ingreedy 0.0.4 → 0.0.5

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: b9b4b6105807082714db4a684b0667635a6c0a82
+  data.tar.gz: 85c9ab384b6c419520723d9719b9f82d6ebca495
+SHA512:
+  metadata.gz: 3321ead5195bc89f49a38e9b9fd762f624e9dfed41486a703d05d4ffade3a5f4f1ea3898b16b99d598a0790ea3f108940851e3b6ab831abc388d1160cf095771
+  data.tar.gz: 545113b12845639fb146fe26d9e263044452bcab8be636f5d83937752b0e74aaa07dbf6050550684e6ce444c5acdb6f3ff74ace23068c3cb7754af46e4ef6053

data/lib/ingreedy.rb

@@ -1,13 +1,20 @@
 path = File.expand_path(File.join(File.dirname(__FILE__), 'ingreedy'))
 
+require File.join(path, 'case_insensitive_parser')
 require File.join(path, 'ingreedy_parser')
+require File.join(path, 'dictionary_collection')
 
 module Ingreedy
   class << self
+    attr_accessor :locale
+
     def parse(query)
-      parser = IngreedyParser.new(query)
+      parser = Parser.new(query)
       parser.parse
-      parser
+    end
+
+    def dictionaries
+      @dictionaries ||= DictionaryCollection.new
     end
   end
 end

data/lib/ingreedy/amount_parser.rb

@@ -0,0 +1,60 @@
+require 'parslet'
+
+module Ingreedy
+
+  class AmountParser < Parslet::Parser
+    include CaseInsensitiveParser
+
+    def initialize(options = {})
+      @key_prefix = options[:key_prefix] ? "#{options[:key_prefix]}_" : ''
+    end
+
+    def capture_key(key)
+      (@key_prefix + key.to_s).to_sym
+    end
+
+    rule(:whitespace) do
+      match("\s")
+    end
+
+    rule(:integer) do
+      match('[0-9]').repeat(1)
+    end
+
+    rule(:float) do
+      integer.maybe >>
+      str('.') >> integer
+    end
+
+    rule(:fraction) do
+      (integer.as(capture_key(:integer_amount)) >> whitespace).maybe >>
+      (integer >> match('/') >> integer).as(capture_key(:fraction_amount))
+    end
+
+    rule(:word_digit) do
+      word_digits.map { |d| stri(d) }.inject(:|) || any
+    end
+
+    rule(:amount_unit_separator) do
+      whitespace | str('-')
+    end
+
+    rule(:amount) do
+      fraction |
+      float.as(capture_key(:float_amount)) |
+      integer.as(capture_key(:integer_amount)) |
+      word_digit.as(capture_key(:word_integer_amount)) >> amount_unit_separator
+    end
+
+    root(:amount)
+
+    private
+
+    def word_digits
+      Ingreedy.dictionaries.current.numbers.keys
+    end
+
+  end
+
+
+end

data/lib/ingreedy/case_insensitive_parser.rb

@@ -0,0 +1,12 @@
+module Ingreedy
+  module CaseInsensitiveParser
+
+    def stri(str)
+      key_chars = str.split(//)
+      key_chars.
+        collect! { |char| match["#{char.upcase}#{char.downcase}"] }.
+        reduce(:>>)
+    end
+
+  end
+end

data/lib/ingreedy/dictionary.rb

@@ -0,0 +1,11 @@
+module Ingreedy
+  class Dictionary
+    attr_reader :units, :numbers, :prepositions
+
+    def initialize(entries = {})
+      @units = entries[:units] || raise('No units found in dictionary')
+      @numbers = entries[:numbers] || {}
+      @prepositions = entries[:prepositions] || []
+    end
+  end
+end

data/lib/ingreedy/dictionary_collection.rb

@@ -0,0 +1,35 @@
+require 'yaml'
+require_relative 'dictionary'
+
+module Ingreedy
+  class DictionaryCollection
+    def initialize
+      @collection = {}
+    end
+
+    def []=(locale, attributes)
+      @collection[locale] = Dictionary.new(attributes)
+    end
+
+    def current
+      @collection[locale] ||= Dictionary.new load_yaml(locale)
+    end
+
+    private
+
+    def locale
+      Ingreedy.locale || i18n_gem_locale || :en
+    end
+
+    def i18n_gem_locale
+      I18n.locale if defined?(I18n)
+    end
+
+    def load_yaml(locale)
+      path = File.expand_path(File.join(File.dirname(__FILE__), 'dictionaries', "#{locale}.yml"))
+      YAML.load_file(path)
+    rescue Errno::ENOENT
+      raise "No dictionary found for :#{locale} locale"
+    end
+  end
+end

data/lib/ingreedy/ingreedy_parser.rb

@@ -1,103 +1,151 @@
-class IngreedyParser
+require 'parslet'
 
