arclight 0.0.1 → 1.6.2

data/arclight.gemspec

@@ -1,28 +1,46 @@
-# coding: utf-8
+# frozen_string_literal: true
 
-lib = File.expand_path('../lib', __FILE__)
+lib = File.expand_path('lib', __dir__)
 $LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
 require 'arclight/version'
 
 Gem::Specification.new do |spec|
   spec.name          = 'arclight'
   spec.version       = Arclight::VERSION
-  spec.authors       = ['Jack Reed']
-  spec.email         = ['phillipjreed@gmail.com']
+  spec.authors       = ['Darren Hardy', 'Jessie Keck', 'Gordon Leacock', 'Jack Reed']
+  spec.email         = ['drh@stanford.edu', 'jessie.keck@gmail.com', 'gordonl@umich.edu', 'phillipjreed@gmail.com']
 
-  spec.summary       = ''
+  spec.summary       = 'A Blacklight-based environment to support discovery and delivery for archives and special collections'
   spec.description   = ''
-  spec.homepage      = 'https://github.com/sul-dlss/arclight'
-  spec.license       = 'Apache-2.0'
+  spec.homepage      = 'https://library.stanford.edu/projects/arclight'
+  spec.metadata = { 'source_code_uri' => 'https://github.com/projectblacklight/arclight' }
+  spec.license = 'Apache-2.0'
 
-  spec.files         = `git ls-files -z`.split("\x0").reject do |f|
+  spec.required_ruby_version = '>= 3.0.0'
+  spec.files = `git ls-files -z`.split("\x0").reject do |f|
     f.match(%r{^(test|spec|features)/})
   end
   spec.bindir        = 'exe'
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
   spec.require_paths = ['lib']
 
-  spec.add_development_dependency 'bundler', '~> 1.14'
-  spec.add_development_dependency 'rake', '~> 10.0'
-  spec.add_development_dependency 'rspec', '~> 3.0'
+  spec.add_dependency 'blacklight', '>= 8.0.0', '< 9'
+  spec.add_dependency 'gretel'
+  spec.add_dependency 'rails', '>= 7.1', '< 9'
+  spec.add_dependency 'traject', '~> 3.0'
+  spec.add_dependency 'traject_plus', '~> 2.0'
+
+  spec.add_development_dependency 'bundler'
+  spec.add_development_dependency 'capybara'
+  spec.add_development_dependency 'engine_cart'
+  spec.add_development_dependency 'i18n-tasks'
+  spec.add_development_dependency 'rake', '>= 12.0'
+  spec.add_development_dependency 'rspec-rails'
+  spec.add_development_dependency 'rubocop', '~> 1.8'
+  spec.add_development_dependency 'rubocop-rails', '~> 2.8'
+  spec.add_development_dependency 'rubocop-rake'
+  spec.add_development_dependency 'rubocop-rspec', '~> 2.3'
+  spec.add_development_dependency 'selenium-webdriver'
+  spec.add_development_dependency 'simplecov'
+  spec.add_development_dependency 'solr_wrapper'
 end

data/bin/rails

@@ -0,0 +1,13 @@
+#!/usr/bin/env ruby
+# This command will automatically be run when you run "rails" with Rails gems
+# installed from the root of your application.
+
+ENGINE_ROOT = File.expand_path('../..', __FILE__)
+ENGINE_PATH = File.expand_path('../../lib/arclight/engine', __FILE__)
+
+# Set up gems listed in the Gemfile.
+ENV['BUNDLE_GEMFILE'] ||= File.expand_path('../../Gemfile', __FILE__)
+require 'bundler/setup' if File.exist?(ENV['BUNDLE_GEMFILE'])
+
+require 'rails/all'
+require 'rails/engine/commands'

