moby 1.0.1beta
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/.gitignore +3 -0
- data/.rvmrc +83 -0
- data/Gemfile +3 -0
- data/Gemfile.lock +18 -0
- data/README.md +83 -0
- data/Rakefile +1 -0
- data/lib/moby/parts_of_speech.rb +133 -0
- data/lib/moby/version.rb +3 -0
- data/lib/moby.rb +10 -0
- data/moby.gemspec +22 -0
- data/share/moby/parts-of-speech/aaREADME.txt +412 -0
- data/share/moby/parts-of-speech/mobypos.UTF-8.txt +233356 -0
- data/share/moby/parts-of-speech/mobypos.txt +233356 -0
- data/spec/moby/parts_of_speech_spec.rb +271 -0
- data/spec/moby_spec.rb +15 -0
- data/spec/spec_helper.rb +6 -0
- metadata +72 -0
data/.gitignore
ADDED
data/.rvmrc
ADDED
@@ -0,0 +1,83 @@
|
|
1
|
+
#!/usr/bin/env bash
|
2
|
+
|
3
|
+
# This is an RVM Project .rvmrc file, used to automatically load the ruby
|
4
|
+
# development environment upon cd'ing into the directory
|
5
|
+
|
6
|
+
# First we specify our desired <ruby>[@<gemset>], the @gemset name is optional,
|
7
|
+
# Only full ruby name is supported here, for short names use:
|
8
|
+
# echo "rvm use 1.9.2" > .rvmrc
|
9
|
+
environment_id="ruby-1.9.2-p290@moby"
|
10
|
+
|
11
|
+
#
|
12
|
+
# Uncomment the following lines if you want to verify rvm version per project
|
13
|
+
#
|
14
|
+
# rvmrc_rvm_version="1.10.2" # 1.10.1 seams as a safe start
|
15
|
+
# eval "$(echo ${rvm_version}.${rvmrc_rvm_version} | awk -F. '{print "[[ "$1*65536+$2*256+$3" -ge "$4*65536+$5*256+$6" ]]"}' )" || {
|
16
|
+
# echo "This .rvmrc file requires at least RVM ${rvmrc_rvm_version}, aborting loading."
|
17
|
+
# return 1
|
18
|
+
# }
|
19
|
+
#
|
20
|
+
|
21
|
+
#
|
22
|
+
# Uncomment following line if you want options to be set only for given project.
|
23
|
+
#
|
24
|
+
# PROJECT_JRUBY_OPTS=( --1.9 )
|
25
|
+
#
|
26
|
+
# The variable PROJECT_JRUBY_OPTS requires the following to be run in shell:
|
27
|
+
#
|
28
|
+
# chmod +x ${rvm_path}/hooks/after_use_jruby_opts
|
29
|
+
#
|
30
|
+
|
31
|
+
#
|
32
|
+
# First we attempt to load the desired environment directly from the environment
|
33
|
+
# file. This is very fast and efficient compared to running through the entire
|
34
|
+
# CLI and selector. If you want feedback on which environment was used then
|
35
|
+
# insert the word 'use' after --create as this triggers verbose mode.
|
36
|
+
#
|
37
|
+
if [[ -d "${rvm_path:-$HOME/.rvm}/environments" \
|
38
|
+
&& -s "${rvm_path:-$HOME/.rvm}/environments/$environment_id" ]]
|
39
|
+
then
|
40
|
+
\. "${rvm_path:-$HOME/.rvm}/environments/$environment_id"
|
41
|
+
|
42
|
+
if [[ -s "${rvm_path:-$HOME/.rvm}/hooks/after_use" ]]
|
43
|
+
then
|
44
|
+
. "${rvm_path:-$HOME/.rvm}/hooks/after_use"
|
45
|
+
fi
|
46
|
+
else
|
47
|
+
# If the environment file has not yet been created, use the RVM CLI to select.
|
48
|
+
if ! rvm --create "$environment_id"
|
49
|
+
then
|
50
|
+
echo "Failed to create RVM environment '${environment_id}'."
