greeb 0.2.1 → 0.2.2.pre1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: d36a8ea4366f3bbc2b14b3f02dd7fadf0f25571e
-  data.tar.gz: 42ecf5ddd8c4b8dac3ffa4394ce8c3211f91d918
+  metadata.gz: bff91e4559c7b7d1f83ef83212455d4b4b282351
+  data.tar.gz: 867bb10aeb676a2608b701dd57c5a68b215eeaa9
 SHA512:
-  metadata.gz: 72282ee20c4566ccceac0a4f51aa3ad8ea213458502d2b5ed97c664c06fbc8aa8a28b5fb1f4e8bb84f0de2d7ea292d54d1500e646524b18c9af192069596e63e
-  data.tar.gz: 1188dc13cac34f31d2536b94d460dfe5948f86095aaebd43c1a85da9f521ba53ae1ebdd94df6cf5702cb2f64df4d0cce37fee6f7a4d308feb230bc78729a79ce
+  metadata.gz: f116b3773ac59a02e17c6cece192cb5d8cc21ae976e2e78778752c602ef87b4e6cfe9c383e4b09150b11e9c4dbc08788b27a34834c15023d315b08d08e922d26
+  data.tar.gz: 2532a596959945be6793265fc062f9c89d7676df2a28a535e1662b8bf4d848ac66f35e3bf299282dc53e3ab72ce9efe76382079ada9476856da14361ce19df53

data/bin/greeb

@@ -6,21 +6,8 @@ end
 
 require 'greeb'
 
-text = STDIN.read
-text.chomp!
+text = STDIN.read.tap(&:chomp!)
 
-tokens = Greeb::Tokenizer.tokenize(text)
-
-extract = proc do |entity|
-  from = tokens.index { |e| e.from == entity.from }
-  to = tokens.index { |e| e.to == entity.to }
-  tokens[from..to] = entity
-end
-
-Greeb::Parser.urls(text).each(&extract)
-Greeb::Parser.emails(text).each(&extract)
-Greeb::Parser.abbrevs(text).each(&extract)
-
-tokens.each do |entity|
+Greeb[text].each do |entity|
   puts text[entity.from...entity.to] unless entity.type == :separ
 end

data/greeb.gemspec

@@ -2,23 +2,23 @@
 
 require File.expand_path('../lib/greeb/version', __FILE__)
 
-Gem::Specification.new do |s|
-  s.name        = 'greeb'
-  s.version     = Greeb::VERSION
-  s.platform    = Gem::Platform::RUBY
-  s.authors     = ['Dmitry Ustalov']
-  s.email       = ['dmitry@eveel.ru']
-  s.homepage    = 'https://github.com/dmchk/greeb'
-  s.summary     = 'Greeb is a simple Unicode-aware regexp-based tokenizer.'
-  s.description = 'Greeb is a simple yet awesome and Unicode-aware ' \
-                  'regexp-based tokenizer, written in Ruby.'
+Gem::Specification.new do |spec|
+  spec.name        = 'greeb'
+  spec.version     = Greeb::VERSION
+  spec.platform    = Gem::Platform::RUBY
+  spec.authors     = ['Dmitry Ustalov']
+  spec.email       = ['dmitry@eveel.ru']
+  spec.homepage    = 'https://github.com/dmchk/greeb'
+  spec.summary     = 'Greeb is a simple Unicode-aware regexp-based tokenizer.'
+  spec.description = 'Greeb is a simple yet awesome and Unicode-aware ' \
+                     'regexp-based tokenizer, written in Ruby.'
 
-  s.rubyforge_project = 'greeb'
+  spec.rubyforge_project = 'greeb'
 
-  s.files         = `git ls-files`.split("\n")
-  s.test_files    = `git ls-files -- {test,spec,features}/*`.split("\n")
-  s.executables   = `git ls-files -- bin/*`.split("\n").map{ |f| File.basename(f) }
-  s.require_paths = ['lib']
+  spec.files         = `git ls-files`.split("\n")
+  spec.test_files    = `git ls-files -- {test,spec,features}/*`.split("\n")
+  spec.executables   = `git ls-files -- bin/*`.split("\n").map{ |f| File.basename(f) }
+  spec.require_paths = ['lib']
 
