uc3-dmp-id 0.0.1

data/lib/uc3-dmp-id/creator.rb

@@ -0,0 +1,11 @@
+# frozen_string_literal: true
+
+module Uc3DmpId
+  class Uc3DmpIdCreatorError << StandardError; end
+
+  class Creator
+    class << self
+
+    end
+  end
+end

data/lib/uc3-dmp-id/deleter.rb

@@ -0,0 +1,11 @@
+# frozen_string_literal: true
+
+module Uc3DmpId
+  class Uc3DmpIdDeleterError << StandardError; end
+
+  class Deleter
+    class << self
+
+    end
+  end
+end

data/lib/uc3-dmp-id/finder.rb

@@ -0,0 +1,127 @@
+# frozen_string_literal: true
+
+require 'uc3-dmp-dynamo'
+
+module Uc3DmpId
+  class Uc3DmpIdFinderError << StandardError; end
+
+  # Methods to find/search for DMP IDs
+  class Finder
+    MSG_INVALID_ARGS = 'Expected JSON to be structured as `{ "dmp": { "PK": "value"} } OR \
+                        { "dmp": { "dmp_id": { "identifier": "value", "type": "value" } }`'
+    MSG_MISSING_PK = 'No PK was provided'
+    MSG_MISSING_PROV_ID = 'No Provenance identifier was provided. \
+                           Expected: `{ "dmp_id": { "identifier": "value", "type": "value" }`'
+
+    class << self
+      # TODO: Replace this with ElasticSearch
+      def search_dmps(**_args)
+
+        # TODO: Need to move this to ElasticSearch!!!
+      end
+      # rubocop:enable Metrics/MethodLength
+
+      # Find the DMP's versions
+      # -------------------------------------------------------------------------
+      def versions(p_key:)
+        raise Uc3DmpIdFinderError, MSG_MISSING_PK if p_key.nil?
+
+        args = {
+          key_conditions: {
+            PK: { attribute_value_list: [Helper.append_pk_prefix(dmp: p_key)], comparison_operator: 'EQ' }
+          },
+          projection_expression: 'modified',
+          scan_index_forward: false
+        }
+        client = Uc3DmpDynamo::Client.new
+        client.query(**args)
+      end
+
+      # Find a DMP based on the contents of the incoming JSON
+      # -------------------------------------------------------------------------
+      def by_json(json:)
+        json = Validator.parse_json(json: json)&.fetch('dmp', {})
+        raise Uc3DmpIdFinderError, MSG_INVALID_ARGS if json.nil? || (json['PK'].nil? && json['dmp_id'].nil?)
+
+        p_key = json['PK']
+        # Translate the incoming :dmp_id into a PK
+        p_key = Helper.dmp_id_to_pk(json: json.fetch('dmp_id', {})) if p_key.nil?
+
+        # find_by_PK
+        resp = by_pk(p_key: p_key, s_key: json['SK']) unless p_key.nil?
+        return resp unless resp.nil?
+
+        # find_by_dmphub_provenance_id -> if no PK and no dmp_id result
+        by_dmphub_provenance_identifier(json: json)
+      end
+
+      # Find the DMP by its PK and SK
+      # -------------------------------------------------------------------------
+      def by_pk(p_key:, s_key: Helper::DMP_LATEST_VERSION)
+        raise Uc3DmpIdFinderError, MSG_MISSING_PK if p_key.nil?
+
+        s_key = Helper::DMP_LATEST_VERSION if s_key.nil? || s_key.strip.empty?
+
+        client = Uc3DmpDynamo::Client.new
+        resp = client.get_item(
+          key: {
+            PK: Helper.append_pk_prefix(dmp: p_key),
+            SK: s_key.nil? || s_key.strip.empty? ? Helper::DMP_LATEST_VERSION : s_key
+          }
+        )
+        return nil if resp.nil? || resp.fetch('dmp', {})['PK'].nil?
+
+        _append_versions(p_key: resp['dmp']['PK'], dmp: resp)
+      end
+
+      # Attempt to find the DMP item by the provenance system's identifier
+      # -------------------------------------------------------------------------
+      # rubocop:disable Metrics/AbcSize
+      def by_dmphub_provenance_identifier(json:)
+        raise Uc3DmpIdFinderError, MSG_MISSING_PROV_ID if json.nil? || json.fetch('dmp_id', {})['identifier'].nil?
+
+        args = {
+          key_conditions: {
+            dmphub_provenance_identifier: {
+              attribute_value_list: [json['dmp_id']['identifier']],
+              comparison_operator: 'EQ'
+            }
+          },
+          filter_expression: 'SK = :version',
+          expression_attribute_values: { ':version': KeyHelper::DMP_LATEST_VERSION }
+        }
+        client = Uc3DmpDynamo::Client.new
+        resp = client.query(**args)
+        return resp if resp.nil? || resp['dmp'].nil?
+
+        # If we got a hit, fetch the DMP and return it.
+        by_pk(p_key: resp['dmp']['PK'], s_key: resp['dmp']['SK'])
+      end
+      # rubocop:enable Metrics/AbcSize
+
+      private
+
+      # Build the dmphub_versions array and attach it to the dmp
+      # rubocop:disable Metrics/AbcSize
+      def _append_versions(p_key:, dmp:)
+        return dmp if p_key.nil? || !dmp.is_a?(Hash) || dmp['dmp'].nil?
+
+        results = find_dmp_versions(p_key: p_key)
+        return dmp unless results.length > 1
+
+        versions = results.map do |version|
+          next if version.fetch('dmp', {})['modified'].nil?
+
+          timestamp = version['dmp']['modified']
+          {
+            timestamp: timestamp,
+            url: "#{Helper.api_base_url}dmps/#{Helper.remove_pk_prefix(dmp: p_key)}?version=#{timestamp}"
+          }
+        end
+        dmp['dmp']['dmphub_versions'] = JSON.parse(versions.to_json)
+        dmp
+      end
+      # rubocop:enable Metrics/AbcSize
+    end
+  end
+end