|
51
|
+
return 1
|
52
|
+
fi
|
53
|
+
fi
|
54
|
+
|
55
|
+
#
|
56
|
+
# If you use an RVM gemset file to install a list of gems (*.gems), you can have
|
57
|
+
# it be automatically loaded. Uncomment the following and adjust the filename if
|
58
|
+
# necessary.
|
59
|
+
#
|
60
|
+
# filename=".gems"
|
61
|
+
# if [[ -s "$filename" ]]
|
62
|
+
# then
|
63
|
+
# rvm gemset import "$filename" | grep -v already | grep -v listed | grep -v complete | sed '/^$/d'
|
64
|
+
# fi
|
65
|
+
|
66
|
+
# If you use bundler, this might be useful to you:
|
67
|
+
# if [[ -s Gemfile ]] && ! command -v bundle >/dev/null
|
68
|
+
# then
|
69
|
+
# printf "%b" "The rubygem 'bundler' is not installed. Installing it now.\n"
|
70
|
+
# gem install bundler
|
71
|
+
# fi
|
72
|
+
# if [[ -s Gemfile ]] && command -v bundle
|
73
|
+
# then
|
74
|
+
# bundle install
|
75
|
+
# fi
|
76
|
+
|
77
|
+
if [[ $- == *i* ]] # check for interactive shells
|
78
|
+
then
|
79
|
+
echo "Using: $(tput setaf 2)$GEM_HOME$(tput sgr0)" # show the user the ruby and gemset they are using in green
|
80
|
+
else
|
81
|
+
echo "Using: $GEM_HOME" # don't use colors in interactive shells
|
82
|
+
fi
|
83
|
+
|
data/Gemfile
ADDED
data/Gemfile.lock
ADDED
@@ -0,0 +1,18 @@
|
|
1
|
+
GEM
|
2
|
+
remote: http://rubygems.org/
|
3
|
+
specs:
|
4
|
+
diff-lcs (1.1.3)
|
5
|
+
rspec (2.8.0)
|
6
|
+
rspec-core (~> 2.8.0)
|
7
|
+
rspec-expectations (~> 2.8.0)
|
8
|
+
rspec-mocks (~> 2.8.0)
|
9
|
+
rspec-core (2.8.0)
|
10
|
+
rspec-expectations (2.8.0)
|
11
|
+
diff-lcs (~> 1.1.2)
|
12
|
+
rspec-mocks (2.8.0)
|
13
|
+
|
14
|
+
PLATFORMS
|
15
|
+
ruby
|
16
|
+
|
17
|
+
DEPENDENCIES
|
18
|
+
rspec
|
data/README.md
ADDED
@@ -0,0 +1,83 @@
|
|
1
|
+
Moby
|
2
|
+
====
|
3
|
+
|
4
|
+
**WIP**
|
5
|
+
|
6
|
+
Ruby interface for [Moby Project](http://en.wikipedia.org/wiki/Moby_Project)
|
7
|
+
[wordlists](http://www.gutenberg.org/catalog/world/results?title=moby+list)
|
8
|
+
|
9
|
+
### Usage
|
10
|
+
|
11
|
+
```ruby
|
12
|
+
pos = Moby::PartsOfSpeech.new
|
13
|
+
```
|
14
|
+
|
15
|
+
### List
|
16
|
+
|
17
|
+
Get word lists by part of speech
|
18
|
+
|
19
|
+
```ruby
|
20
|
+
pos.nouns
|
21
|
+
pos.plurals
|
22
|
+
pos.noun_phrases
|
23
|
+
pos.adjectives
|
24
|
+
pos.adverbs
|
25
|
+
pos.conjunctions
|
26
|
+
pos.prepositions
|
27
|
+
pos.interjections
|
28
|
+
pos.pronouns
|
29
|
+
pos.definite_articles
|
30
|
+
pos.indefinite_articles
|
31
|
+
pos.nominatives
|
32
|
+
```
|
33
|
+
|
34
|
+
`#verbs` takes an optional `:type` parameter since there's three types
|
35
|
+
|
36
|
+
```ruby
|
37
|
+
pos.verbs
|
38
|
+
pos.verbs(:type => :all) # default
|
39
|
+
pos.verbs(:type => :usu)
|
40
|
+
pos.verbs(:type => :transitive)
|
41
|
+
pos.verbs(:type => :intransitive)
|
42
|
+
```
|
43
|
+
|
44
|
+
### Find
|
45
|
+
|
46
|
+
Find parts of speech info for a word
|
47
|
+
|
48
|
+
```ruby
|
49
|
+
pos.find("word")
|
50
|
+
#=> {
|
51
|
+
#=> :word => "word",
|
52
|
+
#=> :code => "Nt",
|
53
|
+
#=> :pos => [:noun, :vert_transitive]
|
54
|
+
#=> }
|
55
|
+
|