data/config/breadcrumbs.rb

@@ -0,0 +1,24 @@
+# frozen_string_literal: true
+
+# Root crumb
+crumb :root do
+  link t('arclight.routes.home'), root_path
+end
+
+crumb :repositories do
+  link t('arclight.routes.repositories'), arclight_engine.repositories_path
+end
+
+crumb :repository do |repository|
+  link repository.name, arclight_engine.repository_path(repository.slug)
+
+  parent :repositories
+end
+
+crumb :search_results do |search_state|
+  if search_state.filter('level').values == ['Collection']
+    link t('arclight.routes.collections')
+  else
+    link t('arclight.routes.search_results')
+  end
+end

data/config/i18n-tasks.yml

@@ -0,0 +1,134 @@
+# i18n-tasks finds and manages missing and unused translations: https://github.com/glebm/i18n-tasks
+
+# The "main" locale.
+base_locale: en
+## All available locales are inferred from the data by default. Alternatively, specify them explicitly:
+# locales: [es, fr]
+## Reporting locale, default: en. Available: en, ru.
+# internal_locale: en
+
+# Read and write translations.
+data:
+  ## Translations are read from the file system. Supported format: YAML, JSON.
+  ## Provide a custom adapter:
+  # adapter: I18n::Tasks::Data::FileSystem
+
+  # Locale files or `File.find` patterns where translations are read from:
+  read:
+    ## Default:
+    # - config/locales/%{locale}.yml
+    ## More files:
+    - config/locales/**/*.%{locale}.yml
+
+  # Locale files to write new keys to, based on a list of key pattern => file rules. Matched from top to bottom:
+  # `i18n-tasks normalize -p` will force move the keys according to these rules
+  write:
+    ## For example, write devise and simple form keys to their respective files:
+    # - ['{devise, simple_form}.*', 'config/locales/\1.%{locale}.yml']
+    ## Catch-all default:
+    # - config/locales/%{locale}.yml
+
+  # External locale data (e.g. gems).
+  # This data is not considered unused and is never written to.
+  external:
+    ## Example (replace %#= with %=):
+    # - "<%#= %x[bundle show vagrant].chomp %>/templates/locales/%{locale}.yml"
+
+  ## Specify the router (see Readme for details). Valid values: conservative_router, pattern_router, or a custom class.
+  # router: conservative_router
+
+  yaml:
+    write:
+      # do not wrap lines at 80 characters
+      line_width: -1
+
+  ## Pretty-print JSON:
+  # json:
+  #   write:
+  #     indent: '  '
+  #     space: ' '
+  #     object_nl: "\n"
+  #     array_nl: "\n"
+
+# Find translate calls
+search:
+  ## Paths or `File.find` patterns to search in:
+  paths:
+   - app/
+   - config/breadcrumbs.rb
+
+  ## Root directories for relative keys resolution.
+  relative_roots:
+    - app/components
+    - app/controllers
+    - app/helpers
+    - app/presenters
+    - app/views
+
+  ## Files or `File.fnmatch` patterns to exclude from search. Some files are always excluded regardless of this setting:
+  ##   %w(*.jpg *.png *.gif *.svg *.ico *.eot *.otf *.ttf *.woff *.woff2 *.pdf *.css *.sass *.scss *.less *.yml *.json)
+  exclude:
+    - app/assets/images
+    - app/assets/fonts
+    - app/assets/videos
+
+  ## Alternatively, the only files or `File.fnmatch patterns` to search in `paths`:
+  ## If specified, this settings takes priority over `exclude`, but `exclude` still applies.
+  # only: ["*.rb", "*.html.slim"]
+
+  ## If `strict` is `false`, guess usages such as t("categories.#{category}.title"). The default is `true`.
+  # strict: true
+
+  ## Multiple scanners can be used. Their results are merged.
+  ## The options specified above are passed down to each scanner. Per-scanner options can be specified as well.
+  ## See this example of a custom scanner: https://github.com/glebm/i18n-tasks/wiki/A-custom-scanner-example
+
+## Translation Services
+# translation:
+#   # Google Translate
+#   # Get an API key and set billing info at https://code.google.com/apis/console to use Google Translate
+#   google_translate_api_key: "AbC-dEf5"
+#   # DeepL Pro Translate
+#   # Get an API key and subscription at https://www.deepl.com/pro to use DeepL Pro
+#   deepl_api_key: "48E92789-57A3-466A-9959-1A1A1A1A1A1A"
+
+## Do not consider these keys missing:
+ignore_missing:
+- '{blacklight}.*'
+
+## Consider these keys used:
+ignore_unused:
+  - '{blacklight}.*'
+# - 'activerecord.attributes.*'
+# - '{devise,kaminari,will_paginate}.*'
+# - 'simple_form.{yes,no}'
+# - 'simple_form.{placeholders,hints,labels}.*'
+# - 'simple_form.{error_notification,required}.:'
+
+## Exclude these keys from the `i18n-tasks eq-base' report:
+# ignore_eq_base:
+#   all:
+#     - common.ok
+#   fr,es:
+#     - common.brand
+
+## Exclude these keys from the `i18n-tasks check-consistent-interpolations` report:
+# ignore_inconsistent_interpolations:
+# - 'activerecord.attributes.*'
+
+## Ignore these keys completely:
+# ignore:
+#  - kaminari.*
+
+## Sometimes, it isn't possible for i18n-tasks to match the key correctly,
+## e.g. in case of a relative key defined in a helper method.
+## In these cases you can use the built-in PatternMapper to map patterns to keys, e.g.:
+#
+# <%# I18n::Tasks.add_scanner 'I18n::Tasks::Scanners::PatternMapper',
+#       only: %w(*.html.haml *.html.slim),
+#       patterns: [['= title\b', '.page_title']] %>
+#
+# The PatternMapper can also match key literals via a special %{key} interpolation, e.g.:
+#
+# <%# I18n::Tasks.add_scanner 'I18n::Tasks::Scanners::PatternMapper',
+#       patterns: [['\bSpree\.t[( ]\s*%{key}', 'spree.%{key}']] %>