data/lib/uc3-dmp-id/helper.rb

@@ -0,0 +1,112 @@
+# frozen_string_literal: true
+
+
+# TODO: Be sure to update the API functions so that they call cleanse_dmp_json before
+#       calling Uc3DmpApiCore::Responder.respond !!!!!!!!!!
+
+
+module Uc3DmpId
+  # Helper functions for working with DMP IDs
+  class Helper
+    PK_DMP_PREFIX = 'DMP#'
+    PK_DMP_REGEX = %r{DMP#[a-zA-Z0-9\-_.]+/[a-zA-Z0-9]{2}\.[a-zA-Z0-9./:]+}.freeze
+
+    SK_DMP_PREFIX = 'VERSION#'
+    SK_DMP_REGEX = /VERSION#\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}\+\d{2}:\d{2}/.freeze
+
+    # TODO: Verify the assumed structure of the DOI is valid
+    DOI_REGEX = %r{[0-9]{2}\.[0-9]{5}/[a-zA-Z0-9/_.-]+}.freeze
+    URL_REGEX = %r{(https?://)?([a-zA-Z0-9\-_]\.)+[a-zA-Z0-9\-_]{2,3}(:[0-9]+)?/?}.freeze
+
+    DMP_LATEST_VERSION = "#{SK_DMP_PREFIX}latest"
+    DMP_TOMBSTONE_VERSION = "#{SK_DMP_PREFIX}tombstone"
+
+    class << self
+      # Return the base URL for a DMP ID
+      def dmp_id_base_url
+        url = ENV.fetch('DMP_ID_BASE_URL', 'https://dmptool-dev.cdlib.org/dmps/')
+        url&.end_with?('/') ? url : "#{url}/"
+      end
+
+      # Return the base URL for the API
+      def api_base_url
+        url = ENV.fetch('DMP_ID_BASE_URL', 'https://api.dmptool-dev.cdlib.org/dmps/')
+        url&.end_with?('/') ? url : "#{url}/"
+      end
+
+      # Format the DMP ID in the way we want it
+      def format_dmp_id(value:, with_protocol: false)
+        dmp_id = value.match(DOI_REGEX).to_s
+        return nil if dmp_id.nil? || dmp_id == ''
+        # If it's already a URL, return it as is
+        return value if value.start_with?('http')
+
+        dmp_id = dmp_id.gsub('doi:', '')
+        dmp_id = dmp_id.start_with?('/') ? dmp_id[1..dmp_id.length] : dmp_id
+        base_domain = with_protocol ? dmp_id_base_url : dmp_id_base_url.gsub(%r{https?://}, '')
+        "#{base_domain}#{dmp_id}"
+      end
+
+      # Convert an API PathParameter (DMP ID) into a PK
+      def path_parameter_to_pk(param:)
+        return nil unless param.is_a?(String) && !param.strip.empty?
+
+        base_domain = dmp_id_base_url.gsub(%r{https?://}, '')
+        p_key = param if param.start_with?(dmp_id_base_url) || param.start_with?(base_domain)
+        p_key = CGI.unescape(p_key.nil? ? param : p_key)
+        p_key = format_dmp_id(value: p_key)
+        append_pk_prefix(dmp: p_key)
+      end
+
+      # Append the :PK prefix to the :dmp_id
+      def dmp_id_to_pk(json:)
+        return nil if json.nil? || json['identifier'].nil?
+
+        # If it's a DOI format it correctly
+        dmp_id = format_dmp_id(value: json['identifier'].to_s)
+        return nil if dmp_id.nil? || dmp_id == ''
+
+        append_pk_prefix(dmp: dmp_id)
+      end
+
+      # Derive the DMP ID by removing the :PK prefix
+      def pk_to_dmp_id(p_key:)
+        return nil if p_key.nil?
+
+        {
+          type: 'doi',
+          identifier: format_dmp_id(value: remove_pk_prefix(dmp: p_key), with_protocol: true)
+        }
+      end
+
+      # Parse the incoming JSON if necessary or return as is if it's already a Hash
+      def parse_json(json:)
+        return json if json.is_a?(Hash)
+
+        json.is_a?(String) ? JSON.parse(json) : nil
+      end
+
+      # Recursive method that strips out any DMPHub related metadata from a DMP record before sending
+      # it to the caller
+      # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
+      def cleanse_dmp_json(json:)
+        return json unless json.is_a?(Hash) || json.is_a?(Array)
+
+        # If it's an array clean each of the objects individually
+        return json.map { |obj| cleanse_dmp_json(json: obj) }.compact if json.is_a?(Array)
+
+        cleansed = {}
+        allowable = %w[dmphub_versions]
+        json.each_key do |key|
+          next if (key.to_s.start_with?('dmphub') && !allowable.include?(key)) || %w[PK SK].include?(key.to_s)
+
+          obj = json[key]
+          # If this object is a Hash or Array then recursively cleanse it
+          cleansed[key] = obj.is_a?(Hash) || obj.is_a?(Array) ? cleanse_dmp_json(json: obj) : obj
+        end
+        cleansed.keys.any? ? cleansed : nil
+      end
+      # rubocop:enable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
+    end
+  end
+end