56
|
+
pos.find("walk")
|
57
|
+
#=> {
|
58
|
+
#=> :word => "walk",
|
59
|
+
#=> :code => "iNt",
|
60
|
+
#=> :pos => [:noun, :verb_transitive, :verb_intransitive]
|
61
|
+
#=> }
|
62
|
+
```
|
63
|
+
|
64
|
+
### Query
|
65
|
+
|
66
|
+
Check a word by part of speech
|
67
|
+
|
68
|
+
```ruby
|
69
|
+
pos.noun?("animal") #=> true
|
70
|
+
pos.verb?("run") #=> true
|
71
|
+
pos.intransitive_verb?("walk") #=> true
|
72
|
+
pos.adjective?("sky") #=> false
|
73
|
+
```
|
74
|
+
|
75
|
+
`#verb?` takes an optional `:type` parameter (like `#verbs`)
|
76
|
+
|
77
|
+
```ruby
|
78
|
+
pos.verb?("abandon", :type => :transitive) #=> true
|
79
|
+
pos.verb?("overurbanized", :type => :intransitive) #=> false
|
80
|
+
pos.verb?("overurbanized") #=> true
|
81
|
+
```
|
82
|
+
|
83
|
+
|
data/Rakefile
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
require 'bundler/gem_tasks'
|
@@ -0,0 +1,133 @@
|
|
1
|
+
module Moby
|
2
|
+
class PartsOfSpeech
|
3
|
+
def find(word)
|
4
|
+
{ :word => word,
|
5
|
+
:found => pos.has_key?(word),
|
6
|
+
:code => pos_code(word),
|
7
|
+
:pos => pos_breakdown(word)
|
8
|
+
}
|
9
|
+
end
|
10
|
+
|
11
|
+
# Word list methods
|
12
|
+
|
13
|
+
def nouns; words(:noun); end
|
14
|
+
def plurals; words(:plural) end
|
15
|
+
def noun_phrases; words(:noun_phrase); end
|
16
|
+
def adjectives; words(:adjective); end
|
17
|
+
def adverbs; words(:adverb); end
|
18
|
+
def conjunctions; words(:conjunction); end
|
19
|
+
def prepositions; words(:preposition); end
|
20
|
+
def interjections; words(:interjection); end
|
21
|
+
def pronouns; words(:pronoun); end
|
22
|
+
def definite_articles; words(:definite_article); end
|
23
|
+
def indefinite_articles; words(:indefinite_article); end
|
24
|
+
def nominatives; words(:nominatives); end
|
25
|
+
|
26
|
+
def verbs(options = {:type => :all})
|
27
|
+
case options[:type]
|
28
|
+
when :all
|
29
|
+
words(:verb_usu_participle) +
|
30
|
+
words(:verb_transitive) +
|
31
|
+
words(:verb_intransitive)
|
32
|
+
when :usu
|
33
|
+
words(:verb_usu_participle)
|
34
|
+
when :transitive
|
35
|
+
words(:verb_transitive)
|
36
|
+
when :intransitive
|
37
|
+
words(:verb_intransitive)
|
38
|
+
end
|
39
|
+
end
|
40
|
+
|
41
|
+
# Query methods
|
42
|
+
|
43
|
+
def noun?(word); pos?(word, :noun); end
|
44
|
+
def plural?(word); pos?(word, :plural); end
|
45
|
+
def noun_phrase?(word); pos?(word, :noun_phrase); end
|
46
|
+
def adjective?(word); pos?(word, :adjective); end
|
47
|
+
def adverb?(word); pos?(word, :adverb); end
|
48
|
+
def conjunction?(word); pos?(word, :conjunction); end
|
49
|
+
def preposition?(word); pos?(word, :preposition); end
|
50
|
+
def interjection?(word); pos?(word, :interjection); end
|
51
|
+
def pronoun?(word); pos?(word, :pronoun); end
|
52
|
+
def definite_article?(word); pos?(word, :definite_article); end
|
53
|
+
def indefinite_article?(word); pos?(word, :indefinite_article); end
|
54
|
+
def nominative?(word); pos?(word, :nominative); end
|
55
|
+
|
56
|
+
def verb?(word, options = {:type => :all})
|
57
|
+