data/config/importmap.rb

@@ -0,0 +1,3 @@
+# frozen_string_literal: true
+
+pin_all_from File.expand_path('../app/assets/javascript', __dir__)

data/config/locales/arclight.en.yml

@@ -0,0 +1,100 @@
+---
+en:
+  arclight:
+    breadcrumbs:
+      aria_label: breadcrumb
+    collection_info_component:
+      collection_id: Collection ID
+      last_indexed: Last indexed
+      more_info: More info
+      online_items: Items available online
+      total_components: Total components
+    hierarchy:
+      view_all: View
+    masthead_heading_html: Archival Collections at Institution
+    repositories: Repositories
+    request:
+      container: Request
+    routes:
+      collections: Collections
+      home: Home
+      repositories: Repositories
+      search_results: Search results
+    search:
+      collections_header: Collections
+      repository_header: 'Collections : [%{repository}]'
+    show_breadcrumb_label: Repository
+    truncation:
+      view_less: view less
+      view_more: view more
+    views:
+      index:
+        all_group_results: view all %{count}
+        all_results: All results
+        group_by_collection: Grouped by collection
+        group_results: Group results
+        group_results_count:
+          one: 1 result
+          other: "%{count} results"
+          zero: 0 results
+        number_of_components:
+          one: component
+          other: components
+        top_group_results:
+          one: Top %{count} result
+          other: Top %{count} results
+      repositories:
+        number_of_collections:
+          one: 1 collection
+          other: "%{count} collections"
+          zero: No collections
+        show: "%{name} - Arclight"
+        title: Repositories - Arclight
+        view_all_collections: View all of our collections
+        view_more: View more
+      show:
+        collection_id: 'Collection ID: %{id}'
+        contents: Contents
+        context: Collection context
+        context_nav:
+          title: Collection overview
+        download:
+          default: Download
+          multiple:
+            ead: EAD
+            pdf: Finding aid
+          single:
+            ead: Download EAD
+            pdf: Download finding aid
+        embedded_content: Online content
+        expand: Expand
+        has_content: Collection contents
+        online_content:
+          description: Some materials from this collection are available online.
+          link_text: Show only online content
+          title: Online content
+        our_collections: Our Collections
+        sections:
+          access_field: Access and use
+          background_field: Background
+          collection_context_field: Collection context
+          component_field: About this %{level}
+          component_indexed_terms_field: Indexed terms
+          indexed_terms_field: Indexed terms
+          related_field: Related
+          scope_and_arrangement_field: Scope and arrangement
+          summary_field: Summary
+        toggle_sidebar: Navigate the collection
+    within_collection_dropdown:
+      all_collections: all collections
+      label_html: Search<span class="sr-only visually-hidden"> within</span>
+      this_collection: this collection
+  blacklight:
+    entry_name:
+      grouped:
+        _root_: collections
+    search:
+      pagination_info:
+        pages:
+          one: "<strong>%{start_num}</strong> - <strong>%{end_num}</strong> of <strong>%{total_num} %{entry_name}</strong>"
+          other: "<strong>%{start_num}</strong> - <strong>%{end_num}</strong> of <strong>%{total_num} %{entry_name}</strong>"

