wiktionary 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 !binary "U0hBMQ==":
-  metadata.gz: 8ab8f85d72bf98e429758c1fa47c139028fc8a14
-  data.tar.gz: 9c926d3537990a2034ad24203cbdbe09f8dbb961
+  metadata.gz: 091eb3a349dd479feed4637a5beaef0900cea651
+  data.tar.gz: ca129fad7fefe18ee4d823d8ae0d94f5e5258536
 SHA512:
-  metadata.gz: e841cd75cbe30ba85c2f2824a95bbee790363595284e79d0b8072d079f17b01fa20e0784965420225fde74c85e6baeab9282a680fac5cec0e2f217b8e27c69cc
-  data.tar.gz: f94050c1d8f61a2ec7fd4dac52fbbaf72386a243a9aaf1365dd208718eba484653de839e3cb34b2dec51ad3aeebe7cde861ee31fa889786423c527a9cd569666
+  metadata.gz: 7d38ae673dcef87169985be6d6f2a189ae588f7d7a27df4e8a8e6944ec8d74394f6969797bd14100cdd230ac624ef164204067f00c6ae9a9544228ffb03111fe
+  data.tar.gz: a93e59d7997fc418488a6deddb22223bd021a287fa5c387b8588ff0f1a6908800256aba7fd77da236260021aa98a0ee58048edd98b0a00918def7dcbe2d3ccbd

data/.gitignore

@@ -0,0 +1 @@
+*.gem

data/Readme.md

@@ -0,0 +1,64 @@
+# Word conversion rules from Wiktionary
+
+The aim of the project is providing ruls for converting English words and basic
+informations about the words morphology. The project is based on the English
+Wiktionary thus the descriptions are pretty accurate and include most of the
+exceptions (e.g. ox -> oxen).
+
+## Basic usage
+
+Installation:
+
+```
+$ gem install wiktionary
+```
+
+So far only nouns are supported.
+
+```ruby
+require 'wiktionary/noun'
+
+nouns = Wiktionary::Noun.new
+nouns.singularize("dogs") #=> ["dog"]
+nouns.singularize("oxen") #=> ["ox"]
+nouns.singularize("feet") #=> ["foot"]
+
+nouns.singular?("dog")  #=> true
+nouns.singular?("dogs") #=> false
+
+
+nouns.plural?("dog")  #=> false
+nouns.plural?("dogs") #=> true
+```
+
+The `singularize` method returns an array since there might be more than one
+base form of a given plural word.
+
+## Data
+
+The `data` directory - contains lists and mappings of noun forms (plural,
+singular, countable, etc.). They are used to identify plural forms and
+transform plural to singular. Generated from Wiktionary dump using
+https://github.com/djstrong/nouns-with-plurals. 
+
+The following files are available:
+* `noun.csv` - countable nouns
+* `noun_countable_and_uncountable.csv` - e.g. http://en.wiktionary.org/wiki/beers
+* `noun_uncountable.csv` - nouns that cannot be used freely with numbers or
+  the indefinite article, and which therefore takes no plural form, e.g.
+  http://en.wiktionary.org/wiki/lycra
+* `noun_usually_uncountable.csv` - e.g. http://en.wiktionary.org/wiki/information
+* `noun_unknown.csv` - nouns with unknown or uncertain plural
+* `noun_pluralia_tantum.csv` - nouns that do not have singular forms, e.g. http://en.wiktionary.org/wiki/scissors
+* `noun_not_attested.csv` - nouns with plural not attested
+
+The following files contain two columns (singular and
+plural form): 
+* `noun.csv`
+* `noun_countable_and_uncountable.csv`
+* `noun_usually_uncountable.csv`
+
+## Credits
+
+Krzysztof Wróbel (djstrong)
+Aleksander Smywiński-Pohl (apohllo)

data/lib/wiktionary/noun.rb

@@ -1,4 +1,6 @@
+# encoding: utf-8
 require 'csv'
+require 'unicode_utils'
 