data/lib/uc3-dmp-id/updater.rb

@@ -0,0 +1,11 @@
+# frozen_string_literal: true
+
+module Uc3DmpId
+  class Uc3DmpIdUpdaterError << StandardError; end
+
+  class Updater
+    class << self
+
+    end
+  end
+end

data/lib/uc3-dmp-id/validator.rb

@@ -0,0 +1,64 @@
+# frozen_string_literal: true
+
+module Uc3DmpId
+  class Uc3DmpIdValidatorError << StandardError; end
+
+  class Validator
+    # Valid Validation modes are:
+    #   - :author --> system of provenance is attempting to create or update
+    #   - :delete --> system of provenance is attempting to delete/tombstone
+    #   - :amend  --> a non-provenance system is attempting to update
+    VALIDATION_MODES = %w[author amend delete].freeze
+
+    MSG_EMPTY_JSON = 'JSON was empty or was not a valid JSON document!'
+    MSG_INVALID_JSON = 'Invalid JSON.'
+    MSG_NO_SCHEMA = 'No JSON schema available!'
+    MSG_BAD_JSON = 'Fatal validation error: %{msg} - %{trace}'
+    MSG_VALID_JSON = 'The JSON is valid.'
+
+    class << self
+      # Validate the specified DMP's :json against the schema for the specified :mode
+      #
+      # ------------------------------------------------------------------------------------
+      def validate(mode:, json:)
+        json = Helper.parse_json(json: json)
+        return [MSG_EMPTY_JSON] if json.nil? || !VALIDATION_MODES.include?(mode)
+
+        # Load the appropriate JSON schema for the mode
+        schema = _load_schema(mode: mode)
+        return [MSG_NO_SCHEMA] if schema.nil?
+
+        # Validate the JSON
+        errors = JSON::Validator.fully_validate(schema, json)
+        errors = errors.map { |err| err.gsub('The property \'#/\' ', '') }
+        errors = ([MSG_INVALID_JSON] << errors).flatten.compact.uniq unless errors.empty?
+        errors.map { |err| err.gsub(/in schema [a-z0-9-]+/, '').strip }
+      rescue JSON::Schema::ValidationError => e
+        raise Uc3DmpIdValidatorError, MSG_BAD_JSON % { msg: e.message, trace: e.backtrace }
+      end
+
+      # ------------------------------------------------------------------------------------
+      # METHODS BELOW ARE ONLY MEANT TO BE INVOKED FROM WITHIN THIS MODULE
+      # ------------------------------------------------------------------------------------
+
+      # Load the JSON schema that corresponds with the mode
+      # ------------------------------------------------------------------------------------
+      def _load_schema(mode:)
+        schema = "#{_schema_dir}/schemas/#{mode}.json"
+        file = schema if File.exist?(schema)
+        return nil if mode.nil? || file.nil? || !File.exist?(file)
+
+        JSON.parse(File.read(file))
+      rescue JSON::ParserError
+        nil
+      end
+
+      # The location of th JSON schema files
+      # ------------------------------------------------------------------------------------
+      def _schema_dir
+        # TODO: Switch this to the gem dirctory, not sure if this is the same as the Layer below
+        '/opt/ruby'
+      end
+    end
+  end
+end

