued and not uploaded files
+      def multivalued_form_attributes
+        attributes.select { |_, v| v.respond_to?(:select) && !v.respond_to?(:read) }
+      end
   end
 end

data/app/actors/curation_concerns/embargo_actor.rb

@@ -0,0 +1,19 @@
+module CurationConcerns
+  class EmbargoActor
+    attr_reader :work
+
+    # @param [Hydra::Works::Work] work
+    def initialize(work)
+      @work = work
+    end
+
+    # Update the visibility of the work to match the correct state of the embargo, then clear the embargo date, etc.
+    # Saves the embargo and the work
+    def destroy
+      work.embargo_visibility! # If the embargo has lapsed, update the current visibility.
+      work.deactivate_embargo!
+      work.embargo.save!
+      work.save!
+    end
+  end
+end

data/app/actors/curation_concerns/file_set_actor.rb

@@ -0,0 +1,200 @@
+module CurationConcerns
+  # Actions are decoupled from controller logic so that they may be called from a controller or a background job.
+  class FileSetActor
+    include CurationConcerns::ManagesEmbargoesActor
+
+    attr_reader :file_set, :user, :attributes, :curation_concern
+
+    def initialize(file_set, user)
+      # we're setting attributes and curation_concern to bridge the difference
+      # between CurationConcerns::FileSetActor and ManagesEmbargoesActor
+      @curation_concern = file_set
+      @file_set = file_set
+      @user = user
+    end
+
+    # Adds the appropriate metadata, visibility and relationships to file_set
+    #
+    # *Note*: In past versions of Sufia this method did not perform a save because it is mainly used in conjunction with
+    # create_content, which also performs a save.  However, due to the relationship between Hydra::PCDM objects,
+    # we have to save both the parent work and the file_set in order to record the "metadata" relationship
+    # between them.
+    # @param [String] upload_set_id id of the batch of files that the file was uploaded with
+    # @param [ActiveFedora::Base] work the parent work that will contain the file_set.
+    # @param [Hash] file_set specifying the visibility, lease and/or embargo of the file set.  If you don't provide at least one of visibility, embargo_release_date or lease_expiration_date, visibility will be copied from the parent.
+
+    def create_metadata(upload_set_id, work, file_set_params = {})
+      file_set.apply_depositor_metadata(user)
+      now = CurationConcerns::TimeService.time_in_utc
+      file_set.date_uploaded = now
+      file_set.date_modified = now
+      file_set.creator = [user.user_key]
+
+      if upload_set_id && file_set.respond_to?(:upload_set_id=)
+        UploadSet.create(id: upload_set_id) unless UploadSet.exists?(upload_set_id)
+        file_set.upload_set_id = upload_set_id
+      else
+        ActiveFedora::Base.logger.warn 'unable to find UploadSet to attach to'
+      end
+
+      if assign_visibility?(file_set_params)
+        interpret_visibility file_set_params
+      end
+      # TODO: Why do we need to check if work is nil? Shoudn't that raise an error?
+      attach_file_to_work(work, file_set, file_set_params) if work
+      yield(file_set) if block_given?
+    end
+
+    # Puts the uploaded content into a staging directory. Then kicks off a
+    # job to characterize and create derivatives with this on disk variant.
+    # Simultaneously moving a preservation copy to the repostiory.
+    # TODO: create a job to monitor this directory and prune old files that
+    # have made it to the repo
+    # @param [ActionDigest::HTTP::UploadedFile, Tempfile] file the file uploaded by the user.
+    def create_content(file)
+      file_set.label ||= file.original_filename
+      file_set.title = [file_set.label] if file_set.title.blank?
+      return false unless file_set.save
+
+      working_file = copy_file_to_working_directory(file, file_set.id)
+      IngestFileJob.perform_later(file_set.id, working_file, file.content_type, user.user_key)
+      make_derivative(file_set.id, working_file)
+      true
+    end
+
+    def revert_content(revision_id)
+      file_set.original_file.restore_version(revision_id)
+
+      return false unless file_set.save
+
+      CurationConcerns::VersioningService.create(file_set.original_file, user)
+
+      # Retrieve a copy of the orginal file from the repository
+      working_file = copy_repository_resource_to_working_directory(file_set)
+      make_derivative(file_set.id, working_file)
+
+      CurationConcerns.config.callback.run(:after_revert_content, file_set, user, revision_id)
+      true
+    end
+
+    def update_content(file)
+      working_file = copy_file_to_working_directory(file, file_set.id)
+      IngestFileJob.perform_later(file_set.id, working_file, file.content_type, user.user_key)
+      make_derivative(file_set.id, working_file)
+      CurationConcerns.config.callback.run(:after_update_content, file_set, user)
+      true
+    end
+
+    def update_metadata(attributes)
+      update_visibility(attributes)
+      # attributes.delete(:visibility) # Applying this attribute is handled by update_visibility
+      file_set.attributes = attributes
+      file_set.date_modified = CurationConcerns::TimeService.time_in_utc
+      save do
+        CurationConcerns.config.callback.run(:after_update_metadata, file_set, user)
+      end
+    end
+
+    def destroy
+      file_set.destroy
+      # TODO: need to mend the linked list of proxies (possibly wrap with a lock)
+      CurationConcerns.config.callback.run(:after_destroy, file_set.id, user)
+    end
+
+    private
+
+      def make_derivative(file_set_id, working_file)
+        CharacterizeJob.perform_later(file_set_id, working_file)
+      end
+
+      # @param [ActionDispatch::Http::UploadedFile] file
+      # @param [String] id the identifer
+      # @return [String] path of the working file
+      def copy_file_to_working_directory(file, id)
+        copy_stream_to_working_directory(id, file.original_filename, file)
+      end
+
+      # @param [FileSet] file_set the resource
+      # @return [String] path of the working file
+      def copy_repository_resource_to_working_directory(file_set)
+        file = file_set.original_file
+        copy_stream_to_working_directory(file_set.id, file.original_name, StringIO.new(file.content))
+      end
+
+      # @param [String] id the identifer
+      # @param [String] name the file name
+      # @param [#read] stream the stream to copy to the working directory
+      # @return [String] path of the working file
+      def copy_stream_to_working_directory(id, name, stream)
+        working_path = full_filename(id, name)
+        FileUtils.mkdir_p(File.dirname(working_path))
+        IO.copy_stream(stream, working_path)
+        working_path
+      end
+
+      def full_filename(id, original_name)
+        pair = id.scan(/..?/).first(4)
+        File.join(CurationConcerns.config.working_path, *pair, original_name)
+      end
+
+      # Takes an optional block and executes the block if the save was successful.
+      # returns false if the save was unsuccessful
+      def save
+        save_tries = 0
+        begin
+          return false unless file_set.save
+        rescue RSolr::Error::Http => error
+          ActiveFedora::Base.logger.warn "CurationConcerns::FileSetActor#save Caught RSOLR error #{error.inspect}"
+          save_tries += 1
+          # fail for good if the tries is greater than 3
+          raise error if save_tries >= 3
+          sleep 0.01
+          retry
+        end
+        yield if block_given?
+        true
+      end
+
+      # Adds a FileSet to the work using ore:Aggregations.
+      # Locks to ensure that only one process is operating on
+      # the list at a time.
+      def attach_file_to_work(work, file_set, file_set_params)
+        acquire_lock_for(work.id) do
+          # Ensure we have an up-to-date copy of the members association, so
+          # that we append to the end of the list.
+          work.reload unless work.new_record?
+          unless assign_visibility?(file_set_params)
+            copy_visibility(work, file_set)
+          end
+          work.ordered_members << file_set
+          # Save the work so the association between the work and the file_set is persisted (head_id)
+          work.save
+        end
+      end
+
+      def acquire_lock_for(lock_key, &block)
+        lock_manager.lock(lock_key, &block)
+      end
+
+      def lock_manager
+        @lock_manager ||= CurationConcerns::LockManager.new(
+          CurationConcerns.config.lock_time_to_live,
+          CurationConcerns.config.lock_retry_count,
+          CurationConcerns.config.lock_retry_delay)
+      end
+
+      def assign_visibility?(file_set_params = {})
+        !((file_set_params || {}).keys & %w(visibility embargo_release_date lease_expiration_date)).empty?
+      end
+
+      # This method can be overridden in case there is a custom approach for visibility (e.g. embargo)
+      def update_visibility(attributes)
+        interpret_visibility(attributes) # relies on CurationConcerns::ManagesEmbargoesActor to interpret and apply visibility
+      end
+
+      # copy visibility from source_concern to destination_concern
+      def copy_visibility(source_concern, destination_concern)
+        destination_concern.visibility =  source_concern.visibility
+      end
+  end
+end

data/app/actors/curation_concerns/lease_actor.rb

@@ -0,0 +1,19 @@
+module CurationConcerns
+  class LeaseActor
+    attr_reader :work
+
+    # @param [Hydra::Works::Work] work
+    def initialize(work)
+      @work = work
+    end
+
+    # Update the visibility of the work to match the correct state of the lease, then clear the lease date, etc.
+    # Saves the lease and the work
+    def destroy
+      work.lease_visibility! # If the lease has lapsed, update the current visibility.
+      work.deactivate_lease!
+      work.lease.save!
+      work.save!
+    end
+  end
+end