 module Wiktionary
   class Noun
@@ -39,7 +41,16 @@ module Wiktionary
 
     # Returns list of possible singular forms of noun.
     def singularize(noun)
-      @plural_to_singulars[noun]
+      first_capital = false
+      if noun =~ /^\p{Lu}/
+        noun = UnicodeUtils.downcase(noun[0]) + noun[1..-1]
+        first_capital = true
+      end
+      result = @plural_to_singulars[noun]
+      if first_capital
+        result.map!{|word| UnicodeUtils.upcase(word[0]) + word[1..-1] }
+      end
+      result
     end
 
     private
@@ -70,13 +81,16 @@ module Wiktionary
     end
 
     def add(singular,plurals)
+      singular = UnicodeUtils.downcase(singular)
+      plurals = plurals.map{|pl| UnicodeUtils.downcase(pl) }
       plurals.each do |plural|
-        @plural_to_singulars[plural] << singular
-        @singular_to_plurals[singular] << plural
+        @plural_to_singulars[plural] << singular unless @plural_to_singulars[plural].include?(singular)
+        @singular_to_plurals[singular] << plural unless @singular_to_plurals[singular].include?(plural)
       end
     end
 
     def add_uncountable(singular)
+      singular = UnicodeUtils.downcase(singular)
       @singular_to_plurals[singular] ||= []
     end
   end

data/spec/noun.rb

@@ -25,5 +25,7 @@ RSpec.describe Wiktionary::Noun do
     expect(@nouns.singularize("politics")).to include "politics"
     expect(@nouns.singularize("oxen")).to include "ox"
     expect(@nouns.singularize("feet")).to include "foot"
+    expect(@nouns.singularize("Dogs")).to include "Dog"
+    expect(@nouns.singularize("Γ rays")).to include "Γ ray"
   end
 end

data/wiktionary.gemspec

@@ -0,0 +1,25 @@
+Gem::Specification.new do |s|
+  s.name = "wiktionary"
+  s.version = "0.1.1"
+  s.date = "#{Time.now.strftime("%Y-%m-%d")}"
+  s.required_ruby_version = '>= 2.0.0'
+  s.authors = ['Krzysztof Wróbel','Aleksander Smywiński-Pohl']
+  s.email   = ["djstrong@gmail.com","apohllo@o2.pl"]
+  s.homepage    = "http://github.com/cycloped-io/wiktionary"
+  s.summary = "Word morphology and conversion based on Wiktionary"
+  s.description = "English words morphological description and basic conversion rules based on the English Wiktionary."
+  s.license = "http://opensource.org/licenses/MIT"
+
+  s.rubyforge_project = "wiktionary"
+  s.rdoc_options = ["--main", "Readme.md"]
+
+  s.files         = `git ls-files`.split("\n")
+  s.test_files    = `git ls-files -- {test,spec,features}/*`.split("\n")
+  s.require_path = "lib"
+
+  s.add_dependency("unicode_utils")
+
+  s.add_development_dependency("rspec", [">= 3.2.0","< 4.0.0"])
+  s.add_development_dependency("rake", [">= 10.4.0","< 11.0.0"])
+end
+

metadata

@@ -1,16 +1,30 @@
 --- !ruby/object:Gem::Specification
 name: wiktionary
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
-- Krzysztofsz Wróbel
+- Krzysztof Wróbel
 - Aleksander Smywiński-Pohl
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-08-21 00:00:00.000000000 Z
+date: 2015-08-22 00:00:00.000000000 Z
 dependencies:
+- !ruby/object:Gem::Dependency
+  name: unicode_utils
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
@@ -60,7 +74,9 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- .gitignore
 - .rspec
+- Readme.md
 - data/noun.csv
 - data/noun_countable_and_uncountable.csv
 - data/noun_non_attested.csv
@@ -72,6 +88,7 @@ files:
 - lib/wiktionary/noun.rb
 - spec/noun.rb
 - spec/spec_helper.rb
+- wiktionary.gemspec
 homepage: http://github.com/cycloped-io/wiktionary
 licenses:
 - http://opensource.org/licenses/MIT