-  s.add_development_dependency 'minitest', '~> 5.0'
+  spec.add_development_dependency 'minitest', '~> 5.0'
 end

data/lib/greeb.rb

@@ -44,3 +44,4 @@ require 'greeb/strscan'
 require 'greeb/tokenizer'
 require 'greeb/segmentator'
 require 'greeb/parser'
+require 'greeb/core'

data/lib/greeb/core.rb

@@ -0,0 +1,45 @@
+# encoding: utf-8
+
+# Greeb::Core is a simple tool that allows to invoke Greeb::Tokenizer and
+# Greeb::Parser facilities together in a convinient and coherent way.
+#
+module Greeb::Core
+  # Greeb::Core uses several helpers from Greeb::Parser to perform
+  # additional analysis using there heuristic methods.
+  #
+  HELPERS = [:urls, :emails, :abbrevs]
+
+  # Recognize e-mail addresses in the input text.
+  #
+  # @param text [String] input text.
+  #
+  # @return [Array<Greeb::Entity>] a set of tokens.
+  #
+  def analyze text
+    Greeb::Tokenizer.tokenize(text).tap do |tokens|
+      HELPERS.each do |helper|
+        Greeb::Parser.public_send(helper, text).each do |parsed|
+          extract_tokens(tokens, parsed)
+        end
+      end
+    end
+  end
+
+  alias_method :'[]', :analyze
+
+  protected
+  # Extact tokens of the specified type from the input tokens set.
+  #
+  # @param tokens [Array<Greeb::Entity>] input tokens set.
+  # @param entity [Greeb::Entity] token to be extracted.
+  #
+  # @return [Greeb::Entity] token to be extracted.
+  #
+  def extract_tokens(tokens, entity)
+    from = tokens.index { |e| e.from == entity.from }
+    to = tokens.index { |e| e.to == entity.to }
+    tokens[from..to] = entity
+  end
+end
+
+Greeb.send(:extend, Greeb::Core)

data/lib/greeb/version.rb

@@ -5,5 +5,5 @@
 module Greeb
   # Version of Greeb.
   #
-  VERSION = '0.2.1'
+  VERSION = '0.2.2.pre1'
 end

data/spec/core_spec.rb

@@ -0,0 +1,42 @@
+# encoding: utf-8
+
+require_relative 'spec_helper'
+
+module Greeb
+  describe Greeb do
+    it 'should do nothing when ran without input' do
+      Greeb[''].must_be_empty
+    end
+
+    it 'should tokenize text when input is given' do
+      Greeb['Hello guys!'].must_equal(
+        [Entity.new(0, 5, :letter),
+         Entity.new(5, 6, :separ),
+         Entity.new(6, 10, :letter),
+         Entity.new(10, 11, :punct)]
+      )
+    end
+
+    it 'should extract URLs' do
+      Greeb['Hello http://nlpub.ru guys!'].must_equal(
+        [Entity.new(0, 5, :letter),
+         Entity.new(5, 6, :separ),
+         Entity.new(6, 21, :url),
+         Entity.new(21, 22, :separ),
+         Entity.new(22, 26, :letter),
+         Entity.new(26, 27, :punct)]
+      )
+    end
+
+    it 'should extract e-mails' do
+      Greeb['Hello example@example.com guys!'].must_equal(
+        [Entity.new(0, 5, :letter),
+         Entity.new(5, 6, :separ),
+         Entity.new(6, 25, :email),
+         Entity.new(25, 26, :separ),
+         Entity.new(26, 30, :letter),
+         Entity.new(30, 31, :punct)]
+      )
+    end
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: greeb
 version: !ruby/object:Gem::Version
-  version: 0.2.1
+  version: 0.2.2.pre1
 platform: ruby
 authors:
 - Dmitry Ustalov
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-07-02 00:00:00.000000000 Z
+date: 2013-10-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: minitest
@@ -44,12 +44,14 @@ files:
 - bin/greeb
 - greeb.gemspec
 - lib/greeb.rb
+- lib/greeb/core.rb
 - lib/greeb/parser.rb
 - lib/greeb/segmentator.rb
 - lib/greeb/strscan.rb
 - lib/greeb/tokenizer.rb
 - lib/greeb/version.rb
 - spec/bin_spec.rb
+- spec/core_spec.rb
 - spec/parser_spec.rb
 - spec/segmentator_spec.rb
 - spec/spec_helper.rb
@@ -69,17 +71,18 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - '>'
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 1.3.1
 requirements: []
 rubyforge_project: greeb
-rubygems_version: 2.0.3
+rubygems_version: 2.1.9
 signing_key: 
 specification_version: 4
 summary: Greeb is a simple Unicode-aware regexp-based tokenizer.
 test_files:
 - spec/bin_spec.rb
+- spec/core_spec.rb
 - spec/parser_spec.rb
 - spec/segmentator_spec.rb
 - spec/spec_helper.rb