data/config/routes.rb

@@ -0,0 +1,6 @@
+# frozen_string_literal: true
+
+Arclight::Engine.routes.draw do
+  get 'collections' => 'catalog#index', defaults: { f: { level: ['Collection'] } }, as: :collections
+  resources :repositories, only: %i[index show], controller: 'arclight/repositories'
+end

data/docker-compose.yml

@@ -0,0 +1,16 @@
+version: "3.7"
+
+services:
+  solr:
+    image: solr:latest
+    volumes:
+      - $PWD/solr/conf:/opt/solr/conf
+    ports:
+      - 8983:8983
+    entrypoint:
+      - docker-entrypoint.sh
+      - solr-precreate
+      - blacklight-core
+      - /opt/solr/conf
+      - "-Xms256m"
+      - "-Xmx512m"

data/lib/arclight/digital_object.rb

@@ -0,0 +1,27 @@
+# frozen_string_literal: true
+
+module Arclight
+  ##
+  # Plain ruby class to model serializing/deserializing digital object data
+  class DigitalObject
+    attr_reader :label, :href
+
+    def initialize(label:, href:)
+      @label = label.presence || href
+      @href = href
+    end
+
+    def to_json(*)
+      { label: label, href: href }.to_json
+    end
+
+    def self.from_json(json)
+      object_data = JSON.parse(json)
+      new(label: object_data['label'], href: object_data['href'])
+    end
+
+    def ==(other)
+      href == other.href && label == other.label
+    end
+  end
+end

data/lib/arclight/engine.rb

@@ -0,0 +1,47 @@
+# frozen_string_literal: true
+
+require 'blacklight'
+require 'traject'
+require 'active_model'
+require 'arclight/exceptions'
+require 'arclight/normalized_date'
+require 'arclight/normalized_id'
+require 'arclight/normalized_title'
+require 'arclight/digital_object'
+require 'gretel'
+
+module Arclight
+  ##
+  # This is the defining class for the Arclight Rails Engine
+  class Engine < ::Rails::Engine
+    config.oembed_resource_exclude_patterns = [/\.pdf$/, /\.ppt$/]
+
+    Arclight::Engine.config.catalog_controller_group_query_params = {
+      group: true,
+      'group.field': '_root_',
+      'group.ngroups': true,
+      'group.limit': 3,
+      fl: '*,collection:[subquery]',
+      'collection.q': '{!terms f=id v=$row._root_}',
+      'collection.defType': 'lucene',
+      'collection.fl': '*',
+      'collection.rows': 1
+    }
+
+    initializer 'arclight.helpers' do
+      config.after_initialize do
+        ActiveSupport.on_load(:action_view) { include ArclightHelper }
+      end
+    end
+
+    initializer 'arclight.assets', before: 'assets' do |app|
+      app.config.assets.precompile << 'arclight/arclight.js'
+      app.config.assets.precompile << 'arclight/oembed_controller.js'
+      app.config.assets.precompile << 'arclight/truncate_controller.js'
+    end
+
+    initializer 'arclight.importmap', before: 'importmap' do |app|
+      app.config.importmap.paths << Engine.root.join('config/importmap.rb') if app.config.respond_to?(:importmap)
+    end
+  end
+end