data/lib/uc3-dmp-id/version.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+module Uc3DmpId
+  VERSION = '0.0.1'
+end

data/lib/uc3-dmp-id/versioner.rb

@@ -0,0 +1,11 @@
+# frozen_string_literal: true
+
+module Uc3DmpId
+  class Uc3DmpIdVersionerError << StandardError; end
+
+  class Versioner
+    class << self
+
+    end
+  end
+end

data/lib/uc3-dmp-id.rb

@@ -0,0 +1,18 @@
+# rubocop:disable Naming/FileName
+# frozen_string_literal: true
+
+require 'json'
+require 'json-schema'
+
+require 'uc3-dmp-id/creator'
+require 'uc3-dmp-id/deleter'
+require 'uc3-dmp-id/finder'
+require 'uc3-dmp-id/helper'
+require 'uc3-dmp-id/updater'
+require 'uc3-dmp-id/validator'
+require 'uc3-dmp-id/versioner'
+
+module Uc3DmpId
+  MSG_DMP_NOT_FOUND = 'The DMP ID does not exist'
+end
+# rubocop:enable Naming/FileName

metadata

@@ -0,0 +1,168 @@
+--- !ruby/object:Gem::Specification
+name: uc3-dmp-id
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+platform: ruby
+authors:
+- Brian Riley
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2023-05-09 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: json
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.6'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.6'
+- !ruby/object:Gem::Dependency
+  name: json-schema
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.0'
+- !ruby/object:Gem::Dependency
+  name: logger
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.4'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.4'
+- !ruby/object:Gem::Dependency
+  name: uc3-dmp-dynamo
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.83'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.83'
+- !ruby/object:Gem::Dependency
+  name: byebug
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 11.1.3
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 11.1.3
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.9.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.9.0
+- !ruby/object:Gem::Dependency
+  name: rubocop
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 1.50.2
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 1.50.2
+- !ruby/object:Gem::Dependency
+  name: rubocop-rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 2.20.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 2.20.0
+description: Helpers for working with JSON that represents a DMP ID
+email:
+- brian.riley@ucop.edu
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- README.md
+- lib/schemas/amend.json
+- lib/schemas/author.json
+- lib/uc3-dmp-id.rb
+- lib/uc3-dmp-id/creator.rb
+- lib/uc3-dmp-id/deleter.rb
+- lib/uc3-dmp-id/finder.rb
+- lib/uc3-dmp-id/helper.rb
+- lib/uc3-dmp-id/updater.rb
+- lib/uc3-dmp-id/validator.rb
+- lib/uc3-dmp-id/version.rb
+- lib/uc3-dmp-id/versioner.rb
+homepage: https://github.com/CDLUC3/dmp-hub-cfn/blob/main/src/sam/gems/uc3-dmp-id
+licenses:
+- MIT
+metadata:
+  rubygems_mfa_required: 'false'
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '2.7'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.1.6
+signing_key: 
+specification_version: 4
+summary: DMPTool gem that provides support for DMP ID records
+test_files: []