cosing 0.1.0

data/lib/cosing/annex/base.rb

@@ -0,0 +1,56 @@
+# frozen_string_literal: true
+
+module Cosing
+  module Annex
+    class Base
+      def initialize
+        @rules = {}
+      end
+
+      def keys
+        @rules.keys
+      end
+
+      def add_rule(params)
+        return unless params[:reference_number]
+
+        @rules[params[:reference_number]] = self.class::Rule.new(
+          **params.merge(annex: self.class.name.gsub("::", " "))
+        )
+      end
+
+      def lookup(reference_number)
+        @rules.fetch(
+          reference_number.to_s,
+          fuzzy_find(reference_number.to_s)
+        )
+      rescue KeyError => e
+        # Known missing data from original source
+        return nil if reference_number == "19" && instance_of?(Annex::VI)
+        return nil if reference_number == "41" && instance_of?(Annex::VI)
+        return nil if reference_number == "31" && instance_of?(Annex::VI)
+        return nil if reference_number == "44" && instance_of?(Annex::VI)
+        return nil if reference_number == "268" && instance_of?(Annex::III)
+
+        raise e
+      end
+
+      private
+
+      def fuzzy_find(reference_number)
+        return @rules[reference_number] if @rules.key?(reference_number)
+
+        candidates = @rules.keys.grep(/#{reference_number}[abcd]/)
+        candidates
+          .map { |candidate| @rules[candidate] }
+          .tap do |candidates|
+            if candidates.empty?
+              raise KeyError,
+                    "Could not find #{reference_number} in #{self.class}, " \
+                    "valid keys are #{keys}"
+            end
+          end
+      end
+    end
+  end
+end

data/lib/cosing/annex/ii.rb

@@ -0,0 +1,11 @@
+# frozen_string_literal: true
+
+module Cosing
+  module Annex
+    class II < Base
+      class Rule < Rule
+        attribute :inn, Types::String
+      end
+    end
+  end
+end

data/lib/cosing/annex/iii.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+module Cosing
+  module Annex
+    class III < Base
+      class Rule < Rule
+        attribute :common_ingredients, Types::Array.of(Types::String)
+        attribute :inn, Types::String
+        attribute :maximum_concentration, Types::String
+        attribute :other_restrictions, Types::String
+        attribute :product_type, Types::String
+      end
+    end
+  end
+end

data/lib/cosing/annex/iv.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+module Cosing
+  module Annex
+    class IV < Base
+      class Rule < Rule
+        attribute :colour, Types::String
+        attribute :colour_index_number, Types::String
+        attribute :maximum_concentration, Types::String
+        attribute :other_restrictions, Types::String
+        attribute :product_type, Types::String
+        attribute :wording_of_conditions, Types::String
+      end
+    end
+  end
+end

data/lib/cosing/annex/rule.rb

@@ -0,0 +1,17 @@
+module Cosing
+  module Annex
+    class Rule < Dry::Struct
+      attribute :annex, Types::String
+      attribute :cas_numbers, Types::Array.of(Types::String)
+      attribute :chemical_name, Types::String
+      attribute :cmr, Types::String
+      attribute :ec_numbers, Types::Array.of(Types::String)
+      attribute :identified_ingredients, Types::Array.of(Types::String)
+      attribute :other_regulations, Types::String
+      attribute :reference_number, Types::String
+      attribute :regulated_by, Types::String
+      attribute :regulation, Types::String
+      attribute :sccs_opinions, Types::Array.of(SccsOpinion)
+    end
+  end
+end

data/lib/cosing/annex/sccs_opinion.rb

@@ -0,0 +1,12 @@
+module Cosing
+  module Annex
+    class SccsOpinion < Dry::Struct
+      attribute :code, Types::String
+      attribute :description, Types::String
+
+      def inspect
+        to_h
+      end
+    end
+  end
+end

data/lib/cosing/annex/v.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+module Cosing
+  module Annex
+    class V < Base
+      class Rule < Rule
+        attribute :inn, Types::String
+        attribute :maximum_concentration, Types::String
+        attribute :other_restrictions, Types::String
+        attribute :product_type, Types::String
+        attribute :wording_of_conditions, Types::String
+      end
+    end
+  end
+end

data/lib/cosing/annex/vi.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+module Cosing
+  module Annex
+    class VI < Base
+      class Rule < Rule
+        attribute :common_ingredients, Types::Array.of(Types::String)
+        attribute :inn, Types::String
+        attribute :maximum_concentration, Types::String
+        attribute :other_restrictions, Types::String
+        attribute :product_type, Types::String
+        attribute :wording_of_conditions, Types::String
+      end
+    end
+  end
+end

data/lib/cosing/annex.rb

@@ -0,0 +1,198 @@
+# frozen_string_literal: true
+
+require_relative "annex/sccs_opinion"
+require_relative "annex/rule"
+require_relative "annex/base"
+require_relative "annex/ii"
+require_relative "annex/iii"
+require_relative "annex/iv"
+require_relative "annex/v"
+require_relative "annex/vi"
+
+module Cosing
+  module Annex
+    module_function
+
+    def parse(path)
+      CSV.parse(
+        File.read(path),
+        headers: true,
+        liberal_parsing: true,
+        header_converters: :symbol
+      ) do |row|
+        row = row
+          .to_h
+          .transform_values do |value|
+            value.to_s.strip.then { |val| val == "-" ? "" : val }
+          end
+
+        cas_numbers = build_cas_numbers(row)
+        ec_numbers = build_ec_numbers(row)
+        sccs_opinions = build_sccs_opinions(row)
+
+        yield row.merge(
+          cas_numbers: cas_numbers.compact,
+          ec_numbers: ec_numbers.compact,
+          sccs_opinions: sccs_opinions.compact,
+        )
+      end
+    end
+
+    def load
+      annex_ii = Annex::II.new.tap do |annex|
+        parse("data/annex.II.csv") do |row|
+          identified_ingredients = transform_array!(
+            row,
+            key: :identified_ingredients,
+            split: ";"
+          )
+
+          annex.add_rule(
+            row.merge(
+              identified_ingredients: identified_ingredients.compact
+            )
+          )
+        end
+      end
+
+      annex_iii = Annex::III.new.tap do |annex|
+        parse("data/annex.III.csv") do |row|
+          common_ingredients = transform_array!(
+            row,
+            key: :common_ingredients,
+            split: ";"
+          )
+          identified_ingredients = transform_array!(
+            row,
+            key: :identified_ingredients,
+            split: ";"
+          )
+
+          annex.add_rule(
+            row.merge(
+              common_ingredients: common_ingredients.compact,
+              identified_ingredients: identified_ingredients.compact,
+              other_restrictions: row[:other]
+            )
+          )
+        end
+      end
+
+      annex_iv = Annex::IV.new.tap do |annex|
+        parse("data/annex.IV.csv") do |row|
+          identified_ingredients = transform_array!(
+            row,
+            key: :identified_ingredients,
+            split: ";"
+          )
+
+          annex.add_rule(
+            row.merge(
+              identified_ingredients: identified_ingredients.compact,
+              other_restrictions: row[:other]
+            )
+          )
+        end
+      end
+
+      annex_v = Annex::V.new.tap do |annex|
+        parse("data/annex.V.csv") do |row|
+          common_ingredients = transform_array!(
+            row,
+            key: :common_ingredients,
+            split: ";"
+          )
+          identified_ingredients = transform_array!(
+            row,
+            key: :identified_ingredients,
+            split: ";"
+          )
+
+          annex.add_rule(
+            row.merge(
+              common_ingredients: common_ingredients.compact,
+              identified_ingredients: identified_ingredients.compact,
+              other_restrictions: row[:other]
+            )
+          )
+        end
+      end
+
+      annex_vi = Annex::VI.new.tap do |annex|
+        parse("data/annex.VI.csv") do |row|
+          common_ingredients = transform_array!(
+            row,
+            key: :common_ingredients,
+            split: ";"
+          )
+          identified_ingredients = transform_array!(
+            row,
+            key: :identified_ingredients,
+            split: ";"
+          )
+
+          annex.add_rule(
+            row.merge(
+              common_ingredients: common_ingredients.compact,
+              identified_ingredients: identified_ingredients.compact,
+              other_restrictions: row[:other]
+            )
+          )
+        end
+      end
+
+      {
+        ii: annex_ii,
+        iii: annex_iii,
+        iv: annex_iv,
+        v: annex_v,
+        vi: annex_vi
+      }
+    end
+
+    def transform_array!(params, key:, split:)
+      params
+        .delete(key)
+        .split(split)
+        .map(&:strip)
+        .reject { |n| n == "-" }
+    end
+
+    def build_sccs_opinions(row)
+      transform_array!(
+        row,
+        key: :sccs_opinions,
+        split: ";"
+      ).map do |opinion|
+        code, *description = opinion.split("-")
+
+        SccsOpinion.new(
+          code:,
+          description: description.join("-")
+        )
+      end
+    end
+
+    def build_cas_numbers(row)
+      transform_array!(
+        row,
+        key: :cas_number,
+        split: "/"
+      ).map do |cas_number|
+        match = cas_number.match(/(?<cas_number>\d{2,7}-\d{2}-\d)/)
+        match[:cas_number] if match
+      end
+    end
+
+    def build_ec_numbers(row)
+      transform_array!(
+        row,
+        key: :ec_number,
+        split: "/"
+      ).map do |ec_number|
+        match = ec_number.match(/(?<ec_number>\d{3}-\d{3}-\d)/)
+        match[:ec_number] if match
+      end
+    end
+  end
+end