data/lib/arclight/exceptions.rb

@@ -0,0 +1,19 @@
+# frozen_string_literal: true
+
+module Arclight
+  module Exceptions
+    # Id's must be present on all documents and components
+    class IDNotFound < StandardError
+      def message
+        'id must be present for all documents and components'
+      end
+    end
+
+    # Unittitle or unitdate must be present on all documents and components
+    class TitleNotFound < StandardError
+      def message
+        '<unittitle/> or <unitdate/> must be present for all documents and components'
+      end
+    end
+  end
+end

data/lib/arclight/hash_absolute_xpath.rb

@@ -0,0 +1,62 @@
+# frozen_string_literal: true
+
+require 'digest'
+
+module Arclight
+  ##
+  # Take a Nokogiri node and get its absolute path (inserting our own indexes for component levels)
+  # and hash that outout. This is intended as a potential strategy for handling missing IDs in EADs.
+  class HashAbsoluteXpath
+    class << self
+      attr_writer :hash_algorithm
+
+      def hash_algorithm
+        return Digest::SHA1 unless defined? @hash_algorithm
+
+        @hash_algorithm
+      end
+    end
+
+    COMPONENT_NODE_NAME_REGEX = /^c\d{,2}$/
+    attr_reader :node
+
+    def initialize(node)
+      @node = node
+    end
+
+    def to_hexdigest
+      self.class.hash_algorithm.hexdigest(absolute_xpath).prepend('al_')
+    end
+
+    def absolute_xpath
+      ancestor_tree = node.ancestors.map do |ancestor|
+        ancestor_name_and_index(ancestor)
+      end
+
+      "#{[ancestor_tree.reverse, node.name].flatten.join('/')}#{current_index}"
+    end
+
+    private
+
+    def current_index
+      siblings.index(node)
+    end
+
+    def component_siblings_for_node(xml_node)
+      xml_node.parent.children.select { |n| n.name =~ COMPONENT_NODE_NAME_REGEX }
+    end
+
+    def siblings
+      @siblings ||= component_siblings_for_node(node)
+    end
+
+    def ancestor_name_and_index(ancestor)
+      if ancestor.name =~ COMPONENT_NODE_NAME_REGEX
+        index = component_siblings_for_node(ancestor).index(ancestor)
+        "#{ancestor.name}#{index}"
+      else
+        ancestor.name
+      end
+    end
+  end
+end

data/lib/arclight/level_label.rb

@@ -0,0 +1,46 @@
+# frozen_string_literal: true
+
+module Arclight
+  ##
+  # A utility class to return a human-readable label for an EAD @level code.
+  # Can use the value from @otherlevel if provided.
+  # Examples from @level: recordgrp = "Record Group"
+  #                       collection = "Collection"
+  #                       subseries = "Subseries"
+  #                       otherlevel = (text provided in @otherlevel)
+  class LevelLabel
+    # @param [String] `level` from the collection or component @level
+    # @param [String] `other_level` from the collection or component @otherlevel
+    def initialize(level, other_level = nil)
+      @level = level
+      @other_level = other_level if other_level.present?
+    end
+
+    # @return [String] the human-readable label
+    def to_s
+      human_readable_level
+    end
+
+    private
+
+    attr_reader :level, :other_level
+
+    CUSTOM_LEVEL_LABELS = {
+      recordgrp: 'Record Group',
+      subgrp: 'Subgroup'
+    }.freeze
+
+    def human_readable_level
+      if level == 'otherlevel'
+        alternative_level
+      elsif level.present?
+        CUSTOM_LEVEL_LABELS.fetch(level.to_sym, level.capitalize).to_s
+      end
+    end
+
+    def alternative_level
+      alternative_level = other_level if other_level
+      alternative_level.present? ? alternative_level.capitalize : 'Other'
+    end
+  end
+end