-  attr_reader :amount, :unit, :ingredient, :query
+require_relative 'amount_parser'
+require_relative 'rationalizer'
+require_relative 'unit_parser'
+require_relative 'unit_variation_mapper'
 
-  def initialize(query)
-    @query = query
-  end
+module Ingreedy
 
-  def parse
-    ingreedy_regex = %r{
-      (?<amount> .?\d+(\.\d+)? ) {0}
-      (?<fraction> \d\/\d ) {0}
+  class Parser < Parslet::Parser
 
-      (?<container_amount> \d+(\.\d+)?) {0}
-      (?<container_unit> .+) {0}
-      (?<container_size> \(\g<container_amount>\s\g<container_unit>\)) {0}
-      (?<unit_and_ingredient> .+ ) {0}
+    attr_reader :original_query
+    Result = Struct.new(:amount, :unit, :ingredient, :original_query)
 
-      (\g<fraction>\s)?(\g<amount>\s?)?(\g<fraction>\s)?(\g<container_size>\s)?\g<unit_and_ingredient>
-    }x
-    results = ingreedy_regex.match(@query)
+    rule(:amount) do
+      AmountParser.new.as(:amount)
+    end
 
-    @ingredient_string = results[:unit_and_ingredient]
-    @container_amount = results[:container_amount]
-    @container_unit = results[:container_unit]
+    rule(:whitespace) do
+      match("\s")
+    end
 
-    parse_amount results[:amount], results[:fraction]
-    parse_unit_and_ingredient
-  end
+    rule(:container_amount) do
+      AmountParser.new(key_prefix: 'container')
+    end
 
-  private
+    rule(:unit) do
+      UnitParser.new
+    end
 
-  def parse_amount(amount_string, fraction_string)
-    fraction = 0
-    if fraction_string
-      numbers = fraction_string.split("\/")
-      numerator = numbers[0].to_f
-      denominator = numbers[1].to_f
-      fraction = numerator / denominator
+    rule(:unit_and_preposition) do
+      if prepositions.empty?
+        unit.as(:unit) >> (whitespace | any.absent?)
+      else
+        unit.as(:unit) >> (preposition | whitespace | any.absent?)
+      end
     end
-    @amount = amount_string.to_f + fraction
-    @amount *= @container_amount.to_f if @container_amount
-  end
-  def set_unit_variations(unit, variations)
-    variations.each do |abbrev|
-      @unit_map[abbrev] = unit
+
+    rule(:preposition) do
+      whitespace >>
+      prepositions.map { |con| str(con) }.inject(:|) >>
+      whitespace
     end
-  end
-  def create_unit_map
-    @unit_map = {}
-    # english units
-    set_unit_variations :cup, ["c.", "c", "cup", "cups"]
-    set_unit_variations :fluid_ounce, ["fl. oz.", "fl oz", "fluid ounce", "fluid ounces"]
-    set_unit_variations :gallon, ["gal", "gal.", "gallon", "gallons"]
-    set_unit_variations :ounce, ["oz", "oz.", "ounce", "ounces"]
-    set_unit_variations :pint, ["pt", "pt.", "pint", "pints"]
-    set_unit_variations :pound, ["lb", "lb.", "pound", "pounds"]
-    set_unit_variations :quart, ["qt", "qt.", "qts", "qts.", "quart", "quarts"]
-    set_unit_variations :tablespoon, ["tbsp.", "tbsp", "T", "T.", "tablespoon", "tablespoons"]
-    set_unit_variations :teaspoon, ["tsp.", "tsp", "t", "t.", "teaspoon", "teaspoons"]
-    # metric units
-    set_unit_variations :gram, ["g", "g.", "gr", "gr.", "gram", "grams"]
-    set_unit_variations :kilogram, ["kg", "kg.", "kilogram", "kilograms"]
-    set_unit_variations :liter, ["l", "l.", "liter", "liters"]
-    set_unit_variations :milligram, ["mg", "mg.", "milligram", "milligrams"]
-    set_unit_variations :milliliter, ["ml", "ml.", "milliliter", "milliliters"]
-  end
-  def parse_unit
-    create_unit_map if @unit_map.nil?
-
-    @unit_map.each do |abbrev, unit|
-      if @ingredient_string.start_with?(abbrev + " ")
-        # if a unit is found, remove it from the ingredient string
-        @ingredient_string.sub! abbrev, ""
-        @unit = unit
-      end
+
+    rule(:container_unit) do
+      UnitParser.new
+    end
+
+    rule(:amount_unit_separator) do
+      whitespace | str('-')
+    end
+
+    rule(:container_size) do
+      # e.g. (12 ounce) or 12 ounce
+      str('(').maybe >>
+      container_amount.as(:container_amount) >>
+      amount_unit_separator.maybe >>
+      container_unit.as(:unit) >>
+      str(')').maybe >> whitespace
+    end
+
+    rule(:amount_and_unit) do
+      amount >>
+      whitespace.maybe >>
+      unit_and_preposition.maybe >>
+      container_size.maybe
+    end
+
+    rule(:quantity) do
+      amount_and_unit | unit_and_preposition
+    end
+
+    rule(:standard_format) do
+      # e.g. 1/2 (12 oz) can black beans
+      quantity >> any.repeat.as(:ingredient)
+    end
+
+    rule(:reverse_format) do
+      # e.g. flour 200g
+      ((whitespace >> quantity).absent? >> any).repeat.as(:ingredient) >> whitespace >> quantity
+    end
+
+    rule(:ingredient_addition) do
+      standard_format | reverse_format
+    end
+
+    root :ingredient_addition
+
+    def initialize(original_query)
+      @original_query = original_query
     end
 