data/lib/cosing/database.rb

@@ -0,0 +1,83 @@
+# frozen_string_literal: true
+
+module Cosing
+  class Database
+    attr_reader :annexes, :ingredients
+
+    def initialize(annexes)
+      @annexes = annexes
+      @ingredients = {}
+    end
+
+    def add_ingredient(params)
+      restrictions = transform_array!(params, key: :restriction, split: "\n")
+      annotation_pattern = %r{([IVX]+)/([IVX]*/)?([\dabcd,]+)}
+
+      regulations = restrictions
+        .flat_map do |restriction|
+          matches = restriction.scan(annotation_pattern)
+          next unless matches.any?
+
+          hits = matches.flat_map do |match|
+            numeral, _, reference_number = match
+
+            reference_number.split(",").flat_map do |number|
+              #debugger if number == "41" && numeral == "VI"
+              @annexes[numeral.downcase.to_sym].lookup(number)
+            end
+          end
+
+          hits.compact
+        end
+
+      cas_numbers = transform_array!(
+        params,
+        key: :cas_number,
+        split: "/"
+      ).map do |cas_number|
+        match = cas_number.match(/(?<cas_number>\d{2,7}-\d{2}-\d)/)
+        match[:cas_number] if match
+      end
+
+      functions = transform_array!(params, key: :functions, split: ",")
+
+      einecs_numbers = transform_array!(
+        params,
+        key: :einecs_number,
+        split: "/"
+      ).map do |einecs_number|
+        match = einecs_number.match(/(?<einecs_number>\d{3}-\d{3}-\d)/)
+        match[:einecs_number] if match
+      end
+
+      @ingredients[params[:reference_number]] = Ingredient.new(
+        functions:,
+        restrictions: restrictions.compact,
+        regulations: regulations.compact,
+        cas_numbers: cas_numbers.compact,
+        einecs_numbers: einecs_numbers.compact,
+        **params
+      )
+    end
+
+    def save(filepath, pretty: false)
+      output = if pretty
+                 JSON.pretty_generate(@ingredients.to_h)
+               else
+                 JSON.dump(@ingredients.to_h)
+               end
+
+      File.write(filepath, output)
+    end
+
+    private
+
+    def transform_array!(params, key:, split:)
+      params
+        .delete(key)
+        .split(split)
+        .map(&:strip)
+        .reject { |n| n == "-" }
+    end
+  end
+end

data/lib/cosing/ingredient.rb

@@ -0,0 +1,24 @@
+# frozen_string_literal: true
+
+module Cosing
+  class Ingredient < Dry::Struct
+    attribute :reference_number, Types::String
+    attribute :inci_name, Types::String
+    attribute :inn, Types::String
+    attribute :ph_eur_name, Types::String
+    attribute :cas_numbers, Types::Array.of(Types::CasNumber)
+    attribute :einecs_numbers, Types::Array.of(Types::EinecsNumber)
+    attribute :description, Types::String
+    attribute :restrictions, Types::Array.of(Types::String)
+    attribute :functions, Types::Array.of(Types::String)
+    attribute :regulations, Types::Array
+
+    def inspect
+      to_h
+    end
+
+    def to_json(...)
+      to_h.to_json(...)
+    end
+  end
+end

data/lib/cosing/types.rb

@@ -0,0 +1,10 @@
+# frozen_string_literal: true
+
+module Cosing
+  module Types
+    include Dry.Types()
+
+    CasNumber = Types::String.constrained(format: /\A\d{2,7}-\d{2}-\d\z/)
+    EinecsNumber = Types::String.constrained(format: /\A\d{3}-\d{3}-\d\z/)
+  end
+end

data/lib/cosing/version.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+module Cosing
+  VERSION = "0.1.0"
+end

data/lib/cosing.rb

@@ -0,0 +1,39 @@
+# frozen_string_literal: true
+
+require "csv"
+require "json"
+require "dry/types"
+require "dry/struct"
+require_relative "cosing/version"
+require_relative "cosing/types"
+require_relative "cosing/ingredient"
+require_relative "cosing/annex"
+require_relative "cosing/database"
+
+module Cosing
+  module_function
+
+  class Error < StandardError; end
+  # Your code goes here...
+
+  def load
+    Database.new(Annex.load).tap do |database|
+      ingredient_file = File.read("data/ingredients.csv").delete("\r")
+
+      CSV.parse(
+        ingredient_file,
+        headers: true,
+        liberal_parsing: true,
+        header_converters: :symbol
+      ) do |row|
+        row =
+          row
+            .to_h
+            .transform_values(&:to_s)
+            .transform_values(&:strip)
+
+        database.add_ingredient(row)
+      end
+    end
+  end
+end

metadata

@@ -0,0 +1,102 @@
+--- !ruby/object:Gem::Specification
+name: cosing
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- Nolan J Tait
+autorequire:
+bindir: exe
+cert_chain: []
+date: 2024-01-03 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: dry-struct
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.6'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.6'
+- !ruby/object:Gem::Dependency
+  name: dry-types
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.7'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.7'
+description: COSING database
+email:
+- nolanjtait@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- ".rspec"
+- ".rubocop.yml"
+- CHANGELOG.md
+- LICENSE.txt
+- README.md
+- Rakefile
+- cosing.gemspec
+- data/annex.II.csv
+- data/annex.III.csv
+- data/annex.IV.csv
+- data/annex.V.csv
+- data/annex.VI.csv
+- data/ingredients.csv
+- lib/cosing.rb
+- lib/cosing/annex.rb
+- lib/cosing/annex/base.rb
+- lib/cosing/annex/ii.rb
+- lib/cosing/annex/iii.rb
+- lib/cosing/annex/iv.rb
+- lib/cosing/annex/rule.rb
+- lib/cosing/annex/sccs_opinion.rb
+- lib/cosing/annex/v.rb
+- lib/cosing/annex/vi.rb
+- lib/cosing/database.rb
+- lib/cosing/ingredient.rb
+- lib/cosing/types.rb
+- lib/cosing/version.rb
+homepage: https://github.com/inhouse-work/cosing
+licenses:
+- MIT
+metadata:
+  homepage_uri: https://github.com/inhouse-work/cosing
+  source_code_uri: https://github.com/inhouse-work/cosing
+  changelog_uri: https://github.com/inhouse-work/cosing
+  rubygems_mfa_required: 'true'
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '3'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.4.10
+signing_key:
+specification_version: 4
+summary: COSING database
+test_files: []