case options[:type]
|
58
|
+
when :all
|
59
|
+
pos?(word, :verb_usu_participle) ||
|
60
|
+
pos?(word, :verb_transitive) ||
|
61
|
+
pos?(word, :verb_intransitive)
|
62
|
+
when :usu
|
63
|
+
pos?(word, :verb_usu_participle)
|
64
|
+
when :transitive
|
65
|
+
pos?(word, :verb_transitive)
|
66
|
+
when :intransitive
|
67
|
+
pos?(word, :verb_intransitive)
|
68
|
+
end
|
69
|
+
end
|
70
|
+
|
71
|
+
private
|
72
|
+
def pos
|
73
|
+
path = %w{share moby parts-of-speech mobypos.UTF-8.txt}
|
74
|
+
pos = File.open(File.join(Moby::base_path, *path))
|
75
|
+
|
76
|
+
@pos ||= Hash[
|
77
|
+
pos.readlines.map {|ln|
|
78
|
+
ln.split('\\').map {|p| p.strip }
|
79
|
+
}
|
80
|
+
]
|
81
|
+
|
82
|
+
@pos.default = ""
|
83
|
+
@pos
|
84
|
+
end
|
85
|
+
|
86
|
+
def pos?(word, pos_name)
|
87
|
+
find(word)[:pos].include?(pos_name.to_sym)
|
88
|
+
end
|
89
|
+
|
90
|
+
# Get words by pos name
|
91
|
+
def words(pos_name)
|
92
|
+
pos.select {|w, c| c.include?(pos_code_map.key(pos_name)) }.keys
|
93
|
+
end
|
94
|
+
|
95
|
+
# Get part of speech code for word
|
96
|
+
def pos_code(word)
|
97
|
+
pos[word]
|
98
|
+
end
|
99
|
+
|
100
|
+
def pos_code_map
|
101
|
+
@pos_code_map ||= {
|
102
|
+
"N" => :noun,
|
103
|
+
"p" => :plural,
|
104
|
+
"h" => :noun_phrase,
|
105
|
+
"V" => :verb_usu_participle,
|
106
|
+
"t" => :verb_transitive,
|
107
|
+
"i" => :verb_intransitive,
|
108
|
+
"A" => :adjective,
|
109
|
+
"v" => :adverb,
|
110
|
+
"C" => :conjunction,
|
111
|
+
"P" => :preposition,
|
112
|
+
"!" => :interjection,
|
113
|
+
"r" => :pronoun,
|
114
|
+
"D" => :definite_article,
|
115
|
+
"I" => :indefinite_article,
|
116
|
+
"o" => :nominative
|
117
|
+
}
|
118
|
+
end
|
119
|
+
|
120
|
+
def pos_codes
|
121
|
+
pos_code_map.keys
|
122
|
+
end
|
123
|
+
|
124
|
+
def pos_names
|
125
|
+
pos_code_map.values
|
126
|
+
end
|
127
|
+
|
128
|
+
# Convert POS code to array of descriptive symbols from #pos_code_map
|
129
|
+
def pos_breakdown(word)
|
130
|
+
pos_code(word).chars.to_a.map {|c| pos_code_map[c] }
|
131
|
+
end
|
132
|
+
end
|
133
|
+
end
|
data/lib/moby/version.rb
ADDED
data/lib/moby.rb
ADDED
data/moby.gemspec
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
# -*- encoding: utf-8 -*-
|
2
|
+
$:.push File.expand_path("../lib", __FILE__)
|
3
|
+
require "moby/version"
|
4
|
+
|
5
|
+
Gem::Specification.new do |s|
|
6
|
+
s.name = "moby"
|
7
|
+
s.version = Moby::VERSION
|
8
|
+
s.authors = ["Dan Richert"]
|
9
|
+
s.email = ["dan.richert@gmail.com"]
|
10
|
+
s.homepage = ""
|
11
|
+
s.summary = %q{Ruby interface for Moby wordlists}
|
12
|
+
s.description = %q{Ruby interface for the world's largest free phonetic database}
|
13
|
+
|
14
|
+
s.rubyforge_project = "moby"
|
15
|
+
|
16
|
+
s.files = `git ls-files`.split("\n")
|
17
|
+
s.test_files = `git ls-files -- {test,spec,features}/*`.split("\n")
|
18
|
+
s.executables = `git ls-files -- bin/*`.split("\n").map{ |f| File.basename(f) }
|
19
|
+
s.require_paths = ["lib"]
|
20
|
+
|
21
|
+
s.add_development_dependency "rspec"
|
22
|
+
end
|