-    # if no unit yet, try it again downcased
-    if @unit.nil?
-      @ingredient_string.downcase!
-      @unit_map.each do |abbrev, unit|
-        if @ingredient_string.start_with?(abbrev + " ")
-          # if a unit is found, remove it from the ingredient string
-          @ingredient_string.sub! abbrev, ""
-          @unit = unit
-        end
+    def parse
+      result = Result.new
+      result[:original_query] = original_query
+
+      parslet_output = super(original_query)
+
+      result[:amount] = rationalize_total_amount(parslet_output[:amount], parslet_output[:container_amount])
+
+      if parslet_output[:unit]
+        result[:unit] = convert_unit_variation_to_canonical(parslet_output[:unit].to_s)
       end
+
+      result[:ingredient] = parslet_output[:ingredient].to_s.lstrip.rstrip #TODO cheating
+
+      result
+    end
+
+    private
+
+    def prepositions
+      Ingreedy.dictionaries.current.prepositions
+    end
+
+    def convert_unit_variation_to_canonical(unit_variation)
+      UnitVariationMapper.unit_from_variation(unit_variation)
     end
 
-    # if we still don't have a unit, check to see if we have a container unit
-    if @unit.nil? and @container_unit
-      @unit_map.each do |abbrev, unit|
-        @unit = unit if abbrev == @container_unit
+    def rationalize_total_amount(amount, container_amount)
+      if container_amount
+        rationalize_amount(amount) * rationalize_amount(container_amount, 'container_')
+      else
+        rationalize_amount(amount)
       end
     end
-  end
-  def parse_unit_and_ingredient
-    parse_unit
-    # clean up ingredient string
-    @ingredient = @ingredient_string.lstrip.rstrip
+
+    def rationalize_amount(amount, capture_key_prefix = '')
+      return unless amount
+      integer = amount["#{capture_key_prefix}integer_amount".to_sym]
+      integer &&= integer.to_s
+
+      float = amount["#{capture_key_prefix}float_amount".to_sym]
+      float &&= float.to_s
+
+      fraction = amount["#{capture_key_prefix}fraction_amount".to_sym]
+      fraction &&= fraction.to_s
+
+      word = amount["#{capture_key_prefix}word_integer_amount".to_sym]
+      word &&= word.to_s
+
+      Rationalizer.rationalize(
+        integer:  integer,
+        float:    float,
+        fraction: fraction,
+        word:     word
+      )
+    end
+
   end
 end

data/lib/ingreedy/rationalizer.rb

@@ -0,0 +1,41 @@
+module Ingreedy
+
+  class Rationalizer
+
+    def self.rationalize(options)
+      new(options).rationalize
+    end
+
+    def initialize(options)
+      @integer  = options.fetch(:integer, nil)
+      @float    = options.fetch(:float, nil)
+      @fraction = options.fetch(:fraction, nil)
+      @word     = options.fetch(:word, nil)
+    end
+
+    def rationalize
+      if @word
+        result = rationalize_word
+      elsif @fraction
+        result = @fraction.to_r
+        if @integer
+          result += @integer.to_i
+        end
+      elsif @integer
+        result = @integer.to_r
+      elsif @float
+        result = @float.to_r
+      end
+
+      result
+    end
+
+    private
+
+    def rationalize_word
+      Ingreedy.dictionaries.current.numbers[@word.downcase]
+    end
+
+  end
+
+end