data/lib/arclight/missing_id_strategy.rb

@@ -0,0 +1,21 @@
+# frozen_string_literal: true
+
+require 'arclight/hash_absolute_xpath'
+
+module Arclight
+  ##
+  # A class to configure a selected MissingIdStrategy.
+  # Defaults to Arclight::HashAbsoluteXpath
+  # This can be updated in an initializer to be any other class
+  class MissingIdStrategy
+    class << self
+      attr_writer :selected
+
+      def selected
+        return Arclight::HashAbsoluteXpath unless defined? @selected
+
+        @selected
+      end
+    end
+  end
+end

data/lib/arclight/normalized_date.rb

@@ -0,0 +1,49 @@
+# frozen_string_literal: true
+
+module Arclight
+  ##
+  # A utility class to normalize dates, typically by joining inclusive and bulk dates
+  # e.g., "1990-2000, bulk 1990-1999"
+  # @see http://www2.archivists.org/standards/DACS/part_I/chapter_2/4_date
+  class NormalizedDate
+    # @param [String | Array<String>] `inclusive` from the `unitdate`
+    # @param [Array<String>] `bulk` from the `unitdate`
+    # @param [Array<String>] `other` from the `unitdate` when type is not specified
+    def initialize(inclusive, bulk = [], other = [])
+      @inclusive = (inclusive || []).map do |inclusive_text|
+        if inclusive_text.is_a? Array # of YYYY-YYYY for ranges
+          # NOTE: This code is not routable AFAICT in actual indexing.
+          # We pass arrays of strings (or xml nodes) here, and never a multidimensional array
+          year_range(inclusive_text)
+        elsif inclusive_text.present?
+          inclusive_text.strip
+        end
+      end&.join(', ')
+
+      @bulk = Array.wrap(bulk).compact.map(&:strip).join(', ')
+      @other = Array.wrap(other).compact.map(&:strip).join(', ')
+    end
+
+    # @return [String] the normalized title/date
+    def to_s
+      normalize
+    end
+
+    private
+
+    attr_reader :inclusive, :bulk, :other
+
+    def year_range(date_array)
+      YearRange.new(date_array.include?('/') ? date_array : date_array.map { |v| v.tr('-', '/') }).to_s
+    end
+
+    # @see http://www2.archivists.org/standards/DACS/part_I/chapter_2/4_date for rules
+    def normalize
+      result = []
+      result << inclusive if inclusive.present?
+      result << other if other.present?
+      result << "bulk #{bulk}" if bulk.present?
+      result.compact.map(&:strip).join(', ').presence
+    end
+  end
+end

data/lib/arclight/normalized_id.rb

@@ -0,0 +1,30 @@
+# frozen_string_literal: true
+
+require 'arclight/exceptions'
+
+module Arclight
+  ##
+  # A simple utility class to normalize identifiers
+  class NormalizedId
+    # Accepts unused kwargs from the ead2_config.rb id to_field directive
+    # (:title and :repository) so that applications can provide a custom
+    # id_normalizer class to traject to form the collection id from these attributes.
+    def initialize(id, **_kwargs)
+      @id = id
+    end
+
+    def to_s
+      normalize
+    end
+
+    private
+
+    attr_reader :id
+
+    def normalize
+      raise Arclight::Exceptions::IDNotFound if id.blank?
+
+      id.strip.tr('.', '-')
+    end
+  end
+end