data/lib/ingreedy/unit_parser.rb

@@ -0,0 +1,20 @@
+module Ingreedy
+
+  class UnitParser < Parslet::Parser
+    include CaseInsensitiveParser
+
+    rule(:unit) do
+      unit_variations.map { |var| str(var) | stri(var) }.reduce(:|)
+    end
+
+    root :unit
+
+    private
+
+    def unit_variations
+      UnitVariationMapper.all_variations
+    end
+
+  end
+
+end

data/lib/ingreedy/unit_variation_mapper.rb

@@ -0,0 +1,31 @@
+module Ingreedy
+  class UnitVariationMapper
+
+    def self.all_variations
+      # Return these in order of size, descending
+      # That way, the longer versions will try to be parsed first, then the shorter versions
+      # e.g. so '1 cup flour' will be parsed as 'cup' instead of 'c'
+      variations_map.values.flatten.sort { |a, b| b.length <=> a.length }
+    end
+
+    def self.unit_from_variation(variation)
+      return if variations_map.empty?
+      hash_entry_as_array = variations_map.detect { |unit, variations| variations.include?(variation) }
+
+      if hash_entry_as_array
+        hash_entry_as_array.first
+      else
+        # try again with the variation downcased
+        # this is a hacky way to deal with the abbreviations for teaspoon and tablespoon
+        hash_entry_as_array = variations_map.detect { |unit, variations| variations.include?(variation.downcase) }
+        hash_entry_as_array.first
+      end
+    end
+
+    private
+
+    def self.variations_map
+      Ingreedy.dictionaries.current.units
+    end
+  end
+end

data/lib/ingreedy/version.rb

@@ -1,4 +1,4 @@
 module Ingreedy
-  VERSION = '0.0.4'
+  VERSION = '0.0.5'
 end
 

metadata

@@ -1,59 +1,169 @@
---- !ruby/object:Gem::Specification 
+--- !ruby/object:Gem::Specification
 name: ingreedy
-version: !ruby/object:Gem::Version 
-  prerelease: 
-  version: 0.0.4
+version: !ruby/object:Gem::Version
+  version: 0.0.5
 platform: ruby
-authors: 
+authors:
 - Ian C. Anderson
 autorequire: 
 bindir: bin
 cert_chain: []
-
-date: 2011-09-05 00:00:00 -04:00
-default_executable: 
-dependencies: []
-
-description: Natural language recipe ingredient parser that supports numeric amount, units, and ingredient
-email: 
+date: 2015-10-17 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: parslet
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 1.7.0
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.7.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 1.7.0
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.7.0
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.9'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0.9'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.9'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0.9'
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 3.3.0
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 3.3.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 3.3.0
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 3.3.0
+- !ruby/object:Gem::Dependency
+  name: rspec-its
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 1.2.0
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.2.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 1.2.0
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.2.0
+- !ruby/object:Gem::Dependency
+  name: coveralls
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.7.0
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.7.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.7.0
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.7.0
+- !ruby/object:Gem::Dependency
+  name: pry
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+description: Natural language recipe ingredient parser that supports numeric amount,
+  units, and ingredient
+email:
 - anderson.ian.c@gmail.com
 executables: []
-
 extensions: []
-
 extra_rdoc_files: []
-
-files: 
+files:
+- lib/ingreedy.rb
+- lib/ingreedy/amount_parser.rb
+- lib/ingreedy/case_insensitive_parser.rb
+- lib/ingreedy/dictionary.rb
+- lib/ingreedy/dictionary_collection.rb
 - lib/ingreedy/ingreedy_parser.rb
+- lib/ingreedy/rationalizer.rb
+- lib/ingreedy/unit_parser.rb
+- lib/ingreedy/unit_variation_mapper.rb
 - lib/ingreedy/version.rb
-- lib/ingreedy.rb
-has_rdoc: true
 homepage: http://github.com/iancanderson/ingreedy
 licenses: []
-
+metadata: {}
 post_install_message: 
 rdoc_options: []
-
-require_paths: 
+require_paths:
 - lib
-required_ruby_version: !ruby/object:Gem::Requirement 
-  none: false
-  requirements: 
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
   - - ">="
-    - !ruby/object:Gem::Version 
-      version: "0"
-required_rubygems_version: !ruby/object:Gem::Requirement 
-  none: false
-  requirements: 
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
   - - ">="
-    - !ruby/object:Gem::Version 
-      version: "0"
+    - !ruby/object:Gem::Version
+      version: '0'
 requirements: []
-
 rubyforge_project: 
-rubygems_version: 1.6.2
+rubygems_version: 2.4.8
 signing_key: 
-specification_version: 3
+specification_version: 4
 summary: Recipe parser
 test_files: []
-
+has_rdoc: